2.2.1 The Systemic Nature of Interpretation

From the doctrinal standpoint, the interpretation of the sources of international law usually begins with the rule on treaty interpretation codified in Article 31 of the VCLT.Footnote ²⁰ This provision provides that international treaties shall be interpreted in good faith and in accordance with the ordinary meaning of the terms used in the treaty.Footnote ²¹ At the same time, the object and purpose of the treaty are also relevant, and the preamble and possible annexes ought to be taken into account.Footnote ²² Furthermore, Article 31 provides that any subsequent treaties between the parties, the practice developed later in the interpretation of relevant obligations and any other rules of international law that bind the treaty parties shall be taken into account.Footnote ²³ Moreover, as McLachlan argues, the Article 31(3)(c) formulation ‘refers to rules of international law in general. The words are apt to include all of the sources of international law, including custom, general principles, and, where applicable, other treaties.’Footnote ²⁴

The provisions of Article 31 appear to be logical: it is necessary to start with the underlying text; the words used in this text should not be attributed any unusual meanings; the treaty provisions should be contextualized within the treaty as a whole, while somehow maintaining a balance between the intentions of the treaty parties at the time of the conclusion of the treaty and the subsequent practice. This is where a problem can arise: how to correctly determine the balance between the intention of the treaty parties at the time of the conclusion of the treaty and contemporary practice.

There seems to be no general rule regulating this, and the balance depends on the particular treaty in question. The European Convention on Human Rights (ECHR) is interpreted as a ‘living instrument’ whereby the intentions of the original parties seven decades ago no longer play any role,Footnote ²⁵ while the Charter of the United Nations (UN) is largely still interpreted in accordance with the intentions of the parties at the end of the Second World War.Footnote ²⁶ At the same time, it should be noted that Article 32 VCLT provides that the travaux préparatoires shall also be considered. With respect to some treaties, for example, the UN Charter, this is regularly done decades after their conclusion, while in others, the practice can quickly outgrow the framework of the original text and thus the travaux quite soon become largely irrelevant.

It clearly follows from Article 31 VCLT that no international treaty is an isolated system of rules. When interpreting treaty provisions, it is necessary to take into account the other international obligations of the treaty parties and other relevant rules of international law. In other words, one should not put a treaty or its individual provisions at the top of the hierarchy and subordinate to it either other treaties or other sources of international law, such as customary international law and general principles of law. The rules of interpretation under Article 31 require a systemic integration of the rules of international law, regardless of the source from which such rules may be stemming.Footnote ²⁷

To illustrate, sometimes the question arises of whether international human rights law even applies in armed conflict.Footnote ²⁸ The preferred view nowadays appears to be that human rights law still applies in armed conflict; it only needs to be systematically integrated with the rules of international humanitarian law.Footnote ²⁹ The scope of the relevant rules of human rights law can thus change, mostly becoming narrower, but this does not create a hierarchy or superiority of one treaty regime over another.

It is worth emphasizing that for systemic integration in the sense of Article 31(1)(c) VCLT, treaty interpreters are needed who (still) operate within the system of international law, that is, the interpreters who have not already placed themselves outside the system or see themselves in it with only one foot. An example of such an interpreter would be the Court of Justice of the European Union,Footnote ³⁰ but to a lesser extent also the panels in the framework of the World Trade Organization (WTO).Footnote ³¹ When the interpreter no longer sees itself in the system of international law, this leads to fragmentation and so-called self-contained regimes and hierarchical placement of one treaty regime above the rest of international law. As noted by the Study Group of the International Law Commission (ILC) in the Fragmentation Report, there is a need

However, as Adamantia Rachovitsa cautions:

Indeed, systemic integration is no panacea. It is not a solution that would harmonize international law or even constitutionalize it. Nevertheless, systemic integration can provide for certain conceptual solutions and rudimentary public law tools for the interpretation of private law type of legal obligations. While a number of doctrinal writings have assumed that a system can only function if it is constitutionalized or otherwise hierarchical in nature,Footnote ³⁴ McLachlan explains that ‘the importance of the rules of customary international law and general principles of law … [in the systemic integration process] is not in their overriding character’.Footnote ³⁵ Indeed, systemic integration is not about hierarchy and overriding, it is about reconciling two or more obligations and shaping their scopes by reading one into another.

2.2.2 From Fragmentation to Systemic Coherence

In his seminal work The Concept of Law, H. L. A. Hart argued that international law was a simple set of primary rules without any systemic coherence significant for municipal legal orders.Footnote ³⁶ The Report of the Study Group of the ILC on Fragmentation of International Law came to a similar conclusion: ‘The international legal system has never enjoyed the kind of coherence that may have characterized the legal orders of States.’Footnote ³⁷

The lack of systemic coherence can lead to problems in the practical application of the rules of international law. As the Study Group of the ILC further observed:

International legal scholarship has tried to overcome the lack of systemic coherence by introducing certain concepts and principles from municipal legal systems. The project on international constitutionalism has been particularly prominent in this regard. Its proponents have indeed tried to conceptualize international law as a constitutionalized ‘homogenous, hierarchical meta-system.’Footnote ³⁹ But critics have pointed out that any instances of a hierarchy of norms in international law are merely perceived, while the practical application of legal norms remains non-hierarchical in nature.Footnote ⁴⁰ This line of thought originates in the Hartian description of international law as a non-system, and then seeks to achieve at least some systemic coherence through the development of the so-called conflict rules.Footnote ⁴¹ These are the rules that resolve conflicts between different norms and sub-regimes of the international legal order. This chapter challenges the assumption that no homogenous meta-system is available to ensure a systemic coherence of the international legal order. But such a meta-system does not need to be hierarchical in nature, and the overarching systemic framework does not need to be constitutionalism or constitutional theory.

International law is a legal system derived from states. This is not to say that actors other than states do not exist in this system. But states are the actors that have the full international legal capacity. This also follows from the formal sources of law: it is states that conclude international treaties in the sense of the VCLT;Footnote ⁴² and it is state practice, combined with opinio juris (again derived from states) that forms customary international law.Footnote ⁴³ From a purely international perspective, international law as law between states operates in a horizontal paradigm. However, the inter-state legal obligations are only one level of international legal regulation.

International law-making is reminiscent of the concept of creation of legal obligations in municipal private law. For example, international treaty obligations are conceptually similar to contractual obligations domestically. Indeed, some languages do not even employ separate terms for contracts and treaties, and it is understood that treaties are in their essence contracts between states. Some treaties even use the term high contracting parties. Despite these private law foundations of international law, certain public law concepts have nevertheless been introduced into the international legal order. Yet it remains ever-elusive how these concepts operate. In the Barcelona Traction case, the ICJ pronounced that certain obligations are ‘the concern of all States. In view of the importance of the rights involved, all States can be held to have a legal interest in their protection; they are obligations erga omnes’.Footnote ⁴⁴ Following this logic, when certain norms are breached, it is not a single state but the international community as a whole who is injured. This is well-illustrated also in the position of the Study Group of the ILC on Fragmentation of International Law: ‘If a State is responsible for torturing its own citizens, no single State suffers any direct harm … such action violates some values or interests of all, or … [of] the international community as a whole.’Footnote ⁴⁵

Indeed, as Sarah Thin has explained:

At least theoretically, international legal doctrine has thus established a distinction between horizontal bilateral legal relations between states and vertical legal relations between the state and the international community. This is similar to the methodological divide between private law and public law known in municipal legal systems. Private law governs the horizontal legal relations between different persons of law (contract and torts),Footnote ⁴⁷ while public law regulates vertical relations between the community (the state) and the subjects of law, as well as the relations between the organs of this public community (constitutional law, administrative law, criminal law).Footnote ⁴⁸

In international law, the distinction between private and public international law means something quite different. Public international law has traditionally developed as a system of rules governing obligations between states, which are for the most part reminiscent of contracts and torts domestically. Private international law has developed as a system of rules governing the relationship between the individual and foreign law/foreign states (e.g. marriage concluded by nationals of different states). The divide between public and private international law is therefore conceptualized on the basis of the identity and nature of actors in a municipal legal system. To put it simply, where the rules govern relations between states, they are called public and where they regulate relations between individuals (e.g. application of foreign law), they are called private.

A theoretical divide between private law and public law akin to the one known in municipal legal systems also exists within the normative framework of the international legal order. In other words, the public/private law divide exists even within the set of rules traditionally – and somewhat misleadingly – termed as public international law. Yet this divide has not been thoroughly constructed in international legal theory. This is where the chapter turns next. It will be argued that international law functions as a system precisely because the private law type of obligations stemming from international treaties and custom are interpreted in line with certain systemic elements that provide for a rudimentary public law umbrella in the system of international law.

2.3.1 Vertical Elements in a Horizontal System

In municipal settings, public law and political theorists have been investigating the relationships between (i) institutions of the state and (ii) the state and the individual. This is necessarily different in international law where the actors involved differ from those at domestic level. Arguably, the relationship between the institutions of the international community have been addressed by the studies on fragmentation of international law.Footnote ⁴⁹ These studies have addressed the problem of the relationship between the different regimes of international law built around different institutions (e.g. WTO law, international criminal law, international human rights law, international investment law) and have proposed a robust system of conflict rules to regulate the relationship between these regimes.Footnote ⁵⁰ At the same time, the vertical relationship between the state and the individual has remained either understudied or overtly influenced by municipal theories that are not always readily transposable to the international level.

