17 - Practical Reasoning and Interpretation of Customary International Law

Summary

When we interpret customary international law (CIL), what is it that we actually interpret? The most promising option is that interpretation of CIL is an interpretation of legal practices. However, the dominant two-element doctrine of CIL assumes that opinio juris generates both legality and normativity of practices, when the latter have no independent normative significance. The chapter challenges this view by drawing on jurisprudential ideas that define practices as inherently normative. This view allows to differentiate between two instances of interpretation of CIL. One is focused on clarification of the normative content of state practices, another is focusing on identification of these practices as legal ones. Interpretation as clarification involves assessing structures of practical reasoning. Interpretation of CIL implies establishing connections between first- and second-order reasons that form the practice and give it meaning. This entails that interpretation of CIL focuses on dynamics of reasons, their inclusion and exclusion within an existing normative framework. Interpretation as identification involves showing that existing second-order reasons meet a threshold of legal validity, which may take the shape of opinio juris, but may as well be functional, when legality of a practice is linked to its relevance in a broader set of legal practices

1 Introduction

Interpretability of customary (international) law belongs to the class of jurisprudential problems that entangle and intertwine almost all thorny theoretical and practical issues. It is especially visible against the background of debates around whether customary international law (CIL) can be interpreted, and if so, how this differs from its identification; are there or should there be some rules for CIL interpretation, and what would be the difference between such rules and those guiding interpretation of treaties?

This chapter aims at addressing some of these issues. It seeks to suggest a meaningful way of seeing the process of CIL interpretation through the perspective of practical reasoning. By doing so, it purports to disentangle one of the theoretical knots of CIL interpretation: what is the difference between the identification and interpretation of rules of CIL, considering that both processes concentrate mostly on state practices?¹ For the purposes of this chapter, by ‘state practices’ I mean a slightly different concept than the one being typically used in international legal scholarship. I defend the view that any practice is normative by definition, otherwise it is not a practice at all. This goes against the commonly accepted view that ‘mere’ state practices are but collections of actions and fail to constitute a (legal) norm. I use the concept in the plural because the growth of normativity within practical engagements of states is typically dispersed in terms of the subject-matter. Different practical engagements converge into different normative practices, rather than constitute one continuous state practice.

Section 2 addresses the issue of duality of CIL within the doctrine of the container/content distinction, which is of fundamental importance to the theory of sources of international law. Section 3 suggests a view on (state) practices as being inherently normative, which implies the differentiation between tests for normativity and legality when patterns of behaviour are concerned. Section 4 provides a more detailed analysis of customary normativity. The concluding Section 5 highlights the difference in interpretation of state practices depending on their container/content perception and will therefore attempt a differentiation between the interpretation for the purpose of identification and interpretation for the purpose of clarification/application of a rule of CIL.

2 What Is This Thing We Interpret When We Say That We Interpret CIL?

It is at the core of most contemporary doctrines of legal interpretation that interpretation of something is interpretation of something. In order to interpret a thing, this thing must already be there, and so its existence, meaning, and function are in principle independent from an act of interpretation. This primary intuition allows to differentiate between interpretation and creation or invention.² But it also assumes that locating a thing and interpreting it are two distinct enterprises; identifying a rule of CIL and clarifying its meaning are supposedly not the same.³ In this regard, legal interpretation is tightly linked to the doctrine of sources of law; interpretation of law presupposes that one knows where to find it and how to identify it amongst other forms of social normativity.

The doctrine of sources is a groundwork of legal positivism. That a legal order rests on certain sources entails that a specific class of utterances or actions qualify as generating or communicating the law, as long as they match the criteria of validity that emerge from within this legal order. Thus in domestic law we often say that, for instance, statutes or precedents are sources of law in a sense that certain activities of certain bodies (parliament, courts, etc.) within a certain procedure create legal obligations for all or some groups of persons. In international law, it is generally agreed that treaties, customary law, and general principles of law perform that very same function; they create, impose, or generate legal obligations for states.

