2. The structure of CAH in law

The central task of this article is to confront the predominant philosophical literature on the normative concept of CAH to the fast-developing law of CAH. Accordingly, the first step is to interpret the structure of CAH as found in the Rome Statute (Art. 7) and in the case law of international criminal courts. To interpret the structure of CAH, and later to navigate through the philosophical literature, I distinguish between two core elements of the definition: the nature of the crimes (E1), which captures what is done to the victims (each individual victim being attacked in a distinctive way), and the nature and capacity of the agent (E2) committing the crimes, which distinguish the kind of agent that lies behind the crimes and the broader context that enables them.

As far as E1 is concerned, CAH include the most abominable acts of violence against individuals such as murder, enslavement, rape, torture, or deportation, among others. What brings commonality to all those offences is their extreme level of violence and suffering (including death) endured by the victims both physically and psychologically. Therefore, for one to be a victim of a crime against humanity, one must be attacked in a particularly atrocious way. The extreme nature of the offence certainly raises the concern of the international community. As Renzo puts it, ‘the seriousness of these crimes, in turn, is what normally accounts for the fact that they are international crimes. It is because these crimes are particularly serious that they cannot be left unpunished when domestic courts fail to prosecute them’.^{Footnote 12} As the Preamble to the Rome Treaty puts it, those crimes refer to ‘unimaginable atrocities that deeply shock the conscience of humanity’.

As to E2, CAH must be committed as part of a ‘widespread or systematic attack directed against any civilian population, with knowledge of the attack’ (Art. 7(1)). This is what has been called the ‘contextual element’^{Footnote 13} in the literature and in the case law. The contextual element also requires that the agent is aware both that her actions are part of the larger attack and that those actions would result in the acts constitutive of the crimes (mens rea requirement). The contextual element has remained central to qualify the crimes in the case law and to distinguish international criminal law from domestic criminal law. As Jalloh explains, the negotiations leading to the Rome Statute clearly indicated that the clause ‘was one way to distinguish attacks of a widespread or systematic nature, rather than acts carried out by some random persons or bands of criminals acting on their own initiative for their own purely selfish motives’.^{Footnote 14} As we shall see in more detail later, the ‘systematic’ criterion plays a more important role than ‘widespread’ in distinguishing the class of international crimes.

In addition, Art. 7(2)(a) specifies what is implied by an ‘attack directed against any civilian population’: ‘a course of conduct involving the multiple commission of acts referred to in paragraph 1 against any civilian population, pursuant to or in furtherance of a State or organizational policy to commit such attack’.^{Footnote 15} This is what is referred to as the ‘policy element’ in the literature. This clause further circumscribes the kind of agent (state or state-like organization) responsible for the planning and the execution of the attack. There is a continuous debate among international lawyers, however, on the appropriate nature of the entity (e.g., a state or a state-like entity), on the nature of the policy (e.g., formalized or not, ‘policy’ or ‘plan’) and on whether the clause constitutes an independent requirement (e.g., in contrast to just supporting evidence that the ‘systematic or widespread’ attack occurred).^{Footnote 16} Cherif Bassiouni, who chaired the drafting committee at the Rome Conference (leading to the adoption of the Rome Statute), argued that Art. 7(2)(a) ‘clearly refers to state policy, and the words ‘organizational policy’ (…) do not refer to the policy of an organization, but the policy of a state. It does not refer to non-state actors’.^{Footnote 17} The debate is still ongoing among more recent contributions^{Footnote 18} and some voices suggest that the policy requirement can be inferred from the systematic nature of the crimes, as we shall later see.

Over the last two decades the proliferation of courts adjudicating CAH (international, ad hoc or national) have significantly specified the core elements of the definition. Rather than providing a comprehensive overview of the case law, at this stage I want to highlight one crucial standard that the ICC has identified to further specify both the contextual and the policy elements of the definition (E2). As far the contextual element is concerned (the ‘widespread or systematic’ attack), the ICC held in the Bemba case that the attack must be ‘massive, frequent, carried out collectively with considerable seriousness and directed against a multiplicity of victims’.^{Footnote 19} More importantly, it held that the attack must be ‘made by groups of persons who govern a specific territory or by any organization with the capability to commit a widespread or systematic attack against any civilian population’.^{Footnote 20} In other words, the perpetrator’s identity is defined neither by its formal nature (state or non-state) nor limited to the capacity to commit a multiplicity of odious acts (murder, rape, deportation, enslavement, etc.), but by the capacity to control an entire territory (hence the people) on which the attack eventually occurs. This capacity clearly makes the agent of CAH akin to state or state-like entities.