Inspired by the rise of international human rights law and international criminal law after the Second World War, some scholars have started to place the individual in the centre of the international legal system and have conceptualized a domestic-like vertical relationship between the individual and the international community.Footnote ⁵¹ But international human rights and international criminal law are still derived from the law-making powers of states. The scope of these legal regimes depends on whether or not a particular state is a party to a certain treaty, whether it has filed reservations to that treaty and sometimes also on how states have voted in the UN Security Council. A good example is the Rome Statute of the International Criminal Court (ICC). While the court was set up to deal with criminal responsibility of individuals, its jurisdiction depends on the will of states and is governed by certain treaty-law mechanisms (e.g. ratification of the Rome Statue, non-party acceptance of jurisdiction, UN Security Council referral). In other words, the position of the individual in international law is not regulated by a ‘public law of the international community’ but by a private law system of treaty-created obligations between states.

If a ‘public law of the international community’ exists, the vertical relationship in this system is the one between the international community and the state. On the international plane, it is the state that takes the position of the individual in a municipal public order. International human rights and international criminal law may well play an important role in the political theory governing the vertical relationship between the international community and the state, but this is because the protection of human rights and punishing international crimes belong to the fundamental values of the international community as a whole.

International human rights and international criminal law in this public law model do not operate from the level of an individual but from the overarching level of the community vis-à-vis the state. It needs to be asked whether the notion of ‘the international community as a whole’ can be conceptualized as a legal fiction to represent an international sovereign authority and develops a political theory of international law that defines the vertical relationship between the fictitious international sovereign – the international community as a whole – and the state.

2.3.2 The Public Law Concepts in the International Legal Order

The use of phrases such as ‘the international community as a whole’ may be seen as a silent adoption of the idea of public law within the international legal order. Such phrasings introduce a sense of a community, but does this have any implications for how international law operates? The debate seems to have been squeezed between two somewhat difficult concepts: obligations erga omnes and norms of jus cogens.Footnote ⁵² The norms most commonly associated with these two concepts are the prohibitions of torture, slavery, racial discrimination, aggression, genocide and the right of self-determination (or rather prohibition of colonialism).Footnote ⁵³ Instead of looking for hierarchical – or even constitutional – superiority of these concepts,Footnote ⁵⁴ international legal theory needs to identify their theoretical foundations and ask what these foundations tell us about the idea of public law that is lingering in the background.

In one explanation, all non-bilateral obligations have an erga omnes character.Footnote ⁵⁵ This view is problematic as it ignores the ICJ’s reference to the ‘importance of the rights involved’, contained in the Barcelona Traction dictum. The ‘importance’ is a substantive issue and can only be defined by the special character of the obligations at stake. Maurizio Ragazzi argues that the obligations of this character have two components: ‘the moral content’ and the ‘required degree of support by the international community’.Footnote ⁵⁶ The concept of obligations erga omnes thus reflects the notion of a value-loaded international community interest. However, when establishing importance for the international community as a whole, the ICJ has only given circular references and rather generally invoked norms and principles of international law. In East Timor, the Court accepted the erga omnes character of the right of self-determination by arguing that self-determination was ‘one of the essential principles of contemporary international law’.Footnote ⁵⁷ East Timor thus defines the ‘importance’ identified in Barcelona Traction as something that is ‘essential’. Such circularities are of little help where one needs to explain which norms are the concern of the ‘international community as a whole’, why that is so, and how is that manifested in the international legal system.

The ICJ has also been unable to identify obligations erga omnes, their content and underpinnings, on the basis of the formal sources of law alone. As Maurizio Ragazzi has put it, obligations erga omnes reflect ‘an exceptionless [sic] moral norm (or moral absolute) prohibiting an act which, in moral terms, is intrinsically evil (malum in se)’.Footnote ⁵⁸ According to Ragazzi, obligations erga omnes are binding not only because states agree that they are, but even more importantly, ‘because nobody can claim exceptions from moral absolutes’.Footnote ⁵⁹ The second claim, that ‘moral absolute’ operates as a direct source of international law, remains debatable. It is nevertheless undisputed that the concept of obligations erga omnes has its underpinnings in strong moral values and it is these underpinnings that shape international law-making.

Although the concept of obligations erga omnes is value-loaded, it is not seen as a hierarchically superior international law. The Report of the Study Group of the ILC on Fragmentation of International Law defines the concept along the following lines:

Obligations erga omnes may thus be seen as a legal manifestation of certain shared community values in the international community, or of an international public order. While the concept may encompass ‘moral absolutes’, it does not take any hierarchical precedence over other rules of international law. The legal effects of these obligations remain unclear. It thus also remains unclear how these shared community values (the international public order) are accommodated within the international legal system. It does appear that obligations erga omnes and their community interest are conceptual in nature; they reflect a sense of a system and community values that stretch beyond the pure bilateral or contractual logic of international law.

There is a great deal of criticism that, after introducing obligations erga omnes in Barcelona Traction, the ICJ did not develop this concept further and its normative effects thus remain unclear.Footnote ⁶¹ Although no specific rule may follow from the concept of erga omnes obligations directly, the concept introduces a systemic way of thinking and creates a vertical relationship between the states and the community. This element of public law influences the interpretation of treaties and custom: the obligations they generate are not necessarily only bilateral in nature, and international law also imposes certain systemic, community-oriented boundaries on the creation and interpretative scope of international legal obligations.

A strong sense of an international community is also inherent in the concept of jus cogens. For a long time, the ICJ had been reluctant to employ the term jus cogens and was referring to an erga omnes character as virtually a synonym for jus cogens.Footnote ⁶² It is now somewhat difficult to separate the two concepts. The ICJ has not explored the content of obligations erga omnes beyond the overlap with jus cogens, which is itself a somewhat mysterious concept.Footnote ⁶³

It is in the nature of jus cogens norms that they have an erga omnes effect.Footnote ⁶⁴ Indeed, we are talking about jus cogens norms and erga omnes obligations. Arguably, jus cogens violations trigger erga omnes obligations – for the international community as a whole. Yet not all obligations erga omnes are to be found on the flipside of jus cogens.Footnote ⁶⁵ Which obligations have an erga omnes but not jus cogens character remains unclear. Christian Tams concludes that ‘[e]rga omnes outside jus cogens is likely to remain uncharted territory until States begin to invoke the concept more commonly in formalised proceedings’.Footnote ⁶⁶ Obligations erga omnes may be seen as a legal manifestation of the international community interest. To the extent of their overlap with norms of jus cogens, they may also be seen as an enforcement mechanism of the latter. It is wrong, however, to see obligations erga omnes as hierarchically superior law. What implications does this have for the idea of an international public order?

As explained by Sarah Thin:

Erga omnes obligations and jus cogens norms thus reflect a shift of paradigm from a strictly bilateral private law–like international law to a more systemic and community-oriented model, which adds at least a very basic vertical relationship between states and the international community.

2.3.3 International Community Values and Limitations on Contractual Freedom

In the VCLT, the concept of jus cogens was for the first time unequivocally mentioned in international treaty law.Footnote ⁶⁸ However, even in this instance it was only given a rather narrow power to void treaties. The Convention also remained silent on the content of the concept. Article 53 of the VCLT, inter alia, provides that ‘a peremptory norm of general international law is a norm accepted and recognized by the international community of States as a whole’.Footnote ⁶⁹ The concept of peremptory norms thus also rests on the presumption of the existence of an international community of states with a shared interest. It is notable, however, that the VCLT qualified the notion of the international community with states, while the ICJ a year later, with regard to obligations erga omnes, only referred to the international community. The Court known for its restrictive interpretations of international law certainly did not omit ‘states’ accidentally.

In the Article 53 definition, a peremptory norm is subject to acceptance – by the international community of states as a whole – of the normative content as well as its peremptory character. Does this mean that jus cogens is an inherently treaty law concept? The concept indeed predates the 1969 VCLT and was invoked by writers even in the pre–Second World War era.Footnote ⁷⁰ At the time, it was unclear whether or not it was a concept generally operating in international law. This has now been generally accepted.Footnote ⁷¹ Even the VCLT reference to ‘general international law’ suggests that jus cogens is a concept grounded outside of treaty law but is capable of having systemic effects on the conclusion and interpretation of treaties.

