But it is his theory of the drone that is even more chilling. It demands that we consider the emergence of a new ethical and political norm of war that is neither war as we know it—nor peace. The ‘principle of immunity for the imperial combatant’ rests on a twisted logic: On the one hand is the achieved capacity of the drone operative (one of many newly installed masters of ‘lethal surveillance’) to move throughout a day between killing fields and coffee breaks, between combat zones and home. On the other hand is the enlisting of a citizenry to accept the ‘moral obligation’ to kill.

It demands that we consider the emergence of a new ethical and political norm of war that is neither war as we know it—nor peace. The 'principle of immunity for the imperial combatant' rests on a twisted logic: On the one hand is the achieved capacity of the drone operative (one of many newly installed masters of 'lethal surveillance') to move throughout a day between killing fields and coffee breaks, between combat zones and home. On the other hand is the enlisting of a citizenry to accept the 'moral obligation' to kill.

The work here is extremely boring. Men pass whole nights watching a screen on which, for the most part, appear unchanging images of another desert on the other side of the planet. Eating Doritos and M&Ms, they wait for something to happen: “months of monotony and milliseconds of mayhem.”

The passengers traveling in three vehicles that, a few hours ago, left their little village in the province of Daikundi have no idea that for quite some time now, dozens of eyes have been watching them. Among those invisible spectators are not only the pilot and sensor operator but also a mission intelligence coordinator, a safety observer, a team of video analysts, and a ground force commander, the last of whom will eventually give the go-ahead for an aerial strike. This network of eyes remains in constant communication with one another. And on this night of February 20, 2010, their conversation is, as usual, recorded

PILOT: They were all PIDed as males, though. No females in the group. … SENSOR OPERATOR: That guy looks like he’s wearing jewelry and stuff like a girl, but he ain’t . . . if he’s a girl, he’s a big one.

Self-preservation by means of drones involves putting vulnerable bodies out of reach. This could be seen as the fulfillment of the ancient desire that inspires the whole history of ballistic weapons: to increase one's reach so as to hit the enemy from a distance before the opponent can launch its own attack. But with drones, the weapon's range (the distance between the weapon and its target) has been increased by the range of the remote control (the distance separating the operator from the weapon). Thousands of miles can now be interposed between the trigger on which one's finger rests and the cannon from which the cannonball will fly.

However, "projection of power" is also a euphemism that obscures the facts of wounding, killing, destroying. And to do this "without projecting vulnerability" implies that the only vulnerability will be that of the enemy, reduced to the status of a mere target. Underlying the palliative military rhetoric, as Elaine Scarry detects, the real claim is that the "successful strategy is one in which the injuring occurs only in one direction. . . . Thus, the original definition, which seems to posit noninjuring against injuring, instead posits one-directional injuring against two-directional injuring." By prolonging and radicalizing preexisting tendencies, the armed drone goes to the very limit: for whoever uses such a weapon, it becomes a priori impossible to die as one kills. Warfare, from being possibly asymmetrical, becomes absolutely unilateral. What could still claim to be combat is converted into a campaign of what is, quite simply, slaughter.

At the time of writing, the American armed forces had at their disposal more than six thousand drones of various kinds; more than 160 of these were Predator drones in the hands of the U.S. Air Force. For both the military and the Central Intelligence Agency (CIA), the use of hunter-killer drones has become commonplace, to the point of being routine. These machines are deployed not only in zones of armed conflict, such as Afghanistan, but also in countries officially at peace, such as Somalia, Yemen, and above all Pakistan, where CIA drones carry out on average one strike every four days. Exact figures are very hard to establish, but in Pakistan alone estimates of the number of deaths between 2004 and 2012 vary from 2,640 to 3,474.

A guiding thread is a thought expressed by the philosopher Simone Weil in the 1930s: "the most defective method possible," she warned, would be to approach warfare and the phenomena of armed violence "in terms of the ends pursued and not by the nature of the means employed." On the other hand, "the very essence of the materialist method is that, in its examination of any human event whatever, it attaches much less importance to the ends pursued than to the consequences necessarily implied by the working out of the means employed." Rather than hastening to seek possible justifications—in other words, rather than moralizing—she advised doing something quite different: Begin by taking apart the mechanism of violence. Go and look at the weapons, study their specific characteristics. Become a technician, in a way. But only in a way, for the aim here is an understanding that is not so much technical as political. What is important is not so much to grasp how the actual device works but rather to discover the implications of how it works for the action that it implements. The point is that the means adopted are binding, and a combination of specific constraints is associated with each type of means adopted. Those means not only make it possible to take action but also determine the form of that action, and one must find out how they do so.

But that is not all. By inventing the armed drone one has also, almost inadvertently, discovered something else: a solution to the central contradiction that for several centuries has affected the modern theory of political sovereignty in matters of warfare. The generalization of such a weapon implies a change in the conditions that apply in the exercise of the power of war, this time in the context of the relations between the state and its own subjects. It would be mistaken to limit the question of weaponry solely to the sphere of external violence. What would the consequences of becoming the subjects of a drone-state be for that state's own population?

In 1964, the engineer John W. Clark produced a study of “remote control in hostile environments”: “When plans are being made for operations in these environments, it is usual to consider only two possibilities: either placing a machine in the environment or placing a protected man there. A third possibility, however, would in many cases give more satisfactory results than either of the others. This possibility employs a vehicle operating in the hostile environment under remote control by a man in a safe environment.” Rather than deep-sea divers or autonomous machines, one could use remotely controlled machines or what Clark, forging an awkward neologism based on ancient Greek roots, called “telechiric machines,” or “technology of manipulation at a distance.”

