Killer Drones, Legal Ethics, and the Inconvenient Referent

The present essay opens with a brief intervention into the debate over whether “antifoundationalist” theories of language such as deconstruction and postmodernism can be invoked to defend ethical standards of justice, rights, and political agency that are not compromised by violence, mystification, or the desire for power. Focusing on the concept/event of war in relation to the law of war, I propose to reconcile poststructuralism’s antiessentialist, contextualist approach to truth with a neo-Kantian or Habermasian pragmatics of “postmetaphysical universalism,” in which validity claims would be the result of rational, transparent, and inclusive procedures of intersubjective deliberation and empirical inquiry; as such, they would be generalizable across heterogeneous belief systems, even as they would also be recognized as historically contingent and potentially fallible forms of knowledge.¹

The second section of the essay offers a preliminary consideration of the language of humanitarian law and ethics, parsing those terms and concepts that tend to be resistant to semantic slippage from those whose denotative and referential elasticity makes them vulnerable to manipulation in certain contexts. I note that it is the unstable or indeterminate legal signifiers that have given rise to a “humanitarian” discourse in which the “lesser evil” of a particular amount of “collateral damage” is weighed against the supposedly “greater good” of a certain amount of anticipated military gain.

In the third section, I show how the logic of lesser-evil humanitarianism was used by the Obama administration to justify, normalize, and perpetuate the deployment of weaponized drones in a campaign of targeted assassinations of suspected anti-US “militants.” Here I look at government speeches and legal documents in which, I argue, the designation of “enemy combatant” is legally problematic, and in which certain fundamental and unambiguous principles of international law pertaining to the use of lethal force (e.g., necessity, imminence, and preemption) are egregiously distorted.

The fourth section reviews the debate over “collateral damage” statistics, pitting the government’s belated claims to transparency and legal-ethical propriety against the more credible findings of civil rights groups and independent organizations.

The fifth section frames the question of statistical death-ratios within a critique of the traditional ethical law of double effect, which underwrites the lesser-evil calculations of military law and which, in certain interpretive contexts, I find, turns the legal tropes of humanity, necessity, discrimination, and proportionality toward semantic, referential, and ethical aporia.

The sixth and penultimate section examines the intentionality clause of double effect, finding it to be problematic both in theory and in practice. Two factors in particular—the drone program’s deliberate and systematic use of inherently flawed targeting techniques and the psychological trauma suffered indiscriminately by populations subjected to continual drone overflight—compel us to interrogate the boundary between the intentional and the foreseeable and, correlatively, between terrorism and jus in bello.

The essay’s conclusion highlights the fetishistic structure of lesser-evil reasoning in the official discourse on civilian casualties, and returns to the issue of unprivileged belligerency, this time in the context of Donald Trump’s expansion of the CIA’s role in approving drone strikes.

Signs and Truth

Both war and the laws that would govern it are products of language, of a medium whose relation to the material world on the one hand and to meaning on the other is not always predictable. War and peace are, after all, correlated modes of interstate and intercultural (mis)communication. The present essay seeks to shed light on what is at stake for theories of just war when their real-world engagements are understood to be an effect of the dialectical tension between pragmatic agency and semiotic indeterminacy, or between performativity and ungovernability, that is latent in every act of signification. Philosophers and legal experts, however specialized their idiom, dispose of the same imperfectly transparent medium as do journalists, historians, military planners, government officials, political scientists, indeed anyone who is charged (professionally or otherwise) with using linguistic artifacts to attempt to convey something of the reality of war. To accomplish that task effectively, however, one must keep in mind the a priori symbolic structure of war itself. Modern warfare, the very referent that our vocabulary aims to describe and/or to regulate, is not only an object but also a creation of knowledge. It is an irreducible amalgam of fact and fiction, contingency and convention—phenomenal reality and “metaphors we live by.”² War is a cultural artifact, a simulacrum that precedes and informs any occurrence and any experience of “real wars,” and that any such experience will either confirm or contradict, reinforce or transform. Newspapers, films, books, and public monuments; toys, television shows, and video games; these and other media translate and communicate human experience while at the same time shaping cultural identity; they not only represent but also mediate war; they determine, that is, what the reality of a war will be, as well as what the material contingencies of a war can mean, in any particular sociopolitical ecology. The tropes of military ethics necessarily seek to ground themselves within this living hermeneutic, this intermedial space of convergence and divergence between War and wars. Not surprisingly, then, the attempt of diplomats, policymakers, and practitioners of the law to justify or to prohibit, to implement or to contain, in short, to adjudicate the violence of war, is a communicative project that places the act of mimesis under enormous stress, relying on inherently aleatory linguistic constructions to anticipate, instantiate, refer to, and manage material contexts that are fraught with uncertainty, and in which errors of articulation and reception can be (and often are) gravely consequential.

Insofar as the meta-languages of military ethics and humanitarian law exist not only to limit but also to enable warfare, they are an integral part of the organized chaos they name. On the one hand, they are illocutionary acts, disciplinary enunciations that perform (that literally make) war; on the other hand, they are mere abstractions within a closed system of differences, signs that are never fully capable of controlling either the events that they set in motion or the multiple interpretations that those events call forth. To say that war is both materially and semiotically excessive is not however to suggest that it is impenetrable to analysis, nor is it to excuse war’s continued existence as a ready-to-hand political convention. Indeed, it is not only because of but also despite the “fog of war,” not only because of but also despite semantic slippage, that the rules of war can be negotiated and revised, that the very existence of war can be challenged. It is, in other words, precisely to the extent that not only the laws of war but also the lived, subjective experience of real wars exist as abstract, semiotic constructs, as interdependent and overlapping systems of meaning, that they can be transformed, turned, one would hope, from the multiplication of pretexts and permissions for war toward the delegitimization of the organized practice of violence as a political institution.

The languages of politics and the law are committed, in principle, to a representational pragmatics of referential accuracy and historical truth. Such a project is in no way inimical to the use of tropes and metaphors, and cannot be reduced to a simple dichotomy between indirect and deceptive modes of figuration on the one hand and direct, literal, and transparent forms on the other. If we accept Lakoff and Johnson’s claim that our very processes of reasoning are themselves grounded in metaphor, it follows that not even the most rigorously logical methods of political or legal persuasion can escape metaphoricity; indeed, they depend on it. Accordingly, abuses of the formal idiom of human rights and international justice, like abuses of political rhetoric, are not the inevitable result of linguistic différance; they are enabled by, but are not endemic to, the arbitrariness of signs and the figurative nature of language.

Every act of intersubjective communication sets in motion an aleatory process of encoding and decoding that brings into fragile balance the various psycholinguistic competencies and sociocultural assumptions of the communicants. Where nonfictional narratives are concerned, truth, whether in a very particular or in a very general sense, always is a function of diacritical coherence, the coherence, that is, of signs in relation to their simultaneously intertextual, intra-textual, and extra-textual contexts—the archive of synchronic meanings that they evoke; the syntagmatic chain in which they are positioned; and the phenomenal, material, or historical objects and events to which they refer. And it is precisely within an especially broad and complex version of context—one that embraces a whole spectrum of cultures and belief systems—that the treaties and conventions pertaining to the conduct of war must formulate and secure their specific truth-claims.

The resulting legal and ethical principles are considered to be universally valid not primarily because states have broadly consented to be bound by them, but because they are guaranteed by rights inhering in individuals and transcending the interests of the states themselves. And yet, in practice, the contractual if not moral legitimacy of international humanitarian law derives precisely from its consensual status, its origination in transparent, publicly validated procedures of reasoning that are governed by context-dependent standards of evidence and impartiality. So, when political or governmental speech acts succeed in swaying public opinion by disseminating information and arguments that lack logical and/or referential coherence, the effectiveness of such propaganda is a function not of the inherent instability but of the distortion of signs, not of the deconstruction but of the destruction of meaning. The selling of the Iraq War to the American people is an obvious case in point. The war was largely the result of an ideological discourse that cultivated the illusion of a “natural” connection between words, meanings, and things while in fact driving a dangerous wedge between them, producing a profound disconnect, a politically useful incoherence, between the administration’s apocalyptic prophesies on the one hand and existing geopolitical realities on the other. Crucially, then, the invasion of Iraq represented a naïve—and/or deliberate—failure on the part of not only the government and the judiciary but also the citizenry and the “fourth estate” to appreciate the extent to which both the integrity and the effectiveness of our foreign policy depend on a general respect for the professional standards of evidence, demonstration, and proof to which political, legal, theoretical, and informational discourses have a special responsibility, particularly when they address the eventuality of war.

