1 The Regulation of Sources of Law

The law on the sources of law in international law, if any, appears to be largely neglected by scholarship. So seems to be state practice of regulation of the sources of law in international law, even where questions about any legal consequences of such practice are not raised. That these two aspects of lawmaking remain understudied, is, in a way, unsurprising: scholarship has remained divided about the very concept of ‘the sources of international law’, as evidenced by vexed controversies about their nature.

Article 38(1) of the International Court of Justice (ICJ) StatuteFootnote ³ almost invariably features in all of these controversies. More importantly, albeit rarely acknowledged in contemporary scholarship, ICJ Statute Article 38(1) is invoked in state practice regarding the sources of law in international law. Two positions about such state practice can be found among scholars and international courts and tribunals. Some accept that regulation patterns are inferable from such state practice, but stop short of drawing any legal consequences from those patterns. Others go further and argue that such patterns in state practice do have legal consequences, amounting to law in the form of rules, particularly general rules of customary international law (CIL). In this vein, ICJ Statute Article 38(1) is said to reflect such CIL general rules. The former position, which the first epigraph epitomises, has become the standard one. The latter position, of which the second epigraph is illustrative, enjoys support among some leading, mostly early, commentatorsFootnote ⁴ and, at least, one international court.Footnote ⁵

‘The sources of international law’, often used in the plural as a set phrase, is a concept which has constantly evaded precise definition. The multiplicity of meanings attributed to it, as Sur has noted, has resulted in contestations of its pertinence.Footnote ⁶ Kelsen, for instance, observed that it designates not only ‘modes’ of lawmaking and ‘reasons’ for the validity of law, but also its ‘ultimate fundament’.Footnote ⁷ According to Truyol y Serra, the linkage of these two aspects of lawmaking accounts for various controversies.Footnote ⁸ As Dupuy correctly notes, it ought to be, and has in a way increasingly been, accepted that a source of law is distinct from the law’s ultimate basis.Footnote ⁹

Notwithstanding the consensus reached among scholars about the distinction between the basis of a legal order and lawmaking within that legal order, various disputes about the nature of the sources of international law stemming from the concept’s polysemy remain unresolved. As Ago noted, other differences over their nature result from persistent reliance on certain assumptions,Footnote ¹⁰ and, as Truyol y Serra observed, variations as to those assumptions result in the intractability of related controversies.Footnote ¹¹ Some of those assumptions, in turn, may involve a conflation of levels of analysis, an implication to which writers of various schools of thought have drawn attention.Footnote ¹² Tunkin, for instance, regarded international law as a ‘multi-level system’.Footnote ¹³ Abi-Saab and Wood, for their part, rightly warn against such a conflation.Footnote ¹⁴

The idea of regulation, as it pertains to the sources of international law, is widely discussed. More broadly, general jurisprudence has also contributed to the understanding of ‘regulation’ in ways which are apposite to this chapter. Without prejudice to a fuller discussion of its general jurisprudential meaning, which falls outside the scope of this chapter, ‘regulation’ is used here to designate the making of rules, whether they have attained or not a legal status, or if so, whether they directly govern conduct or not. First, by encompassing rules which arguably cannot, or, if so, have not yet, attained a legal status, regulation gives expression to the common ground among the aforementioned schools of thought, accepting there is state practice on sources of law. Second, the idea of regulation, as opposed to ‘norm’ in its theoretical sense, is key: while a norm is a rule aiming to guide conduct, other rules may lack such a normative character, and yet still constitute a form of regulation, alongside normative regulation.Footnote ¹⁵ While a distinction between normative and non-normative regulation may ostensibly overlap with Hart’s distinction between primary and secondary rules, a reference to regulation seeks to place emphasis on the legality of non-normative legal regulation, as well as on its place within a legal system as internal to it. In this vein, it is worth recalling that Hart regarded secondary rules on law ascertainment as non-legal and, ultimately, external to the legal system. Third, the elements of any custom giving rise to such CIL general rules may be better understood. Once the assumption that every custom need derive from the same kind of general practice as that leading to the formation of primary, normative, CIL rules is set aside, any custom giving rise to non-normative rules of CIL, including those on sources of law, can be the object of the same legal scrutiny to which any other custom can be typically subjected. Most notably, any enquiries into such non-normative CIL would not be discarded by any misconception confining CIL to rules of CIL derived from practice consisting in ‘physical’, as opposed to ‘verbal’, acts.

