Part IV - Interpretation of Customary International Law

16 Interpreting Customary International Law You’ll Never Walk Alone

Panos Merkouris ^*

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.

17 Practical Reasoning and Interpretation of Customary International Law

Kostiantyn Gorobets

1 Introduction

Interpretability of customary (international) law belongs to the class of jurisprudential problems that entangle and intertwine almost all thorny theoretical and practical issues. It is especially visible against the background of debates around whether customary international law (CIL) can be interpreted, and if so, how this differs from its identification; are there or should there be some rules for CIL interpretation, and what would be the difference between such rules and those guiding interpretation of treaties?

This chapter aims at addressing some of these issues. It seeks to suggest a meaningful way of seeing the process of CIL interpretation through the perspective of practical reasoning. By doing so, it purports to disentangle one of the theoretical knots of CIL interpretation: what is the difference between the identification and interpretation of rules of CIL, considering that both processes concentrate mostly on state practices?¹ For the purposes of this chapter, by ‘state practices’ I mean a slightly different concept than the one being typically used in international legal scholarship. I defend the view that any practice is normative by definition, otherwise it is not a practice at all. This goes against the commonly accepted view that ‘mere’ state practices are but collections of actions and fail to constitute a (legal) norm. I use the concept in the plural because the growth of normativity within practical engagements of states is typically dispersed in terms of the subject-matter. Different practical engagements converge into different normative practices, rather than constitute one continuous state practice.

Section 2 addresses the issue of duality of CIL within the doctrine of the container/content distinction, which is of fundamental importance to the theory of sources of international law. Section 3 suggests a view on (state) practices as being inherently normative, which implies the differentiation between tests for normativity and legality when patterns of behaviour are concerned. Section 4 provides a more detailed analysis of customary normativity. The concluding Section 5 highlights the difference in interpretation of state practices depending on their container/content perception and will therefore attempt a differentiation between the interpretation for the purpose of identification and interpretation for the purpose of clarification/application of a rule of CIL.

2 What Is This Thing We Interpret When We Say That We Interpret CIL?

It is at the core of most contemporary doctrines of legal interpretation that interpretation of something is interpretation of something. In order to interpret a thing, this thing must already be there, and so its existence, meaning, and function are in principle independent from an act of interpretation. This primary intuition allows to differentiate between interpretation and creation or invention.² But it also assumes that locating a thing and interpreting it are two distinct enterprises; identifying a rule of CIL and clarifying its meaning are supposedly not the same.³ In this regard, legal interpretation is tightly linked to the doctrine of sources of law; interpretation of law presupposes that one knows where to find it and how to identify it amongst other forms of social normativity.

The doctrine of sources is a groundwork of legal positivism. That a legal order rests on certain sources entails that a specific class of utterances or actions qualify as generating or communicating the law, as long as they match the criteria of validity that emerge from within this legal order. Thus in domestic law we often say that, for instance, statutes or precedents are sources of law in a sense that certain activities of certain bodies (parliament, courts, etc.) within a certain procedure create legal obligations for all or some groups of persons. In international law, it is generally agreed that treaties, customary law, and general principles of law perform that very same function; they create, impose, or generate legal obligations for states.

The qualification of some social facts as matching criteria of validity does not depend on the content of a purported rule or source. As famously framed by HLA Hart, having criteria of validity for sources of law (‘rule of recognition’) entails that ‘members [of social systems] not merely come to accept separate rules piecemeal, but are committed to the acceptance in advance of general classes of rules, marked out by general criteria of validity’.⁴ This commitment to accept in advance certain classes of rules presupposes that sources of law are merely containers, and their content does not typically play a role in qualifying a source of law as such.⁵ Hence the fundamental postulate of legal positivism is that identifying something as law is separated from assessing its merits.⁶

The container/content duality is of paramount importance for legal interpretation.⁷ One may only engage in legal interpretation if one knows that the normative content one wants to clarify, elucidate, or in any other way meaningfully operationalise, is contained in a valid source of law. In the case of statutory interpretation, a statute is a container of legal rules one wants to interpret. In the case of treaty law,⁸ it is a treaty that is the container, and its provisions form its content. But how about CIL? What is this thing that contains customary rules? This question has no obvious answer, though it is maintained, by the International Law Commission (ILC) for example, that in case of CIL the content/container differentiation still applies.⁹ What, then, is the container one is looking for in order to enquire into the content of a CIL rule?

Apparently, interpretation of CIL is not an interpretation of some texts since it is widely agreed that CIL is an unwritten source of international law. In other words, CIL is not contained in any texts. Certainly, it may have some textual loci in treaties, judgments, statements by state organs, to mention some. Although true, this does not infringe on the fact that linguistic formulas, or certain articulations of customary rules, are not customary rules themselves. They may serve as points of reference, as useful short-hand devices used to communicate and more efficiently engage in the practice that sustains a customary rule, but it would be a mistake to say that a statement of a customary rule by an authority (institutional or academic) is the customary rule itself. In other words, linguistic formulations are but evidences of existence of customary rules, not rules as such. This is true for any type of customary rules, not only legal ones. The same way as judgments merely reflect, articulate, frame customary legal rules that are already somewhere there and exist independently of the fact that a court engages them, manuals of English grammar are also but snapshots of the customary rules of language. Neither of these two can be appropriately used as a criterion for maintaining the practices, and it is actually the other way around: we often discard certain articulations of customary rules as outdated or inaccurate on the basis that this is not how we do it (anymore). Therefore, it is a practice itself which is the ultimate criterion of a customary rule, not its certain pronouncement.

Also, it is difficult to see how rules of CIL can be contained in intentions or positions of states (regardless of whether we treat these as instances of state practice or of opinio juris). That is, interpretation of CIL is not an interpretation of intention or will of a purported author. Unlike treaties, or statutes in domestic law, customary rules cannot be said to have determinate authors. It is a distinct feature of customs that they are matter of what we do, not of what one particular member of a community might intend to do on their own.¹⁰ As put by Gerald Postema, ‘custom is never reducible to what each participant does or to what each says, or thinks, or believes about what each does’.¹¹ Thus, even though it may be the case for some customs that they got intentionally sparked by one action of one particular actor,¹² that actor would not, nevertheless, qualify as its ‘author’. If their action ever rises to a customary rule, this means that it is our rule, not theirs. This, once again, is a feature of customary rules generally, not only legal ones, since what separates them from rules being established externally is that customary rules are rules of a community, not rules for it. They do not get created by someone for the community, rather, they form within the community and define it as such.¹³ Identification of an author of a rule only makes sense when a rule was intentionally designed to bind only particular actors (like in the case of agreements, be it a contract in domestic law or a treaty in international law), or when a rule gets imposed by a lawmaker, since in this situation it is necessary to be able to differentiate between a ‘genuine’ and a ‘fake’ lawmaker. Neither of the two situations are proper descriptions of the context of customary law creation or appearance. Thus, even though it is at times common, in international law specifically, to design a customary rule consciously, this does not suggest that the interpretation of such a rule, when it comes to its application, would be an interpretation of some intentions, or that these intentions would be the container of rules of CIL.

It appears that interpretation of CIL is first and foremost interpretation of state practice. The same way as we interpret other customary rules, say, rules of language, or rules of etiquette, when we interpret CIL, we enquire into what, how, in which circumstances, and so on, participants of a certain practice do and do not do. In the case of CIL, a state practice is the ultimate point of reference one has when clarifying a particular legal rule. I will further define what I mean by state practices in the next section. For now, it suffices to stress that unlike in the case of statutes or precedents in domestic law, or treaties in international law, state practices are not only the containers but also the content of rules one wants to interpret. From the perspective of the doctrine of sources of law, customary rules often appear uneasy to deal with, for they are not only a source of law, they are law themselves.¹⁴ That state practices are both content and containers, however, engenders consequences for what the interpretation of customary rules actually entails.

The content/container dualism of state practices makes them similar to light, as it were, that is, they manifest differently depending on how they are looked at. Light behaves as a wave in one set of conditions of observation, and as particles in another, and as such is, in fact, both.¹⁵ This can also be said about state practices, for when they are interpreted for the purposes of identification of rules of CIL they appear as containers, as something legal obligations are scooped from (see Section 5.1); but when they are interpreted for the purpose of clarifying the meaning of rules of CIL, practices appear as their content, as what the rules are content-wise (see Section 5.2). This dualism of state practices creates a confusion as to how these two instances or cases of interpretation differ. If identification and interpretation are, according to the doctrine of sources, different enterprises, how does one tell the difference between the two if both concentrate on state practice?

Before answering this question, it is necessary to take a closer look at state practices as such, since clarifying their nature is of paramount importance for the further enquiry.

3 State Practices and Normative Deeds

Though it is typically asserted that the concept of opinio juris is far more contested than the concept of state practice,¹⁶ the latter also carries many controversies with it. This is partly so due to its container/content duality, but also due to some conceptual assumptions regarding state practices that are deeply rooted in the doctrines of formation and identification of CIL and are constantly replicated in international legal scholarship.

It is a widespread belief, reflected, among other, in the ILC reports and conclusions, and emerging from the famous North Sea Continental Shelf judgment, that a general practice that is accepted as law is to be distinguished from mere usage or habit.¹⁷ To put it in the words of the ILC, ‘practice without acceptance as law …, even if widespread and consistent, can be no more than a non-binding usage’.¹⁸ A characteristic feature of approaching state practice within the doctrine of identification of CIL, defended also by the ILC, is an all-or-nothingness. It appears that there are only two options: either a state practice is accompanied by opinio juris and then may, if quantitative and qualitative requirements are met, constitute a rule of CIL, or, if it is not, then there exists no obligation for states to act in a certain manner whatsoever. This view on CIL, which Monica Hakimi labels ‘the rulebook conception’, assumes that without opinio juris state practices are mere usages or habits that have no binding force, and that there exist certain clear and formal criteria (i.e., secondary rules) which allow to establish normativity and legality of these practices.¹⁹ This is also articulated by the International Court of Justice (ICJ) that ‘many international acts, eg, in the field of ceremonial and protocol, which are performed almost invariably, but which are motivated only by considerations of courtesy, convenience or tradition, and not by any sense of legal duty’.²⁰ It is, therefore, out of paramount importance that ‘one must look at what States actually do and seek to determine whether they recognize an obligation or a right to act in that way’.²¹ The position of the ICJ and ILC on this matter clearly opposes legal customary rules and their absence, which is reasonable from the point of view of legal logic. What is disturbing, however, is how state practices are thought of when there is no opinio juris. The wording adopted by both institutions not only suggests absence of any obligations within such practices, but also non-normativity of such practices; a view widely supported in the academic literature.²² Opinio juris appears as a magic wand that not only turns the ‘raw material’ of state practices into a norm, but simultaneously into a legal norm.

What seems to be the underlying principle behind such a treatment of state practices rests on two interrelated ideas. First, it is clear that the identification of CIL serves the purpose of establishing the existence of a legal obligation binding upon states. When interpreting state practices for this purpose, one therefore asks questions of legality, that is, whether there exists a norm that provides for legal obligations states must fulfil. What goes alongside it, however, often remains fully or partly unnoticed; namely, that legality is an attribute of a norm,²³ and therefore inquiring into whether there is a legal norm is asking two questions, not one: (1) is there a norm (the question of normativity); (2) if yes, is this norm a legal one (the question of legality)? Importantly, these questions should be answered in this particular order. The question of normativity, though, bears entirely different considerations and should be approached with a distinct methodology and conceptual framework, than the question of legality.²⁴

The language adopted by the ILC and the ICJ, however, makes it seem as if deciding on the legality of certain practices is fundamentally the same as deciding on their normativity; when a practice does not meet the threshold of legality, it is a habit or a usage that creates no obligation or a right, which is basically tantamount to the absence of a norm altogether. This brings the second assumption into play, namely, that state practices are often taken as certain collections of individual actions of states, collections that may or may not feature some pattern (actions that are ‘performed almost invariably’ – as if their performance is a matter of (in)variability, rather than following certain normative consideration). It is thus claimed that ‘the requirement that the practice be consistent means that where the relevant acts are divergent to the extent that no pattern of behaviour can be discerned, no general practice (and thus no corresponding rule of customary international law) can be said to exist’.²⁵

The focus on (in)variability and patterns of behaviour that is so explicit in the reasoning of the ILC and the ICJ seems to neglect the idea that the existence of an observable pattern of conduct is not a relevant marker of there being a practice. Invariability of some actions, even when absolutely consistent, may or may not be evidence of a practice, because it is not the invariability or consistency of actions that matters, but the meaning these actions have for those engaged in them. It is a well-known example by HLA Hart that for an external observer all more or less consistent regularities of behaviour look the same in terms of people doing certain things in certain circumstances. However, that some people go to a cinema once a week does not mean that there is a normative consideration to that effect, that is, that it is somehow socially expected or required from them to go to a cinema once a week.²⁶ On the other hand, that all people lie from time to time (some people more often than others) does not deny the existence of a normative consideration that one must not lie. Thus, that some people go to a cinema once a week is a regularity of behaviour, but not a practice. The only way to differentiate between people engaging in a practice and people simply acting uniformly is to adopt what HLA Hart calls ‘the internal point of view’; practices, unlike mere regularities or patterns of behaviour, feature a critical reflective attitude towards actions, which entails their inherently normative nature.²⁷ So, let us take a closer look at the concept of practice.

