Part I - The Theory of Customary International Law

1 Between Pragmatism and Disenchantment The Theory of Customary International Law after the ILC Project

Jörg Kammerhofer

1 Introduction

There is a fundamental, eternal and unresolvable conundrum at the heart of customary international law (CIL) – the ‘source’, if the pun may be excused, of the enigma that is customary law. It is that we do not know on what we can – or are allowed to – base our arguments for or against one or another concept. Authors frequently note the lack of discipline in our debates on the foundations of this source of international law, even as they fail to show any themselves. Plenty of old wine is poured into new bottles as we seem to periodically rediscover arguments which generations upon generations before us have made – sometimes all the way back to Roman law.

Debates on the theory of CIL continue unabated, inter alia because there is a continuing, strong, urgent and foundational belief that we need CIL in order to keep international law working. Instead of being able to see customary law as a primitive method of norm-creation which is severely limited in its utility and dismissing it – as their domestic colleagues are wont to do – as entirely unsuitable for modern legal orders which tend to be complex and technocratic, an important sub-group of international lawyers wish to see and/or create international law as such a complex legal order. To this group belong practitioners and international legal scholars with a stake in the actual functioning of the law. They imagine customary law to be capable of performing the complex functions analogous to legislation in domestic law.¹ They do so not out of a sense of pride or ego, but because they genuinely believe that we cannot rely on treaties alone, that we must have CIL² (and therefore do) in order to achieve the political goals international society or politics needs to progress (or those which they imagine do). But the question is whether that is reason enough to consciously or subconsciously change the mechanics of customary law to suit these perceived needs and whether CIL has the flexibility to react to these perceived needs.

Two events have prompted my writing of this chapter. The first is that the International Law Commission (ILC) concluded its project on the Identification of Customary International Law in 2018.³ Ably directed by Michael Wood, it has from the very beginning been suffused with the spirit of pragmatism. The project primarily wanted to provide guidance to decision makers, particularly those not professionally trained in international law. Engaging in depth with the theory of CIL was consciously avoided as far as possible. Yet, for all its self-avowed pragmatism, the ILC could not avoid taking a stance on the theoretical aspects of this source, even if only in a roundabout, subconscious manner. On the other side of the equation we find foundational critiques of CIL, with Jean d’Aspremont’s 2018 International Law as a Belief System as well as recent articles on CIL⁴ as excellent recent contributions to this genre. In these writings, CIL is downgraded to a set of doctrines within the canon of folk tales international lawyers tell themselves – our ‘bed-time stories’, so to speak.

Both methods have virtues, but both have very dangerous vices and both, in a sense, contain the seeds of their own destruction. One aim of this contribution will therefore be an effort to show the relative merits and demerits of these two approaches, exemplified in the ILC Report and d’Aspremont’s work. I will focus on what they can tell us about the theoretical foundations of customary law as a source of international law. I am sympathetic to both: CIL is on shakier ground than mainstream writers and practitioners assume, but the point cannot be to employ a brutal reductivism. In this chapter, I will show where the quicksand lies and why our reliance on this source is problematic. To paraphrase Carl Schmitt: whoever invokes customary international law wants to deceive.⁵

The second event to spark this chapter is that at the time of writing fifteen years had passed since I first published an article in the European Journal of International Law on the fundamental ‘uncertainties’, as I called them, of customary international law-making.⁶ This anniversary and the conclusion of the ILC project have prompted me to rethink the argument made then and to reconceptualise the foundation of this source whose importance for international lawyers is eclipsed only by their frustration in the face of the manifold aporia with which they are confronted when wishing to research and/or apply it. My work usually stops at the recognition that we cannot find the law which tells us what the rules on customary international law-making are. In this chapter, I will attempt to go a step further.

Accordingly, Section 2 will summarise what I consider to be the salient features of the two approaches, exemplified by the writings of its two champions, Wood for the pragmatists and d’Aspremont for the iconoclasts. This section is brief because these traits are better discussed using specific examples. Indeed, the example in Section 3 is the pivot point for this chapter, because it is both an illustration of the two approaches as well as an expression of the high-level problem: the ‘meta-meta law’ and the problem of finding what its content is. Section 4 will focus on this problem and will discuss my proposal for a new method for conceptualising this elusive level.

2 Two Approaches to Customary International Law

There are, of course, more than two possible approaches to customary (international) law and the choice of these two is arbitrary. Yet, they are well-known and well-respected archetypes for two essential directions the debates on this topic have taken in the past decade or so – both in terms of the sharp divergences that characterise them as well as the fact that they are surprisingly close on some points. Wood is typical of those scholars and practitioners who wish to construe CIL in a practicable manner from a ‘generalist’ perspective; d’Aspremont is the most adept communicator among the younger generation of scholars who seek to deconstruct the theoretical-philosophical foundations of the stories we tell about custom. Both approaches have merits, but both suffer from significant defects: Wood is right to focus on the positive law, but wrong to dismiss CIL’s problems so easily. His generalist-pragmatic understanding leads to an indistinct view of what CIL is and how it comes about; an impish soul might call him ‘the astigmatic pragmatist’. D’Aspremont is right to criticise that aspect, but the way forward in legal scholarship cannot lie in a reduction of law to collective psychology; he, in turn, could be called a ‘frustrated iconoclast’. My argument is, and has been for more than fifteen years, that both methods have a point, but that we require a combination of factors in order to make headway in international legal scholarship on customary law: it should be a theory-conscious analysis of the positive law in force.

2.1 Astigmatic Pragmatists

From the beginning of the ILC’s project on CIL, Michael Wood as the special rapporteur was committed to pragmatic goals, rather than to exploring theoretical (or even many doctrinal) questions. Wood’s First Report is clear about the project’s goal, namely ‘to offer some guidance to those called upon to apply rules of CIL on how to identify such rules in concrete cases’,⁷ that is ‘especially those who are not necessarily specialists in the general field of public international law’, because it is ‘important that there be a degree of clarity in the practical application of this central aspect of international law, while recognizing of course that the customary process is inherently flexible’.⁸ On that pragmatic level, concerned with the ‘usefulness of its practical consequences’,⁹ the project has to some extent succeeded. In this sense, the ILC’s work has a stabilising function and Wood is to be commended for his contribution. If he had remained on this pragmatic level, it would not have made for a good example for this chapter; however, there are indications that there is more to this mindset. For example, in a 2016 article, Wood (writing in his private capacity), argues: ‘Work on the topic has also shown that several longstanding theoretical controversies related to customary international law have by now been put to rest. It is no longer contested, for example, that verbal acts, and not just physical conduct, may count as “practice.”’¹⁰ One can take issue with statements such as this on several levels. For one, it is less than certain that ‘theoretical controversies’, including the verbal practice problem, have been ‘put to rest’ (which itself can mean a number of things). On another level, however, I submit that this type of statement is indirectly expressive of a particular view popular with practitioners and practice-leaning scholars, mistakenly believing that practice solves theoretical problems. While neither Wood nor the ILC texts openly declare it, one could argue that there is a subconscious belief that the eternal problems of customary law can be solved by the Commission declaring one side the winner – or that it should try. It is trivial to say that the ILC is not a lawmaker which could modify the law on customary international law-making. It is perhaps not so trivial to say that the role of the ILC as epistemic ‘authority’ – as an institution whose pronouncements can be presumed to accurately represent the state of international law – is equally problematic, particularly given the narrow range of sources and arguments on which it, like most orthodox international lawyers, relies.

Partially, this can be explained by the peculiar, if widespread use of the word ‘theory’ in English legal language. Whereas for example in German, Theorie or Rechtstheorie refers to legal theory, in English it tends also to be used for doctrinal statements about the material content of the positive law. The ‘theory of customary international law’ of which Wood writes tends to be concerned with questions like the relative value of domestic court judgments as state practice or the requisite number of instances of opinio juris.¹¹ Those are not the core research areas of legal theorists, but of international legal scholars – Rechtsdogmatik in German. If those topics are ‘theory’, then it is not surprising that legal theory properly so called finds no place in the ILC project and that a pragmatic project assumes that it has made changes to the ‘theory’ of customary law.

Largely, however, it is the culture of orthodoxy¹² which moulds this mindset. Orthodoxy is understood here as respect for conventional authority (acceptance by peers). International lawyers with their largely (but not consistently) ‘positivist’ outlook tend to exhibit three elements as part of the culture of orthodoxy: (1) submission: international lawyers submit to an apology of international tribunals (foremost the ICJ) as almost unquestioned authorities; (2) realist pragmatism: the pragmatic impetus unites with a belief in being ‘realistic’ and accommodating the ‘realities’ of international life, particularly practice – we know that practice is relevant because practice tells us that practice is relevant; (3) problem-solving: their pragmatic bent leads naturally to a tendency to try to solve problems, rather than analyse the law, even when they are not authorised to ‘solve’ the problems themselves.

On this basis, Wood’s reports combine a certain (small-c) conservatism on substantial issues, for example on international legal subjectivity,¹³ with a pragmatic modus operandi. As mentioned above, the problem arises when there is even the tacit assumption that this is the right way to cognise or change the law – the result is an unclear cognition, an astigmatism.

2.2 Frustrated Iconoclasts

Like myself, Jean d’Aspremont has critiqued the naïveté inherent in the ILC project, which cannot escape the theoretical problems of what he calls the ‘monolithic understanding of customary law’.¹⁴ This is obvious in manifold ways, including the central problem of verbal practice which we have both criticised in similar terms. His is a theoretical approach to (customary) international law; it is heterodox in the sense that theoretical coherence is more important to him than his arguments being in line with what is generally accepted. However, even the briefest look at his current theory, summarised in the 2018 book International Law as a Belief System, shows that its radical reductivism borders on non-cognitivism and threatens to destroy more than false assumptions. In this book, d’Aspremont argues with some justification that much of the (orthodox) discourse about what CIL is and how it functions – ‘the articulation of international legal discourses around fundamental doctrines’, as he puts it – has the hallmarks ‘of a belief system’.¹⁵ That, in turn, is characterised in the following manner:

[A] belief system is a set of mutually reinforcing beliefs prevalent in a community or society that is not necessarily formalised. A belief system thus refers to dominant interrelated attitudes of the members of a community or society as to what they regard as true or acceptable or as to make sense of the world. In a belief system, truth or meaning is acquired neither by reason (rationalism) nor by experience (empiricism) but by the deployment of certain transcendental validators that are unjudged and unproved rationally or empirically.¹⁶

The ‘fundamental doctrines’, such as (our talking about) CIL are ‘organised clusters of modes of legal reasoning that are constantly deployed by international lawyers when they formulate international legal claims about the existence and extent of the rights and duties of actors’,¹⁷ which sounds reasonable as a sociological description of the language use by international law professionals. And indeed, on first blush, d’Aspremont seems to carefully guide us through the problems of this deconstructive enterprise. This new view of customary law doctrine as part of a belief system and as a cluster of reasoning is supposedly a ‘heuristic undertaking’ ‘with a view to raising awareness about under-explored dimensions of international legal discourse’. By this method, he posits the possibility of ‘a temporary suspension of the belief system’ and ‘a falsification of the transcendental character of the fundamental doctrines to which international lawyers turn to generate truth, meaning or sense in international legal discourse’.¹⁸ In more conventional terms, by realising that the (dominant) way in which we talk about international law and the widespread acceptance of certain doctrines does not equal ‘truth’, the authority of orthodox assumptions can be questioned. So far so good: questioning the unspoken assumptions of legal scholars is the main task of all legal theory and I would happily count myself among those who participate in this form of ‘radical’ critique in the word’s original sense: pertaining to the radix, the root or foundation, of our knowledge.

Yet at this point the critique turns to iconoclasm, despite d’Aspremont’s avowed aim of avoiding ‘apostasy’, which ‘is neither possible nor desirable’.¹⁹ Let us look at what d’Aspremont does not (wish to) talk about: international law itself and the relationship of the doctrines/belief system to the body of rules/norms that is the law. The open question is as to the reason for this reluctance, which he shares with many other postmodernist international legal theorists. My interpretation of this peculiar state of affairs – peculiar from my theoretical vantage point – is that for d’Aspremont, two foundational beliefs strongly discourage talking about the law itself in any meaningful way: (1) a general noncognitivism; (2) a specific aversion to the ‘ruleness of sources’.

1. (1) For d’Aspremont, the title of his book is enough to show the radical reductivism of this strain of thought: it is international law that is a belief system. Despite considerable vacillation between the possibility of rules/norms and their denial, in the end, international law is identified with and reduced to ‘law-talk’ – the way we talk about the law is the law. The ‘existence’ in any sense of the word of international law as body of rules (as legal order) is half-negated. It seems – although it is difficult to pinpoint in the text – that, on the one hand, substantive rules are rules properly speaking, but on the other hand, sometimes certain parts of the law and the law in general is doctrine. Law is doctrine, law is a socio-psycho-linguistic phenomenon, law is reducible to (a special kind of) facts and apprehensible only by social-scientific methods. Even when d’Aspremont’s approach was still closer to Hartian legal positivism, it tended to favour reductivist, legal realist and anti-metaphysical readings of Hart.²⁰ With this book, this trend is strengthened and he is now closer to the postmodernist orthodoxy in international legal theory.

2. (2) Denying the idea that the law regulates its own creation (i.e. sources sensu stricto), and that sources are not themselves law unites certain post-Hartian and postmodernist theoretical approaches with a long tradition of state-centred thought in international legal doctrine. Whereas the former would rather, as d’Aspremont does, reduce sources to a doctrine – to teachings and to methods of law ascertainment²¹ – the latter see the source of law immediately founded in facts.²²

The problems which this approach engenders are, at least potentially, destructive not just of false orthodox narratives, but of the very idea of law. It does not really matter that d’Aspremont asks us only for ‘a temporary suspension’²³ of the belief system. The very possibility that we can simply suspend belief destroys the underlying concept and is probably self-contradictory – as if we could temporarily suspend belief that half, but not all, of the audience members are in an auditorium. Reductivism of this sort must face up to the enormous problem that it cannot distinguish between the belief system of doctrines about the law and the possibility that the law itself is no more than a belief system. This idea is indeed more than a heuristic tool to critique baseless orthodox taboos and fetishes, it is more than ‘apostasy’; it negates the very possibility of law as something separate from what actually happens in the physical world, its counterfactual nature. ‘The argument that in the end … the “existence” of law “is a matter of fact” is a negation of the very possibility of Ought. Ideals cannot be deduced from reality alone.’²⁴ It does not really matter that perhaps the reductivism which d’Aspremont wishes to promote is not anti-norm-ontic, merely epistemic. As long as the mediatisation of law by way of beliefs and discourse is watertight, law is still reduced to facts. If we adopt such reductivism, we are throwing the baby (the notion that ‘you ought not to kill’ makes sense as a claim to regulate behaviour) out with the bathwater (the true observation that many of our most cherished doctrines have little to do with the content of the positive law). But there is some ‘hope’ that orthodoxy’s serene pragmatism will domesticate and ultimately frustrate this iconoclasm – as it tends to do with all theoretical arguments, whether they are right or wrong.

3 Verbal Practice as Example

How do the two approaches deal with an issue which is not always acknowledged as problem, but which (from a theoretical vantage point) is far from problem-free? From a range of potential topics I have chosen verbal practice, because it allows me to demonstrate the strengths and weaknesses of both approaches introduced in Section 2 – but also because it is an ideal candidate to show the fundamental problem of all CIL theory (Section 4).

Verbal acts have become incredibly important for international law and we have increasingly turned to texts to support our claims to the emergence, change and destruction of customary law. That is because our world has become more complex whereas customary law as ancient law-creation mechanism originally based on raw actions has not. The classical controversy about the role of verbal utterances as practice has abated and it is virtually universally admitted that statements can be state practice.²⁵ In customary international humanitarian law, for example, reliance on verbal practice has far eclipsed ‘battlefield practice’ – take the ICRC study’s almost exclusive use of verbal emanations such as manuals as example for this trend. For example, the ‘Practice’ section for the principle of distinction contains a vast amount of material. As far as I can tell, all of these are statements and not a single instance of battlefield practice is mentioned;²⁶ for example under ‘Other National Practice’, the study quotes the following ‘“[i]t is the opinio juris of the United States that … a distinction must be made”’ – opinio juris is thus made practice. The entire project seems to be aimed at reporting statements, rather than acts.

The pragmatic temperament of colleagues has meant that they are unwilling to exclude any factor that might possibly be useful. Accordingly, verbal acts are now universally recognised, including by Wood. He is dismissive of those who problematise the use of statements; those ‘views … are too restrictive. Accepting such views could also be seen as encouraging confrontation and, in some cases, even the use of force.’²⁷ That is a strongly emotive argument – you better accept verbal practice or we may end up at war – but in terms of a dispassionate legal argument it cannot convince. Yet orthodoxy’s pragmatic impetus pushes Wood and the ILC to focus on the fact of widespread acceptance by peers:²⁸ “it is now generally accepted that verbal conduct … may also count as practice.”²⁹ The only substantive argument is negative; Wood quotes Mark Villiger’s 1997 monograph, which contains the following argument: ‘the term “practice” … is general enough … to cover any act or behaviour … it is not made entirely clear in what respect verbal acts originating from a State would be lacking’.³⁰ Neither argument is particularly strong. Why, on the one hand, should general acceptance by peers be a decisive factor in the creation or cognition of law? Cognition is not a matter for plebiscites; scholarship is not a dictatorship of the majority. This argument has pragmatic value – it is difficult to argue against it, certainly – but is weak in terms of scholarship. On the other hand, a whole school of thought in the classical debate on verbal practice has made it its business to set out what, exactly, this form of practice is ‘lacking’; we are, I think, not really confronted by a dearth of argument against verbal practice.

And, indeed, there are two (partially overlapping) avenues to problematising verbal practice: one is doctrinal and another theoretical. D’Aspremont’s critique is, I submit, rather on the doctrinal than the theoretical side. When he argues that ‘the International Law Commission’s formal acceptance that practice and opinio juris can be extracted from the very same acts collapses the distinction between the two tests’,³¹ I would argue that he is more concerned with a question of language-use, whereby we do not keep apart the two elements and the evidence for them. As I argued in my 2004 article (and as is obvious by the ICRC counting a clear instance of opinio juris as practice), the word ‘practice’ seems to lead a double life.³² Villiger, in another section of the monograph mentioned above, gives us an indication of this double meaning. When he argues that those denying the validity of verbal acts ‘cannot support their views on State practice with State practice’,³³ a critical reading would see him commit a circular argument; on a more charitable reading, however, it is obvious that the two meanings of ‘state practice’ differ markedly.