The qualification of some social facts as matching criteria of validity does not depend on the content of a purported rule or source. As famously framed by HLA Hart, having criteria of validity for sources of law (‘rule of recognition’) entails that ‘members [of social systems] not merely come to accept separate rules piecemeal, but are committed to the acceptance in advance of general classes of rules, marked out by general criteria of validity’.⁴ This commitment to accept in advance certain classes of rules presupposes that sources of law are merely containers, and their content does not typically play a role in qualifying a source of law as such.⁵ Hence the fundamental postulate of legal positivism is that identifying something as law is separated from assessing its merits.⁶

The container/content duality is of paramount importance for legal interpretation.⁷ One may only engage in legal interpretation if one knows that the normative content one wants to clarify, elucidate, or in any other way meaningfully operationalise, is contained in a valid source of law. In the case of statutory interpretation, a statute is a container of legal rules one wants to interpret. In the case of treaty law,⁸ it is a treaty that is the container, and its provisions form its content. But how about CIL? What is this thing that contains customary rules? This question has no obvious answer, though it is maintained, by the International Law Commission (ILC) for example, that in case of CIL the content/container differentiation still applies.⁹ What, then, is the container one is looking for in order to enquire into the content of a CIL rule?

Apparently, interpretation of CIL is not an interpretation of some texts since it is widely agreed that CIL is an unwritten source of international law. In other words, CIL is not contained in any texts. Certainly, it may have some textual loci in treaties, judgments, statements by state organs, to mention some. Although true, this does not infringe on the fact that linguistic formulas, or certain articulations of customary rules, are not customary rules themselves. They may serve as points of reference, as useful short-hand devices used to communicate and more efficiently engage in the practice that sustains a customary rule, but it would be a mistake to say that a statement of a customary rule by an authority (institutional or academic) is the customary rule itself. In other words, linguistic formulations are but evidences of existence of customary rules, not rules as such. This is true for any type of customary rules, not only legal ones. The same way as judgments merely reflect, articulate, frame customary legal rules that are already somewhere there and exist independently of the fact that a court engages them, manuals of English grammar are also but snapshots of the customary rules of language. Neither of these two can be appropriately used as a criterion for maintaining the practices, and it is actually the other way around: we often discard certain articulations of customary rules as outdated or inaccurate on the basis that this is not how we do it (anymore). Therefore, it is a practice itself which is the ultimate criterion of a customary rule, not its certain pronouncement.

Also, it is difficult to see how rules of CIL can be contained in intentions or positions of states (regardless of whether we treat these as instances of state practice or of opinio juris). That is, interpretation of CIL is not an interpretation of intention or will of a purported author. Unlike treaties, or statutes in domestic law, customary rules cannot be said to have determinate authors. It is a distinct feature of customs that they are matter of what we do, not of what one particular member of a community might intend to do on their own.¹⁰ As put by Gerald Postema, ‘custom is never reducible to what each participant does or to what each says, or thinks, or believes about what each does’.¹¹ Thus, even though it may be the case for some customs that they got intentionally sparked by one action of one particular actor,¹² that actor would not, nevertheless, qualify as its ‘author’. If their action ever rises to a customary rule, this means that it is our rule, not theirs. This, once again, is a feature of customary rules generally, not only legal ones, since what separates them from rules being established externally is that customary rules are rules of a community, not rules for it. They do not get created by someone for the community, rather, they form within the community and define it as such.¹³ Identification of an author of a rule only makes sense when a rule was intentionally designed to bind only particular actors (like in the case of agreements, be it a contract in domestic law or a treaty in international law), or when a rule gets imposed by a lawmaker, since in this situation it is necessary to be able to differentiate between a ‘genuine’ and a ‘fake’ lawmaker. Neither of the two situations are proper descriptions of the context of customary law creation or appearance. Thus, even though it is at times common, in international law specifically, to design a customary rule consciously, this does not suggest that the interpretation of such a rule, when it comes to its application, would be an interpretation of some intentions, or that these intentions would be the container of rules of CIL.

It appears that interpretation of CIL is first and foremost interpretation of state practice. The same way as we interpret other customary rules, say, rules of language, or rules of etiquette, when we interpret CIL, we enquire into what, how, in which circumstances, and so on, participants of a certain practice do and do not do. In the case of CIL, a state practice is the ultimate point of reference one has when clarifying a particular legal rule. I will further define what I mean by state practices in the next section. For now, it suffices to stress that unlike in the case of statutes or precedents in domestic law, or treaties in international law, state practices are not only the containers but also the content of rules one wants to interpret. From the perspective of the doctrine of sources of law, customary rules often appear uneasy to deal with, for they are not only a source of law, they are law themselves.¹⁴ That state practices are both content and containers, however, engenders consequences for what the interpretation of customary rules actually entails.

The content/container dualism of state practices makes them similar to light, as it were, that is, they manifest differently depending on how they are looked at. Light behaves as a wave in one set of conditions of observation, and as particles in another, and as such is, in fact, both.¹⁵ This can also be said about state practices, for when they are interpreted for the purposes of identification of rules of CIL they appear as containers, as something legal obligations are scooped from (see Section 5.1); but when they are interpreted for the purpose of clarifying the meaning of rules of CIL, practices appear as their content, as what the rules are content-wise (see Section 5.2). This dualism of state practices creates a confusion as to how these two instances or cases of interpretation differ. If identification and interpretation are, according to the doctrine of sources, different enterprises, how does one tell the difference between the two if both concentrate on state practice?

