Jump to ContentJump to Main Navigation
Targeted KillingsLaw and Morality in an Asymmetrical World$

Claire Finkelstein, Jens David Ohlin, and Andrew Altman

Print publication date: 2012

Print ISBN-13: 9780199646470

Published to Oxford Scholarship Online: September 2012

DOI: 10.1093/acprof:oso/9780199646470.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy). Subscriber: null; date: 12 December 2018



Targeted Killings

Jens David Ohlin

Oxford University Press

Abstract and Keywords

This chapter investigates the tension between national security and civil liberties through a distinctive conceptual framework: what linking principle can be used to connect the targeted individual with the collective group that represents the security threat? The chapter is organized as follows. Section II explains and defends this methodology by demonstrating that no account of targeted killing — whether sounding in jus in bello or jus ad bellum — can be complete without making explicit reference to a linking principle. Section III then proceeds to catalogue five major linking principles — taken from different domains of law including the use of force, international humanitarian law, and criminal law — that could potentially serve that function: direct participation, co-belligerency, membership, control, and complicity/conspiracy. Section IV concludes with a comparative evaluation of the linking principles that exposes their strengths and weaknesses.

Keywords:   national security, civil liberties, security threats, targeted killing, linking principle, direct participation, co-belligerency, membership, control, complicity

I. Introduction

The current debate about targeted killings has revolved around the central divide between jus ad bellum and jus in bello. Either the launching of a drone strike is considered a defensive use of force to be evaluated under the traditional rules of self-defense under Article 51 of the UN Charter, or the drone strike is to be evaluated under the rules of warfare codified in international humanitarian law.1 The prohibition against the killing of civilians is of particular concern here. Of course, the two issues are not mutually exclusive. One can coherently claim that drone strikes satisfy the demands of jus ad bellum but fail to live up to the requirements of jus in bello, and are therefore illegal.2 The reverse is possible as well. One might conclude that targeted killings do not run afoul of international humanitarian law (IHL) but violate the core ad bellum prohibition against the unlawful use of force codified in the UN Charter. These are all logical permutations of the argument.

At a conceptual level, international law is deeply conflicted about how to handle targeted killings; the issue falls between the state-based system of public international law and the individualized system of domestic criminal law. The former contemplates armed conflicts between combatants who open themselves up to the reciprocal risk of killing; the latter contemplates killings in self-defense only when (p.61) the traditional progression of arrest, trial, and punishment is unavailable. Because the terrorist is a non-state actor who falls between these two categories, the current law has had difficulty not only providing a positive rule regarding the legality of targeted killings, but also definitively choosing the correct paradigm. Even the application of traditional rules of IHL to the activity remains contested, since such an application presupposes that one paradigm has been selected over the other.3 It may even be the case that no positive rule of customary international law has crystallized to govern the practice.4Assuming, arguendo, that some form of targeted killing is permissible in some situations, a central and deeply contested question remains: who can be targeted and why? The selection of paradigms again structures our natural intuitions about the answer. Those concerned with national security are inclined to view the question through the lens of the laws of war, where all bona fide combatants are assumed to be targetable with lethal force. Those concerned with civil liberties are inclined to view the question through the lens of the criminal law (or domestic law more generally), where a judge or jury determines outcomes based on a rigorous fact-finding process, and where capture and punishment—not killing—is the default norm. The question of targeting straddles the tension between national security and civil liberties and it is unclear how it can (or should) be resolved.

This chapter investigates the tension between national security and civil liberties through a distinctive conceptual framework: what linking principle can be used to connect the targeted individual with the collective group that represents the security threat? Section II will explain and defend this methodology by demonstrating that no account of targeted killing—whether sounding in jus in bello or jus ad bellum—can be complete without making explicit reference to a linking principle. Section III will then proceed to catalog five major linking principles—taken from different domains of law including the use of force, international humanitarian law, and criminal law—that could potentially serve that function: direct participation, co-belligerency, membership, control, and complicity/conspiracy. Section IV will then conclude with a comparative evaluation of the linking principles that exposes their strengths and weaknesses.

The resulting conclusion will be counter-intuitive to readers accustomed to the standard positions in the literature. Although one would think that criminal law principles, with their strict adherence to conduct rules and culpability, would result in the greatest maximization of civil liberties, this intuition is not realized once the criminal law principles are divorced from their traditional legal process: (p.62) the courtroom. The question of who can be targeted (and the individual's relationship to the collective) requires a more nuanced response, one that uses the legal concepts developed for the law of war, but properly reformulated to take into account the realities of asymmetrical warfare with non-state terrorist organizations. The legal concepts developed for use in criminal trials provide false comfort that one is respecting civil liberties, but ironically they offer fewer protections. In the end, reformulated and redefined law of war principles, with their reliance on status concepts and proxies such as membership, do the job better because the concepts are comparatively more public, transparent, and self-administering than their competitors in the criminal law.

II. The problem of linking

Regardless of which paradigm is selected, there is inevitably a deep conceptual puzzle that straddles both sides of the fundamental divide between jus ad bellum and jus in bello. In both cases, it is unlikely that the single individual who is targeted—in isolation—satisfies the demands of either argument. The individual must be linked to a larger collective—a larger belligerent force—that explains the relevancy of the single individual. This linking requirement is a function of both the jus ad bellum and jus in bello analyses; for example, one cannot simply avoid the linking issue by switching from jus ad bellum to jus in bello or vice versa.

Within the context of jus ad bellum, the traditional argument for a drone attack relies on the international doctrine of self-defense, recognized in Article 51 of the UN Charter but also certainly recognized in customary law as well as the just war tradition.5 The United States has argued publicly that its drone attacks in Yemen, Pakistan, and Afghanistan are supported by the doctrine of self-defense.6 However, under any version of the principle of self-defense—whether expounded by public international lawyers or legal philosophers—the target of the defensive counter-attack must constitute a threat to the United States or its allies.7 The underlying threat makes the defensive force “necessary”—a universally recognized constraint (p.63) on the use of force in self-defense under either basic principles of criminal law or international law.8 The notion that self-defense is a necessary response to a threat is part of the universal structure of self-defense arguments in any legal or moral context.9

My point here is not to advocate for any particular version of what constitutes a “threat”—nor what makes a defensive response to it “necessary.” These are sticky theoretical questions that form the center of most debates about self-defense. Rather, the issue I want to explore is one level deeper. Regardless of one's assessment of what constitutes a threat to a state's interests—territorial integrity, political independence, etc—it is unlikely that a single individual, by himself or herself, can constitute a threat against a state. It is theoretically possible to imagine a hypothetical terrorist who works alone, secretly plotting a devastating attack against a state by procuring weapons and then deploying them without any assistance whatsoever. The Unabomber is one such example, and it is the exception that proves the rule.10

The more common situation involves the existence of a terrorist organization or militia that constitutes a threat by plotting and implementing terrorist or military attacks against a particular state. In such cases, the collective constitutes the threat against the national interest, thus generating the right of self-defense. Furthermore, the individual stands in a certain relationship with the collective, either by belonging to the terrorist organization, contributing to the collective endeavor, or some other mode of participation in the collective group.11 For the moment we must postpone consideration of which linking principle is most appropriate. The point here is simply that individuals acting alone almost never constitute a national threat. Within the War on Terror and the asymmetrical use of targeted killings against non-state actors, an even stronger conclusion is warranted: single individuals never constitute a threat to the United States. The threat comes from organized groups with political or ideological objectives that they seek to bring about by launching attacks against civilians. This is the raison d’être of global terrorism and jihadism.

Shifting the focus to jus in bello does not relieve us of the obligation to find an appropriate linking principle. If terrorists are simply enemy civilians, without any (p.64) relationship to a larger collective, then no operative principle of IHL permits their summary killing.12 It is only when their relationship to a larger collective is considered that the use of force against them may be permissible. Under traditional rules of IHL, combatants may be killed to the extent that they belong to an armed fighting force that is engaged in an armed conflict with the United States.13 Indeed, it is the collective's engaging of the armed conflict with the United States that triggers the operation of the IHL norm allowing combatants to be killed. But it is an open question whether IHL recognizes the existence of an armed conflict with a non-state actor, and whether this is best described as an international armed conflict triggering the Geneva Conventions, a non-international armed conflict triggering Common Article 3 of the same, or neither, thus generating conflict regarding the appropriate default rule in the absence of any governing Geneva Convention regime.14

In this context, there are multiple problems associated with linking an individual to the larger terrorist organization that is engaged in an armed conflict with the United States. First, the United States is currently engaged in an armed conflict (international or non-international) with Al Qaeda, but the individuals targeted by US drones may or may not be card-carrying members of Al Qaeda.15 Indeed, although Al Qaeda may once have been a defined and tightly knit organization controlled by Osama bin Laden, the organization has morphed into an amorphous network of terrorist organizations operating under the common banner of Al Qaeda.16 In rare instances, various local terrorist organizations operating under the name Al Qaeda may share operational or financial support from their parent organization, and may even respond to hierarchical commands issued by bin Laden himself or his commanders.

