Part II - Customary International Law as a Source of International Law

6 The Significance of State Consent for the Legitimate Authority of Customary International Law^*

Andreas Follesdal

experience has taught us, that human affairs would be conducted much more for mutual advantage, were there certain symbols or signs instituted, by which we might give each other security of our conduct in any particular incident.¹

the reasonable expectation produced by a promise … [which is] a declaration of your desire that the person for whom you promise should depend on you for the performance of it.²

1 Introduction

Norms of customary international law (CIL) pose a dilemma for international courts. Rules (and principles) of CIL are unwritten sources of international law with two central constituent features: they form ‘a general practice’ which enjoys ‘acceptance as law’ (opinio juris).³ Among CIL are various jus cogens norms, with a higher rank than treaty law and other CIL.⁴ Judges and scholars may thus disagree about whether alleged CIL indeed is a rule of CIL – whether it satisfies both the two necessary conditions, its specific contents and whether it is jus cogens.

Customary international law appears to challenge the central role of state consent in accounting for the legitimacy of public international law (PIL) in general. Consider the immunity of foreign heads of state, the principle of non-refoulement, or jus cogens norms that outlaw slavery, torture, genocide, aggression, or crimes against humanity. Sovereign states are under a legal obligation to comply with these CIL norms, even though they have not explicitly consented to them. How should international courts accommodate non-consent-based CIL and state sovereignty? The following reflections outline one strategy that avoids or helps address challenges wrought against other attempts to create more consistency and coherence between CIL and the other sources of international law – whilst securing a central role for state consent.

Some argue to weaken the conception of state consent to include various ‘tacit’ forms in CIL. Challenges to these strategies are legio, including how to detect the sort of tacit consent that – unlike forced acquiescence or mindless habituation – can help account for the justificatory binding force of such consent.

Similar challenges arise when resorting to purely hypothetical consent. Thus, some argue that states sometimes have good reason to be bound to CIL norms precisely without their consent, in response to certain collective action problems.⁵ Such functionalist rationales illuminate the reasons states may have to comply with specific CIL norms, but they stop short of explaining or justifying the authority of such norms over sovereign states: why are states bound by such norms, and bound by them rather than by other equally functional rules? Others seek consistency by arguing the reverse: consent-based treaty obligations themselves are but instances of CIL since they rest on custom. Such strategies beg the question of whether there are any relevant differences in kind between treaties and CIL – what role does state consent serve? And what to make of the various constraints on when consent binds states, such as in the Vienna Convention on the Law of Treaties (VCLT) Articles 48–53? Some instead bite the bullet and downplay CIL as a source of international law, because treaties have come to replace and strengthen such unwritten norms.⁶ But such historical and causal claims are contested, and fail to account for all CIL – whether, for example, the authority of pacta sunt servanda only rests on states’ consent ‘all the way down’.

The following reflections pursue another answer to why and when non-consent-based CIL enjoys legitimate authority over states. A plausible account of why states have an obligation to honour treaties they consent to also contributes to justify states’ obligation to honour CIL norms – which they have given normatively significant consent to. The shared normative basis for both sorts of obligations may be a ‘Principle of Non-manipulation’, a norm to not violate intentionally created rightful expectations. The account draws much on Scanlon, MacCormick and Hart.⁷

The aim of these reflections is limited to identifying one shared normative premise for the authority of treaties and of CIL: the principle of non-manipulation. This does not exhaust their normative premises – such as the justification of the jus cogens status of some CIL and their constraints on treaties. The account does not deny that state consent is often a valuable mechanism for states to commit to new obligations – to the contrary, there are reasons to endorse the consent mechanism for states to create legal obligation. But rather than regard it as the centre and normative bedrock of international law with some odd epicycles of exceptions and conditions, we should understand the consent mechanism as embedded in a broader normative structure for international law. Some such normative premises also allow us to justify some non-consent-based CIL.

Section 2 motivates the quest for the grounds for valuing state consent. Section 3 defends state consent as necessary to render certain PIL norms legitimate authorities for international judges and other compliance constituencies. The justification is partly based on a normative principle of non-manipulation, to not harm other actors who have formed reasonable expectations about our future conduct on the basis of our deliberate attempts to foster such expectations. States can use consent to send such complex signals to others, that it henceforth regards some norm as an exclusionary, somewhat content-independent reason for action.

Sections 4 and 5 considers why non-consent-based CIL may also enjoy legitimate authority over states. The same principle of non-manipulation may contribute to justify norms that bind states even though they have not performed any recognisable act of consent.⁸ Section 6 brings this account to bear on some contested conclusions of the International Law Commission’s (ILC) final report on the ‘Identification of Customary International Law’ (Conclusions or ILC Conclusions). This account may also guide the unavoidable broad discretion judges must exercise in discovering and creating CIL.⁹ The concluding section responds to several of the apparent puzzles about the scope and conditions of state consent to valid treaties, and considers objections concerning the relationship between PIL and normative premises such as the principle of non-manipulation: that it is too vague or that the account is unhelpful adds nothing.

2 Some Challenges and Puzzles Concerning State Consent in PIL

Of particular concern here are several apparent puzzles concerning the relationship between CIL and state consent. A wide range of authors have sought to explain the binding nature of all traditional sources of PIL exclusively based on consent. Some claim that there are great benefits to require state consent as the sole basis of the legitimate authority of international law. However, such a requirement also entails some costs: consent by all states serves to prevent agreement to ‘global public goods’, so that important human interests go unmet.¹⁰ Leaving the cost-benefit issues aside, consider various claims about why consent enables states to create new legally binding obligations. Consent is often regarded as a central expression of state sovereignty, yet many challenge the central role of it in endowing international law with normatively legitimate authority in the following sense: how – if at all – and when, does state consent justify the claims of each of the sources of PIL to be legitimate authorities? That is: how does their consent give states new reasons to act otherwise, and judge new reasons when they interpret and adjudicate?

I shall suggest that CIL plays several roles with regard to when the mechanism of state consent creates valid PIL. The relevant conception of state consent should be consistent with and might help justify that source’s claim to be a legitimate authority. That is, what are the scope conditions and background requirements for when state consent creates new rules that enjoy legitimate authority over particular subjects?

I submit that the roles for and constraints on state consent in current CIL may identify a broader normative framework for when CIL (and other sources of PIL) is a legitimate authority for states and international courts (ICs): when do these norms give the actors reason to act differently than they otherwise would, and what is the appropriate role of state consent? On closer scrutiny, it seems that states’ consent is neither necessary nor sufficient for them to have or acquire legal obligations as a matter of PIL.

2.1 Challenges to State Consent as a Necessary Normative Basis for CIL

Several arguments based on the origins of such consent would appear to fail. Consider a frequent form of argument that draws from premises akin to ‘the presumptive ability of State representatives to speak and act on behalf of nations and their citizenry’.¹¹ State consent on this account binds because it expresses the moral autonomy of its citizens. Such a democratic, person-based line of argument appears implausible for PIL, at least without further elaboration. It seems unable to account for how authoritarian non-democracies can be bound – if at all. Their claims to act ‘on behalf of’ their citizenry seem blatantly false in the absence of any signalling or accountability mechanisms linking the executive to the citizenry’s preferences.¹² Some may bite the bullet and conclude that non-democratic states cannot bind themselves by consent to treaties. Yet they – and other states – appear to hold otherwise, and there seems to be good reasons to hold non-democracies to be legally – and arguably morally – obligated to treaties when that is crucial to secure objectives such as combating climate change, piracy, or human rights violations. How can this be?

A second kind of challenge to claims that state consent is the sole reason that PIL binds states as a legitimate authority concerns the risk of infinite regress – a ‘chronological paradox’.¹³ In particular, why is it that their consent binds states? Appeals to the fact that states agreed to the legal norm ‘pacta sunt servanda’ in the VCLT would appear relevant, but is insufficient. Several states have not consented to the VCLT. Moreover, how could states’ act of consent to the VCLT create a state obligation, if state consent had no such magical consequences for the states prior to their consenting to the treaty? The binding force of the practice of state consent itself cannot only rest on the binding force of state consent ‘all the way down’.¹⁴

In order to accommodate CIL as based on state consent, Suarez and many later scholars have appealed to versions of ‘tacit’ or ‘presumed’ consent.¹⁵ Such strategies meet with a range of objections. Who has the authority to determine the specific content of norms that a state tacitly consents to?¹⁶ Are these norms ‘particular’ among a limited group of states only, or general? How can we distinguish normatively binding tacit consent from ‘tacit aquiescence’,¹⁷ or from submission as the only alternative to destruction?¹⁸

The ILC Conclusions explicitly dismiss state consent as a plausible ground for CIL. They note the peculiar nature of CIL as not based on formal consent. The Conclusions include the term ‘opinio juris’ alongside ‘accepted as law’, because ‘it may capture better the particular nature of the subjective element of customary international law as referring to legal conviction and not to formal consent’.¹⁹

Several authors argue that these foundational challenges and various peculiar features of the practice of state consent gives reason to question the significance of consent for issues of legitimate authority.²⁰ These concerns should not lead us to dismiss state consent too quickly. The topic at hand – of the legitimacy of CIL – can benefit from reflections on the reasons that may have led us to believe that consent in general, and state consent in particular, endows parts of PIL with legitimate authority – that is, so that states may have moral obligations to comply with PIL.²¹

2.2 Puzzles of State Consent

There are several striking features of the role of state consent. Standardly, a state’s consent must be informed and voluntary to create an obligation, not ‘procured by the threat or use of force’.²² However, aggressor states may be bound by peace treaties even if they sign only due to threats carried out in accordance with the UN Charter.²³

States are also subject to some legal obligations without their express consent – such as jus cogens norms, ‘accepted and recognized by the international community of States as a whole’, but not necessarily accepted by every state.²⁴ Indeed, the number of non-consent-based CIL suggests that state consent is indeed the exception. Consider the weakened persistent-objector rule, third-party effects of treaties giving rise to a general practice, majority voting within treaty bodies, law making by international organizations and international courts, invalid reservations, or the severability doctrine.²⁵ Sometimes states appear to agree to treaties that crystallise existing CIL, such as pacta sunt servanda as recognised in Article 26 of the VCLT.²⁶ Their consent then does not create new obligations, but the reverse: they consent as a way to recognise and confirm that they regard themselves as already bound. Finally, some CIL norms such as jus cogens circumscribe the material contents of the legal obligations states can create through their consent – even retroactively.²⁷ Indeed, new jus cogens norms may in principle even invalidate treaties retroactively. How can such constraints be justified?

So consent plays important roles when PIL makes claims to be obeyed or deferred to.²⁸ But the mechanism of consent is neither always necessary, nor always sufficient for sovereign states to be under, create, or avoid, legal obligations. State consent thus seems to create new binding obligations only within some material domain and when some procedural conditions are met. Not all these rules that regulate state consent can exhaustively be based on state consent. A better understanding of the present practice calls for further reflections on the reasons to value state consent, thus regulated, in creating legitimate authority.

3 Why Value State Consent?

The practice and puzzles of state consent sketched above calls for further reflection about the role of state consent, so that we neither dismiss CIL and other PIL norms simply because they are not based on state consent, nor rashly approve the tendency that state consent is ‘falling out of favour’.²⁹ The conditions, limitations and incompleteness of state consent as necessary or sufficient conditions for some of the states’ legal obligations may help us identify the justification that such circumscribed state consent provides, to understand why and when international law in general, and CIL in particular, can correctly claim to enjoy legitimate authority.³⁰ The standards for when CIL is a legitimate authority may be different for states, for judges who interpret and adjudicate international law, and for other ‘deference constituencies’.

Such reflections about how well state consent secures certain values may help us assess both criticisms and improvements to the present alleged central role of state consent. On the one hand, the mechanism of state consent may privilege the status quo unduly, or give too many actors untrammelled discretionary power to avoid morally required obligations, and block beneficial or urgently needed treaties.³¹ On the other hand, improvements will presumably be better specified in light of the reasons we have to value international law, so that we can further fine tune the conditions and exceptions for when state consent creates binding legal obligations – including a better understanding of why and when we have reason to value also the consent of non-democratic states.

Attention to our reasons to value state consent may thus also help us acknowledge certain non-consent-based sources of PIL – be it soft law, lawmaking by international organizations and by international courts – and CIL. Their lack of state consent may be open to benefits but also further risks – risks that may be addressed in light of such a broader understanding. So: why might sovereign states, born free, consent to live in chains? What reasons do they have to consent to treaties in order to publicly acquire or acknowledge an obligation to defer to such norms – to subject themselves to their authority?³² In general, state consent grants states some influence over the future within some policy space that may be of value to them and their citizens for several reasons.

3.1 Non-domination

State consent grants states a domain of protected sovereignty that reduces domination: it reduces risks that other actors can arbitrarily determine its options, intentionally or otherwise. Such legal sovereignty of course does not yield perfect protection against domination, as the citizens of Melos realised, and begs important questions about the domains of such sovereignty.

3.2 Some Strategic Control Over the Future

State consent may also increase a state’s control, to better carry out and adapt the policies its government believes it has reason to pursue. If other actors recognise such consent as a state’s attempt to self-bind, the state may secure outcomes otherwise unachievable. The consent must then trigger some new (exclusionary and somewhat content independent) reasons for the state to act otherwise that it has reason to value, and that other actors must understand and respect. Such strategic control is of value only when the state is actually able to identify, assess and select among its options, and when its preferences are not objectionable. Reversely, state consent lacks such value when a state is ignorant of the consequences of alternatives, acts under duress or is unable to reflect about the choice³³ – or if the choice has detrimental effects on other parties – for example, if it violates human rights.

3.3 Predictability

A state’s express consent may increase its future compliance with international law, and hence increase other actors’ ability to predict its actions. This process requires a shared understanding that consent imposes legal obligations that exclude or override some other reasons to act. Note that such predictability is only of value if the state’s future actions are in fact of value for others. And this may be the case also concerning acts of authoritarian, non-democratic states. And predictability may be enhanced even with some escape clauses for emergencies and exemptions – as long as they cannot be too readily abused. Note that such predictions only work if a state succeeds in conveying its ‘intention to abide by a rule’³⁴ by a signal that convinces other states. The practice of state consent is one way to signal such complex intentions – if the actors know that the consenting state generally respects international law.

3.4 Status Equality

A further value of state consent is to express status equality among states.³⁵ So if some states enjoy such legal sovereignty within some domain, it is of value for other states and their citizens to enjoy the same domain for consent. Smaller such domains give rise to two distinct concerns. The state may risk domination by other states, and it expresses their relative inferiority. Oppenheim arguably made the latter point thus: ‘In entering the Family of Nations a State comes as an equal to equals; it demands a certain consideration to be paid to its dignity, the retention of its independence, of its territorial and its personal supremacy.’³⁶ Note that this value does not require any particular domain of issue areas for binding state consent, as long as all states enjoy the same legal domain.

3.5 Some Implications

One important type of cases where state consent may promote some of the values identified above is when consent helps secure coordination for mutual gain. The requirement of consent may help states negotiate on a somewhat more equal basis when they specify shared objectives and select among available means.

So on the one hand, to require state consent may facilitate a fairer bargaining process, reducing domination.³⁷ Yet to require unanimity also risks deadlock by states who simply seek a larger share of benefits. There are important coordination challenges: states may agree on overall objectives but disagree about how to specify them and the requisite strategies.³⁸ Indeed, states often face such two-stage challenges: ‘before States can cooperate to enforce an agreement they must bargain to decide which one to implement’.³⁹ So states’ interest in control over outcomes may hinder ‘global public goods’ which they all have an interest in, such as a sustainable environment.⁴⁰

Such benefits and risks may be why member states of the EU agree to (qualified) majority voting for certain issues.⁴¹ Such decision rules may be more appropriate and more likely when states trust that power differentials will not be exploited, and when all expect to often be in the winning coalitions. Similarly, some multilateral environmental treaties and IMO Conventions have secured simplified consent procedures, or developed opt-out clauses etc.⁴² Such weaknesses notwithstanding, state consent may sometimes be the best way to manage constrained, ‘reasonable’ disagreement in ‘battle of the sexes’ situations. One upshot of these reflections is that the mechanism of state consent is double-edged. It can sometimes promote cooperation on somewhat more equal terms, but it can also hinder cooperation, because some states – often sub-state actors within them – insist on agreements that secure them more benefits and less burdens from cooperation.

These brief points have several implications when we turn to consider the legitimate authority of customary international law. The values of state consent draw on some broader account concerning the domain wherein, and conditions under which, state consent is decisive in creating expectations. States’ own choice, sufficiently informed and uncoerced, among certain options within some domains, may help states secure some of their appropriate objectives.

The limits of those domains and conditions are parts of an account of the value of state consent, not a fundamental challenge to the value of state consent. State consent that contributes in this way to endow treaties with authority is not fundamentally a unilateral act that expresses sovereignty, but rather a complex shared practice. As Peters notes, international treaties are thus not simply

the result of unilateral decisions (rational choices) to bind oneself, but [we] can only understand them as commitments towards other actors. The bindingness of a legal instrument (for example a treaty or a pledge not to use nuclear weapons first or to stop nuclear testing) results from the promise given to the other party, and the normative expectations created thereby in the other, and not from a unilateral decision.⁴³

Specification of the domain and conditions of state consent helps delineate the substantive contents of state sovereignty in international law, not the other way around.⁴⁴ Sovereignty is in part defined as having legal standing to enter into treaties within some domain under certain conditions: ‘the sovereignty of the States may be a consequence of these rules, not the rules a consequence of sovereignty’.⁴⁵

Arguments about the domain and conditions for which we have reason to value state consent may help clarify whether the current specifications under international law render sovereign states legitimate authorities, and how to change such specifications to enhance the states’ legitimate authority.

Several of these reasons to value state consent also hold for non-democratic states. They – and their populations – may have reason to avoid domination by other states, and their compliance may often but not always be of value also for others.⁴⁶ Such commitments increase the likelihood of general compliance and hence are sometimes necessary to ensure the objectives of the treaty. Resolving some such conflicts, for example, to promote human rights, or some forms of trade, also with authoritarian states, may sometimes be of value – also for citizens of those states – and hence legitimate. But the treaties such authoritarian states negotiate may be so unjust as to merely whitewash non-democracies. Consent by non-democratic states may also help address the ‘battle of the sexes’ situations, to help determine which of several possible institutions that might provide the desired service, should actually be implemented and recognised as authoritative. But the consent requirement does not guarantee a fair domestic allocation of the benefits and burdens – neither in democracies nor especially in authoritarian states.

It is an open question whether these values – of predictability, control, non-domination, and status – can also be secured sufficiently or to some extent through other means of international lawmaking that do not rely on state consent, without putting these or other values at risk – or indeed in furtherance of such values. The following sections show that CIL may secure and promote these values, even without state consent.

4 Normative Bases for the Legitimate Authority of Consent and Treaties

This brief sketch has not yet addressed the issue of why and how the mechanism of consent may create a normative obligation for states to defer to PIL norms, even if at that time they have countervailing reasons to act otherwise. What are these normative bases that help render PIL legitimate authority for states? We move to that topic now.

It is submitted here that states seek to use consent and promises more generally to create a moral obligation to honour such agreements, to provide assurance among each other. What normative principle would states violate if they fail to act as they have consented to? Consider a specification of the ‘do no harm’ principle as concerns manipulation, to the effect that we should not frustrate intentionally induced expectations. As background, first consider an alternative basis, the normative principle of fair play, which seems insufficient to account for this case.

