1 Introduction
The TRICI-Law project observes that ‘in the study of customary international law (CIL) there is a critical gap in understanding how CIL can be applied in individual cases once it has been formed’. The project then sets for itself the goal to uncover rules of interpretation of CIL. In the words of the project, if such rules were to exist, CIL need not be induced (ascertained) each and every time, by reference to state practice and opinio juris or asserted by judges.
This chapter attempts to narrow the gap in understanding how CIL is applied and interpreted by domestic courts. Domestic courts are important agents of international legal development,Footnote 1 and they contribute to the entrenchment of the rule of international law, including CIL.Footnote 2 Accordingly, a study of the interpretation of CIL cannot do without an analysis of domestic court practices.
The contribution opens with a critical reflection on the proposed doctrinal shift from mere CIL ascertainment to interpretation of more or less stabilised CIL norms (Section 2). As domestic courts tend to apply pre-existing CIL rather than ascertain CIL de novo,Footnote 3 the author sees a window of opportunity for CIL interpretation. He then goes on to ascertain whether domestic courts also use this window in practice. He does so by analysing a large data set of domestic court decisions (Section 3). The empirical analysis yields a number of ‘true positives’ which suggest that, in admittedly rare cases, domestic courts genuinely interpret relatively stable, pre-existing CIL norms, in particular in the area of international immunities. These courts appear to use methods of interpretation that reflect those used for treaty interpretation, notably systemic interpretation and interpretation taking into account subsequent practice.
2 From CIL Ascertainment to Interpretation
The quest for rules governing the interpretation of norms of international law other than treaty-based norms is not new. Reference can notably be made to the interpretation of the text of resolutions of the United Nations Security Council (UNSC).Footnote 4 The Advocate General advising the Dutch Supreme Court, for instance, recently opined that ‘while Article 31 VCLT did strictly speaking not apply to a resolution of the UN Security Council, its rule of interpretation can be considered as a rule of customary international law’.Footnote 5 The reasoning appears to be that, precisely because Article 31 of the Vienna Convention on the Law of Treaties (VCLT) is of a customary nature, it can also be applied to the interpretation of sources of international law other than treaty law, such as UNSC resolutions. If that is the case, nothing stands in the way of applying the rules of interpretation laid down in Article 31 VCLT to CIL as well.
This line of argumentation is not necessarily convincing, however. There may be little doubt regarding the customary character of Article 31 VCLT,Footnote 6 but that does not make the rules of interpretation laid down in that provision applicable to sources of international law other than the treaties which the VCLT is supposed to govern.Footnote 7 In fact, that the relevant rules of Article 31 VCLT are customary means, in the first place, that they can be applied to other treaties that are not governed by the VCLT, for example, because they predate the entry into force of the VCLT in 1980, because the state party to the relevant treaty has not ratified the VCLT, or because the treaty does not fall within the scope of the VCLT (for instance because it has been concluded in oral form, or between states and other subjects of international law, or between such other subjects inter se).Footnote 8 After all, Article 31(1) VCLT specifically stipulates that ‘[a] treaty shall be interpreted in good faith’.Footnote 9 If that rule has customary character, the parallel customary rule should also state ‘a treaty shall be interpreted in good faith’.
