The endogenous dynamics of the law and secondary rules

Scholars who have sought to grasp the essence of the law tend to converge on a number of minimum requirements, or defining properties of the concept, such as determinacy, consistency and non-arbitrariness. In one of the most influential treatments, Fuller establishes eight essential properties of law, or principles of legality, which, in his view at the same time constitute the inherent morality of a legal system and thus generate fidelity to the law.^{Footnote 10} In Fuller’s view, legal norms must be general, prohibiting, requiring or permitting certain conduct. They must also be promulgated, i.e. accessible to the public, and prospective, not retroactive, enabling citizens to know what the law requires. The law therefore must be clear as well; it should avoid contradictions; it must be realistic and not demand the impossible; its requirements should be relatively constant over time. Finally, there should be congruence between legal norms and the actions of officials operating under the law.^{Footnote 11} These features of the law are necessary in order to be perceived as legally legitimate, i.e. in order to attract adherence and generate fidelity to the law.

To achieve these features of the law in a complex society it is necessary to develop a system of rule-making. Upholding Fuller’s eight principles of legality thus endogenously implies a tendency for the development of secondary rules. In his seminal work ‘Fairness in International Law and Institutions’, Thomas Franck pointed out that the perceived fairness of a legal system is contingent upon it being ‘rooted in a framework of formal requirements about how rules are made, interpreted and applied’.^{Footnote 12} These ‘formal requirements’ perform various important functions in a legal system: they solve the conundrum of the law’s validity by establishing rules of recognition; they enable the progressive development of the law, for instance when changed political circumstances call for new norms (rules of change); and they provide means to settle norm conflicts and ways to interpret indeterminate norms (rules of adjudication).^{Footnote 13} One may add that they should also spell out the consequences of norm violations (norms of implementation). These ‘rules about the rules’ are important, since the law cannot stabilize normative expectations – one of its major functions – if it is unclear, for instance, what the law is, who has authority to apply and enforce it, and who can be called upon to resolve norm collisions. To bring two heroes of legal theory together: without secondary rules primary rules can hardly generate fidelity to the law.

For a long time, these secondary rules were only weakly developed at the international level. Against this background, HLA Hart questioned the quality of international law:

Hart considers international law to be a primitive legal system^{Footnote 15} which he contrasts with more advanced normative orders that include secondary rules. In primitive societies, Hart writes, conflicts between primary norms will not be resolved in the legal system itself, because there is no procedure for settling controversies over the interpretation of primary norms: ‘This defect in the simple social structure of primary rules we may call its uncertainty.’^{Footnote 16} A second defect from which primitive systems suffer is the static character of the rules: ‘There will be no means, in such a society, of deliberately adapting the rules to changing circumstances.’^{Footnote 17} The third deficiency is the lack of adjudicatory mechanisms: disputes as to whether a rule has been violated will occur interminably if there is no agent authorized to ascertain the fact of violation.^{Footnote 18} Hart concludes that the remedy for each of these defects primitive societies suffer from is the establishment of rules about primary rules.

In 1961 – when Hart published his book – the international legal order indeed consisted mainly of primary rules, lacking a dense layer of secondary rules, which is why Hart considered it to resemble a primitive society.^{Footnote 19} In the past decades, however, this ‘primitive order’ has developed a rather elaborate set of secondary rules which generate fidelity to the law, even though they are frequently contested and partly underdeveloped.^{Footnote 20}

Secondary rules and related concepts

While the concept of secondary rules is linked to related concepts like the rule of law and constitutionalization, it also differs from these in important aspects. First, we see it as less normatively laden than concepts such as constitutionalization and the rule of law. Second, we argue that the concept of secondary rules necessarily applies to a legal order as a whole which is not necessarily true for the concept of legalization. Both features – the low degree of built-in normativity and the broad scope – qualify secondary rules as the concept most apt for our analytical purposes, which allows us to make sense of current changes within the international legal system.

Normativity: Major legal theorists argue – for other reasons than Fuller – that criteria with a moral meaning distinguish law from other social norms and regulations.^{Footnote 21} As opposed to this view, legal positivists like HLA Hart and Joseph Raz do not assign any independent moral value to the law as such. On this view, law is only a necessary condition for a good political order, not a moral virtue in itself. Theorists writing about the ‘morality of law’ as well as legal positivists however converge in viewing secondary rules as necessary to preserve the virtues (moral or not) of the law in complex legal systems. In this sense, secondary rules possess built-in virtues. The concept of secondary rules is in our view, however, normatively less demanding than cognate concepts such as the rule of law or the notion of constitutionalism.

There is no universally accepted definition of the rule of law.^{Footnote 22} Whereas the minimal version of the concept of the rule of law – such as the one referred to in the Raz citation above – seems more or less identical with the features of law identified by Fuller (and thus with law as such), all other concepts of the rule of law demand a normative quality that goes beyond the mere presence of secondary rules.^{Footnote 23} Beyond the most minimalist version, the grammar of all arguments about the rule of law is roughly the following. The rule of law first of all implies authority of law and not of people. Therefore, people in positions of authority should exercise their power within a constraining framework of public norms that includes limitations on the abuse of public power for private gain. Second, the rule of law requires legal certainty and predictability.^{Footnote 24} This is the requirement that norms and rules be laid down clearly in advance and be accessible to all; it also includes adjudication procedures and the specification of the consequences of norm violations. Third, this concept of the rule of law includes legal equality, i.e. the law must be the same for all legal subjects and must not contain arbitrary distinctions or special personal prerogatives for those in a position of social or political power. Finally, and partially implied in the first three components, the rule of law implies the recognition of some basic individual rights, even though the exact content and scope of these rights may differ. Understood this way, all the variations of the concept of the rule of law – beyond the most minimalist one – are more deeply imbued with normative content than the concept of secondary rules. This does not exclude the possibility to use the notion of an international rule of law as an ideal type and to identify movements in this direction (or away from it),^{Footnote 25} but the international rule of law seems to be no appropriate term to capture the current international state of affairs.