The view that there is a phenomenon of regulation of sources of law in international law, and that such regulation is carried out by a ‘system of sources’ contained within the legal order of international law as a whole, finds some support in international law scholarship. Virally, for instance, considered that legal orders are generally ‘self-regulated’, including as to their own sources of law. Virally’s view that international law, as any legal order, self-regulated its own sources of law was without prejudice to admitting that such autonomy was relative, the legal order of international law being conditioned by the various circumstances within which it operates.Footnote ¹⁶ Virally’s caveat is not contradictory, since it involves a level of analysis other than that of the rules performing self-regulation of the system’s sources of law, namely that of the various wider processes within whose framework the legal system operates.

2 Custom as Source of Rules Arising out of General Practice on Sources of Law

This section examines the place of custom in the regulation of the sources of international law, with a particular focus on custom’s role as source of the law on sources of law, if any, in international law.

The suitability of custom as a source of universal rules has been widely accepted in the literature. An analysis of custom’s suitability as a source of universal rules usually involves a comparative analysis, vis-à-vis other sources of law.Footnote ¹⁷ Such comparative analyses have tended to point out its inherent qualities. For instance, Marek argued that custom’s inherent qualities rendered it ‘superior’ to any treaty as a source of universal rules.Footnote ¹⁸ Marek characterised this superiority as being a form of ‘inherent superiority’ or ‘superiority of quality’, and not a matter of hierarchy among sources of law.Footnote ¹⁹

As it relates to general rules regarding the sources of law, on the other hand, the suitability of custom is widely contested. Those who contest the suitability of custom for these particular purposes often deny the possibility of regulation of sources of law by any rule created by one of the regulated sources of law. Jennings and Watts’ view, partly quoted in the first epigraph to this chapter, furnishes a typical statement of this denial: ‘Article 38 of the Statute of the International Court of Justice … cannot itself be creative of the legal validity of the sources set out in it, since it belongs to one of those sources itself.’Footnote ²⁰ While Jennings and Watts’ denial concerns ICJ Statute Article 38 qua treaty only, the view is usually predicated of any other claimed source-based rules on sources, including custom-based ones. This is exemplified by Dinstein’s view, for whom reliance on CIL rules on ‘how and when custom is brought into being’ inherently involves ‘a petitio principii’.Footnote ²¹ These instances of reluctance to ascribe legality to the regulation of sources of law, which could be collectively called, using Dinstein’s term, petitio principii objections, are, again, without prejudice to the concomitant acknowledgement of the existence and importance of relevant patterns of regulation of sources of law in general practice.Footnote ²²

Leaving aside the petitio principii objections, the only major objection to the idea of regulation of sources of law and its character as law, in the form of general rules of CIL, might arise from various forms of scepticism as to the idea of regulation or, where accepted, its legality.Footnote ²³ This scepticism is not easily amenable to analysis, since it appears to be latent in the respective bodies of scholarship, never being made explicit by virtue of the very view that it would be pointless to engage in any further arguments against the idea of regulation or its legality, if any. The assumed futility of regulation of sources of law or its legality may explain the lack of arguments in the event of a dismissal of a petitio principii objection on the part of scholarship underpinned by this assumption: in a way, this assumption implies that the vacuum which would be left if the respective petitio principii objection were disproved is one which scholarship based on this presumption has chosen to leave unaddressed. This assumption may underlie the view, expressed by the United Nations’ International Law Commission (ILC) Special Rapporteur on CIL identification, Sir Michael Wood, that ‘[i]t is perhaps unnecessary, at least at this stage, to enter upon the question of the nature of the rules governing the formation and identification of rules of customary international law, for example, whether such rules are themselves part of customary international law’.Footnote ²⁴ In support of this proposition, Special Rapporteur Wood quotes Sinclair’s view on ‘the debate on the nature of some rules of treaty law, particularly pacta sunt servanda’, to the effect that such an enquiry involved ‘doctrinal arguments’, ultimately leading the enquirer to ‘metaphysical regions’.Footnote ²⁵

The remainder of this section proposes to set aside, for the sake of argument, the above petitio principii objections, and to focus, instead, on examining patterns of regulation of sources of law as they arise in general practice in which ICJ Statute Article 38(1) is used outside ICJ proceedings.Footnote ²⁶

