The Uses of the Outputs of the International Law Commission in International Adjudication: Subsidiary Means or Artefacts of Rules?

Abstract

This paper examines the methods which international courts and tribunals (ICTs) employ when using ILC outputs for the purpose of determining rules of international law and their content. Specifically, it identifies common patterns in the ways in which ICTs, first, justify their reliance on ILC outputs and, second, deal with their ambiguities. The paper argues in favour of a consistent methodology for the treatment of ILC outputs in international adjudication. Such a framework is based on the distinction between the identification of the status of a normative proposition contained in these texts and the determination of its content or its interpretation. The identification of the status of a normative proposition requires a critical assessment and reconstruction of the evidence leading up to its development taking also into account that these instruments are not a monolith from the perspective of sources. However, the interpretation of a proposition whose status is uncontested follows a line of inquiry akin to treaty interpretation. This observation has broader implications for the process of interpretation in international law. Specifically, apart from the context of treaty interpretation, international courts or tribunals interpret the normative propositions contained in ILC outputs as a methodological shortcut for the interpretation of rules of customary international law or general principles of law. Conversely, the employment of methods akin to treaty interpretation in this context can constitute evidence of the emergence of common rules, principles, or good practices of interpretation that are also applicable to unwritten international law.

1 Introduction

In a recent separate opinion, Judge Tomka expressed his disagreement with the drafting choices of the majority of his colleagues:

The Court occasionally refers to ‘breaches of the Convention’, ‘breaches of Articles’ or ‘violat[ions] of a number of provisions of the ICSFT and CERD’ […] It is rather regrettable that the principal judicial organ of the United Nations does not pay sufficient attention to the precision of the language it uses. Under international law, for an act of a State to be wrongful, such act, consisting of an action or omission, must both be attributable to the State and constitute a breach of an international obligation of the State (Article [2] of ARSIWA).¹

The uncompromising tone of the criticism illustrates a broader paradox in international adjudication. Outputs of the International Law Commission (‘ILC’), such as the Articles on State Responsibility (‘ARSIWA’),² have no binding effect as such.³ And yet, international courts and tribunals (‘ICTs’) refer to ILC outputs in a remarkable number of decisions, so that even linguistic deviations from their text occasionally give rise to censure. To put this into perspective, according to a 2017 report of the United Nations (‘UN’) Secretariat, the aggregate number of references to one ILC output––ARSIWA––in decisions, individual opinions of judges, and the submissions of parties before various international courts, tribunals, and other treaty bodies approached, at the time, 1400.⁴ The International Court of Justice (‘ICJ’) alone has relied on ILC outputs in at least 25 judgments, whereas more than 70 individual opinions cite various ILC outputs.⁵

The apparent discrepancy between the lack of a ‘formal’ status for the outputs of the ILC and their effective ‘authority’ calls for further reflection.⁶ In principle, ILC outputs do not constitute ‘formal’ sources of international law.⁷ The ILC is no law-making body, but a body of legal experts.⁸ Its mission as a subsidiary body of the UN General Assembly is ‘to promote the progressive development of international law and its codification’ by practically feeding its parent body technical reports and recommendations on legal issues.⁹ In principle, outputs of the ILC largely fall into the category of ‘teachings of the most highly qualified publicists’ on which international courts and tribunal may rely ‘as subsidiary means for the determination of rules of law’.¹⁰ The underlying consideration is that states cannot accept rules which are ‘the result of the doctrine rather than of their own will, or of their usages’.¹¹ Yet, ICTs rarely refer to this category when discussing ILC outputs.¹² In the context of international adjudication, the outputs of the ILC operate as ‘material’ sources of international law, that is, as depictions of the substantive content of an applicable rule of law.¹³ Yet, there are multiple ways to justify the use of ILC outputs in international adjudication, each justification having different implications as to the methodology employed when using them.

Besides, another factor complicating the role of ILC outputs in the context of law determination is that they come in quite diverse shapes and forms. On the one hand, final outputs of the ILC may form the basis for the negotiation of a treaty, but they may also take other forms such as published reports or annexes to UN General Assembly resolutions.¹⁴ In fact, it has been almost two decades since an ILC output led to the adoption of a treaty.¹⁵ What is more, the ILC has adopted over time a variable nomenclature for its final outputs: ‘draft articles’,¹⁶ ‘draft principles’,¹⁷ ‘draft guidelines’,¹⁸ ‘reports’,¹⁹ ‘model rules’,²⁰ ‘draft declarations’,²¹ ‘resolutions’,²² ‘draft conclusions’,²³ ‘draft conventions’,²⁴ or ‘draft codes’.²⁵ Whilst the choice of form can constitute an indication of the ILC’s intention regarding the status of its outputs or their content, the question arises whether and how this diversity impacts the decisions of ICTs.²⁶ On the other hand, the consideration of a topic by the ILC is a complex process. In this (often long) process, a multitude of materials are produced: reports of drafting committees, comments by governments, interim versions of articles or commentaries adopted by the plenary, the summary record of discussions, the reports of the special rapporteur, or even reference materials compiled by the UN Secretariat.²⁷ A combined reading of these materials often reveals ‘titanic disagreements’ which are imprinted in carefully articulated final articles and commentaries.²⁸ The traditional label of ‘teachings’ provides little guidance as to how to navigate through all these materials in determining applicable rules of law, as it treats most of these materials indistinctly.²⁹ In this respect, it is important to examine whether the normative propositions contained in these outputs lose their practical value altogether in case of ambiguity or whether a methodology exists for resolving interpretative issues arising from them.

This paper examines the methods which ICTs employ when using ILC outputs for the purpose of determining the existence of rules of international law and their content. As a preliminary note, whereas some ICTs––notably, the ICJ––might have a more revered position within the international legal profession,³⁰ the premise of this study is that all decisions of ICTs have in principle the same formal status in the process of the determination of rules of international law.³¹ As to its methodology, the paper draws from publicly available decisions of the ICJ, the International Tribunal for the Law of the Sea (‘ITLOS’), the International Criminal Court (ICC) and ad hoc international and hybrid criminal tribunals, dispute settlement bodies of the World Trade Organization (‘WTO’), regional human rights courts, and inter-state and investment arbitral tribunals which make explicit reference to outputs of the ILC.³² The paper does not engage in a quantitative analysis,³³ but instead qualitatively categorises and synthesises the reasoning of decisions of these ICTs on the basis of illustrative examples along two axes. In particular, the paper focuses on two questions: (i) how do ICTs justify their reliance on ILC outputs; and (ii) how do they deal with their ambiguities?

The analysis is structured as follows. Section 2 focuses on the ways in which ICTs use ILC outputs within the process of treaty interpretation. Specifically, it argues that ICTs use the rules of treaty interpretation not only as the legal basis that justifies resorting to ILC outputs, but also as the roadmap on how to use them. Section 3 lays out the ways in which outputs of the ILC are relevant in international adjudication for the identification of unwritten international law, namely, customary international law (Sect. 3.1) and general principles of law (Sect. 3.2). This section records the wide range of justifications on the basis of illustrative examples drawn from a wide variety of adjudicative bodies, but also tries to fill the gaps between existing practice and the theory of sources of international law. Apart from why ICTs use ILC outputs in the process of the determination of unwritten international law, another important question is how they make use of them. Section 4 turns to the methods which ICTs employ to determine the content of rules of unwritten international law on the basis of ILC outputs. Specifically, it maps out common patterns in the use of ILC outputs in international adjudication and attempts to highlight their broader systemic implications by reference to instructive examples.

