A. The Comparative Methodology

The discussions on GPL in the Advisory Committee revolved around the origins of this source, among other issues. As Verdross pointed out in 1935, the origins of GPL were controversial, and had not even been clarified by the Institut de Droit International.Footnote ²² While a natural law approach sees GPL as deriving from legal logic or a natural sense of justice, a positivist approach sees it as deriving from the law of national legal systems.Footnote ²³ Article 38 does not, in itself, lean towards one or the other; its wording seems to confirm that GPL must derive from national legal systems, but at the same time, the provision does not determine whether it must be recognized by domestic legislation of civilized nations (they can also be principles of ‘legal conscience’ recognized by the domestic scholarship, for example).Footnote ²⁴

Special Rapporteur Vázquez-Bermúdez considers that, while the practice of States and international jurisprudence ‘admittedly present some divergences …, an overall approach can be drawn from them: the comparative analysis for the purposes of determining the existence of a [GPL] must be wide and representative, covering different legal families and the various regions of the world’.Footnote ²⁵ He also considers that the position that GPL should exist ‘within a sufficiently large number of States’ is ‘generally accepted in the literature and … supported by practice’.Footnote ²⁶ Accordingly, Draft Conclusion 5 follows the positivist approach to GPL:

Thus, to ascertain the existence and content of a GPL, adjudicators should carry out a comparative analysis. This examination consists of an assessment of domestic legislation and national court decisions, and the analysis must ‘be wide and representative, including different regions of the world’.

While there was broad agreement on a first step to ‘determine the existence of a principle common to the various legal systems of the world’, the requirement of a ‘comparative analysis’ to pursue the first step was much debated in the Commission's 2021 meeting. Some Commission members considered that this approach would be ‘too strict’ and that ‘in practice such comparative analysis was not always wide and representative’.Footnote ²⁸

Some scholars have long considered that general principles should reflect the municipal systems of the world and that they should be representative of different legal systems.Footnote ²⁹ In 1927, Lauterpacht considered that ‘Only general principles of private law recognized by the main systems of jurisprudence, ascertained by comparative study … are a suitable object of analogy’.Footnote ³⁰ Moreover, decisions having recourse to domestic legal systems and Roman law to ascertain the existence of GPL can be traced back to practice even prior to the establishment of the PCIJ.Footnote ³¹ Still, the need for a comparative study was not systemically emphasized, and the natural law approach to GPL shared the stage with the view that general principles should reflect national legal systems.

Indeed, the practice of the PCIJ and the ICJ does not consistently reflect the comparative approach.Footnote ³² The World Court's early case law determined the existence of a GPL broadly and without methodological rigour. In the Polish Upper Silesia case, Poland had advanced an objection to the case's admissibility grounded on a legal argument similar to that of lis pendens. However, to make this claim, Poland did not undertake a comparative study of different legal systems. In turn, the Court dismissed the claim, considering that the requirements for lis pendens had not been met, but did not determine whether the concept reflected a GPL.Footnote ³³

The PCIJ's decision in Factory at Chorzów also illustrates this point. Some authors submit that the Court's reference to principles of reparation reflects an instance of the PCIJ's reliance on GPL.Footnote ³⁴ In that case, the Court considered that ‘As regards the first point, the Court observes that it is a principle of international law, and even a general conception of law, that any breach of an engagement involves an obligation to make reparation’.Footnote ³⁵ The expression ‘general conception of law’ denotes the idea of a general principle of legal logic and can assist in determining whether reparation is a GPL. However, the Court does not offer an assessment of domestic legal systems to buttress this.

In a rare direct reference to Article 38(1)(c) in the PCIJ's case law, Judge Anzilotti referred to res judicata as a GPL. He grounded this merely by saying that ‘[n]ot without reason was the binding effect of res judicata expressly mentioned by the Committee of Jurists entrusted with the preparation of a plan for the establishment of a Permanent Court of International Justice, amongst the principles included in the above-mentioned article’.Footnote ³⁶ Here also, no reference to domestic legal systems is made.