There are two bodies of materials in which ICJ Statute Article 38(1) is used outside ICJ proceedings: decisions of international courts and tribunals and state practice itself. Bearing in mind the difference between these bodies of materials is very significant (a question to which Section 3 returns), and although this section is mainly concerned with selected state practice, it is worth recalling that the place of ICJ Statute Article 38(1) in decisions of international courts and tribunals is widely acknowledged.Footnote ²⁷ For instance, Crawford observes that ‘[ICJ Statute] Article 38(1) has been taken as the standard statement of the so-called “sources” of international law for all international courts and tribunals’.Footnote ²⁸ Charney, in a study concerning the proliferation of international courts and tribunals, reached a similar conclusion.Footnote ²⁹ He inferred that uniformity among international courts and tribunals regarding the sources of law shows that the proliferation of international courts and tribunals has not eroded ‘the international law doctrine of sources’.Footnote ³⁰ The aforementioned reliance on ICJ Statute Article 38(1) by contemporary international courts and tribunals gives continuation to the analogous practice of arbitral tribunals constituted prior to the adoption of the ICJ Statute. Those tribunals invoked Article 38 of the Statute of the Permanent Court of International Justice (PCIJ).Footnote ³¹ As Mendelson observes, PCIJ Statute Article 38 had been ‘treated as an authoritative list by various arbitral tribunals’.Footnote ³² Earlier scholars had also recognised the significance of PCIJ Statute Article 38.Footnote ³³

The place of ICJ Statute Article 38(1) in general practice is paramount and more significant than credited in contemporary scholarship. This general practice, whereby states have characterised sources of law through express invocation of, or through statements largely consistent with those contained in, ICJ Statute Article 38(1), is twofold, taking the form of conduct of state organs for international relations, as well as decisions by state judicial organs.

The first category of relevant general practice consists in inter-state arbitration agreements, multilateral treaties beyond matters of dispute settlement, and statements in international organisations, including the United Nations (UN).

A paramount instance of this category is the very adoption of the ICJ Statute. Indeed, it is widely considered that the identity in content between Articles 38 of the PCIJ and ICJ Statutes, except for the opening sentence introduced in the latter, confirms the continuity of the rules stated in both provisions. Furthermore, for several scholars, this continuity evidences that what matters most about the statements contained in Article 38, common to the PCIJ and ICJ Statutes, is not their character as rules qua treaty, but their broader place beyond the confines of dispute settlement by the PCIJ and the ICJ, respectively.Footnote ³⁴ That this wider significance was attributed to ICJ Statute Article 38 is further confirmed by the fact that proposals to modify its content, in order to account for other categories of acts with purported general lawmaking effects, were unanimously rejected in debates leading to the adoption of the UN Charter,Footnote ³⁵ whose preamble expressly states the importance of the ‘sources of international law’.Footnote ³⁶

The reference to ICJ Statute Article 38 in other major multilateral treaties lends additional support to its wider role in the regulation of sources of law. For instance, the reference to ICJ Statute Article 38 in Articles 74(1) and 83(1) of the UN Convention on the Law of the SeaFootnote ³⁷ is considered as a general ground for denying the character of equitable principles as legally binding.Footnote ³⁸ Other major multilateral treaties, in which no reference to ICJ Statute Article 38(1) is made, are widely regarded as having been negotiated on the understanding that ICJ Statute Article 38(1) underpinned the terms used, as exemplified by Article 42 of the Convention on the Settlement of Investment Disputes between States and Nationals of Other States.Footnote ³⁹

Various states have made multiple statements to the effect that ICJ Statute Article 38(1) sets out the sources of ‘positive’ international law exhaustively and satisfactorily.Footnote ⁴⁰ Even more pertinently, in some instances, states may indicate that ICJ Statute Article 38(1) provides a ‘legal basis’ for statements about sources of international law. In doing so, states rely on ICJ Statute Article 38(1) as a ‘legal basis’ for the sources of law, whether seen as a category in wholeFootnote ⁴¹ or with regard to individual sources. Furthermore, states rely on ICJ Statute Article 38(1) in order to deny that a subsidiary source is a proper source of law,Footnote ⁴² or to substantiate their affirmation that individual recognised sources are indeed sources of law proper.Footnote ⁴³

The second category of relevant general practice comprises two forms of state practice, namely decisions of domestic courts and other forms of practice in connection with domestic judicial proceedings. The former subcategory includes decisions constitutive of state practice, capable of giving rise to custom within the meaning of subparagraph (b) of ICJ Statute Article 38(1), by contrast to their other potential role, as a subsidiary means, under subparagraph (d) thereof. The latter subcategory encompasses pleadings by foreign states opposing the execution of arbitral awards before domestic courts of the place where execution is sought.