A principle of ‘fair play’ (or ‘fairness’) prohibits free riding on others’ compliance. As developed by Hart and discussed by Rawls,⁴⁷ it holds that we should do our part as the rules of a practice specify when others have already complied with the practice in ways we have benefitted from. The normative obligations triggered by consent and promises might appear to be of this kind, in that breaches violate a principle of fair play. However, it does not seem correct that states’ obligations first arise when they have received actual benefits from others’ compliance with the practice of promise keeping. Moreover, this account seems insufficient to account for the complex conditions that regulate when state consent to treaties is taken to bind the state. In particular, the principle fails to explain or criticise when and why there are excusing conditions, and the particular scope conditions for binding consent. Such explanations must go beyond simply more careful sociological mapping of the rules of the actual practice, and should instead refer to conditions for, and the domain within which, we have reason to value such a practice of state consent.

An alternative account, developed by Hart, MacCormick, Scanlon and others holds that the normative principle that undergirds promises and contracts is not restricted to existing practices such as the making of binding agreements, but rather concerns manipulation and fidelity. We should not frustrate intentionally brought about expectations of others about our own conduct. They suggest that this is a particular case of the more general principle to do no harm. Fulfilment of such promises is ‘what we owe to other people when we have led them to form expectations about our future conduct’.⁴⁸

Section 3 above indicates several reasons states may have to deliberately seek to influence others’ expectations of their future actions. In particular, we have an interest in constraining our future actions in some ways, in order to have others – states, individuals or other actors – restrict their options and act in certain other ways that benefit us. If we are able to impose upon ourselves constraints that others trust, they constrain their actions in ways that increase our ability to predict and pursue policies we prefer. On this account, to lead others to form such expectations without intending to act accordingly would be a form of manipulation of others for the sake of their own interests.

The consent mechanism, duly specified and constrained, is one effective, precise and valuable mechanism to create such expectations: ‘State consent signals intention to abide by a rule.’⁴⁹ This mechanism enables the promisor to convey quite complex intentions quite precisely, with the aim to establish certain expectations in others. The promisor aims to signal that she would regard it as wrong of her not to satisfy these expectations, indeed even a violation of her legal obligations, except under certain sets of conditions.⁵⁰

One upshot of this account is that it underscores that the limits of state sovereignty is not left for an individual state to decide, and particularly not by express or tacit consent. It is PIL that determines the domain within which states can consent to treaties: ‘sovereignty seems to amount to a large extent to what legitimate international law says it is, and not the other way around, contrary to what consent-based accounts of sovereignty have long defended’.⁵¹ Also note that the conditions for states to enter into legally binding treaties also reflect other values than honouring expectations created under favourable conditions. On this account there are good reasons to have some conditions and domain constraints on such signals. Hart noted some considerations that restrict the conditions where consent can bind, to ensure that the obligations are ones the agents can regard as worth acquiring. There are only under some such conditions that the possibility to acquire obligations and settle expectations is worth having. We can bring some such conditions to light by answering questions such as ‘How important is it to have the selection among these alternatives depend on one’s choice? How bad a thing is it to have to choose under these conditions?’⁵² We now move to consider such conditions on the procedure and the substantive content of the treaties.

5 Normative Bases for the Authority of Customary International Law

Customary international law may serve at least three important roles relating to the principle of non-manipulation, as applied to PIL. Consider first constraints on the procedure of consent, and then on the domain within which states may consent to treaties.

5.1 Procedural Aspects of Becoming Bound by Consenting to Treaties

States often appear to hold themselves to be obligated by ‘pre-existing’ norms such as pacta sunt servanda, and some other norms recognised in the VCLT, such as fraud or coercion as invalidating consent to be bound.⁵³ On this account, these norms serve one role of CIL which states have reason to acknowledge. They help delineate when state consent can provide control over own behaviour and the expectations and behaviour of others worth having for a state. States often have good reason to foster expressions of more precise and shared expectations and have an interest in publicly signalling such expectations to others by means of consent to treaties – but only under certain conditions. For instance, states must have information about some alternatives available to them and their likely consequences, for the mechanism of consent to plausibly signal such complex commitments that states have reason to give. In particular, such signalling is usually of no worth if it takes the form of treaties signed under coercion or due to fraud.⁵⁴

Stating such conditions in the VCLT specifies such obligations in ways that are helpful. But these norms cannot easily derive their binding force from the mechanism of state consent – which they help constitute and specify. Indeed, if these norms were only binding among states who consented to them, this would hinder stability and foreseeability. The role of these norms as helping constitute the valuable practice also explain why new states are bound by them: being a sovereign state entails being able to commit to treaties by consent – which is constituted by such complex rules.

Some such conditions may absolve the actor of any obligations. Some excusing conditions identify when it is not wrong to create false expectations. These include when consent does not sufficiently express the interests of the consenter,⁵⁵ for example when consent is not voluntary, as when a treaty is signed under duress. Similarly, considerations may excuse the agent if they are unable to reflect sufficiently on alternatives and their consequences, for example due to fraud.⁵⁶

Such lack of deliberation and control among the citizenry would be one reason to be wary of consent by non-democratic states. And this is one reason why it may be appropriate to deny non-democratic governments the ability to bind their successors, for example to repay international loans. However, there may be overriding reasons to still grant such states the authority to enter into certain treaties, based on the value of predictability for others, avoiding domination etc.

I submit that the list of excusing conditions for when treaties are not binding may reflect such complex considerations.⁵⁷ The value of honouring expectations may also reasonably be limited under certain drastic, unexpected changes in circumstance – carefully curtailed to prevent abuse and allow credible monitoring.⁵⁸ It seems reasonable that there are circumstances where other parties should accept that their expectations may still be broken. Thus, states are not held responsible for otherwise wrongful acts if they are due to force majeure, in situations of dire duress, or necessary to safeguard its essential interests⁵⁹ – as long as they do not violate peremptory norms. An important issue is who should have the authority to assess such claims of exceptional circumstances, excusing conditions etc., given the risk of abuse by leaving the determinations to the parties of the agreement.

Consider furthermore the particular puzzle of coerced yet binding consent to peace treaties agreed by aggressor states.⁶⁰ First of all, the aggressor state’s interest in continuing the aggression is not a privilege that the state should be permitted to continue to pursue except when consenting otherwise. And an aggressor state that refuses to consent to a peace treaty has no acceptable reason to do so: that option is not one within the appropriate domain of the mechanism of state consent. To the contrary, all parties, even the aggressor state, have an interest in ending armed conflict at some stage before utter destruction. The aggressor state may have an interest in credibly committing to surrender on some terms – if only to prevent further large-scale demolition of its infrastructure and military by other states. Such considerations counsel that peace treaties signed by aggressor states may be recognised as legally and morally binding even when clearly signed under duress.

5.2 Material Aspects of the Domain of Valid Treaties

Treaties are one way to knowingly induce others to form expectations about one’s future conduct. The brief overview above indicates that the possibility of entering into binding treaties has value only under certain conditions. Some CIL norms delimit the domain and conditions of state consent to treaties, to help ensure that this practice of creating legal obligations is of value – and hence legitimate. The value of honouring intentionally created expectations is after all only one of several values, and itself only of value in some cases. We may ask how important it is for the legitimate plans of states to create expectations among others about their future behaviour, and (thereby) for them to be able to form expectations about the behaviour of each state.

Thus consent should not be assumed to bind states to clearly objectionable actions – in which states have no legitimate interest – or when they infringe on the interests of third parties, be they states, individuals, or other entities. The claims to honour others’ expectations do not outweigh or displace all other values. For instance, expectations that would entail human rights violations would not merit much weight. This concern arguably justifies several constraints on the domains where states may consent to treaties, in the form of peremptory norms of CIL – such as jus cogens prohibitions on genocide, torture etc.⁶¹ Just as with the procedural rules, these scope conditions cannot easily themselves be based on state consent for their legitimate authority ‘all the way down’.

5.3 Inducing Expectations

A third role of some non-consent-based CIL are as norms that states have an interest to induce and maintain expectations about, but where state consent is an unsuitable mechanism. To not require universal consent reduces the ‘transaction costs’ of securing general compliance with rules. Moreover, non-consent-based CIL allow states to show that they regard themselves as obligated, and hence that others can be somewhat assured of their future actions – which reduces the commitment problems for some common challenges.

That is: the same general normative principle of non-manipulation explains why these particular customs bind states, namely the obligation to not frustrate intentionally brought about expectations of others about our own conduct. States must take due care to live up to such expectations, and not lead others to form false expectations.⁶² Consider an existing practice with public rules which the practitioners regard as creating obligations for them, so that they do wrong if they violate these rules – except in some overriding circumstances. Such rule-guided practices may also serve as signals about complex intentions about one’s own future action. The practice may induce others to form expectations about the practitioners’ future behaviour, and such expectations may be foreseen and indeed an objective of the participants in the practice. Under such conditions, the same principle on non-manipulation applies to this public practice even though it has not been explicitly consented to. It creates obligations of future compliance among those who uphold the practice.

I submit that states comply with some CIL norms by precisely such practices. Those who have participated in upholding the practice thereby create expectations about their own future actions, and such expectations are part of the rationale of the practice. One aim of CIL norms is to get others to believe that they can expect future behaviour from these parties according to these norms, thus CIL creates and stabilises expectations. This holds true even though states need not explicitly engage the consent mechanism to create such obligations. Why is this, and when might CIL be preferable to consented treaties? On this account, in some cases treaties may be better suited mechanisms for creating such expectations, while in some cases custom may be better – and in yet other cases the two may be equivalent.

There are sometimes reasons to prefer consent-based mechanisms, for example when control over the specifics of the practice is of particular value and can be achieved by the mechanism of state consent, such as when the choice is of contested and different value among the parties, or when the expectations and conditions need to be very precise, or when having the decisions be subject to others’ discretion is objectionable – and where the risk of blocking states is not unreasonable. Cases may include areas when it suffices that a ‘club’ of states move forward, or when the likelihood of non-consent by any state is small.

On the other hand, there are cases where CIL may at least in principle be better than treaties. Some examples are when actual consent is difficult or impossible to secure in advance yet coordination is obviously important for all. Several such cases concern the scope of state sovereignty: the domain of state and diplomatic immunity,⁶³ and the customs of pacta sunt servanda and of how to express the speech act of giving promise in order to engage in the morally binding practice of consent itself indicated above. This is arguably what occurs when states agree to pacta sunt servanda: they recognise, acknowledge and specify a norm that they already regard as binding. Other domains of sovereignty where consent is impractical or impossible concern ways to prevent negative impact on other parties of one state’s choices. The parties may include states, or citizens of that and other states. Thus, some principle of ‘do no harm’ has a long pedigree in international law. For some such norms the stability of the international order is arguably at stake. One example is the right to self-defence, where the issue is paramount yet universal consent is improbable. The ICJ notes that ‘Article 51 of the [United Nations] Charter is only meaningful on the basis that there is a “natural” or “inherent” right of self-defense, and it is hard to see how this can be other than of a customary nature, even if its present content has been confirmed and influenced by the Charter.’⁶⁴

Another set of cases where relying on a custom may be more beneficial than securing consent to a treaty are those where all parties clearly benefit greatly from having one set of shared practices within some domain,⁶⁵ yet where the negotiations about how to distribute benefits and burdens threaten to postpone or even prevent any unanimous treaty. In some such ‘battle of the sexes’ situations, the veto of any state may block any option. One preferable way to identify and implement one of the several possible solutions may then rather be to follow an established custom – as long as it is within that domain of mutually beneficial alternatives. The selection mechanism is thus to comply with the rules that several states already follow – among several that could have been followed. And once a practice is established, there may be a prima facie case to honour expectations according to the principle of non-manipulation – and often a duty of fair play as well.⁶⁶

Scanlon proposes a more precise specification of the reliance on customs as a ‘Principle of Established Practices’, for

cases … in which there is a need for some principle to govern a particular kind of activity, but there are a number of different principles that would do this in a way that no one could reasonably reject. What I will call the Principle of Established Practices holds that in situations of this kind, if one of these (nonrejectable) principles is generally (it need not be unanimously) accepted in a given community, then it is wrong to violate it simply because this suits one’s convenience … it would be reasonable to reject any principle permitting people to violate one of these established practices whenever they wished to do so or preferred some alternative. It would also be reasonable to reject a principle that would require a practice to be unanimously accepted in order to be binding, since if unanimous agreement were required, practices would be very difficult to establish and the needs they serve would be very likely to go unmet. (It is not necessary to insist on unanimity in order to prevent excessively burdensome practices from being made binding, since the Principle of Established Practices supports only practices that themselves cannot reasonably be rejected.⁶⁷

A note of caution is appropriate. There is no reason to hold that any rule generally followed as custom is normatively acceptable. The ‘survival of the fittest’ custom may simply mean that those customs survive that the more dominant states have an interest in maintaining. In particular, the emergence and survival of customs does not reliably depend on their fair treatment of all affected parties. One central reason is that not all states, and certainly not all affected individuals, may have been able to contribute equally to develop the custom to ensure it does not impose unacceptable and avoidable burdens. A wide range of criticisms of PIL make this point – including feminist and Third World Approaches to International Law (TWAIL) arguments.⁶⁸ This condition may arguably support the specially affected states doctrine, and the requirement of the Conclusions that the CIL practice needs to be widespread, representative, constant and uniform, but not unanimous. A further reason to be wary of customs is that they may not adapt well to changing circumstances.

I submit that this argument supports claims that CIL may create moral obligations to comply for states. Some central cases occur when the norms address issues in ways that most affected parties derive large benefits from, without anyone bearing undue costs; and where prior actual universal consent is impossible or impracticable for various reasons. These norms and conditions will have to be specified and identified in certain ways. The delineation in the Conclusions seem generally consistent with this. The specification of conditions may include:

• When universal consent is impossible: that occurs inter alia for the norms that constitute and limit the procedures and domains whereby state consent binds – including those CIL norms that help constitute sovereign statehood; and pacta sunt servanda, jus cogens limitations etc.

• When universal consent is impracticable: in cases where free riding becomes more tempting as more parties join a scheme, or when some parties clearly abuse their power to free ride or to derive undue benefits compared to others.

Such considerations must on the other hand be robust against abuse, for example, when some parties have good reason to not value the alleged benefits secured. This account thus appears to illuminate the ‘puzzle’ of non-consent-based obligations concerning PIL: certain CIL norms may well bind, though they are not consent based.

5.4 The Values CIL Secures

Consider some objections and challenges to this claim, that CIL may be justified on the basis of the same principle of non-manipulation as treaties. One objection is that CIL, thus circumscribed, may not equally well secure the values that the mechanism of state consents, duly circumscribed.

Customary international law which has been in existence for some time prior to the case at hand reduces but does certainly not remove the risk of domination by some states over others. This is one reason to not recognise ‘instant custom’. And this is one reason to have stringent requirements on international courts’ determination of new CIL, to reduce the risk of being subject to the broad discretion of international courts in this regard. Judges’ good faith references to other cases contribute to reducing such suspicion.

Customary international law does not appear to always give states as much strategic control over the future as the veto power granted them by the consent mechanism. While this is correct, as brought out by TWAIL and feminist critics, this comparative advantage may be limited, and only holds for some CIL norms. Some CIL arguably helps constitute important conditions for such control, insofar as CIL (a) helps constitute states and (b) enables states to use consent to enter into binding treaties. Furthermore, the loss of such lack of control has no normative weight not when the discretion lost would anyway only concern others’ legitimate policy spaces, that is, when the jus cogens norms limit options in unobjectionable ways. The disvalue is even smaller when CIL is binding only on some conditions, and for a range of expected circumstances, but allowing certain exceptions – unexpected very high costs, etc. Finally, less powerful states may not even be able to use the consent mechanism to protect themselves against domination by powerful states – as the Melian dialogue showed.

Customary international law that delineates the scope of what other states may do also enables states to predict the future to some greater extent than without such norms – especially if new CIL cannot be introduced suddenly.

Note finally that CIL also preserves the status equality of states, insofar as CIL constitutes and constrains the policy spaces of all states equally and thus does not express relative inferiority of some states. However, the criticism of TWAIL and feminist perspectives remains. The domains of equal treatment may favour certain individuals and states more than others, be it the domains of non-interference or free use of the oceans.

Customary international law secures these values more when the processes of identifying and specifying CIL secures broad representation from diverse states, avoids ‘instant custom’, and when the rules of the practice are clear and not subject to broad discretion by any state.

6 Implications for the Content of CIL

I have suggested that CIL serves several important roles: it delineates some conditions and constraints on both the procedures and the domains for when states’ consent bind them. Customary international law is also a means to create expectations among others that states will be legally obligated. I submit that this account supports and helps specify several of the Conclusions regarding the limited role of consent, and the two criteria for CIL – that there must be ‘a general practice’ which enjoys ‘acceptance as law’ (opinio juris).

6.1 Is There a Role for ‘Tacit Consent’?

On this account, CIL does not require state consent to be binding. What appears to be required is a weaker form of knowing compliance with the rules of a practice, in the form that a range of states (but not necessarily all) participate in a certain practice. A state that participates in such a practice leads others to expect that it will act in certain ways in the future in ways others want to be assured about. And a state that participates in such a practice should be aware that the practice provides such assurance. Other states may join an existing practice, and thereby signal such commitments – without them having any meaningful consent to give or withhold. An exception may be that persistent objectors to certain rules and practices may thereby prevent such assurances about intentions. This is arguably weaker than ‘tacit consent’, and it seems that to use the term ‘consent’ only adds confusion.

6.2 General Practice

This account provides a justification and explication of the requirement of a ‘general practice’ regarding a possible CIL. There is extensive debate on this point. Some urge to replace ‘practice’ due to conceptual problems⁶⁹ and ‘determining the existence of practice is far from self-evident’.⁷⁰ Indeed, critics argue that international judges exemplify Hart’s claim that custom lacks a clear rule of recognition.⁷¹ What is actually practice among states is difficult to discern – and especially in disputes, since parties may point to clusters of states who maintain incompatible practices.

Conclusion 8 maintains that a general though not universal practice suffices – that is, actual consent by all states is not required. But it is important to consider ‘the extent to which those States that are particularly involved in the relevant activity or are most likely to be concerned with the alleged rule (“specially affected States”) have participated in the practice’.⁷² And the practice must be public – known to other states.⁷³ This is one reason why there can be no ‘instant custom’, in the literal sense, though a short period may suffice.

All these specifications seem appropriate requirements to create expectations about compliance with the practice in the future, and to reduce the risk that the practice is a mechanism of domination by some states that is detrimental to others. Note that actual consent is not required – it suffices that states have been in a position to react,⁷⁴ relying on what it is reasonable to assume regarding whether the state has such knowledge.

I submit that these requirements blunt some criticism against the central role of international judges in determining such a practice. They are often said to exercise extensive discretion – firstly in specifying the criteria of CIL,⁷⁵ and then in determining that there is a practice, and which are the rules that govern it.⁷⁶ However, the task of the international judges is to determine that the practice is sufficiently general, especially among those states likely to have a stake in the specific rules of the practice. This is certainly not an easy task, especially because the international judges are also called upon to interpret conflicting evidence about a practice. For instance, verbal practice may easily conflict with what states actually do. And the judges of international courts (IC) must decide which instances of conduct

that are in fact comparable, that is, where the same or similar issues have arisen so that such instances could indeed constitute reliable guides. The Permanent Court of International Justice referred in the Lotus case to ‘precedents offering a close analogy to the case under consideration; for it is only from precedents of this nature that the existence of a general principle [of customary international law] applicable to the particular case may appear’.⁷⁷

What is thus comparable and analogous is to a great extent a matter of intellectual reconstruction, which leaves much discretion to the IC judges. I submit that this discretion is unavoidable. However, there are ways to reduce the possible damage; firstly, to ensure that the judges reflect a broad range of perspectives and are likely to include the impact of different formulations on different states and populations. The risks are also reduced by the requirement that claims that there is such an (emerging) practice must be made public, especially to those states that may have views. And there must be time for states to formulate opinions about such proposals, and possibly express alternative formulations of the rules.