In most cases, however, terrorist organizations operating under the banner of Al Qaeda in some form are part of a much looser confederacy of co-sympathetic jihadists who share common inspiration and rhetoric without sharing a common command (p.65) structure or operational command.17 They are distinct terrorist organizations linked together by a common cause. It is therefore unclear if the existence of an armed conflict with one Al Qaeda organization can translate into an armed conflict with another sympathetic Al Qaeda organization.18 In some instances, both organizations may be sufficiently well developed that each, on its own terms, meets the appropriate standard for being engaged in an armed conflict with the United States. In other contexts, however, the over-arching umbrella between the organizations may be crucial for our legal determination of an armed conflict with the United States. This is particularly true in cases where one terrorist organization is well developed and clearly engaged in an armed conflict, but the second organization is a nascent and burgeoning endeavor that has not yet launched significant attacks.

III. Five possible linking principles

The preceding analysis suggests that both the jus ad bellum and the jus in bello analyses suffer from a deeper confusion about how to relate the individual terrorist with the larger collective. Attacking the problem in this manner will help expose the deeper question of how to integrate the non-state actor—and the individual terrorist—into the inherently collective nature of public international law and the laws of war that arise from it. We should therefore consider all of the possible linking principles and consider which best describes the particular role and function of the individual terrorist. The possible linking principles include: direct participation in an armed conflict, military membership, co-belligerency, control, complicity, and conspiracy.19 A comparative evaluation of the linking principles will cut across the jus ad bellumjus in bello divide.

(a) Direct participation in an armed conflict

Under a standard jus in bello analysis, civilians are generally protected from the reciprocal risk of killing that governs the relations of enemy soldiers.20 Obviously, though, this protection can be opportunistically exploited by civilians who use their protected status to pursue attacks without subjecting themselves to reciprocal risk.21 Such a system of perfidy would create a perverse incentive: soldiers would have no incentive to identify themselves as soldiers—the only consequence of their identification would be one of exposure. Consequently, traditional rules of jus in (p.66) bello deny protected status to civilians who directly participate in the armed conflict.22 The functional justification for this rule is obvious: civilians who engage in combatancy are functionally equivalent to traditional combatants and ought to be treated similarly; that is, ought to be subject to attack. This rule is now codified in Article 51(3) of Additional Protocol 1, which states that “civilians shall enjoy the protection afforded by this section, unless and for such time as they take a direct part in hostilities.”23

The concept of “direct participation” links the individual to the collective fighting force that is engaged in hostilities. The protection is not lost simply by virtue of holding a gun.24 If the linking principle merely required the use of weapons, it would have stated that. Rather, the linking principle establishes a quasi-causal relation between the non-protected civilian and the larger armed conflict. Unfortunately, though, nobody really knows what constitutes “direct participation” in an armed conflict. The term is undefined in the Additional Protocol and there is little case law on the subject. The International Committee of the Red Cross (ICRC) notes that it is clear that the “lawfulness of an attack on a civilian depends on what exactly constitutes direct participation in hostilities and, related thereto, when direct participation begins and when it ends . . . [but] the meaning of direct participation in hostilities has not yet been clarified,” and concedes that a legal definition of the term does not even exist.25 The ICRC Commentary cites the Inter-American Commission on Human Rights for the proposition that the concept of “direct participation” in hostilities means “acts which, by their nature or purpose, are intended to cause actual harm to enemy personnel and material.”26 Although this interpretation of the concept has some intuitive appeal, it reduces it to a causal criterion—not an inherently objectionable result, although the type and closeness of causal relation is left similarly undefined.

As any good lawyer knows, the real issue is never whether causation is present or not, but rather what type of causation (but-for, proximate, etc) and whether the causation between the act in question and the desired consequence is close enough to meet the applicable standard. Many genuinely civilian actions that patriotically (p.67) support a nation's interest would eventually and predictably cause some harm to enemy personnel, but no one would ever suggest that they constitute direct participation in hostilities.27

One can imagine a spectrum of participatory acts. At one end of the spectrum are acts that unquestionably represent acts of combatancy, such as firing a weapon at the enemy. No one doubts that this constitutes direct participation. At the other end of the spectrum, one might place activities such as a civilian seamstress who sews uniforms in a civilian factory that will one day be worn by soldiers. Or consider the cook who resides far from the battlefield and makes frozen food, some of which will be sold to the military for inclusion in MREs (Meals Ready to Eat). This clearly does not rise to the level of direct participation. In the middle of the spectrum are the hard cases: the civilian contractor who repairs a tank on the battlefield, or the civilian defense department employee who helps design or deploy a new weapons system. Are these individuals directly participating in hostilities?28

One way to get a handle on direct participation is to compare it with indirect participation. The ICRC Commentary cites the Inter-American Commission on Human Rights for the proposition that “mere support” of the military effort by civilian personnel—including commercial sales and “expressing sympathy for the cause of one of the parties”—constitutes indirect participation.29 The asserted rationale for this conclusion is that these forms of participation do not involve “acts of violence which pose an immediate threat of actual harm to the adverse party.”30 The concept of immediacy appears to be doing all of the work here, though it is unclear if immediacy is as significant as the Inter-American Commission believes it to be. Similarly, the ICRC notes that a draft statute for the future International Criminal Court defined participating in hostilities to include scouting, spying, and sabotage, but excluded food deliveries and household domestic staff “in an officer's married accommodation.”31

At Nuremberg, Streicher, Goebels, and others who ran the Nazi propaganda effort were held responsible for aiding the Nazi war machine.32 Indeed, Streicher was charged with criminal responsibility for his writings, which in today's legal climate would have been described as direct and public incitement to commit genocide, in (p.68) the words of the Rome Statute.33 When NATO bombed Serb positions in order to pressure Serbia to withdraw forces from Kosovo, the targets included Serbian state television and other elements of the state's communications regime.34 Although reasonable persons can disagree over the permissibility of these attacks, I take it that the disagreement stems more from the civilian nature of the employees at the state television station, rather than the indirect nature of their causal contribution to the war effort. In many of these situations, the causal role played by the non-military civilians is quite substantial and might even be described as direct.35 Perhaps this is the reason that the US Naval Handbook simply concludes that the direct participation standard “must be judged on a case-by-case basis.”36

The ICRC's latest effort, its Interpretative Guidance on Direct Participation in Hostilities, also cashes out the concept in causal terms.37 Indeed, according to the ICRC, the word “direct” in the legal standard explicitly refers to direct causation as opposed to indirect causation.38 According to the ICRC's metaphysics, a direct causal result implies that the “harm in question must be brought about in one causal step.”39 In applying this standard, the ICRC Interpretative Guidance concludes that building or maintaining the fighting capacity of one party to the conflict is not sufficiently direct because it is a two-step process. Even recruitment of combatants and their military training are excluded because they are two-step processes.40 Temporal and geographic proximity may imply causal proximity, but they do not wholly determine it, since an action could (in theory) directly cause a particular harm far removed in time and space.41 (p.69)

The direct participation standard is difficult to apply to terrorists, and there is currently little uniform state practice that would shed light on the content of the alleged customary norm. On the one hand, some nations take a purely causal approach to the notion, whereby any civilian who contributes to the armed conflict loses protected status. For example, India believes that any person who “contributes towards the furtherance of armed conflict” is no longer a protected civilian.42 On the other hand, some countries conclude that “persons who merely provided support to the enemy . . . for example those who supplied it with weapons, food or medicine,” do not lose their protected status.43 In between, some nations recognize the inherent ambiguity and lack of clarity in the standard. For example, Israeli practice notes that the carrying of arms is not a sufficient condition for losing protected status, since in many locations (for example, Lebanon), civilians routinely carry firearms even though they have nothing to do with the hostilities, though the Israel report notes that “when returning fire, it is extremely difficult (and probably unwise from a military viewpoint) to differentiate between those individuals actually firing their firearms and those just carrying them.”44

The ambiguity becomes starker when one considers another linking principle that is often applied to terrorists: providing material support to terrorists. The United States considers this to be a war crime and a violation of both federal and international law.45 Does providing material support for terrorism constitute direct participation in hostilities? Did Hamdan “directly participate” in the hostilities because he was driving Osama bin Laden?46 The thing about providing material support is that it rests squarely on the shoulders of a causal contribution to the larger effort. If the individual's actions make a terrorist attack more likely—for example, if he aids or abets the larger effort—then the individual has provided material support to terrorism.47 Consequently, providing financial support or engaging in advocacy on behalf of a terrorist cause can constitute material support, since terrorist activities require far more than just brute operational support.48 Many other forms of support are required to bring a terrorist plan to fruition. But providing financial support or ideological advocacy is a far cry from a direct participation in hostilities. What is missing is not a causal link, but the right kind of causal link. (p.70)