Before answering this question, it is necessary to take a closer look at state practices as such, since clarifying their nature is of paramount importance for the further enquiry.

3 State Practices and Normative Deeds

Though it is typically asserted that the concept of opinio juris is far more contested than the concept of state practice,¹⁶ the latter also carries many controversies with it. This is partly so due to its container/content duality, but also due to some conceptual assumptions regarding state practices that are deeply rooted in the doctrines of formation and identification of CIL and are constantly replicated in international legal scholarship.

It is a widespread belief, reflected, among other, in the ILC reports and conclusions, and emerging from the famous North Sea Continental Shelf judgment, that a general practice that is accepted as law is to be distinguished from mere usage or habit.¹⁷ To put it in the words of the ILC, ‘practice without acceptance as law …, even if widespread and consistent, can be no more than a non-binding usage’.¹⁸ A characteristic feature of approaching state practice within the doctrine of identification of CIL, defended also by the ILC, is an all-or-nothingness. It appears that there are only two options: either a state practice is accompanied by opinio juris and then may, if quantitative and qualitative requirements are met, constitute a rule of CIL, or, if it is not, then there exists no obligation for states to act in a certain manner whatsoever. This view on CIL, which Monica Hakimi labels ‘the rulebook conception’, assumes that without opinio juris state practices are mere usages or habits that have no binding force, and that there exist certain clear and formal criteria (i.e., secondary rules) which allow to establish normativity and legality of these practices.¹⁹ This is also articulated by the International Court of Justice (ICJ) that ‘many international acts, eg, in the field of ceremonial and protocol, which are performed almost invariably, but which are motivated only by considerations of courtesy, convenience or tradition, and not by any sense of legal duty’.²⁰ It is, therefore, out of paramount importance that ‘one must look at what States actually do and seek to determine whether they recognize an obligation or a right to act in that way’.²¹ The position of the ICJ and ILC on this matter clearly opposes legal customary rules and their absence, which is reasonable from the point of view of legal logic. What is disturbing, however, is how state practices are thought of when there is no opinio juris. The wording adopted by both institutions not only suggests absence of any obligations within such practices, but also non-normativity of such practices; a view widely supported in the academic literature.²² Opinio juris appears as a magic wand that not only turns the ‘raw material’ of state practices into a norm, but simultaneously into a legal norm.

What seems to be the underlying principle behind such a treatment of state practices rests on two interrelated ideas. First, it is clear that the identification of CIL serves the purpose of establishing the existence of a legal obligation binding upon states. When interpreting state practices for this purpose, one therefore asks questions of legality, that is, whether there exists a norm that provides for legal obligations states must fulfil. What goes alongside it, however, often remains fully or partly unnoticed; namely, that legality is an attribute of a norm,²³ and therefore inquiring into whether there is a legal norm is asking two questions, not one: (1) is there a norm (the question of normativity); (2) if yes, is this norm a legal one (the question of legality)? Importantly, these questions should be answered in this particular order. The question of normativity, though, bears entirely different considerations and should be approached with a distinct methodology and conceptual framework, than the question of legality.²⁴

The language adopted by the ILC and the ICJ, however, makes it seem as if deciding on the legality of certain practices is fundamentally the same as deciding on their normativity; when a practice does not meet the threshold of legality, it is a habit or a usage that creates no obligation or a right, which is basically tantamount to the absence of a norm altogether. This brings the second assumption into play, namely, that state practices are often taken as certain collections of individual actions of states, collections that may or may not feature some pattern (actions that are ‘performed almost invariably’ – as if their performance is a matter of (in)variability, rather than following certain normative consideration). It is thus claimed that ‘the requirement that the practice be consistent means that where the relevant acts are divergent to the extent that no pattern of behaviour can be discerned, no general practice (and thus no corresponding rule of customary international law) can be said to exist’.²⁵