Another risk stemming from this fluidity is the risk of domination by stronger states – and arguably the international judges they nominate, even when the requirement of generality is satisfied. State consent grants each state a veto, which more powerful states may use to maintain the status quo.⁷⁸ Similarly also general principles and CIL that have emerged are likely to promote the interest of strong parties rather than those of weaker states.⁷⁹ Thus, some argue that several great powers moved to establish a prohibition against aggressive wars with the Kellogg-Briand/Peace Pact in 1928, precisely at a time when they had finished conquering and had much to gain by preventing other states encroaching on their (new) territories.⁸⁰

These constraints and guides for the discretion of international judges reduce the risk that they exercise undue influence on the development of CIL. However, this is compatible with substantiated concerns that the international judges fail in their tasks. Indeed, some claim there is ‘a marked tendency [of the ICJ] to assert the existence of a customary rule more than to prove it’.⁸¹ This seems highly problematic on the account explored here, especially for ‘emerging’ customs.⁸² A better response by the international judges would be to ensure that any such claims at least are based on evidence drawn from a range of countries with different legal traditions.

6.3 Opinio Juris

The role of opinio juris in determining whether a norm is CIL is that states must accept this practice as law. That is, the practice in question must be undertaken with a sense of legal right or obligation. It is implausible and anthropomorphic to hold that states ‘have a sense of’ – or that they ‘feel’ or ‘believe’ that there is such an obligation. The present account does not require such metaphors and need not refer to mental entities. The main point is rather that the practice must signal to others that the practitioners limit their action and policies because this is what some norms require or prohibit under certain conditions. These norms are thus somewhat content independent and exclusionary – hence enjoy the sort of authority over the state that law does. I submit that this provides a helpful explication of accepting a practice ‘as law’, or ‘conforming to what amounts to a legal obligation’.⁸³ Evidence of opinio juris should thus not focus on psychological effects, but instead consider states’ responses to reasons of several kinds. What counts as relevant evidence is that states recognise the legitimate authority of such norms as setting other reasons aside. Such self-constraint ‘reflects the existence of a social rule’, arguably showing that the actors ‘take an internal point of view with respect to that behaviour’.⁸⁴ Note that this does not amount to the states actually expressing consent to the norm. This behaviour serves to induce expectations among other parties that the practitioners will constrain their future conduct in similar ways. I submit that there are several ways states can send such signals, including several noted in the Conclusions and by ICs: deliberate abstentions, excuses for breaches, etc.

Evidence of the practice may include inaction – but only when clearly deliberate. Conclusion 6–3 mentions that ‘deliberate abstention from acting may serve such a role: the State in question needs to be conscious of refraining from acting in a given situation, and it cannot simply be assumed that abstention from acting is deliberate’.⁸⁵ Other expressions of state intention may include breaches if they are treated as if they are breaches of apparent legal obligations, and not of etiquette or of mere patterns of behaviour. The relevant state behaviour may include denials of the actions, or arguments concerning excusing conditions, etc. This is in accordance with the Conclusions and the reference to the ICJ’s Nicaragua judgment:

instances of State conduct inconsistent with a given rule should generally have been treated as breaches of that rule, not as indications of the recognition of a new rule. If a State acts in a way prima facie incompatible with a recognized rule, but defends its conduct by appealing to exceptions or justifications contained within the rule itself, then whether or not the State’s conduct is in fact justifiable on that basis, the significance of that attitude is to confirm rather than to weaken the rule.⁸⁶

Evidence would include efforts by states to defend or justify their non-compliance. This indicates that they acknowledge that there is a presumption of deference, violations against which require defence and that such defence may be of the form of plausible specifications and exceptions to the rule or overriding reasons. This account also appears to fit with the Conclusions’ discussion of the persistent objector rule:

the State must express its opposition before a given practice has crystallized into a rule of customary international law, and its position will be best assured if it did so at the earliest possible moment. While the line between objection and violation may not always be an easy one to draw, there is no such thing as a subsequent objector rule: once the rule has come into being, an objection will not avail a State wishing to exempt itself.⁸⁷

Two implications of this are worth noting. This underscores the need for clear statements by those who formulate CIL norms, such as international judges, prior to them being expressed as binding CIL norms. Secondly, new states have no options in this regard – they do not have the possibility to oppose. On this account, that is not an objection to CIL norms being binding, since consent by states is not required to be bound by them.

7 Conclusions: What Is the Added Value of This Account?

The main argument of these reflections is that the reason why CIL can create morally binding legal obligations for states is the same as why states can do so by consenting to treaties: to honour intentionally created expectations among others about one’s future actions. The aim is not to reject several important roles of the mechanism of state consent in current international law, but rather to revise the grounds for its centrality in light of the grounds for valuing such consent. These include the underlying value of fixing expectations and providing assurance of own future conduct. This helps explain why state consent is sometimes but not always appropriate, and neither necessary nor sufficient for a state to create new legal obligations for itself.

States have reasons to value both consent-based obligations, and non-consent based CIL. They are both important signalling mechanisms. This account appears to help resolve several puzzles of the role of state consent. States’ consent may sometimes create binding obligations even under conditions that usually are taken to invalidate consent. Aggressor states may be coerced into signing valid peace treaties,⁸⁸ because we have good reason to allow states to express such commitment to peace and to create expectations among others about such commitments, to avoid further bloodshed.

I submit that this account also helps justify why states are subject to some legal obligations without their express consent. Some CIL norms are binding in part because they are necessary to constitute state sovereignty and to delimit the domain and conditions for the practice of state consent to establish the legitimate authority of new PIL. On this account, that does not seem paradoxical: states have good reason to foster such expressions of more precise and shared expectations, and have an interest in publicly signalling such expectations to others – even without explicit consent.

On the other hand, states cannot by their consent create legally binding obligations to anything without limits. Some jus cogens norms limit the material contents of the legal obligations states can create through their consent.⁸⁹ States also often appear to hold themselves to be obligated by ‘pre-existing’ norms – such as some CIL, for example pacta sunt servanda as recognised in the VCLT.⁹⁰ One main reason is that such limits on the domain of state consent to treaties helps ensure that valid treaties enjoy legitimate authority – jus cogens violations are treaty agreements states have no good reason to pursue, and which they have no good objection to others violating. Furthermore, CIL may serve to help states pursue other shared objectives, especially when there are difficulties in securing universal consent, and when there is low risk involved – for example, as long as the practice is ‘general, meaning that it must be sufficiently widespread and representative’.⁹¹

Consider in closing concerns that such an account is flawed because it draws on moral premises to account for the legitimate authority of PIL. Such contributions are hence of little use, since moral norms about honouring expectations are too vague or contested. Thus Koskenniemi rejects ‘a conception of justice at the root of all customary rules – the principle “legitimate expectations should not be ignored”’.⁹² At least two responses are appropriate. Such claims about vagueness and contested-ness would seem to be comparative. We would need to hear other accounts of the normative grounds for the claim to legitimate authority of PIL – including CIL, which provide a reasonable reconstruction of the practice albeit with critical perspective. We should look closely at such accounts. Secondly, the moral premise used for this account is not a general principle to ‘honour legitimate expectations’. Rather, it is more limited: we should not frustrate intentionally brought about legitimate expectations of others about our own conduct. It is only legitimate expectations that impose an obligation – where the ‘legitimacy’ of these expectations is unpacked in light of what we and others have reasons to pursue and expect of others, for example so that expectation of complicity in torture has no standing as an objection to not fulfilling treaty obligations in this regard. Furthermore, states may follow some patterns of behaviour without any interest in bringing about expectations among others about their behaviour. I submit that the practices that contribute to CIL are different: their rationales and justifications require that other parties take the behaviour as evidence that the states will seek to comply with the rules also in the future. Some may further challenge this account because there is no value added by states solemnly agreeing to norms that are morally binding anyway – be it pacta sunt servanda, or as Koskeniemmi notes:

it is really our certainty that genocide or torture is illegal that allows us to understand State behavior and to accept or reject its legal message, not State behavior itself that allows us to understand that these practices are prohibited by law. It seems to me that if we are uncertain of the latter fact, then there is really little in this world we can feel confident about.⁹³

But Koskenniemi also claims that:

[a] norm is jus cogens … not because it was so decreed by God, or because according to this or that theory it is necessary for the survival of the human species. It is jus cogens if and inasmuch as, to quote Article 53 of the Vienna Convention on the Law of Treaties, it ‘is a norm accepted and recognised by the international community of States as a whole.⁹⁴

One challenge to Koskenniemi’s brief account is cases where states claim that they are recognising ‘pre-existing’ jus cogens norms that have normative authority prior to or independent of their consent. If they are correct, it is not obvious that it is only states’ consent that render such a constraint normatively legitimate, which seems to be Koskenniemi’s claim. Instead, it seems we need to look further. We need not reject state consent, but combine it with something more, to help explain under what conditions the claims to be legitimate authorities on behalf of CIL and other sources of IL are correct.

I thus submit that some general moral norms can also be part of the justification for why jus cogens norms are correct in claiming legitimate authority over states – while accepting that state consent is often also required. Such moral norms must be specified in order to suit the particular circumstances, for example that the actors are states, with new risks of abuse of rules; and the added value of having such norms be public knowledge. The task, on this view, is to dismiss neither moral principles, nor all roles for state consent, nor all non-consent-based IL, but seek to bring somewhat more order into our considered judgments on these complex issues.

7 Custom and the Regulation of ‘the Sources of International Law’

Diego Mejía-Lemos

It is the practice of states which demonstrates which sources are acknowledged as giving rise to rules having the force of law … Article 38 of the Statute of the International Court of Justice … [which] cannot itself be creative of the legal validity of the sources set out in it … is, however, … authoritative generally because it reflects state practice.¹

Article 38(1) of the Statute of the International Court of Justice … is regarded as customary international law.²

1 The Regulation of Sources of Law

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

Article 38(1) of the International Court of Justice (ICJ) Statute³ almost invariably features in all of these controversies. More importantly, albeit rarely acknowledged in contemporary scholarship, ICJ Statute Article 38(1) is invoked in state practice regarding the sources of law in international law. Two positions about such state practice can be found among scholars and international courts and tribunals. Some accept that regulation patterns are inferable from such state practice, but stop short of drawing any legal consequences from those patterns. Others go further and argue that such patterns in state practice do have legal consequences, amounting to law in the form of rules, particularly general rules of customary international law (CIL). In this vein, ICJ Statute Article 38(1) is said to reflect such CIL general rules. The former position, which the first epigraph epitomises, has become the standard one. The latter position, of which the second epigraph is illustrative, enjoys support among some leading, mostly early, commentators⁴ and, at least, one international court.⁵

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

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

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

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

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

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

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

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

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

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

There are two bodies of materials in which ICJ Statute Article 38(1) is used outside ICJ proceedings: decisions of international courts and tribunals and state practice itself. Bearing in mind the difference between these bodies of materials is very significant (a question to which Section 3 returns), and although this section is mainly concerned with selected state practice, it is worth recalling that the place of ICJ Statute Article 38(1) in decisions of international courts and tribunals is widely acknowledged.²⁷ For instance, Crawford observes that ‘[ICJ Statute] Article 38(1) has been taken as the standard statement of the so-called “sources” of international law for all international courts and tribunals’.²⁸ Charney, in a study concerning the proliferation of international courts and tribunals, reached a similar conclusion.²⁹ He inferred that uniformity among international courts and tribunals regarding the sources of law shows that the proliferation of international courts and tribunals has not eroded ‘the international law doctrine of sources’.³⁰ The aforementioned reliance on ICJ Statute Article 38(1) by contemporary international courts and tribunals gives continuation to the analogous practice of arbitral tribunals constituted prior to the adoption of the ICJ Statute. Those tribunals invoked Article 38 of the Statute of the Permanent Court of International Justice (PCIJ).³¹ As Mendelson observes, PCIJ Statute Article 38 had been ‘treated as an authoritative list by various arbitral tribunals’.³² Earlier scholars had also recognised the significance of PCIJ Statute Article 38.³³

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

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

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

The reference to ICJ Statute Article 38 in other major multilateral treaties lends additional support to its wider role in the regulation of sources of law. For instance, the reference to ICJ Statute Article 38 in Articles 74(1) and 83(1) of the UN Convention on the Law of the Sea³⁷ is considered as a general ground for denying the character of equitable principles as legally binding.³⁸ Other major multilateral treaties, in which no reference to ICJ Statute Article 38(1) is made, are widely regarded as having been negotiated on the understanding that ICJ Statute Article 38(1) underpinned the terms used, as exemplified by Article 42 of the Convention on the Settlement of Investment Disputes between States and Nationals of Other States.³⁹

Various states have made multiple statements to the effect that ICJ Statute Article 38(1) sets out the sources of ‘positive’ international law exhaustively and satisfactorily.⁴⁰ Even more pertinently, in some instances, states may indicate that ICJ Statute Article 38(1) provides a ‘legal basis’ for statements about sources of international law. In doing so, states rely on ICJ Statute Article 38(1) as a ‘legal basis’ for the sources of law, whether seen as a category in whole⁴¹ or with regard to individual sources. Furthermore, states rely on ICJ Statute Article 38(1) in order to deny that a subsidiary source is a proper source of law,⁴² or to substantiate their affirmation that individual recognised sources are indeed sources of law proper.⁴³

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

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

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

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

The character of decisions of domestic courts as general practice has raised various questions, which call for some elucidation of their precise nature,⁵³ before turning to the question of the existence of a CIL on sources of law based on states’ general practice primarily in the form of decisions domestic courts. It is common to treat selected judicial decisions as ‘subsidiary means’ under ICJ Statute Article 38(1)(d).⁵⁴ This tends to be the case despite their multiple roles.⁵⁵ One of those roles is as a form of general practice under ICJ Statute Article 38(1)(b).⁵⁶ Lauterpacht had reached the same conclusion regarding PCIJ Statute Article 38.⁵⁷ Several scholars likewise accept that these two roles may be concurrently performed.⁵⁸

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

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

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

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

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

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

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

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

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

Various leading authors have more recently noted that declaratory character is attributed to ICJ Statute Article 38. With respect to ICJ Statute Article 38(1)(c), Jennings and Watts, despite their petitio principii objection, reported on the ‘fact’ that ‘a number of international tribunals, although not bound by the Statute, have treated that paragraph of Article 38 as declaratory of existing law’.⁷⁹ Monaco pointed to the role of PCIJ Statute Article 38 in giving concrete expression to a ‘preexisting practice’.⁸⁰ Sur, likewise, attributes declaratory character to PCIJ Statute Article 38.⁸¹ Pellet, in his 2012 survey of uses of ICJ Article 38, discusses various international instruments, and implies that some refer indirectly to ICJ Statute Article 38.⁸²

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

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

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

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

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

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

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

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

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

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

4 Conclusions

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

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

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

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

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

8 The ILC’s New Way of Codifying International Law, the Motives Behind It, and the Interpretive Approach Best Suited to It

Luigi Crema

1 Introduction

Over the past two decades, the International Law Commission (ILC) has taken an original approach to its ‘progressive codification’ activity.¹ For the most part, in defining international law, the ILC has tended to set aside the attempt to draft articles and pursue binding rules,² expressing instead a descriptive conception of its own work, mostly dedicated to the general framework (i.e., sources, interpretation of norms, responsibility), where specialised, codified in written treaties, regimes operate. This course is not necessarily an independent shift by the ILC itself but reflects the expectations of states nowadays with respect to the commission. To ponder the work of the ILC is an exercise in both understanding how the commission composed by independent experts understands its job and understanding how states look to govern the international community. This chapter aims at reflecting on the activity of the ILC and analysing its transformation over the past decades. In particular, it looks at the way it has, in agreement with the international community of states, interpreted its role, in particular of distilling written restatements of customary law, capturing and crystallising the otherwise often murky and erratic international practice.

The ILC lately tends to avoid the progressive development of law, rather providing restatements of international practice and scholarly doctrines on given topics. In other words, the ILC looks more and more to the codification of customary rules in written, not binding, form rather than aiming at shaping future practice through new treaties. In recent decades, the ILC has come to rarely draft articles aiming at becoming treaties. It occasionally works with this aim, but the success rate of ILC draft articles turning into a treaty in force has been statistically negligible over the past quarter century. Today, the drafting of written material rules tends to happen on different, bilateral or regional, tables, but not at the ILC.

Sections 2 and 3 of this chapter are dedicated to parsing the activity of the ILC and looking at the contemporary revolution in how it carries out its activity of codification, where the key aspect of the activity of codification is not the creation of treaties, but rather the synthesis of existent practice in order to assist the judicial bodies. Notably, its efforts to create general codifications of law have become outdated. In part, the new course of the ILC is geared toward the creation of guidelines, in part toward the scholarly study of topics of general interest, and in part, finally, toward codifying draft articles aiming at becoming treaties – but without going toward any international conference that intends to negotiate a final text.

All these materials have in common the fact of offering written rules, the binding status of which is not clear, and of creating questions about how to assess their contents. As non-binding restatements of international practice, they (both as guidelines and as articles that do not materialise into a convention) amount to a restatement of practice and are, therefore, open for the assessment of the content of the rule they express as is the case for customary rules. Since they are in written form, in an international law document or a document governed by international law, they call for interpretation using the tools of written international law – Article 31–33 of the Vienna Convention on the Law of Treaties (VCLT), which is, nowadays, used not only to interpret treaties, but also every act of international law.³ This chapter has several aims. The first is to illustrate the presence and extent of the shift described here, with a look at the factual grounds. The second is to reflect on the reasons for this transformation. The last is to reflect on the interpretive approach toward the work of the ILC in defining non-binding (but written) guidelines and articles (that do not become treaties).

In order to do this, the next pages will be dedicated to providing an overview of the work and evolution of the ILC approach to codification and development of international law (Sections 2 and 3), its legality under the UN Charter and the ILC Statute (Section 4), and its possible reasons and purposes (Section 5). The chapter will close with an assessment of its effective impact in the work of international courts and tribunals (Section 6), and explain what are the goals and principles which the ILC should consider in carrying out this new course and the interpretive approach that best suits it.

2 A Brief Survey of the Kind of Activities of the ILC Since the Beginning of Its Work: 1949–2020

Since 1949, when the ILC began its operations, it has addressed many different issues, particularly concerning the sources of international law and the law of international relations. As of 1 July 2020, counting its works on the most-favoured-nation clause (MFN clause) as two separate works, the ILC had completed forty-three topics.⁴ Five topics were discontinued or not pursued further.⁵ Eight topics are still under consideration.⁶

Looking only at the concluded topics, in sixteen cases, the ILC approved articles that eventually were brought to a diplomatic conference and culminated in a multilateral treaty. Some of them are well known: the law of treaties, the 1958 conventions on the law of the sea, the works on the succession of treaties, and many others, up to the works on international criminal law, which resulted in the Rome Statute of 1998.⁷ The most recent works of this kind were those dedicated to state immunity, completed in 1991, which led to the adoption of the United Nations (UN) Convention on Jurisdictional Immunities of States and their Property, in 2004.