Nevertheless, this does not mean that Article 31 VCLT has no relevance for the interpretation of norms from other sources of international law. It may have such relevance, as a material source of inspiration, or via reasoning by analogy. In all likelihood, the VCLT rules of interpretation should not be transposed lock-stock-and-barrel to the interpretation of norms derived from other sources of international law, to paraphrase Arnold McNair’s warning in the International Court of Justice (ICJ) South West Africa advisory opinion not to simply import domestic law institutions into international law.Footnote 10 Rather, when considering transposition, one may have to bear in mind the special features of other sources of international law compared to treaty law. Thus, Sir Michael Wood has sympathy for interpreters’ reliance on Article 31 VCLT when interpreting UNSC resolutions, but, given the more political nature of this source of law, invites the interpreter to pay specific attention to the circumstances in which the resolution has been adopted as well as the context of the UN Charter.Footnote 11
That the rules of interpretation devised for treaties can apply mutatis mutandis to UNSC resolutions is in any event understandable to the extent that a binding UNSC resolution is, just like a treaty, a written source of international law. Moreover, UNSC resolutions find their legal basis in a treaty (the UN Charter).Footnote 12 It is less self-evident to apply Article 31 VCLT, with the necessary modifications or not, to the interpretation of CIL norms. Unlike a treaty or a UNSC resolution, CIL is an unwritten source of international law, and it does not, at least not formally, find its legal basis in a treaty. The material source of CIL may sometimes be a treaty, for example, because subsequent to the adoption of a treaty norm, state practice and opinio juris converge on the content of that norm, but at the end of the day, for its legal existence the customary norm is not dependent on the treaty norm.Footnote 13 Because CIL is an unwritten, flexible and protean source of international law, it does not easily lend itself to the transposition of rules of treaty interpretation. What is more, the question may arise whether rules of interpretation of customary law norms serve any purpose at all, as CIL is – at least potentially – in a state of constant flux. Interpretation of norms only makes sense if those norms have a stable existence. In the classic understanding of ascertainment and identification of norms of CIL, legal authorities (law-applying or law-ascertainment agencies) always have to revisit the very existence of customary norms de novo. Although unlikely, it is after all not impossible that customary norms change or form almost overnight (instant custom).Footnote 14
This also appears to follow from the very text of Article 38(1)(b) of the ICJ Statute, which provides that the ICJ (and courts more generally one may well posit) ‘shall apply’ … ‘international custom, as evidence of a general practice accepted as law’. Pursuant to this provision, courts apply a customary norm as soon as they have established its evidence-based existence, without any need for interpretation stricto sensu. This process may perhaps appear interpretative, in that judges interpret evidentiary materials placed before them with the aim of distilling customary norms from those materials. But such interpretation takes place only in an evidentiary rather than normative sense. Judges do not interpret previously crystallised norms by analogy with Article 31 VCLT; they simply ascertain the law. Thus, Merkouris observes that judges do ‘not interpret State practice, they evaluate it, they examine its gravity for the purpose of determining the existence or not of CIL’, whereas ‘interpretation of CIL requires an already existing CIL rule’.Footnote 15
This process of CIL ascertainment or identification has been the subject of many studies, most recently by the International Law Commission (ILC).Footnote 16 In a previous publication with a co-author, this author systematised and categorised the variegated CIL ascertainment techniques used by domestic courts.Footnote 17 Triggered by Stefan Talmon’s earlier finding that, ‘when determining the rules of customary international law, the ICJ does not use one single methodology but, instead, uses a mixture of induction, deduction and assertion’,Footnote 18 it was examined whether similar processes could be witnessed in domestic courts. An analysis of a large number of recent domestic court cases bore out that this is indeed the case. Domestic courts do not normally identify CIL norms on the basis of the textbook method of ascertaining a general practice accepted as law. Rather, they tend to outsource the determination of custom to treaties, non-binding documents, doctrine or international judicial practice. Sometimes, these courts simply assert, without citing persuasive practice, the existence of a customary norm.
While, in principle, ‘other authorities’ only have evidentiary value that should be weighted with other materials which more inductively evidence (or not) the existence of a particular customary norm, one cannot escape the impression that domestic courts are simply giving effect to, or applying pre-existing customary norms, that is, norms which have been identified earlier. But if that is true, there is in principle room for the development of rules of interpretation. As Merkouris observed: ‘[O]nce CIL has been identified as having been formed, its continued manifestation and application in a particular case will be dependent on the deductive process of interpretation. In this manner, interpretation focuses on how the rule is to be understood and applied after the rule has come into existence and for its duration.’Footnote 19 If domestic courts are in fact interpreting customary norms when applying them in given cases, our earlier publication’s lament that domestic courts failed to engage in a serious CIL ascertainment process (which includes parsing all available materials),Footnote 20 loses some of its force. Indeed, assuming that customary norms existentially stabilise at one point, after which they are simply interpreted, there is no need for an elaborate process of identifying a customary norm de novo. Instead, courts may satisfy themselves with reaffirming the existence of the norm – presumably established by other law-ascertainment agencies at an earlier stage without subsequently being challenged – and instead concentrate on how to interpret the norm in a manner similar to how treaty interpretation takes place. Specific CIL rules of interpretation that are autonomous from the VCLT rules of interpretation can, in principle, develop via the regular customary process, through concurrent state practice and opinio juris.Footnote 21 Merkouris has argued that such rules already exist, and that they themselves are amenable to interpretation.Footnote 22
3 The Practice of Domestic Courts Interpreting CIL
The author’s earlier research on how domestic courts found and applied customary norms was conducted through the lens of ascertainment. The current contribution, revisits relevant court decisions with a view to understanding more in-depth how domestic courts engage in CIL interpretation.