Necessity, Proportionality, and the Logic of the Lesser Evil

Coaxing consensus from (transcultural) heterogeneity, international law has stabilized many of its key terms in the last century or so, defining in clear and comprehensive fashion specific categories of persons (e.g., combatant, civilian), actions (e.g., genocide, torture), objects (e.g., lawful and unlawful weapons), and location (e.g., a munitions factory as opposed to an apartment building or a hospital). This is not to say that the semantic boundaries of such categories can never be contested, or adapted to changing circumstances, but that a wide range of meanings and objects denoted by the vocabulary of wartime ethics has been largely consistent over time, and that the enduring functionality of such classifications is the result of not only their continual reaffirmation in the international forum but also their accessibility to modification.

In view of the relative impermeability of much of the legal lexicon pertaining to the conduct of war, challenges to the standards that that lexicon has established generally rest, as I mentioned above, neither on the unintentional misinterpretation of ambiguous terminology nor on the fine-grained deconstruction of simple binaries; they tend instead to be manifestly sophistic attempts to create and exploit semantic indeterminacy where such indeterminacy cannot plausibly be found, and to circumvent meanings that, despite their historical contingency and lack of metaphysical warrant, are contextually fortified against misconstrual. Legal obfuscation and evasion typically involve various strategies for misrepresenting, euphemizing, or “disappearing” the referent. For example, as I will show in this essay, the techniques of legal and political misprision by means of which the United States government would justify its use of drones to conduct “targeted assassinations” of suspects include the following: the disavowal of clandestine military and intelligence operations carried out by the CIA and the US military’s Joint Special Operations Command; the unilateral imputation of a specious “flexibility” to restrictive legal concepts; the arbitrary or mistaken shifting of people, places, or events from one legal category to another (e.g., from civilian to combatant, schoolhouse to training camp, or wedding procession to military convoy); the refusal to investigate (and the will to underestimate) civilian and combatant “casualties”; the propagandistic exaggeration of the precision and the reliability of new weapons technologies; the targeting of unknown persons solely on the basis of statistical estimates; and the use of the language of military ethics to camouflage what may be war crimes.

Of course, not all humanitarian principles related to the conduct of war are defined with precision and detail. Uncertainty afflicts in particular the juridical concepts of necessity and proportionality, definitions of which have served to justify ad hoc exceptions to normally unequivocal moral constraints on the use of state violence, permitting under certain circumstances the perpetration of a “lesser evil” for the sake of preventing a purportedly greater evil, or of promoting the emergence of a supposedly greater good. In the current practice of drone “warfare,” as we shall see, this has been taken to mean that an extrajudicial assassination comporting a certain statistical probability of “collateral” civilian deaths may legitimately be authorized on the grounds that the executed individual is—or might someday become—a source of danger to the security of the United States. The already legally and morally exceptional means (deliberately killing other human beings), along with its “unavoidable” consequences (inadvertently killing human beings who fall under the category of bystander or civilian), are deemed necessary and proportional in relation to the specific nature of the menace to be averted, the immediate “military advantage” to be gained from the attack (jus in bello), and the overall ends of the war itself (jus ad bellum).³ The problem is, first, that each of these variables considered in isolation—the level of violence to be administered, its intended or unintended consequences, the amount of potential harm it aims to pre-empt, and the amount of good it may be expected to accomplish—is always to some extent subjective, speculative, or unpredictable; and second, that the law invites us to establish equations, or rates of exchange, between variables that, in addition to being highly unpredictable, represent concepts (including properties, quantities, and qualities) that are irreducible to any common standard of measure, hence between which there exists no compelling logical equivalence. Typically, again, the “value” of an estimated number of civilians’ lives is balanced against the “value” of the amount of military gain that is expected to accrue from the attack that will accidentally if foreseeably kill the civilians. In such cases, the equivalencies that underwrite relative or proportional value are abstractions from concrete and particular phenomena, translations of incommensurable objects and meanings into a homogeneous economy of compatible data. We are asked, in effect, to construct metaphorical bridges between hypothetical referents despite evidentiary uncertainties and conceptual differences that make coherent comparisons practically impossible.

The difficulty—or the impossibility—of quantifying and relativizing factors that are objectively and experientially incompatible is partly a function the legal language itself, and partly a function of who has the power to interpret that language—that is, which state, sub-state, or supra-state entity has not only the legitimate political and juridical authority but also the requisite geopolitical influence (or hegemony) to make and impose its own particular assessments. In the absence of clear, detailed, and universally agreed-upon guidelines, and of effective international mechanisms for equitably enforcing them, the United States has arrogated to itself the right to devise “ethical” algorithms for measuring minimally acceptable thresholds of violence, risk, and damage on a case-by-case basis. Accordingly, as Eyal Weizman has shown, the basic benchmarks and ratios in this economy of the lesser evil are formulated “within a closed system in which those posing the dilemma, the options available for choice, the factors to be calculated and the very parameters of calculation are unchallenged.” This self-regulating moral unilateralism has resulted in what Weizman calls “the humanitarian present,” an increasingly normalized condition wherein proponents of human rights and humanitarian law collude with military and political powers in the creation of a “necro-economy” of the least possible means, a moral technology in which the utilitarian claim to moderate violence is part of the very logic by which violence now justifies and perpetuates itself: “less brutal measures,” Weizman explains, “are also those that may be more easily naturalized, accepted and tolerated—and hence more frequently used, with the result that a greater evil may be reached cumulatively.”⁴ The drone program is of course paradigmatic in this regard. The conventional wisdom that the use of weaponized drones not only eliminates the need to put US soldiers’ lives at risk but also generally results in fewer civilian casualties than ground assaults and attempts at capture has served as the decisive rationale for dramatically increasing the frequency and the scope of lethal drone operations, and for prolonging indefinitely this supposedly more ethical form of warfare. While the existence of Weizman’s “humanitarian present” has certainly been confirmed by our protracted and media-managed War on Terror, history has also shown that greater evils do not have to develop cumulatively in order to be well tolerated.⁵ Lesser-evil arguments customarily are invoked to excuse the United States’ bombing of Hiroshima and Nagasaki. In the Bush-Cheney world of infinitely imaginable “unknown unknowns,” they served to justify a host of violations of the law in the name of the law, including indefinite detention, extraordinary rendition, torture, targeted assassinations, and a war of aggression during which additional “necessary” evils—the indiscriminate and excessive use of force, collective punishment, and other war crimes—were committed with impunity.⁶

After becoming president in 2009, Barack Obama did take significant measures to put an end to some of the most inhumane of his predecessor’s practices; however, a number of the unconstitutional institutions and procedures pioneered by the Bush administration have endured,⁷ among the more notable of which is the extension and the acceleration of the use of weaponized drones to carry out “targeted killings” of suspected anti-US combatants both inside and outside of officially declared war zones. Attempts to justify this extralegal policy tend to rely heavily on the  self-comforting yet self-compromising “morality” of the aforementioned “humanitarian present,” portraying the assassination program’s yield of thousands of civilian casualties as an unavoidable lesser evil arising from the United States’ compulsory self-defense against the greater evil of an always imminent threat of international terrorism; or, to put the rationale in terms of the greater good rather than the greater evil, thousands of civilian casualties abroad are alleged to be the necessary result of the United States’ protection and furtherance of the life and liberty not just of Americans but of freedom-loving peoples around the world. My point here is that both the greater good and the greater evil are conceived as infinitely extendable, their positive or negative moral stakes being limited only by the capacity of the imagination. Thus, virtually any amount of force used to repel anticipated terrorist attacks of untellable or unknowable force and frequency will readily appear proportionate to the incalculable threat that it averts, and, correlatively, virtually any number of unintentional civilian deaths resulting from this self-defensive use of force will likely be seen as proportionate, both to the unspecified but always greater number of (other) lives to be saved and to the inestimable value of the principles to be preserved. Within this closed system, the available actions and the foreseeable outcomes, the relevant considerations and the standards for weighing them, are, as Weizman pointed out, unilaterally restricted; however, other variables and criteria are at the same time conveniently left open-ended or vague. Since this peculiar combination of rigidity and elasticity, both of context and of conception, is common currency in the discourse of political bureaucracy, unverified and dehumanizing truisms about the inevitability of “collateral” casualties may even, at first blush, sound logically persuasive (as, for example, when a high-ranking government official casually remarks, with breathtaking sweep and absence of specificity, “sometimes you have to take a life to save even more lives.”)⁸