The concept of constitutionalization also carries more normative baggage than the concept of secondary rules.^{Footnote 26} Constitutionalization in general can be described as a process in which different legal orders are integrated by the establishment of an ultimate legal authority in the form of a (written or unwritten) constitution that serves as higher law and enshrines fundamental values. While the rule of law may apply to each legal sphere, constitutionalism prescribes the way different legal spheres are integrated.

There are significant doubts about a global constitution. Legal pluralists suggest that major components of an international rule of law can be established in the absence of a global constitution. According to this view, a world in which different political authorities exist without being integrated in a hierarchical order must not be detrimental to the rule of law.^{Footnote 27} Similarly, it is argued that a number of possibilities and mechanisms – secondary rules so to speak – exist for solving conflicts between different legal orders. This may happen by means of mutual acceptance of general legal principles, thus allowing for a softer form of constitutionalization.^{Footnote 28} There are also good arguments that these collisions have to be solved outside the legal sphere, because they point to political matters.^{Footnote 29} In any case, the classical notion of constitutionalization seems to be rather demanding regarding the normative quality of primary and secondary rules.

Compared to both the rule of law and constitutionalization, the concept of secondary rules has less built-in normativity. While secondary rules are to some extent a necessary component of both the rule of law and constitutionalization, secondary rules do not necessarily need to have all the normative features of the rule of law or constitutionalization. We can think of legal systems with fully established secondary rules that openly violate principles of legal equality (as in the German Empire before 1914) or individual rights (as in China today) and do not possess the features usually associated with a constitution (the international legal system). The concept of secondary rules thus reduces the normative ambition inherent in other concepts such as the rule of law and constitutionalization, which makes it a more appropriate tool to capture the evolution of international law over the last decades.

Scope: A second feature of the concept of secondary rules is that it speaks at least implicitly to the international legal system as a whole instead of sectoral spheres or isolated legal subsystems. It is this feature which distinguishes it from the concept of legalization.^{Footnote 30} Legalization refers to a process by which it is ensured that regulations fulfil certain criteria. Abbott et al. identify three criteria of legalization: ‘obligation’, meaning that States or other actors are bound by a rule or a commitment; ‘precision’, meaning that the rules accurately and unambiguously define the conduct they require, authorize, or proscribe; and ‘delegation’, meaning that authority has been granted to third parties to implement the rules, including their interpretation, application, dispute settlement and further rule-making.^{Footnote 31}

Like our concept of secondary rules, legalization is normatively thinner than the rule of law. Contrary to the notion of secondary rules, it applies mainly to specific legal spheres, most often sectorally defined by issue areas such as trade or climate change or defined in terms of levels. To the extent that the concept of secondary rules also includes procedures for how to handle legal collisions between sectoral regimes as well as political levels, it reaches beyond specific issue areas and refers to a general framework of formal requirements about how rules are made, interpreted and applied.

In sum, for our specific purpose, accounting for and explaining the dynamics of the international legal system in the last decades, the concept of secondary rules is most apt since it provides a clear-cut descriptor which can be identified relatively easily, without normative assessments, and which refers to the legal system as a whole.

Explanatory accounts focusing on exogenous variables

While theorizing on the causal dynamics of secondary rule-making is still in its infant stages, some authors from both IR-theory and legal backgrounds have – at least implicitly – dealt with aspects of the subject matter under the banner of concepts such as legalization, procedural politics, interstitial law-making, etc. While the emergence and transformation of these norms is usually explained with reference to external forces, some of the accounts point at least implicitly to the need to theorize internal dynamics as well.^{Footnote 33}

Legalization theorists, for instance, have suggested a set of explanatory variables such as transaction costs, uncertainty, implications for national sovereignty, heterogeneity of preferences, and power differentials that affect the ‘hardness’ of legal regimes.^{Footnote 34} The authors show that depending on the circumstances, both the powerful and the weak stand to benefit from legalization. Weak States frequently prefer hard legalization because it reduces possibilities for the arbitrary use of power,^{Footnote 35} while powerful States frequently use legalized arrangements to realize their own preferences as efficiently as possible.^{Footnote 36} Even though Abbott and Snidal stress the disproportionate influence exerted by powerful States on legalized agreements, it would be wrong to equate legalization with the practice of the powerful. The authors equally show that in the process of legalization, actors have to argue within the parameters set by the law itself, that is, they have to frame their argument in such a way that it is consonant with existing norms and techniques of legal reasoning.^{Footnote 37}

Theories of procedural politics and interstitial change also analyse rules that are secondary in character. Procedural politics^{Footnote 38} – ‘everyday politics with respect to rules’^{Footnote 39} – is about how actors use, manipulate, and sometimes modify procedures to achieve their political goals. Jupille stresses that the choice of procedure is an ‘acutely political’ question,^{Footnote 40} which dovetails with our assumption that in secondary rule-making, the stakes are even higher than in primary rule-making, because the former determine in the first place who has a say in shaping the latter. While this work concedes that political actors do not enjoy unlimited wiggle-room but are constrained by existing rules,^{Footnote 41} it has little to say about how exactly these existing structures actually constrain actors. Similarly, Héritier and Farrell, who write about interstitial institutional change, primarily focus on actors’ bargaining strength as a determinant of institutional outcomes.^{Footnote 42} All of these approaches focus primarily on exogenous forces for legal development and in this way leave the endogenous forces such as legal consistency under-theorized.^{Footnote 43} Moreover, the need to construct normative consonance between explicit (black letter law) and the norms and mores of society at large is also neglected.^{Footnote 44}

In contrast to IR theory, international legal scholarship has only recently begun to engage in causal analysis of international rule-making.^{Footnote 45} Charlotte Ku’s and Paul Diehl’s monograph on The Dynamics of International Law probably comes closest to a comprehensive causal theory of secondary rule-making.^{Footnote 46} The authors ascribe two main functions to international law: providing the parameters and mechanisms for cross-border interactions, and shaping the values and goals that these interactions promote.^{Footnote 47} While they call the latter the ‘normative system’ of international law, the former is defined as law’s ‘operating system’ and roughly corresponds to Hart’s concept of secondary rules. Ku’s and Diehl’s goal is to provide an explanation for stasis or change in the operating system by fleshing out a number of factors:

They thus stress the obstructionist capability of powerful States over ‘constructive’ norm entrepreneurship, arguing that ‘most critically, we see the power of leading states as lying in their ability to block operating system change rather than impose such modifications’.^{Footnote 49}

While Ku’s and Diehl’s approach has a lot to offer, we would dispute some aspects of their theory. First of all, the term ‘operating system’ discounts the political character of secondary rules. Secondary rules do not form technical systems, but are essentially the core of a political order. Second, their relative neglect of actors other than States risks missing an important part of the picture, especially when it comes to the forces behind legal consistency. Third, and most importantly, the absence of obstructive power or more generally the absence of adverse political scope conditions is in our view not a necessary condition for the development of secondary rules. The internal consistency drive of the legal system may lead to the development of secondary rules, even when power politics and other external factors militate against it. To use Diehl and Ku’s terminology: necessity – or in our terms: the coherence of law – may trump obstructive power.^{Footnote 50}

In the following, we will flesh out a causal mechanism of secondary rule-making which is based on the assumption that the law’s internal drive towards consistency is backed by cognitive dissonance reduction. In so doing, we do not offer a complete account of secondary rule-making that competes with Diehl and Ku’s theoretical framework. We do not claim that the law’s internal drive towards consistency alone can explain the entire range of possible outcomes in the area of secondary rule-making. Rather, our goal is to explicate a causal mechanism which is an important part of the causal story and which is not sufficiently taken into account in existing theorizing. Without integrating such a mechanism into a complete theory, it is hard to understand why we observe additional, and often only partially successful attempts to strengthen the legal system in spite of rather adverse political scope conditions which materialized after the fall of the Twin Towers in 2001, when unilateralist thinking dominated US foreign policy, new rising powers re-emphasized traditional notions of sovereignty, and domestic resistance against international authority grew almost everywhere in the Western world.

Secondary rule-making as dissonance reduction

In most general terms, the integrity of law is based on coherence. Most people would reject ‘checkerboard laws’ that fail to treat like cases alike and introduce arbitrary differences.^{Footnote 51} According to Luhmann, it is the programming towards consistency that delineates law from other social systems.^{Footnote 52} We proceed from the assumption that the legal process – defined by Fuller as the ‘enterprise of subjecting human conduct to the governance of rules’^{Footnote 53} – exhibits an endogenous tendency towards consistency. If law loses consistency when getting more complex, it would lose one of its major virtues. This leads to a response. It is this built-in dynamic which leads to secondary rules in the face of growing legal density and complexity in the age of political denationalization. At first sight, this sounds like an argument of systemic or normative functionalism. We therefore aim at linking the ‘morality of law’ argument to secondary rules via a theory of social action. We use social processes which are rooted in social normativity and their cognitive correlates as explananda in order to understand the rise of secondary rules in the last decades. The role of cognitive theory in this argument is twofold.

As the legal system is programmed towards consistency it sets on the one hand the conditions for any change of legal substance. Opinio juris elevates a particular practice to a law-generating act. Whereas it is bound by shared understandings in the concerned society and the role hegemonic actors play therein, it is only through the articulation of opinio juris that international law can be (trans)formed: ‘[W]hat only opinio juris can do, only opinio juris can undo’.^{Footnote 54}Opinio juris is key to transforming facts and social norms into law. Yet opinio juris is crafted by individuals. The drive towards consistency of the legal system needs to be translated into actions within this system and is thus akin to the cognitive need of individual actors to avoid dissonance. This connection was also recognized by Ronald Dworkin in his writings on law as integrity, where he argues that ‘[o]ur instincts condemn’ checkerboard laws which fail to treat like cases alike and thus give rise to cognitive dissonance.^{Footnote 55} The opinio juris requirement and its dependence on coherence in turn qualifies the power hypothesis: powerful States cannot shape and reshape legal norms as they please – international legal norms are not commands, but are based on intersubjective standards of appropriateness developed in the legal system. Powerful States thus have to make an effort to create this intersubjective quality against the background of existing legal norms. In this sense, legal rules are both a product of power and a constraint upon its exercise.

On the other hand, the legal system cannot exist independently of shared understandings.^{Footnote 56} A legal system, as Neil McCormick explains, must go beyond legal consistency in a technical sense. Each legal rule and the system as a whole must make sense in the eyes of those who have to act upon it.^{Footnote 57} So there is a double requirement for the integrity of the law: technical consistency with other legal rules and coherence with the overarching principles held by a society. Actors seeking to justify a certain practice would have to show that their act is in line with existing law (i.e. they would have to construct normative consistency), or garner collective support for an amendment of the law that would bring the law in line with shared understandings held by society at large. If this adjustment does not happen, i.e. ‘[w]hen the conflict between a society’s shared understandings and legal rules is too stark, legality can come under significant strain’.^{Footnote 58}

To the extent that the normative features of law are part of the shared understandings, the practice of secondary rules increases the compatibility of the law not only with opinio juris, but also with broader societal norms. Considering the distinct rationality of the law, a theory of secondary-rule making which does not take into account how the need for consistency is translated into a disposition of norm-makers to construct legal consistency would thus be reductionist. Martti Koskenniemi famously articulated the notion of international law as constantly oscillating between apology and utopia. International law, he argued, cannot be understood as completely detached from political interests and State behaviour, because then it would seem utopian and slip into insignificance.^{Footnote 59} Yet on the other hand, a law which lacks distance from State behaviour and is widely regarded as being merely a fig leaf for the pursuit of the selfish interests of powerful States loses its legitimacy which it based on its built-in normativity. The law thus seeks to somehow reconcile two conflicting demands. On the one hand, the law must be responsive to political interests (including those of the most powerful actors in the system), if it does not want to be completely sidelined, yet on the other hand, the morality of law requires that the legal system retain a certain degree of autonomy from the political sphere, and that the law be seen as binding the powerful and the weak alike – even if this takes place to different degrees. Otherwise, the concept of law would be stripped of some of its core components, especially equality and non-arbitrariness – and consequently of its virtues and (if present) morality. The legal system’s presumption in favour of normative consistency translates into the need for secondary rules especially in contexts of high legal density and a fluid and contested normative environment. In these situations, secondary rules can serve the maintenance of legal consonance.