The authoritativeness of ICJ Statute Article 38(1) is widely affirmed by domestic courts and tribunals in the first subcategory of practice surveyed. For instance, the United States District Court for the Southern District of New York, in Presbyterian Church of Sudan v. Talisman Energy, Inc, observed that ‘[t]he Second Circuit has cited Article 38(1) as an authoritative reflection of the sources of international law’.Footnote ⁴⁴ This decision is notable for referring to the character of ICJ Statute Article 38 as a ‘reflection’ of, as opposed to a provision directly governing, sources of law. Other courts have emphasised the character of ICJ Statute Article 38 as a formulation enjoying authority beyond ICJ proceedings. For example, in Handelskwekerij GJ Bier BV & Stichting Reinwater v. Mines de Potasse d’Alsace SA, the Amsterdam District Court quoted ICJ Statute Article 38 and held that it ‘must be taken as an authoritative formulation of the sources of international law, inside or outside the International Court of Justice’.Footnote ⁴⁵

Some domestic courts go on to indicate that reliance on ICJ Statute Article 38 is necessary, and not merely called for given its authoritativeness. The Argentine Supreme Court of Justice in both Simon and others and Arancibia Clavel, quoting ICJ Statute Article 38, observed that ‘[i]t is necessary to determine what are the sources of international law … what is provided for by the Statute of the International Court of Justice has to be taken into account’.Footnote ⁴⁶ This stance was confirmed by the Argentine government in a statement at the UN, concerning the place of ICJ Statute Article 38 in its internal judicial practice.Footnote ⁴⁷ Similarly, the Supreme Court of Chile, in Lauritzen and others v. Government of Chile, invoked ICJ Statute Article 38 in support of its statement that ‘customs and treaties figure among the traditional sources, to which may be added principles’.Footnote ⁴⁸ In other cases, such an invocation is stronger, being qualified to the effect that observance of ICJ Statute Article 38 is not only necessary, in a conceptual sense, but also legally required. For instance, the Indonesian Constitutional Court, in Law 27 of 2004 on the Truth and Reconciliation Commission, considered whether a given alleged general principle of law had been ‘created in accordance with the provisions of the Statute of the International Court of Justice regarding the sources of international law’.Footnote ⁴⁹ For similar purposes, Argentina invoked ICJ Statute Article 38(1)(b) before the Court of Cassation of Belgium in Argentine Republic v. NMC Capital, as a rule of law allegedly breached, in support of one of the grounds for her request for cassation.Footnote ⁵⁰

Before concluding this succinct survey of state practice, it is worth revisiting Jennings and Watts’ discussion of the wider value of ICJ Statute Article 38(1). They reiterate their view that, since ‘[ICJ Statute] Article 38 … cannot be regarded as a necessarily exhaustive statement of the sources of international law for all time … [t]hose sources are what the practice of states shows them to be’.Footnote ⁵¹ And yet, a key aspect of their analysis of the continuing wider relevance of ICJ Statute Article 38(1) lies in ‘[t]he fact that the International Court of justice, in its numerous judgments and opinions relating to international organisations, has always been able … to dispose of the questions arising for decision’.Footnote ⁵² While Jennings and Watts justifiably refer to the practice of the ICJ, since they were concerned with the sufficiency of relying on ICJ Statute Article 38(1) in ICJ proceedings, their reference is notable because it is representative of the tendency to exclusively focus on decisions of international courts and tribunals in spite of general statements to the effect that the primary object of enquiry should be state practice itself. As discussed in Section 3, Jennings and Watts are not alone in their tendency, as the work of earlier scholars who did not raise any petitio principii objection to the possibility of CIL on sources of law shows. Indeed, Section 3 shows that practice-based accounts, whether source-based or not, have heretofore tended to overlook state practice itself, given their assumptions regarding state practice, particularly as to decisions of domestic courts.

The character of decisions of domestic courts as general practice has raised various questions, which call for some elucidation of their precise nature,Footnote ⁵³ before turning to the question of the existence of a CIL on sources of law based on states’ general practice primarily in the form of decisions domestic courts. It is common to treat selected judicial decisions as ‘subsidiary means’ under ICJ Statute Article 38(1)(d).Footnote ⁵⁴ This tends to be the case despite their multiple roles.Footnote ⁵⁵ One of those roles is as a form of general practice under ICJ Statute Article 38(1)(b).Footnote ⁵⁶ Lauterpacht had reached the same conclusion regarding PCIJ Statute Article 38.Footnote ⁵⁷ Several scholars likewise accept that these two roles may be concurrently performed.Footnote ⁵⁸