Any norm of customary international law requires its acceptance by states through state practice and opinio juris. However, the acceptance of the special peremptory character, not only normative content, by ‘the international community of states as a whole’ points to the strong ethical underpinning of these norms.Footnote ⁷² Sandesh Sivakumaran argues that jus cogens represents a minimum threshold of the international value system.Footnote ⁷³ The strong community-oriented ethical underpinning of jus cogens norms has implications for law-making. International law is, in principle, a consensus-based legal system. States create treaty obligations for themselves and at free will. It is state practice and opinio juris that leads to the emergence of customary norms of international law, from which states again have an escape route through the concept of a persistent objector. In principle, it is only new states that become automatically bound by pre-existing customary law and even automatically accede to certain treaties previously governing their territory (e.g. human rights treaties).Footnote ⁷⁴ Yet the peremptory status of certain norms, encompassing the minimum threshold of the international value system, also overrides some fundamental tenets of a consensus-based international law-making.

The ethical underpinning of the peremptory norms can compensate for deficiencies in universal acceptance of these norms. In this context, three examples appear to be particularly instructive: (1) the right to the freedom from torture is supported by very strong opinio juris, yet state practice is rather weak. There is nevertheless little doubt that the freedom from torture has a jus cogens status.Footnote ⁷⁵ (2) Apartheid South Africa claimed that it was a persistent objector to the prohibition of racial discrimination. This claim was universally rejected on the basis that unlike ordinary customary law, peremptory law does not allow for the persistent objector’s status.Footnote ⁷⁶ (3) France used to claim that it had never consented to the concept of jus cogens as such.Footnote ⁷⁷ This argument was rejected and now even France has accepted the binding nature of these norms in terms of both content and character.Footnote ⁷⁸

These examples reveal that with regard to peremptory norms, international law-making may work differently. As Robert McCorquodale argues, ‘some human rights create legal obligations on a state irrespective of whether it has ratified a particular treaty, either because the human right is part of customary international law and so binding on all states or by virtue of a rule of jus cogens, which no state can derogate from or evade by contrary practice’.Footnote ⁷⁹ Moreover, in Furundzija, the International Criminal Tribunal for the former Yugoslavia (ICTY) reasoned: ‘Because of the importance of the values [which the prohibition of torture] protects, this principle has evolved into a peremptory norm of jus cogens, that is a norm that enjoys a higher rank in the international hierarchy than treaty law and even “ordinary” customary rules.’Footnote ⁸⁰

Despite this judicial pronouncement, hierarchical superiority of jus cogens remains a contested issue, particularly so after the ICJ’s Germany v Italy decision on jurisdictional immunities.Footnote ⁸¹ Yet jus cogens can have interpretative effects on the system of international law even in the absence of its hierarchical superiority. Notably, neither Article 38(1) ICJ Statute nor Article 31 VCLT create a pre-determined hierarchy of sources and/or norms. The concept of jus cogens has its value as an interpretative systemic element that reflects community values and overrides state centrism and certain classical tenets of international law-making. The concept changes the traditional paradigm of international law as voluntary law and introduces a set of norms that can be legally binding on states even in the absence of their consent. It thus manifests a strong sense of the international community interest that prevails over the will of individual states. However, they have not been given any effects of hierarchical superiority in international judicial practice.

When jus cogens norms are breached, the concept does not create remedies that would trump other international legal obligations.Footnote ⁸² What jus cogens does is that it prevents a certain obligation to develop either by law-making or subsequent interpretation. For example, Article 53 VCLT would void a treaty by virtue of which two or more states create an obligation to torture, for example, a terrorism suspect. Moreover, jus cogens prevents that the positive obligation under the right to life could be interpreted as an obligation to torture in order to extract information in the so-called ticking bomb scenario.Footnote ⁸³ Such a conclusion would also be in line with Conclusion 20 of the ILC Report on Jus cogens which reads: ‘Where it appears that there may be a conflict between a peremptory norm of general international law (jus cogens) and another rule of international law, the latter is, as far as possible, to be interpreted and applied so as to be consistent with the former.’Footnote ⁸⁴

International legal doctrine has thus introduced certain concepts that imply the presence of a ‘hierarchical meta system’,Footnote ⁸⁵ but it appears that Dino Krtitsiotis is right when he concludes that international law does not know how to accommodate these concepts. He argues: ‘Our “international community” is “deep” enough to have conceived of the idea of jus cogens but not deep enough to know what to do with it. It is caught in the perennial mire of something called erga omnes (or obligations owed to the “community” as a whole), and continues to inch toward so-called crimes and offences against the “international order”.’Footnote ⁸⁶ The problem is that all these concepts envisage a strong sense of community and its values. While such a community sense may well limit the treaty-making capacity of states and influence the interpretation of treaty and customary obligations, these public law elements in international law remain rather limited. It appears that the value-loaded concepts of jus cogens and erga omnes serve a similar purpose as limitations on the freedom of contract in municipal law. In other words, it may well be the case that international law remains a rather shallow system that lacks complexity; but systemic elements are there and shape the interpretation and application of legal obligations stemming from treaties and custom.

2.3.4 The Systemic Elements

General principles of law are a formal source of international law under Article 38(1)(c) of the ICJ Statute.Footnote ⁸⁷ Unlike obligations erga omnes and jus cogens norms, general principles thus have a clear textual basis for their legal authority as a formal source of international law. Their normative value nevertheless appears to be somewhat undefined. Writers have particularly questioned whether general principles directly generate any international legal rules and obligations.Footnote ⁸⁸ International treaty and customary international law can create an obligation in the sense that state A shall not do X, and state B shall do Y. It would be difficult to derive such obligations from a general principle of law only and exclusively. As argued by Gleider Hernandez, ‘general principles normally enter judicial reasoning without formal reference or label’.Footnote ⁸⁹ In other words, general principles of international law are often in the background of reasoning when other sources of international law are interpreted but do not generate direct legal obligations.

In that sense, Hersch Lauterpacht wrote that the concept of general principles of law is ‘in essential harmony with the attitude … of every legal community’.Footnote ⁹⁰ Notably, Lauterpacht thus saw general principles are an element that underpins the sense of a legal community in international law. The reference to a legal community is very similar to what we understand nowadays under the term systemic integration. International law lacks coherence and characteristics of a legal system if it does not adopt certain internal rules that ensure at least a minimum threshold of its coherence.Footnote ⁹¹

Craig Eggett notes in this regard:

It has been argued in this chapter that obligations erga omnes and jus cogens norms provide for a very rudimentary systemic public law element in the private law structure of public international law. Both concepts are built on the understanding of there being an international community underpinned by a minimum threshold of public community values. The role of general principles of law is even more complex in this regard. They bring elements of both public and private law into the international legal system.

International judicial bodies had to import to international law a great deal of procedural principles in order to be able to exercise their judicial function.Footnote ⁹³ The underlying treaties that create such judicial bodies do not and cannot anticipate every possible procedural issue that may arise in the future, and reliance on the general principles of international procedural law is thus instrumental. International criminal tribunals need to resort to general principles of criminal law in their work.Footnote ⁹⁴ With respect of international treaties, certain private law analogies (e.g. general principles of contract law) maybe used in the process of interpretation and application.Footnote ⁹⁵

Discussing the rule of law principle, Thin argues that systemic integration of this principle:

In other words, the application of general principles of law enables international courts and tribunals to act as systemic judicial bodies of the international community rather than mere dispute settlement mechanisms akin to private law arbitration.

Eggett argues that ‘general principles of law play a central role in the operation of the international legal system, so much so that they constitute systemic elements of international law’.Footnote ⁹⁷ Indeed, it is well-established in doctrine and practice that individual sources of law cannot be read in isolation and need to be interpreted in line with other applicable sources in a particular situation.Footnote ⁹⁸ The general principles of law are instrumental in the process of fitting the puzzle pieces together. It is sometimes difficult to distinguish the effects of the general principles of law in the sense of Article 38(1)(c) of the ICJ Statute from those under treaty or even customary international law. This is because a number of general principles are in any case also elaborated in treaties or have also become customary (e.g. pacta sunt servanda and lex specialis) under the VCLT and customary law of treaties. But it is not uncommon in international law that a certain legal rule or doctrine may follow from more than one source of law. After all, it is quite common that multilateral treaties overlap with custom. Although Article 38(1)(c) of the ICJ Statute is not their only point of entry into the international legal system, the general principles of law have an important systemic function in the interpretation and application of the rules and obligations of international law.

The ILC has made an interesting terminological choice in Draft conclusion 2 of the Draft Report on General Principles of Law, which reads: ‘For a general principle of law to exist, it must be recognized by the community of nations.’Footnote ⁹⁹ The ILC explained that

The ILC’s terminological choice is thus not accidental but derived from certain other international legal instruments. The terminological choice of the ILC seeks to stress the diversity of legal systems represented in the world:

The ILC here clearly subscribed to there being an idea of a community at the international level. The distinction between the ‘community of nations’ and ‘international community of States as a whole’ seeks to stress that general principles of law can be derived from all legal systems of the world, and the uniformity of state practice and/or general acceptance requirement is inapplicable to this source of international law. The concept of general principles of law in the doctrine of sources of international law thus reflects the idea of an international public community and supra-national law-making beyond state-centrism.