The hostile zone, for its part, remains a space that is left derelict but which, as a potentially threatening area, definitely needs to be kept under surveillance. It may even be exploited for its resources, but it is not, strictly speaking, to be occupied. One intervenes there and patrols it, but there is no suggestion of going to live there—except to carve out new secured zones, bases, or platforms in accordance with a general topographical schema and for reasons of security.

However, the reader, who was no fool, concluded with a quite different scenario, one that, sad to say, was far more realistic: “Far-flung imperial conquests which were ours because we had the Maxim gun and they had the knobkerry will be recalled by new bloodless triumphs coming our way because we have telechiric yeomanry and they, poor fuzzy-wuzzies, have only napalm and nerve-gas.”

It would be a utopia, with warfare converted into a tournament of machines—battles without soldiers, conflicts with no victims. However, the reader, who was no fool, concluded with a quite different scenario, one that, sad to say, was far more realistic: “Far-flung imperial conquests which were ours because we had the Maxim gun and they had the knobkerry will be recalled by new bloodless triumphs coming our way because we have telechiric yeomanry and they, poor fuzzy-wuzzies, have only napalm and nerve-gas.”

Once the remotely controlled machine becomes a weapon of war, it is the enemy who is treated as a dangerous material. He can be eliminated from afar as one watches on a screen, softly enclosed within a climatized “safe zone.” Asymmetrical warfare becomes radicalized, unilateral. Of course people would still die, but only on one side.

However, it was a long time before drones were to be seen cruising above battlefields. To be sure, the idea dates back quite a while: there were the Curtiss-Sperry aerial torpedo and the Kettering Bug at the end of World War I, and then the Nazi V-1s and V-2s unleashed on London in 1944. But those old flying torpedoes may be considered more as the ancestors of cruise missiles than as those of present-day drones. The essential difference lies in the fact that while the former can be used only once, the latter are reusable. The drone is not a projectile, but a projectile-carrying machine.

Barely two months after the outbreak of hostilities in Afghanistan, George Bush was in a position to declare: “The conflict in Afghanistan has taught us more about the future of our military than a decade of blue ribbon panels and think-tank symposiums. The Predator is a good example. . . . Now it is clear the military does not have enough unmanned vehicles.”

The triggers of moral indignation are quite mysterious sometimes. While the virtual hunting of animals was almost universally condemned as scandalous, the remote-controlled hunting of human beings was at the same moment taking off without any of those same people making any objections.

In 2001, U.S. secretary of defense Donald Rumsfeld had become convinced that “the techniques used by the Israelis against the Palestinians could quite simply be deployed on a larger scale.” What he had in mind was Israel’s programs of “targeted assassinations,” the existence of which had recently been recognized by the Israeli leadership. As Eyal Weizman explains, the occupied territories had become “the world’s largest laboratory for airborne thanatotactics,” so it was not surprising that they would eventually be exported.

Within the United States, not all the high-ranking officers who were informed of these plans greeted them with enthusiasm. At the time, journalist Seymour Hersh noted that many feared that the proposed type of operation—what one advisor to the Pentagon called “preemptive manhunting”—had the potential to turn into another Phoenix Program, the sinister secret program of murder and torture that had once been unleashed in Vietnam.

Contrary to Carl von Clausewitz’s classical definition, the fundamental structure of this type of warfare is no longer that of a duel, of two fighters facing each other. The paradigm is quite different: a hunter advancing on a prey that flees or hides from him. The rules of the game are not the same. “In the competition between two enemy combatants,” wrote Crawford, “the goal is to win the battle by defeating the adversary: both combatants must confront to win. However, a manhunt scenario differs in that each player’s strategy is different. The fugitive always wants to avoid capture; the pursuer must confront to win, whereas the fugitive must evade to win.” The hostile relationship now boils down, as in a game of hide-and-seek, to “a competition between the hiders and the seekers.”

The primary task is no longer to immobilize the enemy but to identify and locate it. This implies all the labor of detection. The modern art of tracking is based on an intensive use of new technologies, combining aerial video surveillance, the interception of signals, and cartographic tracking.

The political rationale that underlies this type of practice is that of social defense. Its classic instrument is the security measure, which is “not designed to punish but only to preserve society from the danger presented by the presence of dangerous beings in its midst.” In the logic of this security, based on the preventive elimination of dangerous individuals, “warfare” takes the form of vast campaigns of extrajudiciary executions. The names given to the drones—Predators (birds of prey) and Reapers (angels of death)—are certainly well chosen.

Optical surveillance is not limited to the present time. It also assumes the important function of recording and archiving. “The idea behind persistent surveillance is to make a movie of a city-size area, with the goal of tracking all the moving vehicles and people,” says John Marion, director of the persistent surveillance program for Logos Technologies. Once such a movie of every life and everything is completed, it could be rerun thousands of times, each time focusing on a different person, zooming in on him or her so as to reexamine that person’s own particular history. One could select scenes, rewind, replay, or fast-forward, navigating as one wished through not only space but also time. Once an event had taken place, one could backtrack to study its genealogy. For example, “if a whole town could be surveilled at once, . . . car bombs could be traced back to their points of origin.”

In an effort to resolve this problem, the Pentagon went to the sports stadium. The production of football broadcasts has resulted in a variety of innovative technologies. In every game, dozens of cameras film the players from every angle. Every sequence is instantly indexed on a database. Thanks to efficient software, the control room staff can run replays from a variety of angles while displaying statistics on the screen. As Larry James, Air Force deputy chief of staff for intelligence, surveillance, and reconnaissance, explains, “When it comes to collecting and analyzing data, sports broadcasters are far ahead of the military.” After sending emissaries into ESPN’s studios, the U.S. military decided to acquire a modified version of the software that it used. After all, their concerns are similar: “While sportscasters want to collect and catalog video on a specific player or a winning shot, the military wants the same capacity to follow insurgents and bombings.”