Imminence, Preemption, Feasibility

The clandestine program of targeted killing first came to the attention of the American public with the publication by NBC in February 2012 of a leaked Department of Justice “White Paper” and by way of a subsequent series of cryptic speeches by administration officials, including, preeminently, President Obama’s talk at the National Defense University in 2013.⁹ Partially declassified for that talk was a legal document called the Presidential Policy Guidance, which was not disclosed until August 6, 2016—albeit in heavily redacted form—pursuant to a federal court order obtained by the ACLU under the Freedom of Information Act.¹⁰ The Policy Guidance and a “Report on Process for Determining Targets of Lethal or Capture Operations” that was issued along with it echo key points in the White Paper, but in their published form they are so extensively censored that the White Paper is still the most substantial of the legal documents released thus far.¹¹

The White Paper was written in 2010 by the Justice Department’s Office of Legal Counsel in order to justify placing the name of a US citizen, Anwar Al-Awlaki, on the government’s secret “kill list.” Although, at the outset, the paper claims to address only the legality of targeting a “senior operational leader of Al Qaeda and its associated forces,” the focus rapidly shifts to include all members of Al Qaeda, as well as anyone suspected of being affiliated with the group. The United States, the paper says, may lawfully deploy its remotely piloted, unmanned aerial vehicles to any foreign geographic location, including to areas that are not recognized as active war zones, in order to track and kill any individual suspected of belonging to Al Qaeda or its “associated forces,” provided that an “informed, high-level government official” determines that the individual poses an “imminent threat of violent attack against the United States,” and that capturing the suspect is “infeasible.”

The paper’s definition of imminence is however so broad as to empty the word of its ordinary, functional meaning. As a subsequent ACLU memo observes, “outside of recognized battlefields, the Fourth and Fifth Amendments prohibit the use of lethal force except as a last resort in the face of a concrete, specific, and imminent threat of deadly harm.”¹² The ACLU’s complaint derives from certain settled principles within International Humanitarian Law and modern International Human Rights Law. In an international armed conflict, or war between states, the armed forces of the states that are party to the conflict can lawfully target one another, and may even target civilians if, and only for such time as, they directly participate in hostilities.¹³ In a non-international armed conflict, or war between a state and a non-state actor, the state may target only the organized armed forces of the non-state party to the conflict. Here too, civilians participating in hostilities in a “spontaneous, sporadic or unorganized way” lose their protection against direct attack if, and only for as long as, they carry out hostile acts. Moreover, civilians belonging to an organized armed group are considered to have assumed a “continuous combat function,” and therefore lose their legal immunity from attack for the duration of their membership in the group.¹⁴ This exception to the rule of noncombatant immunity (or, to put it another way, this shift from civilian to combatant status) is, however, legitimately available only in the context of non-international and, mutatis mutandis, international armed conflicts. Outside of an officially declared war zone, everyone is a civilian; and the right to life recognized in the United States Constitution as well as in the International Covenant on Civil and Political Rights has been interpreted to prohibit the use of lethal force against civilians except in those very limited circumstances where they present “an imminent threat to human life that cannot be otherwise ameliorated.”¹⁵

In light of the last of these provisions, the frequent government and media references to “combatants” and “militants” lack legal foundation, insofar as the alleged adversaries are not party to a legally recognized armed conflict with the United States. Similarly groundless is the Bush administration’s labeling of Al Qaeda suspects as “unlawful enemy combatants.” Once they are deprived of the rights or protections normally afforded combatants under international law, these suspects become civilians, not some less-than-human form of “bare life.” As civilians, they are, under human rights law and the law of armed conflict, liable to attack only at and for “such time as they take a direct part in hostilities.”¹⁶ In this context, it is worth noting—particularly since the White Paper was initially drafted to help inculpate him—that it has never publicly been established that the “radical Muslim cleric” Anwar Al-Awlaki ever assumed an “operational” role within Al Qaeda. While his sermons likely inspired many so-called jihadists, there is no extant evidence suggesting that he ever assumed a function equivalent to that of a terrorist or a traditional combatant. Neither his association with Al Qaeda nor his role as a spiritual motivator of anti-American sentiment were sufficient cause for depriving him of his civilian status and waiving the strict requirement of imminent deadly harm.¹⁷

Instead of offering evidence of a concrete and impending threat, whether from Al-Awlaki or from any other alleged “terrorist,” “militant,” or “combatant,” the Justice Department’s memo refers repeatedly to the inherent dangerousness of individuals who may be engaged in what it nebulously refers to as “continual planning” and “constant plotting” against American persons and interests, bluntly asserting that the United States is not required to have “clear evidence that a specific attack . . . will take place in the immediate future” before using deadly force to defend itself against its suspected enemies. Is the charge of “continual planning” an oblique invocation of the “continuous combat function” (mentioned above) of organized armed groups participating in a non-international armed conflict? If so, the reference is misplaced with respect to non-imminent threats occurring outside of an officially declared war zone. But whatever that particular wording’s intended legal connotation, the memo’s basic message manages to be at once dangerously vague and starkly unequivocal: if an individual is identified in secret, by anonymous government sources, as possibly being involved, in some general way, in plotting attacks against the United States, the accused individual may be killed without warning, at any time and any place, “even if there is no specific evidence of where such an attack will take place or of the precise nature of the attack.”¹⁸

In its positing of false dilemmas and worse-case scenarios, and in its failure to demonstrate directly and convincingly the applicability of the just war principles to which it appeals, the White Paper is uncannily reminiscent of the torture memos of the George W. Bush presidency.¹⁹ The drone memo briefly invokes the principles of necessity, distinction, proportionality, and humanity (the avoidance of excessive civilian casualties or “unnecessary suffering”), but the only of these issues it addresses at length is that of necessity, which it frames very broadly in terms of, first, the President’s constitutional responsibility to protect the nation, and second, the sovereign right to self-defense under UN Charter’s article 51. Such language would appear to bring the drone program under the legal umbrella of pre-emptive military action against an imminent or proximate threat, in accordance with a long established exception to the UN Charter’s general restriction of the legitimate use of self-defensive force to the deflection of an actually occurring armed attack.²⁰ But the picture that emerges from the memo is instead that of a scattershot program of killing that is neither preemptive nor preventive in any accepted legal sense.²¹ The claim to self-defense is substantiated only by appeals to an amorphous definition of imminence, by an implausibly capacious reading of the 2001 Authorization for Use of Military Force, and by threadbare arguments from “analogous contexts” (e.g., the laws of war pertaining to detention, or interpretations of “probable cause” in domestic law enforcement).

The Authorization for Use of Military Force is limited in application to those “nations, organizations, or persons” who the president determines had “planned, authorized, committed, or aided” the terrorist attacks of September 11, 2001, or who had harbored the individuals or organizations responsible for those attacks.²² As an ACLU memo observes, “Congress did not intend the AUMF to be a ‘blank check’ for the use of lethal force without geographic or temporal constraints, or for the use of lethal force against actors who have no substantial connection to the September 2001 attacks.”²³ The restrictiveness of the Authorization was clarified by Congress’ rejection of an earlier version of the bill, which would have licensed George W. Bush to use lethal force to “deter and pre-empt any future acts of terrorism or aggression against the United States.”²⁴ Thus, to fall within the Authorization’s ambit, a group must directly have engaged in conduct leading to the terrorist attacks of September 11, 2001. Yet, what government officials refer to as an “Al Qaeda core” no longer exists, and the groups that are currently targeted under the ill-defined rubric of “associated forces” (the Tehrik-i-Taliban Pakistan, Al Qaeda in the Arabian Peninsula, and Somalia’s al-Shabaab) were founded well after the enactment of the Authorization to Use Military Force.²⁵ Despite adopting the “Al Qaeda” brand name and sharing a general ideological affinity descending from the early Al Qaeda, these groups are for the most part neither intent on nor capable of attacking the United States; they are instead preoccupied with domestic civil wars or regional insurgencies.