Practices, unlike mere regularities of behaviour (like that some people happen to go to the cinema once a week), are of normative nature. In ordinary life, it can be said that at the moment a person steps into a practice, they are expected to accept certain deeds that infiltrate and govern this practice, give it shape, and make it meaningful for the participants. A simple test to be used to determine whether a regularity of behaviour is a practice is whether one may fail in performing or not performing certain actions. This is typically ascertained either through existing mutual expectations that deeds of practice are and will be followed, or through criticism explicated when these deeds are ignored, this criticism being an aspect of the practice concerned.²⁸ For people who happened to go to a cinema once a week, it is not a failure not to go there this week, but go twice the next one instead; no one’s expectations are failed to be met, and no criticism would follow. At the same time, lying to people does usually constitute a failure to meet certain expectations, even when no criticism follows (not all lies get discovered, after all). This latter point also relates to other actions rendered prohibited in the context of existing practices, for example, tortures. Such actions are sometimes colloquially referred to as ‘practices’, but even there we can only meaningfully speak of them as ‘practices’ if there exist normative considerations that somehow make tortures meaningful for those engaged in them (e.g., various utilitarian ‘ticking-bomb scenarios’). Existence of conflicting normative expectations within the domain of the same practice is not at all uncommon, given how much these expectations may depend on underlying reasons (see the next section). This is why some practices may feature uncertainty as to what constitutes a failure in following it.²⁹

The difference between the two examples given above is that there is nothing to be failed in the context of going or not going to the cinema; there are no deeds flowing through the conduct of going to a cinema with a certain regularity, and therefore there is no practice, regardless of the fact that for an external observer this could be the most consistent pattern of behaviour by these people they can observe. In the latter example with lying or torturing, though, there is a certain standard embedded into behaviour, a standard that constitutes a deed that generates certain expectations that other participants of a practice would follow this deed. What differentiates practices from regularities of behaviour, therefore, is the existence of deeds as certain standards that get learned and adopted by the participants of a practice and generate expectations regarding other participants.³⁰ Practices, in such a way, are inherently normative, because the mere existence of deeds as standards constitutes an independent reason for acting in one way and not in another. As emphasised by Gerald Postema,

[Customs] are not (merely) patterns of behavior; rather they set standards for behavior, standards of correct and incorrect behavior, and thus purport to guide that behavior and provide bases for its assessment. Thus, mere regularities of behavior taken alone – the usus or ‘state practice’ of international law discourse – not only fail to constitute customs of international law, they fail to constitute customs of any sort, including those of ‘comity’, because they fail to constitute norms.³¹

From this perspective, customary rules do not and cannot exist separately or detached from practices that sustain them. Besides, that there is a practice, and not just a regularity of behaviour, means that there is a norm that shapes this practice. In other words, to say that there exists a state practice on a certain matter already entails saying that there is a norm on this matter, and n-I. For this reason, it is not entirely accurate to ascertain that when a certain practice fails to qualify as a rule of CIL, there is no moral or social obligation in general binding upon states that flows from the deeds and mutual expectation of participants of such a practice.

This view on state practice was particularly endorsed by the International Law Association (ILA) in its ‘Statements of Principles Applicable to Formation of General Customary International Law’, where it claims that:

‘a rule of customary international law is one which is created and sustained by the constant and uniform practice of States and other subjects of international law in or impinging upon their international legal relations, in circumstances which give rise to a legitimate expectation of similar conduct in the future’.³²

To recapitulate, there are two fundamental considerations flowing from the view expressed above. First, practices (any practices, not just state practices) are inherently normative, otherwise they are not practices at all. The normativity of practices is determined by the character of deeds framing them and by the function these practices perform, as well as by the meaning they have for participants. According to Gerald Postema, the normativity of practices is ascertainable first and foremost from the perspective of those participating in them:

Those who participate in a custom’s practice undertake commitments (a) to judge certain performances as appropriate or correct and others as mistaken; (b) to act when the occasion arises in accord with these judgments; (c) to challenge conduct that falls short of these judgments; and (d) to recognize appeals to the judgments as vindications of their actions or valid criticisms of them.³³

Second, the content and meaning of customary rules can be (and usually is) determined without necessarily assessing the character and nature of the normative claims they exhibit (moral, legal, etc.). Hence, practices always create obligations and endow those participating in them with rights. This does not mean that these obligations and rights are of legal nature, but it is important to bear in mind that absence of opinio juris does not signify absence of any obligation.

With an image of state practice as inherently normative, we may now take a further step and try to clarify how such practice can be reconstructed for the purposes of interpretation. What does the normativity of practice look like and what are the interpretative beacons one may use in order to clarify its meaning?

4 Practice as Network of Reasons

In the previous section, I endorsed the conception that practices are inherently normative, and that getting involved in a practice means accepting and following certain normative standards that are embodied in it and are inseparable from deeds penetrating and shaping it. This view entails, among other things, that practices are sustained by mutual expectations of participants and by more or less implicit normative standards that one in principle is able to fail to meet. Importantly, such character of practices makes them normative, and this normativity may, under certain circumstances, qualify as legal. This characteristic of practices is by and large generic and applies to state practices as well.

Normativity, according to a dominant view, reflects a special ability of law and other social practices to provide those participating in them with reasons for action.³⁴ In other words, practices, such as state practices, are normative in a sense that for those who participate in them the mere fact that they do so is a reason for acting and reacting to the actions of other participants in a certain way. This reason-giving function of practices, in their normative manifestation (i.e., from the internal point of view), entails that they require meaningful participation, and this meaningfulness comprises of participants’ ability to recognise and react to actions of others in a way that is intelligible for the rest of the participants. This is precisely why, even when states do not explicate their position regarding actions of other states, this may still contribute to formation of a new, or sustaining an existing, practice. Even an absence of reaction may, under certain circumstances, get deciphered by other participants of a practice meaningfully either as endorsement or at least as acquiescence.

For such meaningful participation, states must consider practice not only as a reason, but as a network of reasons. It is almost never the case that a practice can in one way or another be boiled down to one standalone reason that states ought to comply with, for each practice gets its function, meaning, and normative significance in a wider context of related activities.³⁵ In fact, especially when we look at a broader scope of social practices, even the simplest ones (such as the practice of eating with a fork and a knife) are only meaningful when taken in the context of a much wider set of considerations which justify the existence of these practices and shape their content. Because of this, the precise normative boundaries of practices may be difficult to define. But what is more fundamental for the purposes of interpretation and for the purposes of identification of a state practice is that reasons comprising a practice vary in nature, function, and strength.

One of the most popular and influential explanations of normativity, developed by Joseph Raz, suggests that even though norms are reasons for actions, not all reasons are norms.³⁶ A reason for action, according to his latest definition, is ‘a consideration that renders its [i.e., an action’s] choice intelligible, and counts in its favor’.³⁷ Reasons as such do not give rise to obligations, but it is nevertheless a basic moral principle that one ought to act according to an optimal balance of reasons one has, all things considered. This equally applies to states, since it is almost never disputed that they are morally accountable agents (were they not, it would have been impossible to defend even a proposition that international law has any function or basis for existence whatsoever). In international relations, states claim reasons for their actions all the time, and some of them are norms. Michael Akehurst, in his influential article on custom as a source of international law, refers to an example of states using white paper for diplomatic correspondence to advance his argument that habits do not create rules of law.³⁸ And indeed, that states almost unanimously use white paper only shows that they do so for a widely shared reason, a reason which, nevertheless, is not a norm. If not all reasons are norms, how is it possible to mark a class of reasons that are norms?

Joseph Raz’s solution to the problem of norms being linguistically inseparable from the rest of the reasons³⁹ suggests that there must be some other criteria according to which we could differentiate between ‘mere’ reasons and norms. Norms are second-order pre-emptive reasons,⁴⁰ and because of this they play a drastically different role in practical reasoning as compared to ordinary first-order reasons.⁴¹ Norms, just like other second-order reasons, are reasons to act or to refrain from acting on some first-order reasons. For example, states may share a wide set of reasons for not using armed force in international relations, and the norm of international law that prohibits the use of force is a second-order reason for acting on all those reasons. But also, and probably most importantly, the existence of a norm prohibiting the use of force is a reason for not acting on certain other first-order reasons. The mere fact of such a prohibition implies that states may not act for reasons that count in favour of using force against other states. In such a way, norms are second-order reasons in the sense that they reinforce some first-order reasons and exclude some other first-order reasons. What this means is that not only are norms reasons for actions they prescribe, but they are also reasons for disregarding reasons for non-compliance.⁴² For example, diplomatic immunity is a norm precisely because it is both a reason for states to refrain from subjecting diplomats to their jurisdiction, and a reason for disregarding any other reasons for acting otherwise, no matter how weighty these may be, such as in the cases when diplomats cause lethal accidents or interfere in the internal affairs of the receiving state.⁴³ This pre-emptive character is what differentiates norms from other reasons for action.

Practices are networks of both second-order reasons, that is, norms, and first-order reasons. This allows for complex and often multidimensional justificatory strategies for one or other course of behaviour.⁴⁴ Apart from this, however, this reflects a feature of norms not only being embedded into practices, but also being virtually inseparable from them. Norms, as intrinsically interwoven into practices, do not ‘hang in the air’ or exist in some metaphysical space, and their justification, therefore, is shaped by, and depends on, a wider network of reasons employed within a certain practice. Norms may be justified in a number of ways; as time- and labour-saving devices, as error-eliminating devices, that is, those subjected to such norms use them as shortcuts in practical reasoning so that if a norm gets accepted it is not necessary anymore to figure out each time an optimal balance of reasons to act upon. Some other norms are justified by recourse to an authority, that is, acceptance of a norm comes as a result of acceptance of authorities issuing them. These (and many more) methods of justification of norms may overlap and supplement each other; in fact, most of the norms by which people are bound have more than just one justification.⁴⁵

In such a way, practices, such as state practice, explicate their normativity as tightly intertwined networks of first- and second-order reasons. Seen as such, their interpretation therefore relates to discovering the interconnection between these two classes of reasons, assessing their balance, and unveiling them in justifications employed by states or implied in their actions.

5 Asking the Right Questions: Re-approaching the Content/Container Duality

Thus far this chapter explored the features and intrinsic qualities of (state) practices as the thing being interpreted within the process of legal interpretation. Now it is time to take this a step further and take a look at this interpretation anew. If state practices are networks of first- and second-order reasons within which states form, manifest, and explicate expectations regarding the actions of other states, how does this affect the nature of state practices for the purposes of legal interpretation? How are these networks of reasons interpreted when looked at as containers, and when looked at as contents?

The theory of normativity as a special case of practical reasoning offers an illuminating perspective on interpretation of state practice as network of reasons. Most importantly, it allows to clearly differentiate two instances of interpretation: interpretation of state practice for the purpose of identification of a rule of CIL, and interpretation for the purposes of clarification of its normative content.

5.1 Interpretation as Identification

The formation of a norm is a process of conversion of reasons and expectations, and it may not always be possible to draw straight lines between a stage when states act for a widely shared reason and do not explicate any expectations, a stage practice emerges, and a stage when it has fully developed. Sometimes, articulation of reasons is enough to generate expectations that these reasons would be followed and other reasons would be excluded, which exhibits a practice being formed (e.g., Truman proclamation on continental shelf). At other times, conversion of reasons into norms does not happen because of difficulties associated with balancing them in a most appropriate way, which causes uncertainties and disagreements as to whether a practice has emerged (e.g., uncertainties in regard of the right to unilateral secession). Yet some features of practical reasoning exploited by actors serve as beacons of there to be or not to be a norm and whether it may qualify as a legal one. Thus, when states’ actions are looked at with the purpose of enquiring whether a new rule of CIL has emerged, the network of reasons appears as a purported container, and what states do and how they react to what other states do get assessed within a logic of sources of law. This, first and foremost, affects the questions through which the interpretation of states’ actions is carried out:

1. (1) do states act for the first-order reasons only, and is there therefore only a semblance of a practice (‘pattern’ or ‘regularity of behaviour’)?

2. (2) or do states act for a second-order reason (i.e., norm), and is there therefore a formed practice, where participants exhibit expectations about reasons being followed or excluded, and shape their conduct based on these expectations?

3. (3) if the latter, then is this second-order reason acted upon and articulated as a part of a wider network of legally relevant reasons, that is, does it conform to certain conventional criteria of validity of custom as legal custom?⁴⁶

Questions (1)–(2) inquire into the existence of a second-order reason that states use as a justification for their action. There is a big difference between justification based on first-order reasons and justification based on second-order reasons, that is, norms. Justification based on first-order reasons does not purport any expectations from other actors and such justification may, as a matter of fact, be implicit and not designed as foreseeing, or matching, such expectations. This, however, is a much rarer situation than it may appear. In today’s world, states are much more often acting within practices than they used to, even when it relates to their internal affairs, and therefore justifications, even when implied, are typically met with expectations from other actors. Hence, it is normal that first-order reasons-based justifications are usually addressed to states’ actions in their domestic realm, though even there second-order reasons embedded into state practices play a more and more significant role.⁴⁷

The existence of a norm manifests in a reason that has a pre-emptive function, that is, a reason that counts for not acting on, and not using as a justification some other reasons. Not only does this mean that certain reasons cannot be legitimately acted upon, but it also entails that other states expect these reasons to be excluded and react accordingly when they are not. This, however, does not in and of itself mean that a norm embedded into a practice is a legal norm. There may exist mutual expectations as to what reasons may or may not be acted on, and what kind of second-order reason bridges them, even when these expectations do not have a manifestly legal character. International relations of states are by and large governed by such second-order reasons, which means that state practices (and hence also norms) are virtually omnipresent.⁴⁸

The first two questions, in such a way, are asked in order to determine the reasons-made boundaries of a container of a customary rule. The third question is quite different, though. It aims at establishing whether this container meets the requirements of validity set by a legal order. It is beyond the scope of this chapter to discuss the intricacies of legality of state practices, if only because they do not, in principle, contribute to the process of interpretation of rules of CIL. The dimension of the legality of customary rules, as was suggested in Section 2, is typically of little relevance to the determination of their content. Let us therefore shift to a different mode of interpretation: that aimed at clarification of the content of customary rules.

5.2 Interpretation as Clarification

A more specific, and more legally charged instance of interpretation of state practices is, certainly, interpretation for the purposes of clarification of the normative content of a rule of CIL. This instance of interpretation, however, tends to adopt a view of state practices as the content of rules, rather than their containers. This, though, does not change the nature of state practices as networks of reasons, and therefore the questions through which interpretation proceeds are again addressed to these networks, but these are very different questions:

1. (1) what first-order reasons does a rule of CIL exclude, that is, what reasons states may not legitimately invoke as justification for their actions within a given practice?

2. (2) what first-order reasons does a rule of CIL reinforce, that is, what first-order reasons does this norm account for, how does it balance them, and whether this balance corresponds to expectations of those involved in a practice?