In twelve cases the ILC produced draft articles that did not later become (or, have not yet become) the object of an initiative toward the adoption of a multilateral treaty.⁸ The most famous example is that of the articles on state responsibility for internationally wrongful acts in 2001.⁹ The more recently concluded works could, in theory, still become the object of a diplomatic conference to complete a multilateral treaty: the draft articles on prevention and punishment of crimes against humanity were just adopted by the ILC, in 2019, although several states, including Egypt, Russia and Turkey, had already manifested opposition to this project.¹⁰ At the moment, however, these twelve works have never undergone the difficult path of an international diplomatic conference aiming at reaching a treaty. So, while they are formulated in articles, as definitions and obligations, their binding effect has never been crystalised by any binding treaty.

Three topics produced guidelines or principles.¹¹ Except for the ‘reservations saga’, where the adoption of guidelines was proposed during the extremely lengthy working sessions, for all the other topics, the ILC intended from the beginning to produce this result.

Six topics culminated in studies. The first, on the evidence of customary law, led by Manley Hudson in 1950, concludes with a series of recommendations on the publication of state and international practice. The others are more recent and concern the so-called fragmentation of international law, the principle of aut dedere aut iudicare and the MFN clause, a highly controversial topic in arbitrational awards on international investments.¹² Two other topics, on subsequent agreements and subsequent practice and on the identification of customary international law, were not discussed by a study group but in the ordinary procedure of work. However, like a study group, they produced a set of conclusions.¹³

The six remaining topics are difficult to place in a single category. They include: those on international criminal jurisdiction and on the multilateral treaties concluded under the aegis of the League of Nations, which were structured as legal opinions and lack the depth of a proper study;¹⁴ the works on the soft codification of the Nuremberg principles and on the ‘model rules’ of arbitration procedure (which, with the proper distinctions, could be grouped with works to establish guidelines);¹⁵ the brief work on the formation process for multilateral treaties; and the earliest study on reservations, which was, de facto, rolled into its work on the law of treaties a few years later.¹⁶

These classifications can be further refined and specified, according to the specific procedure and content adopted for each particular topic,¹⁷ but, for the purpose of this chapter, it is enough to generically differentiate between works dedicated to attempts at codification, and other kinds of works.

At this time, July 2020, the ILC is working on eight topics.¹⁸ The three works on sources are intended to produce soft law; those on the provisional application of treaties are geared toward the formulation of ‘guidelines’; those on jus cogens and general principles are geared toward ‘conclusions’.¹⁹ The three works on the environment are similarly organised: the topic about the protection of the environment during armed conflicts is intended to produce ‘draft principles’; the topic on the protection of the atmosphere aims at producing ‘draft guidelines’; while the ILC on 21 May 2019 established a study group on sea-level rise in relation to international law.²⁰ Only two topics are intended to produce articles: the one on the succession of states in respect of state responsibility, and the one on the immunity of state officials from foreign criminal jurisdiction.

3 A Close-Up on the Work of the ILC in the Third Millennium

If we consider only the time period between the year 2000 and today, the numbers are particularly revealing of a shift toward an express intention of the ILC and the UN General Assembly (UNGA) to produce soft law or treatises on international law, rather than draft articles for new multilateral conventions.

Five works adopted guidelines or principles. These included the guiding principles on unilateral acts and on loss from transboundary harm arising out of hazardous activities, both from 2006,²¹ as well as the guidelines on the reservations to treaties, from 2011. Upon close consideration, the (endless) works on reservations, which lasted a full eighteen years, produced something more than a basic set of guidelines, but rather something closer to scholarship: the Guide to Practice on Reservations to Treaties.²² The works on practice and successive agreements and on the identification of customary law resulted in the definitive adoption of ‘Conclusions’.²³ Three works were intended from the beginning to be eminently scholarly. They included, first of all, that on the fragmentation of international law, which began in 2000 and was completed in 2006²⁴. It contains forty-two final conclusions, and the study group itself, at the close of the working sessions, brought the centrality of the over 250-page overall doctrinal work to attention of the UNGA, on the basis of the conclusions themselves: ‘The Study Group stressed the importance of the collective nature of its conclusions. It also emphasized that these conclusions have to be read in connection with the analytical study, finalized by the Chairperson, on which they are based.’²⁵ Second is that on the aut dedere aut judicare principle, completed in 2014, which produced a slim document of less than twenty pages.²⁶ Last is that on the MFN clause, from 2015.²⁷ Like the case before, this study group produced a brief document, containing a concise commentary, just over thirty pages long, on the practice of states that followed the ILC’s earlier works on the same subject.²⁸ Even Nolte’s work on subsequent practice and agreements began, in 2008, in the form of a study group on ‘Treaties over time,’ falling de facto into the body of work on fragmentation and aiming to complete it with a reflection on the other two parts of Article 31(3) of the 1969 Vienna Convention on the Law of Treaties.²⁹ However, as noted above with regard to Pellet’s work on reservations, Nolte’s and Wood’s works are short commentaries with sets of conclusions which make them halfway between the guidelines, and the conclusions of scholarly studies. Therefore, even if some of these topics were not discussed by dedicated study groups, a total of six of the ILC’s recent works involve studies that are chiefly academic.

Since 2000, only six works culminated in the adoption of draft articles, later approved by the UNGA, but never brought to the preparation of an international convention, nor to the convocation of a diplomatic conference.³⁰ The last international conventions which were entirely based on the work of the ILC or which greatly benefited from it are the already mentioned 2004 UN Convention on immunity of States (which originated in a topic completed in 1991), and the 1998 ICC Rome Statute (the preparatory works of which relied on the parallel works of the ILC on criminal law).³¹ Only two topics, diplomatic protection and protection of persons in the event of disasters, culminated in draft articles that are under discussion (and have been for some years) to potentially become an international convention.

As already briefly described above, of the eight topics that are still open and under discussion in July 2020, six are not connected with any plans to draft articles. All works on sources and on the environment are intended to produce soft law (guidelines, principles, of studies with conclusions), and only two of the topics currently under discussion aim at adopting draft articles (whose legal bindingness still needs to be proved through the test of time and of an international convention, or through the test of courts, recognising them as the expression of customary rules).³²

4 The Legality of the New Path Taken by the Commission

This change of direction by the ILC is striking: the production of draft articles is dwindling. First, as a preliminary observation, it must be noted that, while this ILC approach geared toward academic work and drafting of guidelines is new, it is also legal under the UN framework. Article 1(1) of the 1947 Statute of the International Law Commission provides: ‘The International Law Commission shall have for its object the promotion of the progressive development of international law and its codification.’ This provision echoes the meaning of ‘progressive development of international law’ of Article 13 of the UN Charter³³ and seems to extend to any activity which aims at developing and codifying international law. It is true that Article 15 of the same statute specifies what is meant by ‘progressive development of international law’: ‘In the following articles the expression “progressive development of international law” is used for convenience as meaning the preparation of draft conventions.’

Other ILC activities, not geared toward the conclusion of international conventions, would appear to be excluded. However, the same provision adds, further down, that: ‘the expression “codification of international law” is used for convenience as meaning the more precise formulation and systematization of rules of international law, in fields where there already has been extensive State practice, precedent and doctrine’.³⁴ Article 20 of the ILC Statute directs the commission to prepare draft articles with a commentary containing reference to precedents and other relevant data, such as treaties, judicial decisions and doctrine. In this way, the statute opens up to any activity characterised by in-depth study of international scholarship and practice, even when it does not result in the elaboration of draft articles, but in the adoption of guidelines, conclusions, principles or mere studies.³⁵

While the treaty-oriented work under Article 23 of the commission statute has, in some sense, been shelved because draft articles no longer end up in international conventions, the ILC’s work in preparing non-binding guidelines and draft articles ends up exercising the option described by Article 24 to make the evidence of customary law more readily available. Guidelines are technically non-binding, but their importance should not be underestimated. As written summaries of state practice, which look at selected practice and generalise it, the ILC guidelines have potential to be considered as stating binding customary law.

5 Attempts to Uncover the Reasons for the Transformation: The Context of the Fragmentation of International Law

There are many reasons for this evolution, which touch on a variety of different, though interconnected, planes. The first is a sort of renunciation of the work of imposing solutions on contested topics, instead embracing the work of restating international practice (codification of international law). Any hopes that the ILC would create law and propose (and, therefore, impose) it upon states have been extinguished, and the ILC has settled into taking merely descriptive, rather than prescriptive, positions. Using the terminology of the UN Charter, the ILC has taken a step back with regard to developing law, settling on its mere codification. The decisive event pushing in this direction took place during the long development of the articles on the responsibility of states and its final epilogue. After decades of discussions, James Crawford, pointing to the clashing viewpoints of different governments on the contents of Article 19 of Roberto Ago’s 1996 draft articles (responsibility of states for international crimes), eliminated the article itself and moved the draft forward toward its leaner final version.³⁶

Another reason is the fragmentation of the international community, and of international law in many of its forms. There are more states and, concomitantly, more state practice, and, therefore, there are more interests to try to converge into a single rule.³⁷ Moreover, there is a greater number of states aiming at leading the international community: following after the stability of the bipolar phase,³⁸ and the euphoric moment that followed the fall of the Berlin Wall (during which international law was sometimes framed as the tool of a unipolar world)³⁹ diminished, their place was taken by a multipolar situation that was, at the very least, much more complex than the previous one,⁴⁰ if not actually a full-blown ‘international disorder’.⁴¹ Vast international conferences and multilateral treaties became difficult to imagine.

Related to this enlargement of the actors playing at the international level is the explosion of legal scholarship, which offers several contexts for proposing diverging interpretations of the law, with an ever-growing number of reviews and specialised publications.⁴² On the other hand, international law itself started losing its UN-centred simplicity: specialised international law regimes began to emerge with increasing frequency, with the relative proliferation of courts and tribunals, and the resulting ‘judicialisation’ of international law.

The recent work of the ILC has been dedicated to help international law to find its centre,⁴³ fighting back these centrifugal phenomena. The debate on fragmentation of international law began in 1993, when Edith Brown Weiss reflected on the difficult coordination in environmental law due to what she called ‘treaty congestion’.⁴⁴ It then gained momentum in 1995, when Robert Jennings, former President of the International Court of Justice (ICJ), shifted the attention from the possible clash of treaties to the possible interpretive conflicts between jurisdictions. He publicly warned of the dangers that could potentially flow from the introduction of new international tribunals.⁴⁵ It was the time of the creation of ad hoc criminal tribunals, of the Dispute Settlement Body of the World Trade Organization, of the Tribunal for the Law of the Sea and so on. How could international jurists respond to a danger of this kind? With ‘strong’ international law, made up of hierarchies and dogmas?

To a reader in 2020, the solution proffered in the year 2000 by another ICJ president, the French jurist Gilbert Guillaume, to make the ICJ the ultimate guarantor of the coherence of international law, seems somewhat naïve. While Guillaume did not go so far as to imagine the ICJ as a sort of Supreme Court of Cassation of the international legal order (indeed, he observed that appeals and cassation procedures are utilised only very rarely in international law),⁴⁶ he did believe that the court at The Hague should be accepted, at least, as a superior court, with the power to receive requests for clarification ‘on doubtful or important points of general international law raised in cases before them’, following the method of the reference for a preliminary ruling mechanism used by the Court of Justice of the European Union.⁴⁷

The ILC’s reaction to each of these evolutions and its response to these underlying issues over the past two decades have amounted to a total departure from the hierarchical approach of Guillaume. The ILC took stock of fragmentation, and, rather than seeing the proliferation of courts and tribunals as a threat, it saw it as an opportunity to deal with the explosion in the number of legal regimes, states and of legal scholarships: the ILC shifted its potential audience from governments gathered in a multilateral diplomatic conference to litigants and adjudicators. It provides them with guidelines dedicated to sources useful for adjudication, with a plausible, legitimate, common interpretive background in a body of expert scholarship useful for decoding⁴⁸ potentially conflicting interpretations of given rules.⁴⁹ It also proposes (customary – and therefore binding?) restatements of practice about other general issues of international law, leaving it to those called to adjudicate over a specific dispute to provide the definitive answer on their legal bindingness.⁵⁰

In order to achieve this, the ILC set about weakening general international law, in the philosophical sense that it reflects the thought established by a society that lacks consensus on ultimate values,⁵¹ so to be flexible enough to encompass and serve the new, many, international law regimes. The term ‘weak’ is used here because the ILC has been characterised by an approach to defining the law that disregards the goal of final approval of strong, binding, treaties; weak because the power of orienting the conduct of states and adjudicators does not come from a command, but from persuasion, that is from the quality of a given study; weak because it assumes that the validity and authority of a rule does not come from the precision and clearness of its content, but rather from its structured interpretation.

First, the ILC affirms the unity of the international system, not as a single set of binding, material rules, but by proposing the formal unity of its rules dedicated to the determination and establishment of international law. It furnishes parties to a dispute and adjudicators with a common legal ground of principles and guidelines to deal with the sources of international law. At the same time, the ILC has overcome the threat of the explosion of legal scholarships in order to propose, with its guidelines and commentaries, the unification of reasoning about the law, by creating a unique context for its interpretation: a sort of ‘official scholarship’.⁵² The ILC’s series of works on the sources of international law and the interpretation of written law is very clear on this: it is, quite simply, the attempt to state a single line of orthodoxy on secondary rules of international law (according to the Hartian meaning), and to create a line of official scholarship. It both fosters and creates the internal point of view on law and creates an official legal context in which to assess the contents of rules.⁵³ This official scholarship (a doctrine that distils other doctrine and practice) is: authoritative⁵⁴ (because of the credibility and plurality of its source – the ILC of the UN); easily accessible (an easy internet search is all it takes); and open to be used and adopted by anyone involved in a dispute.

Second, the ILC looks at different courts and tribunals as an opportunity to develop international law outside the traditional codifications⁵⁵ – every kind of international law, including material law, not just law dedicated to the sources and their interpretation. Their work can turn soft law (both when it is truly soft law and when it comes in the form of draft articles) into customary law, even in the absence of global agreements.

The ILC approach can be summarised, in essence, to be that of experts whose chief aim is to study and summarise, without imposing anything and without even challenging governments, the UNGA or the Sixth Committee to elaborate international agreements. They leave it to counsels and attorneys, who represent states or private parties before the international tribunals to cite the texts they have produced. Above all, their work falls to the hands of judges and arbitrators who, in the chambers of their respective tribunals, decide what to use, keep or reject of what they have produced. It is redundant to observe that often this process is facilitated by the efficient shuttle-service between Geneva and the seats of arbitration or the tribunals serving the commissioners and the former commissioners of the ILC, the very-visible college of international lawyers.⁵⁶

In order for something to become law in the new millennium, it does not need to pass through long and fraught negotiations, beholden to political positions that are too far apart and, by this time, nearly irreconcilable. When states and other entities involved in a dispute use the ILC’s work product, the third party called upon to adjudicate specific disputes in the concrete can then grant final prescriptive power to an indication contained in a soft-law text or a text that stalled in the draft articles phase.⁵⁷ This route is ultimately more practical than reopening long and often fraught negotiations with the over 190 invited states, and faces less risk of compromise, or even failure. Not even references for a preliminary ruling are necessary, as Guillaume had predicted. The authority of law does not come from a multilateral effort, nor from a clear command of a rule, or from a theory of sources, but rather from a shared discourse about law, and a recognised authority charged to settle a dispute.

The success of the 2001 Articles on the Responsibility of States is illustrative and paved the road: if the work is useful and well done, there is an ‘audience out there’ ready to adopt it and implement it. Crawford and the ILC had put forward not guidelines, but rather articles – but the outcome is analogous – settling for codifying, rather than developing, international law, and leaving the issues of making the rules binding and any potential development of the law to others – particularly international courts and tribunals.

6 A Look at the Practice of International Courts and Tribunals

Under this new approach, draft articles and works of soft law and scholarship naturally flow into the work of international courts and tribunals. Courts and tribunals use some of the articles and conclusions of the ILC as stating customary rules, or as subsidiary means in the sense of 38(1)(d) of the ICJ Statute.⁵⁸ They also use the conclusions of the ILC without qualification. In other cases, the ideas of the ILC are cited by the parties, but ignored by the court or tribunal. Leaving aside the briefs submitted by parties, which make references useful to make their cases, but looking only at the reasoning of courts and tribunals themselves, we can find some illustrative decisions.

As far as draft articles are concerned, along with the 2001 articles on the responsibility of states mentioned above (and their 1997 predecessors whose use in the Gabčíkovo-Nagymaros case⁵⁹ has been already widely commented),⁶⁰ we can also look at how the ICJ used the draft articles on diplomatic protection, produced by the ILC in 2007, in the Diallo case.⁶¹ In that case, despite the fact that a provision put forward by the ILC drew criticism from some governments (Article 1),⁶² the ICJ applied it as an expression of customary law.⁶³ To determine that something is customary law requires a review of the practice, under which the lack of homogeneity of the practice and lack of consensus among states ordinarily would not have allowed for the formulation of a customary rule. The same draft articles were also cited by an investment tribunal in 2014.⁶⁴

One example of successful regulatory cooperation between the ILC and international tribunals involving soft law is the now consistent case law of investment tribunals with regard to the guiding principles on the unilateral declarations of states capable of creating legal obligations. These have been applied in various cases to interpret unilateral acts, although without a declaration of customary law status.⁶⁵ One International Center for the Settlement of Investment Disputes (ICSID) tribunal, in the case Total v. Argentina, gave extensive consideration to the principles in question and, while it did not determine that they were customary law, dedicated an in-depth analysis to them, underscoring their relevance even when interpreting ‘domestic normative acts relied upon by a foreign private investor’.⁶⁶

Last, if we consider the scholarly activities of the ILC, in a very high number of cases the parties invoked study group reports in their briefs. Even if we only look at the reasoning of the judicial bodies themselves, it is clear that study groups, too, have had an impact on the activities of courts and tribunals. The study on the MFN clause, although brought to a close only recently (2015), has already been referred to extensively. The study was brought to bear, in particular, on the merits of a highly controversial issue, that is, the question of whether or not the MFN clause extends to compromissory clauses, starting with Maffezini v. Spain.⁶⁷

It is not surprising, therefore, that international investment tribunals immediately latched onto the study and incorporated it into their reasoning. For example, in the A11Y Ltd v. Czech Republic decision (2017), the tribunal concisely observed: ‘The Tribunal is of the view that an MFN clause can, a priori, apply to dispute settlement. The Final Report of the ILC Study Group on the Most-Favoured-Nation clause is instructive in this respect.’⁶⁸ In the award Le Chèque Déjeuner v. Hungary, two key passages of the Tribunal’s reasoning on the interpretation of the MFN clause are dedicated entirely to quoting and commenting upon the ILC report.⁶⁹ Even the dissenting opinion by Marcelo Kohen attached to the decision is substantially built around references to the ILC report on the MFN clause.⁷⁰

The study on fragmentation, published in 2006, attracted a great deal of scholarly attention, and a less enthusiastic reception by courts and tribunals. Nonetheless, it is easy to find examples in which they refer to it. In the annulment decision Tulip v. Turkey, for instance, an ICSID Annulment Committee formulated the proper way that human rights obligations should be integrated into the interpretation of a state contract through extensive reference to the ILC report on fragmentation.⁷¹ Moreover, in the award on jurisdiction of the RREEF v. Spain case, the tribunal referred extensively to the scholarly work on fragmentation.⁷²

Not all the work produced by the ILC in its ‘new era’ has met with success. One example is its work on reservations, which culminated in 2011, after nearly twenty years, both in a set of guidelines (soft law) and in a thorough scholarly document (the guide). These documents have had practically no impact on international case law. Neither the ICJ nor the investment tribunals have used it, and the Strasbourg Court has only mentioned it in a single case, already eight years old.⁷³

The work of the ILC can also have effects that are not as easily detected as the effects of a citation: its influence may remain in the form of an undercurrent, or an international court may prefer to apply customary law as it is described by the ILC without making explicit reference to the work of the ILC itself. Consider, for example, the 2012 case Habré, in which the ICJ ruled on the erga omnes nature of the convention’s obligations forbidding torture. The ICJ did not mention the articles on state responsibility, but only the customary international law on that topic.⁷⁴ The reasoning, however, clearly reflects the formulation laid out in the ILC’s work.