One can conceptualize this drive for internal consistency in the legal system by utilizing cognitive psychology. In his classic contribution, A Theory of Cognitive Dissonance, Leon Festinger^{Footnote 60} proceeds from the basic assumption that human beings constantly strive to achieve cognitive consonance, or reduce cognitive dissonance, respectively. The term cognitive dissonance denotes ‘the existence of non-fitting relations among cognitions’, the latter being defined as ‘any knowledge, opinion, or belief about the environment, about oneself, or about one’s behavior’.^{Footnote 61} The theory of cognitive dissonance is part of a larger set of theorems in psychology that emphasize the power of consistency. Nobel laureate Daniel Kahneman conceives of the human mind as divided into two subsystems – system 1, which is intuition-based and thus in charge of ‘fast-thinking’, and system 2, which is tasked with reflection – ‘slow-thinking’ – and thus (ideally) functions as a check on system 1. System 1 strongly prefers coherent causal stories over incoherent ones, and, in its drive towards consistency, also influences our intellectual system 2.^{Footnote 62} Several psychological aspects follow this logic. The so-called halo effect, for instance, testifies to the human mind’s desire for coherence. If people are given identical lists of characteristics of two individuals – one list starting with the positive features and ending with the negative ones, and the other grouping the same characteristics in reverse order – a clear majority of people will judge the first person more positively than the second.^{Footnote 63} Similarly, the tendency to reduce ambiguity as well priming and framing effects are based on this mechanism.

Using these insights, we can identify three archetypes of dissonance as sources of normative change: legal norm–behaviour dissonance; legal norm–legal norm dissonance; legal norm–shared understanding dissonance:

• • First, if a given behaviour violates an existing and accepted rule, this creates cognitive dissonance in the original meaning of Festinger. Norm–behaviour dissonance is a source of institutional change. Assume a powerful State X is criticized for violating a given primary norm. The alleged culprit may respond to these charges by arguing that the primary norm in question was not applicable to her behaviour in the first place; or that it was applicable, but should be interpreted in a way that would render her behaviour legal; or that her behaviour was illegal but legitimate, since it responded to overwhelming non-legal (moral or political) considerations with which the law ought to be made consonant. Since (primary) norms are inherently indeterminate, such conflicts over the interpretation of the law will generate a normative demand for secondary rules such as rules of adjudication, in order to decide among competing claims and thus restore legal certainty and the morality of law. In these discourses over the interpretation of primary norms, actors seeking to reduce dissonance between their behaviour and existing norms tend to get rhetorically entrapped, because other actors will expect a consistent commitment to the enunciated norms in the future, demanding fidelity to the law in a Fullerian sense. Operationally speaking, the justification of given interests in legal terms and its contestation thus precede the rise of secondary norms. Let us call this mechanism the justificatory trap mechanism. ^{Footnote 64} Norm–behaviour dissonance may give rise to secondary rules via a second pathway, for external actors who point to such a norm and justify their criticism with reference to a particular norm begin, according to the logic of reducing dissonance, to further internalize this norm. We know from psychological research that humans have a tendency to convince themselves of the rightness of their doing and in this way internalize norms which were originally (partially) used to legitimize our behaviour. If, for instance, a policy official justifies strategically motivated criticism with reference to a legal norm, the norm may move up in her preference ordering. In the same way, if representatives of international organizations justify intrusions into domestic legal orders with reference to the rule of law or democracy, the international organization is more likely to incorporate these norms into its own policies and procedures. Operationally speaking, to the extent that international organizations are involved in regulating secondary rules of Member States, they tend to internalize these secondary rules. Let us call this the self-persuasion mechanism. ^{Footnote 65}

• • The second type refers to a dissonance between different legal norms. If compliance with a given legal norm leads to non-compliance with another, a desire to resolve this norm–norm dissonance arises. To the extent that the density of primary rules increases, collisions between different primary rules as well as conflicts over the implementation of these rules become more likely. The drive for consonance thus leads to a growing demand for establishing secondary rules in order to handle these conflicts on the legal turf. Operationally speaking, secondary rules should therefore develop in parallel with primary rules. This leads especially to rules about norm-interpretation and norm-implementation. Let us call this the conflict resolution mechanism. ^{Footnote 66}

• • The third type of dissonance arises in situations in which a given understanding of a legal norm violates broader societal norms and understandings. In this case we speak of legal norm–shared understanding dissonance. Assuming a long-standing legal norm – such as open discrimination of certain members of society – does not sit well with the broader normative understandings prevalent in this society anymore. In this case there are essentially two ways to bring the norm in line with shared understandings to maintain the legitimacy of law: one is court decisions through which a reinterpretation of existing law takes place; the other option is legal change via new rule-setting. Both responses require secondary rules. Let us call this the normative adaption mechanism. ^{Footnote 67}

These three types of dissonance may eventually lead to more complex legal systems composed of primary as well as secondary rules. Legal actors dislike dissonance and therefore have an internal tendency to resolve it. This tendency is reinforced by observers who expose and criticize these dissonances and therefore put external pressure on the system to adapt itself. In sum then, we aim at identifying a causal mechanism that is based on an internal dynamic towards the proliferation of secondary rules in international law. This dynamic is embedded in the legal system and triggered by the increased complexity and density of international law. It can lead to the development of additional legalization in the form of secondary rules even when political conditions are not conducive to such institutionalization.

The International Criminal Court and the crime of aggression

Rules of adjudication help to resolve disputes arising out of diverging interpretations of primary norms which result from the law’s inherent indeterminacy.^{Footnote 69} While this is a problem common to all legal systems, it is particularly acute in the pluralist and fragmented global realm. In the international realm rules of adjudication which regulate the relationship between different legal regimes assume more and more relevance to prevent the ‘loss of an overall perspective on the law’.^{Footnote 70} In order to ensure the overall coherence of international law it has been suggested to empower the International Court of Justice (ICJ) to review the decisions of other international tribunals and authoritatively resolve norm collisions.^{Footnote 71} The ICJ was established by the United Nations Charter and has the authority to settle disputes between States which have consented to its jurisdiction, as well as issuing advisory opinions on international legal matters.^{Footnote 72} At the 2010 Review Conference on the Rome Statute establishing the International Criminal Court (ICC) in Kampala, Uganda, which was tasked, inter alia, with clarifying the ICC’s jurisdiction over the crime of aggression,^{Footnote 73} the ICJ was considered as one possible trigger activating the ICC’s jurisdiction over the crime of aggression.