Leaving aside the dual role a domestic court decision may play under subparagraphs (b) and (d) of ICJ Statute Article 38(1), and focusing on the former role, there is some debate as to whether the two elements of custom within the meaning of subparagraph (b) are present. Whether a decision of a domestic court constitutes practice is a question which partially overlaps with debates over whether practice need consist in physical, as opposed to verbal, acts. Those debates have lost currency, since it has become increasingly uncontroversial to regard verbal acts, including in the form of written statements, as a form of state practice. Indeed, both the ICJ,Footnote ⁵⁹ the ILC,Footnote ⁶⁰ and some states commenting on the ILC’s recent work on CIL identification,Footnote ⁶¹ have acknowledged that decisions of domestic courts can constitute general practice for the purposes of custom formation and CIL identification. An alternative rationale for the reluctance to accept state practice in the form of domestic courts decisions might lie in the assumption that CIL only encompasses normative regulation. Such an assumption would confine CIL to primary rules, to the exclusion of non-normative regulation, of which secondary rules, including rules on sources of law, are a notable instance. This assumption would translate into a view demanding that all CIL rules derive from the kind of general practice which underlies CIL primary rules, often derived from practice in the form of physical acts. As discussed in Section 1, this assumption results from a misconception requiring all forms of regulation to be normative. The character of decisions of national courts as a form of acceptance as law, or opinio juris, on the other hand, has raised less controversy.Footnote ⁶² Some accept their role, but qualify which decisions are more suitable to constitute opinio juris.Footnote ⁶³

The question of whether and how the two constitutive elements of custom may be satisfied by a set of statements, including those in domestic court decisions, warrants some further examination. Some scholars have accepted the concurrent character of decisions of national courts as practice and acceptance as law.Footnote ⁶⁴ While the concurrent character of internal judicial decisions may be contested given its potential for so-called ‘double-counting’,Footnote ⁶⁵ a set of separate verbal acts cannot be lightly disregarded as establishing both elements of custom. For instance, Argentina’s position is apposite, as an example of how a variety of separate statements, including domestic court decisions, may constitute general practice in support of a position and acceptance as law of that position. Indeed, while Argentina’s judicial organs engage in actual instances of practice, such as the invocation of ICJ Statute Article 38(1) in decisions of the Argentine Supreme Court, among others, other organs separately issue statements clearly indicating that state’s opinio juris to the same effect, such as Argentina’s unequivocal statements at the UNGA concerning the legal value of ICJ Statute Article 38(1).

3 Customary International Law as Law on Sources of Law in International Law: Custom in Foro or in Pays?

This section examines selected claims of existence of law on sources of law, with a particular focus on major models of CIL rules on sources of law, in international law.

The claim that ICJ Statute Article 38 contains statements regarding CIL rules on sources of law has taken various forms. Some advance the claim unqualifiedly. For instance, Ohlin has recently stated that ICJ Statute Article 38 ‘embodies a customary norm’ regarding the sources of international law.Footnote ⁶⁶ He goes on to argue that ICJ Statute Article 38 is such a ‘direct statement about the sources of law’ that it ‘might be the closest thing one could find in any legal system –domestic or international – to a pure rule of recognition’.Footnote ⁶⁷

Some add that ICJ Article 38, while not directly embodying CIL rules on sources of law, is reflective or declaratory of such CIL rules.Footnote ⁶⁸ While both claims point in the right direction, the view that ICJ Statute Article 38 is reflective, rather than directly constitutive, of CIL is more accurate. Hence, ICJ Statute Article 38 does not in itself ‘embody’ CIL. As Sur explains metaphorically, CIL, albeit ‘invisible’, is reflected in ‘mirrors’, and yet ‘these mirrors are not the rule’ of CIL.Footnote ⁶⁹

The claim that ICJ Statute Article 38(1) has a declaratory or reflective character with respect to sources of law is formulated variously.Footnote ⁷⁰ Some refer to a ‘doctrine’, but not to rules as such. For example, Dolzer refers to ‘the traditionally accepted doctrine of sources, as reflected in the [PCIJ and ICJ] Statutes … (Article 38)’.Footnote ⁷¹ Some do refer to rules as being reflected, but do not indicate their legal character. For instance, Tomuschat simply refers to ‘[t]he rules on law-making, as they are reflected in Article 38(1) of the ICJ Statute’.Footnote ⁷² Other scholars refer to the existence of law and its being declared by ICJ Statute Article 38, without indicating the source of the law declared.Footnote ⁷³

Those who claim that ICJ Statute Article 38 is reflective or declaratory of CIL on the sources of international law may qualify those CIL rules as being of general character. This is illustrated by Abi-Saab, who, noting that ICJ Statute Article 38(1) is commonly perceived to be declaratory of ‘general international law’ on sources, adds that such general international law corresponds to Hart’s ‘secondary rules of change’.Footnote ⁷⁴ Some of those who deem ICJ Statute Article 38 as declaratory sometimes hold this claim in relation to propositions regarding specific sub-systems of international law.Footnote ⁷⁵