The decision whether or not to make explicit reference to the source of inspiration for a given rule clearly falls to the discretion of the adjudicating body, and reveals that the choice to specify the ‘places’ in which certain choices are formed and crystallised falls under the policy for legitimising its own work product that a judicial body decides to adopt. At the same time, it also reveals that the influence of the ILC’s work cannot be assessed by merely tallying up explicit references, but may also extend beyond them.

7 The ‘Principle of Quotability’, the Relevance of Official Scholarship and the Importance of an Interpretive Approach Favourable to the travaux of the ILC

On the whole, we can conclude that, during this time of transformation of international society, the ILC is orienting its activities toward ‘dialogue’ with states, other parties and international judicial bodies during a dispute and aims less to elaborate draft conventions. As for the works dedicated to the sources of international law, unity is not reached through an order, nor a shared bedrock of values, but rather through the construction of a common technical language available to international actors and courts, and through reason, which we all still have in common in a fragmented world.

As for the drafting of material rules of international law, the ILC suggests, litigating entities propose, and the courts pick up what they find to be of value and crystallise it into a customary rule.

The ILC’s shift from being prescriptive in its work to being scholarly and descriptive, in which the commission decided to address chiefly courts and tribunals instead of elaborating new international conventions, created the need for the commission’s activities to attain ‘quotable’ results. Quotability means generating an expression of the law (in the subtle form of guidelines or the stronger form of draft articles) that provides a third party – a state, a private entity invoking the protection of a treaty, a judge or arbitrator – with a reference which, notwithstanding the fact that it is not necessarily binding, is succinct, clear and immediately applicable.

In light of this, we may draw three final observations. First, even when the ILC does not intend to create draft articles, but aims at articulating soft law, it is important that the commission not fall into overly descriptive passages in an attempt to avoid making choices unsupported by practice, resulting in tools too unwieldy to apply. When it comes to the documents’ quotability, the ability to make clear choices in the face of non-homogeneous practice is more important than the exhaustiveness of the studies. It is, above all, in this regard that the activity of the ILC must not spill over into works that fail to take a position on controversial topics. The 2011 Guide to Practice on Reservations to Treaties, which has had very little impact so far in international litigation, is illustrative.⁷⁵ The more the ILC aims to make its work exhaustive, at the expense of choosing preferable solutions, the more difficult it is for those who adjudicate cases to make use of the conclusions it adopts.

Second, it is important for the ILC to address issues that are relevant on the practical (judicial, especially) plane, and not only those related to the general framework of the sources and their interpretation. Its study of the MFN clause is emblematic here. While its scope was more limited than that of other recently undertaken studies on sources, it had an immediate impact on international arbitral awards because it stepped in to provide order and clarity in an issue on which investment arbitrations had run aground with conflicting solutions. From this point of view, at the risk of making a false prediction, highly practical studies on hotly debated issues seem destined to have a greater future impact than scholarly analysis on, say, general principles of law, which would serve only the function of offering an official context for assessing international law sources.

Third, the existence of ILC quotable guidelines or articles spares adjudicators from the heavy work of demonstrating the existence of widespread practice and opinio juris and becomes a practical tool to find a guiding legal principal orienting the decision of the adjudicator, and/or giving legitimacy to it – a reference to an external authority is always more legitimate than taking what would appear to be an arbitrary position by adjudicators.

However, the principle, guideline or draft article can be recalled and applied without an extended analysis of the possible nuances of the text only if the litigating parties are not contesting it. In case of disagreement, the customary nature of the non-binding written provision requires the interpreter not to start from the text or the object and purpose, à la VCLT, but rather from the international practice generating it. Unfortunately, it is not common to find a discussion of the appropriate interpretive method of the non-binding provisions of the ILC in international rulings. In the already mentioned Diallo judgment of 2007, an approach keen on customary law would have pushed the ICJ toward an investigation of the travaux préparatoires of that provision, that is, the practice analysed by the ILC and the reactions of governments to it. This would have brought to light that the provision was hotly contested and far from being customary. But the ICJ in that case did not follow the rationale behind restatements of customary rules, but merely looked at the text of a rule, as if it were dealing with a conventional rule, binding for the parties, whose text was sufficiently clear. While this attitude can be accepted as an expression of the deciding power of courts, it does not reflect the customary nature of the process leading to it.

9 Beyond Formalism Reviving the Legacy of Sir Henry Maine for Customary International Law

Andreas Hadjigeorgiou

1 Introduction

Sir Henry Maine is a curious figure.¹ During his time, he was a most famous legal theorist, whose name and fame did not manage to survive the test of time. Nevertheless, his works gave rise to a legacy whose revival can prove invaluable for international law (IL) and, more specifically, customary international law (CIL). This is especially the case as it concerns ontological questions.

While we live in a ‘post-ontological’ era, ‘where the existence of a genuinely legal international order can be safely presumed’,² not all ontological concerns have disappeared; rather they have shifted. Whereas IL is taken as a ‘given’, questions still remain about its genesis and the place CIL holds within it. Some, for example, deny the existence of CIL as law properly so called, while others urge IL to rid itself of the concept of CIL altogether. Further, this position intertwines with a formalist perspective – that is, that law is always the product of formal design³ or formal ascertainment.⁴

Accordingly, formalism in its moderate form treats formal sources, documents and/or proclamations as ‘better’ tools for both (a) the preservation of existing rules of CIL⁵ and (b) the ‘creation’ of new legal rules. At its more extreme, formalism purports the view that (c) IL (or even CIL) finds its genesis only in formal sources, documents or proclamations.⁶ While some formalisation is undeniably helpful and even necessary, we should be more critical of this formalist paradigm. Moreover, these perspectives touch both upon the genesis of IL, but also on its evolution – that is, how law began and how it ought to be further developed. These positions set the wider themes of the chapter. By reviving the legacy of Sir Henry Maine, it seeks to give a new perspective on the genesis of IL as primarily CIL, as well as a new vision of how it should/could be further developed; a vision that goes beyond mere formalism.

2 The Curious Case of Sir Henry Maine

Sir Henry Maine gave jurisprudence a curious spin, one that ended up awarding him the Chair of Jurisprudence in Oxford (a chair later held by Hart and Dworkin). More than that, though, as the University of Oxford’s website testifies,⁷ the chair was established as an effort to attract Maine to Oxford, while accommodating his ‘peculiar genius’.⁸ While jurisprudence has always been a discipline built upon the intersection of law and philosophy; Maine redefined it.

His work attacked the philosophical approaches to law of his time. His claim was that they were based upon speculative narratives of how law began; designed to explain domestic legal systems as we experience them in isolation at a given point in time (the famous sovereign-legislative structures). Instead, Maine suggested that any conceptualisation of law should be based upon a solid historical understanding of law’s genesis and evolution.

The mistake which they committed is therefore analogous to the error of one who, in investigating the laws of the material universe, should commence by contemplating the existing physical world as a whole, instead of beginning with the particles which are its simplest ingredients. … It would seem antecedently that we ought to commence with the simplest social forms in a state as near as possible to their rudimentary condition. In other words, if we followed the course usual in such inquiries, we should penetrate as far up as we could in the history of primitive societies.⁹

Hence, Maine’s work is not a complete theory of law, but rather blueprints for redefining the way law is philosophically apprehended within jurisprudence, by firmly relating law to its empirical, socio-historical dimensions. Maine’s legal history paves the way for something we today coin as interdisciplinary. This enquiry led Maine to understand law as an evolutionary phenomenon and this account holds an astounding amount of invaluable observations, which have yet to be applied to the international legal order.

3 The Genesis of (International) Law as Customary (International) Law

Maine’s most general theme is that law should not be seen as a static, externally imposed system of commands and/or judgments backed by threats; nor should it be seen as a ‘new’ nor ‘recent’ phenomenon. History makes it clear that order and law predate formal organisation, sovereigns or courts. Law’s existence stretches from ancient times. Further, even domestically, law began as custom.

The simplest truism lies in the fact that later formally ascertained law was itself based upon pre-existing customs, the existence and authority of which was accounted for independently of the persons who came to proclaim them. Even more, the first formal proclamations were not legislative acts, but codifications of customary law.¹⁰ Thus, custom is not merely a source of law, it is the first source. From this perspective, social order is the product of custom, or rather an interconnected network of customary practices, which manifest as one organic whole.¹¹

Law, as customary law, is merely one functional subpart of this greater whole. This evolutionary idea was slowly cultivated into interactionism.¹² Order and law appear first as custom, which arises freely and unassisted through the simultaneous interactions of various individuals.¹³ While law is organically connected to the greater customary network, it is at the same time analytically delineable.¹⁴ Various theorists have made interactionist conceptualisations of the international legal order;¹⁵ however the current section would rather build upon this idea using another theme.

As Maine observed, from the beginning of time individuals were not dispersed/unconnected, they were organised. However, they were not organised in societies/communities of individuals as one would expect; rather individuals were organised in societies of institutions, the primary of which was the family. These family institutions were equal units in terms of authority, and the only thing that stood over and above them was customary law – the thing that constituted them. Within the family institution the ‘law’ was the word of the father, but as Maine wisely reminds us:

Society in primitive times was not what it is assumed to be at present, a collection of individuals. In fact, and in the view of the men who composed it, it was an aggregation of families. The contrast may be most forcibly expressed by saying that the unit of an ancient society was the Family, of a modern society the Individual. … Men are first seen distributed in perfectly insulated groups, held together by obedience to the parent. Law is the parent’s word, but it is not yet in the condition of those themistes [judgments of justice/customary law].¹⁶

Customary law is, then, law properly so because that is the part of the greater normative system which organises individuals in the institutional units to which they belong; the units that they themselves, others and society at large understands them through. From this perspective, the reason the word of father (within each family institution) is not yet on par with the legality of customary law becomes apparent: it is customary law which dictates that all individuals, in order to be recognised members of the community, must belong to a family unit. Further, not every family unit will qualify. It is, thus, customary law which dictates that all recognised units must be connected through common lineage. It is also customary law which designates the word of the father as an authority within them. This becomes a truism once it is realised that in other societies (customary) law designates the word of the mother.¹⁷ As such, without assimilating within our perspective customary law, we are unable to understand a society in the same way that it makes sense of its own self and structure.

This has its own conceptual connection to international and CIL. Whereas looking at CIL from within domestic legal systems, it gives the impression that it is a deficient, problematic or an odd instance of law;¹⁸ through Maine’s evolutionary perspective, these characterisations disappear. Law began internationally, in the same way it began domestically, as customary law. Further, in the same way domestically customary law functioned to establish a society of family-institutions (within which the word of the ‘father’ was law), so did CIL constitute an international society of ‘state-institutions’ – within which the word of the ‘sovereign’ is law. Maine was quite aware of this:

Ancient jurisprudence, if a perhaps deceptive comparison may be employed, may be likened to International Law, filling nothing, as it were, excepting the interstices between the great groups which are the atoms of society. In a community so situated, the legislation of assemblies and the jurisdiction of Courts reaches only to the heads of families, and to every other individual the rule of conduct is the law of his home, of which his Parent is the legislator.¹⁹

Moreover, as already noted, customary law also works to establish these atoms. After all, these family-institutions domestically, as these state-institutions internationally, are not entities which exist, or can be understood, independently of rules which first arise customarily. Rather, these state-institutions are conceptual constructions; the result of various rules working together. These include rules that states should be attached to a territory, together with rules which define such territories. Rules that dictate that each state should speak with only one voice and that only one voice from within each state can be accepted as authoritative at a time. It also concerns rules which attribute actions done by individuals to a ‘state’, etc.

Without such common rules to synchronise how and what various individuals see/understand as ‘states’, there could be no consensus that ‘states’ exist, let alone a whole international community of states. Further, rules such as these need by definition to be voluntarily accepted, and need to remain unchallenged, in order to operate. It is at such a foundational level that we must primarily situate our conceptualisation of CIL; at the level where a community constitutes itself freely and unassisted. This nicely brings to mind a quote from Philip Allott’s ‘The Concept of International Law’:

The social function of international law is the same as that of other forms of law. It is a mode of the self-constituting of a society, namely the international society of the whole human race, the society of all societies. Law is a system of legal relations which condition social action to serve the common interest. … National legal systems (including private international law) are part of the international legal system. International law takes a customary form, in which society orders itself through its experience of self-ordering, and a legislative form (treaties). The state of international law at any time reflects the degree of development of international society.²⁰

From this perspective, sovereignty itself is itself constructed through CIL. Maine proposed this conceptualisation as early as 1888, in his posthumous book on International Law: ‘What really enables states to exercise their Sovereignty in this way is nothing but the legal rule itself.’²¹

This rule can be nothing else but a CIL rule. Maine highlights the primacy we must grant CIL over the state units which are constituted, and regulated, by it. This point of view bears close resemblance to what Kelsen argued about three decades later. It is also a perspective argued by Hart, Maine’s successor in Oxford. In the last chapter of his Concept, Hart writes:

For if in fact we find that there exists among states a given form of international authority, the sovereignty of states is to that extent limited, and it has just the extent which the rules allow. … Hence we can only know which states are sovereign, and what the extent of their sovereignty is, when we know what the rules are. … The question for municipal law is: what is the extent of the supreme legislative authority recognized in this system? For international law it is: what is the maximum area of autonomy which the rules allow to states? … [In this way,] there is no way of knowing what sovereignty states have, till we know what the forms of international law are and whether or not they are mere empty forms.²²

This analysis further highlights that CIL and domestic legal systems, at the most foundational level, are organically connected and work together to bring to life these state-institutions. In the same way individuals in ancient society used custom and customary law to constitute themselves as a society of family-institutions, so have individuals internationally used custom and CIL to constitute themselves as a society of state-institutions – and the two interconnect. State institutions internally came to employ a similar hierarchical structure to what was employed internally by the family institutions, as a prerequisite for participating in this international community.

Thus, the communities which failed to take on this internal hierarchical structure, were historically not understood to be real units of the international community and have failed to contribute to its development. Customary international law dictated a greater vision that each unit had to follow both internally and externally in order to ‘count’ as a true member of the international community. A unit which did not adhere, would not be coerced, but rather ostracised. Thus, CIL’s efficacy rests on the fact that it contains the most agreed upon, and desirable, scheme of international co-existing.

As such, the ontological critiques that CIL has traditionally faced from the perspective of domestic legal systems seem to rest on an ahistorical understanding of the genesis of law. Further, this highlights another interesting argument. Without suggesting that international courts merely interpret and apply existing IL and CIL, to the extent that such courts deal with a pre-existing international community made up (at a minimum) of states, this suggests that an effective CIL is already in operation. It is CIL which (at the most foundational level) constitutes the international society – within which such judgments make sense.

4 The Formalisation of Customary (International) Law

Rather than a special, or a defective, case of law, customary law reveals itself as the most fundamental and primary instance of law. Even the oligarchies, which later came to power, recognised its primacy, since even they only claimed to hold knowledge of customary law:

Before the invention of writing, and during the infancy of the art, an aristocracy invested with judicial privileges formed the only expedient by which accurate preservation of the customs of the race or tribe could be at all approximated to. … The epoch of Customary Law, and of its custody by a privileged order, is a very remarkable one. … [Nevertheless] what the juristical oligarchy claims is to monopolise the knowledge of the laws.²³

Exactly because this customary organisation was more complex than lay individuals could comprehend, is why oligarchies were vested with such power. As Maine notes, this power was surely abused, but it still did not amount to Austinian command-sovereignty. Nevertheless, because there was abuse, once writing was invented societies wrote down, codified and (thus) formalised first those customary rules which were law. The XII Tables of Rome were Maine’s primary example and Hart agreed: ‘In Rome, according to tradition, the XII Tables were set up on bronze tablets in the market-place in response to the demands of the Plebeians for publication of an authoritative text of the law. From the meagre evidence available it seems unlikely that the XII Tables departed much from the traditional customary rules.’²⁴ Nevertheless, while this formalisation undeniably had its benefits this had its own adverse effects:

When primitive law has once been embodied in a Code, there is an end to what may be called its spontaneous development. Henceforward the changes effected in it, if effected at all, are effected deliberately and from without. It is impossible to suppose that the customs of any race or tribe remained unaltered during the whole of the long – in some instances the immense – interval between their declaration by a patriarchal monarch and their publication in writing. … It would be unsafe too to affirm that no part of the alteration was effected deliberately. But from the little we know of the progress of law during this period, we are justified in assuming that set purpose had the very smallest share in producing change.²⁵

What Maine suggests is that insofar as law was primarily customary, it was continuously changing alongside society, driven by new social needs and a changing environment. That is, spontaneously developing through practice and always being ‘up-to-date’, in the following way. The two primary tools that practitioners use to familiarise someone to a customary practice are examples (of occurrences of the practice) and rules²⁶ – usually unwritten, that is customary. The two work together. As Wittgenstein reminds us, both rules and examples interrelate to properly triangulate meaning: ‘Not only rules, but also examples are needed for establishing a practice. Our rules leave loop-holes open, and the practice has to speak for itself.’²⁷

Thus, while rule-governed practices might arise without the use of explicitly formulated rules, rules would still necessarily arise in order to (i) introduce/communicate the practice to others, (ii) settle differences and (iii) better preserve the practice in the participants’ collective memory. As such, at a conceptual level, interpreting customary practices is first the process of conceptualising rules out of an activity, and second the process of relating those rules to examples of it as a practice.

Thus, the two, rules and examples, intertwine; however while the rules are customary, examples of the practice take primacy. For a practitioner, the rules one employs in relation to the practice are accurate, and thus acceptable, if they properly explicate at least parts of the practice (i.e., the behaviour of those who engage in it). Nevertheless, whereas the rules remain the same, the practice implicitly grows and changes (even without the awareness of those who engage with it) as it learns to respond better to the same needs, or it begins to respond to new needs.

However, exactly because on the customary level the practice takes the forefront, as the practice grows, existing rules that once ‘fitted’, slowly begin to seem inaccurate or outdated. From there, new customary rules arise which might describe new or changed parts of the practice, even though the rules themselves might be perceived as ‘new (or better) descriptions’ of the ‘same-old’ unchanged practice. The codification and, thus, formalisation of rules, though, severs its relation to the practice and ‘freezes’ rules in time.

While the practice (or examples of it) might still be used to interpret formalised rules; the evolution of the practice can no longer impact upon the validity of the rules. Whereas at the customary level the practice legitimises the rules, once rules are formalised, the situation reverses. In turn, the formalised rules might stop the customary practice from further evolving and, thus, from better responding to existing or changing needs. The same process applies to the formalisation of CIL. As Thirlway observes:

The difficulty is of course that customary law develops of its own accord, without there being any need for States to do more than continue their day-to-day relations, whereas a treaty regime can only be changed by deliberate act of the parties. Furthermore, as Professor Baxter has observed, ‘The clear formulation of rules in a codification treaty and the assent of a substantial number of States may have the effect of arresting change and flux in the state of customary international law. Although the treaty ‘photographs’ the state of the law at the time of its entry into force as to individual States, it continues, so long as States remain parties to it, to speak in terms of the present.’²⁸

Thirlway goes even further to note that codification might have a halting effect upon the customary practice of even non-parties to the treaty.