The local, scattered, and loosely organized nature of these and numerous other nominal Al Qaeda “branches,” “franchises,” or “affiliates” constitutes yet another impediment to the legitimacy of the United States’ targeting of them. For the United States cannot legally consider itself to be party to a non-international armed conflict (whether or not it in fact does so has never been clarified) if its non-state adversary is not a clearly identified fighting force with a central command and control structure, and if hostilities between that actor and the US have not attained a threshold of duration and intensity exceeding merely sporadic acts of violence. Elucidating the foregoing criteria as set forth in an opinion paper of the International Committee of the Red Cross, Daphne Eviatar concludes,

The scattered nature of these groups, the fact that almost none of them is targeting the United States, and the apparent lack of any command structure linking them weigh strongly against the United States being able to treat them, for the purposes of international law, as an enemy actor, either separately or united, against which the United States is engaged in an armed conflict.²⁶

If neither domestic nor international law can credibly establish the legal existence of an armed conflict between the US and its deterritorialized adversaries, then even the likelihood that Pakistan and Yemen have given the United States tacit permission to conduct some drone strikes on their territory is not sufficient to make those states formal co-belligerents of the US and so does not alleviate the US from the jus ad bellum requirements of imminence, necessity, and proportionality.²⁷ Moreover, by withholding formal consent to the drone strikes, Pakistan and Yemen retain their legal status as neutral, third-party states. Belligerent intervention by the United States is therefore a violation of their sovereignty. Some drone proponents claim that the United States can legally take military action against enemy forces on the territory of a neutral state if that state is (as appears to be the case in Pakistan) “unwilling and unable” to prevent the use of its territory to the detriment of the United States. But here again, the “right to self-help” is subject to the requirements of imminence and necessity; a preemptive use of force, it is dispositive only in circumstances where the intervening power can reasonably adduce, in Daniel Webster’s famous phrase, “a necessity of self-defense, instant, overwhelming, leaving no choice of means, and no moment for deliberation.”²⁸

Of course, the primary rationale for extending the drone battlefield beyond hot combat zones is that, to use Mao Tse-tung’s famous phrase, the targeted insurgents behave as fish, seeking shelter and sustenance in the “sea” of local populations. Historically, it is precisely the embedded nature of the enemy that is responsible for the repeated failure of counterinsurgencies. As foreign policy consultant and historian William R. Polk has remarked, “there is no record that counterinsurgency ever worked anywhere, and it is certainly not working” in the War on Terror. Because military intervention in foreign countries to chase cells of “bad guys” tends to be viewed by the native inhabitants as “unjustified and brutal,” it does not matter how many insurgents are killed; new fighters and their supporters will emerge from among the inhabitants.²⁹ This strategic reality is spelled out in the findings of the Stimson Center’s bipartisan inquiry into the political ramifications of drone strikes, in comments made by General Stanley McChrystal, and in a leaked internal CIA report.³⁰ According to a report in The Washington Post, by 2012, after three years of escalating US drone strikes, the core membership of Al Qaeda in the Arabian Peninsula had grown from 300 militants to 700 or more.³¹ Evidently, stretching the concept of the enemy beyond traditional bounds to include guerillas who may have blended with the civilian population and infrastructure in undeclared fields of conflict does nothing to break the cycle of violence; rather, it raises a host of legal and ethical problems around the traditional concept of collateral damage. Constitutional scholar Owen Fiss succinctly summarizes the pertinent uncertainties as well as the legal and moral prohibitions to which those uncertainties should give rise:

Even if there is no doubt as to whether the target is a leader of a terrorist organization, there may be doubt as to whether capture is feasible, or whether the lethal attack might be deferred until a later date when the killing of civilians would be avoided, or whether the terrorist organization is a co-belligerent of al-Qaeda, or whether the nation in which the suspected terrorist is located has relinquished its claim to sovereignty. These uncertainties seem inevitable and are significantly greater than those associated with the killings that invariably take place in a combat zone or an active theater of armed conflict. As a consequence, the advantage to be gained by killing a suspected terrorist is, as a matter of morality and legality, too slender to shoulder the ensured killing of civilians, under the theory that such killings would be nothing more than collateral damage.³²

Functioning in tandem with what John Brennan, Obama’s counterterrorism chief at the time, called a “flexible understanding of ‘imminence'”—that is, again, the notion that the United States can kill its putative enemies without having clear evidence that they are engaged, or are about to engage, in an armed attack against the United States—is the easily manipulated concept of feasibility, where, as mentioned above, the infeasibility of capture is put forth as grounds for opting to assassinate, rather than to apprehend, interrogate, and try in a court of law, suspected terrorists.³³ Outside of conventional battlefields, international human rights law permits the use of lethal force against an imminent threat only when there is no other available means of self-defense, such as diplomatic efforts, non-lethal incapacitation, or capture. Yet, as Daniel Klaidman reveals in his chronicle of the assassination of Osama bin Laden, Obama and his advisers routinely take the position that capture is infeasible, having found that “killing is a lot easier than capturing”³⁴—”easier” for reasons such as Republican opposition to Federal criminal trials for terrorism suspects; the possibility that the “host” country might refuse to consent to a capture operation; the political sensitivity of putting American boots on foreign soil; and the United States’ reluctance to expose its soldiers to the physical risks of ground operations. Thus, political expediency overrides not only international law but also the Fifth Amendment, with its guarantee of due process and protections against being erroneously deprived of one’s life. Like the majority of the prisoners at Guantanamo, but with consequences that are irrevocable, the accused are given no opportunity to defend themselves against flawed intelligence, equipment failure, software glitches, or human error; nor are they ever given the opportunity to surrender. They are judged, convicted, and summarily executed on the basis of secret determinations made by government officials acting unilaterally, with no legal checks or adversarial testing and no sustained, independent oversight, whether from other branches of the government, the press, or the public.

Necro-Statistics

On July 1, 2016, in a belated bid to project transparency and to defend publicly the alleged legitimacy of the assassination campaign, the Obama administration released its assessment of the number of civilians killed in drone and other “counterterrorism” air strikes in areas outside of “active hostilities” (specifically, in Libya, Pakistan, Somalia, and Yemen) between January 2009 and December 2015. The published figures were met by a wave of skepticism in the US, not only from civil rights groups but also from mainstream news outlets, which noted that the estimates covered only a limited range of time and excluded civilian casualties occurring in ground operations as well as in the declared war zones of Afghanistan, Iraq, and Syria (areas where the civilian death toll is likely to be considerable).³⁵ Moreover, the government’s report provides statistics for the inclusive seven-year period in the aggregate only, instead of offering a breakdown by year, or by location, or by each particular strike. This lack of specificity makes the totals at once unverifiable and implausible by comparison with the much higher tallies published by independent organizations, whose incident-by-incident data is compiled not only from local and international news sources but also from leaked government documents, court papers, and field investigations involving extensive interviews with witnesses, survivors, relatives, and foreign officials.³⁶

The government’s release of noncombatant casualty figures was accompanied by an “Executive Order” intended to codify targeted-killing guidelines for the use of future administrations, and to serve as a template for “foreign partners” who might be developing or already deploying advanced drone technology for similar purposes.³⁷ The document states that “relevant departments and agencies” shall “take feasible precautions in conducting attacks to reduce the likelihood of civilian casualties,” a notably weaker and vaguer standard than the “near-certainty” articulated both in the Presidential Policy Guidance and in Obama’s speech at the National Defense University (“before any strike is taken, there must be near-certainty that no civilians will be killed or injured”).³⁸ The Order also authorizes the office of the Director of National Intelligence to “publicly release” an annual summary of “unclassified” information pertaining to combatant and noncombatant deaths occurring outside areas of active hostilities, and it requires that the US government “acknowledge . . . responsibility for civilian casualties and offer condolences, including ex gratia payments, to civilians who are injured or to the families of civilians who are killed.” Such apparent good will is undermined, however, by the opacity and the dubiousness of the statistics accompanying the Executive Order, by the government’s past record of refusing to acknowledge or to apologize for erroneous air strikes (the sole exception being a 2015 strike that accidentally killed two Westerners), by leaked evidence of US attempts to hide its responsibility for civilian deaths resulting from specific strikes, and by clear instances of government propaganda, such as John Brennan and the CIA’s concerted claim, in 2011, that not a single noncombatant had died in the many hundreds of lethal drone strikes of the preceding year.³⁹