Derek Gregory notes that the ability to integrate data produced by a variety of sources—“combining the where, the when, and the who”—into a three-dimensional array “replicates the standard time-geography diagrams developed by the Swedish geographer Torsten Hägerstrand in the 1960s and 1970s.” This extremely inventive development in human geography set out to draw maps of a new kind, spatiotemporal graphs that would show the course of lives in three dimensions, with all their cycles and itineraries but also their accidents and deviations. In a cruel perversion, this idea of a cartography of lives has today become one of the main epistemic bases of armed surveillance. The aim is to be able “to follow several individuals through various social networks in order to establish a form or pattern of life that conforms with the paradigm of ‘information based on activity,’ which today constitutes the heart of the counter-insurgency doctrine.”

Predicting the future is based on knowledge of the past. The archives of lives constitute the basis for claims that, by noting regularities and anticipating recurrences, it is possible both to predict the future and to change the course of it by taking preemptive action. Such claims are clearly founded upon very fragile epistemological bases, which in no way prevents them from being extremely dangerous but, on the contrary, ensures that they are.

The names given to these devices are very revealing: Argus26 and Gorgon Stare. In Greek mythology, Argus, the figure with a hundred eyes, was also known as Panoptes, “the one who sees all.” Jeremy Bentham’s panopticon, analyzed by Michel Foucault, was originally an architectural contraption. In a carrying forward of this pattern, in recent decades cities have been stuffed with video surveillance cameras. Surveillance by means of drones is more economical, as it involves no spatial alterations, nor does it require anything to be affixed to walls. Air and sky are all that are needed. As in the film Eyeborgs, the cameras are detached from walls and thereupon acquire wings and weapons. We are entering into the era of winged and armed panoptics. As for the gaze of the Gorgon, it turned to stone all those unfortunate enough to encounter it. It was a gaze that killed. At this point, it is a matter no longer of surveillance and punishment but of surveillance and annihilation.

David Rohde, a New York Times journalist kidnapped in 2008 and held in Waziristan for seven months, was one of the first Westerners to describe the effects that this lethal continuous surveillance produced upon the populations subjected to it. Evoking a “hell on earth,” he added: “The drones were terrifying. From the ground, it is impossible to determine who or what they are tracking as they circle overhead. The buzz of a distant propeller is a constant reminder of imminent death.”

One inhabitant of Datta Khel—a place hit more than thirty times by drones in the course of the past three years—says that his neighbors “have lost their mental balance . . . are just locked in a room. Just like you lock people in prison, they are locked in a room.” Drones are indeed petrifying. They inflict mass terror upon entire populations. It is this—over and above the deaths, the injuries, the destruction, the anger, and the grieving—that is the effect of permanent lethal surveillance: it amounts to a psychic imprisonment within a perimeter no longer defined by bars, barriers, and walls, but by the endless circling of flying watchtowers up above.

Apart from these “personal strikes,” there are also “signature strikes,” here meaning strikes authorized on the basis of traces, indications, or defining characteristics. Such strikes target individuals whose identity remains unknown but whose behavior suggests, signals, or signs membership in a “terrorist organization.” … In such cases, the strike is made “without knowing the precise identity of the individuals targeted.” It depends solely on their behavior, which, seen from the sky, appears to “correspond to a ‘signature’ of pre-identified behavior that the United States links to militant activity.” Today, strikes of this type, against unknown suspects, appear to constitute the majority of cases.

Each and every person has a particular form or pattern of life. Your daily actions are repetitive, your behavior has certain regularities. For example, you rise at roughly the same hour and regularly make the same journey to work or elsewhere. You frequently meet up with the same friends in the same places. If you are placed under surveillance, it is possible to record all your movements and establish a spatiotemporal map of all your usual doings. Furthermore, by intercepting your telephone calls, observers can superimpose your social network upon this map, determine which are your personal links, and calculate the importance of each one in your life. As an American army manual explains: “While the enemy moves from point to point, reconnaissance or surveillance tracks and notes every location and person visited. Connections between those sites and persons to the target are built, and nodes in the enemy’s network emerge.” Once this network of places and links in your life is established, it will be possible to predict your behavior: if it is not raining, on Saturday you will probably go jogging in a particular park at a particular time. But an observer may also perceive suspicious irregularities: today you have not followed your usual route, and you have met with someone in an unusual place. Any interruption of the norm that you yourself have established by your habits, any departure from your regular behavior, can sound an alarm bell: something abnormal and therefore potentially suspect is happening.

This method stems from activity-based intelligence, or ABI. From the mass of information collected about a particular individual, group, or place gradually emerge patterns, or traceable themes. Activity becomes an alternative to identity. Once a target has been named, instead of trying to localize it, do quite the opposite. Start by establishing surveillance and gathering information. Next, make large-scale graphs to do an analysis of “big data,” picking out nodular points that, by reason of the position and scale they occupy on the diagram, can be identified as threats that need to be neutralized. “By compiling activity-based association data with its metadata over time and adding analysis and reporting from many analysts,” wrote Keith L. Barber of the National Geospatial-Intelligence Agency, “a rich archive will be formed to harvest patterns of life, networks and abnormalities which may have been overlooked otherwise.” The tools of human geography and the sociology of social networks are now enlisted in the service of a policy of eradication in which “persistent surveillance” makes it possible to pick out dangerous individuals. The painstaking work of establishing an archive of lives progressively gathers together the elements of a file that, once it becomes thick enough, will constitute a death warrant.