Thus so far as the States non-parties to the treaty are concerned, for whom the codifying treaty is only evidence of the state of customary law at a certain moment in time, a ‘photograph’ in Prof. Baxter’s vivid expression, other evidence, in particular of practice since the treaty, may show that for such States the law has not stood still: but the treaty will remain strong evidence, not easy to controvert, that the law is still as the treaty states it, so that the treaty will undoubtedly have a freezing effect on the customary law even for States non-parties to it.²⁹

Even interpretation of codified customary rules might completely detach from the underlying practice, as practitioners get overtaken with a textualist outlook. Such a zeal might even alter the underlying practice. On the customary level interpreting rules happens necessarily in relation to the practice which legitimises the rules and, thus, remains embedded within the greater social/cultural whole. The interpretation of codified/formalised rules, on the other hand, gets overtaken with other things which cannot operate on the customary level.

First, there are the specific words chosen to express the rule, then there is the section in which the rule appears within the overall treaty (or code), as well as the meaning that flows out of a systemic reading of a specific rule with the overall whole. Second, there is the matter of the ‘author(s)’, their imagined intentions, and an aura of being ‘faithful’ to that original plan. This could not be truer for IL, where textualist dogmatism is the norm,³⁰ to the expense of developing competing methods. As Peat and Windsor observe:

Interpretation in international law has traditionally been understood as a process of assigning meaning to texts with the objective of establishing rights and obligations. … As new insights on the practice and process of interpretation have proliferated in other fields, international law and international lawyers have continued to grant an imprimatur to rule-based formalism.³¹ … [Thus,] the focus on rule-based approaches to interpretation, exemplified by the VCLT, means that international law lags behind other fields in which interpretive issues are examined in a more nuanced and theoretically informed fashion.³²

Even where the object of interpretation is the codified rules of a customary practice, the sociological roots the content of the rules have with that customary practice (which is itself organically connected to a greater social whole) often get overlooked. This is of course the result of a fierce debate about the formal meaning of the text. As Bianchi wisely notes: ‘[T]he shackles of both formalism and radical critical approaches have scleroticised the debate by focusing on opposite, yet equally sterile, stances that refuse to take duly into account the more sociological aspects of interpretive processes.’³³Maine’s history cautions that while this formalist-culture arose ‘normally’ in the development of all societies, in most cases its benefits did not outweigh the problems it brought with it. Rather, this formalism imposed a ‘staticness’ and promoted law’s detachment from the greater social whole, which for most societies proved fatal. Only those societies which managed to find ways to continuously alter and evolve their formalised law survived. Thus, Maine focused on those agencies which the ‘progressive’ societies used to develop their formalised customary law.

Before we say a word about these agencies, this evolutionary outlook challenges us to ponder about law’s formalist turn(s). Rather than following the total-formalism domestic legal systems exhibit, the international legal order could follow a more moderate alternative where formalist elements are strategically employed to supplement, rather than replace, custom and CIL. By total-formalism I mean a state where (i) codification stops customary law from growing/evolving, (ii) custom is prevented from yielding new legal norms and (iii) from taking away legality from outdated ones.

This can be prevented by following what I shall call a ‘strategic-formalisation’, which involves a number of interrelated tenets. Despite their deficiencies, custom and CIL are more robust and concrete in producing order and the norms they produce do not suffer from any legitimacy concerns.³⁴ Further, uncodified CIL can be most advantageous in certain instances³⁵ and, thus, codification of CIL should be done selectively – primarily in the areas where practices have ‘matured’ enough and exist in a rather stable environment. Further, these codifications should not be seen as replacements of CIL but as supplements.

This means that codified customary rules should be interpreted closely with the customary practice that underlies them and should be used as aids to properly interpret the practice as well as the function it performs within the greater social whole. It further means that customary practice should be allowed to fill gaps or uncertainties within treaties by producing new customary rules. It also means that we should keep an eye out for the state of the customary practice, and the direction of its development, even after codification. The developments of custom and CIL remain relevant to law even in situations where it is contrary to established law. As Thirlway suggests:

[W]hen custom praeter legem begins, as a result of social development, so to encroach on the existing law’s domain, as to verge on the contra legem, it can nonetheless be regarded, in the light of social development, as still only praeter legem, and as tacit law-making so as to effect a repeal. … This argument, it is suggested, remains a correct picture of how custom can develop law beyond, and eventually contrary to, a codifying instrument. … It is the increasing number of cases in which the codified law ‘does not fit’, in which it is natural and proper to apply a different rule, which eventually gives the new rule the status of law enabling it to over-ride the codified law, on the general principle that lex posterior derogat priori. … Thus there can be little doubt that law deriving from the provisions of a multilateral codifying treaty can be modified by a subsequent general practice constitutive of international custom.³⁶

As such, CIL should not only be allowed to exist next to formalised versions of itself, but under certain circumstances it might be valuable to allow CIL to exist/develop even contra formal law. Custom can not only invalidate existing law, but it might (a) evolve in a different version of itself or (b) create CIL different from what was formally envisioned/agreed. This perspective gives a conceptual and an ontological primacy to CIL. In this way, Maine’s perspective forces us to consider as a viable alternative, what Thirlway posited as a prediction:

[S]o far from supplanting customary law, and reducing its field of operation to a minimum, the codifying of great tracts of international law will, on account of the practical and political difficulties of amending multilateral treaties, whether codifying or otherwise, give over the development of international law almost entirely into the hands of custom, operating upon and beyond the codifying treaties.³⁷

This should not be taken to mean that IL should be merely CIL, rather, we should seek to requalify the position CIL holds within the greater matrix of elements which compose IL over and above CIL. This is a viable position because, despite its various other developments, IL still remains primarily CIL. Again, Thirlway points to a similar position:

[A] ‘code’ of international law produced in the form of multilateral treaty is, except insofar as it represents or becomes customary law, a code of obligations accepted by the States parties to the treaty, but not law. If the field covered by the treaty were one which was more or less unregulated by custom, and regulated in detail by the treaty, the consequence would be that in that field the only ‘law’ between the parties would be the single rule ‘pacta sunt servanda’.³⁸

In this spirit, the last part of this examination will focus on drafting out some first steps towards reconceptualising the interrelations between certain IL components. The aim will be to place CIL at the forefront and find ways to preserve and supplement its operation rather than replacing it. While formal elements will be strategically employed, this scheme escapes formalism by not aiming to alter (or formalise) the primarily customary nature of the international legal order.

5 The Development of Customary (International) Law

While the formalisation of customary law proved fatal for most societies, Maine focused on those societies that found ways to continuously alter and, thus, ‘update’ their formalised customary law. Maine calls these agencies of change: ‘A general proposition of some value may be advanced with respect to the agencies by which Law is brought into harmony with society. These instrumentalities seem to me to be three in number, Legal Fictions, Equity, and Legislation. Their historical order is that in which I have placed them.’³⁹ While legislation, as the deliberate making of a rule, would seem to be the most obvious agent of change, it was the last to appear. Each agent drastically altered the legal practice, enabling it to adjust better to the changing social environment it was embedded in. As such, each was a sign of progress and brought law closer to systematisation. We should not be surprised to find both legal fictions and bodies of equity within IL.

While IL surely lacks the kind of legislation that can be found domestically, this clearly signifies that, despite its mainly customary nature, IL is more developed than mere CIL. Of course, exactly because these agents were employed in a state of total-formalism, as a way to override outdated formalised customary law, they produced new formalised rules which did not necessarily correlate with underlying customary practices. Nevertheless, their relationship can still be amended.

5.1 Legal Fictions

I employ the expression ‘Legal Fiction’ to signify any assumption which conceals, or affects to conceal, the fact that a rule of law has undergone alteration, its letter remaining unchanged, its operation being modified. … The fact is in both cases that the law has been wholly changed; the fiction is that it remains what it always was. … The rule of law remains sticking in the system, but it is a mere shell. It has been long ago undermined, and a new rule hides itself under its cover. … They satisfy the desire for improvement, which is not quite wanting, at the same time that they do not offend the superstitious disrelish for change which is always present.⁴⁰

In its most basic terms, this process can be termed as (radical) reinterpretation of formal rules, although it could be extended to include evolutive interpretation.⁴¹ In such situations, while the rule formally remains the same, the way it is (re)interpreted reshapes its function. When individuals were confronted with new problems and needs, they threw old rules into previously non-envisioned situations, thus recasting them. Maine points to an interesting example.

As mentioned, ancient society was organised in family-institutions that were connected through common lineage. However, the lineage requirement became too restrictive for an evolving society which was confronted with a new problem: the slow growth of the community. In this scenario the lineage rule was thrown in a new situation and it was recast through a process of legal fiction by reinterpreting it together with another newly arisen concept: adoption. Without the rule ever changing, its operation was drastically altered once common lineage could be established through adoption.

This allowed society to grow out of an outdated formalised customary rule without ever formally changing it. This definitely applies to formalised IL, which evolves implicitly both through the practice of individuals (at least those who act on behalf of states) and international courts/tribunals. The ultimate test of whether a specific instance is one of legal fiction or a mistaken interpretation will be the acceptance of the community as a whole – thus, altering CIL. The need for the legal fiction to lead back to CIL is especially important in the case of international courts/tribunals whose judgments in this instance would be sensu stricto ultra vires. As Baker notes:

The problem which arises however is that while neither the ICTY nor ICTR is tasked with ‘making’ international law, but rather simply applying it, it is inevitable (as legal institutions tasked with the implementation of, at times, ambiguous and general legal rules) that their jurisprudence will, at times, fundamentally reshape the law that they are being asked to apply. … [N]ew law often arises, not from lawmaking bodies, but rather from citations of practice where often general and ambiguous rules and statutes are interpreted and put into action.⁴²

Further, as it concerns courts/tribunals, at least internationally, legal fiction should not be limited to this radical or evolutive, (re)interpretation of already existing formalised rules, but it should be extended to encompass one more instance. International courts help IL grow not only by reinterpreting and, thus, breathing new life into old rules, but by also establishing new CIL and altogether new rules. While courts cannot account for the entirety of CIL, as some theorists would claim, they do account for a portion of it. This is not necessarily a bad thing.

As for all concepts, CIL does have a ‘core of settled meaning’, that is rules/practices that are definitely law, and a ‘penumbra of uncertainty’. Within this penumbra, though, the requirements of CIL identification can be most problematic and perhaps restraining, thus halting the development of law. It is within this penumbra that international courts/tribunals can best operate to create new CIL through legal fictions. While this process allows room for abuse, there is a safeguard: looking back for community acceptance and CIL-practice.

If we ensure that the decision has been accepted and embedded within CIL practice, then it is irrelevant whether the decision was a process of mere interpretation or legal fictions. In instances of legal fiction an ultra vires decision can be legitimised and legalised ex post facto, that is through CIL-practice, after it is issued, thus contributing to the growth of CIL. The opposite might also hold true, a decision that is followed by little or no practice, or even uniform but contrary practice, might fail to make IL for the right reasons.

By reconceptualising this process and ensuring that it leads back to CIL, we shorten the gap that exists between legal fictions and the community that has to live by their results. Insofar as the results of these fictions get accepted, they give rise to new practice, which in turn further develops CIL and IL as such.

5.2 Equity

The next instrumentality by which the adaptation of law to social wants is carried on I call Equity, meaning by that word any body of rules existing by the side of the original civil law, founded on distinct principles and claiming incidentally to supersede the civil law in virtue of a superior sanctity inherent in those principles. [Equity] differs from the Fictions which in each case preceded it, in that the interference with law is open and avowed. … On the other hand, it differs from Legislation, the agent of legal improvement which comes after it, in that its claim to authority is grounded, not on the prerogative of any external person or body, not even on that of the magistrate who enunciates it, but on the special nature of its principles, to which it is alleged that all law ought to conform.⁴³

Recasting the process of this agent: it speaks of bodies of rules which exist next to settled IL but displace such settled law (and even manage to become law solely/primarily) due to a robust opinio juris (even in spite of little state practice), which itself flows out of the normativity generated by its (moral) content. While odd at first sight, once we become accustomed to this idea, we see bodies of ‘Equity’ all around us. Human rights are a suitable example of this.

Human rights began as a body of formalised (moral) rules enshrined in formal treaties. Nevertheless, and despite little state practice in certain areas, human rights began making claims as CIL due to the high opinio juris their content amasses. In recognising the need for CIL to develop to include human rights, certain theorists sought to redefine the CIL formation and identification formula which proves to be too rigid to allow CIL to grow in this instance. However, as Baker notes, this was not without its problems:

At its most extreme, this scholarship argues that international treaties, especially those encompassing human rights obligations, actually generate international legal norms, because such conventions are inevitably not simply the codification of existing legal norms but rather the creation of new ones. … This non-traditional scholarship presents a framework which insists that the signing of a convention or treaty by a wide group of countries is, in and of itself, evidence of the creation of new customary legal norms.⁴⁴

The problem Baker describes is a real one. Such a move would bring us closer to total-formalism by detaching CIL completely from the practices which underlie it. This is opposite to the vision the present chapter is suggesting, and Baker seems to agree:

At their core, these push-backs argue that the reinterpretation of customary international law advocated by the non-traditional scholarship, one which, as has been seen, envisages the transformation of conventional international law into customary international law as a seamless process and minimizes the role of state practice as a key component in customary international law formation, poses a danger to the entire concept of customary international law. The reinterpretation of customary international law advocated by the non- traditional scholarship is, according to those who oppose it, one which seeks to move the sources of customary international law (i.e., state practice and opinio juris) away from their ‘practice-based’ methodological orientation and instead employ methods which are completely normative in nature.⁴⁵

This pitfall can be avoided by recognising, as Maine suggests, that we are confronted with two separate processes. The claims human rights are making within CIL should be neither confused nor confounded with traditional CIL formation and identification.⁴⁶ Bodies of equity, such as human rights, get grounded as new CIL differently than slowly arising practice-based CIL. Unlike traditional CIL, human rights are formally designed, but the moral principles they derive their force from hold such persuasive force, which is capable of overriding the lack of practice.

In this way, Maine’s findings highlight the need to conceptualise a new mode of grounding within IL, that of bodies of equity, which run parallel to CIL even if they ultimately seek to be assimilated within it. Creating a separate conceptualisation for the processes of equity allows us to preserve the integrity of CIL and how it has been traditionally perceived, while discerning various other newly arisen normative systems interrelating with it. This, in turn, reveals new ways of perceiving the international legal community and highlights new possibilities for its development.

Furthermore, the operationalisation of the concept of equity in this context does not stop at human rights but extends beyond it. For example, it could be used to reconceptualise the role and legitimation of international organisations (IOs) within the international legal order.⁴⁷ International organisations should then be seen as entities which take abstract values and principles (whether moral or otherwise) and slowly generate concrete bodies of equity, in the form of rules, out of them. Their legitimacy could then be drawn out of these wider values and principles they stand on, as well as out of the relation these values and principles hold with the greater international legal order and the individuals that comprise it, without replacing the concept of CIL.

From this perspective, the legitimacy of IOs is not static, nor uniform, but rather dynamic and dependent upon the wider set of values and principles each operates under. Further, the function of IOs also becomes clearer. By slowly generating rules out of such abstract values/principles, they play a large part in the evolution of IL. This is the case regardless of whether IOs directly create IL through treaties or whether they create bodies of soft law. Both treaties and soft law, in this context, can establish common accepted frameworks which may synchronise state practice and enable new CIL to arise, thus developing IL. Under this vision, IOs, as well as the bodies of equity they produce, do not replace CIL but rather supplement its function and expand upon its operation without necessarily altering IL’s primarily customary nature.

6 Conclusion

Looking at the bigger picture Maine paints, we can begin to see how it can benefit the philosophy of CIL and draw some overall conclusions. First, CIL is law properly so, and it has been so since the very beginning, as it has been the case domestically as well – no concerns arise about its ontology. Second, CIL has as tools not only the traditional process of CIL formation, but agents of change such as Legal Fictions and Equity, which aid its evolution into a legal system. Thus, rather than being formally ‘identified’ and ‘ascertained’, rules get grounded as CIL in a variety of ways, and there is a whole set of intricate tools already in operation which have been hiding in plain sight. In this way, the development of IL has not been so different from that followed by domestic legal systems.

Third, rather than blindly attempting to mimic the total-formalism of domestic legal systems, Maine’s perspective highlights other possibilities. While formalism was a necessary evil in the development of domestic legal systems, CIL benefits from a different, more stable, environment; and we should not be so fast to shed or tamper with its customary skin. Rather than replacing customs and CIL with treaties, in their meaningful combination and collaboration we can find better ways to serve society. Maine’s legacy can prove to be of invaluable assistance in this task.

Further, Maine’s evolutionary narrative speaks of a myriad of underexplored stages in between mere customary law and the current legislative structure of domestic legal systems. The agents of change themselves resemble some of the traits that manifest in different evolutionary stages. This clearly exemplifies that the international legal order is far from the level of mere customary law. Nevertheless, IL is still retaining a primarily customary nature and this might not be such a bad thing.

Maine’s legacy paves the way for a different utilisation of CIL within the international legal order, and it is up to theorists to follow in Maine’s footsteps and re-conceptualise IL along those lines. Thus, the present chapter suggested some first steps to how such a Mainian reconceptualisation of CIL could look, opening doors for new ideas and debates to take place. At its core, this forgotten legacy gives reasons for international legal theorists to resist the total-formalism so heavily promoted by domestic legal systems and its advocates. Furthermore, Maine’s work highlights a new vision where CIL gets supplemented, rather than replaced. In this way, ‘strategic-formalisation’ can enable us to further develop CIL into a legal system, without necessarily altering IL’s primarily customary nature. In turn, this provides some new arguments/positions to old debates.

The present chapter, thus, hopes to be one of the first in a line of papers which seek to revive, refine and apply this lost evolutionary legacy to the international legal order. Maine, and those who followed in his footsteps, present an untapped pool of information, a forgotten school of thought that has never been applied to IL. And this school of thought, as a new way of thinking about law and society, can provide some much-needed theoretical foundation to a decaying CIL. The author is curious to see where this legacy takes us.

10 Enkapsis and the Development of Customary International Law An Encyclopedic Approach to Inter-legality

Romel Regalado Bagares ^*

The jurists are still searching for their concept of law.¹

To be required to think of law in terms of interconnectedness, among many normativities’ ‘ins’, prevents our search for law from being stuck in one of them or in the other, that is, the aut-aut between ‘being in’ or nothing. Legality again surfaces through the cases as a continuum, underlying discrete and separate bodies of law.²

1 Introduction

According to Jean d’Aspremont’s historiography of the four lives of customary international law (CIL), the 2018 Report of the International Law Commission (ILC) on the identification of CIL has all but solidified the ‘formal acceptance’ of the proposition that practice and opinio juris can be extracted from the very same acts. With the report, ‘practice and opinio juris no longer needs to be subject to two distinct tests’, and ‘practice is no longer restricted to conduct (action or in-action) strictly speaking, but also includes verbal acts and what State officials say, the latter having the potential to be constitutive of both practice and opinio juris’.³ Thus, the ILC unwittingly overturned a ninety-eight-year-old tradition on the roots of CIL in Article 38 of the Statute of the Permanent Court of International Justice and ironically recovered its original unitary form.

But perhaps, because d’Aspremont’s essay is essentially a historical reconstruction, he does not offer an account of the philosophical approaches that support the conclusion that CIL is the convergence of state practice and opinio juris. In this chapter, I will argue that such an account cannot be divorced from the larger question of an integrating theoretical account of the sources of law itself, of which CIL is but a part. In fact, the question of CIL brings front and centre the question of the concept of law.