Similarly, disputants and judges in the ICJ do not systemically resort to a comparative approach. In the Corfu Channel case, the first case determined by the ICJ, the Court grounded its assessment on the applicability of the procedural principle of ‘circumstantial evidence’ on the fact that ‘[t]his indirect evidence is admitted in all systems of law, and its use is recognized by international decisions’.Footnote ³⁷ Equally, in their submissions United Kingdom and Albania did not seek to determine whether there was a general principle of ‘indirect evidence’ based on a comparative approach.Footnote ³⁸

The Right of Passage over Indian Territory case is a noteworthy exception to this trend, and arguably a turning point in the approach to the identification of GPL. The dispute is often given as the primary illustration of the use of comparative methodology to ascertain GPL.Footnote ³⁹ In that dispute, Portugal advanced the existence of ‘the principle to be inferred from domestic legislations, which authorize a right of access for the benefit of owners of enclaved property’.Footnote ⁴⁰ India argued that none of the principles invoked by Portugal were general principles of law in the sense of Article 38(1)(c), since these ‘… are to be understood as comprising the principles generally adopted by civilized States within their domestic law’.Footnote ⁴¹ Portugal reacted to this by commissioning a comparative study on the existence of a right of access to enclaves. The claim evoked the ‘analogous situation’ of ‘a piece of land which has no access to the public highway and which cannot be reached in any way other than by passing over the land or lands belonging to some other person or persons’:Footnote ⁴²

The study presented by Portugal compares civil law, common law, ‘communist and socialist’ traditions and Islamic countries, amounting to 64 jurisdictions. For this examination, Portugal commissioned Prof. Max Rheinstein, a scholar from the University of Chicago specializing in comparative law.Footnote ⁴⁴

The Court did not take a position on whether the right of access advanced by Portugal reflected a GPL under Article 38. The Court found that the concept at issue corresponded to a ‘practice well understood by the parties’,Footnote ⁴⁵ and hence it did not ‘… consider it necessary to examine whether general international custom or the general principles of law recognized by civilized nations may lead to the same result’.Footnote ⁴⁶

Unlike the Court's decision, some of the judges’ separate and dissenting opinions addressed the legal nature of the right of access to enclaved properties. Some of these opinions associated the GPL at issue, sometimes interchangeably, with expressions such as ‘a reason deeply rooted in the legal consciousness of all peoples’Footnote ⁴⁷ and a ‘principle of justice founded on reason’.Footnote ⁴⁸ This is noteworthy since, even in the face of an exhaustive comparative study, some did not associate the existence of a GPL with the results of comparative methodology.

It seems that after the Right of Passage decision references to comparative law as a methodology to ascertain GPL started to feature more prominently, though mostly in separate and dissenting opinions rather than in the judgments of the Court.Footnote ⁴⁹ In the context of State submissions, subsequent direct and indirect references to comparative methodology for ascertaining a GPL were not always accompanied by an extensive analysis of municipal legal systems of the sort presented by Portugal in its submissions in the Right of Passage case. For example, in the North Sea Continental Shelf case, the parties disputed whether a ‘just and equitable share’ was a general principle of law. Germany submitted that goods held in common by several parties are to be ‘meted out in accordance with an appropriate standard equally applicable to all of them’.Footnote ⁵⁰ It argued that this principle ‘is a basic legal principle emanating from the concept of distributive justice and a generally recognized principle inherent in all legal systems, including the legal system of the international community’. Yet Germany did not offer a comparative study of different national legal systems. Rather, it argued that the principle ‘ranks among those general principles of law which might be argued as having such an inherent, self-evident, and necessary validity’, a natural law argument.Footnote ⁵¹ In response, Denmark and the Netherlands did not contest Germany's methodology for advancing the existence of a general principle of just and equitable shares.

More recent State submissions before the Court increasingly incorporate comparative studies to support the existence of a GPL under Article 38(1)(c).Footnote ⁵² One notable example is Nauru's memorial in Certain Phosphate Lands in Nauru. This includes an appendix featuring a comparative study specially commissioned to support Nauru's argument that the concept of trusteeship reflects a general principle of law, similar to the study commissioned by Portugal in the Right of Passage case.Footnote ⁵³