The most common use of the word ‘state practice’ is wide. Discussing the special problem of treaties as state practice, Villiger notes that ‘the acceptance of a convention[’s] … significance as State practice, namely as an expression of opinio juris, is lessened for three reasons’.³⁴ Quoting another classical author, Michael Akehurst puts it to us that ‘State practice means any act or statement by a State from which views about customary law can be inferred’.³⁵ We can see that state practice on this reading can express opinio juris; it can be the basis for inferences to what states believe to be the state of customary law. I believe that there is a conceptual (but still doctrinal) case to be made that we need to distinguish between evidences/proofs of custom-forming elements and the elements themselves. Critics are certainly right that this commingling almost inevitably leads to problems. I join d’Aspremont, however, in arguing that – on this level – the problem is more practical and more a question of internal incoherence of the orthodox position:

[T]he Special Rapporteur, while showing some awareness for the problem of double counting, had no qualms defending the idea that practice and opinio juris could be extracted from the very same acts. … it must be emphasized how difficult it is to reconcile the claim made in Conclusion 3 that each of the two elements must be verified separately with the explicit possibility that practice and opinio juris may be extracted from the same acts.³⁶

As pragmatic argument, however, this line of critique is liable to be rebuffed by equally pragmatic assurances that in practice this does not matter and will not be problematic: ‘distinctions between “constitutive acts” and “evidence of constitutive acts” … are artificial and arbitrary because one may disguise the other’.³⁷ Wood’s equivocating between his insistence on the separation of the two elements and his willingness, to admit ‘state practice’ as evidence of opinio juris³⁸ is borne of that pragmatic impulse. Nothing would be easier than to give in to the idea that we could just ‘take some scholars’ use of the term “state practice” with a pinch of salt’.³⁹ This ‘state practice’, as it were, is merely a slightly inaccurate use of the word, given that ‘state practice’ is also the technical term for the objective element of customary law creation. ‘State practice and opinio juris may be categorically different things, but we may look for proof of either element in the same place’⁴⁰ – problem solved.

However, these pragmatic manoeuvres hide the real theoretical problem which only the second avenue of critique is able to clarify. Neither the orthodox international lawyer nor the ‘international law as argumentative practice’ theory espoused by d’Aspremont and others – the ‘post-ontological … mindset’⁴¹ – are likely to be able to distil the fundamental legal theoretical problem of verbal practice. The classical canon of generalist international legal writings is once more on point. I still⁴² believe that Karol Wolfke’s argument expresses the true theoretical problem of verbal practice, whether or not a theoretical, rather than doctrinal argument was intended. For him, admitting verbal practice ‘neglects the very essence of every kind of custom, which for centuries has been based upon material deeds and not words. … customs arise from acts of conduct and not from promises of such acts’.⁴³ Wolfke is correct: the utterance ‘I will do x’ does not mean physically doing ‘x’: ‘repeated verbal acts … can give rise to international customs, but only to customs of making such declarations, etc., and not to customs of the conduct described in the content of the verbal acts’.⁴⁴

The theoretical basis for Wolfke’s argument is that at least in the civil tradition of customary law, despite just about everything being contentious about this source, one thing is reasonably clear: customary law must be based on customs. Customs, usus, actus frequens,⁴⁵ in turn must be an observance of the budding prescription or exercise of the budding right. Without the manifestation in a behavioural regularity of what will become binding, there can be no usage; without usage, the law-making of unwritten laws is not customary law. Customary law must be about customs. It is predicated on customs being or becoming obligatory. Only the behaviour that is (or is to become) the content of the norm – the ‘practical exercise of the legal rule’⁴⁶ – can serve as the objective element. Only doing or abstaining from ‘x’ can count as usus for a customary norm which prescribes ‘x’. Talking about doing/abstaining from ‘x’, in contrast, can be content forming only for a norm which prescribes talking. For norms which specify actual behaviour, none but actual behaviour will do as usus. With respect to state practice, the prohibition of torture is not constituted by states saying that they will not torture, only by the actual omission of torture. Customary law is a primitive form of law-making and cannot do all we ask of it in modern international legal debate.

The force of this argument is not undermined by the theory of speech acts.⁴⁷ Sometimes, speech can be more than descriptive: ‘I name this ship “Queen Victoria”’ or ‘I now pronounce you man and wife.’ One could therefore argue that verbal state practice is largely composed of such acts – the content, rather than the fact of uttering, is determinative. In certain cases, this may be true: it is conceivable that there are customary norms whose usus is a series of speech acts. However, not all speech is speech acts and this supplanting cannot happen for all, or indeed, for most legal rules. For example, ‘I am putting a chicken in the oven to be roasted’ is not a speech act; what is more, the chicken will firmly remain on the kitchen counter once I have uttered the sentence. The putting of the chicken into the oven can only happen in the real world; only when I have physically moved the chicken from the kitchen counter to the oven will the sentence be true. The same applies a fortiori to the usus with respect to customary norm-creation: uttering the words: ‘we are not torturing’ is not a speech act, but a (possibly accurate) description of the behaviour by state organs. It is not the actual omission of torture and – even if we accept the theory of performative utterances – cannot replace it as usus/actus frequens in the process of custom-formation for a norm regulating actual behaviour, as the prohibition of torture is.

However, this argument is predicated on customary (international) law being a type of norm that has two essential criteria: (1) the creation of customary norms requires a repetition of behaviour; (2) this behavioural regularity – the sum-total of behaviours attributed to the law-creation process – is the content of the prescription (Tatbestand) of the norm. Yet how do we know that this is the legally correct predicate? The concept of customary (international) law is not unitary, as d’Aspremont points out for international legal doctrine.⁴⁸ There are strong currents, from Roman law all the way to the drafting of the Permanent Court of International Justice (PCIJ) Statute in 1920, of a different basis for customary law: ‘acts … of a specific kind were … considered as custom-creative … only, because these acts evidenced the consensus tacitus in an adequate form’,⁴⁹ writes Siegfried Brie; the PCIJ/International Court of Justice (ICJ) Statute speaks of ‘a general practice accepted as law’. There are, however, also strong indications that over the course of the development of western (particularly civil) law, usus understood as behaviour which forms the content has been regarded as essential for the very idea of customary norms. Yet, even if we ignore countervailing arguments made over the course of the more than 1,500 years of our debating customary law, the legal-structural question remains: why is this the correct (or incorrect) reading of what customary (international) law is? On this level, we cannot effectively counter the orthodox insistence that verbal practice is part of state practice because it is accepted by all those who matter with the essentialist argument that customary law is necessarily shaped in a specific manner which conflicts with the majority opinion. That would mean succumbing to a metaphysical realism. If we do not wish to absolutise and reify concepts such as this, we must at least admit that this source of law could be different. Bin Cheng’s proposal to rename customary international law to ‘general international law’ once he eliminates practice from the conditions for law-creation, is consistent.⁵⁰ Whether CIL is different, however, is a question for the meta-meta law (Section 4).

4 Source-Creating International Law?

4.1 Law-Identification versus Law-Making

Therefore, as mentioned above, I strongly believe that on a truly legal-scholarly (rather than pragmatic) perspective, the real problem is the exact content of those rules which regulate the creation of CIL. Yet, instead of tackling it head-on, recent writings on CIL, the ILC project among them, engage in what looks to me like an exercise in avoidance: they speak of customary law ‘identification’ or ‘ascertainment’. Law-creation, not to mention the law of law-creation, is not discussed. This is understandable, given the widespread feeling among international lawyers that attempts at solving the problems of customary law have been unsuccessful, also given their resignation that they can ever be solved. Focusing on law-identification may seem like a legitimate alternative, particularly on the basis of the success of anti-metaphysical, radically reductivist and non-cognitivist Anglo-Saxon (legal) philosophy; d’Aspremont’s ‘post-ontological … mindset’.⁵¹

Take the ILC project. The 2018 ILC Conclusions open with the bold statement that they ‘concern the way in which the existence and content of rules of CIL are to be determined’;⁵² the commentary explains that the conclusions ‘concern the methodology for identifying rules of CIL, seeking to ‘offer practical guidance’ regarding this determination, which, in turn, means that ‘a structured and careful process of legal analysis and evaluation is required to ensure that a rule of customary international law is properly identified’.⁵³ For the ILC, the objective and subjective elements, state practice and opinio iuris, respectively, are identification elements: ‘to identify the existence and content of a rule of customary international law each of the two constituent elements must be found to be present’: ‘practice and acceptance as law (opinio juris) together supply the information necessary for the identification of customary international law’.⁵⁴ The ILC conclusions explicitly do not wish to address law-creation, which is seen as different from law-identification: ‘Dealing as they do with the identification of rules of customary international law, the draft conclusions do not address, directly, the processes by which customary international law develops over time.’ Yet the ILC seems to distance itself from law-creation properly (ie legally) speaking even in its disavowal. The ‘consideration of the processes by which [customary international law] has developed’, the ‘formation of rules’⁵⁵ which the ILC does not want to look at, seems to be one of the social or socio-psychological forces at play in physical reality, matters usually studied by legal sociologists, rather than legal scholars sensu stricto.

This raises a number of issues concerning both internal coherence and external justification. I have italicised a number of phrases in the previous paragraph to indicate some of the issues, for example the incoherence of arguing that it is a methodology, yet that we should follow them not merely to ensure proper ‘identification’ or ‘determination’ (a sort of ‘correct cognition’), but also qua rules to be followed: ‘as in any legal system, there must in public international law be rules for identifying the sources of the law’.⁵⁶ This seems to be belied by the project’s aim of (merely) providing ‘practical guidance’ to law-appliers. While the ILC’s pragmatic impetus may excuse some of this theoretical imprecision, d’Aspremont’s taking-on-board of this preference is particularly problematic. Discussing the ILC project in a recent paper, he harks back to an understanding of sources he had held earlier – before his recent sceptical turn – of the ‘sources’ as ‘law-ascertainment’.⁵⁷ In International Law as a Belief System, he mentions ‘the establishment of two distinct facts, that is, practice and opinio juris (acceptance as law)’ which are ‘the dominant modes of legal reasoning to determine the existence and content of a rule of customary international law’⁵⁸ – determine the existence, not create. Law-creation is thus robbed of its legal character, harking back to old, yet still popular ideas of the sources of law as not themselves law.⁵⁹

This sustained privileging of the law-cognising over the law-creative function has wide theoretical implications. If the ‘state practice plus opinio juris’ formula is merely, but necessarily, the only legitimate method for determining the existence and content of CIL, is it not peculiar that this process is so rigid, formal and so much like a form of law-creation? Is it not much more likely that a whole range of (epistemic) ‘methods’, which may not have much to do with the two elements, allow us to cognise whether CIL exists and how it is shaped? Is it not much more likely (and does it not accord better with the mainstream understanding of customary law) that state practice and opinio juris are the two elements of law-creation – the two conditions which international law prescribes for the creation of norms of the type ‘customary international law’ – rather than ‘mere’ law-ascertainment or law-identification?

How likely is it that state practice and opinio juris are purely of epistemic interest, rather than factors of law-creation? On that view, state practice and opinio juris would be the microscope with which we can observe cell division, not the cell division itself. If we assume that state practice and opinio juris are mere epistemic tools, how does CIL come about, then? Do we not need customs and a belief or consent to be bound? Even on the epistemic level, would it not be much more sensible to argue that the cognition of CIL involves, as I put it in earlier writings, ‘a re-creation of its genesis’,⁶⁰ that is, an analysis of the various instances of state practice and opinio juris, precisely because these two elements are legally required to create CIL? If this were a debate about domestic legislation, nobody would be tempted to ascribe the label ‘means of law-identification’ to the approval of the bill by the Houses of Parliament or to the sanction by the head of state.

Neither the ILC nor d’Aspremont tell us what, exactly, this ‘identification’, ‘determination’ or ‘ascertainment’ is. Are they properly part of the cognitive faculties – epistemic processes of law-cognition? How can they then be rule-governed? Probably a similar shift in meaning has taken place as for what I have called ‘interpretation_B’ – a process preparatory to application, to be performed by organs, guided by rules somehow inherent to the legal order, but utterly muddled by confounding it with real cognitive processes (interpretation_A). In our case, we would get custom-identification_A versus custom-identification_B. Probably also, the idea of ‘rules’ of custom-identification is as misguided as the idea of rules of interpretation which can somehow determine the hermeneutic process.⁶¹

4.2 The Problem of Source-Law (Meta-meta Law)

Law-identification is important, but it cannot be part of the law. I believe that the whole debate is either a conscious strategy to avoid tackling the problems of law-creation without a written constitution (as a form of avoidance behaviour) or a subconscious category mistake. Figure 1.1 illustrates the different levels which many confound, partly because it is more convenient to do so, partly because of a genuine belief in the post-ontological and anti-metaphysical mindset. As argued in Section 2.2: in order to be able to see law as counterfactual and in order to be able to speak of ‘you ought to do/abstain from doing “x”’ in any meaningful sense, law (norms) cannot be reducible to facts, whether linguistic, factual or psychological. We cannot supplant law and law-making for ascertainment, for linguistic practices and for an analysis of the way we talk about the law – as much as we need to talk about these issues as well. If we follow this reductivist path, the whole idea of rules becomes precarious. What is left of law as standard setter if all we can do is look at linguistic practices?

Figure 1.1 Levels of discussion.

On a non-reductivist reading, then, there are at least three levels of legal-scholarly discussion: (1) legal scholarship analyses (cognises) which norms are valid, particularly (1a) empowerment norms, that is ‘source law’.⁶² (2) It may be possible to speak of a separate discussion about law-identification where the various proofs or evidences for the validity of various norms (particularly empowerment norms) are discussed. This is unlikely, however, because this pragmatic issue will probably be subsumed under (1) or (3). (3) Legal methodology, then, discusses the (proper) methods which legal scholarship may use/uses in order to correctly cognise the law.⁶³

In contrast, Figure 1.2 shows the different ‘levels’ of the law itself – the hierarchy of empowerment norms (‘sources’) and law created under it.⁶⁴ This is the object of cognition for legal scholarship as ‘structural analysis’⁶⁵ of legal orders (level (1) above). ‘The legal order is not a system of coordinate legal norms existing alongside each other, but a hierarchical ordering of various strata of legal norms,’ writes Hans Kelsen. For him, ‘a norm which has been created according to the terms of another norm derives its validity from that latter norm’,⁶⁶ ‘validity’ being the specific form of existence for norms/law. Also, however, such a derivation is necessary: ‘a norm [is] valid, if and when it was created in a certain fashion determined by another norm’,⁶⁷ because we need to keep apart the norms (Ought) as claims to determine human behaviour from ‘mere’ reality.

Figure 1.2 Levels of the law.⁶⁸

The structure of international law might approximate that given in Figure 1.2:

1. – The level of substantive norms contains a number of CIL norms, like uti possidetis or the right to innocent passage, but also a multitude of ‘cousins’ from other sources, such as the treaty norms like the prohibition on the threat or use of force in Article 2(4) UN Charter or the prohibition of expropriation in Article 13(1) Energy Charter Treaty 1994.

2. – The meta-law contains a number of ‘sources’: the empowerment norms to create substantive norms. One such (complex) norm for CIL must be part of international legal order for substantive customary law norms to be able to ‘exist’, that is: be valid. On the traditional reading, then, this empowerment norm prescribes two conditions – state practice and opinio juris – to be fulfilled in order for a customary norm to be created.

3. – If customary and treaty law as well as general principles are to be equals, yet if all of international law is to be one legal order, a further norm is required, since creation according to the same empowerment norm constitutes unity among a multitude of norms.⁶⁹ This could be what Kelsen calls ‘historically first constitution’ (historisch erste Verfassung):⁷⁰ the hierarchically highest positive norm of a positive normative order. It is what I have called ‘meta-meta law’,⁷¹ the legal determinant for which sources the international legal order contains.

The distinction between these levels is crucial, as our arguments about the content of the empowerment norm for CIL can only be grounded on this level. In other words, if author A were to argue that state practice is not a required element for the creation of CIL and that this is the case because many morally valuable (proposed) norms would not exist otherwise,⁷² A must fail. A could succeed only if they could prove that the meta-law for custom-creation is shaped that way. A’s arguments regarding the meta-law, in turn, depend on the shape of meta-meta law: A’s claim that customary international law-making does not require state practice is true if the meta-law is shaped that way; the meta-law is shaped that way if it has been created according to the meta-meta law.

On a truly legal-theoretical perspective, the solution must lie within the law, as law regulates its own creation. The coming-about of CIL is and has to be based on law. Hence, we must find that meta-law, the law on customary international law-making. Yet the tragedy which I have frequently decried is that while we must proceed in this manner in order to generate legal-scholarly knowledge, we cannot do so. Those ‘empowered to will the highest echelons of international law … are unlikely to ever have’⁷³ created meta-meta law. There are also limits to what we can say about international law at this high level. Our epistemological horizon is too limited to answer this question with more than a presumption. As long as we are presupposing, we could presuppose any ‘method’ to create customary law, even an absurd one. Whatever the case, our knowledge of the content of meta law suffers, because there is little we can do to improve our knowledge of the historically first constitution.

What will never do is to say that this justifies basing our arguments (eg on whether state practice is required) on considerations of pragmatic expediency, political legitimacy or moral necessity: even if there were no meta-meta law (and CIL were to be its own legal order with the Grundnorm: consuetudines sunt servanda), arguments of this type would still be based on a category mistake. What will also not do is the ILC’s ‘optimistic’ approach: the attempt to trivialise and minimise the problems of CIL by maximising the leverage of the most widely accepted positions. This is combined with the acceptance of the factual influence of the orthodox position, foremost the ICJ’s jurisprudence constante. When asked what the ‘rules for identifying the sources of the law’ are, Wood’s answer is this: ‘These can be found … by examining in particular how States and courts set about the task of identifying the law.’⁷⁴ While Wood uses the word ‘identifying’, rather than ‘creating’, identification supplants creation in the ILC project (Sections 2.1 and 4.1).

4.3 Approximately Plausible Empowerment Norm

As mentioned above, this is the point where my analysis usually ends: little more can be said, from a legal-theoretical point of view. We cannot know more; those who purport to do so base their arguments on ineligible grounds. In this, we must face a particular problem for the international legal order which permeates all normative orders: how to proceed when the law is ‘sparse’? In other words, what is the ‘default’ position when there is little in the way of (proven or provable) law? This question may look a lot like a burden of proof in the strict sense (ie which of the parties to a judicial procedure has the burden for specific arguments), but it is not. The default position question may or may not arise in judicial proceedings, but it is found on a different level than the standards of proof required by the procedural law for a particular tribunal. It is connected, in a contingent way, to the scholarly ‘burden of proof’, a burden of proof in a very wide sense: what does scholarship have to do in order for its arguments to satisfy the requirement of generating knowledge about the law? For example, scholarship usually does not have to prove a contention that a proposed norm is not valid (although there may be situations where it does); if a scholar claims, however, that ‘x’ is a norm of, say, CIL, this needs to be proven to be accepted as a ‘true’ legal-scholarly statement, rather than wishful thinking.