Interestingly, the existence of a ‘State or organizational policy’ has also been derived from the distinctive capacities of the agent. To determine whether a group qualifies as an organization within the meaning of Art. 7(2)(a), the ICC held that it has to determine ‘whether the group possesses, in fact, the means to carry out a widespread or systematic attack against a civilian population’.^{Footnote 21} This standard was also taken by the International Criminal Tribunal for the Former Yugoslavia (hereafter, the ICTY) in Tadic, where the ad hoc tribunal held that ‘if the acts occur on a widespread or systematic basis that demonstrates a policy to commit those acts, whether formalized or not and can be deduced from the way in which the acts occur’.^{Footnote 22} In addition, the ICTY made clear contra Bassiouni that the agent neither needs to be a state agent nor even a state-sponsored agent; it can just be tolerated by the state.^{Footnote 23} The same tribunal also held in the Kordic case that the policy element does not need to be explicit or definite but that it remains an ‘important indicative factor’.^{Footnote 24} In the more recent Kenya situation, the ICC held that the existence of a policy should still play a role and should be derived from a series of particular events, among which are found ‘the establishment and implementation of autonomous political structures at any level of authority in a given territory (…)’ and ‘the establishment and implementation of autonomous military structures (…)’.^{Footnote 25} Here again, it appears that the class of agents that satisfy the clause is distinguished not just by the horrendous acts committed but by the wide range of coercive action(s) that they (have the capacity to) undertake on a given territory and its people. This is also confirmed in a recent study of the ICC’s case law: ‘by interpreting the organizational requirement in its context, the chamber found that it is the capacities—meaning the means, resources, and co-ordination capabilities—of the group that are decisive’.^{Footnote 26}

The intermediate conclusion to be drawn is that if CAH do not necessarily imply a legally recognized state, it does imply a state or state-like form of agency. The efforts of the ICC to specify the contextual and policy elements have been heavily scrutinized in the legal literature. More precisely, it has been argued that the extension of the agent of CAH to non-state actors ‘follows a tendency in the more recent international case law to downplay the significance of the contextual requirement of crimes against humanity’.^{Footnote 27} Claus Kress thereby supports the dissent opinion of the late Judge Kaul in the Kenya decision, which points to the risk of endangering the ‘demarcation line’^{Footnote 28} between international and domestic crimes. Judge Kaul suggests a more fine-grain list of standards for ‘organization’ that ‘ should partake of some characteristics of a state’^{Footnote 29}: ‘(a) a collectivity of persons; (b) which was established and acts for common purpose; (c) over a prolonged period of time; (d) which is under responsible command or adopted a certain degree of hierarchical structure, including, as a minimum, some kind of policy level; (e) with the capacity to impose the policy on its members and to sanction them; and (f) which has the capacity and means available to attack any civilian population on a large scale’.^{Footnote 30} Kress supports this restrictive view and suggests that ‘the contextual requirement of crimes against humanity reflects the wish of states that these (and others) rather heavy restrictions on their sovereignty only apply in particular instances of human rights violations’.^{Footnote 31} One may view Judge Kaul and Kress’ views as supporting each other. But while Kress ultimately relies on a positivist standard to justify this limited scope (state intent), Judge Kaul is more attached to capture a conceptual category (by listing the standards of ‘organization’). In the same vein, Robinson holds that the policy element is ‘conceptually valuable’^{Footnote 32} in that in it helps distinguishing CAH from ordinary patterns of crimes.