Yet this trend is not absolute. In the more recent Obligation to Negotiate Access to the Pacific Ocean case, Bolivia drew on the notions of good faith, estoppel and legitimate expectations to claim that Chile had ‘repeatedly and formally committed itself’ to an obligation to negotiate access to the sea.Footnote ⁵⁴ It did not specify whether estoppel and legitimate expectations reflected general principles of law, but in some passages it did advance the claim that legitimate expectations was a GPL. Bolivia did not, however, include any comparative study to support its assertion. Chile challenged Bolivia's claims concerning these principles on several grounds and argued that the applicant's arguments were insufficient to demonstrate that legitimate expectations was a self-standing principle of international law.Footnote ⁵⁵ There was no disagreement specifically on whether legitimate expectations and estoppel reflected GPL under Article 38(1)(c); the disagreement mainly concerned whether the concepts were applicable to the case. Accordingly, the Court did not devote much attention to the issue. The Court concluded that the conditions for establishing an estoppel were not fulfilled and it dismissed Bolivia's claim concerning legitimate expectations in one sentence. The Court found that references to investor-State arbitral awards were not sufficient to demonstrate the existence ‘… in general international law of a principle that would give rise to an obligation on the basis of what could be considered a legitimate expectation’.Footnote ⁵⁶

This overview of the practice of and before the World Court leads to the following conclusions. First, the need for and the resort to a comparative study by States as a means to ascertain a GPL is not found in early case law. This trend seemingly shifted after the Right of Passage case, becoming more common, albeit still in a restricted number of cases, in more recent decisions (ie, in the past three decades). Second, the methodology employed to ascertain GPL may vary according to the importance of the principle to the case at issue. For example, in the Right of Passage case, the existence of a GPL on the right of access between enclaved territories was one of the main arguments advanced by Portugal as applicable law. It was not just a procedural matter, but rather a substantive one. Portugal's emphasis on the principle prompted India to challenge its status as a GPL; equally, India's vigorous rejection of this led to Portugal producing its detailed comparative study on national legal systems. In the Obligation to Negotiate Access to the Pacific Ocean case, even though estoppel and legitimate expectations were central to Bolivia's arguments, there was no express disagreement concerning these being GPL under Article 38(1)(c). Perhaps because of this, neither the parties nor the Court addressed the two-step approach in detail. However, had Bolivia demonstrated that legitimate expectations reflected a principle on the basis of the two-step methodology, one may wonder if the Court would have addressed the claim more thoroughly in order to determine whether a principle of legitimate expectations really did exist in general international law.

The Special Rapporteur's Second Rapport abounds in references to international decisions stressing that a given principle must be widely recognized in domestic legal systems.Footnote ⁵⁷ This notwithstanding, the comparative approach is not the unequivocal methodology used to ascertain GPL in the PCIJ and ICJ's case law. At the same time, more recent State practice before the Court indicates the need for a comprehensive and representative comparative methodology to legitimize claims of a GPL. States may challenge the existence of a GPL when this is lacking. This type of counterargument suggests there is an expectation that this methodology must be followed, even if the Court itself does not necessarily abide by it. Such an expectation to legitimize the invocation of a GPL through a wide and representative comparative methodology may be considered to at the very least to reflect a trend in State practice and opinio juris towards the methodology proposed by Draft Conclusion 5.

B. Transposability to the International Legal System

Once it is determined that a general principle is common to the principal legal systems of the world, the next step is to ascertain whether it is also ‘applicable within the international legal system’,Footnote ⁵⁸ bearing in mind the structural differences between these two spheres of law. The current version of Draft Conclusion 6 states:

Draft Conclusion 6 thus sets forth two cumulative sub-conditions: the first of ‘compatibility’ and the second of ‘applicability’. However, for clarity, they will be referred to here as aspects of the general requirement of transposability.Footnote ⁶⁰

Like the first step, the requirement of transposability is not self-evident from the wording of Article 38(1)(c). Moreover, like the comparative approach that informs the first step, the second step only makes sense if GPL are not considered to be a manifestation of natural law. Indeed, if a given principle is grounded upon certain ‘notions of fairness and justice’ which impart to it a natural law validity, the principle in question should be equally applicable to domestic and international spheres. Conversely, transposing notions from the domestic to the international sphere on a positivist basis almost by definition requires an assessment of transposability.