Let us address the first question. Since exclusion is of crucial importance for differentiating norms from other reasons, interpretation of rules of CIL is primarily concerned with what reasons get excluded by a rule that is being interpreted. For example, is it meaningful within existing state practices to ascertain that a cyber-operation as a single ‘hostile’ act employed by one state against another constitutes an armed conflict within the meaning of customary rules of international humanitarian law (IHL)?⁴⁹ To translate this into the language of practical reasoning, may a state justify the reasons for its act of cyber warfare as not being excluded by the norms of IHL? This question can only be answered by looking at how states accommodate a new reason into existing deeds; whether they discursively assess cyber warfare as an instance of an armed conflict, or as something separate, probably creating an independent deed. By its very nature, exclusion is the function of a norm that renders acting on certain reasons as violation of this norm, and therefore when a new reason emerges from within a practice (a practice of modern warfare, in this example), it is a matter of whether states accommodate this new reason within existing normative deeds, or whether they exclude it from practice and prevent it from becoming its deed. This process of accommodation or rejection typically manifests in how states react to a new reason being invoked as a justification for an action within an existing practice.

Where the first question addresses the external boundaries of a practice, that is, the issues of what kind of reasons count as parts of a practice and what kind of reasons are excluded from it, the second question offers a different perspective. It relates to the justification of norms, briefly touched upon in the previous section. Norms, including legal norms, are typically justified as accounting for a certain balance of the first-order reasons that render a practice intelligible. From this perspective, norms always serve a purpose of simplifying or optimising participants’ compliance with these first-order reasons.⁵⁰ Interpretation, therefore, may not only address the issues of exclusion of some reasons, but also the issues of reassessing or even reshaping the balance of reasons that are included in practice. Thus, it is a matter of interpretation to inquire whether a rule of CIL adequately reflects and accounts for underlying reasons that shape a practice and guide a state’s actions.⁵¹ If, for example, the principle of equidistance as a method of delimitation of continental shelves does not properly account for the reasons that comprise the practice of the use of continental shelf, there may exist a need to rebalance these reasons according to a more fundamental principle.⁵² Such a rebalancing, though made within a wider normative framework of equity, does affect the balance of reasons represented by the equidistance rule; some of the reasons it accounts for are now weightier.

These two questions, though different from those discussed in the previous subsection, build on them. It is in the foundations of legal interpretation to enquire into what considerations and in which particular manner legal norms account for, since it is the core function of legal norms to improve or simplify people’s compliance with reasons. And since in the case of customary rules their content and container are one and the same thing, their interpretation ultimately entails clarifying the boundaries of practices. The need for this clarification reflects that it is in the nature of practices to evolve. The normative deeds comprising the inherent normative standard of a practice are typically on the move; not only do they depend on what participants in practices do, but also on how they react to actions. Thus, those engaged in practices constantly accommodate or reject new reasons that may or may not affect the perception of the normative standard, and this is exactly why interpretation of customary rules is essentially an inquiry into the dynamics of practical reasoning implied within a practice.

It is, therefore, not only possible that rules of CIL allow for evolutive interpretation, it is essential, since it follows directly from the way practices operate and develop. It should be noted, however, that evolutive interpretation in the case of treaties is not the same as in the case of CIL.⁵³ For interpretation of treaties, evolutive interpretation generally relates to the phenomenon that when treaty text remains the same, its meaning is altered in the course of time.⁵⁴ It is argued that evolutive interpretation of treaties is justified when there is evidence that the parties intended, from the outset, that their treaty would be capable of evolving over time, that it can remain effective or relevant in the face of changing conditions.⁵⁵ It is, therefore, essential that evolutive interpretation of treaties is based on the provision of Article 31(1) of the VCLT, according to which ‘a treaty shall be interpreted … in the light of its object and purpose’. When a rule of CIL is in question, though, it seems not entirely accurate to speak of its object and purpose, since rules of CIL cannot be always traced back to some shared intentions (as argued in Section 2), their object and purpose are far less clear and determined than in the case of treaties. In the case of treaties, their object and purpose may be an explanandum for the purposes of interpretation, but in the case of rules of CIL, they rather appear as explanans. In other words, an object and purpose of a customary rule may well be the end point of interpretation rather than its starting point. For this reason, evolutive interpretation of the rules of CIL relates more to the function a certain practice performs and to the meaning its practice has in a wider context of states’ activities. Such an evolutive interpretation, then, focuses on re-evaluating the balance of reasons reflected in a norm, adjusting it to the developing patterns of practice itself.⁵⁶ Every instance of interpretation of a rule of CIL is therefore a snapshot of the balance of reasons currently accepted within a practice. However, since practices are dynamic entities, so are the norms which define them and are sustained by them.

To summarise, the interpretation of state practices as normative networks of reasons takes different shape depending on how they are looked at. If a state practice is approached as a container and is thus investigated for the purposes of identification of a rule of CIL, the main strategy of interpretation will consist in assessing whether states act for a second-order reason (a norm, in this context) and whether it meets the threshold of legal validity. When a state practice is addressed as content, the interpretative strategy will primarily entail determination of those reasons a rule of CIL excludes and assessment of whether those reasons it accounts for are properly balanced.

6 Conclusions

It is in the core of the idea of CIL that it manifests in a peculiar duality; it is a source of international law, and at the same time it is international law as such. By blurring the line between container and content, which is essential for the conventional doctrine of sources, CIL challenges the process of its interpretation too. State practices, which appear as both containers and content of rules of CIL, are subject to interpretation from two different positions – when a new rule is identified, and when an existing rule is clarified. This creates confusions as to how to separate these two instances of interpretation.

This chapter endorses the view on state practices as inherently normative networks of reasons. Approached as such, state practice manifests as comprising of deeds and reasons, the latter existing on two different levels. The normativity of a state practice is explained through there being second-order pre-emptive reasons, that is, norms that bridge a variety of first-order reasons, balancing and mutually rendering them meaningful. The interpretation of these norms embedded into state practices entails discovering connections between different groups and levels of reasons. The interpretation for the purpose of identification of a rule of CIL is primarily concerned with a question of whether there is a second-order reason that systematises expectations and critical stances of states, and whether this second-order reason qualifies as a legal one. The interpretation for the purposes of clarification, in turn, focuses on what reasons a rule of CIL excludes, and what reasons it balances, how well this balance reflects the actual weight of the first-order reasons, and how to ensure that newly formed reasons are properly assessed and accommodated within practice, or get excluded from it.

Such an approach to state practice and interpretation of CIL allows one to distinguish different interpretative stages of a lifecycle of state practice, as well as to conceptualise state practice as a normative network of reasons.

18 Different Strings of the Same Harp Interpretation of Rules of Customary International Law, Their Identification and Treaty Interpretation

Marina Fortuna ^*

1 Introduction

At the heart of all knowledge lies difference – the ability to distinguish one concept from another one.¹ From Heraclitus to Derrida and Deleuze, philosophers have grappled with issues of identity and difference and, though not settling on a single truth, have equipped humanity with a conceptual toolbox to categorise human experience.² Differentiation between concepts and their objects is as important in law as it is in other disciplines and is one of the fundamental instruments in the toolbox of legal scholars in their pursuit to understand the workings of law.

Considering this, the present chapter is a reflection in broad brushstrokes on the differences between three interconnected judicial operations: interpretation of customary rules, identification of customary rules and treaty interpretation. While identification of customary rules and treaty interpretation have been explored comprehensively, interpretation of customary rules is a recent addition to the thread of under-researched and complex topics in international law. Until recently hardly anybody throughout the existence of international law has ever asked whether customary international law (CIL) could be or had been interpreted. Today, when the question has been asked, the community of international lawyers faces a difficult task. The difficulty of this task owes, firstly, to the lack of agreement among scholars on the meaning of the concept of interpretation, a notion which is used in legal scholarship with various meanings. Interpretation is perceived both in its hermeneutic dimension (as the determination of the meaning of and the intention behind an act/words/behaviour), but also as a wider concept including within it legal construction. Secondly, hardly any theoretical account of CIL can coherently explain what it is, how it emerges and how it develops, thus, occupying a sort of sui generis space among other sources. Thirdly, interpretation of customary rules seems to be difficult to distinguish, both in theory and in practice, from identification of customary rules and treaty interpretation, which owes largely to the first two reasons.

Interpretation of customary rules is not always appropriately distinguished from identification of customary rules because (1) customary rules are perceived as being equivalent to elements of custom (state practice and opinio juris) and (2) because, according to some scholars, there is an inherent element of interpretation in identification. This linguistic similarity, instead of leading to more clarity, contributes to a greater confusion between the two types of judicial acts.

The difference between interpretation of customary rules and treaty interpretation is another issue which is addressed in this chapter. Unlike the distinction between interpretation of customary rules and their identification, the potential confusion is not linked so much to terminology, as it is to practice. The fact that interpretation of customary rules and treaty interpretation are two different judicial acts (just like customary rules and treaty rules are two different sources of law) has been overlooked in some cases. Two examples are given where international judges engaged in an act of treaty interpretation to clarify the content of customary rules. This non-recognition of the distinction between the two may lead either to a misapplication of the law or to solutions which do not accurately reflect the content of customary rules.

The working definition adopted in this chapter of interpretation of customary rules is ‘the act of determining/construing the content of customary rules the existence of which is unchallenged’. This definition is inspired from a preliminary analysis of the case law on the subject and on the definition of interpretation of customary rules in the meaning used by Merkouris and Orakhelashvili (see Section 1(b)). Both legal scholars conducted an inquiry into the case law of international courts and tribunals and observed that, firstly, judges do not only gather state practice and opinio juris in order to determine the content of customary rules and secondly, that judges use methods of treaty interpretation or similar methods to establish the substance of customary rules. The fact that judges themselves refer to this latter process as interpretation and given the similarity (and sometimes even identity) of the methods used with those employed in treaty interpretation these two scholars settled on the notion ‘interpretation of customary rules’ as the best description for this process.

This chapter encourages further scholarly reflection on the distinction, both in theory and in practice, between these different judicial acts. While all three make up a kind of unity (especially in legal practice) and as strings of a harp work together to build the content and further the evolution of the content of CIL, they remain distinct and should be, according to this author, recognised as such.

The chapter is structured along three main sections. Section 1 provides a contextual background by addressing the concept of interpretation and the arguments supporting the amenability of customary rules to interpretation. It is followed by Section 2, which examines the differences between interpretation of customary rules and their identification (in particular, interpretation in identification) and Section 3, which focuses on the distinction between interpretation of customary rules and treaty interpretation.

2 Interpretation of Customary Rules: The Concept

This section seeks to unravel the meaning of the concept of interpretation of customary rules, while at the same time demonstrating why interpretation of customary rules is not a contradiction in terms. The section starts off by describing the concept of interpretation of customary rules in legal scholarship (Section 2.1) and the arguments against the interpretability of customary rules (Section 2.2). It is followed by the argument concerning the reasons why customary rules are amenable to interpretation (Section 2.3) and the working definition of interpretation of customary rules (Section 2.4).

2.1 The Concept of Interpretation of Customary Rules in Legal Scholarship

Judging by reference to the hundreds of years of international law’s existence, the concept of interpretation of customary rules is quite novel. Arguably the first legal scholar who discussed interpretation of CIL was Charles de Visscher. In Problemes d’Interpretation Judiciaire en Droit International Public de Visscher examined two dimensions of interpretation with respect to custom: interpretation as part of the customary process of law creation and interpretation of customary rules proper.³ As part of the development/formation of customary rules interpretation was perceived by de Visscher as a value judgement on the content of a customary rule made by a relevant agent subsequently to the observation of patterns of repeated state practice. These patterns of facts were to be evaluated in light of moral and social imperatives.⁴ In contrast, interpretation of customary rules was seen as an act of judicial elaboration or, more precisely, the adaptation of general customary norms to particular situations⁵ that marked the transition from the abstract norm to the concrete norm.⁶ De Visscher believed that there are two types of customary rules: customs, the essential components of which make up a hard core and, thus, are rarely, if ever, subject to dispute, and customs in the case of which a dispute may arise around its nucleus, where a fringe of indeterminacy always remained.⁷ The subject was taken over by Sur who, in line with the views expressed by de Visscher, advanced a tripartite classification of interpretation in relation to international custom: (1) interpretation that establishes the existence of a customary rule, (2) interpretation that establishes the content of a customary rule and (3) interpretation that establishes the scope of a customary rule.⁸ For both scholars, interpretation was indispensable to all stages of the existence and development of custom. This position on the omnipresence of interpretation in the life of custom taken as a whole is shared today by contemporary scholars like Alland and Tassinis (see Section 2).⁹

Around the same time that Sur wrote on the subject of interpretation of customary law Bleckmann, a German legal scholar, published a paper on the identification and interpretation of CIL. According to Bleckmann, customary rules were to be determined by induction – the abstract legal principle being derived from practice – and applied to new factual situations by deduction, which could involve the interpretation of the abstract legal principle.¹⁰ Considering this, customary rules, as abstract legal principles, were subject to grammatical, systemic and teleological interpretation.¹¹ In support of this position Capotorti, in his 1994 general course at the Hague Academy of International Law, stated that rules of interpretation enshrined in the Vienna Convention on the Law of Treaties (VCLT) which have a customary basis also regulate the interpretation of international custom and that of other sources of law.¹² More recently, somewhat similar was made by Merkouris and Orakhelashvili.¹³ Both Merkouris and Orakhelashvili conducted an analysis into the case law of international courts and tribunals and have revealed a plethora of cases where judges either use the notion of interpretation with respect to customary rules, or without doing so, employ methods from treaty interpretation.¹⁴ This led them to conclude that interpretation of CIL is not only possible, but actually happens in practice. Since judges used the term ‘interpretation’ with respect to the determination of the content of customary rules in a way that is different from inquiring into state practice and opinio juris and due to the obvious similarities with treaty interpretation, both scholars reached the conclusion that the process of content determination of previously established customary rules is none other than an interpretative act. Thus, the definition advanced by Orakhelashvili (probably induced from the plethora of practice) is that interpretation of customary rules refers to the clarification of ‘the modes and details of applicability of customary rules to specific situations to which they are designed to apply due to their general scope’.¹⁵

2.2 Arguments against the Amenability of Customary Rules to Interpretation

Two main arguments have been forwarded against the amenability of CIL to interpretation.¹⁶ Firstly, it has been argued that the identification of customary rules is the only operation which establishes its content and, thus, any form of clarification of a customary rule would require a new stage of identification and, secondly, that the object of interpretation can only be written law and, since customary rules are unwritten, they cannot be subject to interpretation.¹⁷ These two arguments depend on (1) the understanding of what customary law is and (2) the definition given to interpretation. Essentially, a new cycle of identification is required each and every time only if international custom is equivalent to its constituent elements. Moreover, interpretation is only confined to written rules depending on the definition of interpretation one adopts.