The paper draws from a framework based on the distinction between the identification of the status of a normative proposition contained in these texts and the determination of its content or its interpretation. The identification of the status of a normative proposition requires a critical assessment and reconstruction of the evidence leading up to its development taking also into account that these instruments are not a monolith from the perspective of sources. However, the interpretation of a proposition whose status has already been established follows a line of inquiry akin to treaty interpretation. The observation that this framework is largely confirmed in the practice relating to the use of ILC outputs in international adjudication has broader implications for the process of interpretation in international law. Specifically, leaving treaty interpretation aside, ICTs interpret the normative propositions contained in ILC outputs as a methodological shortcut for the interpretation of rules of customary international law or general principles of law. Conversely, the employment of methods akin to treaty interpretation in this context can constitute evidence of the emergence of common rules, principles, or at least good practices of interpretation that are applicable to unwritten international law.

2 The Outputs of the ILC as a Means of Treaty Interpretation

ICTs overwhelmingly apply the rules of interpretation laid down in the VCLT either qua treaty rules or as articulations of customary international law.³⁴ However, they are not always straightforward as to where exactly ILC outputs fit within the process envisaged in these rules.³⁵ By design, the ILC has a preparatory role in multilateral treaty-making within the UN system. Apart from this function, the ILC may render its interpretation on the content of treaty rules in a variety of instruments regardless of their final form.³⁶ This diversity impacts the ways in which ICTs make use of the outputs of the ILC. This section starts with an examination of the role of ILC outputs in the interpretation of treaties which originate from such outputs. It then turns to use of ILC outputs in which the ILC interprets a treaty provision incidentally in considering any topic on its agenda. As will be shown, the role of ILC outputs in the process of treaty interpretation is more complex than the category of ‘teachings’ might suggest. Rather, the reasons for and the ways of using ILC outputs in this process may vary depending on the treaty and the specific ILC output in question.

According to its Statute, after the conclusion of its work on a given topic, the ILC may recommend to the General Assembly the calling of an international conference with a view to concluding a multilateral treaty.³⁷ Historically, the ILC has laid the groundwork for several treaties in fields such as the law of treaties and diplomatic and consular relations.³⁸ Obviously, if a treaty is concluded through this process, the source of the obligation is the treaty not the ILC draft. The output of the ILC can be relevant in an indirect way as a means for the interpretation of the treaty. Notably, the ILC outputs can be used together with the diplomatic record as a supplementary means for the interpretation of a treaty specifically as ‘preparatory work of a second order’.³⁹ For instance, in the Jadhav case, the ICJ was faced with the question of whether the obligation of a receiving state under the VCCR to inform and allow access to the consular authorities of a sending state in the case of the arrest or detention of its nationals applied in cases of suspected espionage.⁴⁰ The Court, applying the customary rules of interpretation as reflected in Articles 31 and 32 VCLT, found that no relevant exception could be inferred from the ordinary meaning of the terms of the provision and the object and purpose of the treaty.⁴¹ It then took note of the ILC’s decision not to make any exception for cases of espionage in its own draft of the convention as evidenced by the fact that a member of the Commission raised the issue in the plenary discussion but any reference to espionage was omitted in the final commentaries.⁴² The Court explicitly found that its examination of the ILC output was ex abundanti cautela, since ‘the Court need not, in principle, resort to supplementary means of interpretation’ when the text is sufficiently clear.⁴³ In this process, the ILC outputs could still be considered as ‘teachings’, as they do not form, strictly speaking, part of the negotiation of the treaty or emanate directly from the negotiating states.⁴⁴ Yet, ICTs often explicitly designate final ILC outputs as part of the ‘preparatory works’ or ‘travaux préparatoires’ of the treaty.⁴⁵ Either way, the ILC outputs still operate subsidiarily, since the interpreter is not bound to resort to them, unless the meaning of the text remains ambiguous, obscure, or manifestly absurd after the application of the interpretative means of Article 31 VCLT.⁴⁶

That said, the role of ILC outputs in the context of the interpretation of treaties that are developed with the input of the ILC is not always subalternate to other interpretative materials. To illustrate this point, in the Jurisdictional Immunities case, the ICJ dealt tangentially with the interpretation of a treaty provision pertaining to the so-called ‘territorial tort’ exception to state immunity from proceedings in foreign courts.⁴⁷ The Court took particular note of the ILC’s Commentary to the corresponding provision of the draft convention according to which the provision did not apply to ‘situations of armed conflict’.⁴⁸ The Court emphasised that ‘[n]o state questioned this interpretation’ in the ensuing negotiations and noted the fact that some states parties appended similarly worded declarations upon the ratification of the treaty.⁴⁹ In the end, the Court sided with the interpretation provided by the ILC, notwithstanding the fact that the text of the convention in question—much like the VCCR in the Jadhav case—did not provide for such a qualification to the ‘territorial tort’ exception.⁵⁰ Similarly, several ICTs have relied solely on the ILC’s Commentary to its Draft Articles on the Law of Treaties to define key terms in Articles 31 and 32 VCLT like the notion of ‘subsequent agreements’.⁵¹ In most cases, there is no explicit justification in the decision for according such weight to ILC outputs in the process of treaty interpretation.

Yet, there are reasons to believe that such practice is not entirely extraneous to the customary rules of treaty interpretation. Importantly, unlike the record of negotiations, which often comprises contradictory positions of individual states or groupings of states, the ILC final draft reflects the position of an impartial deliberative body which remains constant throughout the negotiation.⁵² As a result, it could be argued that an interpreter should accord more weight to the views of the ILC than the record of negotiations if the parties made no substantial changes to the ILC draft when adopting the treaty.⁵³ Specifically, according to Article 31(2)(a) VCLT, the ‘context’ of the treaty includes any ‘agreement of the parties in relation to the treaty which was made in connection with the conclusion of the treaty’. Typical examples of such agreements include formal acts temporally coinciding with the conclusion of the treaty like final acts of diplomatic conferences.⁵⁴ However, the provision does not seem to require that such agreements are in any particular form or even that they are explicit.⁵⁵ Rather, the determination as to whether the parties have reached an agreement concerning the interpretation of a treaty seems to be ‘a question of fact’.⁵⁶ So, for instance, in Maritime Dispute, the ICJ did not preclude as a matter of principle that the minutes of diplomatic discussions could evidence the existence of an agreement under Article 31(2)(a) VCLT, even if no such agreement existed in the facts of the case.⁵⁷ More pertinently, according to a WTO panel, ‘uncontested interpretations given at a conference, e.g., by a chairman of a drafting committee, may constitute an “agreement” forming part of the “context”’.⁵⁸ This line of reasoning can extend possibly to interpretations given by the ILC in its final output to the extent that they remained uncontested during the negotiation and conclusion of the treaty. Along similar lines, if the ILC final output attaches a special meaning to a term appearing in the treaty, then that special meaning will apply to the term in question due to the fact that the parties agreed upon, or acquiesced to it by not opposing to it.⁵⁹ Therefore, what elicits the additional impact of the ILC output in the interpretative process of a treaty is the stance of the parties towards that output.