Oxford University Press’s database International Law in Domestic Courts (ILDC) was used as the main resource to find relevant domestic court decisions. ‘Interpretation’ was used as the search term, combined with the generic subject ‘Sources, foundations and principles of international law’. The headnote of the search results subsequently indicates whether CIL was relevant to the domestic court decision.Footnote 23 Also, ILDC marks the search term – in this case ‘interpretation’ – in the summary and text of the decision, which greatly facilitated the research.Footnote 24 Methodologically, a discourse analysis of written texts (judgments) was carried out;Footnote 25 the research analysed to what extent domestic courts explicitly used the term ‘interpretation’ when applying CIL.Footnote 26 Such an analysis has its limitations in that it may discount practices of courts implicitly interpreting customary norms. Accordingly, it also included references to interpretation by the ILDC commentators directly commenting on the judgments. However, the emphasis does not lie on what courts may have meant when applying customary norms, but primarily on what they did in fact: did they consciously consider customary norms to be amenable to interpretation?
The search yielded a number of domestic court decisions which featured both ‘customary international law’ and ‘interpretation’. However, not all of these results pertain to the interpretation of CIL norms proper. Such results are ‘false positives’.Footnote 27 A first category of false positives comprises those decisions in which domestic courts erroneously use the term ‘interpretation’, when they in fact meant something else, in particular ascertainment. A second category of false positive comprises those decisions in which courts do engage in interpretation, but not of CIL, but rather of domestic (statutory) law, although in light of CIL. These two categories of false positives are briefly discussed in Section 3.1. Subsequently, Section 3.2 proceeds to the core analysis of true positives, that is, decisions in which courts genuinely interpret CIL norms.
3.1 False Positives
A number of domestic court decisions in which courts profess to interpret CIL are in fact examples of CIL ascertainment. These cases are false positives as they pertain to the identification of the very existence of a customary norm rather than its subsequent interpretation. For example, in the US Court of Appeals (Second Circuit) judgment in Kiobel, Leval, J., concurring, criticises the majority’s holding that corporate liability does not exist under CIL,Footnote 28 on the following grounds: ‘The majority’s interpretation of international law, which accords to corporations a free pass to act in contravention of international law’s norms, conflicts with the humanitarian objectives of that body of law.’Footnote 29 What the majority in fact did in Kiobel was ascertaining the very existence of a customary norm providing for liability of corporations for violations of international law, rather than ‘interpreting (the body of) international law’. Another example is the following characterisation by the US Court of Appeals (11th Circuit) of the difficulties of determining offences that violate CIL under the Offences Clause of the US Constitution (such as offences of drug trafficking):
The determination of what offenses violate customary international law … is no simple task. Customary international law is discerned from myriad decisions made in numerous and varied international and domestic arenas. Furthermore, the relevant evidence of customary international law is widely dispersed and generally unfamiliar to lawyers and judges. These difficulties are compounded by the fact that customary international law – as the term itself implies – is created by the general customs and practices of nations and therefore does not stem from any single, definitive, readily-identifiable source. All of these characteristics give the body of customary international law a soft indeterminate character that is subject to creative interpretation.Footnote 30
Here, the court refers to evidentiary interpretation, that is, the process of parsing state practice with a view to ascertaining CIL. It does not refer to the interpretation of customary norms that have already come into existence.
A final example is the US trial court judgment in Talisman, where the court held that ‘interpretations of [customary] international law [the law of nations] of the Supreme Court and Second Circuit are binding upon this Court’.Footnote 31 This case also concerned the question of whether corporations may be liable for international law violations, which, as pointed out above, is a matter of ascertainment rather than interpretation of international law. This lower court simply wanted to say that, on the basis of stare decisis, it has little agency in ascertaining CIL.Footnote 32 Of course, this need not totally exclude its interpretation of this law subsequent to its ascertainment – an issue which the court however did not address.