Finally, we should note that the nature and capacity of the agent also distinguishes CAH from other international crimes—most importantly, war crimes and genocide. The same odious acts in times of war (international or non-international conflict) would constitute war crimes (Art. 8 of the Rome Statute). CAH, in contrast, are attacks against a civilian population at the exclusion of combatants. Genocide (Art. 6 of the Rome Statute) understood as the planned extermination of a whole population on racial and ethnic grounds requires neither a widespread attack nor a state or organizational policy. As Robinson notes, ‘genocide focuses on the collective character of the victims (and hence may not formally require a policy element), whereas crimes against humanity focus on the collective character of the perpetrators (and thus require state or organizational policy)’.^{Footnote 33} However, to set ‘a higher threshold of associative activity and effort to coordinate crimes’^{Footnote 34} again does not yet explain why this form of association is worthy of international criminal concern. It is factually agreed that CAH are committed in ‘circumstances where the victims are deprived of any prospect of protection by territorial authorities’.^{Footnote 35} What is still missing is a reflection on what makes those authorities normatively distinctive from the subjects’ standpoint. This is where one needs to turn to the normative theory of CAH.

3. The structure of CAH in the philosophical literature

Having identified the basic legal structure of CAH, I now turn to exploring how the predominant philosophical literature has covered the core elements of the definition (E1 and E2) and whether it can remedy the normative lacuna identified above. For the sake of concision, I concentrate on two predominant accounts that explore the two elements of the definition in depth, those of Massimo Renzo and David Luban. In a recent article, Renzo argues that the current structure of CAH is deeply unsatisfactory. This is because the collective qualification (E2) of the crimes is at odds with our commitment to protect human rights. More precisely, the qualification of the crimes does not fit with the human rights’ core idea to which is assigned ‘the function of protecting the dignity of each human person’.^{Footnote 36} Therefore, the very existence of CAH being contingent upon the existence of a systematic attack together with the requirements that the victims are attacked qua members of a specific group of civilians (the discriminatory intent), does not fit with the individual dimension of human rights, which are discovered independently of any social, political or legal attribute.^{Footnote 37} According to Renzo, the sole infliction of such violence and suffering upon any individual—irrespective of there being a widespread attack or a discriminatory intent—should count as a CAH: ‘each of the acts listed in the definition of crimes against humanity of the Rome Statute and other international documents constitutes a serious attack on the human dignity of the victims—one that can be said to deny their status of human being’.^{Footnote 38} Since ‘dignity’ is an inherently human and fundamental status, every single violation justifies international criminal action. An isolated act of rape committed by any individual would fall under this category: ‘any case of rape, torture, forced prostitution and so on is a crime against humanity, no matter whether committed as part of wider or systematic attack or not’.^{Footnote 39}

It quickly appears that Renzo’s approach to CAH exclusively focuses on exploring the first element in the definition (E1), which he refers to as violations of basic human rights. The threshold to distinguish those basic human rights captures ‘the conditions that are necessary for a minimally decent life’.^{Footnote 40} Arguably what exactly constitutes those conditions is subject to a vast disagreement among normative theorists. The exponential literature in human rights theory indicates how persistent the ‘threshold’ question continues to be.^{Footnote 41} Consequently, the indeterminacy in the notion of human rights is built into the notion of CAH. My critical point is more limited in that Renzo’s project radically disconnects the notion from its legal embodiment. In particular, the state or state-like nature of the agent of CAH, the resources needed to prepare and commit their preceding attack (e.g., territorial control, autonomous political and military structures, etc.), and hence the claim that CAH are confined to ‘particular instances of human rights violations’^{Footnote 42} go in an opposite direction to Renzo’s revisionist project. This should not dismiss the project itself, of course. As Chehtman rightly comments, ‘these considerations might not be fatal to his conception, but they certainly undermine its plausibility vis-à-vis our contemporary understanding of these basic concepts’.^{Footnote 43} Renzo’s suppression of one of the two core elements of the definition (E2) equally impacts on the justification of extra-territorial jurisdiction (Q2). Following Anthony Duff’s relational model of accountability (and in contrast to Larry May’s harm model),^{Footnote 44} the universal character of ‘dignity’ creates a universal moral community to which violators are to be held accountable. CAH are ‘wrongs for which we are accountable to our fellow human beings in virtue of the membership in the wider community of humanity’.^{Footnote 45} In conclusion, Renzo’s answers to both Q1 and Q2 are limited to exploring E1.