Besides, the diversity of ILC outputs further frustrates their wholesale classification into a singular category in the context of treaty interpretation. Indeed, there are several outputs of the ILC which have not led to the adoption of a treaty. The precise role of these outputs in the process of treaty interpretation varies depending on the treaty in question and the ILC output at hand. In the first place, ICTs use such outputs as non-assorted interpretative materials without any explicit reference to the rules of treaty interpretation.⁶⁰ Less often, a decision might attempt to draw some connection between its use of ILC outputs and the determination of the intention of the parties to the treaty. So, for instance, a Trial Chamber of the ICC contrasted the provision of the ICC Statute on aiding and abetting to the corresponding provision of the ILC 1996 Draft Code which included an additional element. The Chamber reasoned that ‘[a]lthough the 1996 ILC Draft was not an official part of the drafting history of the Rome Statute, […] it could be argued that had the drafters intended to include qualifying elements they could have done so explicitly in a similar manner to […] the ILC Draft Code’.⁶¹ Similar recourse to materials which are not binding on the parties and do not constitute, strictly speaking, part of the preparatory works of the treaty is not uncommon in practice as a supplementary means of interpretation under Article 32 VCLT.⁶²

Apart from these situations, ILC outputs that have not led to the adoption of a treaty might attain a more prominent role in the process of the interpretation of a treaty in considering its ‘context’ both as envisaged in Article 31 VCLT and in a broader sense. First, as will be shown in the sections that follow, international courts or tribunals may rely on ILC outputs in the process of the determination of rules of customary international law or general principles of law.⁶³ This has implications for the purposes of treaty interpretation, since an interpreter of a treaty may take into account an ILC output as an articulation of ‘other relevant rules of international law applicable in the relations between the parties’.⁶⁴ In this sense, international courts have also referred to the views of the ILC about the meaning of a specific term reflecting customary international law or the object and purpose of a rule of unwritten international law to support a finding about the ‘ordinary meaning’ of treaty terms or its ‘object and purpose’.⁶⁵ Thus, for instance, in Bosnia Genocide, the ICJ drew from the ILC 1996 Draft Code to find that genocide requires the intent to destroy a ‘substantial’ part of a protected group, despite the Genocide Convention being silent on the issue. The ICJ reasoned its acceptance of the substantiality requirement––proposed by the ILC with respect to the customary crime of genocide––by reference to the object and purpose of the Genocide Convention.⁶⁶ Second, the ILC has embarked on the consideration of topics arising from real or apparent gaps of existing treaties, such as its consideration of the topic of reservations to multilateral treaties or of subsequent agreements and practice in the process of treaty interpretation.⁶⁷ ICTs have relied upon these relatively recent instruments only sparsely without providing any detailed justification.⁶⁸ In principle, the fact that these outputs aim to clarify existing rules, which in part stem from previous treaties, should have no bearing on the weight to be accorded to these outputs according to the rule of treaty interpretation.⁶⁹ However, circumstantial evidence, such as subsequent action taken by the General Assembly or the reactions and comments of states to these outputs, may indicate the existence of a subsequent agreement or practice of the parties that can constitute the ‘context’ of a treaty or supplementary means for its interpretation.⁷⁰

The diversity of ILC outputs also influences how ICTs should use them in the process of treaty interpretation. In the first place, ILC outputs qualify at best as ‘preparatory work of a second order’.⁷¹ In the specific case of incidental interpretative pronouncements by the ILC, this entails that an interpreter should put emphasis on the evidence, upon which such outputs rely, and reconstruct (or deconstruct) the reasoning of the ILC by independently applying the rules of treaty interpretation. In this inquiry, there is no doctrinal reason to accord different weight to the ILC final draft or any other document produced by the ILC, because none of these documents originates directly from states.⁷² Nonetheless, the collective stance of the parties to the treaty towards an ILC output can radically change this configuration. As a corollary, when reliance on ILC outputs is so justified, the emphasis should be put on the interpretative statements contained in the ILC draft which enjoys the states parties’ approbation; that is, in most cases, the final ILC output.⁷³ Conversely, materials reflecting the personal views of members of the Commission, earlier views of the Commission, or even comments of individual states towards earlier drafts of the Commission should be accorded a lesser role.⁷⁴

What emerges from this analysis is that ILC outputs assume variable roles in the process of treaty interpretation. ICTs often accord them in practice more weight than the characterisation ‘preparatory works of the second order’ might suggest.⁷⁵ Such practice can be explained on the basis of the customary rule of treaty interpretation. For treaties originating in ILC outputs, the treaty is the litmus test for establishing the parties’ stance towards the ILC output. Similarly, the rule of treaty interpretation can provide a foothold for the use of interpretative statements contained in ILC outputs that are not part of the preparatory work of a treaty. Less conspicuously, the rule of treaty interpretation also entails a methodology for navigating through the different materials produced within the ILC in the process of clarifying the meaning of a treaty. Specifically, conduct indicating the approbation of a specific ILC output by the parties entails that this output takes precedence over other ILC outputs on the same topic for the purposes of treaty interpretation.

3 The Outputs of the ILC as a Means for the Identification of Unwritten International Law

3.1 The Outputs of the ILC as a Means for the Identification of Customary International Law

The identification of customary international law is another process with respect to which the label of ‘teachings’ might understate the role of ILC outputs.⁷⁶ In fact, even the ILC in its Conclusions on the issue seemed to single out its own outputs from the outputs of other bodies ‘engaged in the development and codification of international law’ as meriting special consideration.⁷⁷ According to the ILC, ‘a determination by the Commission affirming the existence and content of a rule of customary international law may have particular value, as may a conclusion by it that no such rule exists’.⁷⁸ Specifically, ‘the weight to be given to the Commission’s determinations depends, however, on various factors, including the sources relied upon by the Commission, the stage reached in its work, and above all upon States’ reception of its output’.⁷⁹ Whilst this statement might appear somewhat self-aggrandising to an external observer, it is hard to deny that it is firmly based on the practice of ICTs. Indeed, in this context, there is an abundance of evidence that ILC outputs may have a pivotal role in the process of the identification of custom.