Although everyone agrees that direct participation requires the right kind of causal link, distinguishing between a direct and indirect causal contribution is far from easy. The “one-step” view espoused by the ICRC Interpretative Guidance appears to boil down to the idea that the causal contribution must be operational and on the battlefield, while indirect contributions emanate from beyond the confines of battlefield activity as they have been traditionally defined.49 But this is not so obvious.50 Directness appeals to the closeness of the causal route, which may or may not accord with a battlefield movement. It is, for example, possible to envision a close financial connection as well as a remote battlefield connection. Each of these possibilities puts pressure on our intuition that the concept of directness correlates essentially with prototypical battlefield activity.51 In other words, the closeness of the causal connection and the shape of the causal route can slip apart. An individual might engage in activity that has only a remote bearing on the hostilities (for example, bearing a weapon when there is no enemy in sight), but the relation between the action and the hostilities can be seen in a straight line. In contrast, an individual might engage in activity that has a strong correlation with the hostilities (for example, transporting a crucial weapon that will change the tide of the battle), but the relation between the action and the hostilities involves a comparatively more circuitous route. At first glance, it is not clear whether the causal element of the direct participation standard ought to be understood with regard to closeness or shape.52

(b) Co-belligerency under the law of neutrality

Another solution to the linking problem is to employ the doctrine of co-belligerency from the well-traveled law of neutrality.53 Under this doctrine, states (p.71) engaged in an international armed conflict are allowed to consider third-party states as co-belligerents of the enemy and thus subject to attack. However, third-party states must first be given the opportunity to declare their neutrality in the conflict, and only if they refuse to remain neutral can they be declared co-belligerents of the enemy and thereby subject to lawful attack.54 The application of this doctrine can be quite controversial, in particular whether a state can feign neutrality and yet offer limited assistance to an ally and remain free from attack.55 This can be referred to as a form of benevolent neutrality, or the idea that a state may “discriminate” against one side of the conflict without necessarily becoming a full co-belligerent in the conflict.56

The deeper problem with the doctrine of co-belligerency is whether it can be successfully transplanted from the original state-based system of public international law into the new realm of non-state actors like Al Qaeda. Bradley and Goldsmith have argued that terrorists who are “co-belligerents” of Al Qaeda are by extension engaged in an armed conflict with the United States by virtue of their status as co-belligerents.57 However, in Al-Bihani, a U.S. federal court rejected application of the doctrine to the war against Al Qaeda, concluding that the doctrine was rooted in traditional public international law notions of state sovereignty and that any “attempt to apply the rules of co-belligerency to such a force would be folly, akin to this court ascribing powers of national sovereignty to a local chapter of the Freemasons.”58

Indeed, the law of neutrality is based on the idea that states have a duty to declare themselves either officially neutral in a conflict or throw their lot in with one side of the conflict over the other—thus sharing the advantages of victory but also sharing the burdens of defeat. In the words of Francis Lieber, they advance and retrograde together.59 The problem is that irregular fighting forces are not similarly situated with their enemies in an analogous fashion to states within the global Westphalian system.60 All states in the Westphalian system enjoy the sovereignty associated with the formal equality of nation-states; one expression of this sovereignty is the ability to form strategic alliances, declare war, engage in armed conflict, sign peace treaties, and return to peaceful relations with an enemy (p.72) state. Non-state actors are neither sovereign entities nor do they enjoy the capacities that flow directly from this sovereignty. Nonetheless, Bradley and Goldsmith have argued that the U.S. president is permitted to target individual terrorists who are co-belligerents of Al Qaeda.61 The invocation of the concept of co-belligerency allows them to connect the individual terrorist with a fighting force that is currently engaged in an armed conflict with the United States. They invoke this rationale to demonstrate that such targeted killings comply with the congressional authorization that was provided to the president in the Authorization for the Use of Military Force (AUMF) passed after the September 11 attacks.62

The concept of co-belligerency is built around the notion that combatants fighting against a common enemy—even if they are not fighting on a unified front—can be linked together simply by virtue of their common enemy. The old adage that the enemy of my enemy is my friend best expresses the principle. Simply by virtue of standing in the common relationship of belligerency against the same enemy, two entities become co-belligerents.

The key thing to remember about the doctrine of co-belligerency, as it exists in the law of neutrality, is that it is built around the notion of publicity. Co-belligerents are not defined simply around their actions on the battlefield. Rather, third-party states must be allowed the opportunity to publicly declare their neutrality in the conflict, and only if they forgo this opportunity may they be labeled co-belligerents and subject to attack. This publicity criterion works well for sovereign entities such as states that are capable of exercising foreign relations. It is less clear how this translates into the domain of individual terrorists who are defined as co-belligerents of Al Qaeda. They are not given the formal opportunity to declare their neutrality, nor are they given a conventional form of notice that they are being declared a co-belligerent of Al Qaeda, except in the generic sense that the United States has publicly declared that all militants are subject to attack unless they foreswear allegiance to Al Qaeda or the Taliban. But this certainly does not meet the formal requirements of the law of neutrality, nor does it capture its underlying spirit of publicity.

(c) Military membership

The traditional rules of IHL implicitly rely on a principle of membership in order to link an individual combatant with a larger fighting force. The basic criteria for the fighting force—the wearing of a military uniform, the display of a fixed emblem recognizable at a distance, the carrying of arms openly—defines the collective fighting force as a military organization that deserves the protection of (p.73) IHL.63 However, the basic criteria also help define the individuals who belong to the organization. Determining membership is based on the fact that individuals in the military wear uniforms, display fixed emblems, and carry their arms openly (to the extent that they use weapons); this in turn publicly signals to the world that the individual is part of the fighting force.

Membership is important because it provides a public criterion that is comparatively easy to establish.64 The link is established simply by virtue of signing up with the military, being drafted, or donning a uniform. No deeper investigation is required. Indeed, it does not even matter if the combatant actually engages in combatancy. His status as a combatant is established simply by virtue of his joining the military organization, regardless of whether he actually fires his weapon and kills an enemy soldier.65 The link is easily administered, public, and clear for both sides of a conflict (and even third parties) to identify the relevant individuals. So there is comparatively little ambiguity about membership in a military organization.

Unfortunately, membership in a terrorist organization does not demonstrate any of the hallmarks that IHL typically assigns to membership in a military organization.66 Terrorists do not wear uniforms or display fixed emblems, nor do they carry arms openly.67 Perfidy and deception are essential tools that allow the terrorist to complete his deadly craft. It may be the case that membership in a terrorist organization may have other essential attributes, but they are undeniably not the same attributes that IHL assigns to military organizations.68 The standard IHL categories were specifically designed to link the individual soldier with warring collectives that are the traditional subjects of public international law (that (p.74) is, nation-states), and to provide a first gloss on Lieber's assumption that individual soldiers are linked to the collective such that they advance and retrograde together. With these criteria, however, the terrorist remains in limbo.

(i) Form vs. function

One might solve this problem by moving from a formal concept of membership to a functional concept of membership.69 Formal membership is built around formal indicia such as membership lists, the wearing of uniforms, and de jure requirements of domestic law, while the functional concept of membership can be determined by the individual's role and function within the organization.70 For the functional definition of membership, it is particularly relevant whether the individual received and carried out orders from the organization's hierarchy.71 The application of the formal concept of membership, with its emphasis on de jure considerations, may not map onto the “the more informal and fluctuating membership structures of irregularly constituted armed forces fighting on behalf of State and non-State belligerents.”72

In contrast, the functional version of the concept takes that informal structure as given and determines membership based on the individual's place within—and relationship to—that hierarchy, even if that hierarchy is nebulous, irregular, or constantly shifting. The result is a version of the membership concept that can actually be applied to terrorist organizations, even if they are ill-defined and lack the same rigorous structure of state military organizations. Although the functional concept of membership is far less public and transparent than the formal concept of membership, it retains the essential characteristics of a membership criterion insofar as it is nominally based on an individual's status as a member of a terrorist organization.

(d) Control

One might connect an individual terrorist with Al Qaeda—and the armed conflict between Al Qaeda and the United States—with a control test. Under this view, the individual is linked to the collective if Al Qaeda “controls” the actions of the individual. This principle has its genesis in public international law and the standard that the International Court of Justice (ICJ) imposed in the Nicaragua case to determine whether the actions of an armed group could be attributed to a state (p.75) for purposes of assigning state responsibility for the group's actions.73 The court concluded that state responsibility existed in cases of effective control of the group's actions. In Nicaragua, the United States was found not be in control of the contras because, although the US was found to be involved in “planning, direction and support” of the contras’ paramilitary activities, there was insufficient evidence that the United States “directed or enforced the perpetuation of the acts contrary to human rights and humanitarian law alleged by the applicant State.”74

There are other versions of the control principle. The International Criminal Tribunal for the former Yugoslavia (ICTY) famously rejected the ICJ's effective control test and formulated a different standard based on overall control.75 Under this new standard, control by the state requires more than mere financing or providing military equipment, but the standard stops short of the strict standard imposed by the ICJ. The overall control standard is met by the planning and supervision of military activities in general, without requiring that the planning or oversight extend down to the level of specific military attacks.76 A more general level of planning or supervision can constitute overall control of the paramilitary organization even in the absence of specifically directing the organization's military operations.