What is perhaps most troubling about President Obama’s outline of the goals and procedures allegedly governing drone strikes is its implicit acknowledgement that the government itself is often ignorant of the identities of those whom it has killed, a fact that is suggested by the Executive Order’s promise that US defense and intelligence agencies will consult periodically with “nongovernmental sources” and with “partner governments” to address any discrepancies between their post-strike assessments and its own, and to update regularly its estimate of “civilian casualty trends” in the light of new information gleaned from ongoing investigations. Indeed, both the reliability of the recently released statistical data and the candor of the accompanying policy statement’s professions of respect for the rule of law were put into question by the administration’s concurrent confirmation that the institutionalized guesswork of what it calls “signature strikes” was nonetheless expected to continue. Although the laws of war stipulate that, in cases of doubt as to whether persons in a war zone are civilians or combatants, they are to be considered civilians,⁴⁰ drone operators routinely carry out attacks against gatherings of unknown people, based solely on their putatively suspicious patterns of life (their behavioral “signature”). More inexplicably still, the unmanned aircraft also perform “double-tap” strikes, in which a second drone attack indiscriminately kills anyone—including bystanders and professional humanitarian workers—seeking to bring aid to the victims of the initial strike. Rather than acknowledge the enormous collateral toll of these literally blind deployments of military force, the Obama administration reportedly counts any unidentified male of military age who dies as a direct result of an American drone strike as an “enemy killed in action” (EKIA)—unless “intelligence” (from unspecified sources) should happen to prove his innocence posthumously.⁴¹ When one considers the generous latitude that the government has granted itself for making fatal errors—errors that arise quite foreseeably from the arbitrariness of targeting methods based on statistical probability and reductive behavioral modeling, the unreliability of much intelligence, the limited precision of drone imaging technology, and the referential vagueness of “Al Qaeda and associated forces”—it is perhaps not surprising that, as we saw in the previous section, the vast majority of people targeted by US drones have turned out to be not the “high-value” masterminds alleged by the Pentagon but either civilians mistaken for combatants or else anonymous low-level “militants” who, while perhaps ideologically sympathetic to Al Qaeda, had neither the means nor the access to pose a serious threat to the US, and who were engaged in tribal conflicts or local insurgencies against their own government—not in plotting terror against the United States. Furthermore, as we have seen, the evidence suggests that rather than “disrupt, dismantle, and defeat” these armed groups,⁴² the drone strikes have facilitated recruitment to their cause, and have motivated an increasing number of attacks by militants in Yemen, Somalia, Pakistan, and North Africa.⁴³

Echoing verbatim the key terms of the White Paper, the Executive Order invokes the “fundamental principles of necessity, humanity, distinction, and proportionality,” and affirms as well the administration’s “heightened policy standards” for the protection of “vulnerable populations.” Yet civilian casualties, it says, are “a tragic and at times unavoidable consequence of the use of force in situations of armed conflict or in the exercise of the state’s inherent right of self-defense.” While the White Paper and the policy statement both present civilian deaths as necessary, unintentional, and secondary in relation to the legitimate “mission objectives” of the state, those deaths are of course the primary motivation (the “necessity”) behind the government’s cryptic legal pronouncements and its release of data, both of which represent “preemptive” maneuvers in “self-defense” against “imminent” accusations of highhandedness. And indeed, given the geographic remoteness of the drone “battlefields,” the ethnic and cultural remoteness of the peoples who inhabit them, and the immunity of the Pentagon’s joystick warriors to the physical risks traditionally associated with engagement in military combat, frequent and detailed news reports about noncombatant deaths would seem to be the single factor most likely to overcome the American public’s general mood of apathy toward the drone program’s (mostly foreign) casualties, and to focus attention on the relevant issues of secrecy, constitutionality, and international law. But despite the periodic appearance of critical newspaper and magazine articles in the mainstream press, protests by human rights organizations, and a spate of trade books about drones and targeted killing, the matter of civilian casualties has not received the kind of sustained reportage and analysis that might erode public support for the policy. On the contrary, recent polls suggest that the great majority of Americans are persuaded of the necessity and the effectiveness of targeting suspected terrorists, and remain unperturbed by news of the ugly “side-effects” of the air strikes.⁴⁴ Meanwhile, this relatively inexpensive and risk-free tactic enjoys broad, bipartisan support on Capitol Hill, where Democrats and Republicans alike appear unmotivated, politically, to subject the program to scrutiny or to meaningful oversight.

The Duplicity of Double Effect

What this general attitude of complacency suggests, among other things, is that the whole debate over collateral-death statistics is fundamentally misplaced. For it automatically situates the conversation about the relative worth of legally innocent (foreign) lives within the utilitarian framework of the lesser-evil arguments that I discussed at the beginning of this essay. The set of criteria that, since its origins in Thomas Aquinas’ Summa Theologica, traditionally has been invoked to justify lesser-evil exceptions to ordinarily inviolable ethical norms is known in philosophical discourse as the law of double effect. The main tenets of the double effect principle may be summarized briefly as follows: in certain situations it may be necessary and therefore permissible to cause harm in order to achieve a greater good, provided that the harm is neither an end in itself nor the direct and intentional means for achieving the good effect, and provided that the intended good effect of the harmful act outweighs its bad effects.⁴⁵ In military ethics, this precept has been taken to mean that it is morally permissible to kill civilian noncombatants provided that their deaths are a foreseeable but unintentional effect of a legitimate military action. Related requirements, codified in the laws of war, are that due diligence must be exercised to minimize harm to civilians, and that the number of civilian and combatant casualties should not be “excessive,” but rather “proportional,” in relation to the threat that the action repels or to the military advantage it aims to accomplish.

Proportionality is a useful concept in international law when what is reciprocally weighed—what is determined to exist in a state of relative equilibrium or disequilibrium—is the kind and the amount of force used by the opponents, the kind and the amount of architectural or environmental damage they inflict on one another, or the kind (combatant or noncombatant) and the number of casualties they may suffer. It is another matter, however, to estimate the value of a certain number of civilian lives relative to a concept, or a value, so subjective as an anticipated military gain. Between these two fundamentally unlike variables there exists no objective ground for comparison, no common standard of measure on the basis of which to establish moral equivalence or discrepancy. Moreover, as I have said, the variables that are to be quantified, weighed, and translated into an ethical hierarchy may be not only incommensurable but also unpredictable. The immediate effects and long-term consequences of any particular act of violence—the nature or the extent of the “greater good” it will accomplish, of the “greater evil” it will avert, or of the “lesser evil” it will incur along the way—cannot be fully foreseen, measured, or controlled by military, government, and intelligence personnel.

These problems are of course only compounded by the secrecy and the legal uncertainty surrounding the United States’ use of killer drones in foreign wars that are and are not wars, over battlefields that are and are not battlefields, against a vaguely defined enemy, for military and strategic ends that are largely hypothetical. For any given drone strike, or series of strikes, what, we might ask, is the ethically satisfactory ratio, the rate of exchange, between noncombatant lives and the anticipated military advantage for which they may be sacrificed? In US Special Operations airstrikes between January 2012 and February 2013, 35 of the 200 people killed were intended targets, and, during one five-month period of the same campaign (“Operation Haymaker”), nearly 90 percent of the people killed were not the ones intended.⁴⁶ Between November 2002 and November 2014, in Yemen and Pakistan, repeated and often unsuccessful attempts to kill 41 specifically targeted men resulted in the estimated death of 1,147 people.⁴⁷ If these figures seem somehow “excessive,” or “disproportional,” then is the government’s tally—between 64 and 116 noncombatant casualties in 473 air strikes that are said to have killed from 2,372 to 2,581 combatants in undeclared battlefields since 2009—safely within the bounds of moral permissibility?