The default position question may also remind us of the ‘residual negative principle’ and the (in)famous dictum in Lotus that ‘[r]estrictions upon the independence of States cannot … be presumed’.⁷⁵ The voluntarist straw-man which has dominated much of international legal scholarly discussion of this passage is, I believe, mistakenly applied to it and another reading of the dictum is better-aligned with legal theory:⁷⁶ ‘“Restrictions” are only applicable if they are positive law of the normative system “international law”. If there is no law, there is no law.’⁷⁷ Under that reading, the non-validity of norms is the default; it thus provides a partial answer to the question.

If we cannot prove the content of an empowerment norm for CIL, yet arguendo proceed from the presumption that ‘customary international law’ exists and that there is such an empowerment norm, what would be the ‘default’ position? In a 1970 book, Herbert Günther mentions that he would proceed on the basis of ‘the assumption that custom … has the power to create law’:⁷⁸

If the norm empowering customary international law as source is thus called a ‘hypothesis’ or ‘postulate’, it is done only in the sense that this is conditional upon our being correct in that certain acts can create an Ought. The presumptive validity of particular norms of customary international law, derived as it is from the constituting norm [the source], is thus hypothetical as well: it is only possible to cognise a particular norm of customary international law as valid if the hierarchically higher norm on law-creation is seen as law and as norm.⁷⁹

This solution sounds quite Kelsenian. There would be a Grundnorm with the content: consuetudines sunt servanda⁸⁰ for a small legal order, comprising CIL (and potential subordinate sources) only, but not international law as a whole. This is also very close to my proposal ‘to incorporate all conditions for the creation of … customary law … into the postulated Grundnorm of … customary international law’. However, as I argued then, this cannot work, because the Grundnorm ‘cannot create what is not already positive. It only gives validity as existence as Ought’.⁸¹

I propose a much weaker heuristic tool, which I will call the Approximately Plausible Empowerment Norm (AppPEN). If we assume (a) that international law contains a positive empowerment norm for ‘customary international law’ and if we assume (b) that the ‘formal source’ thereby constituted includes a creation mode of the customary type, then certain arguments/forms of regulation are made more plausible and certain others are less plausible. This is informed legal-theoretical speculation that the international legal order is shaped that way, not abstract deductive proof. However, there are degrees of plausibility, because the possible structure of normative orders and the idea of law as norms (Ought) is not completely arbitrary and because we can in some cases see constructs not based on logical fallacies as more plausible than those who celebrate inconsistency. Yet this idea is not orthodox majority following: constructs (such as the two-element theory) are not part of AppPEN because they are widely accepted by peers, but because they are more plausible than single-element theories. AppPEN is thus much weaker than ‘ordinary’ legal-scholarly proof of the validity (vel non) of a specific norm, but it may be the best we can hope for, given our poor epistemic position vis-à-vis the highest echelons of the international legal order.

A few examples for this positive and negative plausibility might show how AppPEN would operate:

1. – It is more likely that CIL is based on customs – repetition of behaviour – than not. It is trivially true that an empowerment norm (‘formal source’) could prescribe norm-creation without requiring regular behaviour as basis, for example domestic legislation. However, whether or not we now propose renaming it,⁸² it is more likely that a source called ‘customary’ law is based on customs than not, particularly since the more than 1,500 years of debate have been reasonably consistent on this point.

2. – It is more likely that CIL is its own legal order than part of a complex hierarchically ordered international legal system. If we cannot prove meta-meta norms incorporating CIL, international treaty law, general principles etc. as part of one legal order, it is more likely that no such norm is valid. Hence, ‘international law’ may refer to a family of legal orders, rather than to one.⁸³

3. – It is more likely that there is one source ‘customary international law’ than a whole range of sources. It is possible that a number of empowerment norms is valid which allow for the creation of a whole range of non-treaty international law. Alfred Verdross proposes a variation on this scheme in a 1969 article: ‘It is impossible to found all unwritten norms of international law on the same basis of validity’; ‘yet, it is likely that there is some truth in all theories’⁸⁴ of how CIL is created. Hence, he argues to accept all those procedures which usually succeed in creating CIL. The theoretical basis for this argument is flawed – we cannot know that CIL has been created unless we know the law on creation, which is exactly what we set out to find. Yet it is also unlikely in terms of Occam’s razor: the creation of one empowerment norm is incredibly difficult, see the endless debates about customary international law-making; it is less likely that a whole plethora of such norms is valid, rather than one (or even none), particularly given that we have traditionally discussed only one.

5 Conclusion

I have only given a first impression of what AppPEN is and how it could be used. In particular, I foresee two types of use. The first is pragmatic, similar to the idea behind the ILC project: it allows those who have little time to study the various theories of and approaches to CIL and international legal theory – like judges of domestic and international courts and tribunals – to circumnavigate some of the problems by weeding out implausible and selecting plausible theories. The second, however, is legal-scholarly: doctrinal (international) legal scholarship cannot always question all its foundations and will have to make a number of assumptions. AppPEN helps to select those which are more plausible. Neither practitioners nor doctrinal scholars are, however, best served by the orthodox modus operandi visible in the ILC report. If, for example, a widely held argument is based on a contradictio in adiecto, the fact of the acceptance by peers cannot be better than an approach which consciously avoids solutions which are logically flawed or which are based on an incoherent legal theoretical stance. Legal theory’s goal is not to provide a balanced theory, that is, a theory likely to be most widely accepted by international lawyers, because this implies that truth is to be found in compromise and majorities. Rather, it is meant to be consistent and consistently legal, a theory which takes the positive law seriously, yet shows where the law ends, where arguments are self-contradictory and where pragmatism becomes a fetish.

If we do not wish to operate on such a provisional basis, however, the most consistent course of action is to learn to live with much less CIL than we are used to imagining. At the very least we must acknowledge that customary law is a primitive mode of law-creation: we can do much less than we commonly assume. Customary law cannot be international law’s ‘saviour’ or its ‘future’ – the Zeitgeist will never walk where it can run.

2 The Custom-Making Moment in Customary International Law

Jean d’Aspremont

1 Introduction

Customary international law (CIL), as it is commonly construed in international legal thought and practice, is grounded in a particular social reality. In fact, the two constitutive elements of CIL, namely practice and opinio juris, correspond to two sides of the social reality which CIL is supposed to be grounded in. This is no new state of affairs. Even in the nineteenth century where CIL was thought to be the product of tacit consent, customary international law was construed as the product of social reality.¹ This social grounding of CIL is certainly neither spectacular nor unheard of. That CIL is grounded in a particular social reality bespeaks a construction that has become rather mundane since the Enlightenment,² and according to which norms are no longer supposed to be received by their contemplating³ addressees but are collectively produced⁴ by them as members of a self-conscious social community.⁵ Such an understanding of the making of CIL as originating in a process of self-production, where the authors and the addressees of the customary norm are conflated,⁶ is a manifestation of modern thinking.

Although simple in principle, the grounding of a norm in a social reality is a construction that commonly calls for a number of discursive performances for this grounding to be upheld in the discourse.⁷ The present chapter zeroes in on one of these discursive performances that is required for CIL to be grounded in social reality, namely the postulation of a moment in the past where the social reality actually engendered the norm. In fact, the grounding of CIL in a social reality captured through practice and opinio juris can only be upheld if there was a moment in the past where the practice and opinio juris of states – and possibly of other actors⁸ – have coalesced in a way that generates customary international law. In other words, for CIL to be grounded in social reality, there must have been a moment in the past where customary international law was actually made.

This chapter argues that, in international legal practice and literature, the actual moment where social reality has engendered a customary norm is never established or traced, but is always presupposed.⁹ According to the argument developed here, the moment CIL is made is located neither in time nor in space. Customary international law is always presupposed to have been made through actors’ behaviours at some given point in the past and in a given place. Yet, neither the moment nor the place of such behaviours can be found or traced. In other words, there is never any concrete moment where all practices and opinio juris coalesce into the formation of a rule and which could ever be ‘discovered’. This means that the behaviours actually generating the customary rule at stake are out of time and out of space. Because the custom-making moment is out of time and out of space, it cannot be located, found, or traced, and it must, as a result, be presumed. This is why current debates on CIL in both practice and literature always unfold as if all actors’ behaviours and beliefs had at some point coalesced into a fusional process leading to the creation of customary international law. However no trace of that fusional moment can ever be found, condemning this original fusion of all behaviours and beliefs to be presumed.¹⁰ This presumption of a moment in the past where the social reality creates the norm is called here the presumption of a custom-making moment. Most debates on CIL in practice and in the literature are built on such a presumption of a custom-making moment. As was said, this presumption of a custom-making moment is necessary for CIL to present itself as being grounded in a certain social reality.

This chapter is structured as follows. It first sketches out some of the main manifestations of this presumption of a custom-making moment (1). It then sheds light on some of the discursive consequences of presuming a custom-making moment, including those consequences for the interpretation of CIL (2). The chapter ends with a few observations on what the presumption of a custom-making moment entails for foundational debates about CIL as a whole (3).

Before elucidating the manifestations of the presumption of a custom-making moment and its consequences, a preliminary observation is warranted in light of the presumptive character of the custom-making moment. It could be claimed that the presumption of a custom-making moment is, like opinio juris,¹¹ yet another fiction around which CIL is articulated. In that sense, the custom-making moment would be a fiction about the origin of CIL. It is submitted here that claiming that the custom-making moment is a fiction says basically nothing about what such a presumption stands for and actually does. Indeed, it could be said that international law’s representations of both the reality and the past are always fictitious constructions.¹² What is more, fictions have always been a common mode of representing the real.¹³ That CIL rests on a fictitious representation of the moment of its making can thus not be demoted to just another fiction of international legal reasoning. It is a powerful discursive performance without which CIL could not do all what it does.

2 The Custom-Making Moment in the International Legal Discourse

The following paragraphs mention a few of the manifestations of the presumption of a custom-making moment in international legal thought and practice. It is, for instance, noteworthy that practice and scholarship continuously set aside the question of the duration of practice, as the determination of a minimum threshold would bring back the question of the custom-making moment.¹⁴ The recurrence of the metaphoric shorthand of ‘crystallisation’ to describe the formation of customary law in the literature similarly epitomizes the continuous avoidance of finding a custom-making moment and the presumption of the latter.¹⁵ In the same vein, it is striking that courts always locate the practice and opinio juris they find in the present,¹⁶ thereby constantly avoiding the tracing of a custom-making moment.

The presumption of a custom-making moment has occasionally been touched on in the literature. For instance, the famous discussion on the chronological paradox of CIL is a question that, although focused only on opinio juris, is all about the abovementioned presumption of a custom-making moment.¹⁷ Yet, those debates on the chronological paradox of CIL never explicitly acknowledge the presumption of a custom-making moment. Reference could also be made to scholarly discussions about the relations between a customary international legal rule with a corresponding existing treaty provision.¹⁸ In this situation, the treaty seems to provide some indication of the time and place of the making of CIL.¹⁹ And yet, here too, despite the treaty providing some vague direction in this regard, there is no effort to identify a custom-making moment, the latter remaining presumed.

It is remarkable that the International Law Commission (ILC), in its work on the identification of CIL, consciously decided not to look into the custom-making moment either. Indeed, as it stated in the commentaries to its 2018 conclusions:

Dealing as they do with the identification of rules of customary international law, the draft conclusions do not address, directly, the processes by which customary international law develops over time. Yet in practice identification cannot always be considered in isolation from formation; the identification of the existence and content of a rule of customary international law may well involve consideration of the processes by which it has developed. The draft conclusions thus inevitably refer in places to the formation of rules of customary international law. They do not, however, deal systematically with how such rules emerge, change, or terminate.²⁰

Interestingly, it is this very choice to exclude the question of the formation of customary law from the scope of its work that entailed a change in the way in which the ILC described its own work.²¹ It is submitted here that such a choice is not informed by the material impossibility to trace the formation of CIL or the irrelevance of the question for custom-identification, but by the very fact that this presumption is at the heart of the contemporary understanding of CIL. Being presumed, it does not even need to be traced. In the discourse on CIL, the question of establishing or tracing the custom-making moment simply never arises.

3 The Custom-Making Moment and its Doings

It is argued here that the presumption of a custom-making moment is not just a move of convenience to evade difficult methodological and evidentiary obstacles pertaining to the identification of CIL. Such a construction is being perpetuated for its many discursive virtues. Indeed, as was indicated above, CIL could not be upheld as being grounded in social reality if it could not be presumed as being made at a certain moment in the past. If it were not presumed to be made at a given moment in the past where opinio juris and practice coalesce and generate CIL, it would not be possible to hold that CIL originates in some form of social reality.

In enabling the grounding of CIL in social reality, the presumption of a custom-making moment simultaneously allows some other discursive moves which are worthy of mention here. In particular, attention must be turned to the way in which the presumption of a custom-making moment enables a two-dimensional temporality in the discourse on CIL. In fact, it organizes the life of CIL around two distinct moments, namely: (i) the (presumed) moment of making of CIL in the past; and (ii) the application of CIL in the present.²² If CIL has a past, albeit presumed, it can have a present distinct from that past. The postulation of a custom-making moment in the past thus allows the postulation of other ‘moments’. In particular, this two-dimensional temporality enables the idea that CIL is a product made in the past and subjected to interpretation in the present. Because one presupposes a custom-making moment in the past, one can think of CIL as a tangible artefact in the present which can therefore be subject to an autonomous and neatly organised interpretive process.²³ This is why those scholars that argue that the interpretation of the content of CIL can be distinguished from the interpretation of its legal existence²⁴ extensively and systematically build on this two-dimensional temporality.²⁵ In that sense, the current scholarly attempts to distinguish the interpretation of the making of the CIL rule from the interpretation of its content can be seen as being predicated on this presumption of a custom-making moment.

There is another important consequence of the presumption of a custom-making moment that ought to be mentioned here. That is, the anonymity and impunity in argumentation about CIL that accompany the presumed custom-making moment and its abovementioned two-dimensional temporality. Indeed, since the custom-making moment is outside time and out of space, and simply presumed, those generating the custom-making behaviours cannot be known. The only possible pedigree of customary international law comes to be reduced to ‘all states at some point in the past’. Being presumed, the custom-making process is actually anonymised. This anonymity is explicitly confirmed by the ILC:

The necessary number and distribution of States taking part in the relevant practice (like the number of instances of practice) cannot be identified in the abstract. It is clear, however, that universal participation is not required: it is not necessary to show that all States have participated in the practice in question. The participating States should include those that had an opportunity or possibility of applying the alleged rule.²⁶

The above statement shows that it is not necessary for the sake of custom-ascertainment to even seek to identify who did (or said) what. The ILC is thus saying that the custom-making moment, because it is presumed, ought not to be traced, named and individualised. Being presumed, the making of CIL can stay anonymous.

The anonymity that accompanies the presumption of the custom-making moment is not benign. In fact, as a result of this anonymity, no one can ever be made responsible for the rule of CIL concerned and what is claimed under its name. In other words, as any customary rule enjoys a life of its own out of time and out of space, what is said under the discourse on CIL cannot be blamed for both the good and the suffering caused in the name of CIL. All those invoking CIL can accordingly present themselves as candid followers and observers who just walk behind CIL, be it for the good or the suffering made in the name of CIL.²⁷

4 Concluding Remarks: The End of Foundationalism?

Customary international law epitomizes the idea of grounding. Indeed, by virtue of CIL, the rules to which a customary status is recognised are supposedly grounded in a past social reality. And yet, as has been argued in this chapter, such grounding can never be traced but can solely be presupposed.

It is submitted at this concluding stage that the question of whether this presumptive grounding of CIL is a satisfactory state of the discourse is irrelevant. It is particularly argued here that foundational debates about whether a customary international law rests on valid or invalid, consistent or inconsistent, legal or illegal, grounded or arbitrary, true or untrue, factual or imaginative foundations are bound to be sterile since the foundations of CIL are condemned to be presumptive.²⁸ As was shown by this chapter, venturing into foundational debates about the validity, truth, legality, consistency and factuality of the foundations of CIL is to condemn oneself to an inevitable defeat.

Yet, it must be emphasised that the limitations of foundational debates about CIL do not entail that one should verse into relativism, nihilism or even discourse vandalism. Although twenty-first-century post-truth delinquents feel they have made a groundbreaking discovery about the origin-less-ness of discourses, it has long been shown that modern discourses cannot meet their own standards in terms of origins and grounding.²⁹ The same holds for international law, and even more so for CIL.³⁰ That does not mean, however, that CIL, or all the discourses that cannot meet their own standards of origin and grounding, ought to be derided, disregarded or vandalised.³¹ On the contrary, a discourse should be appreciated for how it does what it does, and especially for its origin-less and untraced performances. In that sense, it is once one is liberated from foundational debates about CIL that one can measure and appreciate both the discursive splendour and the efficacy of the latter.

3 Misinterpreting Customary International Law Corrupt Pedigree or Self-Fulfilling Prophecy?

Noora Arajärvi

1 Introduction

This chapter explores the misinterpretation of customary international law (CIL) and its practical and normative consequences. I focus on three main questions: (1) what is misinterpretation? (2) how and why do different misinterpretations take place? and (3) what are the potential consequences of misinterpretation of CIL? These all converge in the underlying question of whether there are detectable objective standards for the determination of misinterpretation or whether such observation is a subjective one – anchored on a disagreement on the values which lie at the core of international law.

In exploring these questions, I combine doctrinal study with empirical examples. Additionally, the different potential consequences of misinterpretation call for a normative evaluation: whether misinterpretation renders a norm invalid or illustrates lex ferenda? Could misinterpretation create an authoritative verdict of the status of law, even against its flawed premise – ‘a corrupt pedigree’, a term coined by Fernando Teson?¹ And why does pedigree matter for CIL – is there something beyond institutional formality conferring authority and legitimacy?

By its nature, CIL is constantly evolving – the customary process is continuous. While in its purest form interpretation of CIL may consist of an analysis of an already ascertained rule (and its elements), interpretation of CIL in most cases inevitably includes an element of construction at that particular point in time – it is difficult to distinguish between the formation, identification and interpretation of CIL.² As Anthony D’Amato has noted in relation to CIL: ‘there is an interrelation between law-formation and law interpretation’.³

According to traditional methodology, CIL emerges spontaneously ‘like a path in a forest’.⁴ It has been suggested, somewhat convincingly, that the identification and interpretation of CIL have taken a strategic turn, potentially arising from the proliferation of international interactions and norm-interpreters and -entrepreneurs.⁵ The theories of ‘modern CIL’⁶ have attempted to explain and justify the broadened methodology, which utilises CIL to advance political, ethical, economic and other aims. Some of such attempts may in fact encourage and expand potential misinterpretations of CIL, with reliance and application of elements far removed from the common understanding of state practice and opinio juris. The effect is not relevant only in the methodology but also in the outcomes: with the utilisation of different interpretative methodologies by different courts and other norm-interpreters,⁷ the resulting identification of a rule of CIL and/or its subsequent interpretation could be highly inaccurate, due to either a genuine mistake in the interpretive methodology or an aspiration to apply a progressive norm disguised as customary rule for moral, ethical, policy, or other extra-legal reasons.