It would be blatantly false to hold that Renzo’s revisionist approach speaks for the whole philosophy of CAH. In searching for an account that addresses the deployment of CAH in law, one does much better by examining the older but still very influential account of David Luban. In contrast to Renzo, Luban’s account precisely aims to inductively reconstruct the structure of CAH based on the relevant statutes (the Nuremberg Charter, the Rome Statute) and some important judicial decisions. Among the five distinctive features of CAH inferred by Luban, the nature and capacities of the agent (E2) hold a prominent place: ‘crimes against humanity are committed by politically organized groups acting under color of policy’.^{Footnote 46} Not only does Luban acutely identify the state or state-like nature of the agent of CAH. He also points to the broader role and capacities of that agent in line with the jurisprudential findings presented above: ‘what makes [those] crimes distinctive lies in the fact that they are atrocities committed by governments and government-like organizations toward civilian groups under their jurisdiction and control’.^{Footnote 47} More than just a circumscribed list of odious acts, CAH denotes a larger pathological state of affairs that Luban terms ‘politics gone cancerous’.^{Footnote 48} This pathology captures the perversion of the role of the state in its basic protective mandate. While the state is established to protect its subjects from external threats, CAH captures a situation in which the state’s internal resources are massively exploited to attack and atrociously harm those same subjects: ‘the legal category of ‘crimes against humanity’ recognizes the special danger that governments, which are supposed to protect the people who live in their territory, will instead murder them (…)’.^{Footnote 49} The inversion of the role of the state is also acutely grasped by Vernon in another important article: the attacked group of civilians ‘is absolutely worse off than it could be in the worst-case scenario of statelessness’.^{Footnote 50}

Schematically, Luban’s and Renzo’s approaches diverge over which element of the definition (E1 or E2) makes CAH distinctive. However, I argue that this opposition significantly dissipates when one moves from the question of the core structure of CAH (Q1) to the justification of extra-territorial jurisdiction (Q2). As we have briefly seen, Renzo adopts a relational approach to criminal accountability (championed by Anthony Duff) according to which being accountable implies being ‘answerable’ for something in virtue of her occupying a certain role or holding a certain status. The identification of this role is contingent upon what Renzo takes as distinctive of those crimes, namely the violation of the ‘dignity’ of human beings by any of odious acts listed in the definition (E1). As this status is independent of any social or political affiliation, the relevant community to which wrongdoers account becomes the community of humanity: ‘crimes against humanity are those that properly concern the whole of humanity, where this means that they are wrongs for which we are responsible (i.e., accountable) to the whole of humanity’.^{Footnote 51} Although states would still be responsible for the prosecution of those crimes (for practical reasons), any murder or rape committed anywhere in the world for any kind of reason becomes a CAH.

If one were to apply the relational model of criminal accountability to Luban’s account of CAH, it follows that the agent of CAH is called to account in virtue of the distinctive role it plays—a role rightly identified as the one of state or state-like entities (E2). But Luban’s answer to Q2 does not adopt this logic. This is because Luban is skeptical toward the very idea of states having moral interests of their own. To better understand Luban’s view, it is necessary to address another central feature of CAH listed by Luban, namely that ‘crimes against humanity are inflicted on victims based on their group membership rather than their individual characteristics’.^{Footnote 52} This feature leads Luban to capture in more substantive terms the target of the perpetrators’ attack, which Luban metaphorically defines as the ‘political animal’. The political animal refers to our need to live in and belong to a particular group. Since CAH necessarily imply discriminatory intent, they attack ‘humanness’: ‘living in some group or another is a human inevitability, and attacking people on the basis of their group affiliation violates humanity (…)’.^{Footnote 53} On this account, the reliance upon ‘humanity’ does not lead to extending the scope of the crimes (as in Renzo). It rather identifies the very first reasons why subjects establish a state-like institution—reasons that should apply to everyone.