The focus on (in)variability and patterns of behaviour that is so explicit in the reasoning of the ILC and the ICJ seems to neglect the idea that the existence of an observable pattern of conduct is not a relevant marker of there being a practice. Invariability of some actions, even when absolutely consistent, may or may not be evidence of a practice, because it is not the invariability or consistency of actions that matters, but the meaning these actions have for those engaged in them. It is a well-known example by HLA Hart that for an external observer all more or less consistent regularities of behaviour look the same in terms of people doing certain things in certain circumstances. However, that some people go to a cinema once a week does not mean that there is a normative consideration to that effect, that is, that it is somehow socially expected or required from them to go to a cinema once a week.²⁶ On the other hand, that all people lie from time to time (some people more often than others) does not deny the existence of a normative consideration that one must not lie. Thus, that some people go to a cinema once a week is a regularity of behaviour, but not a practice. The only way to differentiate between people engaging in a practice and people simply acting uniformly is to adopt what HLA Hart calls ‘the internal point of view’; practices, unlike mere regularities or patterns of behaviour, feature a critical reflective attitude towards actions, which entails their inherently normative nature.²⁷ So, let us take a closer look at the concept of practice.

Practices, unlike mere regularities of behaviour (like that some people happen to go to the cinema once a week), are of normative nature. In ordinary life, it can be said that at the moment a person steps into a practice, they are expected to accept certain deeds that infiltrate and govern this practice, give it shape, and make it meaningful for the participants. A simple test to be used to determine whether a regularity of behaviour is a practice is whether one may fail in performing or not performing certain actions. This is typically ascertained either through existing mutual expectations that deeds of practice are and will be followed, or through criticism explicated when these deeds are ignored, this criticism being an aspect of the practice concerned.²⁸ For people who happened to go to a cinema once a week, it is not a failure not to go there this week, but go twice the next one instead; no one’s expectations are failed to be met, and no criticism would follow. At the same time, lying to people does usually constitute a failure to meet certain expectations, even when no criticism follows (not all lies get discovered, after all). This latter point also relates to other actions rendered prohibited in the context of existing practices, for example, tortures. Such actions are sometimes colloquially referred to as ‘practices’, but even there we can only meaningfully speak of them as ‘practices’ if there exist normative considerations that somehow make tortures meaningful for those engaged in them (e.g., various utilitarian ‘ticking-bomb scenarios’). Existence of conflicting normative expectations within the domain of the same practice is not at all uncommon, given how much these expectations may depend on underlying reasons (see the next section). This is why some practices may feature uncertainty as to what constitutes a failure in following it.²⁹

The difference between the two examples given above is that there is nothing to be failed in the context of going or not going to the cinema; there are no deeds flowing through the conduct of going to a cinema with a certain regularity, and therefore there is no practice, regardless of the fact that for an external observer this could be the most consistent pattern of behaviour by these people they can observe. In the latter example with lying or torturing, though, there is a certain standard embedded into behaviour, a standard that constitutes a deed that generates certain expectations that other participants of a practice would follow this deed. What differentiates practices from regularities of behaviour, therefore, is the existence of deeds as certain standards that get learned and adopted by the participants of a practice and generate expectations regarding other participants.³⁰ Practices, in such a way, are inherently normative, because the mere existence of deeds as standards constitutes an independent reason for acting in one way and not in another. As emphasised by Gerald Postema,

[Customs] are not (merely) patterns of behavior; rather they set standards for behavior, standards of correct and incorrect behavior, and thus purport to guide that behavior and provide bases for its assessment. Thus, mere regularities of behavior taken alone – the usus or ‘state practice’ of international law discourse – not only fail to constitute customs of international law, they fail to constitute customs of any sort, including those of ‘comity’, because they fail to constitute norms.³¹

From this perspective, customary rules do not and cannot exist separately or detached from practices that sustain them. Besides, that there is a practice, and not just a regularity of behaviour, means that there is a norm that shapes this practice. In other words, to say that there exists a state practice on a certain matter already entails saying that there is a norm on this matter, and n-I. For this reason, it is not entirely accurate to ascertain that when a certain practice fails to qualify as a rule of CIL, there is no moral or social obligation in general binding upon states that flows from the deeds and mutual expectation of participants of such a practice.

This view on state practice was particularly endorsed by the International Law Association (ILA) in its ‘Statements of Principles Applicable to Formation of General Customary International Law’, where it claims that:

‘a rule of customary international law is one which is created and sustained by the constant and uniform practice of States and other subjects of international law in or impinging upon their international legal relations, in circumstances which give rise to a legitimate expectation of similar conduct in the future’.³²

To recapitulate, there are two fundamental considerations flowing from the view expressed above. First, practices (any practices, not just state practices) are inherently normative, otherwise they are not practices at all. The normativity of practices is determined by the character of deeds framing them and by the function these practices perform, as well as by the meaning they have for participants. According to Gerald Postema, the normativity of practices is ascertainable first and foremost from the perspective of those participating in them:

Those who participate in a custom’s practice undertake commitments (a) to judge certain performances as appropriate or correct and others as mistaken; (b) to act when the occasion arises in accord with these judgments; (c) to challenge conduct that falls short of these judgments; and (d) to recognize appeals to the judgments as vindications of their actions or valid criticisms of them.³³

Second, the content and meaning of customary rules can be (and usually is) determined without necessarily assessing the character and nature of the normative claims they exhibit (moral, legal, etc.). Hence, practices always create obligations and endow those participating in them with rights. This does not mean that these obligations and rights are of legal nature, but it is important to bear in mind that absence of opinio juris does not signify absence of any obligation.

With an image of state practice as inherently normative, we may now take a further step and try to clarify how such practice can be reconstructed for the purposes of interpretation. What does the normativity of practice look like and what are the interpretative beacons one may use in order to clarify its meaning?

4 Practice as Network of Reasons

In the previous section, I endorsed the conception that practices are inherently normative, and that getting involved in a practice means accepting and following certain normative standards that are embodied in it and are inseparable from deeds penetrating and shaping it. This view entails, among other things, that practices are sustained by mutual expectations of participants and by more or less implicit normative standards that one in principle is able to fail to meet. Importantly, such character of practices makes them normative, and this normativity may, under certain circumstances, qualify as legal. This characteristic of practices is by and large generic and applies to state practices as well.

Normativity, according to a dominant view, reflects a special ability of law and other social practices to provide those participating in them with reasons for action.³⁴ In other words, practices, such as state practices, are normative in a sense that for those who participate in them the mere fact that they do so is a reason for acting and reacting to the actions of other participants in a certain way. This reason-giving function of practices, in their normative manifestation (i.e., from the internal point of view), entails that they require meaningful participation, and this meaningfulness comprises of participants’ ability to recognise and react to actions of others in a way that is intelligible for the rest of the participants. This is precisely why, even when states do not explicate their position regarding actions of other states, this may still contribute to formation of a new, or sustaining an existing, practice. Even an absence of reaction may, under certain circumstances, get deciphered by other participants of a practice meaningfully either as endorsement or at least as acquiescence.

For such meaningful participation, states must consider practice not only as a reason, but as a network of reasons. It is almost never the case that a practice can in one way or another be boiled down to one standalone reason that states ought to comply with, for each practice gets its function, meaning, and normative significance in a wider context of related activities.³⁵ In fact, especially when we look at a broader scope of social practices, even the simplest ones (such as the practice of eating with a fork and a knife) are only meaningful when taken in the context of a much wider set of considerations which justify the existence of these practices and shape their content. Because of this, the precise normative boundaries of practices may be difficult to define. But what is more fundamental for the purposes of interpretation and for the purposes of identification of a state practice is that reasons comprising a practice vary in nature, function, and strength.

One of the most popular and influential explanations of normativity, developed by Joseph Raz, suggests that even though norms are reasons for actions, not all reasons are norms.³⁶ A reason for action, according to his latest definition, is ‘a consideration that renders its [i.e., an action’s] choice intelligible, and counts in its favor’.³⁷ Reasons as such do not give rise to obligations, but it is nevertheless a basic moral principle that one ought to act according to an optimal balance of reasons one has, all things considered. This equally applies to states, since it is almost never disputed that they are morally accountable agents (were they not, it would have been impossible to defend even a proposition that international law has any function or basis for existence whatsoever). In international relations, states claim reasons for their actions all the time, and some of them are norms. Michael Akehurst, in his influential article on custom as a source of international law, refers to an example of states using white paper for diplomatic correspondence to advance his argument that habits do not create rules of law.³⁸ And indeed, that states almost unanimously use white paper only shows that they do so for a widely shared reason, a reason which, nevertheless, is not a norm. If not all reasons are norms, how is it possible to mark a class of reasons that are norms?