The crime of aggression has long been the subject of debate among scholars and practitioners alike. After the Nuremberg trials, hopes were high that the new criminal law of individual responsibility for the crime of aggression would soon become part and parcel of the international rule of law, yet these hopes were immediately quashed because States could not even agree on what the term ‘aggression’ actually meant:

These contestation processes pitted the powerful against the powerless, with the former insisting on a narrow definition of aggression that would not prevent them from projecting force abroad, whereas the latter supported a broad definition that would shield them from all kinds of external interference.^{Footnote 75} What was needed was therefore not only a more precise definition of the primary norm of non-aggression, but also secondary rules of adjudication which would move conflicts about whether or not the crime of aggression had been committed from the political terrain to the legal turf and thus contribute to a strengthening of the rule of law.

When the US invaded Iraq in 2003, the perceived dissonance between the invasion of a sovereign State with Article 2(4) of the UN Charter unleashed a fierce debate about what it means to wage a war of aggression, resulting in the newly established ICC being flooded with over 240 communications concerning the legality of the war.^{Footnote 76} Since the ICC at the time did not have jurisdiction over the crime of aggression,^{Footnote 77} the Prosecutor declared that he did not ‘have the mandate to address the arguments on the legality of the use of force or the crime of aggression’ and instead referred to the ongoing deliberations on these issues in the ICC Assembly of States Parties (ICC ASP).^{Footnote 78} These deliberations focused on both primary norms (i.e. what does aggression mean? How should the prohibition on the use of force in Article 2(4) of the Charter be interpreted? Does it cover, for instance, anticipatory self-defence and humanitarian intervention?) and secondary rules of adjudication, i.e. which body should be empowered to decide whether aggression had taken place?

Negotiations at the 2010 ICC Review Conference, which took place in Kampala, Uganda, revolved around three^{Footnote 79} main issues: first, the definition of the crime of aggression; second, the triggers activating the ICC’s jurisdiction; and finally, the role of State consent, which turned out to be intimately linked to the issue of trigger mechanisms. The Review Conferences adopted a draft definition of aggression based on an agreement in 2002, when the ICC Assembly of States Parties had established a Special Working Group on the Crime of Aggression (SWGCA).^{Footnote 80}

The ‘elephant in the room’ at Kampala were possible exceptions to the prohibition on the use of force, such as anticipatory self-defence or humanitarian intervention.^{Footnote 81} The US, having conducted both types of interventions in the past and thus having been confronted with claims that its behaviour was inconsistent with some of the most deeply entrenched norms of international law, had sought to make its behaviour seem consonant with international law by lobbying for a restrictive interpretation of the concept of aggression that would not cover such interventions. A number of other Western States equally opted for a restrictive definition of aggression, whereas most non-aligned countries favoured a broad definition along the lines of the draft provided by the SWGCA that made reference to the list of acts contained in the annex to General Assembly resolution 3314^{Footnote 82} without specifying a particular threshold and thus covering a broad range of forcible measures.^{Footnote 83} In order to garner support for a restrictive definition of aggression, the US pointed out that the proposed definition was at odds with customary international law (norm–norm dissonance).^{Footnote 84} This would in turn undermine the fundamental principle of legality:

In the end, however, the delegations present at Kampala did not share Washington’s concerns, and adopted a broader definition based on the SWGCA’s draft.^{Footnote 86} At the insistence of the US delegation, however, a list of understandings was added to the definition in order to clarify the level of gravity required to establish that the crime of aggression had been committed.^{Footnote 87}

While the definition of the crime of aggression was a highly sensitive question, the issue of secondary rules of adjudication proved to be even more controversial. The debate revolved basically around the question whether the United Nations Security Council (UNSC) should be given a monopoly on adjudicating whether the crime of aggression had taken place and thus trigger proceedings before the ICC, as suggested by the International Law Commission’s 1994 Draft ICC Statute.^{Footnote 88} The ILC approach would have further politicized the work of the Court, however, and was therefore difficult to reconcile with the notion of the international rule of law and the separation of powers. As Summers points out, a

No wonder then that even within the ILC the proposal was contested. Doubts were voiced regarding the proposal’s ability to attract sufficient support within the ICC ASP, which turned out to be justified, as the negotiations at the Review Conference showed. Most delegations were opposed to a secondary rule that would accord the UNSC an exclusive role in triggering the ICC’s jurisdiction over the crime of aggression. This opposition ‘reflected deep dissatisfaction with the Security Council among developing countries, and concern that Security Council decisions would reflect the interests of the five permanent members rather than the interests of the international community as a whole’^{Footnote 90} (norm–shared understanding dissonance). Those who were in favour of a Security Council monopoly (the P5, unsurprisingly), however, had long maintained that allowing other bodies such as the ICJ or the UNGA to trigger the ICC’s jurisdiction would be at odds with the UN Charter. In 2001, already, Washington had voiced concerns that such a transfer of authority would raise ‘profound issues of consistency with the Charter … Neither the General Assembly nor the International Court of Justice may properly infringe upon the role given exclusively to the Security Council by the UN Charter’^{Footnote 91} (norm–norm dissonance). However, the Charter speaks of the Council’s ‘primary’, not exclusive, responsibility for the maintenance of international peace and security; hence it is not surprising that the majority of delegates did not perceive a dissonance between the provisions of the UN Charter and the proposal to endow actors other than the UNSC with the competency to activate the ICC’s jurisdiction.