The search also yielded a relatively large number of potentially relevant cases that pertained to statutory interpretation in light of CIL. These cases are false positives as well, in that they are instances of ‘consistent interpretation’, that is, interpretation of domestic law in light of international law,Footnote 33 rather than interpretation of CIL proper. For instance, the Supreme Court of Appeal of South Africa held that ‘[w]hen interpreting legislation, the courts had to prefer a reasonable interpretation that was consistent with international law [including CIL] over any alternative inconsistent interpretation’.Footnote 34 Another example is the Italian Supreme Court’s interpretation of a provision in the Italian criminal code in light of CIL on the prevention of terrorism.Footnote 35 Also included in this category are a large number of immunity cases from Anglo-Saxon jurisdictions (such as the USA, UK, Canada), which have adopted specific immunity legislation, and whose courts go on to interpret such legislation in light of customary immunity rules.Footnote 36 In the end, however, all these decisions, while interesting in their own right, do not interpret customary international law, but rather statutory law, unless it happens that, when interpreting statutory law, they also explicitly interpret rather than merely apply CIL.
Somewhere on a continuum between false and true positives are instances of ‘reverse’ consistent interpretation. Reverse consistent interpretation can be defined as interpretation of CIL in light of domestic law, meaning that in case of various possible interpretations of a norm, the interpretation that is most consistent with domestic law should be chosen. An Israeli judgment can serve as an example. In a case on the scope of state immunity from jurisdiction, the Supreme Court of Israel held that ‘[a]mong various possible alternatives offered by customary international law, an Israeli court should have chosen the alternative most consistent with the basic values of Israeli law, which, in the present context, favoured the restriction of state immunity’.Footnote 37 While the Israeli court appears to be interpreting CIL, using the method of systemic interpretation, it does so in a very insulated and parochial manner, by paying heed to the values of the domestic legal system rather than to ‘the relevant rules of international law’.Footnote 38 Following Odile Ammann, such an interpretative approach can be considered as disregarding or misapplying the interpretative methods of international law, and thus lacking quality and legality.Footnote 39 Interpretation requires international interaction,Footnote 40 that is, paying attention to how other states apply and interpret customary norms. Accordingly, instances of reverse consistent interpretation can largely be considered as false positives.
3.2 True Positives
The research did not just yield decisions in which domestic courts did not engage in CIL interpretation proper. In some cases, domestic courts appear to truly interpret CIL norms. These are the ‘true positives’ in which we are interested. They demonstrate that domestic courts assume that they can interpret CIL norms,Footnote 41 even if they have not given much thought to the doctrinal underpinnings or normative consequences of CIL interpretation.
Most relevant domestic court decisions relate to immunities. This is not surprising as (1) immunities are normally invoked before domestic courts and (2) the law of immunities, in particular the immunities of states and their officials, is one of the few fields of international law that is largely governed by CIL.Footnote 42 As pointed out above, in Anglo-Saxon jurisdictions, international immunities tend to be laid down in statutes, as result of which statutory law – possibly interpreted in light of CIL – will be applied. However, in other jurisdictions, for example on the European continent, immunities are directly derived from (customary) international law, possibly via a renvoi provision in domestic legislation.Footnote 43
For analytical and pedagogical purposes, these decisions are clustered into three theoretical categories. These categories have been generated inductively through coding, conceptualising and analysing the available data (the court decisions referencing interpretation). In social science, such an approach would be termed ‘grounded theory research’.Footnote 44 As the coding exercise is carried out by human beings, the data may obviously feed into different categories.Footnote 45 However, the generic categories offered here may have particular expository power in that they are also transferable to CIL interpretation by law-interpreting agencies other than domestic courts, for example international courts. They allow us to zoom out of the particular context in which domestic courts apply and interpret law, and to reflect at a more abstract level on the practice of CIL interpretation.
The following analytical categories will be successively discussed: (3.2.1) autonomous CIL interpretation, (3.2.2) deference to CIL interpretation by other (international) courts and (3.2.3) interpreting CIL norms laid down in authoritative (written) documents. In the discussion, particular attention is paid to the method of interpretation applied by the court.
3.2.1 Autonomous CIL Interpretation
The research yielded a number of decisions in which domestic courts appeared to interpret CIL relatively autonomously, that is, without (explicitly) taking their cue from international courts’ interpretations, or from written documents purportedly codifying CIL. Most of these decisions pertain to the immunity ratione materiae of state officials from foreign criminal jurisdiction, which has not been codified, at least not until recently,Footnote 46 and regarding which international courts have given little to no guidance. A Swiss, US and Italian case were considered to be relevant.