Joseph Raz’s solution to the problem of norms being linguistically inseparable from the rest of the reasons³⁹ suggests that there must be some other criteria according to which we could differentiate between ‘mere’ reasons and norms. Norms are second-order pre-emptive reasons,⁴⁰ and because of this they play a drastically different role in practical reasoning as compared to ordinary first-order reasons.⁴¹ Norms, just like other second-order reasons, are reasons to act or to refrain from acting on some first-order reasons. For example, states may share a wide set of reasons for not using armed force in international relations, and the norm of international law that prohibits the use of force is a second-order reason for acting on all those reasons. But also, and probably most importantly, the existence of a norm prohibiting the use of force is a reason for not acting on certain other first-order reasons. The mere fact of such a prohibition implies that states may not act for reasons that count in favour of using force against other states. In such a way, norms are second-order reasons in the sense that they reinforce some first-order reasons and exclude some other first-order reasons. What this means is that not only are norms reasons for actions they prescribe, but they are also reasons for disregarding reasons for non-compliance.⁴² For example, diplomatic immunity is a norm precisely because it is both a reason for states to refrain from subjecting diplomats to their jurisdiction, and a reason for disregarding any other reasons for acting otherwise, no matter how weighty these may be, such as in the cases when diplomats cause lethal accidents or interfere in the internal affairs of the receiving state.⁴³ This pre-emptive character is what differentiates norms from other reasons for action.

Practices are networks of both second-order reasons, that is, norms, and first-order reasons. This allows for complex and often multidimensional justificatory strategies for one or other course of behaviour.⁴⁴ Apart from this, however, this reflects a feature of norms not only being embedded into practices, but also being virtually inseparable from them. Norms, as intrinsically interwoven into practices, do not ‘hang in the air’ or exist in some metaphysical space, and their justification, therefore, is shaped by, and depends on, a wider network of reasons employed within a certain practice. Norms may be justified in a number of ways; as time- and labour-saving devices, as error-eliminating devices, that is, those subjected to such norms use them as shortcuts in practical reasoning so that if a norm gets accepted it is not necessary anymore to figure out each time an optimal balance of reasons to act upon. Some other norms are justified by recourse to an authority, that is, acceptance of a norm comes as a result of acceptance of authorities issuing them. These (and many more) methods of justification of norms may overlap and supplement each other; in fact, most of the norms by which people are bound have more than just one justification.⁴⁵

In such a way, practices, such as state practice, explicate their normativity as tightly intertwined networks of first- and second-order reasons. Seen as such, their interpretation therefore relates to discovering the interconnection between these two classes of reasons, assessing their balance, and unveiling them in justifications employed by states or implied in their actions.

5 Asking the Right Questions: Re-approaching the Content/Container Duality

Thus far this chapter explored the features and intrinsic qualities of (state) practices as the thing being interpreted within the process of legal interpretation. Now it is time to take this a step further and take a look at this interpretation anew. If state practices are networks of first- and second-order reasons within which states form, manifest, and explicate expectations regarding the actions of other states, how does this affect the nature of state practices for the purposes of legal interpretation? How are these networks of reasons interpreted when looked at as containers, and when looked at as contents?

The theory of normativity as a special case of practical reasoning offers an illuminating perspective on interpretation of state practice as network of reasons. Most importantly, it allows to clearly differentiate two instances of interpretation: interpretation of state practice for the purpose of identification of a rule of CIL, and interpretation for the purposes of clarification of its normative content.

5.1 Interpretation as Identification

The formation of a norm is a process of conversion of reasons and expectations, and it may not always be possible to draw straight lines between a stage when states act for a widely shared reason and do not explicate any expectations, a stage practice emerges, and a stage when it has fully developed. Sometimes, articulation of reasons is enough to generate expectations that these reasons would be followed and other reasons would be excluded, which exhibits a practice being formed (e.g., Truman proclamation on continental shelf). At other times, conversion of reasons into norms does not happen because of difficulties associated with balancing them in a most appropriate way, which causes uncertainties and disagreements as to whether a practice has emerged (e.g., uncertainties in regard of the right to unilateral secession). Yet some features of practical reasoning exploited by actors serve as beacons of there to be or not to be a norm and whether it may qualify as a legal one. Thus, when states’ actions are looked at with the purpose of enquiring whether a new rule of CIL has emerged, the network of reasons appears as a purported container, and what states do and how they react to what other states do get assessed within a logic of sources of law. This, first and foremost, affects the questions through which the interpretation of states’ actions is carried out:

1. (1) do states act for the first-order reasons only, and is there therefore only a semblance of a practice (‘pattern’ or ‘regularity of behaviour’)?

2. (2) or do states act for a second-order reason (i.e., norm), and is there therefore a formed practice, where participants exhibit expectations about reasons being followed or excluded, and shape their conduct based on these expectations?