Both the means and the methodology are patently limited. As a former CIA officer admits, “You can only see so much from 20,000 feet.” A drone can distinguish shapes only more or less imprecisely. For example, in April 2011, American drones were “unable to discriminate the highly distinctive combat outline of two Marines (with full battle equipment) from the irregular enemy.” A telling joke made in the corridors of American power went, “When the CIA sees three guys doing jumping jacks, the agency thinks it’s a terrorist training camp.”

On March 17, 2011, an American strike decimated a group of men meeting in Datta Khel, Pakistan, on the grounds that “they acted in a manner consistent with AQ [al-Qaeda]-linked militants.” The manner of their gathering corresponded to that predefined as resembling terrorist behavior. But the meeting observed from the skies was actually a traditional assembly, a jirga, convoked to resolve a disagreement in the local community. Seen from the sky, a village meeting looks just like a gathering of militants. Between nineteen and thirty civilians are estimated to have perished in the attack.

This profiling method works only with schemas. And a single schema may, by definition, correspond to a number of heterogeneous phenomena. Imagine that you see a shadow resembling a huge dog. If you have access only to the shadow, how can you tell with certainty what object created it? It may simply have been made by an arrangement of someone’s hands as part of a shadow play.

It is nevertheless on the strength of such epistemological bases that “signature strikes” are today made by American drones. The authorities have built themselves a theater of shadows, but “the result, way too often, is firing blind based on ‘pattern of life’ indicators, without direct confirmation that the targets are, in fact, who we think they are—killing innocent people in the process.”

With the concept of a “global war against terror,” armed violence has lost its traditional limits: indefinite in time, it is also indefinite in space. The whole world, it is said, is a battlefield. But it would probably be more accurate to call it a hunting ground. For if the scope of armed violence has now become global, it is because the imperatives of hunting demand it.

While warfare is defined, in the last analysis, by combat, hunting is essentially defined by pursuit. Two distinct types of geography correspond to the two activities. Combat bursts out wherever opposing forces clash. Hunting, on the other hand, takes place wherever the prey goes. As a hunter-state sees it, armed violence is no longer defined within the boundaries of a demarcated zone but simply by the presence of an enemy-prey who, so to speak, carries with it its own little mobile zone of hostility.

The two principles of this paradoxical dismemberment are the following: (1) The zone of armed conflict, having been fragmented into miniaturizable kill boxes, tends ideally to be reduced to the body of the enemy or prey. That is, his body becomes the battlefield. This is the principle of precision or specification. (2) In order for the pursuit and surgical strikes to be carried out, this mobile microspace must be able to be aimed wherever necessary—so the whole world becomes a hunting ground. That is the principle of globalization or homogenization. According to the military and the CIA, it is because we can aim at our targets with precision that we can strike them down wherever we choose, even outside any war zone.

Even before the advent of the micromachines of the future, drone partisans are already emphasizing the technological precision of their weapons. But the paradox is that they use this supposed gain in precision to extend the field of fire to take in the entire world. What we find here is a double movement that seizes upon the spatiolegal notion of an armed conflict zone in a way that tends to dislocate it almost completely. The two principles of this paradoxical dismemberment are the following: (1) The zone of armed conflict, having been fragmented into miniaturizable kill boxes, tends ideally to be reduced to the body of the enemy or prey. That is, his body becomes the battlefield. This is the principle of precision or specification. (2) In order for the pursuit and surgical strikes to be carried out, this mobile microspace must be able to be aimed wherever necessary—so the whole world becomes a hunting ground. That is the principle of globalization or homogenization. According to the military and the CIA, it is because we can aim at our targets with precision that we can strike them down wherever we choose, even outside any war zone.

As Derek Gregory points out, one of the problems is that the “legal logic through which the battlespace is extended beyond the declared zone of combat in Afghanistan is itself infinitely extendable.” By redefining the notion of armed conflict as a mobile place attached to the person of the enemy, one ends up, under cover of the laws of armed conflict, justifying the equivalent of a right to execute suspects anywhere in the world, even in zones of peace, illegally and without further procedures, one’s own citizens included.

As the jurist Mary Ellen O’Connell, who describes the present-day drone strikes in Pakistan, Somalia, and Yemen as illegal, reminds us: “Drones launch missiles or drop bombs, the kind of weapons that may only be used lawfully in armed conflict hostilities.” The fact is that “there was no armed conflict on the territory of Pakistan because there was no intense armed fighting between organized armed groups. International law does not recognize the right to kill with battlefield weapons outside an actual armed conflict. The so-called ‘global war on terror’ is not an armed conflict.” These strikes therefore constitute grave violations of the laws of war.

It is fighting by means of terror, and no attempt is made to disguise the fact. Says Dunlap, “American precision airpower is analogous (on a much larger and more effective scale) to the effect that insurgents try to impose . . . through the use of improvised explosive devices.” The point could not be made more clearly: at a tactical level (and setting aside technological sophistication), drone strikes are equivalent to bomb attacks. They constitute the weapons of state terrorism.

Evidence provided by a Pakistani Taliban leader, Baitullah Mehsud, illustrates the apparent truth of that thesis: “I spent three months trying to recruit and only got 10–15 persons. One US attack and I got 150 volunteers.”

The partisans of the drone as a privileged weapon of “antiterrorism” promise a war without losses or defeats. What they fail to mention is that it will also be a war without victory. The scenario that looms before us is one of infinite violence, with no possible exit; the paradox of an untouchable power waging interminable wars toward perpetual war.

The message of these myths is that invulnerability is precisely that, a myth. There is always one unforeseen weak point, one flaw. He has felled a dragon but will die from a fallen leaf. The lesson is not only that invulnerability can never be total, but also that any attempt to achieve invulnerability in turn engenders a corresponding vulnerability. It is by grasping Achilles’ body in order to plunge it into the river that Thetis makes it invulnerable and at the same time produces its vulnerable point, which is the spot at which she grasped it. With regard to invulnerability and vulnerability, these two, far from excluding one another, each summon up the other.