The problem with borrowing either of these control principles and applying them to the War on Terror is that many of the individuals who are targeted by the Administration are not controlled by Al Qaeda, even under the looser version of the standard articulated by the ICTY. In some cases, to be sure, the individual's activities may indeed be directed by Al Qaeda. In other situations, however, the individual will be affiliated with a regional terrorist organization with very loose ties to the Al Qaeda parent group. Originally, Al Qaeda represented a defined organization with specific individuals committed to a particular political objective. But the organization has now transformed into a looser confederation of like-minded fellow travelers, many of whom are fighting separate armed conflicts in different regions of the globe. These conflicts include different enemies, different objectives, and different techniques, though they might share an overarching ideological commitment to violent jihadism. Consequently, in many situations, the parent organization may provide ideological and rhetorical support but no direct or even general operational control over the local terrorist organization.

One solution to this problem is to redefine the armed conflict as not against Al Qaeda per se but rather the long list of more local organizations that are engaged (p.76) in terrorist activities.77 This might alleviate the need to use the control principle in the first place, but the strategy can only be imperfectly applied. To the extent that a pre-existing local organization is involved in a bona fide armed conflict with the United States, the strategy works. However, many of these sub-groups might be so localized that they could not be said to be engaged in a declared armed conflict with the United States. Furthermore, some of these local groups might be so loosely organized that even the local group does not “control”—either effectively or overall—the actions of the individual terrorist.

(e) Complicity and conspiracy

Another solution is to import the doctrine of complicity from the domain of criminal law as a way of linking the individual terrorist to a larger group engaged in armed conflict with the United States. The doctrine of complicity implicitly relies on a causal notion, in the sense that complicity liability is generated by an individual's contribution (or attempted contribution) to a criminal endeavor, just as long as the contribution makes the completion of the crime more likely.78 This broad notion of complicity has increasingly been used as a paradigm to understand an individual's contribution to a national collective endeavor of war-making.79 The importation of a criminal law notion into the domain of public international law may, at first glance, appear strange, but the concept's intuitive appeal is undeniable. At first glance, the only difference between the classical criminal law situation and the situation of a national armed struggle is the size of the collective endeavor to which the contribution is made.80 The other side of the equation—the individual, as well as his relationship to the collective—remains the same. Furthermore, the case under consideration here (the individual contributing to the collective terrorist organization) stands in between the classical criminal law paradigm and the state-based paradigm of international conflicts inherent in public international law. This broad notion of complicity in a collective endeavor is also encoded in Article 25(3)(d) of the Rome Statute, which scholars have interpreted as criminalizing a form of residual complicity in a collective criminal endeavor.81 Although terrorism is not a discrete international crime under the Rome Statute, the mode of liability codified (p.77) in Article 25(3)(d) represents a similar invocation of the concept of complicity in group action. The federal crime of providing material support for terrorism is also built around the notion of complicity.82

The causal element of criminal complicity picks up quite nicely the causal interpretation of directly participating in hostilities. Under this view, it makes sense to target individual terrorists who are complicit in the larger collective conflict (whether one defines the conflict as a criminal conflict or a war) because complicity represents a form of participation. In criminal law, this point is purely definitional; complicity is defined as a form of participation in criminal wrongdoing.83 A party to an armed conflict has every reason to target an individual whose actions contribute to—or were aimed at contributing to—their eventual defeat.

The question, however, is whether the causal element of criminal complicity is sufficiently direct as a linking principle to adequately serve as a gloss on the notion of directly participating in hostilities. Indeed, criminal law scholars often describe aiders and abettors—and other form of accomplices—as having engaged in a form of indirect commission of the crime.84 True, at least some accomplices could be described as direct participants in the endeavor, but the criterion of complicity is notoriously broad and meant to capture a wider scope of participation that plays some causal role in the criminal endeavor, even if that causal role is somewhat attenuated. Even in criminal law, though, the causal role cannot be too attenuated; otherwise criminal liability is usually denied as inappropriate. But even still, the criminal law notion may capture a whole host of individuals whose indirect contributions to the endeavor make them criminally culpable (and hence subject to punishment) but perhaps not subject to the immediate and summary killing implicit in traditional combatancy under the standard rules of IHL.

One might attempt to tighten the complicity link by switching to the concept of conspiracy.85 Conspiracy as a mode of liability is arguably stricter than complicity, because it requires an underlying agreement between the individual and the associated individuals.86 As applied to the terrorist, he would be linked to the terrorist organization because he has jointly agreed with other terrorists to pursue an armed struggle against the United States. Individuals who merely contribute to the cause, (p.78) without an underlying agreement for joint action, would not be linked to the collective under the conspiracy doctrine.87

It makes sense to view terrorism through the lens of conspiracy. Terrorists pursue an unlawful objective through conspiratorial means: agreeing to a course of action, collective pursuit of common goals, secret and underground deliberations.88 Moreover, the entire rationale of the conspiracy doctrine was to create an inchoate offence of preparation for criminality that allows the authorities to intervene quickly in a burgeoning criminal endeavor. Whatever public policy rationale exists for intervening in domestic criminal conspiracies applies with equal or greater force to transnational conspiracies to commit acts of terrorism.

Having sketched out the terrain of possible linking principles, our task is now to evaluate their comparative strengths and weaknesses, both from the perspective of positive law (for example, support in treaty or customary law) and as compliance with the underlying normative principles of international law. That being said, this investigation cannot prejudge the correct paradigm; that is, whether the most appropriate normative principles are those underlying the law enforcement paradigm or the law of war paradigm, or a combination of both. Section IV will pursue this goal by pursuing a comparative evaluation of the linking principles.

IV. A comparative evaluation of the linking principles

When can an individual be linked to a collective group for the purposes of being selected for a targeted killing? A comparative analysis of the linking principles reveals that an individual can be linked either through status alone or by virtue of a more discrete action. So membership in a military organization, by virtue of wearing a uniform or displaying a fixed symbol, confers a status on the individual that links him to the collective fighting force. Similarly, the concept of co-belligerency from the law of neutrality involves a status-like element by virtue of a belligerent's refusal to declare itself neutral in a conflict.

It should come as no surprise that IHL relies on the linking principle of membership in a military organization, given how much is at stake. If individuals are linked for purposes of IHL, they gain the privilege of combatant immunity as well as opening themselves to the risk of reciprocal killing. Individuals who meet these criteria know that they meet these criteria, and moreover, their enemies know this as well. In fact, the public nature of the linking principle is internal to the principle itself, because the link is built around the criteria of uniforms, fixed emblems, and weaponry—all of which are designed to publicly convey to one's (p.79) enemy that the linking principle is fulfilled. When so much is at stake, it makes sense for the linking principle to be self-publicizing and self-applying.

In contrast, the criminal law notions of conspiracy and complicity are causal criteria that are far less public. The individual's actions that link him to the collective are hardly public at all, because the actions of the terrorist are usually conducted covertly, far from the prying eyes of the enemy. Terrorists are more like spies than traditional combatants. Furthermore, the criteria for conspiracy or complicity are usually complicated and require the testing and fact-finding process that dominates the criminal trial. Allowing criminal law concepts to function as a linking principle cuts against the underlying nature of IHL, which necessarily relies on easy-to-administer criteria in the absence of a judicial system.

In light of this insight, section IV(a) will reconsider the virtues of membership as a linking principle, even though criminal law scholars have given it a bad name. Section IV(b) will then consider an updated version of the membership concept—the continuous combat function—that avoids many of the anxieties that criminal law scholars have about membership principles. Finally, section IV(c) will compare status and conduct principles and demonstrate that membership principles can be modified into a “functional membership” concept that represents a hybrid between status and conduct. The result is a legally defensible and philosophically coherent principle to link suspected terrorists with the non-state organizations that are fighting the United States.

(a) Rethinking membership

We are therefore caught between two types of linking principles. The traditional IHL linking principles are both self-applying and public. The traditional criminal law linking principles are neither self-applying nor public, since they require a comparatively larger degree of fact-finding to determine if their standards are met. At which end of the spectrum should we place targeted killings? Should targeted individuals be linked with the underlying principles of IHL or the criminal law?

Functionally, targeted killings are much closer to the summary killings that are inherent to IHL on the battlefield. Although the criminal law concepts of conspiracy and complicity cast a wide net, this looseness is mitigated by the fact that the criminal law system affords defendants a chance to contest the causal linkage before a neutral decision-maker.89 No such right exists on the battlefield, which is precisely why the linking principles used by IHL are much narrower.90 Although many individuals might be causally responsible for helping the war (p.80) effort, the rules of IHL limit automatic killing to soldiers in uniform (and civilians directly participating in hostilities). Although this classification might be seriously limited, the whole structure of IHL is built around the notion that the reciprocal risk of killing should be underbroad rather than overbroad, precisely because there is no opportunity to contest a determination on the battlefield. The uniformed soldier on the battlefield cannot complain that he was killed before he could contest his status, because he was wearing a uniform.

(i) A functional equivalent

Targeted killings represent the same kind of summary killing that traditional combatants face on the battlefield. While conspiracy and complicity are strict enough for a system with a criminal process, they are not appropriate for summary execution outside of the judicial process. This suggests that however we link individuals to a collective for purposes of targeted killing, it ought to be with a linking principle that is closer to the IHL linking principles rather than criminal law linking principles. The correct linking principle would represent a functional equivalent to the IHL linking principle that governs the targeting of traditional combatants. The difference would be that the functional equivalent ought to be tailored for the specifics of the situation: a non-state group composed of individuals who pursue terrorism without a uniform.