2.3 Reasons in Favour of the Amenability of Customary Rules to Interpretation

2.3.1 Customary Rules Distinguished from Elements of Custom

To address the first argument against the interpretability of CIL, a distinction is made between customary rules and elements of custom.

According to Article 38(1)(b) of the Statute of the International Court of Justice (ICJ), CIL is one of the sources of law to be applied by the ICJ. In conformity with the provisions of this article, CIL is ‘general practice accepted as law’, which, according to the case law of the ICJ, is comprised of two elements: state practice and opinio juris.¹⁸ Nonetheless, it is not its only meaning. Even a cursory glance at the discussions surrounding the conclusion of the Statute of the Permanent Court of International Justice reveals that ‘custom’ is also another name for the process of law development. In the words of the chairman of the Drafting Committee, ‘[custom] is a very natural and extremely reliable method of development, since it results entirely from the constant expression of the legal convictions and of the needs of the nations in their mutual intercourse’.¹⁹ Thus, one should distinguish between custom as the process and custom as the product of this process.²⁰

At the same time, custom is also the name for the legal norm derived from general practice accepted as law. This meaning of custom is important to emphasise as some legal scholars oppose the possibility of interpretation of customary rules by reducing custom to its constituent elements. For instance, Bos rejects the amenability of CIL to interpretation on the basis that the content of custom is determined simultaneously with its existence.²¹ But this argument would only hold true if custom was identical to its elements. That this is not the case is evidenced, firstly, by the language employed by the International Law Commission (ILC) in its recent Draft Conclusions on the identification of CIL²² and, secondly, confirmed by the opinions of established international legal scholars, theorists of law and by the ICJ itself (see Section 2).

In its conclusions on identification of international custom the ILC has implicitly supported the division between constituent elements of custom and customary rules when it defined CIL as ‘unwritten law deriving from practice accepted as law’. This implies that customar (the rule itself) is not equivalent to state practice, as the act of deriving means obtaining something from something else – if there is identity between two things, then one cannot derive something from itself; in other words, the act of deriving requires for two different things to be present. What follows logically is the lack of identity between the unwritten rule of customary law and its elements (state practice and opinio juris).

Turning to legal scholarship, influential scholars such as de Visscher and Sur, who (while considering interpretation omnipresent in the life cycle of CIL – a point we shall return to later in the chapter) explicitly talked about ‘interpretation de la regle’.²³ In a similar vein, Kelsen, in his discussion of custom as a source of law made the observation that ‘such is the nature of those particular facts which together constitute the existence of the “custom”, ‘creating the general rule’.²⁴ Needless to say, judgments of international courts and tribunals frequently refer to ‘customary rules’,²⁵ thus confirming the distinction between customary norms and state practice with opinio juris.²⁶ This distinction between state practice and opinio juris as elements of custom and customary rules is aptly portrayed in this volume by Gorobets with the container versus content metaphor.²⁷ Slightly adapting the original metaphor, it is possible to argue that the elements of custom are the containers, whereas the customary norms are the content,²⁸ similarly to the distinction in treaty law between the treaty as instrumentum and the norms contained in the treaty.²⁹

2.3.2 Interpretation in Public International Law as Applicative Construction

The second criticism against the amenability of CIL to interpretation rests implicitly on each author’s understanding of the term ‘interpretation’.

Black’s Law Dictionary defines interpretation as ‘the process of determining what something, esp. the law or a legal document, means; the ascertainment of meaning to be given to words or other manifestations of intention’.³⁰ This definition is different from the ordinary meaning of interpretation/to interpret as ‘the way in which someone explains or understands an event, information, someone’s actions etc’,³¹ ‘to explain or tell the meaning of’ or ‘to conceive in light of individual belief, judgment, or circumstance’,³² ‘an explanation or opinion of what something is’.³³ While the general notion of interpretation is tied to meaning,³⁴ which can be the meaning of any object that is meaningful, legal interpretation necessarily requires (as per its definition) that the object is a law or a legal document (and the words contained in it) or, in any event, a manifestation of a (legal) intention (it is presumed that the dictionary referred not to just any intention but legal intention), as an intention to enter/create legal relations/to produce legal consequences. While philosophers still debate on the meaning of interpretation and the space that intention occupies in it,³⁵ for the purposes of this inquiry it suffices to say that legal dictionaries are reflections of a certain consensus within the epistemic community/interpretative community in the discipline. Therefore, in law interpretation is generally tied to some kind of manifestation of intention and is an act which unravels this intention.

Zooming into the discipline of public international law (PIL), the prototype of interpretation is interpretation of treaties, especially since the VCLT, which codified the rules of treaty interpretation. In an illuminating account of what interpretation is for the community of international lawyers, Kammerhofer noted that, as opposed to interpretation, strictly speaking, in PIL interpretation is the name for ‘an applicative construction of the law’s meaning’,³⁶ which involves both the extraction of legal meaning, but also ‘the concretisation of abstract general norms in individual instances’.³⁷ While in some domestic legal systems a distinction is made between interpretation, as clarification of semantic meaning of legal texts, and construction, as the judicial activity of determining a rule’s scope of application and the resolution of gaps and contradictions,³⁸ this distinction was intentionally dismissed upon the drafting of the first Draft Convention on the law of treaties and, subsequently, of the VCLT.³⁹ Compared to the general legal definition offered by Black’s Law Dictionary, the definition of interpretation contained in the VCLT (as an authoritative document on the matter) is considerably wider and goes beyond the mere determination of intention and, therefore, it is reasonable to assume that the meaning in which judges use ‘interpretation’ is also wider than the stricter meaning of interpretation in law more generally. Thus, interpretation of customary rules is not a contradiction in terms or a misconception even if the analysis performed for the purpose to determine a rule’s content disregards intention and focuses on other reference points within the parameters of the language of our discipline. Similarly to treaties, customary rules can be interpreted in the sense of construing their content on the basis of considerations such as teleology, the interconnectedness of norms in the system of law etc. and, thus, interpretation of customary rules is not a contradiction in terms.

2.4 Interpretation of Customary Rules: A Definition

Considering the aforementioned, the working definition of interpretation of customary rules is ‘the act of determining/construing the content of customary rules the existence of which is unchallenged’. This is the definition which, as previously demonstrated, makes sense from a theoretical standpoint, but also best describes the instances of judicial practice in which the content of customary rules is determined differently than by looking at state practice and opinio juris.

3 Interpretation of Customary Rules versus Identification of Customary Rules

Having in mind the definition given to interpretation of customary rules, this contribution now turns to discussing the differences between interpretation of customary rules and their identification.

According to Merkouris, CIL identification is both a process of law-ascertainment and a process of content determination.⁴⁰ By examining evidence of state practice and opinio juris it seeks to determine whether a customary rule exists and what its content is. Similar to identification, interpretation of CIL is also a process of content determination. However, it is a process of content determination that takes place only after the customary rule has been first identified. This relationship between the two processes can be seen as mirroring (to a certain degree) what happens in treaty law. Firstly, the judge finds the relevant applicable rule (which, strictly speaking, is an act of law-ascertainment), which already has a content embodied in the text, and only then the adjudicator can proceed to the interpretation of the rule.

It is quite common to refer to the judicial act which happens at the stage of identification of a customary rule as interpretation. The term is used in four situations: to describe the conglomerate of state practice and opinio juris,⁴¹ to refer to the process of evaluating the mass of state practice and opinio juris,⁴² to derive/infer the relevant customary rule from the mass of state practice⁴³ or to connote the analysis of a singular sample of state practice and the motivation behind it.

Firstly, unlike the somewhat ideal (and sterile) model which allegedly describes the process of identification of customary rules by way of spotting samples of state practice, identification of customary rules is argued to be rarely just about discretionary data collection and, more often than not, as involving some form of interpretation.⁴⁴

Secondly, interpretation is deemed necessary in situations where there is inconsistent state practice and opinio juris.⁴⁵ Such a problem occurs when there are simultaneously examples of state practice supporting the fact that a customary rule has emerged and equally compelling examples of contrary behaviour on behalf of other states. An example given in this sense is the prohibition of torture.⁴⁶ On the one hand, some states do not engage in acts of torture, whereas, on the other hand, there are examples of states that torture individuals and do so without protest from third states. In such a case, the argument goes, there are two possible interpretations of state practice: (1) torture is permitted and (2) torture is prohibited,⁴⁷ and the decision should ultimately be made on the basis of considerations of morality as an implementation of the Rawlsian theory of reflective equilibrium.⁴⁸

Thirdly, there is also some measure of interpretation at the stage of deriving norms from patterns of state practice as ‘the same set of data can support indefinite series of statements as to what the content of the law is’.⁴⁹ This is connected to the previous use of ‘interpretation’ with the difference that in this case it is argued that interpretation is always present, even if there is no inconsistent practice as such. In this case, ‘interpretation’ means ‘formulations of logical propositions describing the norm that we might infer from such conduct’.⁵⁰

Finally, ‘interpretation’ is used as a synonym for the process of assessing the motivation of a state behind a specific behaviour, such as allowing another state’s warship to enter its port without authorisation.⁵¹ Such an interpretative act then contributes to the understanding of whether opinio juris, understood as a collective agreement on a rule, as opposed to the singular motivation of each state, is present. Yet the analysis, as opposed to the first case, is made at the level of a singular specimen of practice, not at the level of the whole mass of state practice.

Qualitatively these types of ‘interpretation’ are different from interpretation of customary rules. Firstly, they are different by reference to their object as they concern the elements of custom, as opposed to the customary rule itself. Secondly, all of them are concerned with what is not yet law and, therefore, do not squarely fit into the notion of legal interpretation. For instance, interpretation in describing state practice and opinio juris is a form of perceptual evaluation,⁵² and focuses on the cognitive dimension (understanding) as opposed to legal interpretation. What is labelled as interpretation in the case of inconsistent state practice is, although similar to legal interpretation (in the sense that it requires a judgment/decision to be made on alternative propositions), primarily a process of law-ascertainment and not an interpretation of a law the existence of which was previously acknowledged. Additionally, the act itself is more an exercise in judging than it is in interpretation, understood in its legal sense. The same can be said of ‘interpretation’ at the stage of deriving a customary norm from legal practice. Finally, the assessment of the motivation behind an instance of state practice, while similar to interpretation in the sense of an act concerned with deciphering legal intention, is, again, part of an exercise in law-ascertainment, as opposed to legal interpretation, because it is an interpretation of the meaning of facts and not of the meaning of existing law.

Even if these acts could be described as interpretative in nature by reference to the ordinary meaning of interpretation (which is also very general), it is still more beneficial to have them distinguished terminologically. Using interpretation at both stages may create confusion and already does, given the complex nature of CIL, which balances between fact and law. Since the VCLT already codifies (implicitly) an authoritative meaning of interpretation, using it with the same meaning with respect to rules of CIL will contribute to linguistic consistency within the discipline. In other words, since the meaning of interpretation, as derived from its use with respect to sources of international law other than CIL, is both interpretation understood strictly and construction, it is better to confine the notion of ‘interpretation of CIL’ to the posterior content determination of customary rules in a way that mirrors treaty interpretation and brings more unity to the system as a whole.

Another argument in favour of using the notion of interpretation only with respect to the content determination of customary rules (as opposed to the evaluation of state practice and opinio juris) is the difference in the aims of the two judicial acts. The initial content determination process seeks to find the customary rule itself and initially determine its content – to make the inductive generalisation out of a collection of state practice, which, even if requiring some kind of interpretative reasoning as method, does not undermine the fact that it is an exercise of law-ascertainment. The subsequent act of content determination is concerned not with law-ascertainment, but rather with construing the relevant norm in a way that contributes to the solution of a dispute. Thus, it is the position of this author that the different aims of the two judicial acts should be reflected in the name of these processes. This is best done by confining the notion of interpretation solely to the subsequent act of content determination.

4 Interpretation of Customary Rules versus Treaty Interpretation

Another difference worth reflecting upon is the one between interpretation of customary rules and treaty interpretation. Both Merkouris and Orakhelashvili noticed a similarity in the methods that different judges or different international courts used when determining the content of a rule past the identification stage.⁵³ For instance, judges have referenced the technique of interpretation by reference to ordinary meaning in Hadzihasanović.⁵⁴ Two legal issues were raised in this decision: (1) whether the principle of command responsibility applicable to international armed conflict is also applicable to non-international armed conflict and (2) whether a superior can be punished under the principle of command responsibility for acts committed by subordinates prior to the assumption of command. Having found no specific state practice and opinio juris on the principle of command responsibility for acts committed in non-international armed conflict, the tribunal argued that ‘where a principle can be shown to have been so established [on the basis of state practice and opinio juris], it is not an objection to the application of the principle to a particular situation to say that the situation is new if it reasonably falls within the application of the principle’.⁵⁵ The interesting part was that in order to support this argument, the Appeals Chamber of the ICTY relied on the prohibitions contained in common Article 3 of the Geneva Conventions and reasoned that ‘in the absence of anything to the contrary, it is the task of a court to interpret the underlying State practice and opinio juris … as bearing its normal meaning that military organization implies responsible command and that responsible command in turn implies command responsibility’.⁵⁶ Leaving aside the convoluted language and possible concerns regarding the strength of the court’s argument, in order to make the argument concerning the meaning of a previously established customary rule, the court borrows the language of treaty interpretation and although it mentions state practice and opinio juris, it actually mentions the preexistent rule.