Here I turn to the work of the Dutch Christian philosopher Herman Dooyeweerd (1894–1977), former chair of jurisprudence and the history of Dutch law at the Vrije Universiteit Amsterdam (VU), for an integrating theoretical account of custom as a source of international law, based on a systematic concept of law, which he termed the Encyclopaedie van de Rechtswetenschap (Encyclopedia of the Science of Law). In more than 200 publications, Dooyeweerd fleshed out his own approach to philosophical thought while wrestling with the reigning neo-Kantian legal philosophies of his day. He inaugurated what has been called a ‘reformational’ philosophical approach, drawing from insights of his predecessor Abraham Kuyper (1837–1920), polymath Dutch statesman, journalist, theologian and founder of the VU Amsterdam. Kuyper had introduced the sociological principle ‘souvereiniteit in eigen kring’ (sovereign in its own orbit)⁴ to guarantee the independence of various spheres of life from unwelcome state encroachment. Dooyeweered transformed it into an ontological or a ‘cosmological’ principle for a systematic theoretical account of a universal modal structure of reality.⁵

His mature system is contained in the three-volume New Critique of Theoretical Thought, first published between 1953 and 1958.⁶ This was a major revision in English of his Dutch-language De Wijsbegeerte der Wetsidee (WdW), which had been published in the 1930s. However, it was in his lesser known Encyclopedia – transcripts published by the Student’s Council of the VU for use in his jurisprudence and history of law classes – where he first elaborated his philosophical approach.

Dooyeweerd’s Encyclopedia offers an engaging framework for understanding ‘the challenge of inter-legality,’ or the question of ‘the ways through which legal domains end up overlapping due to the interconnection of substantive, material objects’.⁷ In inter-legality, we are confronted with a ‘plurality of legalities’⁸ even if embodied in a single specimen of law. Here, ‘the law surfaces as the composite legal nature of the issue under scrutiny’⁹ demonstrating resilient and reflexive ‘material interconnectedness’¹⁰ ‘among functional fields’.¹¹ Moreover, this composite question arises from ‘the overlapping among regimes and orders’,¹² which are also self-referential and coherent in themselves.¹³

In Sections 2–3, I will present key features of Dooyeweerd’s systematic theory of law as embodied in his Encyclopedia, notably, his theory of the modal aspects and his theory of entities, whose correlation are essential for constructing a comprehensive concept of law. In Sections 4–7, I deploy Dooyeweerd’s concept of (legal) ‘enkapsis’, to show that CIL’s various manifestations exhibit the phenomena of inter-legality, and that enkapsis is a promising guide for understanding the phenomena. To that end, I will present an analysis of concrete examples from two interrelated and celebrated international law cases.

2 Dooyeweerd’s Encyclopedic Approach to the Concept of Law

Dooyeweerd asserts that the question of ‘legal sources’ constitutes ‘the key problem for the entire positive science of law as a specific discipline’.¹⁴ Of course many scholars have posed the problem in different ways. Remarking on eight theories of the sources of law in his time, Dooyeweerd concludes that they reveal ‘the almost chaotic confusion regarding the meaning in which the phrase “legal source” is employed’.¹⁵ These approaches either explain law by reference to one or other non-legal factors (for example, as a function of history, or of social practices, or of logical concepts, thus obscuring the boundary between law and other spheres of human life) or take the opposite direction, removing law totally from its inner connections with the other spheres of life by positing a transcendent source without any further scientific elaboration.¹⁶ Each is founded on a particular ‘cosmonomic idea’ or theory of the ordering of reality and its essential elements. Such a ‘philosophic ground idea’¹⁷ shapes in profound ways the concept of law and the theory of the sources of law.

Dooyeweerd’s ‘transcendental critique of theoretical thought’ asserts that every scientific endeavour is founded on pre-scientific and pre-theoretical commitments that are in the final analysis religious in nature, because they point to an ultimate conviction about the nature of things.¹⁸ This is the inner connection between theoretical or scientific knowledge and religious conviction.¹⁹ Various scientific disciplines must thus be seen in the context of the whole of human knowledge,²⁰ whose various areas have ‘an inner coherence and are not simply related to each other in an external and arbitrary fashion’.²¹ This requires philosophical presuppositions that direct a comprehensive account of the ‘mutual relationship and coherence of the jural aspect with the remaining aspects of reality’.²² Pre-scientific knowing is not invalidated but understood as primary. It enables humans to experience events, acts, things and relations as individual, temporal totalities, their different aspects not separately conceived but encountered in their unbroken and mutual coherence with the whole of reality.²³ In contrast, in theoretical thought, different aspects of reality are analysed and distinguished from one another. The theoretical work of any discipline – the theoretical attitude of knowing, involves the Gegenstand relation.²⁴

3 The Three Interrelated Pillars of The Encyclopedia of the Science of Law

Rejecting an account of the state as the sole lawmaker, Dooyeweerd also proffers a pluralist ontology founded on the philosophical principle of sphere sovereignty as a source of diverse structural-material principles for legal or jural positivisation. Here, there are three interrelated pillars anchored on two horizons of human experience, constituted by his interlocking theory of modal aspects and theory of entities. The first pillar is his modal theory of the jural aspect, which is one of the irreducible yet interconnected universal multidimensional modes or aspects of reality. The second pillar is his theory of entities, which gives rise to law unique to their particular practice (entities as rule complexes, each sovereign in its own orbit, exhibiting a ‘differentiated responsibility’²⁵ and ‘distinctive integrity’²⁶ unique to its nature). The third pillar is the various ways in which entities engage in relations of enkapsis or enkaptic interlacement – resulting in a complex intertwinement of the formal and the material sources of law.

3.1 The Jural Aspect among the Modal Aspects

In Dooyeweerd’s mature systematic philosophy, fifteen universal mutually irreducible but mutually coherent ‘modal aspects’ of temporal reality occur in the following particular order: numerical, spatial, movement (or kinematic), physical, biological, psychical, logical, historical, lingual (or symbolical), social, economic, aesthetic, jural, moral and the pistical (also referred to as faith or certitudinal aspect). His theory of the modal aspects ‘is the distinctive and original element of his philosophical systematics, which includes his legal philosophy’.²⁷ His general modal theory of these aspects accounts for basic diversity in reality and the unity and coherence that can be found within such diversity. The modal aspects are not specific things but the ‘different modes of the universal “how” [that] determine the aspects of our theoretical view of reality’.²⁸ All things, entities, events and relations function in all of the irreducible, universal aspects. Everything displays all of the aspects in some way.

The Encyclopedia as a scientific practice examines the nature of the jural dimension that gives to legal phenomena their jural character,²⁹ distinguishing the jural aspect from all other aspects of reality, and accounting for the former’s internal structure as it is interconnected with those of the other aspects.³⁰ The jural aspect gives the concrete human laws their inherent legal normativity, with an original meaning – or ‘meaning kernel’ – for the jural aspect alone (in the same way that the other aspects have a meaning-kernel or nuclear moment proper to each of their spheres, which cannot be defined by any of the other aspects).³¹ The meaning-kernel of the jural aspect, according to Dooyeweerd, is ‘retribution’, which is not to be confused with its criminal law sense. In his sense, retribution is an irreducible mode of balancing and harmonising individual and social interests. It implies ‘a standard of proportionality regulating the legal interpretation of social facts and their factual social consequences in order to maintain the juridical balance by a just reaction, viz. the so-called legal consequences of the fact related to a juridical ground’.³²

This broader idea of retribution ‘involves an appeal to all of the modal aspects of reality that precede it in the order of the aspects’.³³ Such an appeal expresses an indissoluble relationship between and among aspects through a series of analogies, as ‘no single aspect stands by itself: everyone refers within and beyond itself to all the others’.³⁴ Yet each has its own unique, undefinable and intuitive core that qualifies its nature and character and directs its full expression in its interlocking coherence with all the other aspects. Jural normativity – ‘the perspective of law’³⁵ – as distinct from other normativities such as economic, social, historical, aesthetic or ethical, can neither be replaced nor erased because it is one of the different interlocking modes of being in reality. But according to the principle of sphere sovereignty, the jural aspect’s own meaning-kernel should determine how it uses the analogies from all the other aspects for each concrete situation of jural positivisation.

Dooyeweerd’s modal theory provides the building blocks to a full concept of law. Its complex of aspectual analogies or connections pointing backward from the jural aspect – the retrocipations³⁶ – are a substratum, constitutive spheres or aspects, without which the jural aspect and any legal system cannot exist; meanwhile the post-stratum connections pointing forward – the anticipations³⁷ – are regulative in nature,³⁸ as they deepen the constitutive meaning of the jural aspect by opening-up its ethical (moral) and certitudinal (faith) anticipations in the formative historical process of societal ‘disclosure’.³⁹ So-called primitive societies are ‘closed’ legal systems, valid and working within their own contexts but yet unable to move beyond a most basic system of accountability and punishment, in the absence of disclosed and deepened anticipatory aspects.⁴⁰

This process of disclosure – directed by the certitudinal or faith aspect – may happen for better or for worse, as such leading could also take an ‘apostate direction’,⁴¹ in which certain aspects of reality are deified and absolutised.⁴² Moreover, these analogies are expressed in two correlated grids: of the law or norm side, and subject, or factual side, of reality, such that one is meaningless without the other.⁴³ Every concrete fact is subject to this cosmic law-ordering – the modal aspects in which those facts function. Such law-ordering on the norm side correlated with the factual side is only discernible in their positivisation in concrete legal phenomena.⁴⁴ Yet they are not reducible to each other.⁴⁵

Thus the ‘architectonic’ modal structure of the jural aspect embodies the following analogies: (numerical) legal unity and multiplicity; (spatial) legal area of validity of legal norms and the juridical place of legal facts, legal subjects, etc; (kinematic) legal constancy and legal dynamism of norms and facts subject to them; (physical) legal force of legal validity and legal causality; (biotic) legal life and competent legal organs; (psychical) ordering legal will and legal will of subjective parties; (logical) legal identity, legal contradiction, legal attribution and imputation; (historical) legal power or legal competence and legal form-giving by competent legal organs; (lingual or symbolical) legal declaration, legal signification and legal interpretation; (social) legal intercourse and correlation of communal and inter-individual or coordinational relationships; (economic) legal economy and equilibrium; (aesthetic) legal harmony and proportionality; (moral) legal morality; and (faith) legal faith or legal conviction.⁴⁶ Therefore, the problem of ‘law-ascertainment’⁴⁷ in international law will involve the investigation and application of every analogy involved in the modal structure of the jural aspect.

3.2 Theory of Entities, Enkapsis and the Sources of Law

Dooyeweerd’s theory of the modal aspects is interlocked with his theory of entities. The theory of entities deals with how things, events and relationships exhibit typical functions within the modal aspects. Each entity or relationship displays all modal aspects at the same time but there will always be two aspects which will exhibit and define its particular identity. These are the founding and the qualifying functions. The founding function is the aspect qualifying the process of transformation of an entity.⁴⁸ The qualifying function is an entity’s intrinsic purpose. The intrinsic purpose qualifies the thing’s internal structure.⁴⁹ The qualifying function is also the ‘individual leading function’ that plays a role in an entity’s internal unfolding process,⁵⁰ through which the function acquires ‘an internal structural coherence’.⁵¹

The difference between a modal and entitary perspective is that the latter is focused on the qualifying function of things; the former is not.⁵² An entity’s qualifying function shows that its identity cannot be understood through the theory of modal aspects but that nevertheless, its structural unity expresses itself in all modal functions.⁵³ In this way, Dooyeweerd’s theory of entities accounts for entitary distinctiveness: family, marriage, church, mosque, temple, corporation, museum, university, a humanitarian NGO, an international organisation or state, each of them imbued with an original material competence, with their own sphere sovereignty. Family or marriage is founded on the biotic aspect but qualified by the ethical aspect; a church, a mosque and a temple are all founded in the historical aspect and all qualified by the faith aspect; a corporation is founded in the historical aspect but qualified by the economic aspect; a museum, a university or a humanitarian NGO are all founded in the historical aspect but have different qualifying functions: aesthetic for the first, logical for the second and ethical for the third. The UN and the state are founded in the historical aspect but both are qualified by the jural aspect. Each has an intrinsic jural competence as an entity, community or institution. Their legal personality is not dependent on the grant of recognition by the state or any other institution. The radical implications of this differentiated social ontology for international law may be summarised in the following description: ‘[d]ifferent social relationships have different characters, different kinds of law-making requirements, different foundations’.⁵⁴ The structural principles that arise out of differentiated societal entities, spheres, and relations form the building blocks of Dooyeweerd’s theory of the sources of law: all law displaying the typical individuality structure of a particular community of inter-individual or inter-communal relationship, in principle falls within the material-jural sphere of competence of such a societal orbit, and is only formally connected (in its genetic form) with spheres of competence of other societal orbits.⁵⁵ A legal source is every juridical form in which those organs of a communal or coordinational relationship in the mutual correlation of these communal and coordinational functions, who are competent to form law, positivise legal principles into valid law within a given life context.⁵⁶

Societal structural principles rooted in the differentiated creational order ‘lie at the basis of every formation of positive law and it is only these principles that make the latter only possible’.⁵⁷ In addition, his theory of entities also accounts for their interrelationships. Here, what Dooyeweerd calls ‘enkapsis’ comes to the fore. These interlacements are ‘free forms of positivisation’,⁵⁸ owing to their ‘typical historical foundation’⁵⁹ – or to their development located in the unfolding of the differentiation of society. Enkapsis is the ‘complicated manner in which the simple entities are interlaced with each other by the cosmic order of time and through which they are united, in part, within complex structural totalities’.⁶⁰

Enkapsis happens in the mutual intertwinement of differently qualified societal spheres and relationships, which are ‘pheno-typical’⁶¹ forms. By this, he means that in these relationships, the inner natures of the societal spheres are not at all obliterated by their particular interlacements. There are different types of enkapsis that are entitary and structural, such as the correlative and the unilateral. In a correlative enkapsis, two structures presuppose each other, as in the case of interlacements between communal and coordinational relationships. A variation of correlative enkapsis is territorial enkapsis, where all differentiated societal structures are territorially bound to the state in whole or in part.⁶² Moreover, in concrete expressions – of positivised – law or of formal law proper, there is what may be called ‘legal’ enkapsis. In other words, positivised laws found in the various spheres of competence are interlinked with one another in complex ways.⁶³

A Kelsenian purely formal law in which all positive law is material leads to ‘a radical levelling of the material structural differences amongst the various jural norm-systems’.⁶⁴ Rather, a material classification of law should be based on the ‘typical internal character of the various norm complexes of a legal order’.⁶⁵ A formal law must also be distinguished from material competence. Formal law is determined by the different enkaptic relations that happen in the interlacement of different societal structures and relationships.⁶⁶ Material competence refers to the invariant structural principles of the various societal relationships that are sovereign in their own spheres; the latter are the material sources of law.⁶⁷ One and the same genetic form posivitising jural principles (that is, a formal source of law) may involve an original source of law in one sphere of competence but may be a derived source of law in another sphere.⁶⁸ While a formal law is inextricably bound with a law-forming organ, such an organ is interwoven with various material spheres of competence, so that it can never be the sole source of validity of all positive law.⁶⁹ Validity and positivity are inextricably connected; there is no juridical validity without positivity, just as there is no positivity without the juridical validity of norms.⁷⁰ This is his ‘material juridical meaning-theory of validity’:⁷¹ ‘with its entire sphere-sovereignty every positive legal order rests on the changing meaning of those law-spheres lying at the foundation of the jural sphere’.⁷²

Transformations in the constitutive spheres lead to corresponding transformations in the positivisation of structural principles. It remains the task of the leading formers of law to be properly guided by faith – by their ultimate commitments – in a way that does not distort or deny the jural limits or boundaries between and among societal spheres and relations.

4 Dooyeweerd’s Critique of CIL as an Indirect (Formal) Source of Law

The now familiar two-element test of the crystallisation of customary norms comprising the opinio and longaevus usus requirements, according to Dooyeweerd, proceeds from the assumption that it is the ‘sole indirect mode of legal formation’.⁷³ But such an approach ‘can never qualify custom to become a legal source, because, as we know, a legal source [in the formal sense] is unthinkable without a competent formative legal organ’.⁷⁴

4.1 The Antinomy of Opinio and Usus

To begin with, there is an inner antinomy to the accepted formula ‘opinio and usus’, because the former ‘concerns a jural anticipation within the psychic sphere’, which in the first place ‘presupposes the jural’, which it then attempts to define.⁷⁵ As an earlier and distinct aspect, the psychic aspect cannot define what is normative of the jural aspect, for such would deny the jural aspect its inherent normativity and violate the principle of sphere sovereignty, in which no aspect is reducible to any other aspect. That is, the accepted notion of an opinio is an improper understanding of the analogy to the psychic sphere, which precedes and anticipates (points towards) the jural aspect in Dooyeweerd’s suite of fifteen modal aspects. Rather, in the unbreakable law-side and factual-side correlation, the analogy to the psychic sphere on the law side is the basic concept of the function of the legal will expressed as ‘a legally ordering will of an organ competent to form law’.⁷⁶

On the factual side, the analogy to the psychic aspect is expressed in the ‘will-function of a legal subject correlated with the ordering will of a competent organ and capable of accepting jural responsibility and normative accountability for its acts in legal life’.⁷⁷ Only a competent organ has the legal power (an analogy to the human formative historical aspect) to positivise structural principles into law. This is an important requirement, given current proposals for deformalisation in law-ascertainment in international law.⁷⁸ Law-ascertainment cannot be divorced from the question of who has competence as a legal organ to positivise material-legal principles into law. Thus the work of such a competent organ (itself an analogy to the biotic sphere) should not be confused with the scientific description of norms positivised into law by such competent organ, as in the case of the writings of legal scholars or publicists on CIL.⁷⁹ Also, longaevus usus is not a requirement for the indirect formation of law, as a rule may develop into CIL within just a brief period of time.⁸⁰

An example is the quick adoption of cruiser rules in naval warfare to submarine warfare, following the experience of World War I. In regard to this, ‘a positive piece of the law of nations was formed with regard to the competence to take custody and to bring in with submarines commercial ships, analogous to the existing positive law practice applicable to cruisers’.⁸¹ This came about because of ‘a series of legal actions which succeeded each other relatively quickly’.⁸² He might be referring here to Part IV of the Treaty of London of 22 April 1930, which bound submarines to extant rules applicable to surface warships when dealing with merchant ships. According to that treaty provision, surface warships and submarines may not attack a merchant vessel without having first placed passengers, crew and ship’s papers in a place of safety. The only exceptions are when the vessel refuses to stop on being summoned, or when it resists a visit or search.⁸³ Thus, although CIL is an indirect way of forming law, like a treaty, it presupposes the same original competent organ (here, the states in their mutual consent).⁸⁴ It was adopted by the major naval powers of the era and many others from their experience of World War I, when submarines were first widely used. However, of forty-nine nations that were parties to the treaty, each of the major powers abrogated the treaty as soon as war broke out again.⁸⁵ Here, Dooyeweerd also anticipated by at least three decades the ruling of the North Sea Continental Shelf case that widespread and representative adoption of a conventional norm by non-signatory states, even within a short span of time, may transform it into CIL.⁸⁶

Yet, the rise and fall of this legal regime invokes the various analogies in the architectonic modal structure of the jural aspect: in the human formative unfolding process (historical aspect), new rules have to be drafted to apply to a new horrific method of waging war, spurred by technological developments – submarine warfare. The legal life of the treaty, and the CIL on which it was based, proved short-lived (biotic aspect), enjoying a brief legal constancy but eventually losing to the legal dynamism of lawmaking (kinematic aspect) brought about by new historical circumstances, and resulting in the loss of its legal force (physical aspect) in the waters subjected to the legal regime (spatial aspect), as legal subjects – states, who are also legal ordering organs (psychic aspect) – broke faith (analogy to the faith aspect) with it, ceding moral considerations (moral aspect) to the expediency of war.