Mastery over the airwaves is also necessary. In 2009, the press reported that Iraqi insurgents had managed to intercept the video feeds transmitted by Predator drones. To accomplish this, all they needed was a satellite antenna and software that could be purchased on the Internet for $26. Convinced of their own technological superiority, the American military had apparently not taken the elementary precaution of effectively encrypting their transmissions.

The Israeli army recently realized that as a result of similar negligence, Hezbollah had over the past ten years developed the capacity to intercept video feeds from Israeli drones, which enabled that organization to, among other things, pinpoint the position of the Israeli battalions on the ground, the better to ambush them. Armed surveillance was, without the Israelis knowing it, lending its eyes to the enemy. One of the classic principles of guerrilla warfare is to supply oneself with weapons taken from the enemy camp. It is a rule that today is equally valid for the electromagnetic components of one's arsenal.

If the signals emitted by the drones have been so easily intercepted, it is not beyond the bounds of possibility that the flows of data that control them could likewise be hacked. The air pirates of the future will use software to crack codes and take control from a distance. In 2011 Wired magazine revealed that malware had infected the Creech Air Force Base computers, including those handled by drone operators. This involved a software spy of the keylogger variety, capable of recording keystrokes and transmitting them to a third party in such a way as to make it possible to recover passwords. That threat remained relatively benign, but it is obviously possible to envisage other scenarios. Just like any other connected computer system, the drone is vulnerable to intrusions. A computer army can be paralyzed by a viral attack more efficiently than by bombs.

Walter Benjamin did some thinking about drones, radio-controlled planes that the military thinkers of the mid-1930s were already imagining. He used this example to illustrate the difference between what he called the “first technique,” which could be traced back to prehistoric art, and the “second technique,” which was characteristic of modern industry. As he saw it, the distinction between them was not so much the inferiority or archaism of the one in comparison to the other but rather a “difference of trends.” “The first technique,” he wrote, “engages the human being as much as possible, the second as little as possible. The great technological feat of the first technique is, in a manner of speaking, the human sacrifice; that of the second lies along the lines of remote-controlled airplanes that don’t require any human crew.”

Contrary to the evolutionist schema that Benjamin, in truth, only suggested, the better to subvert it, the kamikaze and the drone—the weapon of sacrifice and the weapon of self-preservation—did not succeed each other chronologically, one following from the other as history follows from prehistory, On the contrary, they emerged together, as two opposed but historically simultaneous tactics.

This antagonism between the kamikaze and remote control reappears today: suicide bombings versus phantom bombings. The polarity is primarily economic. It sets those who have nothing but their bodies with which to fight in opposition to those who possess capital and technology. But these two regimes, the one tactical, the other material, also correspond to two different ethical regimes: the ethic of heroic sacrifice, on one hand, and the ethic of vital self-preservation, on the other.

Richard Cohen, a columnist at the Washington Post, sets out his view of the situation: “As for the Taliban fighters, they not only don’t cherish life, they expend it freely in suicide bombings. It’s difficult to envisage an American suicide bomber.” He asserts: “There is really no such thing as an American suicide bomber. We don’t extol the bomber and parade his or her children before the TV cameras so that other children will envy them for the death of a parent. This is odd to us. This is chilling to us. This is downright repugnant.” Then he adds complacently, “Maybe we have come to cherish life too much.”

Using new means, the drone procures for its operators an even greater sense of invulnerability. Today as yesterday, the radical imbalance in exposure to death leads to a redefinition of relations of hostility and of the very sense of what is called “waging war.” Warfare, by distancing itself totally from the model of hand-to-hand combat, becomes something quite different, a “state of violence” of a different kind. It degenerates into slaughter or hunting. One no longer fights the enemy; one eliminates him, as one shoots rabbits.

But every time such situations arose, certain contemporaries were troubled. Faced with a spectacle of armed violence that so manifestly contravened “the conventional understanding of war as an activity in which human dying and killing are exchanged,” it was not long before they expressed their indignation. But when these people began to express their objections too openly, they were often faced with a very ancient discursive strategy used to calm uneasy consciences and silence the most vociferous: the reassuring discourse of historical permanence. It is a matter of showing, with the backing of many historical examples, that there is nothing fundamentally new about this kind of situation, so it is perfectly acceptable.

Bell might well have added that the desire to take out one’s enemies from a safe distance was never better satisfied than in the “glorious” episodes of colonial wars, in which natives were felled en masse while the armies of the whites were hardly scratched. On the evening following the “battle” of Omdurman, in Sudan, on September 2, 1898, there were forty-eight dead on the side of the Anglo-Egyptian forces under Kitchener’s command, as against the nearly ten thousand dervishes cut down by the bursts of bullets from Maxim machine guns.

If there was any repugnance felt at the use of ignoble means, it was present only in situations where the conflict was between equals, rather than simply a matter of putting down inferiors. As Jünger reflects, “In every age a distinction has been drawn between two styles, a superior and a barbaric form of the law and the conventions of war. . . . In the Middle Ages, the Christian fleets could fire red hot cannon balls only when they encountered Turkish vessels. In the twentieth century, dum-dum bullets, proscribed in European theatres of operation, were used in colonial wars and the reason for this was said to be that lead bullets did not halt the frenzy of attacks by the ‘savages.’”

Freed from the constraints imposed by reciprocal relations, will the drone masters be able to continue to demonstrate virtue and to resist the temptation to commit injustice with virtually no sanction imposed? That question, to which we shall return, raises the matter of moral hazard.