In A v. Swiss Federal Public Prosecutor, the Swiss Federal Criminal Court interpreted the customary norms on state official immunity ratione materiae (functional immunity) as follows, in a case concerning the claimed immunity of a former defence minister of a foreign state regarding a charge of war crime:
It remained to be decided whether A’s residual immunity ratione materiae covered acts performed while in office, and whether it trumped the necessity of establishing his responsibility for alleged grave human rights violations. In light of … developments, it was not clear that this immunity should prevail, as serious crimes against humanity, including torture, were prohibited by customary international law. The Swiss legislature’s commitment to repressing ius cogens violations was an additional reason for denying A immunity ratione materiae, as it would be contradictory to express such a commitment while giving a broad interpretation to this immunity.Footnote 47
Arguably, the Swiss court assumed that a state official’s immunity ratione materiae for official acts had already crystallised as a customary norm and thus had a relatively stable existence.Footnote 48 What mattered now, was how to understand and apply the norm in respect of jus cogens violations. This is an interpretative exercise that mirrors the interpretative rule enshrined in Article 31(3)(c) VCLT, pursuant to which ‘[t]here shall be taken into account, together with the context … any relevant rules of international law applicable in the relations between the parties’.
A practice of interpreting functional immunity in respect of jus cogens violations can also be gleaned from the judgment of the US Court of Appeals for the Fourth Circuit in Yousuf v. Samantar, which concerned the same question of whether a high-ranking government official was immune from suit under head-of-state immunity or foreign official immunity for jus cogens violations, even if the acts had been performed in the defendant’s official capacity.Footnote 49 The case had been remanded by the US Supreme Court, which had held that the Foreign Sovereign Immunities Act did not govern a claim of immunity by a foreign official.Footnote 50 On remand, the Court of Appeals held that the common law, which included CIL, governed such a claim,Footnote 51 and it went on to (arguably) interpret functional immunity, holding that ‘[t]here has been an increasing trend in international law to abrogate foreign official immunity for individuals who commit acts, otherwise attributable to the State, that violate jus cogens norms – i.e., they commit international crimes or human rights violations’.Footnote 52 Admittedly, the court itself did not use the term interpretation, but the ILDC commentator conspicuously did, not only in the analysis of the judgment, but also in the Held section which is supposed to simply restate the court’s reasoning.
There is obviously a fine distinction with law ascertainment here, as it could as well be argued that whether immunity ratione materiae extends to international crimes is itself amenable to customary law formation: can sufficient state practice be identified to buttress the crystallisation of a customary law exception to the immunity ratione materiae of state officials?Footnote 53 However, both courts embraced a deductive approachFootnote 54 which emphasises the relationship of immunity with jus cogens norms. Such an approach can be termed ‘interpretative’, as it gives meaning to an established customary norm in the specific milieu of international crimes. The fact that a court may also cite other state practice (other domestic court decisions)Footnote 55 does not necessarily render the process one of customary law ascertainment, as such practice may well qualify as subsequent practice in the application of the customary norm which establishes the agreement of states regarding its interpretation, to paraphrase Article 31(3)(b) VCLT. Specifically regarding the purported immunity ‘exception’ for jus cogens violations, the systemic integration-based technique of interpretation may also be of particular relevance, that is, the interpretation of a customary norm in light of ‘any relevant rules of international law applicable in the relations between the [states]’, to paraphrase Article 31(3)(c) VCLT; jus cogens norms qualify as such rules.Footnote 56
The fine line between law ascertainment and interpretation is also apparent in another functional immunity case, Abu Omar, before the Italian Court of Cassation. In this case, which pertained to the question of whether, under CIL, state officials who had participated in an extraordinary rendition operation enjoyed functional immunity from the criminal jurisdiction of a foreign state, the court decided as follows:
The problem … consists of checking whether there effectively exists a customary law regulation under international law that also guarantees criminal immunity to the individual-entity of a sovereign state, even when it does not involve Diplomatic and/or Consular officials and high appointments of state.