3.2.1 Ratione Materiae Hybridity

As detailed in Section 3.3 in relation to the materials of comparativism in international investment law and arbitration, the nature of this field is disputed among its stakeholders who are involved in dispute settlement. Such a controversy exists also in academia. Face-to-face, one finds in particular the proponents of the ‘public paradigm’ against those who are supporting and promoting the ‘private paradigm’. Whatever the significance of this controversy, it is worth noting that ratione materiae hybridity does not originate only from dispute settlement and legal scholarship. It also proves to be the outcome of law-making in the field. The hybridity that derives from that process has both a synchronic dimension and a diachronic one, both contributing to explaining the use of comparative reasoning in the field.

Looking at the matter from a synchronic standpoint first, it is fair to say that ratione materiae hybridity is not specific to our time. It has always been a feature of international investment law and arbitration. At all the stages of its history, the regime and its component parts have proven to display diversity, a diversity due to the various legal orders and disciplines it has been connected to. Focusing on the genesis of the regime, for instance, such a characteristic is reflected in the sources of rights and obligations.Footnote ⁵ Customary international law, notably the minimum standard of treatment,Footnote ⁶ attached the field to the (public) international legal order. On the other hand, the lex mercatoria connected it to a transnational legal order.Footnote ⁷ The debate about the legal order in which contracts concluded between host states and foreign investors (‘state contracts’) are grounded provides another illustration of the ratione materiae hybridity of the regime at that time.Footnote ⁸ This diversity of the sources of rights and obligations also reveals that hybridity in the field is a consequence of the choice that is left therein to the parties to choose the law applicable to the settlement of their disputes. Unavoidably, the reference to domestic law, international law, or the lex mercatoria leads the regime in different directions. This contributes to its ratione materiae hybridity and, more generally, this encourages or justifies the use of different analogies as part of comparative reasoning.

This ratione materiae hybridity has also a diachronic dimension, which has impacted both IIAs’ substantive rules and the features/nature of the settlement of investor-state disputes. Evolving views and approaches among policymakers have led to evolutions of key components of the substantive protection conferred by IIAs to foreign investors. This has notably been the result of the politics of bilateral investment treaty (BIT) models. European modelsFootnote ⁹ that were originally predominant have later shared the ‘policy floor’ with other models from developed and developing countries.Footnote ¹⁰ All this is well illustrated by the evolution of the expropriation provision. In this respect, it is worth flagging the approach adopted on the matter by the US in their 2004 BIT Model, which,Footnote ¹¹ as regards expropriation, is directly influenced by the case-law of the US Supreme Court.Footnote ¹² Since then, this approach has spread across treaty practice.Footnote ¹³ Such an influence of domestic constitutional law evidences, not only how the regime has developed a ratione materiae hybridity over time but also how – again – this encourages or justifies the use of different analogies as part of comparative reasoning.

Investor-state arbitration, and more generally the settlement of investor-state disputes, has also experienced a ratione materiae hybridity throughout its history. While originally modelled on commercial arbitration and inter-state arbitration, the former being often said to be a ‘private’ type of justice, investor-state arbitration has developed ‘public’ features as of the 2000s. The move away from confidentiality and the increase in transparency, initiated by the International Centre for Settlement of Investment Disputes (ICSID) Secretariat,Footnote ¹⁴ illustrates this diachronic dimension of ratione materiae hybridity. The diversification of the methods used to settle investor-state disputes offers another illustration. This diversification materialises, for instance, in the choice made by some treaty parties to favour ‘the return of the State’Footnote ¹⁵ and to opt for inter-state arbitration at the expense of investor-state arbitration.Footnote ¹⁶ Such a choice has notable consequences as regards the debate about the direct or derivative nature of investors rights and the comparisons that can be drawn (or not), for instance, with international human rights law.Footnote ¹⁷ As it appears here again, all this encourages or justifies the use of different analogies as part of comparative reasoning.

3.2.2 Ratione Personae Hybridity

As mentioned above, this form of hybridity pertains to the international investment law and arbitration epistemic community.Footnote ¹⁸ It allows us to understand and assess how the profile of the members of that community contributes to hybridity in the field and to the reliance on comparative reasoning therein.

As evidenced by the programmes of conferences, the websites of law firms, or the composition of tribunals made public, professionals in the field have a very diverse scientific background. As detailed below in Section 3.3 regarding the discussion of the materials of comparative reasoning, those backgrounds are manifold. This diversity pertains to legal traditions, legal orders, and legal disciplines, which can all vary – according to various combinations – from one of those professionals to the other. Looking at the literature and arbitral practice, it appears that those scientific backgrounds at least partially explain the diverse analogies and comparative reasonings conducted in the field.Footnote ¹⁹ Arbitral practice about the most-favoured-nation clause seems to confirm this influence.Footnote ²⁰ At the systemic level, those diverse backgrounds generate diverging epistemic forces, which greatly contribute to the ratione materiae hybridity of the regime.

That being said, one should not conclude that scientific backgrounds govern ipso facto reasonings. There are many reasons for not drawing such a conclusion.

First of all, asserting the opposite basically may lead one to deny the ability of arbitrators to take some distance from their own background and to focus, as they shall do, on the factual and legal specifics of the cases they are called upon to decide.Footnote ²¹ In relation to this, such a radical view tends to negate the benefit of the collegiality that characterises most arbitral tribunals.Footnote ²² Indeed, according to this view, there is simply no room for an open-minded discussion and argumentation within the tribunal. There exists no possibility that the dynamics at play between the arbitrators, in the frame of the factual and legal specifics of cases, lead them to broaden their epistemic horizon.

Secondly, it is also of great import not to oversimplify the epistemic landscape and legal practice in our field. Indeed, diversity has not only a collective dimension but also an individual one. A number of stakeholders in international investment law and arbitration have, themselves, a diverse scientific background, rooted in different legal traditions, orders, or disciplines. This is simply the result of the inter-disciplinary and transnational education and experience acquired by those stakeholders.Footnote ²³

Thirdly, it is worth realising that comparative reasoning is not only a matter of scientific background and knowledge, but that it is also one of interests and capacities. The role that each of the international investment law and arbitration professionals plays in the system impacts (or in any case should impact) on the extent to which she builds her work on her background. Acting as an arbitrator in proceedings is a different exercise than writing an article as a scholar. Also, obviously, acting as a counsel and the interests that one is called upon to defend in that capacity inform if and how one relies on one’s own background. Likewise, it indicates the type of comparative reasoning that one conducts.

All these considerations should be taken into account when considering the ratione personae hybridity that no doubt characterises and impacts on the regime, notably in the context of its legal indeterminacy.

3.2.3 Legal Indeterminacy

Legal indeterminacy has been typical of IIAs’ provisions since the adoption of the first IIA between Germany and Pakistan in 1959.Footnote ²⁴

More precisely, among all forms of indeterminacy,Footnote ²⁵ they have been characterised by the indeterminacy of language as notably theorised by Hart.Footnote ²⁶ He made notably a distinction between ‘easy cases’ and ‘hard cases’, in relation respectively to the ‘core’ and the ‘penumbra’ of legal rules. In easy cases, which involve the core of legal rules, their application does not raise any specific difficulty as their meaning appears to be relatively clear in relation to the facts. On the contrary, there is no such linguistic clarity in hard cases that pertain to the penumbra of legal rules. In those hard cases, several plausible interpretations exist, the issue – discussed in Section 3.4 – being then to select one among them. Of course, legal rules and their application can hardly be divided into two clear-cut categories; yet they constitute no doubt the two extreme sides of a sliding scale along which the degree of linguistic indeterminacy varies.Footnote ²⁷ IIA provisions have traditionally been characterised by a pretty high degree of indeterminacy. Well-known examples are the fair and equitable treatment (FET) and indirect expropriation. What do fairness and equity mean and cover as regards the former?Footnote ²⁸ What are the criteria and circumstances to be taken into account to determine the occurrence of an indirect expropriation?Footnote ²⁹ These are key questions and challenges that academics and, most importantly, counsels and arbitrators have been facing. To deal with this indeterminacy, notably when international law rules of interpretation fail to deliver a clear meaning, they often rely on paradigms, legal systems, or specific rules that they deem relevant, be it for functional or strategic reasons. This is well illustrated by the practice of arbitral tribunals concerning expropriation. They have relied on various legal systems, rules, and principles in the process of applying that provision and appraising the existence of indirect expropriations.Footnote ³⁰ Reference has been made in particular to customary international law in relation to the police power doctrine,Footnote ³¹ or to the European Court of Human Rights case law in relation to the proportionality doctrine.Footnote ³²

Of course, drafting has evolved in the IIAs that have been adopted since the 2000s. Treaty provisions in those agreements are often more detailed than they used to be in older ones. Criteria to be taken into account to identify instances of indirect expropriation are included in some treaties,Footnote ³³ while a distinction between expropriatory measures and regulatory measures is made in many of those newly adopted treaties.Footnote ³⁴ Likewise, recent treaty practice appears to be much more specific as regards the content of the FET, be it by listing the components of that standard,Footnote ³⁵ by attaching it to the customary law minimum standard of treatment,Footnote ³⁶ or by providing information about its content. For instance, with regard to legitimate expectations, which is strongly connected to domestic administrative law,Footnote ³⁷ some treaties specify that expectations can only originate from specific representations or commitments.Footnote ³⁸

This trend in treaty practice will no doubt limit the need to rely on analogies and comparative reasoning in the future. That being said, one should remember that the vast majority of the treaties that are the object of current disputes before tribunals form part of the old generation characterised by legal indeterminacy. More fundamentally, that old generation of IIAs is still massively predominant among the treaties currently in force, the new generation constituting only a minority. As a result, comparativism will very likely remain paramount in the reasoning of arbitrators and tribunals in the years to come. It is then very much needed to address the issues that comparativism raises in international investment law and arbitration. Before doing precisely this, it is first necessary to introduce the materials that arbitrators and tribunals rely on when conducting comparative reasonings.