In the first case submitted to the PCIJ, the SS Wimbledon dispute, the applicants invoked the concept of servitude in international law. The Court considered that it ‘[was] not called upon to take a definite attitude with regard to the question, which is moreover of a very controversial nature, whether in the domain of international law, there really exist servitudes analogous to the servitudes of private law’.Footnote ⁶¹ Circumventing the applicability of the concept of servitude to the case in hand, the Court looked to the wording of Article 380 of the Treaty of VersaillesFootnote ⁶² concluding that ‘[it] fe[lt] obliged to stop at the point where the so-called restrictive interpretation would be contrary to the plain terms of the article and would destroy what has been clearly granted’.Footnote ⁶³

The PCIJ's 1925 judgment in Polish Upper Silesia tackles a similar problem in a similar way. As explained above, Poland had invoked the concept of litispendence, and the Court considered that

The Court concluded that there were not two identical disputes pending before international tribunals and that the parties were not the same; therefore, there was no litispendence, even if this concept was applicable in international law.

The Court did not explicitly address the transposition of a GPL to the international legal order in either dispute, it simply dismissed the applicability of the domestic law concept. Still, the Court's reasoning in both judgments offers an embryonic form of the transposability stage of the test, as it queried whether domestic law concepts were applicable to international law.

The requirement of transposability takes an explicit form in the famous dictum by Lord McNair in his Separate Opinion in the South West Africa Advisory Opinion. Lord McNair examined the legal nature of the League of Nations’ mandate system, ‘based on the analogy of the contract of mandate in private law’.Footnote ⁶⁵ He considered:

It is necessary to refer to this in full: it is the textbook quotation explaining the requirement of transposability in the context of GPL.Footnote ⁶⁷

After the South West Africa Advisory Opinion, references to the second step become more frequent and explicit in the ICJ case law. In the South West Africa cases, South Africa contested that the concept of non-discrimination advanced by the applicants as was a GPL under Article 38(1)(c). South Africa argued that GPL ‘are taken from the realm of municipal law, they are elevated by analogy from that law into international law relationships and applied there …’.Footnote ⁶⁸ The Court ‘disposed of the case on different grounds’,Footnote ⁶⁹ and did not examine the merits of these arguments concerning non-discrimination. However, the Court did examine the notion of actio popularis, finding that ‘although a right of this kind may be known to certain municipal systems of law, it is not known to international law as it stands at present’.Footnote ⁷⁰

In the North Sea Continental Shelf case, Denmark and the Netherlands contested Germany's invocation of a general principle of a ‘just and equitable share’ on the basis that it conflicted with the ‘whole legal approach to the determination of boundaries in international law’.Footnote ⁷¹ Denmark and the Netherlands argued that it violated the equality of States and also sought to demonstrate that State practice did not support the existence of a just and equitable share when delimiting boundaries of coastal States.Footnote ⁷²

In Certain Phosphate Lands in Nauru, Australia claimed that the notion of ‘trust’ in domestic law would ‘mistake completely the fundamental elements of the United Nations Trusteeship System’. Additionally, Australia referred to McNair's dicta pointing to ‘the inappropriateness of seeking to apply private law institutions directly to an international institution’.Footnote ⁷³

The exercise of determining the compatibility of a domestic GPL with the underpinnings of the international legal order became more frequent after the South West Africa Advisory Opinion. Still, this did not seem to reflect transposability as a step, that is to say as a part of a structured methodology to ascertain a GPL applicable under Article 38(1)(c). Liechtenstein's Memorial in the Certain Property case is perhaps the clearest example of recourse to the two-step methodology, and the condition of transposability in particular.Footnote ⁷⁴

Liechtenstein, the claimant, explicitly invoked the two-step methodology to demonstrate the applicability of the principle of unjust enrichment. It argued that a ‘rule must be considered as a general principle of law (i) if it is applied in the main systems of municipal law and (ii) if it is “transposable” in international law, ie, it is not inconsistent with any general principle of or applicable rule of public international law’.Footnote ⁷⁵ After arguing that the unjust enrichment ‘is a foundational principle underlying restitution or compensation in numerous domestic legal systems’,Footnote ⁷⁶ it claimed that ‘the principle is received at the international level’.Footnote ⁷⁷ Liechtenstein invoked Lord McNair's opinion in the South West Africa Advisory Opinion before arguing that there were no incompatibilities between the principle and general international law.Footnote ⁷⁸ The Court's decision that it lacked jurisdiction meant that it did not consider these arguments.