3. (3) if the latter, then is this second-order reason acted upon and articulated as a part of a wider network of legally relevant reasons, that is, does it conform to certain conventional criteria of validity of custom as legal custom?⁴⁶

Questions (1)–(2) inquire into the existence of a second-order reason that states use as a justification for their action. There is a big difference between justification based on first-order reasons and justification based on second-order reasons, that is, norms. Justification based on first-order reasons does not purport any expectations from other actors and such justification may, as a matter of fact, be implicit and not designed as foreseeing, or matching, such expectations. This, however, is a much rarer situation than it may appear. In today’s world, states are much more often acting within practices than they used to, even when it relates to their internal affairs, and therefore justifications, even when implied, are typically met with expectations from other actors. Hence, it is normal that first-order reasons-based justifications are usually addressed to states’ actions in their domestic realm, though even there second-order reasons embedded into state practices play a more and more significant role.⁴⁷

The existence of a norm manifests in a reason that has a pre-emptive function, that is, a reason that counts for not acting on, and not using as a justification some other reasons. Not only does this mean that certain reasons cannot be legitimately acted upon, but it also entails that other states expect these reasons to be excluded and react accordingly when they are not. This, however, does not in and of itself mean that a norm embedded into a practice is a legal norm. There may exist mutual expectations as to what reasons may or may not be acted on, and what kind of second-order reason bridges them, even when these expectations do not have a manifestly legal character. International relations of states are by and large governed by such second-order reasons, which means that state practices (and hence also norms) are virtually omnipresent.⁴⁸

The first two questions, in such a way, are asked in order to determine the reasons-made boundaries of a container of a customary rule. The third question is quite different, though. It aims at establishing whether this container meets the requirements of validity set by a legal order. It is beyond the scope of this chapter to discuss the intricacies of legality of state practices, if only because they do not, in principle, contribute to the process of interpretation of rules of CIL. The dimension of the legality of customary rules, as was suggested in Section 2, is typically of little relevance to the determination of their content. Let us therefore shift to a different mode of interpretation: that aimed at clarification of the content of customary rules.

5.2 Interpretation as Clarification

A more specific, and more legally charged instance of interpretation of state practices is, certainly, interpretation for the purposes of clarification of the normative content of a rule of CIL. This instance of interpretation, however, tends to adopt a view of state practices as the content of rules, rather than their containers. This, though, does not change the nature of state practices as networks of reasons, and therefore the questions through which interpretation proceeds are again addressed to these networks, but these are very different questions:

1. (1) what first-order reasons does a rule of CIL exclude, that is, what reasons states may not legitimately invoke as justification for their actions within a given practice?

2. (2) what first-order reasons does a rule of CIL reinforce, that is, what first-order reasons does this norm account for, how does it balance them, and whether this balance corresponds to expectations of those involved in a practice?

Let us address the first question. Since exclusion is of crucial importance for differentiating norms from other reasons, interpretation of rules of CIL is primarily concerned with what reasons get excluded by a rule that is being interpreted. For example, is it meaningful within existing state practices to ascertain that a cyber-operation as a single ‘hostile’ act employed by one state against another constitutes an armed conflict within the meaning of customary rules of international humanitarian law (IHL)?⁴⁹ To translate this into the language of practical reasoning, may a state justify the reasons for its act of cyber warfare as not being excluded by the norms of IHL? This question can only be answered by looking at how states accommodate a new reason into existing deeds; whether they discursively assess cyber warfare as an instance of an armed conflict, or as something separate, probably creating an independent deed. By its very nature, exclusion is the function of a norm that renders acting on certain reasons as violation of this norm, and therefore when a new reason emerges from within a practice (a practice of modern warfare, in this example), it is a matter of whether states accommodate this new reason within existing normative deeds, or whether they exclude it from practice and prevent it from becoming its deed. This process of accommodation or rejection typically manifests in how states react to a new reason being invoked as a justification for an action within an existing practice.

Where the first question addresses the external boundaries of a practice, that is, the issues of what kind of reasons count as parts of a practice and what kind of reasons are excluded from it, the second question offers a different perspective. It relates to the justification of norms, briefly touched upon in the previous section. Norms, including legal norms, are typically justified as accounting for a certain balance of the first-order reasons that render a practice intelligible. From this perspective, norms always serve a purpose of simplifying or optimising participants’ compliance with these first-order reasons.⁵⁰ Interpretation, therefore, may not only address the issues of exclusion of some reasons, but also the issues of reassessing or even reshaping the balance of reasons that are included in practice. Thus, it is a matter of interpretation to inquire whether a rule of CIL adequately reflects and accounts for underlying reasons that shape a practice and guide a state’s actions.⁵¹ If, for example, the principle of equidistance as a method of delimitation of continental shelves does not properly account for the reasons that comprise the practice of the use of continental shelf, there may exist a need to rebalance these reasons according to a more fundamental principle.⁵² Such a rebalancing, though made within a wider normative framework of equity, does affect the balance of reasons represented by the equidistance rule; some of the reasons it accounts for are now weightier.