The attribution of the character as declaratory or reflective of CIL to ICJ Statute Article 38 is not entirely novel, since this was equally predicated of PCIJ Statute Article 38. Verdross cited approvingly a 1928 arbitral award holding that, in the event of ‘silence of the compromis on the sources of law, every international arbitral tribunal must apply the rules of the law of nations, taking into account the definition contained in Article 38 of the [PCIJ] Statute’. Verdross implied that custom served as legal basis for findings like this one. In fact, he inferred from the ‘long history’ of arbitral tribunals’ invocation of general principles of law (a source of law already included in PCIJ Statute Article 38) without ‘special authorisation’, that ‘the application of such principles has been sanctioned by international custom’.Footnote ⁷⁶ It is notable that Verdross, unlike Jennings and Watts, did not see any inconsistency in relying on a source of law, such as custom, as basis for the legal character of another source of law, such as general principles of law. Lauterpacht also deemed PCIJ Statute Article 38 as declaratory of ‘custom expressed by a long series of conventions and arbitral awards’.Footnote ⁷⁷ Lauterpacht added, with particular reference to PCIJ Statute Article 38(3) (which would become ICJ Statute Article 38(1)(c)), that it was ‘purely declaratory’ since, prior to the PCIJ Statute, both ‘arbitral practice and arbitration agreements’ recognised general principles of law.Footnote ⁷⁸

Various leading authors have more recently noted that declaratory character is attributed to ICJ Statute Article 38. With respect to ICJ Statute Article 38(1)(c), Jennings and Watts, despite their petitio principii objection, reported on the ‘fact’ that ‘a number of international tribunals, although not bound by the Statute, have treated that paragraph of Article 38 as declaratory of existing law’.Footnote ⁷⁹ Monaco pointed to the role of PCIJ Statute Article 38 in giving concrete expression to a ‘preexisting practice’.Footnote ⁸⁰ Sur, likewise, attributes declaratory character to PCIJ Statute Article 38.Footnote ⁸¹ Pellet, in his 2012 survey of uses of ICJ Article 38, discusses various international instruments, and implies that some refer indirectly to ICJ Statute Article 38.Footnote ⁸²

Some have gone further, holding that ICJ Statute Article 38 codifies the ‘sources’ of international law,Footnote ⁸³ or, more precisely, the CIL rules governing the ‘sources’ in international law.Footnote ⁸⁴ Supporters of the view that ICJ Statute Article 38 is codificatory include earlier scholars, such as Lauterpacht.Footnote ⁸⁵ Along similar lines, Conforti referred to the role of PCIJ Statute Article 38 as a codification of the ‘practice followed by international tribunals’.Footnote ⁸⁶ Lepard, for his part, not only claims that ICJ Statute Article 38 is codificatory, but also attributes to it authoritativeness as a statement of CIL rules on sources of international law directly.Footnote ⁸⁷ Lepard’s statement is notable, since most contemporary writers who regard ICJ Statute Article 38 as authoritative fail to indicate whether it is so qua treaty or qua statement of a separate rule, including any CIL rule.

The foregoing discussion has shown that, in essence, there are two models of CIL on sources of law in scholarship. Before delving into these two models, a discussion of some conceptual underpinnings is warranted. In particular, Bentham’s distinction between custom in foro and custom in pays sheds light on the nature of these two models.Footnote ⁸⁸

This distinction has been recently revisited by Lamond, as part of his critique of Hart’s conception of the rule of recognition. Lamond argues that Hart’s characterisation of the rule of recognition as a form of ‘collective social practice of officials’Footnote ⁸⁹ fails to account for its additional, and more important, character as ‘a form of customary law’.Footnote ⁹⁰ Significantly, Lamond advocates the importance of characterising rules on sources of law as customary, not merely as practice-based. While Lamond does not put forward a source-based model of customary rules of recognition, his analysis of the character of the rule of recognition as a form of law standing on an equal footing with other forms of law within a legal system is pertinent.Footnote ⁹¹

The following discussion focuses on Lamond’s conception of customary rules of recognition, its transferability to an analysis of international lawmaking, and the various aspects of his portrayal of that kind of customary law as internal, systemic and foundational.