Other references to the requirement of transposition can be found in Judge Tanaka's Dissenting Opinion in South West Africa cases,Footnote ⁷⁹ Judge Simma's Separate Opinion in the Interim Accord caseFootnote ⁸⁰ and (implicitly) in Judge Greenwood's Separate Opinion in the Delimitation between Nicaragua and Colombia.Footnote ⁸¹

The transposability condition is also often accompanied by McNair's dictum rather than an inference from Article 38(1)(c) or even presented as a logical inference from the concept of GPL deriving from domestic legal systems). The nudge to consolidate transposability as a step/condition for ascertaining the existence and applicability of a GPL thus seems to have been derived from, or at least solidified by, the South West Africa Advisory Opinion.

The Special Rapporteur's references to ‘evidence confirming transposition’ as a separate step when ascertaining GPL are less abundant and more ambiguous than those relating to the first step.Footnote ⁸² There are few instances in the Court's jurisprudence which expressly highlight the requirement of transposability, and State practice before the ICJ does not seem consistent enough to suggest that this is established. Similarly, the ICJ references described above are not unanimous on how to determine transposability, and this requirement is more often referred to in separate opinions. Even more scarce are references to the twofold analysis within the transposability step as set out in Draft Conclusion 6. Therefore, as far as PCIJ and ICJ practice is concerned, the procedure described by Draft Conclusion 6 and commentaries appears to be an example of progressive development.

A. The Two Steps in International Legal Scholarship

The first step, ie the need to investigate national legislation in order to ascertain GPL, can be found in early scholarly works.Footnote ⁹⁰ The idea of using analogy to derive concepts of international law from domestic legal systems features in scholarly works preceding the PCIJ Statute.Footnote ⁹¹ However, the use of analogy varied, and scholars did not address the scope of ‘comparative’ methodology in detail. The comparative approach to determining the existence of a GPL in the sense of Article 38(1)(c) emerged in scholarly works in the second half of the twentieth century. One of the first significant contributions to the idea that comparative law as a structured methodology should be used for this purpose is Schlesinger's 1957 proposal concerning ‘Research on the General Principles of Law Recognized by Civilized Nations’.Footnote ⁹²

Yet the trend was not uniform. Rosenne's The Law and Practice of the International Court, to this day one of the main commentaries on the case law of the ICJ, considered in its 1965 edition that the Court's references to GPL ‘… show that the “general principles of law recognized by civilized nations” are not so much generalizations reached by application of comparative law … , as particularisations of a common underlying sense of what is just in the circumstances’.Footnote ⁹³ Rosenne therefore viewed GPL with a hint of a natural law perspective.Footnote ⁹⁴ Moreover, the idea that a comparative analysis should be ‘wide and representative’ is also a recent development that arguably is still not implemented in practice.Footnote ⁹⁵

Conversely, scholarship seldom identified the requirement of transposability before the South West Africa Advisory Opinion.Footnote ⁹⁶ Prior to the 1950s, some scholars did refer to the different features of domestic and international law and the need to acknowledge these differences. Yet even those works did not view transposability as a separate step in the process of identification of GPL, and the scope of this requirement was not clearly delineated.Footnote ⁹⁷ For example, in his 1927 Private Law Sources and Analogies of International Law, Lauterpacht wrote that ‘… of all attempts to apply to relations between States conceptions taken from private law, none has caused more confusion or has brought the recourse to analogy into more disrepute than the efforts made to introduce the conception of servitudes into international public law’.Footnote ⁹⁸ Still, when discussing the issue of transposition, he did not expressly mention the need to consider the transposability of the private law principle of servitude into the international sphere. Rather, Lauterpacht considered that ‘… it should be resorted to only when there is no doubt that the parties intended it to be a permanent relation independently of who is the sovereign of the entitled or encumbered territory’.Footnote ⁹⁹ Lauterpacht's reasoning focused on the need to consider the parties’ will. He thus seemed to accept the existence of a general principle of private law generally recognized by the majority of States, but which should be applied by means of an in dubio mitius approach (similar to the Court's reasoning in SS Wimbledon).Footnote ¹⁰⁰