3.3.1 Systemic Comparative Reasoning

This type of comparative reasoning consists in the comparison of the international investment law and arbitration regime with other legal systems and paradigms. While this comparison most of the time leads arbitrators and scholars to focus on and favour one of them, they sometimes adopt a pluralistic approach as they consider the regime as a hybrid system. As a result, they rely on various legal systems and paradigms that are seen as being each relevant for different aspects of international investment law and arbitration.Footnote ⁴⁰ Whatever the approach used by those stakeholders, the choice of either of those systems and paradigms is dictated by the analogy that they establish between the situation and problems that arise in international investment law and arbitration and those that arise in those other systems and under those paradigms. This is well illustrated by those who make a comparison between, on the one hand, the foreign investor–host State relationship and, on the other hand, the relation between private persons and the executive.Footnote ⁴¹ Sometimes, this is the analogy in terms of legal structure and substance that drives comparative reasoning, as is illustrated by those who draw an analogy between international investment law and arbitration and international human rights law.Footnote ⁴²

Those legal systems and paradigms that are relied upon in international investment law and arbitration are well-known, in particular the public-private law paradigms. They have already been studied in great detail in the literature.Footnote ⁴³ While they are well-known, one should yet be cautious when looking at the systemic and paradigmatic landscape to ensure that one catches all its nuances. This matters as oversimplifying that landscape deprives us from the possibility of getting a sound understanding of systemic comparative reasoning and, most importantly, can lead in practice to making wrong analogies. Two main reasons call for such a cautiousness.

From a methodological point of view, first, the landscape appears to be actually messy. This is mainly due to the fact that there are different ways, in particular in the literature, to map the systemic and paradigmatic landscape. Notably, to develop their typology, some focus on the ‘nature’ of the body of law, that is, public or private, or on, so to speak, the ‘localisation’ of the legal order, that is, domestic, international, transnational. For instance, some authors who rely on or discuss the public-private divide put under the paradigmatic umbrella of the ‘public’ both domestic public law and public international law.Footnote ⁴⁴ On the contrary, others make a distinction between them.Footnote ⁴⁵ Likewise, classifications are based, explicitly or implicitly and alternatively or in combination, on the nature of investment arbitration and on the functions that it performs.Footnote ⁴⁶

The second reason that calls for cautiousness consists in the fact that those systems and paradigms display, themselves, diversity. This is well illustrated by the public international law paradigm. It encompasses analogies with general public international law,Footnote ⁴⁷ international human rights law,Footnote ⁴⁸ or WTO law.Footnote ⁴⁹ This diversity matters as the analogy with each of them can lead to different outcomes in international investment law and arbitration. This is well illustrated by the debate about the derivative or direct nature of investors rights by analogy with the traditional model of diplomatic protection and human rights, respectively,Footnote ⁵⁰ which has an impact on concrete issues such as the characterisation of IIAs breaches as countermeasures.Footnote ⁵¹

Irrespective of the need to approach the matter cautiously, it is worth insisting that systemic comparative reasoning is not a mere academic exercise. The choice of either of those systems and paradigms has substantive practical consequences. A very telling example is provided by the issue of state consent to arbitration. The way it is construed by arbitrators depends largely on the paradigm they rely on. And this greatly contributes to explain the diverse and conflicting views they adopt on specific legal issues, such as the status and nature of pre-conditions to arbitration, or the application of the most-favoured-nation clause to IIAs’ dispute settlement provisions.Footnote ⁵²

3.3.2 Specific Comparative Reasoning

While systemic comparative reasoning pertains to comparisons with other legal systems and paradigms considered as a whole, specific comparative reasoning consists in the comparison with specific principles, rules, judicial decisions, and practices forming part of such other legal systems. As it appears, the latter is closely connected to the systemic understanding that international investment law and arbitration’s stakeholders have of the regime. For instance, those who make an analogy between international investment law and international human rights law will have a strong tendency to turn to the case law of human rights courts.

Specific comparative reasoning is seen here in the context of the interpretation of IIAs’ provisions, which constitutes the focus of the chapter. In that context and as discussed in Section 3.4, the materials of that reasoning can be used for various purposes that need to be distinguished as they raise different issues. At this stage, it suffices to say that they are mainly called upon to play a role in evidencing the existence and/or the content of international law rules. This is true notably when materials, notably past judicial decisions, are relied on directly in the process of determining the content of those IIAs’ provisions. But this holds true also in those situations where materials are relied upon in relation to other sources, which are themselves used in the interpretation of IIAs’ provisions. For instance, domestic rules and principles are used to determine GPL but also as state practice for the identification of customary international law.

Before that, it is worth noting that arbitral tribunals often do not specify or justify how and why they can rely on those materials and sources. Yet, most of the time, it suffices to refer to the gateways that public international law and international arbitration offer to make this possible. Article 31(3)(c) of the 1969 Vienna Convention on the Law of Treaties is one of them as it provides that, when interpreting a treaty, ‘there shall be taken into account … any relevant rules of international law which is applicable in the relations between the parties’.Footnote ⁵³ Referring to judicial decisions as a subsidiary means for determining rules of law – as coined in Article 38 of the International Court of Justice Statute – is also relevant in this respect.Footnote ⁵⁴ Last but not least, the applicable law chosen by the disputing parties – when international law is part of it – or the law applicable pursuant to Article 42(1) of the ICSID Convention are also of interest to justify the reliance on those materials and sources.Footnote ⁵⁵

3.4.1 Legality Perspective

As explained above, the theory of sources, and what it teaches us about the determination of the rules of international law, is highly relevant for the purpose of this chapter as all international law sources play a role in the interpretation of IIAs’ provisions. Instances where rules emanating from those sources have been relied on by tribunals in that interpretative process are manifold. Customary international law has been used, for example, in the course of the interpretation of expropriation provisions in applicable IIAs,Footnote ⁵⁶ or in the context of the interpretation of fair and equitable treatment provisions in relation to the minimum standard of treatment.Footnote ⁵⁷ GPL have been relied on, for instance, in that latter context in relation to legitimate expectations.Footnote ⁵⁸ As for the case-law, it has been used – often implicitly – as a subsidiary source in the context of the interpretation of many provisions, such as the umbrella clauseFootnote ⁵⁹ or the most-favoured-nation clause.Footnote ⁶⁰ Relying on those sources requires to make comparisons between state practices, domestic rules and principles, and past ‘judicial’ decisions, respectively. Let us appraise the practice of tribunals thereon, focusing successively on the comparisons between judicial decisions and between domestic rules/principles.

Among the judicial decisions rendered by international courts and tribunals, past arbitral awards or decisions are by far the main type of ‘judicial’ decisions relied on by tribunals in investor-state arbitration. Depending on the case at hand, tribunals rely on one precedent, on a few of them, or on jurisprudence constante, which are made of series of precedents.Footnote ⁶¹ In all those situations, the issue of similarity arises. More precisely, the issue is twofold. First of all, it is worth flagging that similarity does not amount to identity. This begs the question of the degree of similarity or, to put it differently, the question of the conditions under which similarity is deemed to be sufficient. Secondly, it matters to realise that similarity should characterise two sets of relations. Past awards/decisions should first display similarity between themselves. Also, they should be (sufficiently) similar to the factual and legal specifics of the case at hand, in particular to the IIA provision being interpreted.

A review of arbitral practice shows that tribunals hardly ever enquire into similarity in practice. This holds true as regards the similarity that should characterise the two sets of relation mentioned above. Ipso facto, this also applies to the degree of similarity. Such a trend is obviously an issue in methodological terms. But, more fundamentally, this is an issue inasmuch as this may lead to situations where the interpretation that derives from relying on precedents conflicts with the intent of treaty parties, by failing to respect the textual, the teleological, or the contextual specifics of the treaty at hand.Footnote ⁶²

Turning to the discussion of arbitral practice pertaining to the reliance on domestic rules and principles, attention should first be drawn to selectivity. Tribunals most of the time refer to a very limited number of domestic legal orders and rules thereof in the process of interpreting IIAs’ provisions.Footnote ⁶³ Also, they usually hardly enquire into the similarity between those rules and principles. Is it an issue? The lack of discussion about similarity is no doubt one, as this is a conditio sine qua non to testify of their collective substantive significance. As for selectivity, the answer to the question depends actually on the function assigned to relying on those domestic rules and principles. When the intent is to fill a gap by trying to evidence the existence of a GPL, this seems to be an issue indeed. It is so because the theory of sources teaches us that such general principles should be widespread across domestic legal systems and be characterised by representativeness.Footnote ⁶⁴ On the contrary, when the intent of the tribunal is to confirm an interpretation reached on other grounds, it is less of a concern as those domestic rules and principles relied upon do not aim at evidencing such a general principle.Footnote ⁶⁵ Such a reasoning does not run against the theory of sources, which appears to be irrelevant in that context. From a legality point of view, nothing prevents such a use of domestic rules and principles. On the other hand, this appears to be more problematic seen from the standpoint of legitimacy, to which the chapter now turns to fully grasp the issues raised by comparative reasoning in investor-state arbitration.