Following the ‘CIL as a path’ metaphor, the interpreter of a norm may misidentify, say, a dried-up stream as a path, designate a minor trail as a fully-fledged path, find a path where there is none, or call a man-made walkway a path. These present examples of misinterpretations without delving into their underlying motives. It is, however, useful to analyse reasons behind a misinterpretation as they may have a direct bearing on the consequences flowing from it, how it is received and responded to by the international community, and for the determination of whether it is representative of ‘a corrupt pedigree’ or a matter of sluggish methodology.

While Articles 31 to 33 of the Vienna Convention on the Law of Treaties (VCLT) serve as the starting point for addressing interpretation in international law – and these rules have crystallised as part of CIL in their own right⁸ – they provide little guidance on how to interpret CIL rules. Just as well, courts have adopted and adapted their own approaches and methodologies on how to go on about interpreting CIL, and not always consistently even within the same institution. In this chapter I intentionally avoid delving into the discussion of what constitutes interpretation: this is accomplished by other authors in this volume.⁹ One may criticise my approach for cutting corners or reversing the analysis while pursuing exactly the same result: exploring what interpretation is not. I accept that the critique may be warranted. The purpose here is, however, not to provide an ample understanding of the misinterpretation of CIL but to initiate conversation on how interpretation may go awry and what it may do to the validity and legitimacy of CIL.

Capturing the definition and examples of misinterpretation is like chasing a moving target – as with interpretation, the elements may be in flux, the circumstances and narratives changing, and the line between genuine and fake CIL – and correct interpretation and misinterpretation – often fluid. With CIL identification (and possibly subsequent interpretation), one can usually find evidence to support what one is looking for – but so can the opposite party. Transplanting a correct interpretation reached at a given point in time into a later case may in fact provide the very premise for misinterpretation even when the methodology of the initial interpretation has been accurate per se: the act of interpreting CIL requires the interpreter to analyse the practice and opinio juris at a specific point in time. By definition, CIL can develop continually and therefore the interpreter needs to look beyond the matter or dispute at hand to get a broader vision of the applicable evidence of the elements of CIL. This argument runs somewhat parallel to Article 30 of the VCLT on the application of successive treaties relating to the same subject matter and to its Article 31 (3) (a) and (b) on subsequent agreement and practice in the interpretation of a treaty: the commentaries call for consideration of what is appropriate in particular circumstances and for caution in resorting to effective interpretation, noting that ‘even when a possible occasion for [principles and maxims’] application may appear to exist, their application is not automatic but depends on the conviction of the interpreter that it is appropriate in the particular circumstances of the case’,¹⁰ and that to resort to extensive or liberal interpretation ‘might encourage attempts to extend the meaning of treaties illegitimately on the basis of the so-called principle of “effective interpretation”’.¹¹ The commentaries also touch upon consequences of such extensive interpretation, with a reference to the 1950 Interpretation of Peace Treaties Advisory Opinion of the International Court of Justice (ICJ), where the court emphasised that to adopt an interpretation which ran counter to the clear meaning of the terms would not be to interpret but to revise the treaty.¹² In a similar manner, to interpret CIL ‘effectively’ or in a way that runs counter to established practice and opinio juris, could either constitute a revision of the rule of CIL (if it is shown that new practice and opinio have emerged) or, as is the focus here, result in misinterpretation (if practice and opinio do not sufficiently support the new interpretation). As noted by the arbitrator in the Island of Palmas case: ‘The same principle which subjects the act creative of a right to the law in force at the time the right arises, demands that the existence of the right, in other words its continued manifestation, shall follow the conditions required by the evolution of law.’¹³

Moreover, methodological questions on whether a solid definition of ‘interpretation’ ought to be stipulated in order to address misinterpretation cannot be dismissed. While the purpose of this chapter is not to extensively delve into these questions, they run in the background of this inquiry and occasionally surface; must we pre-determine the conditions of validity of CIL before we can analyse its misreadings? How, by whom and why should the interpretation of CIL rule be deemed invalid? What can the examples of misinterpretation of CIL tell us about the rules of interpretation?

In discussing the substance of CIL, I use ‘norms’ when the legal validity is uncertain or they appear in a space of conceptual ambiguity – is it CIL or not, is it a legal rule or simply social practice or aspiration? ‘Rules’ refer to those norms which have, to the best of our knowledge and assessment, materialised or crystallised as a part of CIL.

2 From Methods of Interpretation to Misinterpretation

Interpretative exercises in customary international law have been described as ‘methodological mayhem’ resting on the flexibility of methodological uncertainty,¹⁴ and creating an environment of ontological doubt.¹⁵ The same goes for misinterpretation. The orthodox purpose of interpretation is to clarify the intentions of parties. While in treaty law this may be a feasible – if not an easy – task, in CIL identifying and clarifying the intentions of parties is practically impossible, not least for the absence of records of travaux preparatoires. This may depend, however, on the theory of formation of CIL: whether one accepts that CIL forms ‘like a path in the forest’ or whether CIL may arise through a focused, intended and continued practice of actors in international law. In the latter occasion, tracing cognisant practices may be possible, even if based on speculation.

As we know, in the North Sea Continental Shelf cases¹⁶ the ICJ evaluated the basic parameters for CIL based on Article 38 (1) (b) of the ICJ Statute, which have become the reference point for the traditional account of CIL. The ICJ articulated the elements of CIL as follows: ‘Not only must the acts concerned amount to a settled practice, but they must also be such, or be carried out in such a way, as to be evidence of a belief that this practice is rendered obligatory by the existence of a rule of law requiring it.’¹⁷ This was adapted in the Nicaragua case,¹⁸ which saw a more flexible approach, in particular regarding the relationship and the chronological order of emergence of practice and opinio juris. The ICJ reaffirmed the two-element approach in the Jurisdictional Immunities of the State case, stating, with a reference to its previous case law,¹⁹ that ‘the existence of a rule of customary international law requires that there be “a settled practice” together with opinio juris’.²⁰ Now, beyond the basic identification of CIL and its elements, the ICJ has not contributed a great deal to the science of interpretation of CIL, nor – luckily – does it offer obvious examples of misinterpretation either,²¹ although some have claimed that ‘the identification practice of the International Court of Justice for customary norms deviates from the traditional definition of customary law in Art. 38 (1) lit. b of the ICJ Statute’.²² It has been suggested that ‘progressive determinations of CIL [by courts] are generally unproblematic when States are in the dock [as opposed to international criminal law proceedings]’.²³ This statement is simplistic and misleading. First, determinations and interpretations of CIL, especially when delivered by a court with a high authority such as the ICJ, unavoidably influence not only the development of international law but also methods and techniques on how CIL is interpreted subsequently by other international and national courts. Second, while the principle of legality has a heightened relevance in (international) criminal proceedings,²⁴ it is not redundant in inter-state adjudication. In addition, the ICJ – and any other court deciding inter-state cases – has a duty to uphold core tenets of the rule of law, such as consistency, predictability and non-arbitrariness. Adopting progressive determinations of CIL has the potentiality to obstruct these fundamental principles in international judicial decision-making, regardless of whether it is states or individuals in the dock.

Some have argued that the Draft Conclusions on Identification of Customary International Law by the International Law Commission (ILC) constitute a ‘statement of the principles guiding the interpretation of CIL’.²⁵ This characterisation is inaccurate. The difference between ‘formation’ and ‘identification’ has been discussed at length in the ILC’s Special Rapporteur’s First and Second Reports on CIL,²⁶ but ‘interpretation’ has simply not received similar attention; while the work of Sir Michael Wood is invaluable in setting out the issues, controversies and principles for the identification of CIL, it does little to inform the interpretation of existing rules of CIL. This caveat was highlighted also, inter alia, in the Comments and Observations by the Government of the Netherlands to the Draft Conclusions in 2018, which notes that ‘it does not become clear whether the process for identifying the existence of a rule is the same as the process for determining the content of that rule’.²⁷ As has been discussed in this chapter and extensively elsewhere over the course of the lifespan of international legal scholarship, the process of CIL interpretation, nevertheless, overlaps with the process of identification in a complex manner.

In my previous work,²⁸ I have addressed different categories of identification and interpretation of CIL: First, courts may find customary international law by employing the traditional method of assessing state practice supported by opinio juris. Second, they may place more weight on opinio juris over practice – often in this context understood to include a broad spectrum of different considerations. Third, they may deduce customary rules from treaties, national legislation and other (binding or nonbinding) documents. And fourth, courts may refer to previous case law as a confirmation of the customary status of a norm, without in fact assessing the actual findings of practice and opinio juris at that point in time.²⁹ The main criticism regarding the CIL methodology concerns the lack of proper analysis of the elements of custom: courts rendering assertions without justifications, either intentionally, negligently, or, as often seems to be the case, rather casually. This serves as a background for the ensuing analysis of misinterpretation of CIL, which bridges the methodological considerations with the underlying rationales. Identifying examples of misinterpretation is to a large extent related to the way one perceives the functions and limits of international law. For those adopting traditional reading of CIL,³⁰ many more cases of misinterpretation may be detectable than to those with leanings towards the ‘modern approach’³¹ or ‘the sliding scale approach’.³²

There are two dimensions to misinterpretation. Misinterpretation can refer to, on the one hand, to the process and the outcome of the process, and on the other, to the law ascertainment and content determination:³³

1. (1) The act of misinterpretation and its motives – affects institutional legitimacy.

2. (2) The substance of the misinterpretation and its consequences – affects the validity of the norm, depending on its reception by relevant actors, and hence, depends on the conditions of validity imposed by the normative framework.

While misinterpretation has an inherently negative sound to it, it can be a necessary stage in the development and normative change of CIL rules. For an existing CIL rule to change, the practice and/or opinio juris ought to differ from, contradict or go beyond lex lata, providing initially a consideration of lex ferenda. Now, even when practice and/or opinio juris may not have changed or the evidence thereof is mixed, the interpreters (usually courts) may pitch in to spearhead the change in a catalytic manner: such activity may constitute an example of misinterpretation of CIL, closely related to the misidentification and misevaluation³⁴ of the elements of CIL and leading to a potential misrepresentation of a norm. The interpreters may find evidence of practice and/or opinio juris where there is none, exaggerate their prevalence and impact, or declare a norm as CIL without further ado. The breadth, depth, scope and applicability of CIL may be incorrectly set out: for instance, a regional custom may be (mis)interpreted as universally applicable, a general principle of law may be mistakenly awarded customary status, or a jus cogens norm may be characterised as CIL even in the absence of widespread and consistent practice.³⁵ The same could occur in reverse: downplaying practice and/or opinio juris, to hinder the emergence of an undesired rule of CIL even when the elements would point to its crystallisation.

Does the finding of misinterpretation presuppose a cognisant misinterpreter? No: misinterpretation by definition is not concerned with motivations, but it simply refers to ‘the act of forming a wrong understanding of something that is said or done, or an example of a wrong understanding’.³⁶ In any case, evidence of deliberate misinterpretation of CIL is rare and mostly misinterpreters have adopted a lazy methodology or ignored rules of interpretation in evaluating practice and/or opinio juris.

When analysing the notion of misinterpretation in CIL, we can also break it down to the elements: is it the CIL rule as a whole, or practice or opinio juris that is being misinterpreted? What is the relative relevance of misinterpreting practice or opinio juris? Misinterpreting the element of practice – being usually more quantifiable – may be more obvious than that of the more fluid, subjective, element of opinio juris. It may also be more consequential, as it is viewed – at least by many of us – as the very bedrock of custom.

Now, it is not terribly difficult to find cases of national courts taking liberties in their interpretation of the concept and rules of CIL,³⁷ although most often these courts ‘simply assert, without citing persuasive practice authority, the existence of a customary norm’.³⁸ At the international level one may expect to see more cross-referencing and recognise potential consequences of misinterpretation for international law. The next section sets out categories, with selected examples, where international courts and tribunals have overstepped methodological limits to the effect that may constitute misinterpretation.

3 Categories of Misinterpretation

Based on my initial research findings, three types of misinterpretation are identifiable. First, the extension (or reduction) of CIL through exitus acta probat – the end-justifies-the-means approach – where analysis of the elements of CIL is modified to fit the desired outcome and the elements are substituted or complemented with resort to extra-legal tools and concepts. This approach often finds support among the more modern-liberal theories of CIL. Second, I have identified the negligent interpretation, which may amount to misinterpretation when the norm-interpreter labels a norm as CIL without further analysis of the elements and where in fact opposite practice and opinio juris might be observable. Finally – and luckily evidence of this remains scarce – there is the fallacious method of misinterpretation, where the interpreter finds false CIL or considers flawed or incomplete evidence of its elements. All the three categories contain overlapping dimensions – it may be hard to distinguish whether the norm-interpreter was merely negligent or plain wrong, or where the end-justifies-the-means approach crosses over to the delivery of fallacious interpretation. In particular, regarding the latter example, ideological leanings may cloud the legal astuteness of determining whether the interpreter has acted in good faith or not, and wherein lies the line between an actual legal error and consequentialist bending of the rules of interpretation to achieve a morally desirable outcome. Also, a fallacious interpretation of CIL has a much higher chance of success to flourish through subsequent interpretations and practice when it leads to ‘good’ outcomes – for instance, an ‘effective’ interpretation expanding the scope of a human right can be expected to be met with more praise than an argument to the opposite effect.

A misinterpretation may be discoverable in subsequent proceedings by the same or another court. In May 2010, the Extraordinary Chambers in the Courts of Cambodia (ECCC) held that the mode of responsibility of Joint Criminal Enterprise (JCE III)³⁹ did not exist under CIL in 1975–79, and consequently was not applicable in the proceedings of that court.⁴⁰ The ECCC limited JCE III by declaring that there was not enough evidence of its customary nature, at least not in 1975–79, thus dismissing the ICTY’s argumentation in Tadić by illustrating that the Tadić court had in fact invented that category of criminal responsibility.⁴¹ In analysing the concept of JCE, the ECCC first noted that it must consider ‘not only whether JCE existed under customary international law at the relevant time, thus being punishable under international criminal law, but also whether it was sufficiently foreseeable and accessible to the Charged Persons’.⁴² It then examined the findings in Tadić, other ICTY cases, and case law dealing with the crimes committed in World War II, stating in relation to JCE III that ‘[h]aving reviewed the authorities relied upon by Tadić in relation to the extended form of JCE (JCE III), the Pre-Trial Chamber is of the view that they do not provide sufficient evidence of consistent State practice or opinio juris at the time relevant to Case 002’.⁴³ This approach was further confirmed in November 2016 by the Supreme Court Chamber of the ECCC.⁴⁴

Earlier the same year, however, the Appeals Chamber of the ICTY reconfirmed its interpretation of JCE III as a mode of liability under CIL by noting that ‘the third category of joint criminal enterprise has existed as a mode of liability in customary international law since at least 1992 and that it applies to all crimes consistently confirmed in the Tribunal’s subsequent jurisprudence’.⁴⁵ In responding to the defendant’s challenge of the customary nature of JCE III, the Appeals Chamber stated that ‘this contention is essentially premised on his suggestion to depart from the existing jurisprudence on the basis of his misconstruction of the law’.⁴⁶ This could be viewed as the Appeals Chamber’s reaction to the debate surrounding JCE III and presents a clear expression of its position vis-à-vis interpretation of CIL by the ECCC and many scholars.

Even when acknowledging that nearly twenty years passed between the commission of crimes in Cambodia and in the former Yugoslavia – and the applicable rule of CIL needs to be determined with reference to those points in time – these cases nonetheless show CIL’s ambiguity and the challenges it poses to interpretation. The drastic departure of the ECCC from the ICTY jurisprudence brings uncertainty on the actual status of the rule and raises the question of the implications of such diverse interpretations for future cases dealing with modes of criminal liability.⁴⁷ This goes on to illustrate that there may not be objectively one right answer to how and what to interpret. The method and consequently the outcome may depend on the interpreter’s approach to CIL: which element carries the most weight and which evidence is included in the assessment of practice and/or opinio juris. Both interpretations of the status of JCE III can be objectionable on these grounds. As noted by Verdier and Voeten, ‘an attempt to justify a breach by reference to the rule’s ambiguity is likely to be interpreted as a violation by the counterparty and (some) third parties’.⁴⁸ Regarding the treatment of the JCE III neither of the tribunals can be accused of sluggish methodology even if one could be persuaded to view ICTY’s early judicial activism as falling into the exitus acta probat category. The interpretation of JCE III as CIL has, however, been repeated in the jurisprudence of the ICTY in numerous subsequent cases. Below, under ‘Consequences of Misinterpretation’, I will discuss the relevance of repetitive judicial practice in the context of potentially ‘corrupt pedigree’ of CIL.

Similar discourse involving possible misinterpretations of CIL took place internally between different chambers of the ICTY and the Special Court for Sierra Leone (SCSL), in relation to the requirement of specific direction as a part of the actus reus of aiding and abetting.⁴⁹ The ICTY and the SCSL, within a space of less than a year, interpreted this requirement under CIL reaching opposite outcomes: Whilst the ICTY Appeals Chamber held in Perisić that specific direction is a part of actus reus of aiding and abetting, it did not make any explicit reference to its status under CIL. The SCSL Appeals Chamber in Taylor, on the other hand, stated that:

[i]n the absence of any discussion of customary international law, it is presumed that the ICTY Appeals Chamber in Perisić was only identifying and applying internally binding precedent. … [T]he ICTY Appeals Chamber’s jurisprudence does not contain a clear, detailed analysis of the authorities supporting the conclusion that specific direction is an element of the actus reus of aiding and abetting liability under customary international law.⁵⁰

In effect, the SCSL called out the ICTY Appeals Chamber’s misinterpretation of CIL. The ICTY Appeals Chamber, in a decision shortly after, adopted the SCSL interpretation, mending its own prior misinterpretation.⁵¹ Hence, this is an example of the second category of misinterpretation where the ICTY had discussed a rule, presuming it part of CIL without a further analysis of its elements and when in fact a further analysis would have uncovered opposite practice and opinio juris.