In this context, the South West Africa Advisory Opinion appears to have caused a ripple effect not only in the Court's jurisprudence, as described above, but also in international legal scholarship. Rosenne's 1957 The International Court of Justice refers to McNair's dicta, noting the ‘dangers of drawing hasty analogies’ from municipal law into international law.Footnote ¹⁰¹ Schwarzenberger's 1957 International Law goes as far as to imply that the requirement of transposability was a development brought about through the practice of the ICJ: ‘Although it does not follow from the text of Article 38(1)(c) of the Statute, a fourth condition of the applicability of a general principle of law on the international level has to be considered’. Schwarzenberger then quotes McNair's dicta, and explains that

After the South West Africa Advisory Opinion, the two requirements then progressively mutate from scant, unsystematized references in the literature to a systematized two-step methodology.Footnote ¹⁰³ It is interesting, for example, to consider the slight shift from de Visscher's 1953 Théories et réalités en droit international public to his 1960 third edition of the same monograph. While the former simply refers to ‘procedure of abstraction’,Footnote ¹⁰⁴ the latter explicitly mentions a two-step process.Footnote ¹⁰⁵ The 1960 Dictionnaire de la Terminologie du droit international considered that dominant scholarship viewed GPL as propositions ‘enshrined in the internal law of civilized nations insofar as they are transposable to the international order’.Footnote ¹⁰⁶ In more recent scholarship, important contemporary handbooks and texts on GPL refer to the two-step methodology as a given,Footnote ¹⁰⁷ even if the natural law versus positivist divide is still acknowledged.Footnote ¹⁰⁸

In 2018, the ILA issued a report on ‘The use of domestic law principles in the development of international law’.Footnote ¹⁰⁹ Following an analysis of different international jurisdictions and international law bodies (such as the ILC), the ILA concluded that to identify a GPL, the ‘comparative method could be used’ and ‘the identified general principle should fit into the international context and be able to address the specific legal problem at international level’.Footnote ¹¹⁰ Interestingly, the cases described in the ILA report do not strictly follow this methodology.

There also seems to be a prevalence of a positivist approach to GPL in contemporary scholarship. Some recent works go beyond the procedure of a comparative study, denouncing a ‘Westernized approach’ to the source in international courts and tribunals and arguing for a more significant role to be given to different domestic legal systems in the process of ascertaining GPL.Footnote ¹¹¹ The overall trend reaffirms the two-step approach. The ILA and ILC then reinforce this trend by adopting the same methodology.

B. Remarks on Scholarship Concerning The Two-Step Methodology

This analysis supports the conclusion reached in Section II that the two-step methodology was not initially viewed as the undisputed method for the identification of GPL in the early years of the Statute of the PCIJ. Instead, this approach evolved over time. Bearing in mind the caveats previously mentioned,Footnote ¹¹² the evolution of international legal scholarship leads to two conclusions relating to each of the two steps.

First, the idea that GPL should be derived from a comparative analysis of national legal systems was present from the early years of the PCIJ. However, this did not require an examination of a broad sample of domestic legal systems or to be representative. This is hardly surprising, as natural law approaches to GPL were dominant at the beginning of the last century. Even when not associated with natural law, GPL would often be defined as ‘certain broad rules and maxims, common to the good sense and conscience of civilized mankind, which are found in all great systems of law’ by some authors.Footnote ¹¹³ It seems to be only around the second half of the twentieth century that more refined considerations of how to conduct the comparative analysis started to emerge. This trend seems to have been pushed by comparativists more than by international legal scholars.Footnote ¹¹⁴ The timing coincides with the Right of Passage case, examined in Section II.A, Portugal's submission remaining to this day a unique example of an exhaustive study of domestic legal systems in support of an alleged GPL.

Secondly, international legal scholarship rarely acknowledged the requirement of transposability until Judge McNair's Separate Opinion in the South West Africa Advisory Opinion. After the publication of that dictum, not only did transposability gain more space in commentaries to Article 38(1)(c) (often referring to the ‘lock, stock and barrel’ formula), but it was also increasingly acknowledged as a formal requirement for the determination of a GPL deriving from domestic legal systems. The consolidation of the two-step methodology, as such, seems to be a very recent phenomenon, subsequent to the maturing of the requirement of transposability.