Although it is difficult to sketch out the exact contours of this hypothetical linking principle, it ought to lie somewhere between the doctrine of co-belligerency and membership in a military organization. The doctrine of co-belligerency, as understood by the law of neutrality, has the advantage that it is based on both publicity and self-declared consent; the co-belligerent nation publicly refuses to affirm its neutrality and is therefore declared a co-belligerent. The very same publicity and self-declared consent is performed by the individual soldier who dons a uniform. Both are then subject to summary attack under the laws of war, though one norm flows from jus ad bellum and the other flows from jus in bello. But the structure of both is remarkably similar.

The functional equivalent in cases of targeted killings would link the individual to the collective terrorist group if the individual is a card-carrying member of a terrorist organization or a self-declared enemy of the United States.91 Membership might be established in a number of ways, not simply by attending an Al Qaeda training camp.92 We are therefore left with the following linking principle: (p.81) voluntary membership in an organization engaged in an armed conflict with the United States.93 This linking principle might at first glance sound too narrow, because terrorists might opportunistically avoid declaring their allegiances in order to avoid being targeted—an example of lawfare to be sure. But the anxiety is misplaced. The very concept of terrorism hinges on publicity—publicity for a cause and a political objective, neither of which can be easily disowned without doing damage to the theater of violence implicit in terrorist attacks.94

(ii) The transitory requirement

This conclusion is more than just normative-philosophical. It is also a legal conclusion, in the sense that it can be understood as a gloss on the concept of direct participation in hostilities, the original requirement of jus in bello that explains when a civilian loses his or her protected status under IHL. On this point, one might object that this understanding—direct participation in hostilities in terms of self-declared membership in an organization engaged in an armed conflict with the United States—conflicts with another aspect of the “direct participation” linking principle. The Additional Protocol withdraws protection from civilians “for such time” as they are directly participating in hostilities.95 The flexible and temporal work performed by the concept of “for such time” suggests that the associated status (protected civilian vs. unprotected combatant) shifts constantly depending on the actions of the particular individual. He can fall in and out of protection at each moment in time, depending on his conduct—without a reified status that endures throughout the individual's existence. This approach was famously discussed by the Israeli Supreme Court in its Targeted Killings decision.96

Is this transitory requirement of the Additional Protocol consistent with membership in an organization engaged in an armed conflict with the United States? Or is the latter far too status-oriented—that is, not sufficiently transitory and flexible—to accord with the “for such time” standard?97 It strikes me that the notion of self-declared membership is, in fact, consistent with the transitory nature of the “for such time” standard. Individuals join and leave organizations all the time—just as (p.82) they join and leave criminal conspiracies—and such decisions are both legally and morally significant. The individual terrorist is subject to the risk of being killed “for such time” as he is a member of Al Qaeda, though he regains the core protections of IHL if and when he permanently leaves Al Qaeda. At that moment in time he becomes a subject of the criminal process again. This solution avoids some of the most perverse aspects of the revolving door problem that is, the risk that terrorists will launch terrorist attacks but fall back into civilian status to shield themselves from the enemy.98 If the “for such time” criterion is linked to membership in the organization, such opportunistic shifts are dramatically more difficult.

(b) The continuous combat function standard

This membership principle is arguably what the ICRC was getting at in its Interpretative Guidance on the Notion of Direct Participation, which explicitly recognized the significance of engaging in a continuous combat function.99 According to the Interpretative Guidance, membership in an armed group of a non-state party to a non-international armed conflict depends on whether the individual engages in a “continuous combat function.”100 The point of introducing the new continuous combat function criterion is to distinguish between, on the one hand, “members of the organized fighting forces of a non-State party from civilians who directly participate in hostilities on a merely spontaneous, sporadic, or unorganized basis, or who assume exclusively political, administrative or other non-combat functions.”101 The functional consequence of this distinction is to carve out a category under IHL that treats soldiers in a non-state military organization in analogous fashion (for example, according to membership) to soldiers in a more traditional state-party military organization.

How is this distinction to be made? An individual is deemed to be engaged in a continuous combat function, as opposed to the more transitory and fleeting direct participation in hostilities, if their “continuous function involves the preparation, execution, or command of acts or operations amounting to direct participation in hostilities are assuming a continuous combat function. An individual recruited, trained and equipped by such a group to continuously and directly participate in hostilities on its behalf can be considered to assume a continuous combat function even before he or she first carries out a hostile (p.83) act.”102 If one applies this standard to Al Qaeda, there is a plausible argument that these terrorists are trained to continuously operate as terrorists with the goal of pursuing attacks against the United States and its allies. Moreover, there is a lasting integration of the individual into the collective, on whose behalf the individual is acting.103 Although many of these members have not yet finalized an attack, they are engaged in the process of preparing, planning, or training for an attack. Their status as Al Qaeda terrorists therefore makes them subject to military attack.

The ICRC standard of engaging in a continuous combat function was (and remains) highly controversial when it was adopted by the Red Cross working group.104 Some scholars disapproved of the membership-oriented nature of the concept and believed that the concept of direct participation in hostilities ought to remain transitory and based solely on the actions of the individual at each moment in time.105 Furthermore, these scholars rejected the rationale that armed groups of a non-state party to an armed conflict ought to have a functional analogue to membership in a state's military organization.106 On the other hand, other scholars, including some who participated in the ICRC working group that developed the continuous combat function standard, criticized the proposal from the opposite direction; that is, sacrificing the principle of military necessity for the principle of humanity.107 These criticisms were a natural outgrowth of a pre-existing anxiety about how IHL treats organized armed groups differently depending on whether they are a state party or not. Members of a non-state armed organization receive the added protection of the “for such time” limitation (and are consequently immune from targeting part of the time), while members of a state party's military organization are subject to attack purely on the basis of membership.108 Why should members of a non-state armed organization receive more protection under the customary rules of IHL, rather than less?

The continuous combat function standard was meant to be a solution to that problem. In fact, the ICRC Interpretative Guidelines apply the continuous combat function criterion both to non-international armed conflicts and international (p.84) armed conflicts, such that membership is limited to those individuals who display a continuous combat function as opposed to those who, like reservists, have a combat function that is “spontaneous, sporadic, or temporary” or “assume exclusively political, administrative or other non-combat functions”109 The problem with the ICRC's particular proposal is that it did not go far enough. According to at least some scholars, the requirement set up a different legal regime that provided an unfair and unwarranted advantage to insurgent groups.110 Only members of an organized armed group who evidence a continuous combat function could be lawfully targeted; all other members of the group can only be targeted for such time as they are directly participating in hostilities. By contrast, all members of a state's military apparatus are subject to lawfully targeting, even a cook, regardless of whether they are directly participating in hostilities or not.111 From the point of view of this criticism, the proper remedy is to normalize the standard across all armed groups, whether state actors or non-state actors. In other words, membership in both domains could be limited to those who display a continuous combat function or, in the alternative, membership in both domains could be expanded to all individuals and include the proverbial cook in both the state military and the insurgent group, so as to eliminate the unfair advantage conferred on the insurgents.112 This the Red Cross proposal does not do.

However, even if one sticks with the Red Cross proposal and applies the continuous combat function requirement just to insurgents, it may be the case that some insurgent groups are so entirely focused on planning and perpetrating military attacks that every member of the group is engaged in a continuous combat function.113 (p.85) The U.S. administration has taken a similar view in habeas corpus proceedings in federal court arising out of Guantanamo Bay detentions.114 According to the Obama Administration, Al Qaeda is a military organization through-and-through, such that all members of the group are dedicated to planning, supporting, or executing future attacks in some way or another.115 Unlike other insurgent armed groups that also perform some political or civilian functions (for example, Hamas in Gaza or the Taliban in Afghanistan),116 Al Qaeda exists solely to plot terrorist attacks against designated targets; it has no positive political program of its own, nor does it aspire to directly control territory through the operation of an Al Qaeda syndicate government. Is it therefore possible that all members of Al Qaeda and similar groups are engaged in a continuous combat function in some way or another?

(c) Status rules vs. conduct rules

Whether one accepts this argument or not, the real point is to emphasize that the entire discussion of the continuous combat function requirement takes place within the general context of membership as a linking principle. As good criminal law scholars, we are supposed to favor conduct rules over outcomes based on status alone. As criminal law professors, we assign our students Martin v. State and drive home the proposition that the principle of culpability requires that we punish individuals solely for their blameworthy actions, not their status.117 This argument is particularly relevant for the War on Terror, where the government arguably uses status to determine who should be declared an unlawful combatant, interned at Guantanamo Bay, tried before a military commission, or even summarily killed by a drone attack.118 To some critics, this represents an unwarranted infringement on civil liberties in order to protect national security. Under this view, if draconian consequences are required to protect our nation, they should only be visited upon an individual suspect if he has engaged in (p.86) proscribed conduct. Anything less represents a fundamental betrayal of the civil liberties enshrined in our constitutional structure.