If the drone is represented as being virtuous, that is because it makes it possible to rule out any possibility of casualties in one's own camp. That argument was recently summed up in a British report: "the use of unmanned aircraft prevents the potential loss of aircrew lives and is thus in itself morally justified." A comparison between this thesis of drones, which are deemed virtuous because they spare their agents any confrontation with death, and the classic maxims according to which military virtue is the precise opposite, is enough to reveal the scope of the revolution taking place in the field of values.

Although they do not risk their bodies in those operations, they do risk their mental health. This would be a specific form of bravery, defined no longer by the exposure of one's physical vulnerability to enemy violence, but by exposure of one's psychic vulnerability to the effects created by one's own destructiveness.

Alfred de Vigny wrote, "The paid man, the soldier, is a glorious pauper, victim, and executioner." A soldier wields violence and is also exposed to it; he is both an executioner and a victim. But what does he become once the very possibility of being exposed to violence is removed? The fatal conclusion is that he becomes a mere executioner. And that is why he must somehow continue to be a victim if he wishes to continue to be called a soldier. There is a difficulty, though: what could he be the victim of? The only possibility is that he is psychically the victim of having to act as an executioner. That is the only way he can be defined in his own eyes, as in those of society, as the fighter that he no longer is.

Nevertheless, there is a shadow darkening the media picture of empathetic drone operators suffering psychic trauma: it has no empirical basis. The military psychologist Hugo Ortega recently conducted a vast investigation into the subject. He subjected drone operators to psychological tests in order to determine their levels of stress and discover whether they might be affected by post-traumatic stress disorder. His conclusion:

Military psychologists could well save themselves both time and money, as there is no point in carrying out lengthy and costly inquiries in order to discover whether these pathologies, thus defined, are to be found among drone operators. For it is, by definition, impossible: the technology radically rules out or substantially modifies the only stress factors that are covered by the existing categories of disease.

Closer to our own times, the psychologist Rachel McNair has suggested expanding the overly narrow notion of PTSD by defining a condition called "Perpetration-Induced Traumatic Stress" (PITS). Noting that recent literature has focused almost exclusively on the traumas inflicted on passive victims by external forces, she tries to isolate the active component of the anxiety, the one that stems from the fact of having been the agent of violence, in fact a perpetrator of it. Within the mixed experience of a soldier, it is hard to sort out one particular element from the rest, but McNair studies cases of pure perpetration, for example the nightmares of executioners haunted by images of the last moments of their condemned prisoners. She does not cite the case of drone operators, as her book was published too early for that, but it looks as if that might be a good means by which to test out her idea, for it presents a case of pure perpetration, armed violence reduced solely to its active aspect and without any vital threat to its perpetrator.

The rapid development of new techniques of violence at a distance will doubtless reorient the psychoethical modes of problematizing the experience of war in Western societies. The first signs of such a reorientation are already detectable. In a state supplied with largely dronized armed forces, one will probably move on inexorably from a study of psychic traumas linked to violence personally suffered to a study of psychic wounds linked to violence personally perpetrated. A kind of clinic for executioners would thus develop alongside psychotherapies for assassins, all of which would be designed to deliver them from their unease.

This new combination of physical distance and ocular proximity gives the lie to the classic law of distance: the great distance no longer renders the violence more abstract or more impersonal but, on the contrary, makes it more graphic, more personalized.

The filtered nature of perception, the figurative reduction of the enemy, the nonreciprocity of the fields of perception, and the dislocation of the phenomenological unity of the action are all factors that, when combined, produce a strong “moral buffering” effect. Thus, offsetting the visual proximity, the device presents its operators with powerful means of distancing.

The filtered nature of perception, the figurative reduction of the enemy, the nonreciprocity of the fields of perception, and the dislocation of the phenomenological unity of the action are all factors that, when combined, produce a strong "moral buffering" effect. Thus, offsetting the visual proximity, the device presents its operators with powerful means of distancing. But this form of experience also presents a second important characteristic: the fact that the violence of warfare is being exercised from a peaceful zone.

And what if drone psychopathology lay not where it is believed to be, in the possible traumas of the drone operators, but in the industrial production of compartmentalized psyches, immunized against any possibility of reflecting upon their own violence, just as their bodies are already immunized against any possibility of being exposed to the enemy?

Simone Weil offered a very fine interpretation of the Platonic myth and reformulated it definitively for the present day. What is invisible, she said, is not the person who wears the ring, but the ring itself. “The ring of Gyges that has itself become invisible—that is precisely in what consists the act of setting apart. It is setting apart oneself and the crime one commits; not establishing the connection between the two.” What Gyges says is, “I have become king and the other king has been assassinated. The two things are totally unrelated. Here is the ring.” One sets something aside and forgets that one has set it aside, one compartmentalizes—and “that faculty of setting apart makes all crimes permissible.”

“No body bags,” “a phobia of losses,” “an aversion to risk”: in the late 1990s, massive use was made of all these expressions in the United States. They all conveyed the same idea: the tendency to subordinate the use of military force to a political imperative in order to preserve American military lives. What is specific about contemporary imperial violence has less to do with the asymmetry of force and the resulting unequal distribution of vulnerability—a classic feature of all the “small wars” of history—than with the type of norms that now shape the great Western “democratic” powers’ exercise of that violence. If there is anything new about the situation, it might lie in the fact that the practical quasi-invulnerability of the dominant camp was, at the end of the twentieth century, set up as the dominant ethical and political norm.