In the Orić case, where the ratio decidendi in Hadzihasanović was the object of contention,⁵⁷ Judge Schomburg, as one of the dissenting judges, argued that the customary principle of command responsibility must be interpreted by giving ‘consideration to the purpose of a superior’s obligation to effectively make his subordinates criminally accountable for breaches of the law of armed conflict’.⁵⁸ He then emphasised that ‘considering thus the purpose of superior responsibility, it is arbitrary – and contrary to the spirit of international humanitarian law – to require for a superior’s individual criminal responsibility that the subordinate’s conduct took place only when he was placed under the superior’s effective control’.⁵⁹

In Furundžija⁶⁰ the court faced a question concerning the definition of rape and the forms of behaviour that fall under this offence (in particular, whether oral penetration can qualify as rape). The Trial Chamber, firstly, stated that the prohibition of rape in armed conflict has evolved into a norm of CIL,⁶¹ yet found that international law (either treaty or custom) contains no definition of rape.⁶² Then, it scrutinised national legislation and found major discrepancies between the criminal laws of various countries as to the definition of rape and whether oral penetration qualifies as rape or a different type of sexual assault.⁶³ Lastly, it resorted to the principle of respect for human dignity to interpret the crime of rape. The Trial Chamber noted ‘it is consonant with this principle [principle of protection of human dignity] that such an extremely serious sexual outrage as forced oral penetration should be classified as rape’.⁶⁴ As the statement reveals, the Trial Chamber did not apply the principle of protection of human dignity to the case directly,⁶⁵ but it determined the definition of rape in consonance with this principle.⁶⁶ This example can be taken as a form of interpretation similar to systemic interpretation in treaty interpretation.

What runs like a red thread through these examples are the arguments typically resorted to for the purposes of treaty interpretation (it would not be far-fetched even to argue that judges apply the same interpretative techniques by analogy). Depending on the interpretative method used, some cases raise important questions concerning the relationship between interpretation of customary rules and treaty interpretation. For example, in the previously mentioned Hadzihasanović case one of the Appeal Chamber judges appended a dissenting opinion where he noted that ‘any interpretation [of the customary rule] can be made by reference to the object and purpose of the provisions laying down the doctrine’.⁶⁷ By the same token, in the North Sea Continental Shelf case, in his dissenting opinion, Judge Sørensen observed:

If the provisions of a given convention are recognized as generally accepted rules of law, this is likely to have an important bearing upon any problem of interpretation which may arise. In the absence of a convention of this nature, any question as to the exact scope and implications of a customary rule must be answered on the basis of a detailed analysis of the State practice out of which the customary rule has emerged. If, on the other hand, the provisions of the convention serve as evidence of generally accepted rules of law, it is legitimate, or even necessary, to have recourse to ordinary principles of treaty interpretation, including, if the circumstances so require, an examination of travaux preparatoires.⁶⁸

In legal scholarship these and other similar examples have been frowned upon either as a failure to distinguish between treaty interpretation and identification of custom⁶⁹ or as a disregard for the fact that customary rules possess an independent rationale and should be assessed by reference to it, rather than by reference to a treaty’s object and purpose.⁷⁰ The main point behind these criticisms is the need to keep interpretation (or identification) of customary rules separate from treaty interpretation. The danger is that using reference points related to the treaty counterpart of the customary rule may lead to a misapplication of the law — to the application of a treaty rule which is not clearly established as a customary rule or the usage of considerations which are related to the treaty, but not, as such, connected to the customary rule.

According to this author, the answer should be nuanced depending on the type of customary rule involved — a question which ties to the relationship between customary rules and treaty rules more generally. Generally speaking, the relationship between custom and treaties is a multifaceted one. On the one hand, treaties may codify, crystallise or lead to the creation of customary rules. On the other hand, treaties may be used to confirm the existence of a customary rule in the process of identification. According to the empirical study conducted by Choi and Gulati, treaties are the dominant form of evidence in the ascertainment of customary rules.⁷¹ Not only the existence, but also the content of customary rules may be determined by reference to treaty provisions, which includes the situation when the content of a customary rule is determined posterior to the acknowledgement of its existence. Otherwise said, treaties can be an important reference point in the interpretation of customary rules. According to the ICJ itself, ‘multilateral conventions may have an important role to play in recording and defining rules deriving from custom’.⁷² In a similar vein to what Bleckmann (one of the forerunners of the concept of interpretation of customary rules) argued with respect to using state documents for the purposes of grammatical interpretation of custom, treaties can be used to define concepts contained in CIL, when their meaning is disputed. Moreover, just as customary rules have been used for the purposes of treaty interpretation under Article 31(3)(c), treaties (and general principles of law) can be used for interpreting customary rules as a form of systemic interpretation. For the purposes of interpretation of customary rules, judges may use both codification treaties which contain provisions with content similar to that of the customary rule or on the same subject matter, or, equally, treaties that are neither codifications of customary rules, nor belong to a different (albeit, possibly related) subject matter. A relevant example in this sense is Judge Guillaume’s suggestion in the Advisory Opinion on Nuclear Weapons that rules of jus ad bellum may aid the clarification of the rules of the jus in bello. However, the problem arises at the level of argumentative reference points such as the ordinary meaning of terms, context, travaux preparatoires, intention of the parties or object and purpose and the answer as to whether each one of these reference points may be used for the purpose of a customary rule’s content determination should depend on the type of customary rule involved. As rightfully pointed out by Judge Jennings in his Dissenting Opinion to the Nicaragua case:

[t]o indulge the treaty interpretation process, in order to determine the content of a posited customary rule, must raise a suspicion that it is in reality the treaty itself that is being applied under another name. Of course this way of going about things may be justified where the treaty text was, from the beginning, designed to be a codification of custom; or where the treaty is itself the origin of a customary law rule.⁷³

When a treaty is a codification of customary rules, either completely or preponderantly, it could be imagined that the judge heavily relies on the text of the treaty, since having a text as a reference point allows for a more straightforward interpretation of a rule, which might be more acceptable to the subjects of law because of its predictability (achieved through the written/codified nature of the rule). However, ideally, this should only be permissible provided that there are no indications that the customary rule has evolved posterior to its codification.⁷⁴ In the case of crystallised rules or rules which passed into CIL from treaties the same considerations apply.

With regard to other reference points, such as context or travaux preparatoires, matters are slightly different. For instance, in the case of a customary rule codified in a treaty a contextual interpretation may use the context of the treaty if other rules are also a codification of CIL. It is more likely to admit an interpretation which uses the context of the treaty as a reference point in customary rules crystallised or which evolved into a rule of custom from a treaty, as the treaty serves as their springboard. Also, it is doubtful whether it is possible to use the context of the treaty (especially other provisions of the same treaty) when the customary rules (either codified, crystallised or evolved from a treaty) do not form an organic unity or unity of origin with the other provisions. Such a unity may be created by the fact that two norms belong to the same sub-branch of international law. A multilateral treaty which contains provisions from different fields of international law and which does not contain other customary rules except the one which is under scrutiny will be an unlikely candidate as a reference point for interpreting the customary rule in question. This is unless these other types of rules are used as a form of systemic interpretation justified by the fact that they are somehow related to the dispute and, thus, to the customary rule which is interpreted.

Travaux preparatoires may be used for the purposes of analysing the content of customary rules crystallised from or evolved from treaties as they may aid in determining the precise meaning and, thus, scope of a customary rule, again, unless there is evidence that the content of the customary rule has changed through time. As for object and purpose, the treaty’s rationale can hardly be a valid reference point, unless the treaty as a whole is a codification and there is some kind of organic unity in its provisions. This is because the object and purpose of the treaty may be much wider than the subject matter to which the customary rule refers to. For instance, the object and purpose of a regional treaty between a handful of states which declares in the Preamble that its aim is the maintenance of friendly relations between the parties can hardly be used as an interpretative reference point to interpret a customary rule on environmental protection (even if it has emerged from this treaty) as it does not aid in clarifying the content of this rule. In any event, automatic application of reference points from treaties to CIL is not advised and alertness should always be present when a treaty is used to aid the interpretation of customary rules.

5 Conclusion

This chapter was a reflection on the differences between, on the one hand, interpretation of customary rules and their identification, and interpretation of customary rules and treaty interpretation, on the other. Section 2 examined the concept of interpretation of customary rules by firstly distinguishing between elements of custom and customary rules proper and, secondly, by presenting the different meanings that the term interpretation may have. While the general definition of interpretation is ‘understanding’, the legal definition is limited to the determination of meaning of words or other manifestations of intention. Even more importantly, in PIL the term ‘interpretation’ is not limited to its hermeneutical dimension but can better be described as a form of applicative construction.

Section 3 examined the difference between interpretation of customary rules and their identification. Firstly, there is a difference in the object of the analysis and, secondly, there is a qualitative difference in the process, which, while possible to be regarded as ‘interpretation’ in its ordinary meaning, does not fall within the notion of ‘legal interpretation’.

Section 4 discussed the differences between interpretation of customary rules and treaty interpretation. While the methods of interpretation may be similar, using reference points from treaty interpretation will not always be a sensible solution and judges should remain alert to the differences between the two sources of law.

Taking the points made in this chapter as a whole, the crux of the matter is that the processes that have been analysed are part of the same palette that judges use when giving a solution on a case. Nonetheless, these operations are in meaningful ways different from each other, just like different strings of the same harp, and it is important to remain alert to these differences, both in theory and in practice.

19 Customary International Law Identification versus Interpretation

Riccardo Di Marco

1 Introduction

When dealing with a difficult issue such as the theory of interpretation,¹ the first obstacle to be faced concerns the nature of the object under examination: is interpretation relevant to a point of law or not?² Each doctrinal orientation would give a different answer. Some scholars consider that interpretation is an intellectual operation;³ others define interpretation as a creative activity;⁴ still others argue that interpretation is a linguistic issue, maybe even a methodological one, but, in any case, not a legal matter.⁵ On the contrary, some scholars incorporate the study of interpretation into positive law:⁶ by perceiving the legal character of the object, they act on the ground of the so-called rules of interpretation.⁷ It is impossible to give an exhaustive picture of such a debate in only a few lines.⁸ I will confine myself to note that international law writers consider the matter under a different light compared to scholars of other juridical systems. In fact, with respect to public international law, a clear position has already been taken: I refer to the Vienna Convention on the Law of Treaties (VCLT)⁹ that, while codifying the law of treaties,¹⁰ included certain rules of interpretation.¹¹ Even though sometimes slightly modified, these rules of interpretation have been constantly applied by international tribunals. Internationalists, usually hindered by the soft formalism of the international legal order, in this matter enjoy a privileged position.

To interpret a rule means to seek and understand its exact meaning, and, as a consequence, to clarify its scope, in order to be able to correctly apply it to the material case. In fact, since a rule is susceptible to different applications – because of its character of generality and abstractness – that content must be specified from time to time for the particular case. To determine the meaning of a rule, thus, the interpreter must accomplish a task of cognition (or recognition). This creative activity also raises practical issues: to which types of rules can interpretation be applied? Which theoretical-methodological tools should the interpreter use? With regard to customary rules, is it possible to separate the two distinct processes of identification and interpretation?

Bearing in mind the horizontal nature of the international legal system¹² as well as the important role played by customary rules in public international law, it is worth considering the following question: is it possible to apply to custom the international rules of interpretation (that, on their turn, are customary too)? In other words, is it possible to interpret customary international law (CIL) or can it only be identified? Hence, how can internationalists distinguish interpretation from identification with respect to customary rules? Has the International Court of Justice (ICJ or ‘the Court’) provided some methodological tools in this regard?

The recent codification promoted by the United Nations, in relation to the identification of customary rules,¹³ has prompted the author to reflect about such questions.¹⁴ At the end of its work, the International Law Commission (ILC) reached highly practical draft conclusions.¹⁵ Indeed, pointing out that the determination of the existence of a customary rule and of its content would be simultaneous processes,¹⁶ the ILC seemed not to have independently dealt with the content-ascertainment issue of CIL, nor with the similarly interesting topic of its meaning-determination. Namely, whether a particular unwritten rule could be interpreted (even or exclusively?) after its identification. It is also worth noting that the relation between customary rules and rules of interpretation – the latter being usually considered relevant only for written rules – has been scarcely investigated in international legal literature.

In this chapter I shall draw a schematisation of the differences (many) and similarities (very few) between the processes of identification and interpretation of an international rule: in particular CIL.¹⁷ By following a positivist approach – which reflects, at the same time, the reality of the social phenomenon to which international law refers and its historical evolution – I will try to take into account the close connection with the dynamics of international relations, proper to the relationship between the international community and the law which regulates it. This chapter will therefore aim to present international law as it results from the practice of international actors on the one hand and, on the other, as it is interpreted by international jurisdictions, in particular by the ICJ.

My argument is developed in two parts. After providing a plausible definition of interpretation in international law, I will investigate – by taking as main example the Jurisdictional Immunities of the State case – both legal and logical differences between the two distinct moments of identification and interpretation of a customary rule.

2 A Fundamental Preliminary Definition

The interpretation of international law in general¹⁸ poses a multitude of challenges:¹⁹ one of these is that its rules are often extremely indeterminate. In fact, sometimes they are unwritten,²⁰ like CIL. Unwritten rules present, especially in public international law, a peculiar issue of interpretation. There is no text and, despite this, they would appear to be constantly interpreted. In fact, the very fact that the customary rule is not written, makes this rule even more subject to a heterogenesis of meanings. It is therefore very difficult not to ask the fundamental question: is CIL subject to the interpretative rules of international law? And by consequence, in practice, are customary rules interpreted or are they only identified? It should also be noted that interpretation, being a ubiquitous and helpful activity for the intricate nature of the discipline of international law, can potentially produce conflicts between rules too. Yet even if it is taken as a ubiquitous activity, it does not mean that interpretation is a homogeneous and unitary phenomenon. According to the interpretative process, judges interpret the rule which they are empowered to apply, with a view to determining (or creating, according to a Kelsenian account²¹) the normative guideline for the case of which they are seized. This activity consists in an interpretation for meaning-determination purposes, which is surely not an activity reserved only to the judges. In fact, any professional dealing with international law will undertake this operation.²² Nevertheless, it is within the context of adjudication that the interpretative activity is the most visible. Excluding those who in no way allow customary law to be interpreted, I now refer to those who argue that the interpretation of a custom is contextual to its identification. The main point to be made here is that our understanding of interpretation of a customary rule should not be limited to its identification process. This particular distinction between the content-ascertainment process and the scope-determination process of a customary rule is, in my opinion, essential to understand the concept and the practice of interpretation as well as the general concept of law. Mainstream studies of interpretation in international law look almost exclusively at the content-determination of a customary rule. However, what allows a rule to be applied involves an act of interpretation. When applying a custom, the judge, the practitioner, or the academic necessarily try to clarify the meaning of some pre-existing – thus, already identified – customary international rules.²³ Hence, to fully understand the distinction – in my opinion not only terminological – between identification and interpretation that I will try to outline in this chapter, it seems appropriate first to define what is meant by interpretation of a rule: ascertainment of content or determination of meaning? If the scholarly debate does not preliminarily agree on the definition to give to the interpretative activity, it seems useless to carry on.²⁴ This is precisely the point that deserves a preliminary, more careful reflection. If by interpretation we mean determination of content, it seems natural to affirm that the interpretative process of a customary rule is absorbed in its identification process and that, by consequence, it takes place at the same time as the ascertainment of its existence. If, on the contrary, we define interpretation as the operation by which the meaning of a legal provision is reconstructed, in order to understand its scope, it would seem logical to maintain that such activity is carried out at a different time from that of its identification. As a result, if the second definition of interpretation is accepted, it would appear that the answer to this question does not raise too many difficulties and that it is therefore possible to clearly distinguish between the activity of identification and that of interpretation.