After prolonged haggling, a two-track procedure for activating the Court’s jurisdiction over the crime of aggression was adopted. Under the first track, the ICC’s jurisdiction would be activated by a Security Council referral.^{Footnote 92} Under the second, the Court’s jurisdiction would be activated by State referral or the Prosecutor’s exercise of her proprio motu powers.^{Footnote 93} However, before launching an investigation, the Prosecutor must consult the UNSC and may only go ahead with the investigation if the Council either determines that an act of aggression has occurred, or fails to make a determination on this issue.^{Footnote 94} In the absence of a Security Council determination, the Prosecutor must receive a green light from the ICC’s Pre-Trial Chamber before opening an investigation.^{Footnote 95} So far, the secondary rules governing the ICC’s jurisdiction over the crime of aggression resembles the regime governing the Court’s jurisdiction over the other three core crimes of the Rome Statute. However, they depart from the latter in important ways, namely by providing for an opt-out mechanism for State parties and by positing that the ICC may not exercise jurisdiction in respect of a crime of aggression committed by the nationals or on the territory of States that are not parties to the Rome Statute (except, of course, in cases in which the UNSC has referred a situation).^{Footnote 96} This means that officials from States not parties to the Rome Statute who possess veto rights in the UNSC will be effectively shielded from the ICC’s prosecution. Had the US not exercised its leverage, it is unlikely that the decision that was ultimately adopted would have exempted non-parties from key portions of the secondary rules regime governing the ICC’s jurisdiction.^{Footnote 97} This significant concession to geopolitics notwithstanding, the preservation of the ICC’s powers to act in the absence of a Security Council referral was greeted as an important achievement in ‘detaching prosecution from political whims’.^{Footnote 98} The conference outcome also shows the normative adaptation mechanism at play, as the perceived dissonance between a UNSC monopoly and the interests of the international community at large was reduced by crafting a new secondary rule that accorded the UNSC a prominent, yet not exclusive role in activating the ICC’s jurisdiction.

In sum, different types of dissonance were involved in the debate over giving the ICC jurisdiction over the crime of aggression. The debate was initially spurred by the perceived incompatibility of Washington’s invasion of Iraq with the rules governing the use of force (norm–behaviour dissonance). When States met at Kampala to design a regime of primary and secondary rules governing aggression proceedings before the ICC, the US justified its position in the following manner. With regard to the primary norm defining the crime of aggression, Washington argued that the proposed definition was at odds with customary international law (norm–norm dissonance). With regard to the secondary rules of adjudication, the US – seconded by the other permanent Security Council members – argued that allowing bodies other than the UNSC to trigger aggression proceedings would be inconsistent with the UN Charter (norm–norm dissonance). This viewpoint was called into question by most other delegations who pointed to the Council’s lack of legitimacy in representing the interests of the international community at large (norm–shared understanding dissonance). In the end, the secondary rules agreed upon at Kampala reflected a complex compromise which naturally did not eliminate the dissonances perceived by each actor or group of actors entirely, but at least mitigated some of them, and was greeted as an important achievement in strengthening the international rule of law.^{Footnote 99}

The UN Security Council and the responsibility not to veto

The question of whether the five permanent members of the UNSC’s sway over the making and application of international law is consonant with the rule of law was not only a hotly debated issue at Kampala, but is also at the heart of the debate over the responsibility not to veto (Rn2V), an offshoot of the responsibility to protect (R2P). The history of both concepts illustrates the justificatory trap mechanism at play, i.e. it shows how powerful States became entrapped in their own normative rhetoric and how contestation about the interpretation and application of the emerging primary norm of R2P triggered demands for a secondary rule of implementation.

The enormous selectivity of humanitarian interventions authorized by the UNSC was seen increasingly as undermining the legitimacy and ‘morality’ of those interventions.^{Footnote 100} A new secondary rule – the responsibility not to veto – is meant to reduce norm–behaviour dissonance by ensuring consistency in the application of R2P, thus heightening the legitimacy of the UNSC as a protector of civilians and bolstering its fidelity to the law. At the same time, the debate over Rn2V exposed another type of dissonance, that is, legal norm–shared understanding dissonance. While the UN Charter imposes few legal limits on the use of the veto, an intersubjective consensus has emerged in recent years that even though those who pay the costs of an intervention should be accorded a privileged position in the Council’s decision-making, they should not (ab)use this privilege to further their narrow national interests. The concept of Rn2V has not yet consolidated into a secondary rule. This does not mean, however, that pressure on the P5 to explain and justify the (ab)use of their veto rights has subsided, on the contrary. In a remarkable volte-face, one P5 member – France – affirmed its support for the responsibility not to veto, thus fuelling hopes that the other four permanent members will follow suit.

The concept of Rn2V emerged from the debate over the responsibility to protect. The gist of the concept of the responsibility to protect is that ‘state sovereignty implies responsibility, and the primary responsibility for the protection of its people lies with the state itself’.^{Footnote 101} However, if a State proves unwilling or unable to live up to its responsibilities, i.e. if it violates basic human rights or does not prevent such violations, the international community has a residual responsibility to act. The principle of non-intervention thus yields to the international responsibility to protect.^{Footnote 102} The discourse about R2P was triggered by a decade of (mostly) Western interventionism on behalf of human rights in the post-Cold War era, culminating in NATO’s controversial bombing of Serbia in 1999. Even though the intervening parties did not consistently assert a legal right to humanitarian intervention but mostly justified their actions on moral grounds, the war against Serbia did set off a process of self-entrapment, as a normative expectation began to emerge that those same States that had intervened in Serbia to safeguard human rights would display a consistent commitment to the emerging (yet still opaque) norm of atrocity prevention in the future in other places of the world as well.

A string of humanitarian crises which followed the Kosovo intervention and the publication of the ICISS report has demonstrated that patterns of accepted justifications in the field of atrocity prevention have indeed changed, and that the language of R2P has provided proponents of intervention with a powerful vocabulary to press for tough action against human rights violators. Yet those States which had intervened in Kosovo in 1999 have not been consistent in their commitment to atrocity prevention, thus giving rise to dissonance between their proclaimed (legal or moral) responsibility to protect civilians from mass atrocities on the one hand, and their actual behaviour on the other hand. One of the reasons for this implementation deficit is that the content of the primary R2P ‘norm’ remains opaque;^{Footnote 103} another reason is that the secondary rules governing the implementation of R2P are contested, in particular those regarding the UNSC’s responsibility to react to atrocity crimes. With regard to the latter, the concept of the responsibility not to veto has recently gained traction internally.^{Footnote 104} The origins of Rn2V date back to the ICISS report, whose authors proposed that ‘a permanent member, in matters where its vital national interests were not claimed to be involved, would not use its veto to obstruct the passage of what would otherwise be a majority resolution’.^{Footnote 105} The main challenge faced by proponents of Rn2V is that in order to modify – through an amendment of the UN Charter – the existing secondary rule which grants the P5 veto powers over implementing R2P, they will need the consent of just those actors whose privileges they seek to curtail, because a redistribution of authority for the maintenance of international peace and security requires the consent of a two-thirds majority of UN Member States, including the permanent members of the Security Council. The possibilities for normative innovation are thus circumscribed by existing procedures, which grant a handful of powerful States a veto over the revision of these procedures.