In other words, from inferring the categorical agent of CAH (E2), Luban moves to providing an independent account of what the label ‘humanity’ stands for—in analogy to Renzo’s ‘dignity’. This move has been noted in the literature. As Chehtman puts it in his review, ‘Luban explicitly admits that he came up with these features by proceeding inductively from the various relevant statutes and decisions. However, these features seem to have a deeper meaning in his conception than simple abstractions or generalizations about the law’.^{Footnote 54} If this ‘deeper’ investigation amounts to reconstructing the fundamental reasons why human beings enter the civil condition, then it should be welcome. As suggested in the last section, the legal literature points to a lacuna when it comes to explain what is normatively cogent about state-like entities. What is problematic is how Luban extends the relevant jurisdiction of CAH (Q2) to any ‘vigilante jurisdiction’. Luban holds that

The thought is that we are all potential victims of CAH because we are all in need of state- or state-like protection (we are all ‘political animals’). Therefore, we all have an interest in repressing CAH and any ‘vigilante jurisdiction’ can prosecute those crimes provided that it respects the procedural tenets of ‘natural justice’.^{Footnote 56} For Luban, the fact that states or international tribunals conventionally have jurisdiction follows from the assumption that they can guarantee those procedural principles: ‘the primary reason is to protect the accused against mistakes and not because states have a direct interest in punishment’.^{Footnote 57} For this extension of extra-territorial jurisdiction (Q2) to hold states and vigilantes must be constructed as equivalent. They are equivalent in the normative sense that they both embody the universal interest in state or state-like protection (derived from ‘political animality’). As Luban puts it, ‘putting perpetrators on trial means putting their politics on trial’.^{Footnote 58} And they are equivalent in the qualified sense that they both meet the standards of natural justice. But what this equivalence entails is that states are just the surrogates of their subjects. In a more recent contribution, Luban specifies that ‘perpetrators of infamies offend not against state interests but human interests—the perpetrators are indeed ‘enemies of all mankind’—and that states prosecuting infamies under UCJ act merely as surrogates’.^{Footnote 59} This assumption is puzzling. Presumably, if states really have no ‘direct interest in punishment’^{Footnote 60} and in fact have no moral interest of their own, then the Preamble’s reference to the ‘most serious crimes of concern to the international community’ does not confer any existence to such community. The fact that the jurisdiction for CAH falls upon state or state-derived entities (the ICC, an ad hoc tribunal or a national court) specifically reinforces this problem. Alternatively the established role of the agent of CAH (state or state-like entities) in the case law and the entity legally responsible (state or state-derived entities) indicates a striking similarity of status. Normatively, it invites one to adopt Duff’s relational account and construct a relation of criminal responsibility in which a normative community calls their members to account. I develop this approach in the next sections.

4. Concentrating on the preparatory conditions of CAH

In the last section, I concluded that despite strong divergences over which element of the legal definition (E1 or E2) should count as distinctive of CAH. Renzo and Luban converge when it comes to identify the relevant jurisdiction of CAH (Q2), namely humanity. This overlapping conclusion is reached via a different route, though. In Renzo’s case, it is the emphasis on the odious acts and the identification of ‘dignity’ as the inherent status or value violated by them that creates the universal moral community to which wrongdoers have to answer. In Luban’s case, it is the universal interest in ‘healthy politics’ and the strong skepticism toward states having moral interests that triggers the jurisdiction of any ‘vigilante’ embodying ‘humanity’. In this section and the next, I aim to further test those predominant answers to Q1 and Q2 against the law of CAH. In this section more specifically, I first return to the case law and identify what I call the ‘preparatory conditions’ of CAH, which denote the concrete organizational steps and institutional patterns established for the ‘widespread or systematic’ attack preceding the odious acts to take place.

To introduce the PCs of CAH, let me start by briefly returning to Luban’s illuminating idea of ‘politics gone horribly wrong’. CAH epitomize cases in which the state’s internal resources are used to atrociously attacked harm the subjects they are supposed to protect. An important step of this process (required by Art. 7(1)) is that the odious acts are part of a ‘widespread or systematic’ attack. The ICC takes this attack as indicative of a ‘State or organizational policy’ (required by Art. 7(2)). What I call the PCs of CAH denote the broader institutional conditions that the agent of CAH has established in preparation of and/or in conducting this attack. Luban rightly explains that the attack ‘suggests [that] something dynamic, something moving and on-going—a persecution in the process of getting worse (…)’^{Footnote 61} but he does not elaborate on this formative process and does not explore the subjects’ internal point of view in depth. My study of the PCs aims to fill this gap. I distinguish below three interrelated PCs before drawing some important implications for the normative theory of CAH:

5. Refining the structure and revisiting the relevant jurisdiction of CAH

Having presented a typology of the PCs of CAH, I need now to explain how this legal-empirical inquiry can provide two benefits for the purpose of theorizing the notion of CAH. First, such inquiry helps building a more accurate account of the prototypical conditions conducive to the odious acts (in legal-empirical terms) and thereby refines the structure of those crimes (Q1). Those conditions denote the systematic and persecutory exercise of physical control (or coercion) over an entire civilian population with the help of extra-ordinary (including lethal) resources (material and human) by an agent enjoying de facto authority over a large portion of territory in the name of a particular policy or ideology. The range of coercive actions this agent may undertake in those conditions is enormous. This disposition has been mentioned (but not developed) in the literature. In the words of Vernon, the agent of CAH ‘can apply unilateral force; it can prevent you from escaping it; it can prevent outsiders from coming to your aid; it can induce many of your fellow citizens to take part in your persecution or at least to stand by ineffectually, out of deference to authority’.^{Footnote 75} My legal-empirical inquiry concretely and precisely typifies the PCs that conduct to the odious acts and therefore fills the missing gap between the two core elements of the definition (E1 and E2). There is not simply coincidence or correlation between the systematic control and the odious acts. The very possibility of the latter depends on the former.

But refining a chain of events does not suffice to articulate the normative significance of CAH. If the odious acts are inseparable from the PCs (spatio-temporally, causally and intentionally), then one should explain what is normatively distinctive about this sequence from the subjects’ standpoint. The core idea of ‘role inversion’ has been brilliantly ascertained by the Luban-Vernon view. Vernon argues that the triad of ‘administrative capacity, local authority and territoriality’^{Footnote 76} distinguishes the travesty of states when they attack their subjects: ‘when, therefore, they play an essential role in an attack on a group of a state’s subjects, that group is absolutely worse off than it could be in the worst-case scenario of statelessness’.^{Footnote 77} From the subjects’ standpoint, however, if one just considers the odious acts then the idea of role inversion loses its significance as ‘to be enslaved or tortured or killed by a state is not very different from suffering the same thing at the hands of a sadist or serial killer’.^{Footnote 78} We therefore need to more precisely explain where and how the perversion occurs from a subject-based perspective (in addition to the odious acts, of course). Reconstructing the PCs better explain from this standpoint what it is about states that those crimes deeply pervert. This specification however requires saying a bit more about what states stand for in the first place.

Most of us constantly live under the authority of states that de facto enjoy a monopoly of institutional coercion. Most of us presume, however, that the systematic patterns that state authorities establish with regard to this monopoly serve our collective interest in physical security. If we have to constantly evaluate the soundness of the state’s actions in this basic domain, the very purpose of having a state would be in question. Instead of trying to coordinate our efforts on our own and decide what we each need to do, we establish a coordinating authority that helps us comply with the reasons we all already have. The authority then establishes its legitimacy by systematically improving our conformity to those reasons. This view is owed to Joseph Raz and his Normal Justification Thesis (NJT). For an authority to be legitimate is to give its subjects pre-emptive and content-independent reasons for action. Those reasons are pre-emptive in that the authority’s directive replaces all the reasons that count against the action. And those reasons are content-independent because there is no ‘direct connection between the reason and the action for which it is a reason’.^{Footnote 79} The directive is followed without the subject evaluating the soundness of the directive. This is again because subjects better comply with the reasons that apply to them if they accept the directive as authoritative.^{Footnote 80} If any, political authorities must provide this service with regard to the basic domain of institutional coercion over an entire civilian population—the domain that the PCs are specifically concerned with.

I suggest adopting the Razian concept of authority to reconstruct the internal point of view of the subjects placed under the PCs of CAH. Viewing the PCs through this lens allows us to capture the extent to which the agent of CAH perverts the subjects’ relation to political authority. The PCs strikingly mirror the systematic and pre-emptive role of the state (conditions a) to c) above). The agent holds and is able to maintain a monopoly of institutional coercion over an entire civilian population—a population over which it must have the ‘capacity to promulgate an ideology (…)’.^{Footnote 81} The agent thereby not only holds a de facto authority but also makes a claim to legitimate political authority (before, during, and after the odious acts). But while the agent plays the same systematic role with the same pre-emptive intention with regard to the same basic domain, this role serves to persecute, terrorize and finally odiously attack its subjects. Subjects need to evaluate every single directive issued to them while at the same time being at the mercy of this highly coercive agent. This applies for instance when victims of CAH are selected based simply on a perception that they form part of a targeted group. This is seen for instance in Bosco Ntaganga, in which it is reported that ‘the UPC/FPLC adopted an organisational policy to attack civilians perceived to be non-Hema’.^{Footnote 82} Similarly, in the Kenya situation the ICC held that ‘as a whole, the evidence shows that the criterion used by ‘the perpetrators to identify and attack their victims was essentially their perceived political affiliation with the PNU’.^{Footnote 83} In this sense, the PCs epitomize the perversion of the authoritative role of states.