3.4.2 Legitimacy Perspective

Legitimacy – understood here as the acceptance of a power exercised by an entity – does matter when appraising the use of comparative reasoning in arbitral practice, as it does more generally as regards the existence and functioning of arbitral tribunals.Footnote ⁶⁶ This results from the power that they exercise and that impacts on disputing parties and beyond, in particular on local populations. This power consists primarily in the settlement of the disputes submitted to tribunals. But it also materialises into the law-making function that they are de facto called upon to perform when interpreting vague IIAs’ provisions.Footnote ⁶⁷ Indeed, their legal indeterminacy, discussed in Section 3.2.3, requires from tribunals to make choices as regards the substance that they give to those provisions in light of the facts of cases. To beef up the acceptability of those choices – and more generally of the power they perform – tribunals have to provide arguments in support thereof. As it appears, this de facto law-making function, and interpretation in the context of which this function is formally exercised, boils down to an argumentative practice that is intended to convince the tribunal’s audiences of the soundness of the choices made.Footnote ⁶⁸ Comparative reasoning being one method used by tribunals to exercise this de facto law-making function, the way it is conducted should contribute to convince the tribunals’ audiences.

This is where, in particular, the selectivity of arbitral tribunals as regards the domestic principles or rules that they use can be seen as an issue. Even though such a selectivity cannot be challenged from a legality standpoint when they are relied upon to confirm an interpretation, a ‘pick and choose’ approach opens a Pandora’s box of criticism from that legitimacy point of view. Critics denounce notably the legal hegemony and the cultural biases across the investment law and arbitration community that this selectivity reflects.Footnote ⁶⁹ All this undermines the acceptability of the choices made by tribunals, the acceptability of awards or decisions, and, more fundamentally, the acceptability of investor-state arbitration.

Beyond the academic debate among comparativists between the proponents of functionalism and culturalism as regards the significance of cultural specificities,Footnote ⁷⁰ the difficulty in investor-state arbitration is obviously that the audience is diverse.Footnote ⁷¹ Distinctions can be made between disputing parties and a broader audience and between states and investors. Another distinction can be made between states depending on whether they belong to the Global South or the Global North. Local populations and civil society are also seen as constituting an additional audience of arbitral tribunals.Footnote ⁷²

This diversity of the audience brings a diversity of expectations. Its component parts often have diverse and sometimes conflicting expectations of arbitral tribunals and, more generally, of international investment law and arbitration. In relation to the discussion conducted in Section 3.3.1 about systemic comparative reasoning, it appears that those expectations are governed by the respective system or paradigm those component parts of the audience rely on. In light of that context, arbitral tribunals seem to have little chance, when conducting their systemic comparative reasoning, to succeed in convincing all of them that they made the right choice in their interpretation of vague IIAs’ provisions.

4.3.1 No-harm Obligation

This section explores the stance of international courts towards the customary rule of no-harm in IEL. The approach followed is a chronological one because in this way it is easier to follow the movement of the interpretative efforts of the jurisprudence. Ironically enough, the first proclamation on non-conventional IEL is to be found in the Trail Smelter case between the US and Canada regarding air pollution where the Arbitral Tribunal famously stated that

The irony lies in the methodology of the Tribunal because it did not make that proclamation after reviewing the previous case law of international courts since such case law did not exist. The Tribunal made the point that it could not even draw an analogy with water pollution since there was no international case law on that issue either.Footnote ⁴⁶ Therefore, it was forced to come up with a different solution since the applicable law on the case was international law, relevant rules of the law of the US, and the Tribunal was also directed to ‘give consideration to the desire of the high contracting parties to reach a solution just to all parties concerned’.Footnote ⁴⁷

The Tribunal came up with the first iteration of the no-harm rule after reviewing the jurisprudence of the US Supreme Court on transboundary intra-state pollution. Ιt used as a legal basis US law, and it declared that the same principles could be part of international law by analogy especially since there was no international rule contradicting them.Footnote ⁴⁸ Moreover, the Tribunal, in the process of establishing the no-harm rule, also made explicit reference to the desire of the parties to reach a just solution. It appears that it took this to mean that it could go beyond existing international law offering an extra justification to its decision to draw an analogy between US and international law. Therefore, the Tribunal propounded the above dictum on the basis of the ‘principles of international law’ as well as the law of the United States. It is not clear how the Tribunal uses the term ‘principles’ in the text. It could be that it wanted to denote that there is a general principle of lawFootnote ⁴⁹ or, more probably, it just used the term ‘principle’ interchangeably with the term ‘rule’. The language in the pertinent paragraph indicates that it probably meant to give the term a normative value. The Tribunal refers to the ‘right’ of States, and it also refers to ‘injury’. It, crucially, also includes the threshold of the application of the principle by stating that the consequences of the harm must be serious, and it also declares a standard of proof. While all this could be just the only wording the Tribunal had at its disposal, it most probably means that it viewed the principle as a legal rule. This becomes even clearer a few lines further down the text where the Tribunal holds that, on the basis of the dictum, it finds Canada responsible in international law.Footnote ⁵⁰

The crucial question is whether the Tribunal simply established the existence of the rule, or it also determined its content. It is obvious that the Tribunal did both, at least to some degree and with a twist. This is not a case where the judges look at the two elements of customary law, state practice and opinio juris and then decide whether a rule has come to existence. The Tribunal decided that a rule existed without looking at the two elements (they were not there) or even mentioning its nature, leaving it to be inferred from the wording of the decision. Then, it went on to shape its contours. Therefore, it did interpret the rule. Based on its reading of the jurisprudence of the US Supreme Court, it did set the threshold of application of the rule (serious consequence) and it did set the evidentiary standard required for the rule to be applied.

This is a striking dictum in that what can be seen as interpretation of the no-harm rule also serves as evidence that the Tribunal indeed meant to establish an international customary rule that does not come from international law as such. Rule identification and content determination, or rather interpretation, are intertwined in the dictum and still there is a way to distinguish them. For both identification and interpretation, though, there is no methodological consistency or clarity in the decision. The identification of custom is unorthodox to say the least, and the interpretation also comes without any strong backing in practice, opinio juris, or other elements that could lead to evolutive or systemic interpretation. The Tribunal looked for an interpretative solution outside the confines of classic customary international law and therefore it worked its way through the analogy with US intrastate law. The award remains a testament to the efforts of the Tribunal to come up with a viable solution to a pressing problem while, at the same time, it brought the no-harm rule to life. And this is no small feat.

The general application of the rule was affirmed, albeit in a context outside IEL, by the ICJ in Corfu Channel.Footnote ⁵¹ It is submitted that it cannot be safely assumed that the ICJ had in mind the Trail Smelter judgment when formulating the dictum in Corfu Channel. In Corfu Channel, the ICJ held that every State has an ‘obligation not to allow knowingly its territory to be used for acts contrary to the rights of other States’.Footnote ⁵² This is a less nuanced repetition of Trail Smelter. There are two points of contact between Trail Smelter and Corfu Channel dicta. First, it was clear in the Trail Smelter dictum that the rule can be breached through omission; and in Corfu Channel, the ICJ applied exactly this aspect of the rule since the harm was indeed caused by omission. Second, and more importantly, the rule in Corfu Channel becomes one of general application and is not confined to cases of injury by fumes. These points show that the ICJ did not move towards interpretation but that it simply generalized the Trail Smelter no-harm iteration. It also, importantly, did not adopt the extra elements of the Trail Smelter dictum, namely the requirement of seriousness and the evidentiary standard. This is not to say that there was any kind of judicial dialogue at play. It is merely a reading of the two cases with the benefit of hindsight. At the same time, the ICJ did include a criterion of knowledge that was not there in Trail Smelter. It is hard to see how this is an interpretative move at least not in the sense of refining the Trail Smelter dictum. The ICJ simply stated clearly what was implicit in Trail Smelter thereby deducing an obligation of notification. In Trail Smelter, the facts of the case did not leave any doubt as to the element of knowledge; therefore, it was only natural that the Tribunal would not linger on the issue. The operation in Trail was known to both parties and what was left to be determined was responsibility and remedies. In Corfu Channel, the issue of knowledge was central to the facts and the ICJ gave more prominence to it.Footnote ⁵³

What seemed like a promising discussion regarding the no-harm rule in the 1940s became a non-issue in the following decades. When the issue resurfaced before the ICJ almost fifty years later in Nuclear Weapons, IEL had already exploded.Footnote ⁵⁴ Both the Stockholm and Rio Declarations were adopted, the ILC had been working on the topic of injurious consequences arising out of acts not prohibited by international law for over a decade and most major environmental conventions were in place.Footnote ⁵⁵

Despite the developments in IEL, there was no actual environmental litigation before the ICJ or any other forum with general international law jurisdiction. It was only in 2010 in Pulp Mills that the ICJ had a chance to revisit the no-harm rule. In Pulp Mills, the ICJ took a surprising course that not only left the content of the rule opaque but also sparked some unneeded confusion. The dispute between Argentina and Uruguay concerned the possible environmental impact of the construction of pulp mills on the river Uruguay, as the river was a shared resource between the two States. The legal basis for the dispute was the Statute of the River Uruguay (CARU) that laid down the rules for the management of the river by the two parties before the Court. At the same time, however, the ICJ recognized that in order to interpret the obligations arising out of the Statute, it had to resort to general international law as well as subsequent agreements between the parties.Footnote ⁵⁶ It was in this vein that the Court went on to discuss customary IEL.