4.2 A Dooyeweerdian ‘Communitarian Semantics’

Secondly, the existence of CIL is determined according to ‘other formal sources of knowledge’⁸⁷ – in other words, formal knowledge of the practice of states, or of any competent organ for that matter. In principle, formal knowledge of sources would embrace an entire host of indicators, whether written or unwritten, given that CIL is an indirect way of formalising law. It is important that the indicators are traceable to competent organs. This process of law-ascertainment for CIL will then involve the lingual (or symbolical) analogies of legal declaration, legal signification and legal interpretation. The ordering will of competent organs is expressed in a legal declaration – not a social fact. Such declaration may well be a correlation of written and unwritten expression, or by conclusive lawmaking behaviour, rebus ipsis et factis.⁸⁸ The practice (legal declaration) of competent organs already embodies and implies a belief (by an ordering legal will and accepted as such by the will of other legal subjects) that the rule followed in such practice is legally binding. Thus, it would be circuitous to assert that rules crystallise into CIL through the confluence of opinio juris and state practice.⁸⁹ The norms embodied in CIL are understood, amplified, communicated, and applied in legal intercourse among legal subjects (analogy to the social aspect).⁹⁰ Legal norms do not exist by themselves. In the dynamic process of lawmaking, they are norms of legal intercourse and interaction in inter-individual or inter-communal contexts.⁹¹ Yet their meaningfulness becomes a legal question in any legal dispute, in which case their legal signification will require the legal interpretation by ‘law-applying authorities’:⁹² domestic courts and international tribunals.

In interpreting the legal declaration at issue, courts and tribunals must apply rules of logic in a jural way, considering legal identity, legal contradiction, legal attribution and imputation. They also must observe legal economy and legal harmony, with an opened sense of legal morality. Their legal interpretation of the legal declaration in question will help deepen or otherwise clarify the norms at issue. In time, continued legal intercourse may serve to deepen the international legal order’s appreciation and commitment to the norms embodied in the legal declaration, and as interpreted by law-applying authorities, thus resulting in heightened faith or trust in its legitimacy (faith aspect) by legal subjects from various spheres. It is against this backdrop where d’Aspremont’s social thesis on the need for a ‘communitarian semantics’⁹³ is better understood: domestic courts and non-state actors generate social practice to illuminate the meaning of the law-ascertainment criteria of the international legal system, thus participating in the reinforcement of the possibility for the international legal system of producing a vocabulary enabling ascertainment of the rules of which it is composed.⁹⁴

4.3 CIL as a Formal Source of Law and the Promise of Pluralist Ontology

Customary international law merely describes the form of a source of law (formal source). The concept ‘customary law’ itself lacks a juridical delineation that refers to its specific type of positive (jural) material content. What is needed is a granular examination of the material source of norms embodied within the formal source, founded on the modal sphere sovereignty of the jural aspect with its internal connections to the substratum of the preceding modal aspects.⁹⁵ But crucially, the material content of valid customary law, as with the material content of all formal sources of law, is also founded upon the multiplicity of the different entities and their enkaptic jural relationships.

This ‘thick’ account of sources of law amounts to a radical legal pluralism, which implies that the material sources of CIL are not mere invention of states. In their formation of CIL, competent organs ought to recognise the multiple forms of the spheres involved and their respective underlying principles that are each unique. The same rule applies to other formal sources of international law. On the domestic front, there are various normative complexes, the state being only one of them. Each of them has an irreducible material sphere of competence unique to their nature, with their own jural sphere sovereignty, and forming a jural unity within a multiplicity of norms.⁹⁶ Moreover, if pursued to its logical conclusion, the crystallisation of CIL must be appropriate to the structure and nature of the diversity of spheres, entities and relations. At the very least, on the international level, states or other competent organs should respect their differentiated responsibilities and distinctive integrity in the development of CIL. Palombella would recognise this as a desirable characteristic of the plurality of legalities he writes of with keen approval:

The plurality of legalities is ensured if law is not monopolized by one single hand, and if it stems from a variety of those capable of triggering the generation of law. Accordingly, rulers (be they sovereign States or omnipotent global regulators) should better be also respectful to a law that is not their own, if the rule of law, not of men, has to thrive. Such premises, then, suggest that the plurality of sources (and one might add, in a wider ultra-state setting, legalities) qualifies, in principle, as a legal precondition of non-domination through law and of liberty.⁹⁷

In such theory of inter-legality, the law and liberty of spheres appear to be an external, sociological, explanation. Dooyeweerd’s theory of differentiated responsibility with its distinctive integrity, however, leads to a truly radical view of international legal personality, where the normative claims of each sphere are drawn from the givenness of its inner nature. The very idea of the irreducibility and coherence of the modal aspects resists the overreach of one sphere over the others, or the undue dominance of one entity over the others; every sphere or entity is to regard the others’ respective differentiated responsibility and distinctive integrity.

5 Inter-legality: Overlapping of Formal and Material Sources in Legal Enkapsis

Thus, legal enkapsis may take place in the international realm between two different genetic forms, or between different formal sources of law. An example is the already well-known intersection between treaty law and CIL demonstrating a ‘duality of norms’. It can be the case that a treaty may also embody principles long considered as binding – those longstanding principles of CIL.

This has been recognised in Nicaragua v. United States.⁹⁸ Here, the International Court of Justice (ICJ) held that the UN Charter cannot be said to exhaust all principles dealing with the use of force, as the latter is also addressed by CIL, which exists contemporaneously with treaty law – yet the former may also be grafted into the latter by way of codification. A ruling predating the Nicaragua case and stating the same principle has been expressed in the Philippine case of Kuroda v. Jalandoni,⁹⁹ which concerned the trial of a Japanese officer for war crimes during World War II. In this case, lawyers for the Japanese officer argued that, under the principle of legality, he could not be tried for war crimes under the Hague Convention on Rules and Regulations covering Land Warfare because the Philippines was not a party to it. However, the Philippine Supreme Court held that while the Philippines was not a party to the Hague Convention, it was nevertheless bound to it because its provisions are also part of CIL, which in turn became domestic law through the Incorporation Clause of the 1935 Philippine Constitution. This analysis is limited only to the level of formal sources.

One illuminating example of the legal enkaptic interlacement of two different material sources in one genetic form of law may be taken from the field of international humanitarian law (IHL). There is a host of treaties embodying the customary practice of protecting religious beliefs and convictions in situations of armed conflicts,¹⁰⁰ which shows how ecclesiastical practice and religious freedom intersect with the sphere that is for the most part the concern of states. Another example is the customary practice of environmental protection, where there is also a host of instruments, declarations, military manuals and pieces of national legislation providing for certain forms of protection for the environment in situations of armed conflict.¹⁰¹ These examples show the correlation of two distinct interests embodying CIL – the regulation of the means and methods of warfare, with the protection of the environment in times of armed conflict. This illustrates bi-layered legal enkapsis: firstly, the intersection of different formal sources – national legislation, treaty law and CIL – and secondly, the meeting of two different material sources providing material-legal principles; that is, environmental protection and the international public regulation of armed conflict in one instrument: treaty law. The norms in question pertain to IHL but also involve norms taken from other spheres; moreover, these latter derived norms do not in any way invalidate the nature of the positive law as IHL.

6 Community Interests in International Law: South West Africa and Barcelona Traction

South West Africa and Barcelona Traction illuminate the reach of the Encyclopedia of the Science of Law, not only in understanding the formation of CIL in particular but of the development of international legal processes in general. After ruling in 1962 in the preliminary phase of South West Africa¹⁰² that it had jurisdiction to hear the challenges brought by Liberia and Ethiopia against apartheid practiced by South Africa in South West Africa, the ICJ made a surprising turnaround in 1966. There, it dismissed the applications on the ground that the applicants lacked locus standi, having possessed no subjective legal right or interest in the subject matter of their claims against South Africa.¹⁰³ It held that the applicants have no subjective rights under the concept of the ‘sacred trust of civilisation,’ as such ground may only be understood within the particular organs and mechanisms established by the League of Nations, which by then were already defunct.

The ICJ’s analysis of how the notion of ‘sacred trust’ was drafted into the legal structure of the mandates is a useful study in how law relates to morality and faith in an analogical way. While agreeing that all states have an interest in seeing the realisation of sacred trust, the ICJ nevertheless held that such a realisation cannot be merely a ‘moral or humanitarian ideal’;¹⁰⁴ it must assume a particular legal form – a (legal) positivisation – to grant rights and obligations, as for example, the UN trusteeship system or the charter’s provisions on non-self-governing territories, which are expressly provided for in relevant texts.¹⁰⁵ Thus, it ‘must be given form as a juridical regime in the shape of that system’.¹⁰⁶ The ICJ held that in the structure of the mandates there was ‘no residual juridical content [that] could, so far as any particular mandate is concerned, operate per se to give rise to legal rights and obligations outside the [mandate] system as a whole’.¹⁰⁷

The mandates provided two modes of rights entitlement of members. The first was the ‘conduct’ provision in the mandates, pertaining to the mandates as a whole, and the second was the ‘special interests’ provisions, which were granted to the states as individual members and their nationals.¹⁰⁸ The court held that the applicants founded their claims on the conduct provisions, but this could not be maintained, because the sacred trust as a right was positivised in a different way. The form it had taken meant that individual member states could not, on their own – on the basis of subjective rights not granted by the League’s charter – directly intervene in the work of administering the mandatories. This right belonged only to the league’s organs under the conduct provisions.

In Dooyeweerd’s modal theory, the ethical (moral) aspect stands immediately next or anticipatory to or above the jural aspect; immediately succeeding the ethical aspect is the upper limit of the aspects, the faith or certitudinal aspect. These two latter regulative aspects open up, deepen and disclose a fully orbed meaning of the jural aspect as expressed in concrete legal systems through a developed idea of justice. Thus, here we see how the deepened principles of jural morality may come into being: in the case of apartheid, what would have been relevant is the principle of legal personality as guided by the ‘regulative jural principle of the value of the human being (dignitas humana)’,¹⁰⁹ which acquires a significant role in the disclosure or differentiated development of the public spheres of jural freedom of the human personality within the state as well as within the non-political spheres of life.¹¹⁰ The moral aspect expressed as dignitas humana may be incorporated into the jural aspect, not in the original sense but in an analogical manner. Because the jural and the moral constitute distinct aspects that nevertheless cohere with each other, one may not be reduced into the other.¹¹¹ Moreover, this legal use of the idea of sacred trust appeals to notions of the sanctity of faith in the highest ideals of civilisation. In his dissent, Judge Tanaka alludes to its Christian theological roots by saying the principle has been present as far back as Vitoria.¹¹²

The faith aspect in turn also refers back to, and builds on, the earlier moral aspect, which to begin with, evinces the jural-moral duty not to violate the legal faith concept of sacred trust by causing others injury, and the notion of dignitas humana as basis for fundamental equality of races. In fact, there is strong evidence that the mandate system was not the sole juridical expression of such a sacred trust, but is traceable as already well-established in nineteenth-century international law, both as ‘consensus of opinion of all civilized states’¹¹³ and as a treaty norm, in particular, as embodied in the General Act of the Berlin Conference of 26 February 1885.¹¹⁴ The mandates could therefore have also embodied CIL. The South West Africa Cases palpably demonstrate a court yet unprepared to recognise an international society’s deepened and disclosed understanding of the principle of equality among races as to vest in any state the right – equivalent of an actio popularis¹¹⁵ – to hold South Africa accountable for its practice of what would today be a violation of a jus cogens norm.

Four years later in the Barcelona Traction case, it would make an about-face, at least with respect to the idea of states being able to make a claim on behalf not just of their particular interests but also in the name of common interests, in the unlikely case of a private commercial dispute.¹¹⁶ I approach Barcelona Traction from two levels. Firstly, there is the question of the diplomatic protection proper and its implications. The ICJ found customary norms for the international legal dispute from municipal law, one traceable to a specific sphere – that of business, where commercial law has developed rules germane to the life-world of corporations. It thus provided what may well be a precedent in international law on the separate and distinct legal personality of a corporation from that of its shareholders, as well as on the ascertainment of the nationality of a corporation.¹¹⁷ It recognised the domain occupied by the corporation as a domestically governed entity, where municipal rules under which it has been created are relevant to the international dispute.¹¹⁸ Secondly, there is the now famous obiter dictum of the ICJ in Barcelona Traction pertaining to obligations erga omnes, or those ‘owed to the international community as a whole’¹¹⁹ arising from CIL.¹²⁰

7 Legal Enkapsis, Coordinational Interests and the Public-Private Divide in International Law

On the first point, we can speak of enkaptic interlacements between the commercial interest involved in Barcelona Traction and international public interest involving states. Indeed, one specific instance of a relevant public interest is that of a state that may have been injured by the failure of another to afford legal protection to the former’s nationals. This is separate from the interest of the municipal corporation itself.¹²¹ Here, too, we see how public international law intersects with the realm of private law, in this case, the domestic law of corporations, in a way that might in other circumstances be regarded as being in conflict. This analysis also brings to sharper relief a more nuanced, legal enkaptic treatment of the public-private divide obtaining in the correlation of domestic law-international law in a phenomenon of inter-legality.

Moreover, there is also a retrocipatory connection by international law to the historical and economic aspects, on account of the rapid changes in international commerce, which have transformed and widened the original scope of diplomatic protection ‘to municipal institutions, which have transcended frontiers and have begun to exercise considerable influence on international relations’.¹²² On the second point, an expanding complexity of international legal processes has led to the deepening of the international legal order’s understanding of international law – the anticipatory analogy to the moral aspect¹²³ – where certain violations of CIL now implicate the interest of the international legal order as a whole and not just that of directly affected states. Thus, the ICJ in Barcelona Traction recognised the existence of that distinct class of obligations erga omnes, which makes ‘an essential distinction’ ‘between the obligations of a State towards the international community as a whole, and those arising vis-a-vis another State in the field of diplomatic protection’.¹²⁴ The case demonstrates the reality of a plurality of inter-legalities: (1) the derivation via legal enkapsis of CIL from municipal practice on the legal personality and nationality of corporations, itself sourced from private (corporation) law, which was then (2) applied to a (public) international law dispute involving states; and (3) the diverse conflicting interests involved – of the corporation shareholders as against the corporation itself, of states with conflicting claims to the right to invoke diplomatic protection; and of the international legal order itself, which requires predictability, stability and fairness in the resolution of such conflicting claims.

In Palombella’s terms, here we encounter a single specimen of law, CIL, bringing the composite legal nature of a dispute to the surface. The dispute implicates functional fields of law that embody material interconnectedness and at the same time concern overlapping legal regimes and legal orders: (private) commercial law, national law, (public) international law, domestic orders, transnational and international legal orders.¹²⁵ Finally, these regimes and orders are self-referential, reflexive, resilient and coherent in themselves.¹²⁶ As a retrocipatory analogy to the numerical aspect, these variegated multiplicity of legal interests are woven into a jural unity as coordinated by the enkaptic relations of states in international law as coordinational law. Indeed, Barcelona Traction also underlines the need for jural harmonisation of interests implicated in the coordinational relationship of states (the retrocipatory analogy to the aesthetic aspect) in a proportionate manner (retrocipatory analogy to the economic aspect) that any ruling in a legal dispute must consider.

Jural harmonisation eschews any excessive pursuit of a legal interest over against the others and finds an optimal balance between and among them according to a retributive measure of proportionality.¹²⁷ The coordinational relationship between and among states is correlatively enkaptic in nature, as opposed to one involving a part-whole relation characterised by relations of hierarchy and subordination, or of a foundational relation, where one is the basis of another. In coordinational relationships, even where the issues involved pertain to fundamental norms of international law, states never lose their identities as states! This is an essential inference made out of the doctrine of societal sphere sovereignty. The failure to properly understand the distinction between coordinational and communal relations all too often gives rise to unmet, if not unreasonable expectations of what international law can do.¹²⁸ It is also at the heart of a proper understanding of the phenomenon of inter-legality as legal enkapsis.

8 Conclusion

Recent codification work of the ILC has devalued the two-factor formula of state practice and opinio juris to identify CIL. This gives rise to the question of how to account philosophically for this development. I suggest that such account is properly a question of the concept of law – one often elided, papered over, or otherwise not recognised as a foundational issue, in the continuing debates about the nature of CIL. I have shown that a fruitful alternative approach to understanding CIL is the Encyclopedia of the Science of Law developed by Herman Dooyeweerd. It distinguishes the jural aspect from all other aspects of reality, accounting for the former’s internal structure as it is interconnected with that of the fourteen other aspects, in ascending (anticipatory and regulative) and descending (retrocipatory and constitutive) analogical relations. Dooyeweerd’s approach examines the nature of the jural dimension through three interrelated pillars.

The first pillar is his modal theory of the jural aspect, which is one of the irreducible yet interconnected universal multidimensional modes or aspects of reality, and through which the second and third pillars – entities and enkapsis – are viewed and understood as legal phenomena. The second pillar is his theory of entities, which give rise to pluralistic legal ontologies unique to their particular practice or sphere (entities as rule complexes, each sovereign in its own orbit and exhibiting a differentiated responsibility and differentiated integrity). The third pillar concerns the various ways in which entities and relations engage in enkapsis or enkaptic interlacements, resulting in complex intertwinements of the formal and the material sources of law with profound implications on the private-public (law) distinction. Meanwhile, in ‘legal enkapsis’ different material sources of law display a mutual interrelationship (enkaptic interlacement) that bind and limit without altogether cancelling one another – a process accounted for by a growing and contemporary movement in legal anthropology as ‘inter-legality’. Enkapsis combines the two pillars of his philosophy – the theory of modal aspects and the theory of entities, into a comprehensive and integrative concept of law, thus providing a better systematic and coherent account of inter-legality.

From an encyclopedic modal analysis, there is an inner antinomy to the well-known two-factor formula: the received interpretation of an opinio is an analogy to the psychic sphere, which precedes and anticipates (points towards) the jural aspect. The psychic aspect cannot define what is normative of the jural aspect, for such would deny the jural aspect its inherent normativity and violate the principle of sphere sovereignty, in which no aspect is reducible to any other aspect. Rather, opinio should be interpreted as the ordering will of a competent organ correlated with the legal will of legal subjects accepting legal responsibility arising from material-legal principles positivised into law by the competent organ. Moreover, this must be seen against the background of the entire constellation of the architectonic modal structure of the jural aspect, in which various analogies come into play in the proper understanding of the making of CIL. Further, I have shown the relevance of legal enkapsis through an examination of CIL as a formal source of law and how it actually weaves together material legal principles in concreto. Thus, CIL may be an original source of law in one sphere of competence but may be a derived source of law in another. In other words, its material bases may lie in the particular enkaptic interlacements involved, displaying a multilayered inter-legality of structural-material legal principles. As such, CIL becomes an embodiment of differentiated material legal principles derived from other fields and spheres of law.

Finally, I have demonstrated this phenomenon of differentiated inter-legalities in CIL through concrete examples: in the duality of treaty and CIL norms, IHL, international economic law and state responsibility. Dooyeweerd’s theory of the sources of law is a prescient and comprehensive approach to inter-legality in both domestic and international law, in which we are able to account: ‘for the very fact that several times either the norm to be applied in a specific case, and controlling it, derives from a different context and from a different regime of legality, or the norm to be applied results from a concurring/competing legality’.¹²⁹