Likewise, an example of misinterpretation, subsequently identified by another authority, can be found in the EU case law. The rules which constitute an expression of customary international law are binding upon the EU institutions and form part of the EU legal order and as such: ‘[CIL] is regularly interpreted and applied by the Court as an “integral part” of EU law’.⁵² De Burca has observed that ‘CIL was cited by the CJEU in twenty-one cases’ (as of October 2015).⁵³ A recent search on EUR-LEX reveals that the number stands now at thirty.⁵⁴ Interestingly, the Court of Justice of the European Union and the Advocate General have made some remarks about misinterpretation of CIL. For example, without delving further into EU case law, in Front Polisario, the Advocate General considered and accepted⁵⁵ the argument put forth by the council and commission that ‘the General Court misinterpreted customary international law, as it did not cite any legal basis requiring the EU institutions to verify that the other party to the agreement has complied with the principle of permanent sovereignty over natural resources and the primacy of the interests of the inhabitants of non-self-governing territories’.⁵⁶

A striking example of a misinterpretation of CIL – or misidentification as the limits may be fluid – through flawed methodology undercutting the element of practice took place in the ICTY Trial Chamber judgment Prosecutor v. Kupreškić, discussing the prohibition of reprisal attacks against civilians:

Admittedly, there does not seem to have emerged recently a body of State practice consistently supporting the proposition that one of the elements of custom, namely usus or diuturnitas has taken shape. This is however an area where opinio iuris sive necessitatis may play a much greater role than usus, as a result of the aforementioned Martens Clause. In the light of the way States and courts have implemented it, this Clause shows that principles of international humanitarian law may emerge through a customary process under the pressure of the demands of humanity or the dictates of public conscience, even where State practice is scant or inconsistent.⁵⁷

The court, going against the traditional understanding of CIL and its elements, allowed for inconsistent practice to suffice in its finding of CIL and suggested that opinio juris is of a higher value than state practice. It relied almost exclusively on opinio juris in its interpretation of the CIL rule on reprisals, while simultaneously broadening the internal nature of opinio juris and its limits to include extra-legal considerations, such as ‘elementary considerations of humanity’, which, the ICTY considered, should be fully used when interpreting and applying loose (customary) international rules.⁵⁸ It further noted a customary rule of international law had emerged ‘due to the pressure exerted by the requirements of humanity and the dictates of public conscience’.⁵⁹

Such misinterpretation originates, most likely, from two objectives: on the one hand, the court saw evidence of a horrific act which was, however, not explicitly covered by the rules of international law at the time and felt a moral duty to rectify this – to bring the perpetrators to account for their actions, to deliver justice for the victims, to contribute to deterring future atrocities and to enhance international criminal law as a social pedagogical imperative, and, possibly, to contribute to the development of the law. At the same time, the mandate of the ICTY limits its jurisdiction to the application of international humanitarian law that is ‘beyond any doubt part of customary law so that the problem of adherence of some but not all States to specific conventions does not arise’.⁶⁰ Hence, the ICTY could only apply law that it considered to have already crystallised as CIL, which then led it down the path of broadening the methodology of identifying and interpreting norms so as to fit them under the umbrella of CIL – at times like in Kupreškić amounting to misinterpretation of CIL at the intersection of an exitus acta probat (end-justifies-the-means) approach and the intentionally deceiving method of misinterpretation.

Another example of a potential misinterpretation of CIL stems from the 6 May 2019 judgment of the Appeals Chamber on the International Criminal Court (ICC) in the Jordan Referral in the Al Bashir case. The court discussed the customary status of Article 27 (2) of the Rome Statute and concluded that Head of State immunity under customary international law does not apply in international courts and tribunals.⁶¹ It stated that: ‘[t]here is neither State practice nor opinio juris that would support the existence of Head of State immunity under customary international law vis-à-vis an international court. To the contrary, such immunity has never been recognised in international law as a bar to the jurisdiction of an international court.’⁶² The decision was not unanimous and in their joint dissenting opinion Judges Luz Del Carmen Ibáñez Carranza and Solomy Balungi Bossa stated that ‘[i]t is thus clear that the international community as a whole has consistently rejected the invocation of Head of State immunity for the commission of international crimes’, continuing somewhat contentiously, that ‘[u]nder customary international law, immunity can never result in impunity for grave violations of the core values consolidated in international human rights law’.⁶³

Despite the ICC’s approach that there is no CIL rule supporting the applicability of immunities for international crimes in an international court, there is no general understanding on the status of this rule, as illustrated by the several amicus curiae briefs submitted on the issue at the ICC, strong political resistance from the African Union,⁶⁴ and some previous jurisprudence of the court itself. For instance, in the South Africa Decision in the Al Bashir case, the ICC Pre-Trial Chamber II held that ‘[t]he Chamber is unable to identify a rule in customary international law that would exclude immunity for Heads of State when their arrest is sought for international crimes by another state, even when the arrest is sought on behalf of an international court’.⁶⁵ Further to this, the practice of states on the matter remains far from consistent, constant and uniform, and expressions of opinio juris are few and far between.⁶⁶

The Al-Bashir case and its discussion on immunities touches on a core tenet of interpretation: it all hinges on the expectations. What tasks and results should international courts, other norm-interpreters and international law in general, deliver? For instance, while many scholars debate the ICC’s recent decision on Heads of State immunity under CIL,⁶⁷ advocates for ending impunity, human rights organisations and several scholars have cheered at the decision, viewing it as very much the correct interpretation of CIL of immunities in international tribunals.⁶⁸ This is to show how different levels and categories of misinterpretation will be most definitely welcomed by one audience or another. Interpreters, of course, are aware of this and can strategically adjust the method of interpretation and, consequently, the ensuing norm, to address the target audience.⁶⁹ In this context, Andrea Bianchi has noted that ‘interpretive strategies [are] adaptable and flexible enough to serve different purposes’.⁷⁰

Finally, misinterpretations may be politically motivated with little legal foundation or objective. They may be just plain wrong, with flawed methodology and conclusion. As noted above, examples of such outrightly erroneous misinterpretation of CIL at the international level remain scarce. One example could be devised from the torture debate in the early 2000s in the USA.⁷¹ Similarly, in vain, a few scholars tried to argue in 2003 that the Bush Doctrine – as some still recall meaning preventative war with pre-emptive strikes and so on – had developed into CIL, or even into instant CIL.⁷² These misinterpretations did not come from internationally authoritative sources nor did they do much more than stir some debate and yield material for scholarly articles. They remain cautionary tales of getting CIL wrong but also attest to the resilience of traditional sources theory and the approach of international legal scholarship in preserving the core of what constitutes CIL, even when recognising that due to their morally distasteful nature they were bound to invoke strong opposition.

4 Consequences of Misinterpretation

The implications and consequences of misinterpretation vary depending on the original form of misinterpretation. Corrupt pedigree and self-fulfilling prophecy are by no means mutually exclusive but may complement one another in a fertile environment. They may be reflections of different stages of the customary process: originating from a corrupt pedigree, but resulting in a self-fulfilling prophecy when viewed with subsequent hindsight. What, then, are the consequences of misinterpretation?

Practice – as some examples set out in Section 3 and many in this volume illustrate – suggests that the interpretative rules of CIL may be fluid and open to interpretation themselves (just like Articles 31–33 VCLT have been given various interpretations). Hence, the methods do differ and even when employing the same set of methods or rules of interpretation, different interpreters may reach different outcomes. Interpretation is inevitably connected to cognitive frames and social needs,⁷³ and extra-legal considerations are omnipresent and impact the process and result. It depends, not least, on the interpreter’s position towards CIL how explicitly these considerations inform his or her interpretative methodology.

Fernando Teson offers Nicaragua’s determination of the rule of non-intervention as an example of CIL based on corrupt pedigree, in which the court failed to cite practice, precedent or consensus.⁷⁴ The ensuing practice, precedents and consensus consolidate and perpetuate the legal error, and themselves become sort of precedents creating a corrupted chain of legal justification. As opposed to Teson, who claims that the only kind of fake custom that has the power to generate genuine custom is the false legal statements made by states if they are then clearly followed widely by the international community,⁷⁵ I argue that the same must go for decisions of international courts, but only if subsequent practice confirms the rule. This idea can be implicitly found also in the ILC Draft Conclusions, which note that while the practice of international courts and tribunals is not state practice, pleadings by states in those forums can be.⁷⁶ Hence, when a state would refer to the case law containing the ‘corrupt pedigree’ in front of a court, this could be viewed as constituting such a confirmatory subsequent practice. As Harlan Cohen notes, ‘precedent must be understood within practice [– or community of practice –] of international law [and its] force derives solely from the desirability of the rule reflected in it’.⁷⁷ If the precedent is not cited, followed or endorsed in any way, it has very little authority on its own. At best, it can produce a strong presumption that the interpretation is in fact the rule, creating a compliance-pull. This depends on multiple factors: the quality of legal reasoning, the clarity of interpretation, adherence to prior interpretations, and how well the interpretation fits within the broader legal framework, the aspirations of the parties and the potential burden it imposes on those parties.⁷⁸ Thus, even with corrupt pedigree, a norm may under favourable conditions eventually spawn into real custom.

Antonio Cassese has suggested that acquiescence to a misinterpretation would have the same effect as affirmation through practice and consensus, as is the case with the formation of CIL⁷⁹ – where ‘silence equals consent’. I am sceptical of this position for it sidesteps any requirement of actual practice, basing the existence of CIL merely on ‘combination of a string of decisions … coupled with the implicit acceptance or acquiescence of all the international subjects concerned’.⁸⁰ Teson argues against such flex CIL methodology, stating that ‘citing a multitude of non-binding documents does not turn a proposed norm into a binding customary norm because is neither anchored in state practice nor is the object of a universal and specific consensus’.⁸¹ Repetition alone does not confer normativity or legitimacy! In order to sustain a level of legitimacy, the interpreter cannot exclusively refer to their own practice and create a cyclic self-asserting method of interpretation – an external confirmation or affirmation is required, even when the end result may be what I refer to as a self-fulfilling prophecy. While it may be likely that with multiple decisions discussing the same rule at least some would enter a detailed analysis of practice and opinio juris, we cannot base a coherent conceptual analysis of CIL merely on judicial practice, neglecting evidence of practice or its absence. This would place a completely new burden on states to react to decisions of courts (and potentially other norm-interpreters) in order to ensure they will not be bound by the so-called CIL arising from their decisions.

Subsequent state practice and consensus can legitimise the misinterpretation, which has then served as a catalyst for change. When this is the case, the discovery of past misinterpretation – classifying a norm under CIL when it has not yet so crystallised – does not denounce its subsequent normative validity, if it has been followed as if it were already part of CIL. So, in determining the consequences of misinterpretation, we must go back to the roots of CIL before and after the act of misinterpretation, to look at the practice and opinio juris, and to assess whether sufficient affirmation exists which renders the legal basis of the CIL rule. This affirmative consequence of misinterpretation was noted already at the International Military Tribunal for the Far East in 1953, with Judge Pal stating that ‘law also can be created illegally otherwise than by the recognized procedures … any law created in this manner and applied will perhaps be the law henceforth’.⁸² Misinterpretation can, naturally, lead to positive as well as negative outcomes. It may be that a progressive approach and dynamic interpretation, even when considered incorrect either methodologically or substantively, directs the development of practices and beliefs towards more just processes and outcomes.⁸³ Even if the rule was not customary at the initial point, the subsequent practice may override the initially faulty interpretation as the norm gains wider usage, which is supported by illustrations of opinio juris.⁸⁴

5 Concluding Remarks

Is it possible to avoid misinterpretation in international law and if not, how can the negative consequences be mitigated? For CIL to develop, some instances of misinterpretation may be inevitable, serving as test cases on what states and other actors perceive to be the acceptable limits of the law at a given time. Determining the relevance of misinterpretation is a retrospective exercise, and in that sense, if courts deliberately enter unknown, uncertain or even outright incorrect legal terrain, they risk facing accusations of judicial activism or ‘effective interpretation’, which may affect their institutional legitimacy; at the same time, if all plays out well, their (mis)interpretation may instigate the crystallisation of a new (and maybe better!) CIL rule.

Identifying misinterpretation is a challenging task, which depends on the underlying approach to CIL. If one concedes that CIL forms through various ways and that its elements rest on a broad range of evidence beyond settled understanding of what amounts to state practice and opinio juris, one is likely to also accept the wide-ranging methods of interpretation of CIL – and consequently, find less occurrences of misinterpretation. To remain faithful to the traditional notion of CIL – which is embodied in practice – and to preserve legal certainty and predictability, it is crucial to recognise that courts are not infallible, sometimes lacking the requisite methodological tools, and occasionally just getting CIL wrong. The indeterminacy of CIL renders futile the attempts to pin down precise conditions for its validity, and simultaneously, leaves open the definition of misinterpretation of CIL. Misinterpretation, in general, diminishes foreseeability and consistency of the law. It may also, however, push the actors to develop the law. With the inescapable constructive dimension of CIL, courts implicitly serve a key function in the development of CIL through their interpretations and their interpretative methodology.

The misinterpretation of a customary norm, which is subsequently followed by states and other entities as if it were part of CIL, creates a self-fulfilling prophecy – a self-generating crystallisation of a rule. Even if the rule was not customary law embedded in practice and opinio juris at its ‘inception’, the subsequent practice and acceptance eradicates the mishap of the initial faulty interpretation and legitimises the rule as part of CIL. On the other hand, as examples from international criminal tribunals illustrate, a later decision may denounce the misinterpretation and correct the course of the customary process and norm development, or the misinterpretation will remain an unfortunate but soon forgotten misstep, neither to be restored nor repeated.

4 The Logic of Absence in Customary International Law An Open-System Approach

Anna Irene Baka

1 Introduction

Customary international law (CIL) bears an ab initio element of absence and thus abstractness: the lack of written formality, which, as such, can spur multitudinous interpretative debates. The profound ambiguity surrounding all elements of CIL particularly as regards the subjective, psychological element of opinio juris is further accentuated by the prevailing element of absence, silence or non-action and their often-monolithic interpretation as non-objection or, even, acquiescence. There also appears to be a fundamental presumption against the existence of semantic voids in CIL – a presumption that attaches negativity to silence and positive value to affirmative propositions. Indeed, negative premises appear to be less valuable and less informative than affirmative ones, while affirmatives are given semantic priority and added value over negatives. But is that true, according to the rules of informal logic? If a positive statement corresponds to a positive affirmation, to what state of affairs does a non-statement refer or correspond? What is a negative fact? What is a non-fact? What is the value of non-doing? Non-acting or abstaining? Non-believing towards the formation of a certain opinio juris? Is every absence, or negation, necessarily a denial of a state of affairs?

International law does not provide any clear guidance as regards the legal effects that follow from state silence. This produces further difficulties with the polysemous nature of silence, which may have several meanings, from tacit agreement to absence of view or simple lack of interest. The legal positivist eagerness to evaluate and attach negativity to absence has its roots, on the one hand, in the Wittgenstenian, contextual and consensual origins of legal positivism, assumed in HLA Hart’s theory and his subsequent rejection of metaphysics, that is the premise that there is no meaning outside communitarian semiotics.¹ On the other hand, Kelsen’s Grundnorm theory assumes a complete normative order consisting only of positive norms, even if those positive norms are negatively deduced.² However, according to the rules of logic and the canons of reasoning, absence may correspond to multiple values, a variety of propositions and modalities, which in international jurisprudence have been either equated or largely ignored. The mainstream interpretation of CIL overlooks the quantifications and varieties of meaning in non-appearances, such as the conceivable neutrality of absence.

The modalities of absence are not mere academic exercises. They affect the rationality and soundness of international legal doctrine and even have a real impact on international relations when overlooked. This repositions the whole enquiry to the proper place of informal logic in international legal and judicial reasoning. The chapter suggests that the rational deficit in international legal reasoning has led to, or has been enhanced by, persuasive-teleological argumentation, in the sense that the person or agency elaborating on silence aims at a certain end and is thus characterised by a certain ‘argumentative orientation’ towards a preferred conclusion. In this spirit, the ICJ has developed several techniques of superficial, persuasive argumentation, teleologically governed by the non liquet principle, the containment of international crises and the effective resolution of international disputes. This is a form of judicial interventionism, further accentuated by the demonstrated judicial or scholarly difficulty to ‘translate’ silence and/or the absence of state practice by virtue of some justification that transcends a particular case, is intrinsic to the legal system and is construed logically, that is by virtue of specialised rules of deductive thought which rely on a highly logical systemisation. An open-system approach could shed light on these inconsistencies and/or political manoeuvres.

2 Setting Up the Standards: Is International Legal Reasoning a Scientific Method of Reasoning?

The answer to this question necessitates a twofold examination, namely (a) how science and the scientific method of reasoning are generally defined and (b) whether law as a discipline, and international legal reasoning in particular, fit into these definitions.

To respond to the first question, science is traditionally considered to refer to any kind of methodical study that ‘has a definite subject matter, is systemic and comprehensive and … its aim is to discover the truth as far as possible’,³ whereas scientific method ‘is just taking things in order, simplifying as far as necessary and possible, endeavouring to leave out nothing that ought to go in, and distinguishing true from false’.⁴ Science is characterised by systematisation, linguistic and conceptual accuracy, the application of rules of logic and a concrete methodology for the purpose of generating knowledge. The scientific method consists of techniques of ‘argument, conceptual clarification, logic and discussion’⁵ or, if the subject matter is the investigation of natural phenomena, the application of the empirical method, namely qualitative or quantitative techniques and the process of hypothesis testing and verification.⁶ The process of concept formation is an essential part of scientific knowledge, which traditionally consists of a logically ordered, hierarchical pyramid of concepts.⁷ In most disciplines this pyramid takes the form of axioms, principles and derived theories that subsequently produce valid deductive inferences, provided that the working concepts are clear and unambiguous.⁸

Regarding the second question, on the scientificity of international legal reasoning, it is necessary to first inquire into the object and method of law in general. One should first distinguish between such terms as ‘the sociology of law’ or ‘sociological approaches to law’ or ‘socio-legal studies’, and such prepositions as ‘law as a social science’. Whereas the former form part of a distinct discipline that examines law as a social phenomenon⁹ the latter investigates the scientificity of law as such. In universities, law is traditionally classified among the social sciences. It is, however, questionable whether such attribute is accurate. There have been arguments in favour, namely that law ‘is not just a social science but one that is central to social thought in general,’¹⁰ as well as arguments against, that law cannot be categorised as a social science because ‘it is preoccupied with normative judgments and not with human interaction and behaviour’.¹¹ Another argument against law as a social science is that its theories cannot be falsified, according to Karl Popper’s falsification test.¹²

It is useful, at this point, to apply an insider’s approach and investigate how legal theory has dealt with this problem. Kelsen’s formalism and his Pure Theory of Law have separated legal doctrine from social, moral and political theories. In his Concept of Law, HLA Hart has been somewhat less stringent in that he placed particular emphasis on the ‘social functions’ of law and considered his work to be both a legal theory as well as ‘an essay in descriptive sociology’.¹³ He applied a so-called internal and external attitude to law in different contexts and pointed out that there are elements of social psychology behind legal concepts.¹⁴ Dworkin and Raz acknowledged the importance of the social scientific method for the study of legal institutions but both drew a clear line between jurisprudence and ‘legal sociology’ or ‘sociological jurisprudence’ as distinct disciplines.