Specifically, ICTs very rarely use explicitly the label of ‘teachings’ with respect to ILC outputs. In numerical terms, only one out of a sample of 409 decisions designates ILC outputs as ‘teachings’, whereas two more include this characterisation within the range of justifications for relying upon ILC outputs.⁸⁰ What seems to drive this tendency is the ensuing discrepancy between, on the one hand, the relative value which ‘teachings’ are to be accorded generally in the determination of applicable rules according to the ICJ Statute and, on the other hand, the actual use of ILC outputs in the context of the decision. In principle, the characterisation as ‘teachings’ entails that ILC outputs have no evidentiary value in and of themselves for the establishment of state practice and opinio iuris which constitute the customary rule.⁸¹ At least in the first place, these outputs should be approached with great caution focusing on the evidence they rely upon to establish such state practice and opinio iuris, rather than the normative propositions they contain.⁸² What is more, contrary to the ILC’s conclusions, there is no reason to distinguish between the ILC’s final outputs, their commentaries, and the normative propositions contained in previous drafts and reports.⁸³ All such propositions only embody the opinions, and often the disagreements, of learned jurists.⁸⁴ Even assuming that such propositions were developed with the input of governments, what formally counts as evidence of state practice or opinio iuris is the comments of governments as such, not the ILC output itself. In this respect, ICTs occasionally cite ILC outputs on a par with scholarly writings to support a determination that a certain normative proposition found in judicial pronouncements or other sources reflects a rule of international law.⁸⁵

What ‘above all’ may increase the value of ILC outputs for the determination of rules of customary international law is their reception by states.⁸⁶ This is most conspicuously the case with respect to treaties in the drafting of which the ILC had a role. In this context, the ILC’s views are particularly relevant in establishing whether a treaty codifies a rule of customary international law or its negotiations have led to the crystallization of such a rule.⁸⁷ In principle, it is the practice of states in the negotiation and conclusion of a treaty and not the ILC outputs that constitutes state practice for the purposes of the formation of customary international law.⁸⁸ Yet, the ILC’s views construed as part of the preparatory work of the treaty can constitute evidence for the determination of the opinio iuris of states with respect to the character of a treaty provision.⁸⁹ For instance, in North Sea Continental Shelf, the ICJ resorted to the ILC outputs leading up to the adoption of the 1958 Continental Shelf Convention so as to determine the ‘opinio [i]uris on the matter of delimitation’.⁹⁰ The ICJ unreservedly attached decisive weight to the ILC outputs explicitly admitting that ‘the status of the rule in the Convention […] depends mainly on the processes that led the Commission to propose it’.⁹¹

At the same time, the ILC hints that there can be an alternative way to justify reliance on the normative propositions contained in its final outputs for the purpose of determining rules of customary international law regardless of whether they lead to the adoption of a treaty. Specifically, if the General Assembly takes action with respect to a final draft of the Commission, such as annexing them in a resolution and commending them to states, such action can be considered an instantiation or evidence of state practice and opinio iuris.⁹² According to the ILC, such action does not constitute ‘conclusive evidence’ for the customary character of the normative propositions contained in ILC final outputs.⁹³ What the ILC scheme seems to envisage for such normative propositions after the approbation of the General Assembly is a liminal space between emerging and existing law.⁹⁴ On the one hand, the normative propositions contained in the ILC final output can no longer be considered mere ‘teachings’, as they also constitute ‘important evidence’ of the collective opinion of virtually all states as to the existence and content of rules of customary international law.⁹⁵ On the other hand, an international court or tribunal may still have to justify its reliance on ILC final outputs on the basis of further evidence, since the General Assembly lacks the competence to impose binding rules on states. What counts as state practice in the context of the identification of customary international law is not the General Assembly resolution as such, but states’ conduct in connection to that resolution.⁹⁶ Hence, the importance of the General Assembly’s commendation can be undercut by circumstantial evidence suggesting a lack of general practice or opinio iuris, such as the adoption of a resolution with partial support or little substantive discussion or the Assembly’s decision to maintain the topic considered by the ILC on its agenda for further consideration.⁹⁷

That said, the practice of ICTs is much less methodical than these considerations might suggest. Very frequently, ICTs apply the normative propositions articulated in ILC outputs because they ‘codify’,⁹⁸ ‘lay down’,⁹⁹ ‘reflect’,¹⁰⁰ ‘state’,¹⁰¹ ‘restate’,¹⁰² ‘express’,¹⁰³ ‘formulate’,¹⁰⁴ ‘articulate’,¹⁰⁵ ‘represent’,¹⁰⁶ ‘are declaratory of’,¹⁰⁷ or ‘are part of’¹⁰⁸ customary international law. Overwhelmingly, these findings are couched in axiomatic terms without any further explanation or are reasoned in such vague terms so as to amount to little more than assertions.¹⁰⁹ When they do reason such findings, ICTs tend to uphold the authority of the ILC outputs on the basis of various justifications including:

• (i) Vague references to the evidence they rely upon¹¹⁰;

• (ii) The mandate of the ILC and the particularities of its drafting process¹¹¹;

• (iii) Their subsequent reception in the practice of states including, more often, subsequent UN General Assembly action¹¹²;

• (iv) The pronouncements of other ICTs finding that certain provisions laid down in an ILC output reflect customary international law¹¹³;

• (v) The stance of the parties to the dispute towards the provision proposed in the ILC output in question.¹¹⁴

Whatever the specific line of reasoning, the common thread between these decisions is the finding that an ILC output or a specific provision proposed by the ILC has decisive value for the identification of customary international law on a certain matter. This is so notwithstanding the fact that the UN General Assembly might have technically reserved a specific topic for further consideration.¹¹⁵

The key takeaway from this analysis is that ILC outputs can assume different roles in the context of the identification of customary international law. Whilst there are firm doctrinal reasons to consider them as merely subsidiary in principle, practice suggests that they can often obtain important value as evidence of customary international law. Indeed, ICTs might accord to the normative propositions contained in ILC outputs decisive value so that they are treated as having the status of––or, more precisely, as materially identical to––rules of customary international law. The implications of this approach for the use of ILC outputs will be further explored in Sect. 4.

3.2 The Outputs of the ILC as a Means for the Identification of General Principles of Law

Besides custom, ‘general principles of law recognized by civilized nations’ may offer an alternative justification for the reliance on normative propositions contained in ILC outputs.¹¹⁶ Yet, apart from the undeniable status of general principles of law as ‘formal’ sources of international law, it is ‘a slight exaggeration to state that there is agreement on little else regarding their ascertainment, content and function’.¹¹⁷ The ongoing work of the ILC on general principles of law is testament to these ambiguities. One existential issue is whether general principles of law stem only from domestic laws or they can also emerge independently within international law.¹¹⁸ Second, whilst the requirement of ‘recognition by civilized nations’––or less anachronistically, ‘by principal legal systems of the world’––of a proposition as a general principle of law seems to have been established, it is unclear what it entails or how it is distinct from the process of the identification of customary international law.¹¹⁹ Third, it is not readily apparent whether common standards exist to determine the ‘general’ character of a principle, so as to enable its ‘transposition’ either from domestic laws to international law or, conceivably, from one context of international law to another.¹²⁰ Fourth, the function of general principles as independent sources of binding legal obligations is still contested. In this respect, general principles are often viewed as norms of a general character that do not impose a specific course of conduct, but which operate as ‘gap fillers’ or interpretative aids to avoid lacunae in international law in case no applicable rule can be found in treaties and customary international law.¹²¹ Overall, the crux of the contention seems to be that the requirements of recognition and transposability could imply a less robust requirement of state consent than the standards relating to treaties and customary international law. This inevitably brings to the fore the fundamental question of who has the final say about the validity of a normative proposition qua general principle of law if not states.