3 Identification versus Interpretation

First of all, it is necessary to provide some tools in order to deal with the peculiar distinction between ‘identification’ and ‘interpretation’ of a rule in general, and, in particular, of a customary rule.²⁵ With respect to customary rules, in fact, the confusion between the two concepts is at the root of numerous misunderstandings and essential divergences. As far as treaty law is concerned, interpretation and identification are two, clearly separate, processes. Treaties are generally easy to identify and in most cases, once their identification is completed, it is possible to interpret their content with ease. Instead, when dealing with unwritten rules, specifically with customary rules, this distinction does not seem to be so evident. In this case, the analysis seems to concern two groups of elements: those relevant to the emersion process of the rule (state practice and opinio juris), on one side and the written and/or verbal formulations of the rule (generally retrospective, but sometimes programmatic or even concomitant) defined by a number of actors (judges, diplomatic chancelleries, scholars, etc.), that spare no efforts to express with words the customary rule, on the other.²⁶

Both identification and interpretation processes have been the object of formalisation by international legal scholars. International lawyers have long attempted to balance the uncertainty of the meaning of rules through a definition of the techniques and methods of the interpretative process. The process of such formalisation has not followed the same path for interpretation and identification, the two concepts being substantially distinct. With regard to interpretation, scholars have tried to delineate its criteria, finding a compromise between intentional, purposive and textual methods. On the one hand, the VCLT can be seen as the epitome of this effort to delineate the techniques of interpretation.²⁷ On the other hand, as to identification, recent works of the ILC on ‘identification of customary international law’ can be considered the embodiment of such an attempt to formalise the recognition methods of customary law. The suggested dichotomy implies a practical discrepancy between interpretation and identification,²⁸ each of these processes accomplishing a peculiar operation. The former seeks to explicate the meaning of rules with a view to establishing the standard of conduct, hence, the scope of the rule. The latter intends to determine how a given rule is a part of the international legal order. This means that interpretation is supposed to define meanings and standards of behaviour, while identification is meant to build a double architecture of ascertainment that differentiates law and non-law. Consequently, as far as both customary and treaty rules are concerned, while ‘identification’ seems to be an intellectual phenomenon, ‘interpretation’ appears a purely legal operation. More precisely, the first seems to consist in ‘representing’ a rule, the second in ‘building it’²⁹ or, to put it in another way, to rebuild it on the basis of certain legal methods.³⁰

Nevertheless, it seems likewise appropriate to admit that both processes of interpretation and identification of a customary rule can share some comparable characteristics. Such similar features may justify the fact that – with respect to customary law – they are often confused one for the other. The difficulty in categorising them and, by consequence, in denying the possibility to interpret the ius non scriptum,³¹ is also intensified by the fact that in practice, according to many authors, they may be performed at the same time.³² Nonetheless, by accepting the above-mentioned conceptual dissimilarities between the two operations, it seems difficult to argue that the process of identification of a rule is indistinguishable from the one of its interpretation, even in the case of an unwritten rule.³³ It is true that, in the case of a written rule, the determination of its content is clearer. That is evident. However, it is also true that although a written rule has (apparently) a clear content, this should be interpreted in the subsequent moment of the rule application. And the same operation, in my view, takes place with reference to customary rules too. These, in fact, once identified, have a (more or less) clear content. Afterwards, at the moment of the application to the particular case, this (the content) needs to be interpreted in order to exactly understand the scope of the rule. Consequently, to deny the possibility that such an operation is also applicable to customary rules would be detrimental to the correct and consistent application of the whole international law. This point of view intends, in fact, to assure the maintenance of a reasonable (logical and juridical) flexibility in the application of rules in general. Hence, in the application of customary rules too.

In order to better understand my perspective, I will refer to the ‘dynamic’ of customary rules. Such a ‘dynamic’ is obviously tied to the existence of the rule (formation and identification), but it can also involve the interpretation of the same (i.e., meaning and scope determination aimed at the rule application). In my opinion, those two ‘dynamics’ operate in a totally independent way to one another. In fact, they refer to two distinct operations: one thing is to investigate the dynamic of the existence of a customary rule (identification), and another is to analyse – once dealing with an already consolidated customary rule – the dynamic of its application, hence, its scope (interpretation).³⁴ Although in legal literature it is widely considered that the only logical path to follow is: first, identification (thus, the simultaneous interpretation); second, application of a customary rule, from my point of view, it would seem difficult to deny that identification and interpretation take place in two distinct moment of the ‘dynamics’ of a customary rule. As a result, after the customary rule formation, by means of both a consistent and general international practice by states and a subjective acceptance of the practice as binding by the international community, once the rule is identified (i.e., its existence and its content are ascertained) – through an evaluation of its two constitutive elements – this can be applied to a particular case only after a preliminary interpretative operation. An important premise must be made to fully understand this point of view: by interpreting a customary rule I explicitly refer to an already identified rule, properly understood (i.e., unwritten) and not to its constitutive elements, nor to its written reformulation.

Its existence being totally uncontested, I will take as a main example the customary rule of state immunity in order to investigate whether and to what extent this distinction occurred in practice by exploring the thin border between rule modification (related to the dynamic of its existence) and rule interpretation (related to the dynamic of its application). The practical relevance of this matter has been particularly evident with regard to the Jurisdictional Immunities of the State case.³⁵ The object of the litigation dealt with ‘the scope and extent’ of the customary rule, whose existence was recognised by Italy as well as by Germany, regarding foreign states’ immunity from civil jurisdiction. Indeed, both parties admitted ‘that States are generally entitled to immunity in respect of acta jure imperii’,³⁶ but they disagreed on the scope of such a norm. Italy invoked the application of the so-called tort exception – that is, the absence of immunity in case of actions having caused death, personal injury or damages in the territory of the host state – also in relation to acta jure imperii. On the contrary, Germany – by giving a different interpretation of such rule, that is, by considering that this particular case did not fall within the rule’s scope – denied the application of such an exception of the customary international rule. The ICJ itself stated that the parties’ agreement on the existence and/or the content of a rule would not, after establishing the existence of this international custom (i.e., identifying it), exempt it from making its own evaluation on the scope and extension of state immunity (i.e., to make its own interpretation).³⁷ Hence, in order to decide this case, did the court interpret or identify the customary rule under consideration? More generally, when a judge deals with a modification of a customary rule, does he identify the rule or does he interpret it? Both stances could be convincingly supported.³⁸ Nevertheless, in the Jurisdictional Immunities of the State case, by ruling upon the so-called tort exception, the ICJ seemed to confine its assignment to the identification of the existence of an exception from the general rule and, thus, stated the inexistence of such exception. However, the Court could have operated in a different manner. In fact, as asserted in the judgment, the ICJ task could also have been understood as an interpretation of the customary rule under consideration.³⁹ Without searching for the two constitutive elements of the customary rule on state immunity, aimed at confirming or not the existence of the tort exception, the ICJ could have interpreted the customary rule on state immunity – already identified and uncontested by the parties – in order to establish the scope of the same: that is, whether and to what extent it could have been applied to this specific case. As mentioned above, since examining state practice and opinio juris reveals the existence and the content of the rule and does not explain whether this rule is applicable or not to the particular case, in order to apply a rule to a specific case, it seems crucial to investigate the scope and the extent of the same (to interpret it), and not anymore its existence (to identify it).⁴⁰ In fact, any operation by which a rule is applied requires a prior interpretative activity. The application to a particular case of a general and abstract rule, logically implies the determination of its meaning too. Without such operation, it would not be even possible to understand all the legal consequences resulting in that particular case. In other words, the problem of legal interpretation cannot be circumvented since it is always indispensable (and propaedeutic) for the rule application. Therefore, by taking the Jurisdictional Immunities of the State case as main example, my purpose is to highlight how in practice identification and interpretation processes can both be easily performable and, by consequence, often confused. This case is particularly relevant for my argument since here it is evident how thin the line between the two operations can be, one related to the ‘dynamic’ of a rule existence, and the other related to the ‘dynamic’ of a rule application.

The logical correlation between the two moments of interpretation and application, with respect to customary rules too, can also be grasped by observing the conduct of the actors obliged to comply with the customary rule provision: the states. The customary rule, already identified, conditions their behaviour through an intellectual operation (interpretation) intended to clarify the correct meaning in the specific case. This means that customary rules would require the state whether it is or not in the situation (the particular case) provided for by the rule itself.⁴¹ This intellectual operation – aimed at verifying whether in a particular case the conditions provided by the customary rule are satisfied – can, indeed, determine state observance of customary provisions. It can also lead to a conflict of evaluations between two or more states,⁴² to a rule infringement,⁴³ possibly also to an impartial, third-party evaluation.⁴⁴ The spontaneous observance, the impartial evaluation as well as the enforcement of a customary rule, all belong to the application of CIL. The practical implication is the safeguard of a reasonable flexibility in the process of customary rule application. In fact, excluding any interpretative activity with reference to custom would artificially restrict the interpreter’s necessary task.⁴⁵ Hence, in the application of a well-established custom, the legal operator must take into account the content of the rule in order to understand its meaning (interpretation).⁴⁶ This, of course, without affecting its content (established at the time of identification) by modifying it.

At the end of this short analysis, it should also be emphasised that this practical and theoretical distinction raises the question of the admissibility of analogy⁴⁷ or restrictive⁴⁸ interpretation of customary rules too. Indeed, one should not wonder what and how the international community members would have decided in a specific matter by going to investigate the constitutive elements of a customary rule, such as state judgments, domestic laws or diplomatic notes. In the search for the meaning of the prescription of the customary rule, it would not seem to be relevant, nor it would seem to lead to any reliable result in the interpretation of the rule itself. On the contrary, this is an evaluation on whether the content of the customary rule (established through the identification process) can be applied to the new particular case too, for example, through its analogy with the hypotheses regulated by the customary rules in question. This will widen, narrow down or otherwise correct the scope of the rule already formed for the generality of the affiliates.

4 Concluding Observations

To differentiate the two operations of identification and interpretation is essential to correctly determine the scope of a rule. This is true for a written rule and, in my view, is even more true for an unwritten rule. For a written rule it can be considered that, exactly because it is written, it is relatively simple to separate its content-ascertainment moment from that of its meaning and scope-determination. By contrast, for an unwritten rule – and in particular, for a customary rule – this may not be evident. As is well known, in a legal system as little organised as the international one, given the importance of customary rules as well as the lack of specific bodies for the formation and manifestation of collective will – and therefore for the formation and manifestation of law – the need to distinguish these two operations seems even more important.

Hence, this distinction is evident for both categories of rules (treaty and customary), being, even if at times confusing, two operations logically and chronologically clearly divergent. As I tried to highlight with respect to the Jurisdictional Immunities of the State case, the interpretative activity takes place at a time subsequent to that of the identification of the content. That is, when the rule is applied to the particular case. In fact, for customary rules, as well as for treaty rules, the search for the scope is an indispensable operation, accomplished after the identification and preliminary to that of the application of the rule to the particular case. In other words, it is the application of the rule to a particular case that, indeed, forces the legal practitioner to interpret its content. The interpretative activity, thus understood, is therefore inherent to the moment of the rule application to the particular case. If this were not the case, there would never be a problem of interpretation, neither with regard to treaty rules nor with regard to customary rules. This would be the same as arguing that any content of a law is so clear and so specific that it is able to precisely reproduce every case that will occur in the future.⁴⁹ A rule will never be so clear as to be directly applied to a particular case without further logical steps. Furthermore, to support the assimilation of the interpretative process of a customary rule to its identifying process would lead to the paradoxical scenario in which a customary rule would require to be identified each and every time it needs to be applied.⁵⁰ This begs the question of whether a customary rule can be interpreted.⁵¹ Consequently, according to such an approach, whenever a dispute concerning a customary rule is brought before a judge, he should constantly – by making reference to both state practice and opinio juris – take into account the existence, development and manifestation of customary rules. According to such a perspective, a judge should identify the customary rule each time he applies it to the particular case. In a similar conception there would exist an infinity of customary rules, all different from each other but each of them extremely specific and very particular, being applicable to only one specific case: the one in which it was identified. This would defeat the very function of having a rule and it would no longer be useful to have a system composed of general and abstract rules. It seems extremely difficult to argue that a previously established customary rule could be applied to new cases falling within its scope, regardless of the general principles of interpretation. Such a theoretical approach would seem to conform to the logical requirements of the whole dynamic of customary rule.

However, several doubts remain. For example, how did the international actors deal with the issue of the interpretation of CIL? Has it been differently addressed in the various cases? According to the ICJ, what would it mean to interpret a customary rule? Has the Court provided the theoretical-methodological tools needed to interpret a customary rule? And to distinguish the two logical operations of interpretation and identification? What are the principles established in this regard by the ICJ? As pointed out before, in the Jurisdictional Immunities of the State case the Court could have interpreted the customary rule on state immunity? Or it could exclusively have identified it? Both stances could be convincingly supported. Further study and analysis of the topic might try to answer some of these questions.