On this point, jurisprudence is divided, because alongside those authorities that recognise the existence of a customary law regulation of this kind, there are others that recognise this only in respect of the activities authorised by the foreign country where these take place, while there are still others that maintain that the benefit of immunity is recognised according to specific regulations only to certain categories of entities in exercising the functions that are typical of their office.
This Court believes that this last interpretation is the more correct one, because it takes into account the developments in international relations, which as already stated, the Nato [London] Convention [Agreement between the Parties to the North Atlantic Treaty regarding the Status of their Forces] and the [Vienna Convention on Consular Relations] are valid examples.Footnote 57
At first sight, in this case, the Italian court appears to ascertain the very existence of a customary norm on functional immunity (‘checking whether there effectively exists a customary law regulation’). However, the court’s use of the term ‘interpretation’ is not necessarily misguided, as what the court may actually be doing is to interpret the scope of functional immunity, without casting doubt on the principled customary existence of functional immunity (the ‘core norm’). The judgment could be read as affirming the principled existence of customary functional immunity, while denying its blanket application to all categories of state entities exercising official functions. To reach the conclusion that functional immunity under customary law ‘only’ applies to certain categories, the court appears to have recourse to contextual interpretation, when it states that it ‘takes into account the developments in international relations’.
Finally, there is a decision by the Belgian Court of Cassation with respect to immunity from execution, which is particularly relevant from a conceptual perspective. In this decision, the court held as follows:
Il ne résulte pas de [l’article 38, § 1er, b), du Statut de la Cour internationale de Justice] que le juge étatique qui identifie et interprète une règle coutumière internationale est tenu de constater, dans sa décision, l’existence d’une pratique générale, admise par une majorité des États, qui soit à l’origine de cette règle coutumière.Footnote 58
What the court states here is that domestic courts identifying and interpreting a CIL norm are not required to establish the existence of a general practice accepted by a majority of states which is at the origin of the CIL norm. As the court uses the terms ‘identifying’ and ‘interpreting’, it is apparent that the court is not conflating law ascertainment and law interpretation. Arguably, the court uses the term ‘interpretation’ in response to the applicant’s subsidiary argument that the lower court:
[N]e justifie pas légalement sa décision en rendant applicable aux comptes d’ambassade la règle ne impediatur legatio, à supposer celle-ci établie, sans constater d’abord qu’une majorité des États admet que la règle ne impediatur legatio consacre également une immunité d’exécution diplomatique autonome des comptes d’ambassade (violation de la règle coutumière internationale ne impediatur legatio).Footnote 59
Thus, the applicant assumes, arguendo, that the CIL norm ne impediatur legatio has already crystallised,Footnote 60 and then proceeds to argue that the majority of states still need to accept that this norm also provides for autonomous diplomatic immunity from execution of embassy bank accounts.Footnote 61 The Court of Cassation rejects this argument. While in the context of law identification, this holding may possibly be problematic,Footnote 62 it is far less so in the context of law interpretation insofar as the core CIL norm has already crystallised and no proof of existence needs to be adduced. The Court of Cassation ultimately does not state what rules govern the interpretation of CIL norms (the principle of ne impediatur legatio in particular), nor does the lower court.Footnote 63
3.2.2 Deference to CIL Interpretation by International Courts
In Conclusion 13(1) of its draft conclusions on identification of CIL, the ILC states that ‘[d]ecisions of international courts and tribunals, in particular of the International Court of Justice, concerning the existence and content of rules of customary international law are a subsidiary means for the determination of such rules’.Footnote 64 And indeed, domestic courts tend to look to international courts for guidance when ascertaining international law.Footnote 65 However, they may also refer and defer to international courts which have interpreted CIL. Such domestic court decisions are relevant in that they confirm the methodological validity of interpreting CIL.