The internal character of the rule of recognition is an important aspect, and its denial, particularly where justified by the comparison between games and law, warrants further examination. Lamond contests the relevance of this comparison, which holds that, since players of games are not required to determine the rules of the game as part of the game, those second-order rules are not part of the rules of the game. Instead, Lamond argues, the assumption that game-playing necessarily excludes the creation of rules of game-playing does not hold true with respect to legal systems, which are precisely concerned with a wide range of regulation, ‘including crucially the activities of law-identification and law-creation themselves’.Footnote ⁹² Lamond’s critique of the widespread reliance on the ‘rules of the game’ comparison to justify a segregation of the rule of recognition from other rules of law is apposite, in the sense that it reminds that a separation in all respects of secondary rules from primary rules, including as to their making and identification, is mistaken. This, as discussed in Section 1, would be as unjustified as assuming that all regulation need be normative in nature. Indeed, leaving aside functional differentiations, secondary and primary rules may partake in the same properties, as rules of the same legal system, including their source-based creation and identification. This is all the more relevant in international law, given its so-called horizontality whereby, among others, even jus cogens rules need not ‘displace [the] application’ of certain non-peremptory rules.Footnote ⁹³

The systemic character Lamond attributes to custom, is also of relevance to international lawmaking. Lamond’s systemic account of acceptance aptly introduces the idea of levels of analysis. Indeed, while acceptance may occur at the separate levels of an individual rule and that of the legal system, acceptance ultimately performs the same function, including at the second-order level where a rule of recognition operates.Footnote ⁹⁴ A similar approach is defended by Mendelson in his analysis of the place of consent in the formation of custom. In particular, Mendelson aptly critiques voluntarist theories for importing consent ‘[a]t the most general, systemic level’ into the ‘identification-of-sources level’.Footnote ⁹⁵

The source-based character of a customary rule of recognition is an area where Lamond’s model may be of limited relevance, and, in a way, partake in the shortcomings of Hart’s conception of the rule of recognition, for the purposes of international lawmaking. For Lamond, the rule of recognition’s foundational character does not detract from its place as internal to the legal system. Indeed, he argues, ‘the fact that the rule of recognition is the ultimate basis for source-based standards in the system, and does not owe its status to satisfying the criteria in some further standard, does not show that it is not internal to the system of laws’.Footnote ⁹⁶ He appears to accept the idea that the non-application of the standard demanded by the rule of recognition for the creation of custom to the very creation or identification of the rule of recognition is not at odds with the character of the rule of recognition as a customary rule which is internal to the legal system. This seems to be justified by Lamond’s apparent view that source-based and non-source-based law are equal parts of a legal system. This view is evidenced by his conception of custom in foro. He defines custom in pays as ‘the custom of non-officials recognized by the law … not … the custom of the officials themselves’. By contrast, he defines custom in foro as ‘customary law’ which ‘rests on being applied in the practice of the courts’. In particular, he lays out four features of custom in foro: ‘customary legal standards … are: (a) authoritative for the courts; (b) not validated by another legal standard; (c) depend for their existence on being applied in the practice of the courts; and (d) belong to a system of laws’.Footnote ⁹⁷ Nevertheless, establishing a source of law for rules governing sources of law is key to properly determine how such rules may be created, changed or terminated. Lamond himself seems to hint at the importance of characterising a rule as customary, including the rule of recognition, since such an understanding determines ‘[a]t a practical level, … how we think they can or cannot be altered’.Footnote ⁹⁸ Yet, it is unclear why understanding the rule of recognition as a form of customary law, without treating it as a form of source-based law, derived from a custom, suffices to determine how that customary law can be changed, let alone how it is created or terminated. As the present author has shown elsewhere with regard to strands of constitutional theories of international lawmaking, theories which do not identify a source of law for rules regulating sources of law are bound to fail in their attempts to properly account for changes of those rules.Footnote ⁹⁹

It is fitting to now discuss a caveat to the distinction between custom in foro and custom in pays. Lamond states that ‘[c]ustomary international law is most similar to the custom of non-officials recognized by the law (custom ‘in pays’), not to the custom of the officials themselves’.Footnote ¹⁰⁰ This caveat is partly accurate. On one hand, it may misrepresent the dual character of states as both law-addressees (and, in that sense, non-officials) and lawmakers (and, thus, officials, to the extent that they have lawmaking power), a duality encapsulated in Scelle’s concept of ‘dédoublement fonctionnel’.Footnote ¹⁰¹ On the other hand, Lamond’s caveat correctly reminds that custom is a source through which the general practice of subjects of law becomes law (custom in pays), independently of the practice international courts and tribunals develop as they settle disputes (akin to a custom in foro). Indeed, the ‘officialdom’ of international courts and tribunals, in their capacity as dispute settlers, is typically limited to the confines of the jurisdiction to which the subjects of law submitting their disputes have consented.