However, the interplay between conduct and status is rich and complex and not so black and white. Status is often a shortcut for a history of repeated conduct, such that the status of being a drug addict or the status of appearing drunk in public are both, with limited exceptions,119 the product of component actions (consuming alcohol or drugs) that we would naturally classify as conduct. Similarly, the building blocks of IHL demonstrate a complex relationship between conduct and status. Although membership in a military organization is usually described as a status, once one inquires about how this status is determined, one learns that the component requirements are wearing a uniform, the display of a fixed emblem recognizable at a distance, and the carrying of arms openly—all examples of conduct par excellence.120 It is rare, then, to have a case of status all the way down.

(i) Functional membership as a hybrid concept

This is even more true when one considers a functional version of the concept of membership, which looks to the individual's relationship to an organizational hierarchy and whether he receives and carries out orders from that command structure.121 Unlike a formal version of membership, which relies more heavily on status criteria, the functional concept is half way along the road to a conduct rule. It relies on the status concept of membership but cashes out that standard by reference to what the individual is actually doing—not necessarily at each discrete moment in time, but rather from the broader perspective of a longer time period: taking orders from commanders, engaging in military operations at the behest of commanders, etc.122 In fact, one might describe the functional version of membership as a hybrid concept that straddles the distinction between status and conduct—an appropriate (p.87) result for the context of terrorist organizations and other irregular armed groups. The result is hardly Solomonic; rather, it merges the best of both worlds.

That being said, it would be an exaggeration to say that the distinction between conduct and status is wholly illusory. There is a fundamental difference between them, albeit one that is often obscured. A status usually represents a proxy for lower-level conduct. Proxies usually get a bad name in both law and philosophy, because it is natural to presume that if the lower-level facts generate the moral or legal significance, one ought to eliminate the higher-level proxy and deal exclusively with the lower-level elements. Under this view, the identification of a proxy suggests eliminativism as the proper course of action. This is a hasty conclusion because one ought to distinguish between crude proxies and successful proxies. Crude proxies take a rough set of intuitions and create a shortcut that obscures the real significance of the underlying elements; what is gained in administrability and convenience is outweighed by the loss of accuracy.123 By contrast, successful proxies link together a diverse set of lower-level elements, solve evidentiary problems, and help root out inconsistencies.124 The question is whether the status concept under consideration in this chapter—membership in a terrorist organization engaged in a self-declared armed conflict—is the former or the latter.

There is a plausible argument that the status concept that we have deployed here illuminates more than it obscures. First, it has obvious evidentiary value. Self-declared membership in an organized armed group is public and transparent; those who join a group dedicated to jihad can understand the position of conflict that they have placed themselves in. Second, third parties can monitor compliance with this norm with relative ease. By contrast, limiting targetability based on the conduct of the targeted individual at each cardinal moment in time is comparatively less transparent and very difficult for third parties to monitor. These are precisely the considerations that originally sparked the use of status concepts such as membership in traditional IHL norms.125 (p.88)

(ii) Preserving civil liberties

We are left, then, with a somewhat surprising result. The traditional dichotomy of national security vs. civil liberties turns out to be illusory.126 When viewed through the lens of domestic criminal law, the use of status concepts appears to threaten the principle of culpability and suggests that the proposed scheme impermissibility infringes civil liberties. But when viewed through the lens of IHL, the use of status concepts reveals itself to be entirely consistent with the conceptual structure of IHL—a structure that is based largely on status concepts, and for good reason. To insist yet again that pure conduct alone should determine targetability is to import criminal law linking principles into a legal terrain—the battlefield—where the preferred linking principles are publicly observable and self-administering status concepts such as membership. Moreover, shifting to a hybrid status-conduct concept such as functional membership goes even further towards ensuring that truly innocent civilians fall outside the scope of legitimate targets.

How could this standard be administered? One might object that it is difficult—if not impossible—to prove that any given individual is truly a member of a terrorist organization engaged in an armed conflict with the United States. After a targeted killing, who is to say that the killing did not live up to this standard? There are two important answers here. Such problems of proof are endemic to all IHL norms governing civilians, and the current problem will be comparatively easier to administer when compared against a more transitory revolving door scheme. Second, the concept of joining and leaving a criminal organization is well worked out in the literature and case law on conspiracies, which in some jurisdictions imposes stringent requirements on individuals seeking to leave a criminal organization and escape the consequences of their membership.127 These standards sometimes require a public repudiation of the enterprise—either to the leaders of the enterprise or to the relevant authorities.128 This is a high standard to meet, and appropriately so in the case of domestic criminal law.129

Applied to terrorists, the standard would require a public declaration repudiating the armed conflict against the United States before they could regain their protected (p.89) status.130 It is unlikely that any jihadist terrorist would opportunistically exploit this standard in order to falsely gain protected status. Even despite this fact, however, there are strong reasons to defend a modified standard for abandonment. Given that the criteria for membership is our previously identified hybrid concept of functional membership, abandonment or renunciation would be demonstrated by the continued non-existence, for a sustained period of time, of the very factors that led to the finding of functional membership in the first instance. If, for example, the individual no longer receives and carries out orders from the command hierarchy, this would necessarily entail that the individual is no longer a functional member of the terrorist organization. With this caveat, then, the hybrid concept should offer bona fide comfort to civil libertarians committed to conduct rules.

V. Conclusion

This new standard has the virtue that it avoids the “revolving door” problem noted by Justice Barak in the Israeli Supreme Court decision. In fact, the standard is more permanent than the transitory standard offered by Justice Barak, yet it is not so permanent that it runs afoul of the “for such time” requirement of the Additional Protocol. The linking principle is easy to administer, self-applying, and based on semi-public criteria, which makes it a functional equivalent to being a member of a military organization. True, this new linking principle is not as easy to administer as the traditional IHL linking principle of being a member of a military organization, but it is certainly easier to apply than the criminal law notions of conspiracy and complicity that require intensive fact-based determinations by a neutral decision-maker. The linking principle is consistent with the underlying legal principles embedded in the laws of war, as well as the legal instruments that codify them. Although the linking principle may not be as permissive as some governments would wish, it is better to utilize a narrow linking principle that is legally and philosophically justified, rather than a looser linking principle that cannot be justified.


(1) See,Mary Ellen O'Connell, “Unlawful Killing with Combat Drones: A Case Study of Pakistan, 2004–2009” in Simon Bronitt ed., Shooting to Kill: The Law Governing Lethal Force in Context (Hart Publishing, forthcoming)Report of the Special Rapporteur on Extrajudicial, Summary or Arbitrary Execution: Study on Targeted Killingsdelivered to the Human Rights Council

(2) Cf.Jordan J. Paust, “Self-Defense Targetings of Non-State Actors and Permissibility of U.S. Use of Drones in Pakistan,” 19 J. Transnational Law and Policy (2010) 237

(3) Gabriella Blum and Philip Heymann, “Law and Policy of Targeted Killing,” 1 Harvard National Security Journal (2010) 145

(4) S.S. Lotus (France v. Turkey), 1927 P.C.I.J. (ser. A) No. 10, at 18

(5) Compare Legal Consequences of the Construction of a Wall in the Occupied Palestinian Territory, Advisory Opinion, (Advisory Opinion on the Wall) 2004 I.C.J. 136, 189, 194 (July 9, 2004) (no international right of self-defense against non-state actors), with Armed Activities on the Territory of the Congo (Dem. Rep. Congo v. Uganda), 2005 I.C.J. 168, 222–6 (December 19). See also Mary Ellen O'Connell, “The Legal Case Against the Global War on Terror,” 36 Case Western Reserve Journal of International Law (2004) 349.

(6) See Harold H. Koh, U.S. Department of State, The Obama Administration and International Law, Annual Meeting of the American Society of International Law, Washington, DC (March 25, 2010), available at 〈http://www.state.gov/s/l/releases/remarks/139119.htm〉 accessed November 4, 2011.

(7) Ian Brownlie, Principles of Public International Law, 7th edn (Oxford University Press, 2008) 732–3

(8) Ibid. at 734Caroline

(9) G.P Fletcher and J.D. Ohlin, Defending Humanity: When Force is Justified and Why (Oxford University Press, 2008) 91–6

(10) See,George P. Fletcher, “The Indefinable Concept of Terrorism,” 4 J. International Criminal Justice (2006) 894, 907–08

(11) Christopher Kutz, Complicity: Ethics and Law for a Collective Age (Cambridge University Press, 2000)

(12) Dieter Fleck, The Handbook of International Humanitarian Law, 2nd edn (Oxford University Press, 2008) 46

(13) Ibid. at 82. See also Richard Murphy and Afsheen John Radsan, “Due Process and Targeted Killing of Terrorists,” 32 Cardozo Law Review (2009) 405, 416.

(14) See, e.g., Hamdan v. Rumsfeld, 548 U.S. 557, 630 (2006) (conflict with Al Qaeda is a non-international armed conflict falling under Common Article 3). For a discussion, see D. Glazier, “Full and Fair by What Measure?: Identifying the International Law Regulating Military Commission Procedure,” 24 Boston University International Law Journal (2006) 55, 60 (“Recognizing that the terrorism conflict does not fit particularly well with traditional classifications of either ‘international’ or ‘non-international’ armed conflict, it concludes that this war is instead best defined as ‘transnational’.”)