These two questions, though different from those discussed in the previous subsection, build on them. It is in the foundations of legal interpretation to enquire into what considerations and in which particular manner legal norms account for, since it is the core function of legal norms to improve or simplify people’s compliance with reasons. And since in the case of customary rules their content and container are one and the same thing, their interpretation ultimately entails clarifying the boundaries of practices. The need for this clarification reflects that it is in the nature of practices to evolve. The normative deeds comprising the inherent normative standard of a practice are typically on the move; not only do they depend on what participants in practices do, but also on how they react to actions. Thus, those engaged in practices constantly accommodate or reject new reasons that may or may not affect the perception of the normative standard, and this is exactly why interpretation of customary rules is essentially an inquiry into the dynamics of practical reasoning implied within a practice.

It is, therefore, not only possible that rules of CIL allow for evolutive interpretation, it is essential, since it follows directly from the way practices operate and develop. It should be noted, however, that evolutive interpretation in the case of treaties is not the same as in the case of CIL.⁵³ For interpretation of treaties, evolutive interpretation generally relates to the phenomenon that when treaty text remains the same, its meaning is altered in the course of time.⁵⁴ It is argued that evolutive interpretation of treaties is justified when there is evidence that the parties intended, from the outset, that their treaty would be capable of evolving over time, that it can remain effective or relevant in the face of changing conditions.⁵⁵ It is, therefore, essential that evolutive interpretation of treaties is based on the provision of Article 31(1) of the VCLT, according to which ‘a treaty shall be interpreted … in the light of its object and purpose’. When a rule of CIL is in question, though, it seems not entirely accurate to speak of its object and purpose, since rules of CIL cannot be always traced back to some shared intentions (as argued in Section 2), their object and purpose are far less clear and determined than in the case of treaties. In the case of treaties, their object and purpose may be an explanandum for the purposes of interpretation, but in the case of rules of CIL, they rather appear as explanans. In other words, an object and purpose of a customary rule may well be the end point of interpretation rather than its starting point. For this reason, evolutive interpretation of the rules of CIL relates more to the function a certain practice performs and to the meaning its practice has in a wider context of states’ activities. Such an evolutive interpretation, then, focuses on re-evaluating the balance of reasons reflected in a norm, adjusting it to the developing patterns of practice itself.⁵⁶ Every instance of interpretation of a rule of CIL is therefore a snapshot of the balance of reasons currently accepted within a practice. However, since practices are dynamic entities, so are the norms which define them and are sustained by them.

To summarise, the interpretation of state practices as normative networks of reasons takes different shape depending on how they are looked at. If a state practice is approached as a container and is thus investigated for the purposes of identification of a rule of CIL, the main strategy of interpretation will consist in assessing whether states act for a second-order reason (a norm, in this context) and whether it meets the threshold of legal validity. When a state practice is addressed as content, the interpretative strategy will primarily entail determination of those reasons a rule of CIL excludes and assessment of whether those reasons it accounts for are properly balanced.

6 Conclusions

It is in the core of the idea of CIL that it manifests in a peculiar duality; it is a source of international law, and at the same time it is international law as such. By blurring the line between container and content, which is essential for the conventional doctrine of sources, CIL challenges the process of its interpretation too. State practices, which appear as both containers and content of rules of CIL, are subject to interpretation from two different positions – when a new rule is identified, and when an existing rule is clarified. This creates confusions as to how to separate these two instances of interpretation.

This chapter endorses the view on state practices as inherently normative networks of reasons. Approached as such, state practice manifests as comprising of deeds and reasons, the latter existing on two different levels. The normativity of a state practice is explained through there being second-order pre-emptive reasons, that is, norms that bridge a variety of first-order reasons, balancing and mutually rendering them meaningful. The interpretation of these norms embedded into state practices entails discovering connections between different groups and levels of reasons. The interpretation for the purpose of identification of a rule of CIL is primarily concerned with a question of whether there is a second-order reason that systematises expectations and critical stances of states, and whether this second-order reason qualifies as a legal one. The interpretation for the purposes of clarification, in turn, focuses on what reasons a rule of CIL excludes, and what reasons it balances, how well this balance reflects the actual weight of the first-order reasons, and how to ensure that newly formed reasons are properly assessed and accommodated within practice, or get excluded from it.

Such an approach to state practice and interpretation of CIL allows one to distinguish different interpretative stages of a lifecycle of state practice, as well as to conceptualise state practice as a normative network of reasons.