20 ‘And in the Darkness Bind Them’ Hand-Waving, Bootstrapping, and the Interpretation of Customary International Law after Chagos^*

John R. Morss and Emily Forbes

1 Introduction

The essential role of interpretation in relation to national statutes has, especially in common law jurisdictions, given rise to a complex apparatus of guidelines, axioms and indeed further statutes (‘Interpretation Acts’). Somewhat by analogy or by extension, the interpretation of international conventions (treaties) has long been recognised as itself both complex and immanent to the process of application of treaties, not least when the effect of a treaty is in dispute. It is true that despite well over a century of international jurisprudence on the interpretation of treaties, including the development of the Vienna Convention on the Law of Treaties (VCLT) itself, complexities and unresolved difficulties remain in that sphere.¹ However the processes of disciplined interpretation of written statements of law, whether municipal (national) or international, and the problems that arise therefrom, are at least familiar.

The role of interpretation in the second of the sources of public international law – ‘international custom, as evidence of a general practice accepted as law’ – is much less familiar. Scholarship in this area is just beginning.² It has been proposed that interpretation in this context is of most significance in the application of customary international law (CIL) to new situations, rather than in its initial discernment as such. Without disputing the relative importance of interpretation in the application of CIL, vis-à-vis interpretation in other aspects of the technique of CIL, it may still prove of value to widen the scope of enquiry in that respect. There is thus room for the view that interpretation is intrinsic to the definition, articulation and implementation of CIL just as it is for the law of treaties, for ‘general principles’, for teachings of the publicists or for previous decisions of international tribunals; and indeed for considerations ex aequo et bono.³ Interpretation may of course be applicable in somewhat different ways to these various genres of ‘source’ (by which term they are collectively and colloquially known) in international law.⁴ Put this way, interpretation is ubiquitous; and while this is itself significant, care must also be taken to distinguish modes or genres of interpretation. To draw attention to a role for interpretation in relation to CIL is thus only a very preliminary step, as is of course recognised by scholars.⁵

In this chapter we argue that some of the most important aspects of the role of interpretation in the context of CIL can be expressed in the following way: namely to claim at least for the sake of argument that the most characteristic phrase concerning CIL in the discourse of public international law is the phrase ‘this may represent customary international law’.

It is the word ‘may’ on which we lay emphasis. This hypothetical or tentative assertion of the existence and validity of a particular CIL, is intrinsically interpretive. What we will argue is that this hypothetical or tentative gesture, thought of as a species of interpretation, takes us to the heart of CIL. In other words, what is central to any statement about CIL is the attribution of the possibility of customary justification for some conduct or prohibition, irrespective of whether this has ever been or will ever be, tested.⁶ Despite the gravity of the rights and obligations connected with the norm thus speculated upon, there is a certain archness to the trope. The kind of uncertainty thereby conjured is a kind of uncertainty quite different to what one finds with other genres of source. A wished-for consensus of the most qualified publicists or of judicial decisions, on a particular point, might expose uncertainty of a somewhat humble variety and one that is in essence empirical. A survey of relevantly common municipal regulation, entered into in the spirit of the Barcelona Traction dispute, would be uncertain in somewhat the same manner, as of research. To refer in an open-ended manner to the possibility of there existing a presently unknown, written agreement between sovereigns would seem absurdly speculative yet this is only the case because treaty making has become a public affair. In principle all such hypothesised sources might be enquired into with a reasonable expectation of establishing either their existence or non-existence. Often CIL will be sought out in a similarly empirical manner yet the gesture seems intrinsic to this source of norms in a way unmatched by the other sources. This gesture, almost a gesture to a higher realm of the transcendent, might be said to locate CIL in some grey zone between lex lata and lex ferenda: a zone we might call lex hypothetica.⁷

Approached in this somewhat sceptical manner, the essence of CIL seems to be ‘bindingness’ (legal obligation) combined with opacity – what might be called ‘blindingness’ – because the transparency that comes with treaties (albeit, only since the mid-twentieth century) is necessarily absent. Accountability might be said to be dramatically lacking.⁸ The combination of these two factors might be referred to by borrowing, with apologies and with some poetic license, the words of J. R. R. Tolkien: ‘And in the darkness bind them.’ Only retrospectively and vicariously, that is to say in the decisions or advisory opinions of tribunals, is CIL endowed with a measure of transparency. This takes place by means of the ‘translation’ of the CIL into written form. Thus CIL, which is defined as unwritten, paradoxically only has normative force when it is written. Up until that point at least it is something of a will o’ the wisp. The process seems a little like the recognition of states under the declarative mode, or like the announced discovery of a common law principle or a maxim of equity: more alchemy than chemistry or, to adjust the metaphor, more priestly than Priestley. A qualitative change takes place, a transubstantiation or saltation.⁹ Correspondingly, the circularity in argument or ‘bootstrapping’ aspect seems problematic as one intangible step leads on to another. Of course the enigmatic if not paradoxical character of CIL is widely recognised, for example in the arcane form of the persistent objector to an emerging customary norm.¹⁰ Here a sovereign is held retrospectively to have been sufficiently cognizant of an emerging customary norm as between relevant sovereigns, of which that sovereign is one, that the sovereign’s historical protests constitute a kind of negative prescription by means of which his or her putative obligations are nullified. Given the relationships between prescription in the international law of territory and the common law principle of adverse possession, the persistent objector would seem to be claiming something like an ‘adverse immunity’.¹¹

Further below, we will frame and motivate our comments on the role of interpretation in CIL by means of an enquiry into the role played by CIL in the Chagos Advisory Opinion.¹² On 25 February 2019 the International Court of Justice (ICJ) handed down its Advisory Opinion in relation to the continuing administration by the UK of the Chagos Archipelago in the Indian Ocean. In exerting control over the territory, which prior to the independence of Mauritius was part of that non-self-governing entity, the UK had forcibly transferred its population. A population of Chagossians had been dispossessed by the British government and its military in the 1960s, and relocated to mainland Mauritius, giving rise to various deleterious consequences for that population both material and affective.¹³ The UK had undertaken to ‘return’ the Chagos Archipelago to Mauritius if and when it was no longer needed for defence purposes.¹⁴ Subsequent General Assembly resolutions consistently condemned the continuing administration of the Chagos Archipelago by the UK.¹⁵ While at pains to avoid the appearance of treating the question before it as a dispute between two parties, namely Mauritius and the UK, the majority clearly endorsed the postcolonialist argument that was proposed on behalf of Mauritius and also on behalf of many other states contributing to the proceedings.¹⁶ Thus the process of decolonisation was found not to have been ‘lawfully completed’ in 1968 when Mauritius acceded to independence.¹⁷ Merits of arguments submitted to the court will not be rehearsed or evaluated.¹⁸ This Advisory Opinion has already generated considerable commentary and debate in relation to the continuing administration by the UK of the Chagos Archipelago.¹⁹

Customary international law plays an important role in the Chagos opinion. Has Chagos advanced our understanding of CIL? Has it ‘developed’ CIL? Has it clarified the relationships between CIL and such denizens of international law as the peremptory norm, obligations erga omnes, the technique of uti possidetis juris, the general principle of international law or the inexorable (e.g. anticolonial) trend of modern history recognised as a matter of fact by the court? Has it ‘stabilise[d] history’ as, it has been suggested, was the aspiration behind the UN Charter itself?²⁰ In other words what, against the background of our observations above, does Chagos tell us about the role of interpretation in CIL? Before engaging with this question however, some wider questions need to be addressed.

2 On Structure, Depth and Explanation in International Legal Discourse

There is no a priori reason to treat interpretation as so intimately connected to written text that norms based on unwritten conduct fall outside of its scope. Certainly the academic disciplines of hermeneutics and other interpretive techniques evolved from practices of the glossing of written texts, namely texts of Holy Writ.²¹ That history clings to the techniques in ways that we sometimes notice and sometimes do not notice, just as international law as a whole is contaminated but not entirely determined by the colonial oppression inflicted on the globe by the hegemonic princes of past centuries.²² And just as hermeneutics has no intrinsic, constrained identity with particular texts in that sense, being relevant to a variety of profane as well as religious writings (and to non-Christian as well as Christian religious materials, it goes without saying), interpretation has no intrinsic constrained identity with the printed word in any language. Not to recognise this would be among other things to entirely misunderstand Derrida’s dictum (or was it a ratio?) to the effect that ‘there is nothing outside the text’. What ‘text’ meant for Derrida was not restricted to cold print on the page, as in the myth of ‘black letter law’ so derided by critical writers in legal theory; and without neglecting the significance of the challenge made by Derrida to hermeneutics, to phenomenology, and to other brands of interpretive discipline, his point ranged more widely. To approach this point from another direction, the science or discipline of semiotics is patently concerned with meaning-making well beyond the written (or indeed the spoken) word. Famously Roland Barthes analysed dress codes.²³ Semiotics in that respect is cognate with structuralist anthropology from Levi-Strauss to Margaret Strathern and hence with structuralism in general.²⁴

2.1 Bootstrapping and Hand-Waving

Interpretation thus includes any appeal to particular frameworks of meaning beyond the specific text, conduct or pattern that is observed. Across the many forms and disciplines of interpretation, it can be generally said that coherence is a significant virtue. By the application of meaning systems, propositions gain a kind of validity from that coherence independently of other forms of legitimation. Coherence becomes a kind of authority.²⁵ However, to develop an argument based primarily on coherence might in some circumstances be unkindly referred to as ‘bootstrapping’. To attempt to lift oneself up by one’s own bootstraps is a telling metaphor. It is of course the reflexivity that is the problem. As Archimedes noticed, an external point of leverage is called for by the aid of which the lifting becomes possible. Again, if one’s bootstraps are sufficiently robust and the boots themselves of the correct size, an external agent standing on terra firma may well be able to lift one up in such a manner, however undignified that would be.

Coherence is not to be lightly dismissed. After all, any use of logic or of mathematics involves reliance on a system of coherence, and is usually not considered the worse for that. But neither logic nor mathematics is self-executing. In a context relevant to the discussion below, it has been claimed that the principle of uti possedetis according to which administrative boundaries of the colonial power are retained by new neighbours after independence, is ‘logically connected with the phenomenon of the obtaining of independence, wherever it occurs’.²⁶ Just as the application of mathematics or logic can lead the scholar astray at times, without it always being obvious exactly when the wrong step was taken or the wrong connection made, so the cumulative construction of claims about the conduct of international entities as generating normativity – the bread and butter of CIL as usually understood – may risk the inevitable fate of a house of cards. It is merely high school level physics to learn how error estimates accumulate in the laboratory, such that a modest error range on each of two or three independent parameters (temperature, weight and so forth) may accumulate to a hefty ‘known unknown’ when those readings are combined in sequence. If the discourse of CIL involves a sequence of ‘ifs’ and ‘maybes’, as is surely so often the case, then it may be a tottering tower of claims that is constructed. If apples and oranges are on occasion pressed into service, so to speak – so that what is being built comprises somewhat different elements – the fragility is again manifest. A telling analogy is the conceit of the chain of counterfactuals across history, so that the retrospective adjustment of one event suggests consequences which themselves serve to generate further and equally fictitious consequences. Historian Niall Ferguson has attempted such a conceit.²⁷ ‘Hand-waving’ refers to the argumentative practice of deliberately indicating that one is evading difficult questions or traversing fragile steps of a thesis. In combination these discursive gestures or techniques have a tendency to generate conclusions that leap well ahead of any substantive basis in interconnected claims either empirical, conceptual or legal. To change the metaphor yet again, the conclusions thus arrived at may appear somewhat like the rabbit produced from the hat of the conjuror. If that comparison is in any way apt, then there would be cause for concern.

Before exploring this idea further, other distinctions between CIL and other genres of international norm should be considered. The unwritten aspect of CIL is often considered its essential characteristic. But it is only in contrast to international treaties or ‘conventions’ that the unwritten aspect of CIL seems to be noteworthy in the context of the sources of international law. Treaties are written documents, signed by representatives of polities. ‘General principles’ of international law are not, as such, reduced to writing. The most canonical examples of such principles, at least in terms of the jurisprudence of the ICJ, consist of the appeal to common practices among relevant States vis-à-vis their national legal frameworks on the status of corporate entities. The fact that these national laws (in Belgium, Spain and so on) are themselves, and in a variety of languages, written law in a very traditional sense, does not affect the conclusion that the general principle abstracted from them in the case of the Barcelona Traction, is not itself written. The status of so-called principles of international law in a wider sense (beyond the definition in Article 38(1)(c)) is problematical in any event. Text-book writers have been trying to tie down such generic principles for several centuries.

One extreme example – extreme in its scale and self-confidence, and perhaps in its practical import – is the encyclopaedic package of claims made on behalf of the International Committee of the Red Cross (ICRC) concerning regulation of armed conflict in the form of international humanitarian law.²⁸ That project has given rise to extensive online resources that attempt to maintain and update a body of knowledge on both conduct-based (‘objective’) and opinio-based (‘subjective’) grounds for the identification of CIL. While the specialised focus of those claims takes the ICRC exercise outside the substantive scope of this chapter, it might be borne in mind as a marker for the potential scope of an industry of speculation relating to CIL.

Optimism is one might say ‘in the DNA’ of the ICRC, given its Sisyphus-like exertions in the face of human conflict. What this chapter is concerned with is a more general process. It is the Chagos Advisory Opinion as paradigm for the discerning of ‘general’ CIL, that is, rules of CIL valid for all states as such.²⁹ A familiar example of such a general or pan-state CIL would be immunity from prosecution of heads of state.³⁰ Such rules must be distinguished from rules jus cogens that are asserted, on grounds other than customary observance, to have peremptory force as such. At this point it could be observed that one of the points of contact between the norm jus cogens (the peremptory norm) and CIL is precisely in this modality of the hopeful hypothetical.³¹ The substance of posited norms jus cogens is typically weighty in an ethical sense, as compared to the typical CIL on boundaries or access to fish, but while they diverge in various other ways these two challenging forms of international norm do seem to share this gesture; and of course, self-determination as a putative CIL does indeed involve weighty ethical issues. What will be suggested here is that the step by which a general rule of CIL is identified in the Chagos Advisory Opinion comprises an extrapolation from at best a combination of majoritarian avowal on behalf of sovereigns, and the sovereign ratification of international instruments. As we shall see, the gap between ‘ought’ and ‘is’ is bridged by a mixture of hand-waving and bootstrapping even if the expressed view of the majority of the bench is much more restrained in this regard than some of the separate opinions.