The dominant view from both the legal and the sociological perspective is that law cannot be considered to be a social science (a) because of its very narrow subject matter which is distinct from the one of sociology and the sociology of law, and (b) because legal doctrine does not apply the traditional methods of the social sciences, that is, the qualitative and quantitative techniques. The narrow reading of law as a closed system of knowledge has attracted serious criticisms due to its isolation from the social and political settings, as well as its autopoietic nature and stringent self-referentiality.¹⁵ Law as a closed, isolated system of knowledge inevitably leads to ‘a body of knowledge [that] has nothing to contribute, epistemologically speaking, to our knowledge of the world as an empirical phenomenon’, whereas it is a narcissistic discipline that ‘is of little interest intellectually speaking to those outside [it], save perhaps to those social scientists interested in studying the corps of lawyers as a social phenomenon itself’.¹⁶

No matter how one looks at it, law is not a social science. From this assertion alone it does not follow, however, that international legal reasoning should not conform to rules of informal logic. The scientific authority of law is a quasi-logical requirement for internal coherence and causality, which, together with other logical principles, have built a closed system of logic.¹⁷ In this closed system of logic deductive inferences are produced from a matrix of consented legal axioms. This process forms the ‘scientific’ authority of the legal syllogism.¹⁸ It is obvious that, from this perspective, legal reasoning is at best quasi-logical.

Apart from the quasi-logical nature of legal reasoning, the ambiguity of language, as well as the various legalist approaches that normally complement the legal syllogism, such as functional, hermeneutical and dialectical approaches, lead to an obscure model of reasoning that is not open to testability.¹⁹ The legal syllogism is complemented by an erratic series of variables that include (a peculiar understanding of) logic, interpretation, functionalism and systematisation, as well as an abstract appeal to general principles such as democracy, legal certainty and the rule of law.²⁰ All these variables attract dialectical instrumentalism, inasmuch as they impose an additional burden to the legal theorist to be consistent ‘with the multitudinous rules’ of legal systems which ‘should [also] make sense when taken together’.²¹ It follows from the above that legal reasoning does not concur with logical reasoning. Then, our initial question needs to be reformulated thus: should this closed system of logic with its demand for coherence operate at the expense of logical rationality? And how is this logical rationality to be measured?

Kelsen himself claimed that law is a normative science and it is necessary for legal norms to be logically explained and connected.²² One can therefore observe a traditional association of law with the requirement of objective-external rationality. And rightly so: without objective standards of logic, legal theory and jurisprudence are contemned to drift into speculation. From the perspective of both legal theory and jurisprudence the requirement of rationality is always relevant. Such requirement, however, cannot be considered fulfilled at the narrow level of internal coherence.

To return to our reformulated question, namely whether law should be governed by something more than the superficial requirement of coherence, the chapter answers in the affirmative. Among the variables that determine the legal syllogism, namely interpretation, systematisation, functionalism and the appeal to abstract principles, the rules of informal logic appear to be the crucial constant, in the mathematical use of the term, which can direct the legal syllogism to rational, that is syllogistically sound, conclusions. It is only by transcending the closed logic of law and the limitations of legal formalism and coherence that the syllogistic credibility of law can be restored. Law cannot be rational if it operates autonomously and self-sufficiently. Law should not operate beyond logic; legal theorists need to resort to the classical understandings of logic in order to avoid superficial formal rationality (which may or may not coincide with logical rationality) as well as unscientific instrumentalist thinking. It has been argued that law has teleology and is both ‘natural, in the sense that it has to be found out and is not made by any arbitrary act of will and rational because it is not solely a fact of observation’.²³

The laws of thought and the so-called canons of reasoning²⁴ operate in accordance with a natural mind process, which is independent from social institutions. Whether such laws of thought can be applied in social settings in the same way that they are applied in natural sciences has been the object of a heated debate, famously initiated by Hume with his fact-value/is-ought distinction. Although the purpose of the chapter is not to get into the details of the debate, it should be nonetheless noted that, even for traditional logicians like John Stuart Mill, the ‘mathematical inexactness’ of the humanities and the social sciences is not a conviction to unscientificity. For Mill ‘whenever it is sufficient to know how the great majority of the human race or of some nation or class of persons will think, feel, and act, these propositions are equivalent to universal ones. For the purposes of political and social science this is sufficient’.²⁵ Likewise, in his Novum Organum, Francis Bacon insisted that his method is applicable to both normative and factual issues alike.²⁶

3 Methods and Techniques for the Interpretation of Silence in ICL by the ICJ: Errors and Inconsistencies

International law does not provide any clear guidance as regards the legal effects that follow from state silence. This produces further difficulties with the interpretation of the polysemous nature of silence, which may have several meanings, from tacit agreement to absence of view or simple lack of interest.

Two scholarly debates are of particular relevance here. The first debate is a systemic one, relating to the nature of international law as a normative system and the question whether it is an open or a closed system of norms. Should we follow the hypothesis that international law is an open system, then one could then fathom the possibility of an absence of law, which could then open up the possibility for a non liquet declaration. For Jörg Kammerhofer the question is whether silence is simply a gap or a gap in law.²⁷ Kammerhofer argues that, since normative systems consist of positive norms, it is unthinkable to have a situation where there is an absence of norms within the legal system.²⁸ Accordingly, when we face a situation of non-regulation in international law, then the Lotus principle, that is, the presumption in favour of state liberty to act, cannot be sustained because this liberty is essentially a factual state of affairs that falls completely outside the normative order.

The second debate relates to the epistemological tools that the international judge has at hand for the adjudication of a case: the inductive and the deductive methods of reasoning. For the purposes of law ascertainment, induction may be defined as empirical generalisation, ‘as inference of a general rule from a pattern of empirically observable individual instances of State practice and opinio juris’.²⁹ The deductive method, on the other hand, is defined as inference of a specific rule from an existing and generally accepted rule or principle, that is, the process of deriving the specific from the general. With the exception of the common law tradition, legal academic reasoning is mostly based on deductive syllogisms, namely the application of general laws and principles to concrete cases. Indeed, both Kelsen’s Grundnorm system and HLA Hart’s model of the Rule of Recognition portray a stringently hierarchical arrangement of axiomatic concepts, which presumably produce a series of safe, deductive inferences. Interestingly, this is not the view of Georg Schwarzenberger who, in his Inductive Approach to International Law, famously praises the application of the inductive method and attacks the, as he says, eclectic and unreliable results of the deductive method of legal reasoning.³⁰ The rationale behind this paradoxical – from the logical point of view – thesis may be summarised as thus: the derivation of lesser axioms, the process of legal interpretation and the application of general principles to concrete cases can all end up being extremely subjectivist and logically misleading for they are unverifiable and often based on speculation and an arbitrary ‘picking up and choosing’ from both natural and positive law.³¹ The speculative and eclectic nature of international legal deduction is, according to Schwarzenberger, underpinned by the obscure positivist borderline between lex lata and lex ferenda.³² With respect to the naturalist approaches to law, Schwarzenberger is equally suspicious and notes that the ‘law-finding’ process of naturalist deduction is often a ‘law-making’ process in disguise.³³ He defends the inductive method of international legal reasoning on grounds that it is an empirical device that secures international legal theory from ‘the subjectivism of deductive speculation and eclectic caprice, and the vested interests prone to use – and abuse – both’.³⁴ He therefore treats all ‘deduction, speculation, or intuition’ as mere hypotheses until they are all inductively verified by reference to the ‘law-creating processes’ and the ‘law-determining agencies’ which are enumerated in Article 38 of the Statute of the ICJ.³⁵ Despite his extensive eulogy to inductive reasoning Schwarzenberger’s formalism does not embrace an unrestricted use of induction in international legal theory. Instead, he submits all logical methods, the inductive method included, to the requirement of consistency and systemic coherence, as well as the standard verification process of ‘the three law-creating processes of international law’ which all come down to the principle of consent.³⁶

Although, in practice, both the inductive and deductive methods are employed in judicial syllogistics, the ICJ rarely states explicitly the methodology that it uses for the determination of CIL, and, as we will see in Section 5, it is often the case that it applies the two methods of reasoning erroneously. In fact, it appears that there is a lot of confusion among jurists and legal theorists vis-à-vis the proper definition and application of the two logical methods of reasoning. It has been argued, for instance, that induction is employed in the application rather than the determination of the applicable law, which is a deviation from the typical definition of induction from the scope of informal logic.³⁷ A justification for this deviation is that logical reasoning should not be equated with legal reasoning, which is governed from an internal logic, a logic of its own. In the same vein, judicial deduction is regarded as not being the same as logical deduction.

There is widespread agreement that CIL is, as a rule of thumb, ascertained by means of induction, since according to the mainstream, or traditional, legal doctrine the two elements of CIL are gathered in an empirical and inductive way. Because this is not a mathematical exercise – and against Schwarzenberger’s theory on the merits of the inductive method in international law – it has been suggested that the application of the inductive method for customary law ascertainment is prone to subjectivity, selectivity and law creation.³⁸ Since it is practically impossible to gather and assess the practice and opinio juris of states, the ascertainment of any customary rule entails a selection that is often ‘supportive of a preconceived rule of customary law’.³⁹ Besides, it is for the ICJ to assess what counts as state practice, what counts as opinio juris, whether the state practice is consistent and uniform etc.

The two above-mentioned debates, that is, the question whether international law is an open or a closed system of norms, as well as the application of induction as the prominent tool for the ascertainment of CIL, intertwine in a new theoretical trend that involves deduction and assertion as alternative, or additional, methods for customary law ascertainment. In this new and ongoing debate there is a distinction between traditional and the so-called modern deductive CIL – a distinction between customary law that results from the traditional, inductive method of reasoning, and customary law that arises in instances where the inductive method is considered ‘impossible to use’ because state practice is non-existent, the legal question is too new and has not been dealt with etc.⁴⁰ In the latter case, it has been assumed that, because international law is a closed system of norms and non liquet is simply not a possibility for the ICJ, international legal theorists and judges are left either with deduction of customary law from other international legal norms and principles or, even, simple assertion of CIL, that is, statements regarding the existence of customary rules that are ungrounded or not properly explained. Both deduction and assertion emphasize opinio juris rather than state practice, and often reveal value judgements.⁴¹

Deduction and assertion are not only limited to ‘positive’ customary rules, but also the negation or absence of custom, where the ICJ simply denies the existence of customary law due to the, presumably, lack of (uniform) state practice and/or opinio juris. This is particularly true for cases of omission, abstention and absence of either state practice or opinio juris. One could recall, for instance, the Gulf of Maine case, where it was held that the lack of state practice precludes the formation of a customary rule.⁴² In the North Continental Shelf cases, Judge Sørensen argued that ‘[i]n view of the manner in which international relations are conducted, there may be numerous cases in which it is practically impossible for one government to produce conclusive evidence of the motives which have prompted the action and policy of other governments’.⁴³ In other words, whereas the absence of state practice and/or opinio juris ‘discourages’ the effective application of the inductive method of reasoning as a tool for the ascertainment of CIL, the new category of the so-called deductive CIL allows for state silence to be interpreted on the basis of rules that are deduced from general principles such as the sovereign equality of states etc. Inevitably, the deductive or assertive character of this new CIL implies value judgements or even the personal preferences of the adjudicating judge, exposing the type of subjectivism and eclecticism that Schwarzenberger so viciously criticised.

From the above, it follows that there is no straightforward answer or at least an interpretative formula as regards the reading of absence of state practice or the silence of states in the process of CIL ascertainment. In relations among sovereign states, the lack of explicit protest often equals recognition, or at least formal non-objection to a certain legal state of affairs that is under law-creation. There is, for instance, the notion of acquiescence in custom formation and change, such as with territorial claims. Acquiescence is a negative concept related to state inaction or silence, whereby a state is faced with a situation constituting an infringement or threat to its rights. It could be the case that acquiescence be inferred from states’ failure to react to certain claims or acts that call for a positive reaction from their part. Such failure to react thus signifies a non-objection to these claims or acts. In this context, passivity or state silence is tantamount to absence of opposition. The concept has particularly arisen in ICJ proceedings relating to border disputes, asylum, maritime claims and consular rights. For instance, in the process of annexing a new territory, the exercise of formal protest means that the objecting state does not acquiesce in the situation, and that it has no intention of abandoning its territorial rights over the region. Conversely, when a state does not raise an objection, such silence may often be considered as acquiescence.

A question that obviously arises is whether passivity or non-denouncement equals implicit approval. In the Pedra Branca/Pulau Batu Puteh case, the ICJ found, by means of induction, that the absence of reaction conveys acquiescence provided that the conduct of the other state calls for reaction.⁴⁴ This is part of the condition si loqui debuisset ac potuisset (if one can and must act) that was previously articulated in the Temple of Preah Vihear case.⁴⁵ In the same vein, states whose rights are directly affected by a certain act are naturally expected to react. On the contrary, the ICJ ruled in the Asylum case that, as far as regional customary law is concerned, silence on the part of a state vis-à-vis an emerging regional practice equals objection or protest.⁴⁶ This goes against the general presumption implied in acquiescence and the persistent objector doctrine that states should be explicit if they wish not to be bound by an emerging international legal norm.

One could therefore conclude that, when interpreting state silence or inaction for the determination of rules of CIL, the case law of the ICJ is characterised by inconsistencies and an improper, that is non-technical, use of the inductive and deductive methods of reasoning. These disparities are hardly coincidental. Whereas there is, arguably, unfamiliarity among international jurists with the two methods of reasoning, both the interpreter of CIL and the CIL enforcer are often driven by a certain legal purposefulness: governments are naturally tempted to interpret state inaction and silence in a self-serving way, while the ICJ is driven by a combination of systemic considerations, such as the non liquet principle, and legal expediency, such as the preservation of the legal status quo or the management and dealing with international crises. This has been, for instance, the case with the Kosovo advisory opinion and the Asylum case, where the ICJ changed the normal calculus and opted for ad hoc solutions.

4 Lack of Formal Rationality and Recourse to Persuasive Argumentation

From the scope of informal logic, acquiescence is, in principle, quite problematic a concept since the absence of opposition to a state of affairs does not necessarily equal tacit approval. In fact it could be precisely that: absence of opposition. Although the ICJ aimed at addressing the deficiencies of the principle by construing a theory of intentional silence connected to the (natural law) idea that states are willing Leviathans, no robust methodology has been so far produced due to the inevitable subjectivism ensuing from the abstract psychologism pertaining to the will theory. Moreover, the mainstream opting for a closed system of norms precludes the proper application to international legal reasoning of informal logic thus undermining the external rationality of international legal syllogisms.

The lack in international legal reasoning of external rationality and even internal consistency has paved the way to persuasive-teleological argumentation. Argumentation is teleological, in the sense that the person or agency producing the argument aims at a certain end and is thus characterised by a certain ‘argumentative orientation’ towards the preferred conclusion. In this spirit, the ICJ has developed several techniques of superficial, persuasive argumentation, teleologically governed by the non liquet principle, the containment of international crises and the effective resolution of international disputes. For instance, in the Burkina Faso/Mali Frontier Dispute, and without any substantive justification, the ICJ made a leap and asserted the general scope of the uti possideti juris although at that time the principle had only been applied in the context of Latin America and Africa: ‘[i]t is a general principle, which is logically connected with the phenomenon of the obtaining of independence, wherever it occurs’.⁴⁷ In the Land, Island and Maritime Dispute the uti possideti juris was extended to offshore islands and historic bays and in the Territorial and Maritime Dispute in the Caribbean Sea, to the territorial sea.⁴⁸ No substantive justification was sought in the Construction of a Wall case, where the ICJ asserted that the right of peoples to self-determination is a right erga omnes.⁴⁹ At no stage did the court examine the practice and opinio juris of states. Indeed, it is quite often the case that the court simply ‘asserts’ the rules of CIL.

The semantic abstractness of absence and silence constitutes the perfect ground for ‘magic’ argumentative tricks. A characteristic example of this is the Asylum case,⁵⁰ where the ICJ aimed at containing the global expanse of a regional custom in Latin America, namely a regional customary rule requiring a host state to grant safe passage from the embassy where a political refugee has sought diplomatic asylum to the asylum state. In order to suppress the international distillation of the regional custom, the ICJ reversed, without any substantive justification, its settled jurisprudence and ruled that, where a regional custom was concerned, state silence in the face of an emerging regional practice meant that states’ opinio juris was to object/protest to the emerging rule. This assertion, however, ran counter to the general, customary law presumption that states have to raise objections if they wish to avoid being bound by an emerging custom. A year later, in the Fisheries case,⁵¹ Norway had attempted to claim ocean areas by mapping them through ‘straight baselines’, drawn from points along its coastline, and asserted that the enclosed areas were exclusively Norwegian. Norway’s argument was also based on Britain’s lack of protests, which according to Norway meant that Britain had waived its rights by not objecting. However, the ICJ asserted that Norway’s straight baselines were not against international law, for the additional reason that ‘[t]he general toleration of foreign States with regard to the Norwegian practice is an unchallenged fact. For a period of more than sixty years the United Kingdom Government itself in no way contested it’.⁵²

It seems that the ICJ rulings regarding the formation of CIL are particularly troublesome, and, with the exception of regional CIL, they favour aggression and proactiveness in staking claims, while other states’ absence or silence is, as a rule, taken as acquiescence or implicit approval. However, the ICJ’s argumentation techniques often lead to irrational or even absurd results. A typical example is the Kosovo advisory opinion, where the ICJ committed, among others, typical informal fallacies due to argumenta e silentio, as well as argumenta ad ignorantiam.

5 Logical Fallacies in the Kosovo Advisory Opinion: An ‘Open-System’ Approach

Absence and silence are not monosemic. They may signify a variety of things, from acceptance to opposition, or they may have no significance at all. I also need to clarify the following: by referring to absence, I distinguish – yet I mean both – (a) silence qua non-expressed opinio juris and (b) absence proper, qua the lack of (positive) state practice/action. The debate falls into the broader discourse relating to arguments from silence, or argumenta e silentio, as well as arguments from ignorance, or argumenta ad ignorantiam. These are normally classified as informal logical fallacies or weak arguments (weak types of induction) that are somewhat strengthened when evidence is produced at a later stage. Arguments from silence occur when someone interprets someone’s silence as meaning anything other than silence, basically arguing that silence is either communicating implicit approval or disapproval. On the other hand, the fallacy ad ignorantiam occurs when someone argues in favour or against something, in our case state practice, because the opposite has not been proven to be the case.⁵³ In other words, something is said to be true because we do not know whether it is not true. The issue typically has to do with the so-called burden of proof or onus probandi: the ignorance fallacy is a dialectical manoeuvre aiming at unfairly shifting the burden of proof. Normally, in a legal debate between two parties, when one makes a claim that the other party disputes, then the party who makes the claim or assertion has the burden of proof, that is, needs to prove, justify or substantiate the claim.