The role which ILC outputs can have in the context of the determination of general principles of law is inextricably linked to these questions. The traditional view of ILC outputs as ‘teachings’ entails that they are relevant in the first place as ‘a subsidiary means for the determination of general principles of law’.¹²² Therefore, much like the identification of customary international law, it is the evidence upon which the ILC outputs rely that is important for the identification of the general principle of law.¹²³ The normative propositions, which are formulated by the ILC in its final outputs should be approached with circumspection, since there is no doctrinal reason to accord them any more value than other ‘teachings’, including discussions and reports within the ILC. To illustrate this point, according to the ILC, a general principle derived from both the private and public law of most states is that an act does not constitute a breach of an obligation unless the actor is bound by the obligation in question at the time the act occurs.¹²⁴ However, domestic legal systems of private and criminal law deal with the problems arising from the application of this principle in vastly different ways. For instance, a previous draft of the ARSIWA introduced a distinction—inspired mainly from Romano-Germanic criminal law¹²⁵—between ‘obligations of result’, whose breach consists of a failure to achieve a result regardless of the conduct followed, and ‘obligations of conduct’, whose breach consists of a failure to undertake the prescribed course of conduct.¹²⁶ This distinction came under severe criticism that originated mainly from French scholars. They noted that in domestic private law systems—viz. French private law—obligations of result included prescriptions of specific conduct, whereas obligations of conduct required the taking of an effort to achieve a result.¹²⁷ Although the distinction was abandoned in the final draft of ARSIWA,¹²⁸ it still sporadically appears in judicial pronouncements and individual opinions predominantly in the form suggested by the French scholars.¹²⁹ This goes to show that certain ICTs or individual judges accorded little weight to both the ILC’s initial findings and its final decision to reject the distinction between ‘obligations of conduct’ and ‘obligations of result’.¹³⁰ They rather found other ‘teachings’ more convincing.

Yet, it is not uncommon for ICTs to declare that a certain normative proposition of the ILC is generally recognised in domestic legal systems without engaging in any detailed comparative examination or independently assessing its transposability in international law.¹³¹ This brings to the fore the question of whether the normative propositions contained in ILC final outputs can attain a more prominent role in the identification of general principles of law. Whilst with respect to customary international law it is clear that reception by states has an amplifying effect, the situation in the case of general principles of law is less straightforward. If states accept a normative proposition as law in their practice, this would indicate the existence of a rule of customary international law that is formally distinct from a general principle of law.¹³² That said, it is still possible to maintain that what enhances the value of ILC outputs in the process of the identification of general principles of law is their subsequent reception by states, particularly subsequent action by the UN General Assembly. In this respect, the Assembly’s action could be construed as a form of recognition by states of ‘the existence of certain principles intrinsically legal in nature’.¹³³ In other words, such approbation by states embodies the recognition that such principles exist. It also represents the determination that such principles are transposable into the international legal system or from one strand of international law to another, in a way that mirrors the establishment of opinio iuris in the case of custom.

That said, the institutional characteristics of the Commission—such as its composition, its mandate, and the thoroughness of its procedures—suggest that its determination as to the existence or not of a general principle of law and its content should have even more weight than its determinations relating to rules of customary international law.¹³⁴ First, unlike other ‘publicists’, the mandate of the ILC originates directly from a collective expression of state consent, namely the UN Charter and the UNGA action establishing the ILC and electing its members.¹³⁵ Second, the ILC Statute explicitly requires representation by the ‘principal legal systems of the world’ amongst its membership.¹³⁶ Third, legal experts seem better placed than any political body to deal with the essentially juridical task of determining whether ‘a principle is common to principal legal systems of the world’ and whether the systemic conditions exist to allow the application of such prescription on the international level.¹³⁷ The same is true for the systematization of seemingly disparate principles underlying rules established in treaties and customary international law to the extent that such principles can also be considered ‘formal’ sources of international law qua general principles of law.¹³⁸ Following the same line of reasoning, the traditional category of ‘subsidiary means’ also understates the role of decisions of––at least some––international courts for the determination of general principles of law.¹³⁹

On the basis of these considerations, one can construct an alternative line of justification for relying on the ILC’s outputs as depictions of general principles of law. Specifically, what amplifies their value as ‘material’ sources of general principles of law is their reception in the context of international dispute settlement by ICTs. On the one hand, it is not in doubt that judicial decisions or other outcomes of dispute settlement are only binding on the parties before the court or tribunal and only with respect to a particular dispute.¹⁴⁰ On the other hand, it is hard to argue that the support of certain normative propositions by multiple ICTs has only a ‘subsidiary’ value in practice. The power to have recourse to general principles of law is explicit in the mandate of some ICTs and arguably implicit in their mandate to resolve disputes before them and avoid a non-liquet.¹⁴¹ In this respect, it is difficult to overlook the fact that most ILC outputs put forward normative propositions that are for the most part drawn from judicial pronouncements.¹⁴² Conversely, it is also an undeniable fact that ICTs tend to adduce evidence for justifying their reliance on ILC outputs primarily from previous international judicial decisions.¹⁴³ To be sure, such decisions very often affirm the customary character of certain normative propositions. However, less frequently, judicial pronouncements can be understood as alluding to principles which draw their validity, or at least authority, from the aggregate of practice of ICTs. For instance, in the Chorzów Factory judgment, the Permanent Court of International Justice (‘PCIJ’) held that it is a ‘principle, which is accepted in the jurisprudence of arbitral tribunals’ that ‘in estimating the damage done by an unlawful act, only […] the damage done to whom is to serve as a means of gauging the reparation, must be taken into account’.¹⁴⁴ When the ICJ was called upon to elaborate on this principle in the Diallo judgment, the ICJ ‘t[ook] into account the practice in other international courts, tribunals, and commissions […] which have applied general principles governing compensation when fixing its amount’.¹⁴⁵ On both of these occasions, the World Court accorded much more value to the practice of ICTs for the identification of general principles and the determination of their content than the gloss of ‘subsidiary means’ as the PCIJ/ICJ Statute suggests. This goes to show that ‘general principles of law’ may offer an additional foothold within the theory of sources for the treatment of a normative proposition contained in an ILC output as a statement of a binding rule of law. Specifically, a consistent pattern of the use of such propositions in international dispute settlement can offer important evidence as to their ‘formal’ status as general principles of law.¹⁴⁶

To conclude this section, the category of general principles of law can provide––and, indeed, has provided––another justification for the use of ILC outputs in international adjudication. Yet, even in this context, these outputs may be relied upon in various ways. Whereas the starting point remains that these instruments constitute ‘subsidiary means for the determination’ of general principles of law, in practice they can attain more weight in this process. In fact, much like in the case of customary international law, ICTs may end up treating normative propositions of the ILC as materially identical to a general principle of law. ICTs rarely spell out the reasons for such additional value. This section has speculated whether the particular institutional and procedural characteristics of the ILC can imply a more prominent role for its outputs in the determination of general principles of law that is bolstered through their widespread use in international dispute settlement. However, it is hard to reconcile this idea with traditional accounts of the creation of international law based on state consent. In Sect. 4, I attempt to offer a more consistent account based on the analytical distinction between identification and interpretation.