Three decisions with respect to the immunity of states, the scope of which the ICJ clarified in Jurisdictional Immunities of the State,Footnote 66 stand out. In Simoncioni, the Italian Constitutional Court cited the ‘interpretation by the ICJ of the customary rule on state immunity for acts iure imperii’ in Jurisdictional Immunities of the State.Footnote 67 By the same token, in Alessi, the Florence Court of First Instance held that the Italian court is not permitted ‘an interpretation of the binding, inescapable validity of the jus cogens rules of international law, the area in which the International Court of Justice has absolute and exclusive jurisdiction’.Footnote 68 In the context of state immunity from execution, a commentator commenting on a decision of the German Federal Court of Justice somewhat similarly pointed out that the distinction between state property used for sovereign purposes and property not so used ‘corresponded to the interpretation of customary international law on immunity from enforcement given by the International Court of Justice (‘ICJ’) in Jurisdictional Immunities’.Footnote 69
That the ICJ interpreted customary law in Jurisdictional Immunities of the State is itself an interpretation by domestic courts, for that matter. Indeed, in Jurisdictional Immunities of the State the ICJ did not explicitly use the term ‘interpretation’ in the context of immunities under CIL. Still, the judgment contains indications that the ICJ did actually interpret rather than ascertain CIL on immunities, in line with how the aforementioned domestic courts construed the ICJ’s judgment. First, with respect to immunity from jurisdiction, the ICJ stated in respect of Article 12 of the UN Convention on Jurisdictional Immunities that ‘[n]o state questioned this interpretation’,Footnote 70 that is, the interpretation that military activities are not covered by the territorial tort exception. While it may appear that the ICJ interpreted the convention and thus simply applied rules of treaty interpretation – in this case having recourse to the travaux préparatoires of the convention per Article 32 VCLT – it bears emphasis that the convention had not yet entered into force. The territorial tort exception being of customary law character,Footnote 71 the ICJ may instead have interpreted the CIL equivalent of the conventional exception. The stabilised ‘core’ customary norm is that immunity in principle does not extend to territorial torts, whereas interpretation of that norm may yield the identification of the limited circumstances in which immunity does extend to territorial torts. Second, with respect to state immunity from execution, the ICJ may have used the term ‘find’,Footnote 72 which may suggest ascertainment rather than interpretation of the law,Footnote 73 but it is of note that ‘find’ has other meanings too. The most relevant are ‘to discover’ and ‘to determine and make a statement about’,Footnote 74 the latter approximating the meaning of ‘to interpret’ as ‘to conceive in the light of individual belief, judgment, or circumstance’.Footnote 75 Accordingly, what the ICJ possibly did was to interpret a core customary norm on state immunity from execution on the basis of ‘subsequent practice in the application of the [customary norm] which establishes the agreement of [states] regarding its interpretation’, to paraphrase Article 31(3)(b) VCLT. Besides, the customary norm on state immunity could also be interpreted in light of international human rights law, in particular creditors’ rights to a remedy and to property.Footnote 76 Such an interpretation would give effect to the CIL equivalent of Article 31(3)(c) VCLT. Arguably, the relevant core customary norm is that state immunity from execution is not absolute, but relative. Under what precise circumstances state immunity does not apply will then be amenable to interpretation.Footnote 77
3.2.3 Interpreting CIL Norms Laid Down in Authoritative (Written) Documents
A third category is made up of those decisions that indirectly interpret CIL norms by interpreting the written documents in which they have been laid down. Insofar as CIL is laid down in an authoritative written text, courts will be more likely to have recourse to customary law interpretation than to customary law ascertainment, as supposedly the norm has already crystallised, black-on-white. It is the very codification of customary law which gives this body of law a more stable existence and shifts the focus to subsequent interpretation. Methodologically speaking, reliance on codification treaties to understand the meaning of CIL rules is a form of systemic interpretation mirroring the interpretative rule laid down in Article 31(3)(c) VCLT, the written text being a ‘relevant rule of international law’.Footnote 78
The most obvious written documents serving such a purpose are treaties. Thus, it is no surprise that the ICJ relied on, and arguably interpreted Article 12 of the UN Convention on Jurisdictional Immunities as CIL, as discussed in Section 3.2.2. Another example is offered by US courts’ reliance on the UN Convention on the Law of the Sea (UNCLOS), to which the US is not a party, for purposes of applying parallel CIL of the sea with the same content.Footnote 79 The application of such CIL also has an interpretative dimension, as is borne out by the Sea Shepherd case. In this case, the US Court of Appeals for the Ninth Circuit interpreted the ‘private ends’ requirement of piracy by taking the UNCLOS definition of piracy (Article 101 UNCLOS) as the starting point for its investigation of whether ‘private ends’ include those pursued on personal, moral or philosophical grounds, such as the NGO Sea Shepherd’s professed environmental goals.Footnote 80 The court held as follows: ‘Belgian courts, perhaps the only ones to have previously considered the issue, have held that environmental activism qualifies as a private end. … This interpretation is “entitled to considerable weight”’.Footnote 81 What the court was in fact doing was to interpret the ‘private ends’ variant of the CIL definition of piracy, which just happens to be codified in UNCLOS. The interpretative rule applied by the court was arguably the one based on subsequent practice, echoing Article 31(3)(b) VCLT.Footnote 82