The ICJ stated that ‘that the principle of prevention, as a customary rule, has its origins in the due diligence that is required of a State in its territory’.Footnote ⁵⁷ It then went on to cite Corfu Channel and Nuclear Weapons in support of this statement and it held that ‘[a] State is thus obliged to use all the means at its disposal in order to avoid activities which take place in its territory, or in any area under its jurisdiction, causing significant damage to the environment of another State’.Footnote ⁵⁸ To begin with, the wording of the no-harm obligation as it appears in Pulp Mills is different than the wording used in previous decisions. While the content remains the same, it is unnecessarily confusing. It is not clear why the Court referred to due diligence. Whether it did so because it considered due diligence a general obligation, rule, or principle is not clear. Due diligence speaks to the nature of an obligation denoting that it is an obligation of conduct and not of result.Footnote ⁵⁹ It simply characterizes the obligation in question. In this instance, the no harm obligation is an obligation that requires due diligence on behalf of the State of origin of the activity. Second, the ICJ does not explain the relationship between the principle of prevention and due diligence. It could be that the Court really meant that there is a general obligation of due diligence, irrespective of the existence of a primary obligation like the no-harm rule in the context of international environmental law. It could be that the Court pointed to the dictum in Corfu Channel as establishing such a general principle. The fact that the Court cites Corfu Channel in the preceding sentence could point towards that direction. It is highly unlikely, however, that the Court would proceed to such an important statement without offering any support. Moreover, and just like in Nuclear Weapons, the ICJ did not make any use of the actual development and refinement of the customary obligation so as to interpret it in a new light. It could either decide that the rule has changed through state practice and opinio juris, it could interpret dynamically the rule and bring it up to modern standards, or it could look at the development of IEL as a whole and go for a systemic interpretation. The ICJ did not look at international practice, it did not take into account the work of the ILC, it did not look at relevant treaties or the work of their treaty bodies. It neither seemed to bother too much with the current understanding of environmental harm in science and technology.

Nonetheless, the ICJ did follow the developments in IEL in one significant respect by introducing the threshold of ‘significant harm’. The ILC in its Prevention Articles had already lowered the threshold from ‘serious consequence’ to ‘significant harm’ and it has not been seriously challenged before international courts. The ILC was clear that the threshold is that of ‘significant harm’.Footnote ⁶⁰ The ILC based its decision on the Brundtland Commission’s Article 10, the Commission being a group of experts under the UN that produced the seminal ‘Our Common Future’ report.Footnote ⁶¹ The ILC went on to further define ‘significant’ as meaning ‘something more than detectable but need not be at the level of “serious” or “substantial”’.Footnote ⁶² The ILC makes a clear choice between the two thresholds. The Commentary to Article 2 acknowledges that the threshold of ‘serious’ is used both in Trail Smelter and in Lake LanouxFootnote ⁶³ arbitrations.Footnote ⁶⁴ Both judgments were handed down decades before the ILC adopted the Articles on Prevention; and in between, IEL had developed in an explosive manner covering almost, if not all, areas of human activity and its connection to the environment. The ILC follows the development of the law towards a more progressive path.

The Court accepted the progress signalled by the ILC regarding the threshold of environmental harm, but it proceeded to introduce fresh confusion in the content of the no-harm rule. The ICJ cryptically stated that there is a ‘functional link’ between the substantive obligations of states (that presumably include the no-harm rule) and the procedural obligations of states (that include notification, EIA, the obligation to monitor the relevant activity, to consult and to cooperate, etc.).Footnote ⁶⁵ The Court came to this conclusion interpreting the Statute of the River Uruguay and not a customary law obligation. Nonetheless, it is submitted that it is useful to look at the interpretation of the Statute and the distinction between substantive and procedural obligations therein as significant because the same distinction is reflected in the customary obligation of no-harm. There is the general substantive no-harm rule that prohibits States from using their territory so as to cause harm to another State.Footnote ⁶⁶ And then, there are procedural obligations that provide guidance to States as to the steps they must take to avoid causing such harm.Footnote ⁶⁷

The Court’s dictum on the functional link between these two types of obligations is apparently promising in that it looks like it interprets the no-harm rule in a nuanced manner. However, it probably means, according to the judgment, that if the substantive obligation is not breached, a breach of a procedural obligation (in Pulp Mills, the obligation to notify)Footnote ⁶⁸ is not important from the point of view of reparation. Indeed, the Court declared that the recognition of the breach in the judgment itself constitutes satisfaction. Yet the judgment does not explain in detail what this functional link is – a fact not lost on judges Al Khasawneh and Simma who criticized this lack of any explanation in their joint dissent.Footnote ⁶⁹ The dissent also criticizes the view of the Court on the procedural obligations as unsatisfactory because it renders the procedural obligations essentially unimportant.Footnote ⁷⁰ It is only natural that the ‘functional link’ statement has not been followed up in the case law that followed Pulp Mills. On the contrary, Judge Donoghue directly rejected it in her separate opinion in Costa Rica/Nicaragua stating that even ‘[i]f at a subsequent phase, the failure of a State to exercise due diligence in the implementation of a project causes significant transboundary harm, the primary norm that is breached remains one of due diligence, but the reparations due to the affected State must also address the material damage caused to the affected state’ therefore it is not ‘[u]seful to draw distinctions between “procedural” and “substantive” obligations’.Footnote ⁷¹

Pulp Mills was indeed a rare opportunity both from the standpoint of the substance of the case as well as from the standpoint of the evolution of the methodology of the ICJ in identifying and elaborating upon customary international law. The Court followed a path wherein it simply repeated its previous jurisprudence and declined to explain, at the very least and even as obiter, what the functional link between procedure and substance is. This is even more frustrating if one takes into account the fact that Argentina claimed the breach of both substantive and procedural rules. When the ICJ went on to look at whether the mills had already caused harm to the river Uruguay it did so solely with reference to the Statute of the river as its legal basis.Footnote ⁷² It essentially dismissed all of Argentina’s claims based on lack of proof of harm. The outcome of the decision is not unrelated to the lack of elaboration of the no-harm principle. A better enunciated rule could have been more useful in properly assessing the impact of the mills.Footnote ⁷³

The ICJ did not offer any more significant insights in its very brief mention of the no-harm rule in Costa Rica/Nicaragua where again it simply affirmed its existence.Footnote ⁷⁴ It did not even look at the more elaborate definition of the rule and the insight to some of its content by the ITLOS in its Deep Seabed Advisory Opinion.Footnote ⁷⁵ The ITLOS Special Chamber went on at some length in discussing the no-harm rule and it made a few interesting points along the way. First, it admitted that the content of due diligence is difficult to pin down.Footnote ⁷⁶ Second, it gave an idea of the mechanics of the rule. The Chamber claimed that the ‘obligation becomes stricter or more lenient depending on the activity’.Footnote ⁷⁷ Despite not breaking new ground, the Chamber did interpret the no-harm obligation as one that does not remain constant in its application, but it must be applied differently depending on the circumstances. This is, according to the Chamber, how the nature of the obligation – that is, due diligence – affects the concrete application of the obligation. The ICJ did not engage with the Advisory Opinion and, like it had done in Pulp Mills, ignored the other developments in IEL. The only notable, and important, difference between Pulp Mills and Costa Rica/Nicaragua is that the ICJ dropped, without explanation, its idea of a functional link between procedural and substantive obligations, and that was a positive step.