2.2 Alternative Reference Classes for Customary International Law

The reference class of the term CIL can be defined in a wide variety of ways. One alternative is of extreme breadth but little precision. As James Crawford indicates, in the tradition of Ian Brownlie and others, the practice of states understood most generally can be what is indicated.³² Without neglecting the essential distinction between CIL and mere comity or courtesy, a distinction that goes back at least to the beginning of the nineteenth century,³³ it is still the case that customary forms of international law represent a kind of oceanic backdrop for much that is more narrow, more technical or more specified in terms of the conduct and the expectations of parties. Thus in what might be called the Brownlie-Crawford approach, attention is paid to what are taken to be deeper and slower-moving features of the international legal landscape, generative of ‘principles’. This is an extensive rather than an intensive approach to what might loosely be termed custom-based conduct. It is exemplified by the observation that ‘the state is itself a customary law phenomenon’.³⁴ This recognition does not claim that all such conduct is CIL in a substantive, dispute-resolving sense. It is in some ways a Kelsenian argument, pointing to an infrastructure of norms that might be said to be logically necessary in order for substantive CIL to exist and function. That is to say, these oceanic customs are understood as disconnected from the CIL that according to the jurisprudence of the ICJ, is provided in Article 38 (1)(b). Nor is it claimed in any systematic way that such background customs constitute necessary appurtenances of states. In other words there is something of a policy-oriented positivism in operation here: ‘this is what states do’.³⁵

As suggested in the preceding, a reference class for CIL may be the class of incidents of statehood. This ‘ontological’ variety of the sense of custom at the international level, tied to the definition of statehood in recursive ways, has an extensive lineage even if its inadequacies are patent. Clearly the reference class of customs that apply to all states, whether or not that relationship is thought of as involving inherence, is an important variety of reference class. So as well as the background or process-oriented framework, we have the statehood-intrinsic (or ontological) and the statehood-generic as additional, distinct but overlapping reference classes. Specially affected states as constituting the relevant set might be thought of as yet another option; and the admittedly old-fashioned view that hegemonic or elite states play a special role in generating and legitimising CIL, might be further added.³⁶ Finally, reference to CIL may narrowly refer to a CIL that has been found by an appropriate tribunal – and this really means the ICJ – to meet its technical and forensic requirements, as famously laid out for example in the North Sea Continental Shelf Cases. That is to say, CIL may refer specifically and narrowly to legal obligations discerned by the ICJ (or similar) in dispute resolution or in the context of advisory opinions. This narrowly defined category would correspond to an approach to the notion of peremptory norms on the basis of accepting only those norms that the ICJ has in fact identified as having jus cogens status.³⁷

3 Interpretation in Chagos: A Customary International Law of Self-Determination?

So what of the Chagos Advisory Opinion? In the opinion crafted by the majority, considerable reliance is placed on CIL.³⁸ It could be said to be the most significant basis for the outcome, always granted the overriding role of the UN Charter. The latter of course refers to the ‘self-determination of peoples’ but without much clarity.³⁹ In the Chagos Advisory Opinion, the right to self-determination is defined as CIL and moreover, one that is ‘binding on all States’.⁴⁰ Respect for such a right ‘is an obligation erga omnes’.⁴¹ On that basis, the majority finds that the territorial integrity of a former non-self-governing territory is violated if part of that territory is excised from the territory of the newly independent entity.⁴²

It should be noted that the majority is at pains to emphasise its restrained approach in contrast to the enthusiastic and even crusading, ex cathedra values-based approach advocated by the Separate Opinion of Judge Cançado Trindade. The majority emphasises its continuity of approach with previous findings and opinions of the ICJ. If anything the majority presents itself, if only strategically, as conservative and cautious. Thus in its Chagos Advisory Opinion the majority of the ICJ bench carried out an exercise in the interpretation of CIL in the context of the self-determination of peoples and of the administration of non-self-governing territories in terms of Chapter XI of the UN Charter. However politically welcome the outcome in respect of a snub for a former colonial power, and ipso facto a snub for that power’s powerful client, the United States in this case (since the United States had leased the Diego Garcia location as a naval base), the reasoning of the court in terms of its reliance on CIL stands in need of interrogation.⁴³

In the Chagos Advisory Opinion it appears that the court provided an object lesson in the ascertainment and application of CIL in twenty-first century international law. While insisting that the ascertainment of the content of the putative rule of CIL (the ‘what’) as well as the ascertainment of the chronology of its coming into effect (the ‘when’) are exercises limited to the specific customary rule in question, more general methodological assertions can certainly be identified. The unpacking of the putative customary rule and the closely related investigation of its provenance are anchored to previous findings of the court in relation to quite different kinds of CIL such as the allocation of access to offshore resources among adjacent coastal states. The parameters of difference are such matters as human rights norms versus access to resources norms; but also, pan-state norms versus ‘regional’ or ‘specially affected state’ norms. Although the court’s primary steps in enquiring into a putatively salient customary rule in this context may be said to be formal ones, to the extent that finding the relevant and adequate combination of conduct and opinio juris is a formal exercise, the investigation into the content and thus the consequences or effects of the customary rule is unambiguously a matter of interpretation. (In any event form and content are intermingled, if not circular, in the context of CIL.) The extent to which ‘reading up’ of the CIL takes place – the widening and the increased weighting of the obligations which are said to flow from it – is perhaps much greater than would usually be the case, for example with another North Sea Continental Shelf situation. An advisory opinion is indeed more appropriate to such an expansive exercise than a dispute between states. The CIL that is examined in the Chagos Advisory Opinion, with its UN Charter connections, its General Assembly contributions and its world-historical resonances, might be said with some justification to be sui generis. But the building of the edifice of the CIL of the decolonising of non-self-governing territories is still remarkable. The method of interpretation employed by the court enables it to successively unpack this CIL into what one might describe as an articulated and systematic project management scheme governing the decolonising process as generic historical transition. It is such a process, read by the ICJ majority into the combined effect of the UN Charter and Resolutions of the General Assembly, that according to the bench was applicable to the case of the Chagos Archipelago but manifestly dishonoured by successive UK governments.

According to the majority, a right to self-determination based on CIL may be discerned, crucially one that was in existence before the time at which the UK government purported to excise the Chagos Archipelago.⁴⁴ The court saw it as its task to ascertain ‘when the [right to self-determination] crystallised as a customary rule binding on all States’.⁴⁵ In this light, General Assembly Resolution 1514 (XV) of 1960, the Declaration on the Granting of Independence to Colonial Countries and Peoples, was ‘a defining moment in the consolidation of State practice on decolonization’; it ‘clarifies the content and scope of the right to self-determination’.⁴⁶ It has the character of a declaration of a right to self-determination as ‘a customary norm’.⁴⁷ This right to self-determination is a ‘basic principle of international law’ and ‘its normative character under customary international law’ was also confirmed by General Assembly Resolution 2625 (XXV) of 1970, to which was annexed the Declaration on Principles of International Law concerning Friendly Relations and Co-operation among States in accordance with the Charter of the United Nations.⁴⁸

Moreover, a right to the territorial integrity of a non-self-governing territory is itself a customary law.⁴⁹ In this context the court observed, in a methodological vein, that ‘State practice and opinio juris … are consolidated and confirmed gradually over time’.⁵⁰ Thus it is proper for the court to consider ‘the evolution of the law of self-determination’ since Resolution 1514.⁵¹ But ‘confirmed gradually’ seems odd, as does the language of ‘evolution’. How can ‘confirming’ be gradual? Perhaps it might be said that, as observed by Criddle and Fox-Decent, evidence going to the identification of a rule of CIL may be said to accumulate. Thus in the context of a posited rule of CIL requiring state cooperation over human rights breaches, ‘[a]s evidence of state practice and opinio juris continues to accumulate in the future [the existence of such a rule of customary international law] may eventually become the prevailing view’.⁵² Now despite the attractiveness of this innocent-looking proposal, complexities still arise in this formulation. As discussed above, a postulated customary norm is not amenable to empirical investigation and the accumulation of data in the way that other forms of source may perhaps be. Common strategies in municipal law across the globe might well accumulate in a factual sense, thus building up the argument for a ‘general principle’ à la Barcelona Traction should a suitable international dispute arise. Despite their subsidiary status, judicial decisions or the writings of the publicists might accumulate across time in relevantly patterned ways, to which international tribunals may well pay attention. Suitably trained agents might keep track of such data. Nor would processes of interpretation be entirely absent from the epistemological projects involved, for the categorisation of a second or subsequent statute, judicial finding or scholarly conclusion as adding weight to a first rather than starting its own pile, must always involve interpretation.

Yet, CIL just does not seem amenable to such scientistic accretion of data. Especially in the domain of the opinio, where the data would have to take the form of evidence of obligations understood by sovereigns as binding, the customary norm must first be postulated and in effect promulgated in order for the data to be defined. There is a circularity here which the discerning of other forms of international norm can evade. It is true of course that the doctrine of ‘intertemporal law’ requires retrospective assessments of international norms, that is to say ascertainments of applicable norms from a previous era.⁵³ Various sources of international law and various forms of evidence for them might be investigated in that historical mode. Quite deliberately, and in effect as a ‘legal fiction’, the bench transports itself as in a time machine to that past era in a quasi-archaeological investigation. The ‘synchronic’ findings, such as the conclusions on legitimate modes of acquisition of territory in the late nineteenth century, may subsequently be drawn on by scholars interested in defining ‘diachronic’ trends across historical time in relation to such norms. But that would be an entirely separate and so to speak parasitic exercise. The application of intertemporal law does not yield knowledge in the diachronic domain. The time machine travels strictly between ‘then’ and ‘now’, it does not traverse the times before or between. Nor does it generate comparisons even between ‘then’ and ‘now’; the ‘now’ is no more than a launching pad for the shuttle which returns to base after its sample of the core. This methodology for dispute resolution does not seek to trace longitudinal patterns, developments or ‘evolution’. It does of course involve a leap of the institutional imagination but that leap is strictly constrained; it is not a leap of faith.

The term ‘evolution’ therefore does not assist. The term is a flexible one, as it has been across many disciplines and several centuries.⁵⁴ But it certainly connotes gradual improvement in a manner that is in some sense natural and certainly not the direct result of human agency, even if the Darwinian model of natural selection is closely modelled on the systematic interventions of the breeder of domesticated animals, that persistent objector to the customary reign of Mother Nature. In an otherwise carefully crafted opinion, the term may be readily excused. It is perhaps a harmless nod to the grand historical narrative of the postcolonial. Yet the implication that the customary form of international legal norm is in some sense a natural emanation, deserves a little more investigation. It is of course straightforward to connect such an appeal to the natural law tendencies of some of the world’s most influential jurists. A progressivist and even triumphalist tone is not difficult to discern. But this attitude would seem to have a particular connection with CIL if that variety of norm is thought of as the expression of an organic and inarticulate global conscience of mankind, growing or unfolding, slowly yet inexorably, across time.⁵⁵ And of course such treatment would be interpretation, indeed.

Finally, the implications of the Chagos Advisory Opinion for other putative customary norms should be briefly discussed. The principle of uti possidetis juris in international dispute resolution has famously been applied in postcolonial Africa in treating as default international boundaries between newly independent states, the administrative boundaries drawn up by former colonial sovereigns. As with any reference to a ‘principle’ of international law, the question must always be put as to what kind of source such a posited norm may be. The claim that uti possidetis has customary status is questionable.⁵⁶ It seems to exist in that penumbra of the quasi-customary along with procedural norms of wider significance such as pacta sunt servanda. In any event to the extent uti possidetis represents the dead hand of colonialism, the Chagos Advisory Opinion represents if anything the revenge of the principle. Here, the colonial boundary manifested by the inclusion of the Chagos Archipelago into a larger Mauritius entity by France, is now relied upon to the detriment of the residual administrative power (the UK), so that it is hoist by its own imperialist petard.⁵⁷ Of course any reference to uti possidetis raises questions of self-determination.⁵⁸ The formula is in effect an alternative to self-determination and consigns the latter to the ‘too hard’ basket, in favour of ‘nation building’.⁵⁹ To the extent self-determination is coterminous with peoplehood, peoplehood is recognised as flowing across national borders so that pluri-peoplehood within one territory is implicit, typically in the form of one or more minority populations. If two or more peoples are clearly identified within one state territory, then territorial integrity might become a burden rather than a virtue from the point of view of the self-determination of ‘peoples’.⁶⁰ Thus the key finding in the Chagos Advisory Opinion that territorial integrity of a non-self-governing territory is the essence of the CIL of self-determination in decolonisation, serves to undermine self-determination in favour of territorial integrity.⁶¹ There still does not seem to be a substantive contribution from CIL to the vital question of ‘what is a people?’⁶²

In a methodological sense, the account now provided by the ICJ of the international legal norms governing self-determination constitutes primarily an act of interpretation of CIL. Disappointing or not in its achievements in that regard, it is a reminder of the significance of such interpretive processes. Customary norms of international law weigh heavy on minority populations and there is no route to resolving such injustice save through interpretation of those norms in a variety of senses of ‘interpretation’. Paradoxically again, it may be that interpretation in CIL turns out to be even more important than it is in its familiar ‘comfort zone’ of treaties.

4 Conclusion

Sceptical remarks have been made above concerning the rhetorical devices employed in the discourse of CIL. Especially in the writings of commentators, one of the myriad ways in which interpretation is thus involved is in the ubiquity of the proposal that such-and-such a conduct or prohibition ‘may be’ CIL or ‘may be emerging’ as such. This mode of speculation has been said above to be of the essence of what CIL means in the public international law of the present era. Opacity and regulation sit uncomfortably together either at the municipal or the international level. At the municipal level democratic arrangements, however fragile and imperfect, play an important role in battling the forces of obscurantism. At the international level the barriers to transparency must be dismantled by eminent jurists, assisted by commentators. To express it generously, the goal of hermeneutics is interpretation in the interest of enlightenment. That goal is an honourable one and therefore the systematic investigation of the role played by interpretation in the theory and practice of CIL is essential. International law like all law in the real world is made and remade by humans, albeit in complex ways; and binding in darkness belongs only in fantasy fiction.