The veto rights of the P5 have long been anathema to the ‘significant majority of Member States’^{Footnote 106} who believe that this places a small group of powerful States above the law and who have sought to rewrite the secondary rules governing decision-making in the UNSC. In 1993, the UN General Assembly established an open-ended working group to consider ‘all aspects of the question of increase in the membership of the Security Council and other matters related to the council’.^{Footnote 107} The working group organized several rounds of consultations on these matters which tackled, inter alia, the issue of limiting the use of the veto in situations involving mass atrocities.^{Footnote 108} However, consensus on these matters was not forthcoming, as the P5 insisted that any reform of the Security Council leave the essence of their veto powers intact.^{Footnote 109} The consultations organized under the auspices of the working group thus quickly reached a dead end.^{Footnote 110}

In light of the almost insurmountable procedural hurdles of amending the UN Charter, a coalition of smaller States (the S5 – Costa Rica, Jordan, Liechtenstein, Singapore and Switzerland), seconded by civil society, has begun to explore a different avenue for changing the rules of implementation for R2P through a non-binding General Assembly resolution, where rule-making power is much more evenly distributed than in the UNSC. This resolution would establish a code of conduct for the P5, asking the latter to abstain from using their veto in situations involving mass atrocities and also requesting those States which cast a veto nonetheless to publicly explain the reasons for doing so, ‘in particular with regard to its consistency with the purposes and principles of the Charter of the United Nations and applicable international law’.^{Footnote 111} The S5 sought to make their proposal palatable to Member States by highlighting its consistency with the provisions of the UN Charter and the 2005 World Summit Outcome Document.^{Footnote 112} At the same time the S5 proposal was geared towards forcing those who do cast a veto to acknowledge the dissonance between their actions and widely accepted norms of international law (norm–behaviour dissonance). Hence, Jordan’s Ambassador Prince Zeid claimed that while the UN Charter accords the Security Council primary responsibility for the maintenance of international peace and security, the Charter also requires that the Council’s decisions be consonant ‘with the principle of justice and international law’.^{Footnote 113} Genocide and mass slaughter, he continued, are certainly not consistent with those principles.^{Footnote 114}

Even though the S5’s push for enhanced Security Council accountability in principle seemed to enjoy a significant measure of support among Member States,^{Footnote 115} when the S5 decided to put their draft resolution to a vote, they encountered great opposition on the part of the P5.^{Footnote 116} When it became clear that the draft resolution could not muster the requisite majority in the General Assembly – what exactly constituted the ‘requisite majority’ in this case was subject to debate, because the secondary rules governing UN General Assembly decisions on matters relating to the Security Council are themselves contested^{Footnote 117} – the S5 decided to withdraw the resolution. They announced, however, to further consult with those Member States that had expressed support for the draft resolution and, should these consultations lead to a new dynamic, to return to the matter.^{Footnote 118}

And indeed, the S5 initiative gathered new momentum in 2012, when charges of hypocrisy where increasingly levelled against the UNSC for its failure to hold the Syrian regime responsible for the atrocities committed against its own people. Referencing draft resolution A/66/L.42/Rev.2 on the responsibility not to veto, the Singaporean delegate, for instance, denounced the behaviour of

Remarkably, at the same debate one of the P5 publicly announced that it would henceforth support the S5 push for limiting the use of the veto in cases involving mass atrocities: ‘France supports the permanent members of the Council voluntarily and jointly foregoing the use of the veto in situations under the Council’s consideration in which mass atrocities are being committed and, more generally, which pertain to the responsibility to protect’, the French representative proclaimed, thus increasing justificatory pressure on the other permanent members to explain the dissonance between their verbal commitment to R2P on the one hand, and their failure to live up to their responsibilities in the context of the current crisis in Syria on the other hand.^{Footnote 120} The general tone of the debate was one in favour of a Rn2V, with only Russia explicitly speaking out against it. It thus seems fair to conclude with Liechtenstein’s ambassador that ‘the accountability needs and the relevant pressure on the Council are increasing, as is illustrated in connection with the widespread and systematic crimes committed against the civilian population in Syria’.^{Footnote 121} In the case of France, the justificatory trap mechanism has resulted in a remarkable policy change and has nourished hopes that a stable normative expectation will emerge in the future that the international community will intervene on behalf of endangered civilians whose government manifestly fails to meet its sovereign responsibilities. In 2013 a new initiative of 22 UN Member States (ACT) was launched which took over the baton from the S5 and now pushes for increased accountability, coherence, and transparency of the UN Security Council’s decision-making processes.^{Footnote 122} The initiative has already had some success in driving a wedge between the P5, as not only France is sympathetic, but apparently also the UK, which participated in the launch of ACT last May.^{Footnote 123} Observers moreover believe that ‘pressure is likely to mount on the US, given that the incoming UN Ambassador, Samantha Power, is a prominent advocate of mass atrocity prevention’.^{Footnote 124}

In sum, the debate over Rn2V tells us several things about the role of dissonance reduction in legal systems, and here especially norm–behaviour dissonance/legal norm–shared understanding dissonance. For one, it shows that even the most powerful members of the international community have come under pressure to answer to those who are affected by their decisions, and to demonstrate that their behaviour is consistent with fundamental primary norms of international law. Whereas the supporters of Rn2V have not prevailed so far, at least the requirement of reason-giving enshrined in Rn2V seems to be established by now. The debate over Rn2V has demonstrated that these processes of reason-giving have triggered a dynamic of its own which may at the end of the day lead to the adoption of a new secondary rule. While no further adjustment of the secondary rules of implementation has taken place yet, the P5 have come under increasing pressure to justify their privileges, especially if these prerogatives lead to a failure to act decisively in the face of mass atrocities.