Despite being evidently inconsistent with the tenet of the primacy of state practice, as primary subjects of international law, the large majority of existing theories of CIL on the sources of law are largely modelled after a form of custom in foro. As such, those theories seek to establish the existence, scope and content of rules on sources of law without due consideration of any relevant general practice of states. Instead, custom is frequently ascertained primarily by reference to decisions of international courts and tribunals. Exceptionally, selected decisions of domestic courts may be examined, provided that they can be regarded as more authoritative than regular national judicial decisions and placed on the same plane as international decisions. This is the model followed by various leading writers supportive of CIL accounts of regulation of sources. In fact, Lauterpacht and Verdross reached their conclusion as to the declaratory status of PCIJ Statute Article 38 solely on the basis of its use by international tribunals other than the PCIJ, without placing the various arbitration agreements underling those treaty-based arbitrations at the centre of their respective claims of existence of CIL then declared by PCIJ Statute Article 38. This model is replicated in more recent accounts which accept the idea of CIL rules on sources of law, as exemplified by Tams’ analysis of ‘meta-rules’ on sources of law, which focuses mainly on ICJ decisions.Footnote ¹⁰²

By contrast, a model of CIL whereby custom is created through (and ascertained by reference to) state practice proper, is what the notion of custom in pays calls for. By means of such general practice, states make determinations as to sources of law in various contexts, whether they engage in lawmaking or in law-identification in connection with dispute settlement proceedings, on the basis of rules which they accept as law. This kind of custom is a model of CIL which is advocated by some scholars, albeit the degrees to which they substantiate their claims varies. Major examples of claims of CIL on sources of law which intend to rely on state practice, but do not fully substantiate those claims by reference to actual instances of state practice, can be found in the work of Henkin. He argued that customary law was a default category including diverse bodies of law which would not necessarily meet the requirements for the existence of custom proper. Nevertheless, and interestingly for the purposes of the present chapter, Henkin claimed that there exists custom regulating the sources of law, and considered that it constituted a form of custom proper, one based on actual general state practice and accepted as law.Footnote ¹⁰³

4 Conclusions

This chapter has explored various aspects of the idea that there may exist a law on the sources of law, as opposed to a mere theory or doctrine thereof, in the form of CIL. In doing so, the chapter has provided an overview of the existing theories, both in general jurisprudence and international law scholarship, and assessed their limitations and merits.

Section 1, by way of background, has indicated that, in spite of sharing important common ground, as with many other vexed issues in the theory of the sources of international law, there is a controversy as to the idea of regulation of sources of law and the legality of rules on sources of law, if any, among major schools of thought. At the core of debates concerning this idea, it was submitted, is the fact that regulation tends to be conflated with normativity, unjustifiably denying the legality of non-normative regulation.

Section 2, in particular, has shown that the standard position, albeit professing to be practice-based, rejects a source-based account of regulation of sources of law, on grounds of a petitio principii objection. It has proposed to set aside, for the sake of argument, any petitio principii objection(s), and suggested that the standard account, seemingly assuming the futility of the idea of regulation of sources, has neglected the task of addressing alternative arguments against a source-based account of the regulation of sources of law. In addition, it has examined the idea of custom’s inherent features, which make it especially suitable to create general rules of universal scope. Furthermore, it has provided an overview of state practice invoking ICJ Statute Article 38(1) in contexts other than proceedings before the ICJ. It has further shown that there is evidence that this general practice is accepted as law. In addition, it has argued in favour of reaffirming the centrality of state practice itself, including in the form of decisions of domestic courts, as opposed to the typical exclusive reliance on decisions of international courts and tribunals with respect to the regulation of sources of law.

Section 3 has provided an overview of some accounts of CIL on sources of law and has proposed to conceptualise these accounts in terms of two models, along the lines of Bentham’s distinction between custom in foro and in pays, as revisited in recent general jurisprudence literature. In this vein, it has discussed in detail Lamond’s recent critique of the rule of recognition which, relying on the distinction between custom in foro and custom in pays, argues in favour of treating the rule of recognition as a rule of customary law having the same properties as any legal rule of the legal system. Section 3 has argued that, to the extent that this critique is apposite to international law, the insights it contains shed light on the nature of a CIL regulating the sources of law in international law.

To conclude, the chapter suggests that the persistence of custom in foro as a model for CIL regulation of the sources of law, a matter calling for further research, may detract from the potential for establishing the existence of a CIL regulating the sources of international law on the solid grounds of proper state general practice and acceptance as law thereof.