To this end, pilots were forbidden to fly below an altitude of 15,000 feet, a security distance that meant they were nearly out of reach of the enemy antiaircraft defenses and were almost untouchable. William Cohen, Bill Clinton’s defense secretary, later declared: “The paramount lesson learned from Operation Allied Force is that the well-being of our people must remain our first priority.” The NATO planes carried out 38,004 raids in seventy-eight days without sustaining a single loss among the members of their crews. The Americans had indeed managed to invent warfare that inflicted zero deaths in their own camp. However, this did not rule out contradictions. For the very altitude that ensured that the lives of the pilots were not exposed also meant a potential loss of accuracy for the air strikes. NATO officials questioned on this matter by Amnesty International acknowledged this: “An aircrew flying at 15,000 feet would be able only to identify whether the objective was the intended one according to the planning preparations, but would be unable to tell whether, for example, civilians had moved within its vicinity. The 15,000-feet rule thus effectively made it impossible for NATO aircrews to respect the obligation to suspend an attack once circumstances had changed on the ground rendering the objective no longer legitimate.”

It was a tension that journalist Michael Ignatieff summed up as follows: “High tech warfare is governed by two constraints—avoiding civilian casualties and avoiding risks to pilots—that are in direct contradiction. To target effectively you have to fly low. If you fly low, you lose pilots. Fly high and you get civilians.” Which of those two imperatives should win out? It was a problem of priorities, a hierarchization of norms. NATO responded with little hesitation, awarding priority to the lives of pilots, even if it increased the “collateral” risk of wounding and killing civilians. In the name of preserving military lives, the risk of producing more casualties among the civilians was accepted, even though it was those very civilians whom the operation was designed to protect. This amounted to admitting that, in military-political reasoning, the life of a Kosovar civilian was of less value than that of an American service member.

Philosophers familiar with theories of “just warfare” were faced with this upsetting aspect of “warfare without risk.” Among those who detected a fundamental normative swing here, reactions were strong, even scandalized. The fact was that such a choice overturned the very principles of the standard war ethic. Wrote Jean Bethke Elshtain: “American officials described that intervention as a moral imperative. Yet before the conflict had even ended, observers were wondering if the United States had turned moral tradition on its head, with combatants rather than non-combatants provided immunity from the effects of fighting.” In her indignation, she had seized upon an essential point. What had emerged here, very clearly, was the preeminence of a tacit normative principle that was alien to the law of armed conflicts but implicitly regarded as superior to it, namely a sui generis principle of the immunity of imperial combatants.

That principle of immunity for the imperial combatant, which was presented in a more or less implicit, pragmatic fashion in the 1990s, has since been theoretically formalized. The doctrine was elaborated by Israel: “When senior Israeli Defense Forces [IDF] officers are asked about the killing of hundreds of Palestinian civilians during the fighting in the Gaza strip, they almost all give the same answer: The use of massive force was designed to protect the lives of the soldiers, and when faced with a choice between protecting the lives of Israeli soldiers and those of enemy civilians . . . , the soldiers take precedence.” That response was by no means merely improvised but was based on “an ethical theory, developed several years ago, that justifies its actions.” What’s the use of moral philosophy? Among other things, it facilitates the waging of war.

In 2005, he, together with Major General Amos Yadlin, coauthored an article entitled “Military Ethics of Fighting Terror.” Their aim was to revise, from top to bottom, the established principles underlying the ethics and law of armed conflicts. The most radical of their attacks concerned the principle of noncombatant immunity: “According to the ordinary conception underlying the distinction between combatants and non-combatants, the former have a lighter package of state duties than the latter. Consequently, the duty to minimize casualties among combatants during combat is last on the list of priorities. . . . We reject such conceptions because we consider them to be immoral. A combatant is a citizen in uniform. . . . His blood is as red and thick as that of citizens who are not in uniform. His life is as precious as the life of anyone else.” From this, one was supposed to conclude that the preservation of the lives of the subjects of the nation-state is the supreme duty, which must in all cases and at whatever cost be placed above the duty to minimize the losses of noncombatants in the enemy camp. This meant that in a war situation, minimizing risks for an Israeli soldier took precedence over the duty to minimize “collateral risks” for a child in Gaza. The life of the former, armed to the teeth though he was, was ruled to matter more than the life of the latter. This was established in the style of cold violence peculiar to this kind of “ethical” discourse, which mimes the formal rigor of analytical philosophy.

The argument is paradoxically supported by a rhetorical reminder of the equal value of all lives. Because, however, that equality holds only among citizens, the result is a hierarchization of bodies based on whether they are subjects of the nation-state or not. The operation consists in replacing the primacy of the structural distinction between civilians and combatants by another that simultaneously redefines it, one that boils down to a hierarchical separation between nationals and foreigners—and all in the name of an “ethic” that is just a polite euphemism for a nationalism of the most ferocious kind.

We need to assess the full scale of this assault. The project is nothing less than a dynamiting of the law of armed conflict as it was established in the second half of the twentieth century: an evisceration of the principles of international law in favor of a nationalism of self-preservation. And, as we shall see, that is also the primary principle of the necroethics of drones.

According to this view of military morality, to kill while exposing one's own life is bad; to take lives without ever endangering one's own is good. That first principle of drone necroethics is, paradoxically, vitalist. And it is in accordance with this logic that the drone can be said to be a "humanitarian" weapon: the humanitarian imperative is to save lives. And the drone does indeed save our lives. It is therefore a humanitarian technology. QED.

According to him, the drone is not just morally acceptable but "morally obligatory." If you wish to kill in conformity with moral law, you have to use a drone. His thesis is based on what he calls the "principle of unnecessary risk," according to which it is "wrong to command someone to take on unnecessary potentially lethal risk." His reasoning runs as follows: "I argue that we have a duty to protect an agent engaged in a justified act from harm to the greatest extent possible, so long as that protection does not interfere with the agent's ability to act justly. UAVs afford precisely such protection. Therefore, we are obligated to employ UAV weapon systems if it can be shown that their use does not significantly reduce a warfighter's operational capacity."