Second, does this view also suggest a different answer to Q2? In a nutshell, the predominant rationale (in Renzo and Luban) is that the accused perpetrators of CAH are answerable to ‘humanity’ despite the disagreement about the relevant agent of those crimes. In Renzo’s account, this extension obtains by identifying a substantive and universal feature of human beings that CAH reach—a purely ‘victim-based’ approach—which also extends the scope of the notion itself. In contrast, Luban does circumscribe the agent of CAH to state or state-like entities but extends the relevant jurisdiction to any ‘vigilante jurisdiction’ provided that it meets the procedural standards of ‘natural justice’. The normative basis for this extension is simply that we all share an interest in ‘healthy politics’. We are all political animals in need of state or state-like protection. But this normative basis cannot hold when one examines how the vigilante establishes its authority vis-à-vis the accused perpetrator(s) of CAH. The question of authority has been at the center of the recent literature. In the words of Duff,

In others words, establishing the relevant jurisdiction of CAH first implies identifying a normative community to which wrongdoers are ‘answerable’. This approach requires viewing the criminal procedure as doing justice not only to the victims, but also to the perpetrators. Rather than straightforwardly depict the accused perpetrators of CAH as ‘everyone’s legitimate enemy’ in Luban’s vein, it emphasizes the defendant’s status as responsible member of a distinctively political community—as Renzo does with regard to dignity. In this sense, I suggest taking the law of CAH as community- or polity- creator. The premises of this community are explicit in the Rome Statute where the drafters refer to the ‘most serious crimes of concern to the international community’ (Preamble). Yet the predominant approaches to CAH do not explore the law further to identify what could possibly form this distinctively inter-state community. Renzo favours the community of humanity by revising the definition itself. In turn, Luban cannot assign any meaning to this community because states cannot have moral interests of their own. Finally, Duff examines the idea that the ICC’s authority obtains on behalf of the political community in which these crimes were committed—as ‘delegated jurisdiction’^{Footnote 85}—but rapidly dismisses it. This is because in universal jurisdiction procedures triggered by national courts or the Security Council, there is no delegation being formally authorized by the state in which CAH occur.^{Footnote 86} In all cases, however, the relevant community is determined independently of the notion’s deployment in law.

This is precisely where I want to stop and place the conclusions I reached in refining my answer to Q1. By cataloging, documenting and interpreting the PCs of CAH, I have shown that the directives the agent imposes on its subjects patently make a claim to legitimate political authority before atrociously attacking them. I have inferred that in examining potential agents of CAH, international criminal courts identify that claim as distinctive of the agent committing the odious acts. I have also concluded that the agent perverts the reason-giving role of the state. Arguably this refinement reinforces the view that even in ‘unauthorized’ cases, the perpetrators become ‘answerable’ to the normative community of states by having played that particular role thanks to which and in the name of which the odious acts are committed. I therefore suggest viewing extra-territorial jurisdiction as doing justice not only to the victims (by proving the crimes) but also to the accused perpetrator by treating him as responsible member of the normative community of states—in Duff’s words, ‘we respond to his wrongdoing, however terrible and ‘inhuman’ it was, not by simply destroying him, but by trying to bring him to answer for it’.^{Footnote 87} This account of CAH informed by the PCs can thereby better explain why state or state-derived entities (the ICC, an ad hoc tribunal or a national court) can take action against accused perpetrators of CAH. When states in which CAH occur are unwilling or unable to prosecute and punish, the community calls its responsible members to account and organizes an extra-territorial trial in the name of their common membership. This marks the evolution of the normative aspiration of this inter-state community from the exclusive right to rule (external sovereignty) on which public international law was traditionally founded.