The fallacy of ignorance occurs when the burden of proof is arbitrarily and unjustifiably reversed, that is, shifted towards the party who disputes the claim. The fallacy of ignorance assumes that something is the case because it has not yet proved to be false or vice versa. This is essentially a false dichotomy providing for forced options, inasmuch as it excludes the possibility that the truth is simply unknowable – not necessarily true or false – or that there has been insufficient investigation of the matter. A typical example in most legal traditions is the presumption of innocence: there is a benefit of assumption, that is, the accused is presumed to be innocent until, and if, evidence is produced to the contrary. Those who are accused of committing a crime are not burdened with proving themselves innocent. One can never shift the burden of proof, which generally rests on the one who sets forth a claim. In criminal proceedings, it is the prosecutor who must show, beyond reasonable doubt, that the accused person is guilty. Not providing adequate evidence of innocence is irrelevant to the verdict. Therefore, an ad ignorantiam fallacy of the type ‘the defendant is guilty because he could not prove his innocence’ would never stand in a criminal court. As we will see later on, in CIL the fallacy ad ignorantiam occurs when there is a judicial misinterpretation of the absence of evidence, that is, instances of state practice, and is normally tightly connected to the fallacy of silence.

From the above it follows that the lack of evidence, in our case, state practice, is not necessarily neutral. There are times when the absence of evidence may prove or disprove a claim. In that case, however, one needs to take into account the context of the case: suppose that John needs to rent an apartment in Groningen, Netherlands, but he needs to make sure that the house has no cockroaches. He hires a specialist who, after investigating the apartment, reaches the conclusion that it does not have any swarms or cockroaches or other insects. The lack of evidence in this case is not neutral. In the evaluation of evidence, the authority that makes a certain claim is taken into account. Moreover, although it appears as though we have a typical case of argument ad ignorantiam, the truth is that the negative inference (absence of cockroaches) is based on a positive evaluation of evidence. A fallacy ad ignorantiam occurs when there is no evidence and no proof whatsoever is offered for the claim, that is, when one argues that there are no cockroaches in the apartment simply because they have not seen any. The argument that there is no God simply because one cannot see Him, and vice versa, the argument that there is God because the atheists cannot disprove His existence, are both arguments from ignorance, and thus informal fallacies.

To bring this back to the Kosovo advisory opinion, the ICJ implicitly applied the Lotus principle and reformulated the legal question. Instead of examining whether unilateral declarations of independence are in accordance with international law, the court, without providing any substantive justification for this choice, decided to examine whether international declarations of independence are forbidden under international law, thus substantially changing the question, while at the same time committing the fallacy of false alternatives. Moreover, the judicial argument did not entail any substantial evaluation of evidence of state practice, opinio juris, or any substantial evaluation of absence or silence, but merely took note of the historical fact that:

In no case … does the practice of States as a whole suggest that the act of promulgating the declaration was regarded as contrary to international law. On the contrary, State practice … points clearly to the conclusion that international law contained no prohibition of declarations of independence. During the second half of the twentieth century, the international law of self-determination developed in such a way as to create a right to independence for the peoples of non-self-governing territories and peoples subject to alien subjugation, domination and exploitation … A great many new States have come into existence as a result of the exercise of this right. There were, however, also instances of declarations of independence outside this context. The practice of States in these latter cases does not point to the emergence in international law of a new rule prohibiting the making of a declaration of independence in such cases … For the reasons already given, the Court considers that general international law contains no applicable prohibition of declarations of independence. Accordingly, it concludes that the declaration of independence of 17 February 2008 did not violate general international law.⁵⁴

The ICJ examined the general law applicable to the case before it and asserted that there is no general rule of international law – either treaty law or customary law – that prohibits declarations of independence and that ‘[i]n no case … does the practice of States as a whole suggest that the act of promulgating the declaration was regarded as contrary to international law’.⁵⁵ Judges Yusuf and Simma criticised the court’s conclusion on the ground that the Lotus principle is an outdated doctrine and the silence in international law should be understood and interpreted more broadly. Judge Simma asserted that according to the Lotus principle ‘restrictions on the independence of States cannot be presumed because of the consensual nature of the international legal order’.⁵⁶ He criticised the court for being too formalistic in equating ‘the absence of a prohibition with the existence of a permissive rule’ and drew the attention to ‘the possibility that international law can be neutral or deliberately silent on the international lawfulness of certain acts’.⁵⁷ The advisory opinion on Kosovo received robust criticisms and extensive commentaries from both the judiciary and the international legal scholarship.

We have seen that arguments from silence occur when someone automatically interprets someone’s silence as meaning anything other than silence, basically arguing that silence is either communicating implicit approval or disapproval. Generally, when we are dealing with a silent authority (i.e. a state) we should ask ourselves: would the silent authority have known about the claim and consciously chose to remain silent? Is the silent authority definitely aware of the claim? Is the silent authority most likely to be honest about the claim? Do we have a complete record of everything written/done by the authority? Is this record true and reliable record, and not just a presumption based on lack of evidence? If the answer to any of the above questions is negative, it is quite possible that we are dealing with a fallacious argument from silence. However, even if we answer in the affirmative, even a good argument from silence is a weak argument that should be treated as inconclusive or uncertain when no other evidence is provided.

From the perspective of informal logic, the argument from silence is also tightly connected to the relation between negation and belief. Let us assume that someone does hold a certain belief or opinio juris. Is it monosemic or could it express various modalities? Indeed, there are various modalities governing the belief-universe, a misunderstanding/misapplication of which could generate logical fallacies and lead to distortions. For instance, there are the so-called internal and external negations of belief, and thus, opinio juris. Let us diagnose the fallacy:

1. 1. John believes that God does not exist. (Internal negation)

2. 2. John does not believe that God exists. (External negation)

3. 3. John believes that God exists.

These three examples depict the so-called withhold/deny fallacy.⁵⁸ The fallacy is to read 1 and 2 as meaning the same, whereas according to informal logic 1 entails 2, but not vice versa. Accordingly, the denial of 3 is sometimes wrongly taken to be 1 (case of false alternatives), whereas the contradictory/true denial of 3 is 2. In other words, one’s denial to hold a belief does not affirm that one holds the opposite belief. Not believing does not amount to disbelieving. This is what distinguishes agnosticism from atheism: the choice between belief and disbelief is not a forced choice: there is a third way, the way of withhold or non-belief. In everyday argumentation, it is quite often the case that we commit the withhold/deny fallacy for the sole reason that the practical consequences are seemingly indistinguishable. However, that would only make sense if the object of belief was entirely factual/practical rather than conceptual. Generally, the occurrence of the withhold/deny fallacy also produces the fallacy of false alternatives: that is, a state either accepts a regional custom or not.⁵⁹ There is no in-between. The fallacy of false alternatives in CIL has been formally incorporated in legal doctrine via the Lotus principle, as it manifests itself in the Kosovo advisory opinion, among others. Let us assume that the assertion ‘Anna believes in ghosts’ is ‘Ag’. The variations of negation can be further symbolised as those contained in Figure 4.1.

Figure 4.1

The richness of belief. There is also the problem of belief itself. Let us also take as a given that a state is an entity that can be conceptualised as a Leviathan who thinks and reasons, which is of course not the case, so the induction is already arbitrary so to speak. The state is an enormous political-bureaucratic machine, and so one may naturally wonder how many beliefs by state-agents, legal advisors and high-ranking officials need to coordinate towards a certain belief or idea. Whose belief is of greater value, if so? How many views are considered enough to formulate the so-called opinio juris as a belief-reservoir? Or, even more profoundly, how will these individual beliefs be measured, attested and evaluated? Should we resort to official archives? Either official or unofficial communications? General Assembly Resolutions? How intense or strong should the negation or affirmation be in order to qualify as a positive or negative belief regarding the perceived bindingness of a norm? And what about plain indifference? What type of formality should be attached to this set of beliefs? Moreover, we cannot simply assume that a certain belief – opinio juris – is always in full awareness.

The richness of silence. Accordingly, we, as international lawyers, may indeed have to deal with either conscious silence or unconscious silence: intended or unintended silence. Should we assume that there are no variations in silence itself? What if silence qua the consciously or unconsciously omissive passage of time is not semantically homogenous throughout (the silent) time, that is, transforms into something semantically different at some point, given new circumstances? Judge Sørensen in the North Sea Continental Shelf noted ‘[i]n view of the manner in which international relations are conducted, there may be numerous cases in which it is practically impossible for one government to produce conclusive evidence of the motives which have prompted the action and policy of other governments’.⁶⁰

Some of these questions have been addressed by the ICJ but most of them have not. For instance, in the 1951 Fisheries case, the ICJ seems to have taken into deeper consideration the context of British silence and ruled that ‘[t]he notoriety of the facts, … Great Britain’s position in the North Sea, her own interest in the question, and her prolonged abstention would in any case warrant Norway’s enforcement of her (straight baseline) system against the United Kingdom’.⁶¹ Accordingly, in the North Sea Continental Shelf Cases, the court stated:

[w]ith respect to the other elements usually regarded as necessary before a conventional rule can be considered to have become a general rule of international law, it might be that, even without the passage of any considerable period of time, a very widespread and representative participation in the convention might suffice of itself, provided it included that of States whose interests were specially affected.⁶²

Normally, however, the evaluation of silence is much more superficial and narrow. In the Military and Paramilitary Activities in and Against Nicaragua, the court ruled that opinio juris may also be deduced from the attitude of states towards certain General Assembly resolutions.⁶³ This was confirmed in the Nuclear Weapons advisory opinion.⁶⁴ In the Lotus case, the Permanent Court of International Justice deduced from the freedom of the seas’ principle that ‘vessels on the high seas are subject to no authority except that of the State whose flag they fly’.⁶⁵

We have seen that arguments from silence occur when someone interprets someone’s silence as meaning anything other than silence, basically arguing that silence is either communicating implicit approval or disapproval. Again, when we are dealing with a silent authority, such as a state, we should ask ourselves: would the silent authority have known about the claim and consciously chose to remain silent? Is the silent authority fully aware of the claim? Do we have a complete record of everything written/done by the authority? Is this record true and reliable, and not just a presumption based on lack of evidence? If the answer is ‘no’ to any of the above, it is quite possible that we are dealing with a fallacious argument from silence. On the other hand, the fallacy ad ignorantiam occurs when someone argues in favour or against something, in our case state practice, simply because of lack of evidence, and not because of a positive evaluation of the absence of evidence.

In the Kosovo advisory opinion, the ICJ committed both fallacies. Without any substantial argumentation and with the ultimate goal to solve the Kosovo puzzle, the ICJ erroneously interpreted the absence of state practice as also implying a neutral opinio juris by the vast majority of states vis-à-vis unilateral declarations of independence. The question that obviously arises here is this: how is it even possible for states to concur in a situation that would put their very existence in danger? One therefore notices a typical example of persuasive-teleological argumentation, oriented towards the effective resolution of an international dispute. The ICJ also construed the controversial perceived intent argument, that is, the argument that the authors of the declaration of independence did not seek to act within the constitutional framework of the interim administration for Kosovo (i.e., as the Assembly of Kosovo), but instead ‘acted together in their capacity as representatives of the people of Kosovo’.⁶⁶ In particular, the ICJ held that ‘the authors of that declaration did not act, or intend to act, in the capacity of an institution created by and empowered to act within that legal order but, rather, set out to adopt a measure the significance and effects of which would lie outside that order’.⁶⁷ Regarding the Serbian constitutional order, the ICJ concluded that the constitutional laws of Serbia were not applicable insofar as the object and purpose of the 1244 Security Council Resolution was to establish a temporary legal regime that would supersede the Serbian Constitution. In other words, the ICJ opted for the mainstream, closed-system approach, that is, departed from the premise that international law is a closed normative system, and treated the unilateral declaration of independence as a factuality lying outside that system. It also ‘extended’ the will theory to a non-state actor. By doing so, the ICJ also reached another paradoxical conclusion: one can act outside a normative order and thus avoid liability, simply by saying so.

Apart from the erroneous interpretation of silence and absence, the ICJ also committed an important syllogistic fallacy. Inference, or formal inference, is the logical process of understanding what is implied in a certain proposition; the process of deriving general or particular propositions, on the basis of something previously assented to, namely the ‘derivation of one proposition, called the Conclusion, from one or more given, admitted, or assumed propositions, called the Premise or Premises’.⁶⁸ The objective of inference is the objective of reasoning: the examination of the validity of a statement by reason of certain facts or statements from which it is said to follow. Syllogism is the narrow concept of mediate inference, namely the inference for the completion of which we necessarily employ a medium or middle term. The term ‘syllogism’ derives from the Greek words σύν (together) and λόγος (thought) and means the bringing together in thought of two Propositions in order to compose a third proposition, commonly referred to as Conclusion. For logician William Minto, ‘[t]he main use of the syllogism is in dealing with incompletely expressed or elliptical arguments from general principles’.⁶⁹ It is often the case where elliptical arguments are put forward, also known as enthymemes, whereby one premise is explicit and the other suppressed, namely held in the mind.⁷⁰ In this case, the purpose of the syllogism is practical: to expose the implications of the hidden premises in the most explicit, convincing and undeniable way possible, and challenge what is otherwise considered to be self-evident. There is such a fundamental, hidden syllogism in the Kosovo advisory opinion.⁷¹ The syllogism goes as thus:

1. Premise 1 (P1): All what is not forbidden (to states) is allowable.

2. Premise 2 (P2): All declarations of independence (by non-state actors) are not forbidden.

3. Conclusion (C): All declarations of independence are allowable (for non-state actors).

From the above scheme one immediately notices that the problem with the judicial syllogism does not only rest with the fallacious interpretation of absence and silence in international law according to the Lotus principle but also with a serious syllogistic fallacy. The conclusion of the deductive judicial syllogism is logically unsound because it does not follow from the premises. The syllogism suffers from the logical fallacy of equivocation. Equivocation is not a formal fallacy but a verbal or material fallacy, which implies that the same word or phrase is used in two different ways. The predicate term ‘what is allowable’ has a dual meaning: in P1 it means ‘allowable for states’ whereas in C it means ‘allowable for non-state actors’. This is a typical sophist fallacy: ‘an elephant has a trunk; a car has a trunk; therefore, an elephant must be a car’. This is equivocation. The rationality crisis is camouflaged because the judicial critique (i.e., the declarations, separate and dissenting opinions) focuses on the interpretation of silence.

Considering that the judicial syllogism does not have any external rationality, let us now turn to its internal consistency. As mentioned above, silence was attributed to the authors of the unilateral declaration of independence, it is therefore important to examine whether, from the scope of international law, silence could concern the conduct of non-State actors as well. In other words, the crucial question from an internal point of view is whether the Lotus principle applies to states and non-state actors alike. This particular question was neither posed nor addressed by the ICJ. At a first glance, it is debatable whether the principle applies to non-state actors at all due to their limited and derivative legal personality. In fact, in the same advisory opinion the ICJ concluded that the principle of territorial integrity is not applicable to, and thus does not bind, non-state actors. However, from an internal perspective and based on Kammerhofer’s interpretation of Kelsen, that is, that a normative order can only be composed of positive norms, one can also go as far as to regard silence by states as well as non-state actors as merely factual, that is, as lying outside the normative order. From this perspective, what the ICJ then did was simply to acknowledge a factual state of affairs, namely a freedom that is normatively indifferent. However, it is difficult to argue in favour of such an interpretation, given that the ICJ did not simply acknowledge a freedom that is factual, but actually a freedom that is normative, insofar as it is accorded concrete legal consequences. Indeed, according to the ICJ, in the absence of a prohibitive rule, states (and non-state actors alike) are legally free to do as they wish. One can hardly argue that such an assertion is normatively indifferent, given that it clearly entails a positive legal permission as well as a corresponding legal entitlement.

6 Conclusion

It has been demonstrated that international law does not provide any clear guidance as regards the legal effects that follow from state silence. The prominent closed-system approach goes against the rules of logic and the canons of reasoning, according to which absence may correspond to multiple values, a variety of propositions and modalities. It has been argued that, in international jurisprudence, these modalities have been either equated or largely ignored. In the same spirit, the mainstream interpretation of CIL overlooks the quantifications and varieties of meaning in non-appearances. It has been suggested that an open-system perspective could shed light on inconsistencies and/or erroneous interpretations.

The modalities of absence affect the rationality and soundness of international legal doctrine and even have a real impact on international relations when overlooked. Due to the scarcity of proper inductive arguments in the process of CIL ascertainment, the striving for discursive truth and reason has been limited to an examination of superficial rationality, that is, a mere analytical and superficial testing of consistency. Even this internal consistency, though, is not a given. There is, for instance, a new distinction between the traditional and the so-called modern deductive CIL, that is, a distinction between customary law that results from the traditional, inductive method of reasoning, and customary law that arises in instances where the inductive method is considered impossible to use because state practice is non-existent, the legal question is too new and has not been dealt with etc. In those instances, deduction and assertion are often used. Both the absence of state practice and state silence are thus often interpreted on the basis of rules that are either deduced from general principles, or expressing simple assertions. Inevitably, the deductive or assertive character of this new CIL implies value judgements and/or the personal, political preferences of the adjudicating judge.

The lack in international legal reasoning of external rationality and even internal consistency has paved the way to persuasive-teleological argumentation. The ICJ has developed several techniques of superficial, persuasive argumentation, teleologically governed by the non liquet principle, the containment of international crises and the effective resolution of international disputes, thus producing and reproducing serious rationality deficits in the judicial treatment of silence in the framework of CIL. The existence of external and internal rationality deficits as well as the corresponding rhetorical manoeuvring increase the need for the legal system to appeal to a concrete legitimizing basis for the explanation of derogations, exemptions, ad hoc solutions or whatever argumentation games and gaps cannot be justified by virtue of the normative structure of the system itself or some generalised imperative of system maintenance, such as a state of emergency. This has been, for instance, the case with the Kosovo advisory opinion and the Asylum case, where the ICJ changed the normal calculus in the interpretation of silence and opted for ad hoc solutions. The anomaly in the Kosovo advisory opinion was pointed out by Judge Tomka, who in his Declaration argued that:

[t]he legal régime governing the international territorial administration of Kosovo by the United Nations remained, on 17 February 2008, unchanged. What certainly evolved were the political situation and realities in Kosovo. The majority deemed preferable to take into account these political developments and realities, rather than the strict requirement of respect for such rules, thus trespassing the limits of judicial restraint.⁷²

However, unless either an open-system approach is applied or robust coherence and consistency is systematically and methodologically pursued within a closed system, ad hoc and arbitrary judicial responses to non-appearances will persist.

5 Schrödinger’s Custom Implications of Identification on the Interpretation of Customary International Law

Markus P Beham

One can even set up quite ridiculous cases. A cat is penned up in a steel chamber, along with the following diabolical device (which must be secured against direct interference by the cat): in a Geiger counter there is a tiny bit of radioactive substance, so small, that perhaps in the course of one hour one of the atoms decays, but also, with equal probability, perhaps none; if it happens, the counter tube discharges and through a relay releases a hammer which shatters a small flask of hydrocyanic acid. If one has left this entire system to itself for an hour, one would say that the cat still lives if meanwhile no atom has decayed. The first atomic decay would have poisoned it. The ψ-function of the entire system would express this by having in it the living and the dead cat (pardon the expression) mixed or smeared out in equal parts.