4 The Interpretation of the Outputs of the ILC as a Proxy for the Interpretation of Unwritten International Law

4.1 The Text of the Outputs of the ILC as the Artefact of Unwritten International Law

The previous section has shown that ICTs often justify their reliance on ILC outputs with the gloss of the identification of rules of unwritten international law. This section turns to the practical implications of a judicial determination that a normative proposition of an ILC output has decisive value for the identification of a rule of unwritten international law. In the case of treaties, the determination whether a text or statement has the formal hallmark of a treaty entailing binding obligations, on the one hand, and the determination of the meaning of a binding treaty provision, on the other, clearly involves different considerations so much so that it is possible to speak of two distinct juridical operations governed by different rules.¹⁴⁷ In the context of unwritten international law, the distinction between identification and interpretation is still contested. In this respect, it has been maintained in theory that it is impossible to identify a rule of unwritten international law without, at the same time, determining its content.¹⁴⁸ Conversely, rules of unwritten law are not amenable to interpretation, this operation presupposing the existence of a text.¹⁴⁹ As a corollary, the determination of the content of a rule depends on the very same means as the identification of a rule and requires the establishment of state practice and opinio juris or of recognition and transposability, as the case may be.¹⁵⁰ This section (and the sub-sections that comprise it) shows that these theoretical considerations can only partially explain the practice of ICTs relating to ILC outputs. ICTs only start with testing the legal pedigree of a normative proposition contained in an ILC output. In fact, this determination allows ICTs to treat the normative proposition found in the ILC output as the written artefact of the rule.¹⁵¹ This methodological movement enables the interpretation of rules of unwritten international law, more conspicuously, through the implementation of a textual approach.¹⁵²

To start with, there are two widespread tendencies which clearly show that normative propositions of the ILC, whose legal pedigree has been affirmed, are treated as the written artefacts of rules of unwritten international law. However, the role of interpretation in determining the content of applicable rules is often less discernible, as it intertwines with the ways in which tribunals use ILC outputs in this process. First, ICTs very often proceed to apply normative propositions of the ILC to the facts of a case as self-explanatory. For instance, they routinely invoke ARSIWA with respect to attributing to the state the conduct of persons having the status of organs according to its domestic law.¹⁵³ Second, ICTs often identify in the terms of an ILC output an applicable rule of law and then refer to judicial pronouncements as a means to determine the meaning of that rule. For instance, in Jan de Nul, the tribunal found that Article 8 ARSIWA on attributing to the state the conduct of private persons acting under its control constituted ‘a statement of customary international law’.¹⁵⁴ It then went on to hold ‘[i]nternational jurisprudence […] requires both a general control of the State over the person or entity and a specific control of the State over the act the attribution of which is at stake; this is known as the “effective control” test’.¹⁵⁵ Subsequent awards have reproduced the Jan de Nul formula more or less verbatim.¹⁵⁶ That said, it is difficult to discern what precise juridical operation is at play in these decisions.

As to the first tendency, it is possible to argue that the lack of any separate analysis on the content of the applicable rule is suggestive of the absence of an intermediate step between the identification of a rule of customary international law or a general principle of law and its application.¹⁵⁷ Yet, this argument fails to fully convince. In most cases, ICTs do not even purport to engage in an independent analysis of state practice and opinio juris or a comparative survey.¹⁵⁸ Rather, they proceed to apply the formulations of the ILC to the facts of the case as if they were a binding text. The conciseness of analysis can also be construed as an emanation of a textual approach towards ILC outputs in a way that parallels known approaches of treaty interpretation. In other words, the tribunals’ line of reasoning consists conceivably of the application of the terms of a provision whose source of legal validity (customary international law or a general principle of law) has already been determined, because they deem its ordinary meaning to be sufficiently clear.¹⁵⁹

Indeed, the role of interpretation becomes more apparent in cases where the precise content of the normative proposition of the ILC is contested, but not its legal pedigree. In this context, ICTs have engaged in textual analysis. Thus, in Gabčikovo-Nagymaros, the ICJ confirmed that the conditions for the invocation of necessity laid down in the––then unfinished––ILC draft on state responsibility reflected customary international law.¹⁶⁰ The Court proceeded to deduce the content of the customary rule in the following terms:

The word ‘peril’ certainly evokes the idea of ‘risk’ […]. But a state of necessity could not exist without a ‘peril’ duly established at the relevant point in time […]. It could moreover hardly be otherwise, when the ‘peril’ constituting the state of necessity has at the same time to be ‘grave’ and ‘imminent’. ‘Imminence’ is synonymous with ‘immediacy’ or ‘proximity’ and goes far beyond the concept of ‘possibility’. […] That does not exclude, in the view of the Court, that a ‘peril’ appearing in the long term might be held to be ‘imminent’ as soon as it is established, at the relevant point in time, that the realization of that peril, however far off it might be, is not thereby any less certain and inevitable.¹⁶¹

Similarly, in Tulip, the tribunal accepted that ‘the ILC Articles constitute a codification of customary international law with respect to the issue of attribution of conduct to the State’.¹⁶² Turning to Article 8 ARSIWA, the tribunal focused on its text and decided that ‘[p]lainly, the words “instructions”, “direction” and “control” are to be read disjunctively’.¹⁶³ In the subsequent annulment decision in Tulip, the committee upheld the analysis of the tribunal finding that ‘[it] correctly interpreted Article 8’.¹⁶⁴ From a traditional perspective on the determination of unwritten international law, such findings seem untenable, if not plainly absurd. It is clear that the relevant pronouncements not only engage in a textual interpretation of the ILC output qua artefact of the rule of unwritten international law, but also such a textual approach is virtually dispositive for the determination of the meaning of that rule.¹⁶⁵

Turning to the second tendency identified above, it is possible to argue that the reliance on judicial pronouncements can be construed as an extension of the determination of state practice/opinio juris or recognition/transposability, as the case may be, albeit implicitly and on the basis of secondary evidence.¹⁶⁶ After all, judicial decisions, much like ILC outputs, constitute ‘subsidiary means’ for the determination of applicable rules.¹⁶⁷ However, in the context of the use of ILC normative propositions whose legal status is uncontested, such recourse is better understood as an interpretative operation. What I mean by this is that ICTs remain mindful that such previous decisions do not identify applicable rules but merely interpret such rules. Whilst ICTs are rarely explicit about their methodological choices, there is nonetheless evidence in judicial practice. So, for instance, in El Paso, Argentina argued that the tribunal had exceeded its powers by relying on case law to identify ‘fair market value’ as the applicable standard of ‘full’ reparation under the law on state responsibility in cases of violations of fair and equitable treatment, despite judicial decisions’ lack of binding status beyond the confines of a specific case.¹⁶⁸ The annulment committee dismissed this claim on the basis that ‘[a]rbitral tribunals must resort to different methods of interpretation to decide the dispute’ before them and, in the event, the tribunal relied on previous case law only ‘to be helped in its interpretation’.¹⁶⁹ In other words, the identification of a normative proposition of the ILC as the artefact of the rule of unwritten international law was a stepping stone, which allowed the tribunal to refer to international jurisprudence as an interpretative aid.