Universal jurisdiction and the concept of immunity

The ascendancy of the contested concept of universal jurisdiction has led to collisions with long-established primary norms of international law, such as the norm of immunity (norm–norm dissonance). The history of the concept illustrates how the increasing density of primary norms generates a demand for secondary rules (such as rules of implementation) that can resolve such collisions between primary norms (conflict resolution mechanism). Just like in the other cases described above, the struggle over the scope and meaning of the concept of universal jurisdiction pitted powerful against weaker actors – in this case, the European Union against the African Union (AU).

The controversy came to its head when the ICJ handed down a highly controversial judgment in the Arrest Warrant case. The diplomatic haggling that accompanied the proceedings before the ICJ is indicative of the contestedness of the primary norms of universal jurisdiction and immunity. In 2000, the Democratic Republic of the Congo (DRC) initiated proceedings against Belgium before the ICJ in respect of a dispute concerning an arrest warrant issued by Belgium against Congolese Foreign Minister Abdulaye Yerodia Ndombasi, charging him with war crimes and crimes against humanity.^{Footnote 125} The arrest warrant was issued on the basis of Belgium’s 1993 statute providing for jurisdiction over war crimes and crimes against humanity wherever they are committed.^{Footnote 126} The principle of universal jurisdiction, which underlies the statute, posits that certain crimes are so abhorrent that every State in the world is entitled (or even obliged) to prosecute them, regardless of where the crime was committed and regardless of the nationality of the victims, perpetrators, or any other type of connection with the prosecuting country.^{Footnote 127} In its application, the DRC claimed that ‘[t]he universal jurisdiction that the Belgian State attributes to itself’ was dissonant with established principles of international law, namely the ‘principle that a State may not exercise its authority on the territory of another State’, the principle of sovereign equality, as well as Yerodia’s immunity.^{Footnote 128} Subsequently, however, the DRC reduced the scope of its application to the alleged violation of Yerodia’s immunity.^{Footnote 129} The Court therefore confined itself to making a pronouncement on the issue of immunity while leaving aside the question of universal jurisdiction. This decision was criticized by a number of individual judges, who observed that ‘there can only be immunity from jurisdiction where there is jurisdiction’.^{Footnote 130} In its application, the DRC argued that incumbent foreign ministers enjoy absolute immunity from criminal prosecution.^{Footnote 131} Belgium, by contrast, contended that immunity does not shield an official from prosecution for acts committed not in performance of his or her functions, particularly if the act in question violated international jus cogens.^{Footnote 132}

The Court ruled that incumbent foreign ministers enjoy immunity from prosecution in foreign national courts for acts committed in an official capacity,^{Footnote 133} and that no exception to this rule exists in respect of war crimes or crimes against humanity.^{Footnote 134} It emphasized, however, that immunity should not be equated with impunity, as an incumbent foreign minister could still be tried by his or her home State or before an international tribunal.^{Footnote 135} The Court also stated that after a person ceases to hold the office of Minister for Foreign Affairs, he or she will no longer enjoy all of the immunities accorded by international law in other States. Provided that it has jurisdiction under international law, a court of one State may try a former Minister of Foreign Affairs of another State in respect of acts committed prior or subsequent to his or her period in office, as well as in respect of acts committed during that period of office in a private capacity.^{Footnote 136}

Yet the Court did not take a stand on whether war crimes and crimes against humanity fall into the latter category. In her dissenting opinion, ad hoc Judge Van den Wyngaert challenged the Court’s interpretation of the scope of immunity, claiming that foreign ministers do not enjoy immunity for violations of jus cogens.^{Footnote 137} She also regretted that the Court had dealt with the issue in a very technical way, thus bypassing the underlying fundamental norm–norm dissonance:

Van den Wyngaert concluded that international law indeed permits universal jurisdiction over international core crimes^{Footnote 139} and that universal jurisdiction could also be exercised in absentia, contrary to the DRC’s assertion.^{Footnote 140} Even though Van den Wyngaert’s view was not shared by the majority of judges on the bench, her dissenting opinion exposed the norm–norm dissonance underlying the case, and flagged the need to (re)balance the diverging interests protected by the norms at issue.

The Yerodia incident prompted the AU to mount a flurry of diplomatic activity aimed at preventing what was perceived as a politicization of the principle of universal jurisdiction. African leaders felt that the (public) issuance of arrest warrants for sitting African State officials was dissonant with fundamental principles of international law such as the sovereign equality of States, the officials’ immunity, and the presumption of innocence: ‘For African states, this evokes memories of colonialism.’^{Footnote 141} The AU therefore called for secondary rules that could help reduce this dissonance by clarifying the scope of the concept of universal jurisdiction, thus reducing possibilities for politicizing this concept and re-establishing legal certainty. More specifically, the AU proposed the establishment of an international regulatory mechanism ‘with competence to review and/or handle complaints or appeals arising out of abuse of the principle of universal jurisdiction by individual states’.^{Footnote 142} Moreover, following intense lobbying by the African Group, the UN General Assembly eventually adopted a resolution calling for further discussion and clarification of the concept.^{Footnote 143}

While no international regulatory mechanism has been established yet which would enshrine rules of adjudication/implementation that could provide guidance as to how to weigh universal jurisdiction versus immunity, the matter continues to be high on the agenda of global governance, as the heated debate sparked by the issuance of an ICC arrest warrant against Sudanese President al-Bashir has shown.^{Footnote 144} To the extent that the number of countries which have adopted a concept of limited immunity is now in the majority and still growing, this will not change. Currently, the AU is seeking an ICJ advisory opinion on the scope of the immunity principle through the UN General Assembly, hoping to thereby clarify the contours of this contested primary norm of international law. In sum then, the immunity of State officials from criminal proceedings in foreign national courts remains a contested issue.^{Footnote 145} The underlying norm–norm dissonance has triggered demands for the creation of secondary rules institutionalized in a new international regulatory mechanism.