The shift from ascertainment to interpretation, facilitated by CIL having been laid down in a written document, may not be limited to situations of CIL norms codified in a treaty. It may also extend to situations of such norms being derived from authoritative, although non-binding written documents. An example of a court apparently interpreting CIL laid down in such a document is the Haifa District Court (Israel), which held that the non-binding San Remo Manual on International Law applicable to Armed Conflicts at Sea (1990)Footnote 83 was recognised as reflecting CIL, and thus that the authority for confiscating a vessel, at issue in the case, derived from CIL.Footnote 84 The court then proceeded to find that most states required legal adjudication for an act of confiscating a vessel and also required a speedy court procedure,Footnote 85 thereby apparently interpreting the provisions of the San Remo Manual on prize law (which do not set forth a court procedure) by resorting to subsequent practice. Admittedly, the court itself did not use the term interpretation, but the ILDC commentator did, observing, in addition, that ‘any maritime court would have to address the potential impact of human rights law on the interpretation of the right to capture blockade-runners under traditional prize law’ (thus favouring systemic interpretation taking into account other norms of international law).Footnote 86 The Israeli Court decision suggests that law interpreters may consider CIL norms that have been laid down in authoritative non-binding documents to lead a relatively stable existence, and thus to be amenable to interpretation.Footnote 87
4 Concluding Observations
By and large, domestic courts, just like international courts, hew to the fiction that they find, identify or ascertain CIL. Earlier research has demonstrated that domestic courts have only limited agency in identifying CIL.Footnote 88 Instead, they tend to simply apply pre-existing CIL. However, when domestic courts apply such CIL, they may also interpret and develop CIL, as any application of law, almost out of necessity, also involves a measure of interpretation and legal development.Footnote 89 This contribution supports the TRICI project’s methodological premise that CIL norms, just like treaty norms, can be interpreted. Interpretation will notably take place after a ‘core’ CIL norm has crystallised and stabilised, after which the penumbra of that rule – its precise scope, its exceptions – are amenable to mechanisms of interpretation.
This chapter has analysed a large dataset of domestic court decisions relevant to CIL, and found that, indeed, domestic courts at times engage in CIL interpretation, even if they largely refrain from using that term. Domestic courts may interpret CIL autonomously, may defer to and validate international courts’ CIL interpretations, or they interpret written documents, such as treaties, codifying CIL norms. Such practices bear out that domestic courts may consider some core CIL norms to be relatively stable and amenable to further refinement through interpretation.
When interpreting CIL, domestic courts appear to resort mainly to systemic interpretation and interpretation on the basis of subsequent state practice. This reflects earlier findings by Panos Merkouris with respect to CIL interpretation by international courts.Footnote 90 In particular, domestic courts apply by analogy the canons of construction laid down in Article 31(3)(b) VCLT (‘any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation’), and Article 31(3)(c) (‘any relevant rules of international law applicable in the relations between the parties’). The analysis of state practice under the CIL equivalent of Article 31(3)(b) VCLT tends to be cursory, however. As Odile Ammann has pointed out, mainly in respect of CIL ascertainment, domestic courts tend to refer to their own practice or the practice of their own state, a process which she characterises as self-referentiality and circularity.Footnote 91 Cast in terms of interpretation, such a process may take the form of ‘reverse consistent interpretation’: domestic courts may choose this interpretation of a CIL norm which is mostly in line with domestic law and practice.Footnote 92
In future cases, domestic courts deciding cases on the basis of CIL may in any event want to be more explicit regarding whether they engage in de novo CIL norm-identification or rather in the interpretation of pre-existing and stabilised customary norms. In case of interpretation, they may want to improve the methodological quality of their reasoning by pinpointing the canon of construction which they apply (e.g., systemic interpretation; interpretation on the basis of subsequent practice). Finally, in the case of both CIL identification and interpretation, they should make sure that they rely on sufficient international (including foreign) legal practice, in keeping with the methodological requirements of CIL identification and interpretation.