(15) See O'Connell, supra n. 1, 10–11.

(16) Manooher Mofidi and Amy E. Eckert, “‘Unlawful Combatants’ or “‘Prisoners of War’: The Law and Politics of Labels,” 36 Cornell International Law Journal (2003) 59, 82

(17) ,United States v. Mustafa, 406 Fed. Appx. 526 (2nd Cir. 2011)

(18) Curtis A. Bradley and Jack L. Goldsmith, “Congressional Authorization and the War on Terrorism,” 118 Harvard Law Review (2005) 2047, 2112

(19) The list of linking principles is not meant to be exhaustive, but rather to include a representative cross-section of the relevant types.

(20) See Fleck, supra n. 12, 96–7, 237–8.

(22) See International Committee of the Red Cross, Customary International Humanitarian Law (Cambridge University Press, 2005) vol. I, 19–24 (hereafter cited as ICRC Commentary).

(23) See Protocol Additional to the Geneva Conventions of August 12, 1949, and Relating to the Protection of Victims of International Armed Conflicts (Additional Protocol I), adopted June 8, 1977, art. 51(3), 1125 U.N.T.S. 3.

(24) Even civilians retain the right of individual self-defense, which might be one reason to retain small arms even in a conflict zone. This complicates the ascription of combatancy to individuals carrying weapons—a particular problem during the recent fighting in Libya. See, e.g., Thom Shanker and Charlie Savage, “NATO Warns Libyan Rebels Against Attacking Civilians,” New York Times (March 31, 2011).

(25) See ICRC Commentary, supra n. 22, vol. 1, 21.

(27) Cf.Michael Walzer, Just and Unjust Wars, 3d edn (Basic Books, 2000) 146

(28) The United States Naval Handbook states that guards, lookouts, and intelligence acts all meet the direct participation standard. See ICRC Commentary, supra n. 22, vol. 1, 22.

(30) Ibid

(31) ICRC Commentary, supra n. 22, vol. II, 116.

(32) Reifenstahl might also be included in that list, though she was never prosecuted for her films.

(33) SeeJudgment, Streicher, “International Military Tribunal at Nuremberg; D.F. Orentlicher, Criminalizing Hate Speech in the Crucible of Trial: Prosecutor v. Nahimana,” 21 American University Int'l L. Rev. (2006) 557, 582–3

(34) Final Report to the Prosecutor by the Committee Established to Review the NATO Bombing Campaign Against the Federal Republic of Yugoslavia, 39 I.L.M. 1257 (2000).

(35) Cf.Sandoz, Swinarski and Zimmerman, Commentary on the Additional Protocols (Martinus Nijhoff Publishers, 1987) 619

(36) See ICRC CommentarysupraSeeibid., vol. II, 117

(37) ICRC Interpretative Guidance (International Committee of the Red Cross, 2009), 1019 (requiring a “direct causal link between a specific act and the harm likely to result either from that act, or from a coordinated military operation of which that act constitutes an integral part.”).

(39) Ibid

(42) Ibid.

(44) Ibid. at vol. II, 120–1See alsosupra

(45) 18 U.S.C. §§ 2339A and 2339B.

(46) Cf.George P. Fletcher, “On the Crimes Subject to Prosecution in Military Commissions,” 5 Journal of International Criminal Justice (2007) 39

(47) Ibid

(48) Holder v. Humanitarian Law Project, 130 S. Ct. 2705, 2720 (2010)

(49) See ICRC Interpretative Guidance, supra n. 37, 1021 (defending one-step causal criterion over allegedly wider alternatives such as “materially facilitating harm”).

(50) The one-step view of causation was controversial among the ICRC working group members. Compare, e.g., Michael N. Schmitt, “Deconstructing Direct Participation in Hostilities: The Constitutive Elements”, 42 N.Y.U. Journal of International Law & Policy (2010) 697, 727, with Melzer, “Keeping the Balance,” supra n. 67, 865–8 (defending one-step causal relation). In particular, Melzer concludes that Schmitt's more permissive definition of causation amounts to an “unlimited causal chain” that would extend as far downstream as the causal relation extends, including individuals who design, manufacture, and store weaponry. Ibid. at 868. Melzer concludes that although this wide causal criterion would be appropriate for ex post determination of criminal responsibility, it is inappropriate for an ex ante determination of combatancy under the direct participation standard. Ibid.

(51) The ICRC Interpretive Guidance, supra n. 37, 1022, goes part of the way to understanding this issue by noting that the concept of directness must be understood within the context of the collective nature of the hostilities, such that individual actions may produce little causation on their own, but when aggregated together, contribute to the collective hostilities. However, even the notion of collective hostilities does not resolve the tension between directness and shape of the causal route.

(52) The laity's common-sense understanding of the concept of directness arguably includes an ambiguity with regard to closeness vs. shape.

(53) See Fleck, supra n.12, 576–7.

(54) Ibid

(55) W. Heintschel von Heinegg, “‘Benevolent’ Third States in International Armed Conflict,” in M. Schmitt and J. Pejic, International Law and Armed Conflict: Exploring the Faultlines (Nijhoff Leiden, 2007) 543–68

(56) Ibid

(57) See Bradley and Goldsmith, supra n. 18, 2112.

(58) Al-Bihani v. Obama, 590 F.3d 866, 873 (D.C. Cir. 2010)D.C. Circuit Rejected “Co-Belligerency” in Al-Bihanihttp://opiniojuris.org/2010/10/17/dc-circuit-rejects-co-belligerency/

(59) US General Order No. 100, April 24, 1863 (the Lieber Code), art. 20.

(60) See also L. Oppenheim, International Law: A Treatise (1906) vol. 2, § 74.

(61) Bradley and Goldsmith, supra n. 18, 2113.

(62) Ibid

(63) See, e.g., Geneva Convention Relative to the Treatment of Prisoners of War of August 12, 1949, opened for signature August 12, 1949, 6 U.S.T. 3316, 75 U.N.T.S. 135.

(64) See,William Bradford, “In the Minds of Men: A Theory of Compliance with the Laws of War,” 36 Ariz. St. L.J. (2004) 1243, 1269

(65) supraProsecutor v. Halilović, ICTY Trial Chamber, No. IT-01-48-T (November 16, 2005) para. 34

(66) See, e.g., Bensayah v. Obama, 610 F.3d 718, 725 (D.C. Cir. 2010); Awad v. Obama, 608 F.3d 1, 11 (D.C. Cir. 2010) (membership in “command structure” is a sufficient but not necessary condition for legal determination that detainee is a member of Al Qaeda). For a discussion, see also John B. Bellinger III and Vijay M. Padmanabhan, “Detention Operations in Contemporary Conflicts: Four Challenges for the Geneva Conventions and Other Existing Law,” 105 AJIL (2011) 201, 220 (discussing need for workable criteria for detention of unlawful combatants based on their status).

(67) Nils Melzer, “Keeping the Balance Between Military Necessity and Humanity: A Response to the Four Critiques of the ICRC's Interpretive Guidance on the Notion of Direct Participation in Hostilities,” 42 NYU Journal of International Law & Politics (2010) 831, 843

(68) See Program on Humanitarian Policy and Conflict Research, Harvard University, “IHL and Civilian Participation in Hostilities in the OPT,” Policy Brief, October 2007, 10 (“The end of membership must be objectively communicated, posing the same intelligence problems as the affirmative disengagement approach above, especially given that many groups may not have official rosters of membership, uniforms, or centralized housing.”).

(69) See Interpretive Guidance, supra n. 37, 1005 (concluding that membership in military organizations is based on “formal integration into permanent units distinguishable by uniforms, insignia and equipment” but that membership in irregular groups requires functional criteria).

(70) Al Warafi v. Obama, 704 F. Supp. 2d 32, 38 (D.D.C. 2010)Hamlily

(71) Al Warafi v. Obama, 704 F. Supp. 2d 32, 38 (but noting that knowledge and intent is required and excluding those who “unwittingly become part of the apparatus”).

(72) See Melzer, supra n. 67, 845 (defending relevance of functional criteria for membership).

(73) Military and Paramilitary Activities In and Against NicaraguaNicaragua v. United States), June 27, 1986, ICJ Reports (1986)

(75) See Prosecutor v. Tadic, ICTY Appeals Chamber, Case No. IT-94-1-A, para. 137.

(76) Ibid

(77) The concept of the “War on Terror” represents an even wider solution, where the enemy is terrorism itself. However, this is just as nonsensical as declaring a War on War or a War on Enemies, with the opponent being defined as anyone who threatens aggressive action. This eviscerates the notion of an armed conflict against a defined enemy.

(78) Compare Sanford H. Kadish, “Complicity, Cause and Blame: A Study in the Interpretation of Doctrine,” 73 California Law Review (1985) 323, 343 and John Gardner, “Complicity and Causality,” 1 Criminal Law and Philosophy (2007) 127, with Christopher Kutz, “Causeless Complicity,” I Criminal Law And Philosophy (2007) 289.