If proportionality depends, as would seem to be the case, as much on the quantification of value as on the valuation of quantity, should the combatant to noncombatant ratio not be adjusted (sometimes it is) for a “high-value target” as opposed to a “low-level militant”? Under the general rubric of “civilian casualties,” shouldn’t there be subcategories to which greater or lesser weight may be assigned? Pondering the mysteries of collateral proportionality, an Israeli military lawyer asks, in all seriousness, “How do you count women in relation to men? How do you count the death of children? Does one dead child equal one dead grownup, or does he equal five dead grownups?”⁴⁸ Taking the implications of this computational logic a step further, shouldn’t we assign different numerical values to different civilians’ lives based not only on their age and sex but also (in deference to President Trump’s worldview) on their social status, ethnic identity, or religious affiliation? What about a “casualty’s” overall intelligence or attractiveness, health, or disability? Shall we consider as well his or her indispensability to the financial resources of a particular household, or to the political stability of his or her local community?

In the end, none of these metric refinements would help us to come up with a universally or even broadly consensual evaluative framework for ethically measuring the statistical probability of a certain “cost” in civilian life against the largely unknowable military, strategic, or humanitarian “benefit” to be derived from the violence unleashed in its name. How much, and what kind of, “good” results from the targeted assassination of a person who is suspected of belonging to a group that is believed to be affiliated with—or merely sympathetic to—Al Qaeda or ISIS, whose members may be, or may someday become, involved in plotting unfathomable harms to American interests?⁴⁹ Lesser-evil humanitarianism would require some concrete knowledge of how many American lives this particular suspect’s death is likely to save, and would dictate that we compare that figure with the number of civilian lives likely to be lost in the strike or strikes that ultimately eradicate him. But even if we could magically produce such information with accuracy, that would not resolve the ethical aporia produced by our attempt to apply a mathematical balancing test to the inherent value of human lives. And further unknowns would abide. What immediate military advantage would this individual’s death bring about? If he is indeed a high-level militant, does killing him throw the terrorist network to which he belongs into a state of panic, paralysis, or general disarray; or does another leader simply rise to take his place, accompanied by new fighters effortlessly recruited from among the victim’s friends, family, or tribe? What long-term casus belli is served by targeting him and at the same time inadvertently killing bystanders? Is the War on Terror thereby closer to being won?

Let us suppose, for the sake of argument, that lesser-evil calculations can and do yield acceptable ethical guidelines. If, as a result, we Americans are indeed safer now than we were before the armed drones came to our rescue, then should our government not be willing and able to explain to us, in empirical terms, just how much safer we actually are, and to demonstrate to other countries as well that its transnational program of targeted killing sufficiently advances the interests and the security not only of the US, but also of the global community, to justify the life-risks it levies on civilians? My point here is not that we should embrace lesser-evil humanitarianism, but simply that, having opted for a lesser-evil interpretation of international law, the United States has, in the course of its drone campaign, failed to make a sincere effort to apply such a framework. Since those standards have been formally adopted by the US, both law and morality demand not that they be invoked vaguely or tendentiously, but that they be formulated and adhered to as rigorously as possible. On the evidence thus far, however, and within the parameters of the legal humanitarianism that it professes, the government-intelligence nexus is capable of exercising far greater caution and producing far greater empirical certainty than it in fact does. This does not mean that, by increasing our margin of “due care” (and hence reducing somewhat the number of innocents killed), we would thereby be relieved of the ethical burden posed by the hyper-positivistic assumptions and the inevitable empirical indeterminacies of a “humanitarian” necro-calculus. It means only that the US has the responsibility, within the just-war framework it claims to respect, to produce a much higher level of predictability and verifiability for its kill ratios and collateral damage estimates. A greater degree of certainty in this regard, while constituting a lesser evil, would not however eliminate the pitfalls of incommensurability and error, nor would it escape the moral challenge raised by the civilian casualties for which it allows. At the same time, in keeping with the principle of necessity at the level of jus ad bellum, we need to ask ourselves if the freedoms and the core values of the world’s sole remaining superpower could truly be so fragile, so permanently subject to “imminent” destruction, that innocent strangers a world away must continue to die for them. We certainly do not ask the civilian who is about to become a statistic what he thinks the relative value of his life ought to be on the American market of political ideas. Yet it is not at all obvious that he does not have an inherent right not to be killed, or not to be exposed to the risk of being killed, for ends that he himself has not freely chosen.

Despite the hermeneutic room for maneuver afforded by the fog surrounding double-effect calculations of proportionality in cases where fundamentally incommensurable variables are mathematically converted and compared, government lawyers, as we have seen, have been unable to make a sound case for the practice of targeted killing other than by hiding or distorting certain facts and by bending legal principles beyond recognition. I have noted in particular the casual conflation of local insurgencies with global terrorism, the failure to investigate compelling testimony and material evidence, the refusal to define feasibility, and the subtraction from imminence of precisely those semantic components (urgency, necessity, and last resort) that give the term its distinctive meaning as a legal trigger for defensive military action. Had government lawyers undertaken instead to perform a “close reading” of the pertinent conceptual binaries—to locate the “in-between” of, say, imminence and latency, or feasibility and infeasibility, or proportionality and excess—the adopted critical methodology would have required that they formulate subtle distinctions and precise definitions grounded not only in reasoned inferences from juridical intertexts but also in the analysis of material evidence and empirical contexts. The endpoint of such an exercise would be the provisional (re)construction of meaning, of a precisely circumscribed semantic field derived from, and subject to, both legal and public consensus. In the administration’s rhetoric, on the contrary, discipline-specific terms of art are made to signify whatever is asked of them, independently of the normal constraints of both semantics and pragmatics. Too often, moreover, the connection between the new, idiosyncratic sense (or non-sense) and any corroborating situational reality is either unilaterally asserted or else obscurely implied, rather than being extrapolated from verifiable facts, in accordance with transparent and inclusive procedures of deliberation.

The Alibi of Intention

Before concluding my reading of the Obama administration’s reiterated claim that its remotely guided executions are conducted in a manner fully consistent with the “fundamental principles” of the international laws of war, I would like briefly to consider the legal ramifications of the intentionality clause of the double effect rule. Again, double effect stipulates that it may be necessary to cause harm in order to achieve a greater good, provided that the harm is not the direct and intentional means for achieving the good effect, and provided that the intended good effect of the harmful act outweighs its bad effects. Correspondingly, in the definition of proportionality as a legal principle, the legitimacy of trading civilian lives for military gain is contingent on the actor’s intentions regarding the foreseeable outcomes of an attack: under the Statute of the International Criminal Court, it is a war crime “intentionally” to launch an attack “in the knowledge that such attack will cause incidental loss of life or injury to civilians or damage to civilian objects . . . which would be clearly excessive in relation to the concrete and direct overall military advantage anticipated.”⁵⁰ Indeed, in order to distinguish between terrorism—definitions of which generally stipulate the deliberate use of violence, or of the threat of violence, against civilians for the purpose of achieving political objectives⁵¹—and collateral casualties, we need to be able to tell the difference between what an agent “directly intends” and what he or she “foresees” as an incidental, accidental, and unavoidable by-product of violence that is intended to bring about a legitimate military goal.