What emerges from this exposition is an emergent analytical distinction between the determination that a normative proposition of the ILC reflects a rule of unwritten international law and the determination of the content of such rule through the use of the ILC normative proposition as an artefact for interpretation. Reliance on previous jurisprudence clarifying the ILC proposition and, more overtly, a textual analysis of that proposition constitutes instantiations of this process of the interpretation of unwritten international law. The section that follows expands upon the ways in which the interpretation of ILC outputs operates as a proxy for the interpretation of unwritten international law through the employment of other means of interpretation.

4.2 The Interpretation of the Outputs of the ILC through Means Akin to Treaty Interpretation

A judicial finding that a normative proposition contained in an ILC output is materially identical to a rule of unwritten international law also enables an ICT to employ other means of interpretation with a view to determining its content. The key question is whether and how this process differs from the mainstream view about the rules on the identification of unwritten international law. To this end, this sub-section starts with an exposition of relevant pronouncements of the ILC relating to the identification of unwritten international law. It then turns to discuss tendencies in practice by reference to illustrative examples which controvert the ILC’s views and point to the existence of a process of the interpretation of unwritten international law which is analytically distinct from identification.

To start with, the ILC envisages the process of the identification of unwritten international law largely as an inductive process of the examination of evidence of state practice and opinio juris or of recognition and transposability, as the case may be.¹⁷⁰ But, even according to the ILC, it is not limited to induction. So, with respect to the identification of customary international law, the ILC has concluded that ‘the two-elements approach does not preclude an element of deduction as an aid’ particularly ‘when considering possible rules of customary international law that operate against the backdrop of rules framed in more general terms that themselves derive from and reflect a general practice accepted as law or when concluding that possible rules of international law form part of an “indivisible regime”’.¹⁷¹ More perplexingly, the current special rapporteur on general principles of law has opined that ‘deduction is […] the main criterion to establish the existence of a legal principle that has a general scope’.¹⁷² The terminology of deduction within the framework of the identification of rules of unwritten international law is imprecise. First, it opens up the possibility that rules––especially general principles of law––can emerge from any ‘preconceived ideas’ whatsoever.¹⁷³ Second, it understates the role of interpretation and creates unnecessary confusion as to the precise means of interpretation and their relationship to each other. This is particularly important in the context of the use of ILC outputs because the mainstream view is of little help for navigating through the diverse materials produced in the consideration of a topic by the ILC.¹⁷⁴

In fact, the means employed by ICTs to determine the meaning of a rule of unwritten international law after a finding that this rule is materially identical to a normative proposition of the ILC are not random. Apart from the text, a finding that a normative proposition reflects a rule of unwritten international law allows a consideration of the immediate and broader context of that normative proposition in determining its content.¹⁷⁵ For instance, in Bosnia Genocide, the ICJ affirmed that Article 8 ARSIWA relating to the attribution of the conduct of private persons under the instructions, direction, or control of the state reflected customary international law.¹⁷⁶ Yet, a purely inductive analysis brought to the fore a conflict between its own previous pronouncements on the notion of control and the findings of other courts applying a laxer test.¹⁷⁷ To resolve the impasse the Court referred to the context of the rule consisting of ‘the fundamental principle governing the law of international responsibility: a State is responsible only for its own conduct, that is to say the conduct of persons acting, on whatever basis, on its behalf’.¹⁷⁸ Similarly, in Diallo, the ICJ affirmed the customary character of Article 1 of the Draft Articles on Diplomatic Protection (‘DADP’) as a rule relating to the implementation of state responsibility.¹⁷⁹ It then went on to find that a state could exercise diplomatic protection with respect to human rights violations: ‘owing to the substantive development of international law over recent decades in respect of the rights it accords to individuals’.¹⁸⁰ Whilst this practice largely corresponds to the ILC’s understanding of deduction in the process of customary law identification, the nomenclature is somewhat misleading.¹⁸¹ In methodological terms, the ICJ seems to refer to other rules of international law, which it deemed relevant for the interpretation of the rule reflected in ARSIWA or DADP, in a way akin to the context of a treaty.¹⁸²

Along similar lines, it is not uncommon for ICTs to refer to the object and purpose of an ILC proposition qua artefact of the rule of unwritten international law. For instance, several decisions invoke the stability of international obligations as a stepping stone for a restrictive interpretation of the customary defence of necessity as reflected in Article 25 ARSIWA.¹⁸³ Another set of illustrative decisions declare that Article 38 ARSIWA on the award of interest reflects an applicable rule on compensation. Even though the provision is silent on the matter, the same decisions emphasise that the purpose of an award of interest is to ‘ensure full reparation’ and proceed to award compound interest.¹⁸⁴ Findings alluding to the terminology of the identification of general principles of law can also be possibly understood in this light. For instance, in Quiborax, the tribunal referred to Articles 34 and 37 ARSIWA and enunciated that ARSIWA ‘restate customary international law and its rules on reparation have served as guidance to many tribunals in investor-State disputes’.¹⁸⁵ It specified that ‘the remedies outlined by the ILC Articles may apply in investor-State arbitration depending on the nature of the remedy and of the injury which it is meant to repair’.¹⁸⁶ In this respect, it cautioned that ‘some types of satisfaction as a remedy are not transposable to investor-State disputes’.¹⁸⁷ In particular, it held that ‘the type of satisfaction which is meant to redress harm caused to the dignity, hono[u]r and prestige of a State, is not applicable in investor-State disputes’.¹⁸⁸ These examples further corroborate that the terminology of deduction and law identification is misguiding and unnecessarily vague. These findings seem to evoke the object and purpose or the ratio of the ILC normative proposition qua artefact of the rule of unwritten international law in order to determine the meaning of the applicable rule in a way that parallels known approaches to treaty interpretation.¹⁸⁹

In fact, less commonly, ICTs not only distinguish between the process of the identification and interpretation of unwritten international law, but they are also explicit about the interpretative principle that they apply. Most notably, an investment tribunal pronounced that ‘every rule […] of international law must be interpreted in good faith’.¹⁹⁰ It then went on to apply this rule of interpretation to the requirement of the exhaustion of local remedies under customary international law.¹⁹¹ The tribunal held that ‘[t]his rule is interpreted to mean that applicants are only required to exhaust domestic remedies that are available and effective’.¹⁹² Similarly, an annulment committee remarked with respect to Article 25 ARSIWA on necessity that the ‘the concept of “only means” is open to more than one interpretation’.¹⁹³ It held that ‘[i]n the light of the principle that necessity is an exceptional plea which must be strictly applied (a principle expressly stated in paragraph 1171 of the Award), […] “only” means “only”; it is not enough if another lawful means is more expensive or less convenient’.¹⁹⁴ Conversely, another tribunal noted that the text of Article 8 ARSIWA only mentioned ‘persons or group of persons’, but made no reference to ‘entities’ like, for instance, Article 5 ARSIWA establishing also a rule on the attribution of conduct.¹⁹⁵ The tribunal observed that ‘it would make no sense to impose a restrictive interpretation that would allow a State to circumvent the rules of attribution by sending its direction or instruction to a corporate entity rather than a physical person or group of physical persons’.¹⁹⁶ Instead, it chose a different interpretation considering that the instructions or direction would be received and acted upon by natural persons even in the case of corporations (i.e. the directors and agents of the corporation).¹⁹⁷ From a doctrinal viewpoint, the tribunal chose out of two available interpretations the one that gave full effect to Article 8 ARSIWA in what appears to be a straightforward application of the interpretative principle of effectiveness (ut res magis valeat quam pereat or effet utile).¹⁹⁸ All these illustrative examples suggest that ICTs do not identify applicable rules by deduction, but merely interpret rules which they have found already to exist. What is more, they do so by reference to specific considerations that resemble the process of treaty interpretation rather than vague preconceived ideas or values.