(79) ,Christopher Kutz, “The Difference Uniforms Make: Collective Violence in Criminal Law and the Law of War,” 33 Philosophy and Public Affairs (2005) 148

(81) The provision was interpreted by the ICC in Lubanga, ¶ 337.

(82) Norman Abrams, “The Material Support Terrorism Offenses: Perspectives Derived from the (Early) Model Penal Code,” 1 Journal of National Security Law and Policy (2005) 5

(83) SeeGeorge P. Fletcher, “Complicity,” 30 Israel Law Review (1996) 140

(84) SeeSee alsoF. Jessberger, “On the Application of a Theory of Indirect Perpetration in Al Bashir: German Doctrine at The Hague?,” 6 Journal of International Criminal Justice (2008) 853

(85) Conspiracy as a mode of liability is sometimes viewed as a separate doctrine from complicity, and occasionally as a subcategory of complicity (with accomplice liability being the other subcategory). This ambiguity is immaterial for our purposes here.

(86) 18 U.S.C. §371.

(87) However, they would be guilty of providing material support.

(88) J.D. Ohlin, “Group Think: The Law of Conspiracy and Collective Reason,” 98 J. Criminal Law and Criminology (2007) 147, 201

(89) Larry May, Global Justice and Due Process (Cambridge University Press, 2011) 117

(90) Cf.Richard Murphy and Afsheen John Radsan, “Due Process and the Targeted Killing of Terrorists,” 31 Cardozo Law Review (2009) 405, 409supra

(91) In his UN report, Philip Alston denies that membership alone can be sufficient to identify a terrorist as an appropriate target for a killing. See U.N. Human Rights Council, Report of the Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions, Addendum: Study on Targeted Killings, U.N. Doc. A/HRC/14/24/Add.6 (May 28, 2010) paras 65–6 (criticizing the ICRC standard of “continuous combat function” for its overreliance on membership and other status-based concepts). For a complete discussion of the ICRC notion of the continuous combat function, see infra section IV(b).

(92) ,United States v. Hassoun, 2007 U.S. Dist. LEXIS 85684 (D. Fla. 2007)

(93) For a discussion, see Program on Humanitarian Policy and Conflict Research, Harvard University, IHL and Civilian Participation in Hostilities in the OPT, October 2007, 10 (comparing “membership approach” with “limited membership approach” that restricts targeting to fighting members of armed groups).

(94) Fletcher, supra n. 10, 909.

(95) See Additional Protocol I, supra n. 23, 174, art. 51(3).

(96) Public Committee Against Torture v. Israel (“Targeted Killings Case”), HCJ 769/02 (2005)

(97) CompareBill Boothby, “‘And for such time as’: The Time Dimension to Direct Participation in Hostilities,” 42 NYU Journal of International Law & Politics (2010) 741, 764–5supra

(98) Targeted Killings Case, supra n. 96, para. 40 (discussing problem of revolving door and citing 1 Kings 1:50 and Numbers 35:11).

(99) See Interpretative Guidance, supra n. 37, 991, 1007–9. The document's principal author was Nils Melzer, ICRC Legal Advisor, and was adopted by the Assembly of the International Committee of the Red Cross on February 26, 2009.

(101) Ibid

(104) See Mezler, supra n. 67, 831, 834.

(106) Ibid

(107) ,Kenneth Watkin, “Opportunity Lost: Organized Armed Groups and the ICRC ‘Direct Participation in the Hostilities’ Interpretive Guidance,” 42 NYU Journal of International Law & Politics (2010) 641

(108) Compare Watkin, supra n. 107, 644, with Melzer, supra n. 67, 851.

(109) ICRC Interpretative Guidance, supra n. 37, 1007.

(110) See, e.g., Michael N. Schmitt, “The Interpretative Guidance on Direct Participation in Hostilities: A Critical Analysis,” Harvard National Security Journal (2010) 5, 23. See also Adam Roberts, “The Equal Application of the Laws of War: A Principle Under Pressure,” 90 International Review of the Red Cross (2008) 931.

(111) Schmitt, supra n. 110, 23. Melzer contends that the asymmetry is justified because even cooks in a traditional army “are not only entitled, but also trained, armed, and expected to directly participate in hostilities in case of enemy contact and, therefore, also assume a continuous combat function.” See Melzer, Halilović supra n. 67, 852. The ICTY apparently disagrees. See, e.g., Prosecutor v. Halilović, No. IT-01-48-T (November 16, 2005) para. 34 (noting that while “membership of the armed forces can be a strong indication that the victim is directly participating in the hostilities, it is not an indicator which in and of itself is sufficient”). However, the only two counter-examples offered by the ICTY Trial Chamber include non-mobilized reservists and civilian police officers incorporated de jure into the armed forces by domestic statute. Ibid., para. 34 n. 78. For a brief discussion, see Ryan Goodman, “The Detention of Civilians in Armed Conflict,” 103 AJIL (2009) 48, n. 41.

(112) But see Melzer, supra n. 67, 851.

(113) For a discussion of the ambiguity in applying this criterion in these situations, compare Schmitt, supra n. 50, 727 (noting difficulty with defining “capacity-building” activities such as recruitment of suicide bombers, procurement of materials, and assembly and storage of explosives), with Melzer, supra n. 67, 865–6 (“whether an act constitutes a measure preparatory or otherwise integral to a specific hostile act or operation, or whether it remains limited to general capacity-building, must be determined separately for each case, and it is clear that the same objective criteria must apply to all civilians, regardless of whether they happen to support an unsophisticated insurgency or a technologically advanced State.”).

(114) I am indebted to Marty Lederman on this point.

(115) See, e.g., Al Bihani v. Obama, 594 F. Supp. 2d 35, 39–40 (concluding that despite petitioner's contention that he was only a cook, he was also carrying a rifle and ammunition and taking orders from an Al Qaeda military commander).

(116) See Schmitt, supra n. 110, 23 (noting that Hamas and Hezbollah have political or social wings but also concluding that “while membership in an organized armed group can be uncertain, it may also be irrefutable”).

(117) Martin v. State, 31 Ala. App. 334, 17 So. 2d 427 (Alabama 1944); Robinson v. California, 370 U.S. 660 (1962) (status of being a drug addict). But see Powell v. Texas, 392 U.S. 514 (1968) (upholding public intoxication statute).

(118) ,Hamlily v. Obama, 616 F. Supp. 2d 63, (D.D.C. 2009)

(119) There are a few examples of status categories that are not reducible to an individual's own actions, such as an infant drug addict who suffered from fetal intoxication in utero. In that case, the individual's status is causally reducible to an individual action, but it is someone else's action—the parent.

(120) ICRC Interpretative GuidancesupraIbid

(121) See supra section III(c)(i) for a complete discussion of formal vs. functional membership.

(122) See, e.g., Al-Bihani v. Obama, 590 F.3d 866, 872–3 (functional membership based on “accompanying the brigade on the battlefield, carrying a brigade-issued weapon, cooking for the unit, and retreating and surrendering under brigade orders” even “in the absence of an official membership card”).

(123) Cf.Felix S. Cohen, “Transcendental Nonsense and the Functionalist Approach,” 35 Colum. L. Rev. (1935) 809

(124) Similarly, see Jeremy Waldron, “‘Transcendental Nonsense’ and System in the Law,” 100 Colum. L. Rev. (2000) 16 (concluding that metaphysical concepts in the law provide meaningful explanations when their explanatory circle is sufficiently large). See also Jens David Ohlin, “Is the Concept of the Person Necessary for Human Rights?,” 105 Colum. L. Rev. (2005) 209 (invoking Cohen and Waldron and concluding that metaphysical concepts often link together diverse propositions to promote coherence and root out inconsistencies in doctrine).

(125) Robert Chesney and Jack Goldsmith, “Terrorism and the Convergence of Criminal and Military Detention Models,” 60 Stan. L. Rev. (2008) 1079, 1084

(126) Cf.S. Macdonald, “Why We Should Abandon the Balance Metaphor,” 15 ILSA J. International & Comparative Law (2008–2009) 95

(127) Compare Hyde v. United States, 225 U.S. 347 (1912) with United States v. U.S. Gypsum Co., 438 U.S. 422, 464–5 (1978) (“Affirmative acts inconsistent with the object of the conspiracy and communicated in a manner reasonably calculated to reach co-conspirators have generally been regarded as sufficient to establish withdrawal or abandonment.”).

(128) See, e.g., Al Ginco v. Obama, 626 F. Supp. 2d 123, 128 (D.D.C. 2009) (“prior relationship [with] al Qaeda . . . can be sufficiently vitiated by the passage of time, intervening events, or both”). See also Eldredge v. United States, 62 F.2d 449, 451 (10th Cir. 1932) (conspiracy).

(129) See, e.g., ICRC Interpretive Guidance, supra n. 37, 1008 (“In practice, the principle of distinction must be applied based on information which is practically available and can reasonably be regarded as reliable in the prevailing circumstances.”).

(130) Cf.ibid