The first and most obvious obstacle to establishing the nature of an agent’s intentions is that we have no way of gaining direct access to subjective states of mind. Proving another’s intention or lack of intention with regard to the commission of a specific act requires making logical inferences from external signs regarding motives that are by no means necessarily logical, and whose origins may be to varying degrees conscious or unconscious. Determining legal and moral responsibility for the civilian casualties of drone strikes is further complicated by the various ways in which the new weapons technology, as well as its means of implementation, tend to obscure accountability by spreading it among multiple agents, both human and nonhuman. For example, drone missions and their pilots generally fall under two chains of command, one located in the U.S. and the other in the distant “battlefield.” When a mistake is made, the pilot could be at fault, but so could either of the commanding officers. And it is often unclear which command chain has the greater authority, hence responsibility, for the outcome of a particular strike. Moreover we may not know exactly who, using what computer algorithm and what kind of data, may have been involved in making pre-strike collateral risk assessments; or who—which “informed, high-level government official”—may ultimately have been responsible for deciding that the planned strike qualified as a “necessary” and “proportional” response to an “imminent” danger. Consider also that the weapon’s targeting programmer may inadvertently have entered the wrong coordinates; the intelligence may have been flawed; or the on-screen images and icons might have been in some crucial way either ambiguous or misleading. Perhaps there was a bug in the software; perhaps the system was improperly designed. Should, then, a programmer or an engineer be held responsible?⁵²

Errors of this kind may be unintentional, but they still leave open the question of whether the resultant civilian deaths are “clearly excessive in relation to the concrete and direct overall military advantage anticipated” or, for that matter, in relation to the greater good that the war itself is expected to secure. If it is true that drones generally can be more discriminating than either manned aircraft or ground troops, then it is only fair to ask why so many civilian casualties seem “unavoidably” to result from the use of this exceptionally precise technology. Are the significant civilian death counts indicative of a certain recklessness in the execution of the strikes, or of negligence at some stage in the decision-making process leading up to them? If instead we assume that, in the course of most drone sorties, the precautions taken to protect civilian life are successful in yielding no more than an anticipated and proportionally scaled number of unintended casualties, does the military advantage procured by these missions still outweigh the “unavoidable” lesser evil of killing the maximum permissible number of civilians—especially when we dispose of both the technology and the know-how to kill far fewer innocents than we actually do? How unintentional, then, are those supposedly proportional deaths? In such circumstances, is the indirect and unintentional—but also foreseeable—evil of slaying and maiming civilians really morally distinguishable from the direct, purposive, and “necessary” means (the resort to lethal force) from which it is practically inseparable?

In questioning the fault-line between foreseeability and intention, I am not primarily concerned with the relative morality or legality of sporadic errors occurring within an otherwise controlled humanitarian necro-economy; nor am I particularly interested in discovering and evaluating specific casualty estimates, or the formulae for calibrating them, in classified “collateral damage forecasts.” I am concerned, more broadly, with the targeted-assassination practices that are a matter of national policy, and that therefore cannot properly be called unintentional. These include assassinations authorized on the basis of often unreliable cellphone metadata; the arbitrary and indiscriminate use of double-tap strikes; and the inference of guilt (including necessity and imminence) from suspected ideological affinities or pattern-of-life analyses. The laxness of such methods, and the enormous risks that they pose for noncombatant populations, are fully understood by those who implement them. To the extent, then, that the “accidental” civilian deaths incurred by these inherently imprecise military and political practices are not the result of an unintentional failure to foresee harmful consequences that might reasonably have been foreseen and avoided, but rather of the deliberate and systematic application of practices that are known to comport an unusually high risk of error, these practices cannot be characterized as “mere” recklessness or negligence (which are forms of liability not necessarily bound to intentionality).⁵³ The systematization, codification, and institutionalization of scarcely informed best-guessing signals an intention to kill combatants and noncombatants alike, without regard to customary and feasible standards of due caution and discrimination. This does not mean that US operatives necessarily like to kill civilians, or that they want to kill civilians—only that they are willing to kill civilians in both foreseeable and unforeseeable numbers, for equally foreseeable and unforeseeable ends. They intend to kill more combatants than civilians, of course, but they do not know for certain that that is what they are doing; indeed, the vagueness of their own published statistics suggests that they know very well that that is not what they are doing.

Another significant sense in which the government’s extrajudicial program of targeted killing may be said to flirt with terrorism lies in the direct and foreseeable effect that the sniper planes have on the civilian populations living beneath them—on the men, women, and children who involuntarily witness the carnage of each strike, and who in the aftermath live with the incessant fear of being the next target of what they justifiably see as a merciless and unpredictable killing machine. Let us recall, in this context, that definitions of terrorism stipulate the calculated use not only of violence but also of fear, intimidation, or the threat of violence, for political, religious, or ideological ends.⁵⁴ While politically coercive terror is certainly not the avowed purpose of the United States’ global network of lethal drones, to pretend that it is not a purpose thereof, and indeed a significant one, is surely disingenuous. It is, in fact, difficult to see how the constant, hovering presence of these weaponized aircraft could not be intended as a permanent threat, not only to suspected militants but also to civilian noncombatants who “could” be thinking of aiding the enemy, or of becoming enemy combatants themselves. In addition to the wanton destruction of property and bodies that is attributable to US drone strikes, forensic evidence and the testimony of witnesses have revealed severe psychological and social harms, including depression, PTSD, and a sharp increase in suicides among persons subject to the unrelenting gaze of the lethal aircraft. In response to the “signature strikes” that have massacred groups of civilians at schools, funeral processions, wedding parties, and meetings of tribal elders, many Afghani and Pakistani parents have removed their children from school, while local inhabitants generally seek safety by avoiding social gatherings altogether.⁵⁵

But we need neither empirical studies nor a sophisticated theory of mind to understand that the indigenous populations living under unremitting panoptic surveillance by military drones must suffer acutely from their unrelieved subjection to the imminent risk of being killed, nor to infer therefrom that, whether they perform their tasks with reluctance or with zeal, the men and women who plan, command, and execute these strikes are implicated perforce in the crime of collective punishment and intimidation.⁵⁶ Since they cannot be unaware that the coercive effects of the unmanned aircraft on policed populations are wholly indiscriminate, drone operators necessarily participate, knowingly and intentionally, in the killing and wounding of “some” despite the traumatic consequences for the “many.” If they neither expect nor wish thereby to create even more enemies for the US, then it is likely, albeit by no means imperative, that they believe that the killings will serve as a lesson, a warning that may deprive the militants of popular support, weaken their resolve, and shift allegiances. Ultimately, however, it does not matter what conscious or unconscious reasoning may have motivated a particular drone-team agent or commander to (intentionally) launch an attack or a series of attacks. It cannot logically and in good faith be maintained that the shock, distress, and paralyzing fear that the attacks, as instruments of a foreign policy, inflict on whole communities of defenseless men, women, and children are purely incidental and accidental, much less necessary and proportional. The continued use of this weapon of terror despite its tendency to turn whole populations against the US suggests either that policymakers are too afraid of being perceived as “soft on terrorism” to change a course of action that they know is counterproductive, or that the intended pedagogy of drone violence (the will to control uncooperative peoples by cowing them into submission) is so intuitively commonsensical to the majority of Americans that their belief in its effectiveness will not yield to the geopolitical reality. If, as we have seen, the violence of counterinsurgency is more likely to breed further violence than to bring about peaceful surrender,⁵⁷ then the whole drone campaign may be understood as an egregious psycho-political miscalculation, one that, under the momentum of state, corporate, and techno-military investments, has taken on a life of its own, stubbornly refusing to correct itself even in the face of its evident failure to reduce global terrorism. Perhaps the violent backlash that we are witnessing among those who are condemned to live in the shadow of these menacing aircraft would appear less inexplicable if one day an ally of the US, following the precedent that we have established, were to adopt the same murderous and error-prone technology to hunt its suspected terrorists on American soil, say in a populous city like New York or Los Angeles.