Erwin Schrödinger, ‘The Present Situation in Quantum Mechanics’¹

1 Introduction

Normative efforts in international law – including interpretation – must be grounded on a sound ascertainment of the sources of legal obligation. What might appear as a comforting truism for followers of black letter law seems an almost unattainable quest when it comes to the identification of international custom. This chapter proposes a pragmatic positivist approach to the identification of non-consensual, unwritten law: Schrödinger’s custom. If the classic textbook ‘two-element’ theory of customary international law (CIL) is valid – and the ILC still seems to think it is² – then at least half of the identification process consists of an empirical assessment. It requires to look at – to follow the title of Louis Henkin’s seminal work³ – how nations behave. Under the classical view of realism, states act according to a set of inherent interests. These may provide a compass for orientation through the haze of normative propositions. The chapter begins by characterising CIL among the sources from which international rights and obligations arise (Section 2). It then moves on to depict the process of identification referred to here as Schrödinger’s custom including its implications for the issue of custom interpretation (Section 3). On that basis, it discusses how international relations theory may help predict the outcome of the identification process (Section 4). A conclusion rounds it all off. (Section 5).

2 Custom as a Source of Legal Obligation

If one accepts the catalogue of manifestations of international law in Article 38(1)(a)–(c) of the Statute of the International Court of Justice as an expression of universal state consensus,⁴ one must look for ‘international custom, as evidence of a general practice accepted as law’. Without overthinking the implications of the wording referring to custom as the evidence as opposed to being evidenced by ‘a general practice accepted as law’,⁵ one could simply take the provision at face value: references to custom as a source of legal obligation imply the existence of evidence as to ‘a general practice accepted as law’. Invocation of custom proposes the possibility of identifying both state practice and opinio juris, making ‘international custom’ and ‘evidence of a general practice accepted as law’ synonymous sides of an equation.⁶ In this view, ‘international custom’ comprises both the process of identification (Rechtserkenntnis) and the underlying acts of law creation (Rechtserzeugung).⁷

How may these underlying acts be characterised? The element of ‘practice’ reaches directly into international relations as they are conducted on a daily basis. In addition, for something to constitute custom, the respective behaviour must follow a sense of legal obligation. This has led certain strands in the literature to equate custom with tacit agreements.⁸ Yet the general principles of law ut res magis valeat quam pereat⁹ and favor contractus¹⁰ carry the assumption that states adopting Article 38 of the Statute of the International Court of Justice intended to give meaning to its words. That custom is unwritten, unless it is codified, seems uncontested. But following ‘international conventions, whether general or particular, establishing rules expressly recognized’, ‘a general practice accepted as law’ must also mean something other than a consensual agreement. Unless one were to stretch the word ‘convention’ beyond its ordinary meaning under international law (for the lay use of the term might actually be synonymous with custom), Article 38(1)(a) refers to ‘agreements’. As such, these may be written or unwritten, even implicit in the form of a tacit agreement.¹¹ When the Statute of the International Court of Justice requires that they establish ‘rules expressly recognized by the contesting states’, this refers to the express recognition inherent in the process of reaching an ‘agreement’.

If one were now to equate ‘a general practice’ with ‘international conventions, whether general or particular’ and ‘accepted as law’ with ‘establishing rules expressly recognized’, no sense would be given to the two separate provisions included in Article 38(1) of the Statute of the International Court of Justice. In short, opinio juris cannot simply be equated to an oral or tacit agreement. Still, it represents a quasi-consensual element, in that states could equally choose to answer the question ‘did you just behave that way because you thought there is a legal obligation to do so’ negatively.¹²

3 Identification

The state of CIL is in constant flux. The paradox that a customary norm must first be broken in order for a new one to arise, follows the Linnaean urge of scholars to sort and categorise their surroundings. But this approach does not do justice to the dynamic nature of a set of norms that is largely dependent upon the interaction of states.¹³ While an awareness of certain trends within a particular area of law is both useful and necessary to satisfy expectations towards the rule of law, a full assessment is only necessary once a specific argument is put forward. Like a snapshot photograph, CIL is identified at a certain point in time, be it within judicial proceedings or in a scholarly publication.¹⁴ Since custom implies both identification and creation, their temporal dimensions collapse. The view that CIL is made in the past becomes a myth.¹⁵

Custom forms only in the present, once it is invoked and an observer is introduced. Explanatory aid may be sought from the famous thought experiment of Austrian physicist Erwin Schrödinger.¹⁶ In his (for pet lovers luckily only theoretical) experimental set-up, a cat is placed in a steel chamber together with a vial of deadly acid that is released the moment an atom from a piece of radioactive material decays. However, it is equally probable that the radioactive material does not decay. Without an observer, there is no knowing whether the atom has decayed. Until that point in time, both the living and the dead cat must be assumed to exist. They are ‘mixed or smeared’ together.¹⁷ What Schrödinger intended as an illustration of the paradox between reality and theoretical quantum-mechanics may easily be transposed to the problem of CIL formation. Until an observer is introduced, it is unclear how many states have already engaged in practice accompanied by opinio juris.

This should not be mistaken with the identification of an exact point in time at which a particular norm of CIL was created. The question is only as to the present existence of ‘evidence of a general practice accepted as law’. As Maurice Mendelson pointedly illustrated,

it makes no more sense to ask a member of a customary law society ‘Exactly how many of you have to participate in such-and-such a practice for it to become law’ than it would to approach a group of skinheads in the centre of The Hague and ask them, ‘How many of you had to start wearing a particular type of trousers for it to become the fashion – and, indeed, de rigeur – for members of your group?’ … The customary process is in fact a continuous one, which does not stop when the rule has emerged, even if one could identify that exact moment. To illustrate the point, I would like to introduce a simile. … My simile is the building of a house. It is often not easy or even possible to say exactly when a house has been created. Clearly, it is not when the first foundation stone is laid. But it is not when the last lick of paint has been added either. It is problematic at exactly what point we could say ‘This is a house’. Do we have to wait for the roof to go on, for the windows to be put in, or for all of the utilities to be installed? So it is with customary international law.¹⁸

Rather than the point of formation, the observer will ‘take a still photograph, so to speak, of the state of the (customary) law at a given moment’,¹⁹ the lex lata. The relevant question in practice – and in scholarship, for that matter – will mostly be the application of a certain rule to a particular set of circumstances, rather than a historic narrative of when and how a rule has formed.²⁰ In Charles De Visscher’s words, ‘[i]n international relations more than elsewhere, the fact precedes its classification’.²¹ The result is simply a manifestation of the dynamic character of international relations.

In our experiment, what do we imagine this observer to look like? Obviously, it cannot be a lobbyist or policymaker, nor an idealist international lawyer.²² So, should it be a judicial robot, an algorithm fed with empirical data? While this idea of an objective assessment seems attractive at first, it is hard to see how this would deliver equitable results; more likely, such a sterile approach to law identification – which ultimately relies on the interaction of states as raw data – might result in a ‘Bizarro World’ image of international law. The fact that states torture with the conviction that they have a legitimate basis for doing so – one must only think of the ‘ticking time bomb’ scenario²³ – would result in a permissive rule allowing torture under such circumstances.

The analysis requires an underlying human corrective. It is in the same sense that Andreas Paulus and Bruno Simma speak of the need for an ‘enlightened positivism’.²⁴ It would seem fitting to rely on the proverbial man on the Clapham omnibus. This reasonable – we might also imagine ‘extra-terrestrial’ – is neither an idealist, nor a cynic, neither a revisionist, nor an innovator. As little is he driven by a particular national interest, as by the ideal of the international community as a civitas maxima. Admittedly, this is a ‘you know it, when you see it’ approach, but in combination with the identification of CIL restricted to a certain point in time it will surely allow for a more grounded assessment of the body of CIL than any elaborate game theory model or natural law-based impulse. Occam’s razor will easily help in the identification of state practice and opinio juris.²⁵

What does this imply for the act of interpretation? If the temporal dimensions of creation and identification collapse, it can only result in ‘instant interpretation’. As custom is frozen in the moment of its invocation, any statement about its future application becomes meaningless. Instead, custom must be repeatedly reassessed, unless there is a good faith assumption that the original invocation still constitutes ‘evidence of a general practice accepted as law’. Any subsequent practice always paves the road towards new custom. Taking the example of a codification, if one were to ‘interpret’ its content for purposes of clarification, one would either be interpreting ‘subsidiary means for the determination of rules of law’²⁶ to help identify such ‘international custom, as evidence of a general practice accepted as law’,²⁷ or again engage in the identification of the underlying acts of law creation, thereby acting as an observer to ‘Schrödinger’s custom’.

4 State Interest

Are the implications of this mode of identification on the interpretation of CIL that it simply becomes unpredictable? As Malcolm Shaw writes, ‘[c]hange is rarely smooth but rather spasmodic’.²⁸ If state practice follows day-to-day world affairs, international relations theory might help. As Louis Henkin convincingly laid out in his seminal work How Nations Behave, states act according to carefully calculated interests and dependent upon the consequences of their conformity to or violation of international law.²⁹ This approach is, generally, quite similar to the economic theory of negligence that ‘[w]hen the cost of accidents is less than the cost of prevention, a rational profit-maximizing enterprise will pay tort judgments to the accident victims rather than incur the larger cost of avoiding liability’.³⁰

The term ‘interest’ derives from the Latin interesse, which carries the meaning ‘to differ’ or ‘to make a difference’. The interest is something that makes a difference to someone – or, if speaking of a juridical entity, to something. In discussing these issues, one is always confronted with the problem of anthropomorphising states.³¹ Some writers have gone as far as to argue that states are not capable of holding such interests, ‘as if artificial entities could have discernible motivations’.³² However, this position overlooks the idea of statehood as represented through the collective of individual actors with a common agenda. Just as what makes a difference for an individual employee does not necessarily make a difference for a corporation, it does not necessarily make a difference for a state.³³ Each entity, the natural person as well as the juridical body, carries distinct goals and purposes. Some may correlate, some may differ. It is the nature of the respective actor that determines the interest.

The expression ‘state interest’ or ‘national interest’,³⁴ as it is sometimes found in the literature, confers the idea that there must be a common set of factors that are important to the existence of the abstract entity of the state. At the same time, it has been suggested that ‘no agreement can be reached about its ultimate meaning’.³⁵ Still, it seems to be an important factor in decision-making of political stakeholders,³⁶ best reflected in the anecdotal quotes of Charles De Gaulle and Henry Kissinger that their respective states had ‘no friends’ but ‘only interests’. That states ultimately strengthen and enrich themselves at the cost of others cannot shock an international lawyer since Emer de Vattel’s 1758 publication of Le Droit des Gens.³⁷

The idea that law formation follows the interplay of interests is also not particularly new. Carl Schmitt – the Dooyeweerd of German people³⁸ – already argued that public international law in the nineteenth century rested less on ideas of sovereignty than on a selection of specific state interests.³⁹ Jean d’Aspremont found that ‘[e]ven liberals and constitutionalists agree that States first strive to promote their own interests’ and that ‘they naturally act to maximize the interest of their constituency given their perception of the interests of other States and the distribution of State power’.⁴⁰ Martti Koskenniemi has called reference to this fact a ‘truism, present since Vattel’.⁴¹ Richard Steinberg convincingly showed how different schools of international legal thought and international relations theory resorted to realism whenever they dealt with states.⁴² Today, Martin Dixon begins his introductory textbook on international law by finding that ‘[i]t is true of all legal systems that vital interests of its subjects may prevail over the dictates of the law’.⁴³ According to Malcolm Shaw, the motivation behind an act of a state lies within the way in which ‘it perceives its interests’, which again depends upon ‘the power and role of the State and its international standing’.⁴⁴

What are these supposed interests that determine the probability of state action? For any realist, states are driven by two principal considerations: first, national security, comprising the protection of statehood, territorial integrity, as well as sovereignty, and, second, a functioning economy. Gerhard Hafner identified five traditional areas of state interest: ‘the protection of statehood, territorial integrity, sovereignty, security and economic wealth’.⁴⁵ Nicholas Onuf speaks, in the Hobbesian tradition, of ‘standing, security, and wealth’.⁴⁶ Recalling the definition of what constitutes a state, these ‘traditional’ interests are inextricably linked to its ‘survival’.⁴⁷ Each student of international law knows that the ‘primary subjects’⁴⁸ of international law consist of a permanent population, a defined territory and a government.⁴⁹ Recalling this definition, these ‘traditional’ interests are inextricably linked to the ‘survival’ of a state. In a sense, to anthropomorphise states once more, this feature is not so different from the ‘survival instinct’ of individuals. The latter are equally interested in escaping the Hobbesian bellum omnium contra omnes before all else. The social contract that allows for this escape wants to be upheld. Thereby, state interest is equated with the survival of the state.⁵⁰ Without territory, without governmental control, it lacks its constitutive elements.

While states require individuals to take action on their behalf, these ‘do not act on their own account but as State officials, as the tools of the structures to which they belong’,⁵¹ a view that is further reflected in the rules of attribution in the International Law Commission’s Articles on State Responsibility.⁵² The state organs are limited by the framework that is the respective state, even if this is little more than the collectivity of individual decisions. Its economy, social structure, and cultural heritage will largely determine what is opportune. Thus, states may weigh their interests differently and in accordance with additional factors such as ideology, be it liberal democracy, socialism, or some pan-territorial or ethnic component.⁵³ Still, the definition of the state is tainted by the fact that individuals act on its behalf. The way it is externally perceived is shaped by its successive governments. Therefore, it is important to differentiate between the state, its organs, and its population in making any determinations as to its character. Brierly defined the state exactly along these lines as ‘a system of relations which men establish among themselves as a means of securing certain objects, of which the most fundamental is a system of order within which their activities can be carried on’. At the same time, he cautioned that the state ‘should not be confused with the whole community of persons living on its territory’, as ‘it is only one among a multitude of other institutions, such as churches and corporations, which a community established for securing different objects’.⁵⁴ Yet this is little different from the way that multi- or transnational corporations such as Walmart, Royal Dutch Shell, or ExxonMobil are perceived against the background of a change in the board of directors. Only in extreme situations such as a revolution, is it likely that states entirely change their character on the initiative of a government or other persons or groups of persons exercising authority. States, as all other legal entities, are fictions to express the idea that individuals may come together to create an entity that pursues goals not necessarily representing their own and vice-versa. Even Immanuel Kant, one of those authors most championed for the cause of lofty values, pointed out that the wellbeing of the state – the ‘Heil des Staates’ – does not necessarily correspond with the wellbeing or happiness of its respective citizens.⁵⁵

A number of structural arguments have been brought against this view. For example, the need of states ‘to include [NGOs] in their foreign policy analysis and respect their interests in the process of creating norms of international law’ as a result of ‘the power exercised by them through the use of media and similar means’.⁵⁶ However, these are means to an end:⁵⁷ the survival of states and, in this case, governments. These will likely set acts in the name of a state that aim at preventing civil unrest, cultivating a happy electorate,⁵⁸ attracting investment and highly skilled labour, securing development aid, gaining admission to an international organisation – the list goes on.⁵⁹ There is also still a certain impetus of morality determining action in the face of mass human rights violations or unrestrained warfare.⁶⁰ But this altruistic impulse seems often by itself too weak to spur any form of meaningful intervention.⁶¹ Notwithstanding, the constitutionalist or Kantian argument still stands strong within international legal scholarship, spurred by Wolffian ideas of a civitas maxima.⁶² Its moral superiority is, after all, compelling.⁶³ Equally, state interest is not a one-way street. Interests of other states must be taken into account at some level, in particular in an international relations reality that has become dominated by a universal international organisation that is the United Nations.⁶⁴ Yet, this is a simple outcome of the discourse within which international relations take place,⁶⁵ already identified and incorporated by structural realism.⁶⁶ Yet ‘subsidiary interests’ will not necessarily predict what states will do, when competing core interests of survival arise. In such cases, states will usually resort to ‘Realpolitik’.⁶⁷ They will, generally, not compromise on their interests out of altruistic motives – in this case vis-à-vis states – or out of concern for public opinion.⁶⁸ Even Gerhard Hafner, who takes a position that emphasises the role of the individual in international law, concedes that states take all the weight in this balance of interests when he writes that ‘the reflection of the – nevertheless increasing – individual-oriented interests in norms of international law still depends on the will of states’.⁶⁹

Equally, the constitutionalist argument does not stand empirical scrutiny. Just as states will bulldoze over public image considerations, whenever their survival interests are at stake, states will limit their activism with regard to jus cogens and erga omnes obligations to situations in which their own interests are concerned.⁷⁰ In the competition of ‘first-order reasons’, to borrow Joseph Raz’s terminology,⁷¹ interests related to the survival of the states will, naturally, prevail. In absence of an exclusionary rule, a state will balance these interests in accordance with their respective ‘strength’ or ‘weight’.⁷² A ready example is the primacy that states accord to national security considerations over basic citizens’ rights in the face of terrorism.⁷³ Altruistic obligations, in particular, do not seem likely candidates for custom.

How can these considerations on state interest help identify possible trends in CIL? Add to this effectivity and reciprocity, the catalysers of international law formation,⁷⁴ as the vertical and horizontal angles for the realist’s theodolite and a credible prediction should be the likely result. After all, it is not just international law that guides the behaviour of states, but politics of interest. In turn, interest determines the formation of international law.⁷⁵ There might also exist areas of law in which compliance is not necessarily rewarded by reciprocal behaviour, but it seems that CIL will, at least, likely reflect an equilibrium of interests.⁷⁶

5 Outlook

The ‘cliché’⁷⁷ two-element theory of CIL can provide a simple solution against the legion of alternative theories. As a manifestation of international law that does not directly spring from the ‘will’ or ‘consent’ of states, it reflects their perpetual international relations. States do not voluntarily form a will at the international level but consciously or unconsciously influence its creation through their actions. Following the metaphor of ‘Schrödinger’s custom’, until an observer is introduced to determine what the particular customary rule is in a certain moment, CIL remains ‘mixed or smeared’.

Once an observer is introduced and the temporal dimensions of creation and identification collapse, ‘interpretation’ can only mean the assessment of ‘evidence of a general practice accepted as law’ at a certain point in time. Subsequent practice will always only ever pave the road towards new custom.

This should not suggest a nihilistic view of custom. While the literature may already now concede the instructive value of realism when dealing with states, stronger attention should be given to the interplay of this ‘truism’ with the formation of CIL. It is obvious that parties bring their interests to the table when negotiating a treaty. Strangely, it appears less obvious whenever scholars seek to harness custom for the normative project of international law. More even than other sources, CIL will most likely reflect an equilibrium of interests.