Another, less overt, indication that ICTs engage in the interpretation of unwritten international law through the proxy of ILC outputs is how they tend to navigate through the diverse materials produced by the ILC in the consideration of the topic. Unlike treaty interpretation, the ultimate aim of the interpretation of ILC outputs qua artefacts of rules of unwritten international law is not the determination of the intention of its drafters.¹⁹⁹ The formal foundation of the validity of the normative propositions contained in ILC outputs continues to be, in the final analysis, the assent of states either in the form of acceptance as law or of recognition as a general principle. These general considerations have a bearing on how ICTs use ILC outputs in the context of the interpretation of rules of unwritten international law.

First, one issue that often arises in practice is the relationship between an ILC final output and its commentary. In this respect, most decisions seem to accord great value to the ILC’s Commentary in interpreting the terms of a normative proposition of the ILC.²⁰⁰ For instance, in Gabčikovo-Nagymaros, the ICJ referred to the ILC Commentary to the draft articles on state responsibility to establish ‘the meaning given to [an] expression in Article 33 [now 25] of the Draft of the International Law Commission’.²⁰¹ However, when no such textual foothold exists or when there is an irreconcilable discrepancy or conflict between the normative proposition found to reflect unwritten international law and the ILC Commentary, there is a tendency to favour other interpretative materials or means of interpretation.²⁰² One conspicuous example is the award of compound interest in the context of investment arbitration. Investment tribunals reason such finding on the basis of a purpose-driven interpretation of ARSIWA,²⁰³ despite the ILC Commentary clearly favouring the award of simple interest.²⁰⁴

Second, in the context of testing the legal pedigree of a normative proposition, ICTs do occasionally resort to ILC materials other than those reflecting the views of the plenary of the ILC or past versions of ILC outputs.²⁰⁵ By contrast, when it comes to disambiguating a normative proposition of the ILC found to reflect a rule of unwritten international law, such references are very infrequent in practice and are virtually always used to confirm an interpretation reached by other means.²⁰⁶ Thus, for instance, the Appellate Body of the WTO, after affirming that Article 28 VCLT on the non-retroactivity of treaties reflected a general principle of law, went on to elaborate its content.²⁰⁷ Specifically, it first referred to the notion of continuous acts in Article 14 ARSIWA arguably as a provision reflecting a relevant rule of international law.²⁰⁸ It then turned to the ILC Commentary on its draft on the law of treaties and the views of the special rapporteur on the law of treaties to confirm its interpretation on the principle of the non-retroactivity of treaties with respect to continuous acts.²⁰⁹

It is possible to draw certain overarching conclusions from this indicative exposition. The determination that a certain normative proposition of the ILC reflects a rule of unwritten international law entails certain methodological implications. In this respect, the language of identification and deduction is unhelpful and imprecise, because it implies an unfettered scope of judicial creativity. In fact, ICTs tend to disambiguate the content of a normative proposition of the ILC qua artefact of the unwritten rule through interpretation. In so doing, they make use of means of interpretation which are similar to treaty interpretation. The employment of such means also allows ICTs to organise their analysis and navigate through the diverse outputs of the ILC in the context of interpretation.

5 Conclusion

It is trite that the ILC possesses a revered role amongst international scholars and practitioners alike. Arguably, this role is reinforced by social norms and expectations, not least broader desiderata about the capacity of the legal profession to shape reality. In this light, the feedback loop between the outputs of the ILC and the decisions of ICTs can also be attributed at least in part to institutional or social forces such as the intrinsic characteristics of the ILC or, more mundanely, the existing or past affiliation of international judges and arbitrators with the ILC.²¹⁰ Yet, at the very least, a closer look at the methods employed in the use of ILC outputs in judicial practice is instructive from a broader perspective as to the ways in which claims to authority take a legal form and gradually turn into limitations.

As shown, the use of ILC outputs in international adjudication is not random, but takes place against the background of rules and principles relating to international law identification and interpretation. In this light, a wholesale classification of ILC outputs under the label of ‘teachings’ is overly reductive. Rather, there are multiple footholds in the theory of sources that suggest a more consequential role for these materials in the context of law identification and interpretation. In particular, the rule of treaty interpretation provides not only a justification for the use of ILC outputs in international adjudication, but also a blueprint on how to use them. However, the use of these outputs in the context of the determination of unwritten international law is perplexed by gaps within the existing theory of the identification of these rules and the neglected role of the interpretation of unwritten international law.

A synthesis of the practice of ICTs relating to the use of ILC outputs reveals an analytical distinction between the identification and interpretation of unwritten international law. In this respect, ICTs commence their analysis by testing or asserting the legal pedigree of a normative proposition of the ILC. When a justification is offered, this part of the reasoning focuses on the establishment of state approval with respect to the specific outputs of the ILC. To this end, an international court or tribunal does not only take into account the evidence which the ILC adduces to support its proposition, but, more importantly, it relies on circumstantial evidence of approval by states like subsequent action by the UN General Assembly. A positive finding in this respect entails that the ILC output is accorded decisive value for the identification of a rule of unwritten international law or, in fact, it is treated as materially identical to that rule.

A key insight gained from this survey is that this determination is only a starting point. ICTs frequently resolve disputes about the content of the rule of unwritten international law by interpreting an ILC normative proposition through the use of interpretative means. This process of interpretation is in a way the inverse of the identification of such rules. First, the affirmation of the legal pedigree of a normative proposition of the ILC enables its treatment as the written artefact of the rule. Accordingly, ICTs have occasionally engaged in a literal or grammatical interpretation of a rule of formally unwritten international law. Second, whilst ICTs confirm the legal pedigree of normative propositions of the ILC in a piecemeal fashion, their interpretation takes into account their immediate and broader context and object and purpose. Third, whilst they seem to accord particular value to an ILC’s determination denying binding status to a certain normative proposition, they use materials produced in the run-up to the adoption of an ILC output only exceptionally and in a supplementary fashion to confirm an interpretation of a rule of unwritten international law.

The insights gained from this survey about the use of ILC outputs in international adjudication also raise broader questions about the mainstream understanding of the process of the identification of rules of unwritten international law. The ILC in its relevant outputs opined that the process of the identification of rules of unwritten international law is not limited to an inductive examination of evidence but also involves an element of deduction. Whilst the scope of this study was limited in this respect, as it has dealt exclusively with the use of ILC outputs by ICTs, it has adduced some evidence which concretise this aspect of the determination of unwritten international law. The process at play seems not to be the identification, still less formation, of rules from preconceived ideas, values, or principles. Rather, the practice seems to be indicative of a more robust process which involves the interpretation of a normative proposition whose legal pedigree has already been determined through an inductive examination of evidence. This process of interpretation is based to a very large extent on interpretative means akin to treaty interpretation and is structured in comparable ways.