Conclusion

It has been my methodological contention here that theorists and practitioners of ethics and the law can acknowledge that the mimetic effects of language are not always pragmatically or performatively predictable, without thereby renouncing the effort to achieve a viable consensus as to the accuracy of their disciplinary vocabulary, particularly in morally uncertain and referentially complex situations. By remaining alert to the specific types of verbal and situational contexts in which this specialized language may incline toward either reductionism or ambiguity, we are in a position better to understand how and why certain ethical concepts become vulnerable to abuse (as does the idea of intention, when it is characterized either as inherently indecipherable or, at the opposite extreme, as always simple and immediately transparent). We have seen in particular that, when it is bent to the utilitarian ends of the laws of armed conflict—which, we recall, exist not only to deter but also to permit the prosecution of wars—the rule of double effect, despite (or indeed because of) its principles of just cause, right intention, proportionality, and discrimination, can become a mere apology for the rule of double standards. In certain applications, as we have seen, double effect’s “doubleness” becomes duplicity, a rhetoric of sovereign self-exemption from ethical normativity, structured in accordance with the discourse of the fetish: We know that we should not intentionally kill or terrorize civilians, but just the same, we are to be excused for intentionally deploying a tactic that we know will kill some of them and terrorize many of them, for to do so may break the enemy’s will to resist, and in the end might possibly help to save American lives.⁵⁸ We know that it is wrong intentionally to kill and terrorize innocent noncombatants, so we do our best to kill and terrorize them only as much as our soldiers’ safety and our projected military advantage requires, and always in due proportion to the greater good that we anticipate our victory will bring about. Incongruously abstract, speculative, and hyper-positivistic, thinking of this kind underlies the Pentagon’s quasi-mystical belief in the capacity of war games, “human terrain” forecasts, statistical “threat profiling,” and cybernetic models of the “target” culture to transform the messy business of counterinsurgency into an objective, antiseptic, manageable process.⁵⁹ Starting from similar assumptions, legal minds in the government, the military, and the intelligence community set themselves to quantifying good and evil, loss and gain, life and death, as if they were processable units of data, information to be converted by magical algorithms into “humanitarian” reasons for assassinating terrorist suspects, and at the same time for killing and terrorizing civilians.

I said earlier that the whole debate over the relative accuracy of different civilian casualty estimates has been largely misplaced. This is in part because statistics alone, to the extent that they are devoid of narrative, imagistic, and referential “content,” lack moral weight; they invite us to reduce the meaning of each noncombatant’s life to an anonymous unit of exchange that is abstractly equivalent in value to any other within the same general category of “civilian noncombatant,” then morally to evaluate the aggregate of those lives. But in relation to what? The problem is not only that nameless, faceless body counts are unlikely to arouse empathy, or that the flat-toned, journalistic synopses that accompany them mask gruesome and morally discomfiting details. Even if such details should happen to be forthcoming, the question remains: how we are to interpret them; by what ethical standards are we to give meaning to the details? Once a photograph, an article, a newscast, or a blog “fills in” the statistical unit with human substance (with individual character traits, a life story, personal aspirations, and so on), we begin to sense that the human particularity of that life—its moral “weight,” as it were—is beyond measure. Imagined, through an act of empathic identification, in all its uniqueness, and experienced, from the “inside,” as at once uncannily like and irreducibly unlike all that is “me” or “mine,” the intentionally/unintentionally extinguished human life becomes an absolute. It confronts us with the untranscendable distance between the firsthand, first-person experience of tragedy on the one side and the languages of bureaucracy, technocracy, and necro-economics on the other.

This is not to say that we should simply stop (if, indeed, we have truly even begun) keeping careful records of casualty figures; those figures belong to an epistemological context without which it would be impossible to get any kind of ethical conversation off the ground. Rather, it is our moral responsibility, even as we count dead bodies, to recognize the fundamental nonequivalence of each human life. If, that is to say, each human life is equal in value to every other human life, it is in the sense that the “other” life is just as precious—just as irreplaceable in the eyes of certain others—as our own life is to us, to our family, our friends, or our fellow citizens. From this identificatory, or “subjectivist,” standpoint, the difference between ten dead civilians and one hundred, or between one hundred and a thousand, is—like the value of each and every individual—inherently incalculable. By all means, then, let us count dead bodies, recognizing that, by any empirical ethical calculus, to cause less civilian death is always better than to cause more civilian death. But let us at the same time recognize the futility, indeed the perversity, of quantifying and rationalizing death in an empty economy of means and ends, where means and ends are highly unpredictable and fundamentally irreducible. Holding simultaneously in mind these two irreconcilable but crucial perspectives—the one experiential, the other analytical—we are, I think, less likely to commit the blunder of sacrificing clear legal definitions and deontological moral principles to hubristic and self-defeating political causes.

In his previously mentioned remarks at the National Defense University, President Obama, referring to the contending reports of civilian casualties from drone strikes, offers in passing, in a speech otherwise devoted to an objectivist moral economy, an acknowledgement of the subjective, experiential ontology: “it is a hard fact that U.S. strikes have resulted in civilian casualties,” says the president, noting that such deaths are “a risk that exists in every war.” “And,” he continues, “for the families of those civilians, no words or legal construct can justify their loss. For me, and those in my chain of command, those deaths will haunt us as long as we live, just as we are haunted by the civilian casualties that have occurred throughout conventional fighting in Afghanistan and Iraq.” The president quickly moves on, however, from his brief confession of private torment over the unjust human toll of “every war.” “But as Commander-in-Chief,” he says, “I must weigh these heartbreaking tragedies against the alternatives.” Thus, as if suddenly rising above an obligatory moment of sentimental self-indulgence, Obama conjures away the “heartbreaking” referent and, returning to the language of “hard facts” and dispassionate calculation, reaffirms the power of elastic “legal constructs” to “weigh” the “risk” of civilian casualties against the somehow foreordained “alternative” of “far more civilian casualties” at the hands of terrorists. By asserting the priority of the hypothetically calculable over the experientially incalculable, President Obama’s words invert the logic of the moral sensibility and lend authority to a regime of legal opacity.⁶⁰

Let us add, by way of a coda, a few observations regarding the military adventurism of Obama’s successor to the White House. With the accession of Donald Trump to the presidency, any hint of compassion, moral conscience, or legal restraint has been superseded by crass bellicosity and self-congratulatory callousness. Trump has pledged to “bomb the shit out of” the Islamic State—”I’d blow up every single inch; there would be nothing left.”⁶¹ In the first counterterrorism operation Trump authorized after taking office, thirty people, including at least eight women and seven children, were killed as US helicopter gunships, Reaper drones, and Special Operations forces launched a dawn raid against “Qaeda militants” in Yemen. Among the dead were two Americans—one, a US Navy SEAL, the other, the eight-year-old daughter of Anwar Al-Awlaki.⁶² In March of 2017, there were approximately forty drone strikes in Yemen, including twenty-five on a single day. And whereas Obama conducted one strike every 5.4 days, Trump averaged one strike or raid every 1.25 days, an increase of 432 per cent over his predecessor.⁶³

In this atmosphere of escalating belligerence, even the modest guidelines of Obama’s Executive Order are being rescinded, as senior officials of the National Security Council conduct a review aimed at lowering the threshold for “acceptable” civilian casualties and at scaling back constraints such as the requirement of a continuing and imminent threat to the US.⁶⁴ In the interest of “streamlining” the decision-making process, moreover, the authority to approve individual drone strikes has been transferred from the White House to the Pentagon and the CIA, an agency that operates under official secrecy and outside the bounds of the laws of war. Indeed, since the CIA personnel and private contractors who are engaged in staging drone strikes are civilians and, as such, are not subject to the Uniform Code of Military Justice, their participation in the deployment of lethal force violates customary international law. Specifically, the 1923 draft Hague air warfare regulations stipulate that belligerent rights may be exercised only by military aircraft bearing external signs of their nationality and of their “military character,” and that these aircraft must be under the command of duly commissioned military personnel and manned by an “exclusively military” crew wearing “a fixed distinctive emblem.”⁶⁵

As civilians participating in hostilities, then, drone operators (including those involved in tactical reconnaissance and intelligence gathering) lose both their immunity from attack and their “combatant’s privilege,” that is, their immunity from domestic laws prohibiting and penalizing the use of lethal force. Accordingly, not only the operators themselves but also the government officials overseeing the drone program are potentially liable to prosecution for murder, both in the US and in the home state of the victim(s).⁶⁶ Even worse, if the US government’s assumption that “unprivileged belligerency constitutes a war crime per se” should happen to be correct, then “all those responsible for the CIA program both may be tried by any country exercising universal jurisdiction over war crimes and may be prosecuted by military tribunals rather than ordinary civilian courts.”⁶⁷ As Gary Solis puts it, “those C.I.A. agents are, unlike their military counterparts but like the fighters they target, unlawful combatants. No less than their insurgent targets, they are fighters without uniforms or insignia, directly participating in hostilities, employing armed force contrary to the laws and customs of war.”⁶⁸ Thus, by scrapping Obama’s already largely rhetorical “playbook” and shifting the power of life and death toward the least transparent and least accountable of bureaucracies, the Trump administration ensures that we will recognize ourselves in the image of the enemy we seek to destroy.