What emerges from this picture is that the ICJ did not feel the need to expand upon, analyze, or identify the content of one of the cornerstones of IEL. Why the court made that choice is a matter of speculation. Especially when there is a rich customary as well as treaty background available to the ICJ. The outcome therefore can be described as regressive interpretation of the rule. This concept comes from the realm of human rights. The European Court of Human Rights (ECtHR) has since its early days treated the European Convention on Human Rights as a ‘living instrument’, that is, it interprets the Convention by adapting its content to the current political and social conditions.Footnote ⁷⁸ In the following years, the ECtHR has interpreted the Convention following the ‘living instrument’ doctrine, but the interpretation has not been unidirectional towards the expansion of human rights. Research has convincingly shown that the ECtHR has also adopted a regressive interpretative approach overturning previous ‘expansionist’ decisions.Footnote ⁷⁹ Methodologically speaking, there are significant differences between the two fields when it comes to studying the interpretation of the respective rules. The ECtHR has an impressive number of judgments that offer ample samples for qualitative and quantitative research.Footnote ⁸⁰ The cases that involve IEL before the ICJ, or arbitral tribunals for that matter, are few and far between. Moreover, the ECtHR has the benefit of interpreting a Convention instead of customary international law; therefore, it has a more secure footing from the start. The ICJ, on the contrary, often identifies the customary IEL rule and interprets it at the same time.Footnote ⁸¹ The concept of regressive interpretation is used for what it actually is: a concept. It is not intended to serve as an analogy to human rights law or the human rights jurisprudence of the ECtHR. With few decisions and fewer dissenting or separate opinions, the concept is tested, for the most part, against settled doctrine.

With regressive interpretation in mind, the prehistory of IEL interpretation, traced back to Trail Smelter, offers some insights. The Arbitral Tribunal in Trail Smelter took very bold steps: it had to come up with analogies and essentially ‘invent’ a rule of international law that did not exist at the time. While it took time for IEL to develop and grapple with the difficulties of the no-harm rule, eventually the development did come about through the work of the ILC, through conventions that reflect the customary rule, as well as through elaborate doctrinal work. And then the ICJ started moving almost backwards in what can be seen as regressive interpretation. It did not interpret the rule in a way that captured the changes that had occurred in international law. It did not touch upon the threshold of the obligation; it did not offer insights regarding the nature of the rule (the declaration on the due diligence nature aside); and it certainly did not add anything to the rule. On the contrary, the ICJ added unnecessary confusion with the employment of the ‘functional link’ concept and the decoupling of the procedural obligations from the overall obligation to not cause transboundary harm. Overall, it seems that it regressed to something that falls behind the Trail Smelter dictum – let alone further developments in IEL – in terms of boldness and inventiveness.

4.3.2 The Obligation to Conduct an Environmental Impact Assessment

The situation regarding the treatment of the customary obligation to conduct an EIA is even more complex than that surrounding the no-harm rule. This section shows that the case law slowly, but surely, adopted an evolutive interpretation on two levels: first, on the level of the relevant legal framework (national/international) within which it must be conducted and, second, regarding the content of the obligation and the extent to which it can be accurately sketched by international courts.

The obligation to conduct an EIA features prominently in several IEL treaties, and it has become a rule of customary law.Footnote ⁸² ΕΙΑ is the platform upon which states usually concentrate their disputes before international courts and tribunals.Footnote ⁸³ Almost every environmental dispute that finds its way before a court has at least one of its aspects dedicated to EIA. What is more, the ITLOS has specifically instructed parties to disputes to cooperate in order to assess the risk of the activityFootnote ⁸⁴ or to appoint experts to study the environmental impact of the activity.Footnote ⁸⁵

The beginning of the courts’ involvement with the obligation to conduct an EIA was not promising. In Gabčikovo/Nagymaros, the ICJ had an opportunity to discuss EIA in detail. Despite the fact that the Court shied away from this discussion, it provided some insight as to how international environmental rules should be interpreted. The ICJ suggested that obligations geared towards monitoring transboundary risk are amenable to evolutive interpretation and that they cannot be treated as static norms.Footnote ⁸⁶ This finding is important enough in and of itself and without any connection to the EIA or the no-harm rule. Yet, beside this dictum (better classified as obiter), the ICJ did not go any further. This attitude, even at the time of the Gabčikovo/Nagymaros judgment, is peculiar. It is not as if the parties to the dispute had not stressed the EIA, quite the contrary.Footnote ⁸⁷ While the Espoo Convention lays down detailed conditions for the conduct of an EIA, the numerous treaties that contain the obligation do not go into such a detail. At the same time, the ILC was not clear at all in its commentary to the Prevention Articles.Footnote ⁸⁸ It follows that the snapshot the courts attempt to take of the obligation to conduct and EIA is of great importance as it informs the content of a ubiquitous obligation that seeks clarification.

The ICJ in Pulp Mills affirmed the customary nature of the obligation to conduct EIA coupled with an ongoing obligation to monitor the operation of the activity. After it set out the ground rule of customary law, the ICJ went on to discuss the obligation to conduct an EIA. The ICJ states that the CARU (the Statute of the River Uruguay) does not provide any content to the obligation.Footnote ⁸⁹ It follows that up by stating that the scope and content of the obligation is not defined in general international law either.Footnote ⁹⁰ This grossly inaccurate statement is the only way the ICJ can reach its conclusion that ‘it is for each State to determine in its domestic legislation or in the authorization process for the project, the specific content of the environmental impact assessment’.Footnote ⁹¹ To be fair, this is the only time that Pulp Mills seems to follow the ILC in its assessment that each State determines the process and content of the EIA. But this is not entirely true. Indeed, the Commentary to Article 7 of the Prevention Articles of the ILC states that the specifics of what ought to be the content of the EIA are to be decided by the State that conducts the assessment.Footnote ⁹² This statement is however qualified further down in the Commentary. The ILC clarifies that the assessment must include the effects of the assessment. It goes on to add that the content of the assessment is dependent by the type of energy used for the activity, the activity’s size, as well as its location.Footnote ⁹³

Therefore, the meaning of the Pulp Mills dictum on EIA merits closer attention. If the ICJ only meant that it is through domestic legislation that each State gives effect to its international customary obligation to conduct an EIA, then there is nothing further to discuss. This is true, yet it is also self-evident. If the ICJ meant just that, then it simply did not add anything new to the discussion. It is also true, however, that if national legislation is to indeed give effect to the international obligation, it cannot merely dictate the conduct of an EIA without any kind of prerequisites. If this was the reasoning of the ICJ behind this proclamation, then this is obviously unsatisfactory for two main reasons. First, the ICJ does not bother to look either to the ILC Articles or at the numerous international instruments that provide for an EIA and at the same time provide for a rudimentary framework. It is only natural that the technical details should be determined in the domestic legislation, but the overarching framework and broad directions could, relatively easily, be inferred from the ILC Articles, IEL conventions and practice. Second, the conduct of the EIA itself is highly complex. It can easily lead to disagreements between the State of origin and the potentially affected State meaning that a deadlock in the discussions is not unprecedented.Footnote ⁹⁴ A clear framework could at least anticipate and address some of the most obvious obstacles in the process.

In Nicaragua/Costa Rica, the Court changed course in a spectacular way. It interpreted the obligation of an EIA in a rather open and progressive way, breaking away from Pulp Mills. It held that several factors are relevant in examining the proper adherence of the State to the obligation to conduct and EIA. The ICJ included the magnitude of the project, its context, the scale, location, as well as the geographical conditions.Footnote ⁹⁵ This is clearly an interpretation of the obligation that incorporates elements found in environmental agreements, albeit without any such acknowledgement whatsoever. The Espoo Convention, being the most comprehensive international instrument regulating the conduct of EIA, offers significant guidance. Appendix II of the Convention sets out a number of criteria that determine the conduct of an EIA. The Appendix includes the size and the location of the activity as well as the nature and kind of damage that may occur. The ILC Commentary also adheres to these criteria.Footnote ⁹⁶ They are very similar, practically identical to those propounded by the ICJ. In other words, the Court caught up with the evolution of international environmental law in order to define the scope and content of the obligation. Whether the ICJ moved in that direction because it looked into relevant environmental law instruments, state practice, or other developments is not possible to know.

It is the lack of elaboration on behalf of the ICJ that leaves behind a question mark as to whether this can be classified as evolutive interpretation, or it is merely the application of the obligation to the facts of the case. On the one hand, the ICJ in Nicaragua/Costa Rica quoted (not with disapproval) the Pulp Mills dictum that the State determines the content of the EIA obligation in its domestic legislation. At the same time, the ICJ also looked at the fact that Costa Rica’s domestic legislation provided for an emergency provision exempting the government from the obligation to conduct an EIA.Footnote ⁹⁷ It determined that the emergency provisions could not trump the obligation of Costa Rica to carry out an EIA under international law.Footnote ⁹⁸ This might be a hint of differentiation from Pulp Mills. Be that as it may, had the ICJ merely applied the Pulp Mills dictum as a definitive statement on the status as well as on the content of the obligation, it would have deferred any criteria on the content of the EIA to the relevant domestic legislation in question. The ICJ did something different. It interpreted the bare obligation as found in Pulp Mills. It gave flesh to its bones by interpreting it as containing specific criteria that the State of origin has to take into account when it transposes the international obligation to its domestic legal order. This, of course, is only one of the reasonable readings of the judgment, yet the most convincing. State practice, international agreements and the work of the ILC were all there when the ICJ decided Pulp Mills. It may be true that, rather than evolution, it seems more like catching up a few years later; the judgment is an example of expansive interpretation at least compared to the Court’s past jurisprudence.