What is emerging here, more fundamentally, is a regime of military violence that claims to be humanitarian but might equally be called humilitarian. It is a power that both kills and saves, wounds and heals, and it performs those double tasks in a single gesture, in an integrated manner: an immediate synthesis of a power of destruction and a power of caring, murder at the same time as care.

Lives are saved. But saved from what? From oneself, from one's own power of death. The violence could have been worse, and since one tried in good faith to limit its deadly effects, one acted morally.

As Eyal Weizman has shown, this type of justification is essentially based on a logic of the lesser evil: "the essence of our humanitarian present, obsessed with the calculations and calibrations that seek to moderate, ever so slightly, the evils that it has largely caused itself." He notes that Hannah Arendt too warned against this type of argument: "Politically, the weakness of the argument has always been that those who choose the lesser evil forget very quickly that they chose evil."

However, that type of comparison is erroneous. To evaluate it properly, the drone should be set alongside weapons currently available for the same tactical function. For liquidating Osama bin Laden, the choice was between a drone and a commando raid, not between a drone and a Dresden-like bombing of Abbottabad. If one avoids being misled by some external attribute, the right form of comparison involves not a similarity of forms but an equivalence of functions. The drone is not a means of carpet bombing, and World War II bombers were not weapons designed for targeted assassination. The useful comparison here is not between a present-day flying weapon and an aerial weapon of the past, leading to the conclusion that progress has been made, but rather between this weapon and other present-day means for functions of the same order.

However, that type of comparison is erroneous. To evaluate it properly, the drone should be set alongside weapons currently available for the same tactical function. For liquidating Osama bin Laden, the choice was between a drone and a commando raid, not between a drone and a Dresden-like bombing of Abbottabad. If one avoids being misled by some external attribute, the right form of comparison involves not a similarity of forms but an equivalence of functions.

But the matter is still muddled on account of another confusion, a semantic one. Three close but not synonymous notions are blithely confused under the term “precision”: the accuracy of the firing, the extent of its impact, and the adequacy of the identification of its target.

It is estimated that the AGM-114 Hellfire fired by the Predator drone has a “kill zone” of 15 meters—which means that all those who happen to be within a radius of 15 meters around the point of impact, even if they are not the designated target, will die together with the target. As for the “wound radius,” that is estimated to be 20 meters.

However, those two criteria, direct participation in the hostilities and posing an imminent threat, are rendered completely ineffective by the exclusive use of drones. How can there be any direct participation in hostilities if there is no longer any fighting? How can there be any imminent threat if there are no longer any troops on the ground? By depriving the enemy of any possibility of participating directly in hostilities, one also deprives oneself of the surest means of recognizing an enemy as such. The paradox is that the drone, so highly praised for its great ability to make out the difference between combatants and noncombatants, in practice abolishes the very condition for that differentiation, namely combat. It is rather as if one were using an extremely powerful microscope whose visualization techniques deactivate the phenomenon it is supposed to be observing.

The paradox is that the drone, so highly praised for its great ability to make out the difference between combatants and noncombatants, in practice abolishes the very condition for that differentiation, namely combat. It is rather as if one were using an extremely powerful microscope whose visualization techniques deactivate the phenomenon it is supposed to be observing.

What we now have is an insidious switch from the category of “combatants” to that of “suspected militants.” This combatant = militant equivalence has the effect of extending the right to kill well beyond the classic legal boundaries and conferring an indefinite elasticity on the concept of a legitimate target.

The statistical miracle lay, as it often does, in the counting system employed. It was a simple but redoubtable trick, as New York Times journalists Jo Becker and Scott Shane revealed: “It in effect counts all military-age males in a strike zone as combatants . . . unless there is explicit intelligence posthumously proving them innocent.”

Once people’s minds are implanted with the false proof that the drone is in principle a more precise weapon and, on that account, supposedly more in conformity with the principle of distinction, the necroethics of the drone abandons any discussion of fundamental issues and steers critics toward a debate on numbers in which, by reversing the onus of proof, those critics now have to establish empirically that this weapon, which they have just conceded in theory to be a priori more ethical, has in practice produced quite the opposite results—a fact that, once one has swallowed the original postulate, remains totally inexplicable other than by contingent circumstances such as human misuse and errors in deployment. The instrument, in other respects, remains good in itself.

But what is not permissible, it seems to me, is what NATO did in the Kosovo war, where its leaders declared in advance that they would not send ground forces into battle, whatever happened." Then he drives the point home: "This is not a possible moral position. You can't kill unless you are prepared to die." The right to kill in this war thus seems in principle to depend upon a willingness to expose the lives of one's own soldiers, or at least not a priori excluding that possibility.

What he lays his finger on is an extreme case that he has to reject. But the reason it is so scandalous is not solely moral but also theoretical: the gut reaction of a theory faced with a phenomenon that, if accepted wholeheartedly, would threaten to destroy it utterly.

It is worth noting that the position that Walzer defines here traps the war leaders in a kind of double bind: on one hand, they are morally bound to minimize the risks to their own soldiers, but on the other, they are morally prohibited from eliminating that risk altogether. As soon as one succeeds in fully satisfying the first imperative by achieving zero risk, one hits bottom, in every sense of the term, since that virtuous minimization turns into its contrary and becomes the most extreme of moral scandals. This would be reaching a forbidden limit. However, that is not exactly what Walzer says. For him, this applies less to what soldiers do in an effort to eliminate risk for themselves than to the way their leaders declare that they will not put their soldiers in harm's way—which leaves open the question of whether doing so without declaring it would be more allowable. Anyhow, as he sees it, the problem consists in setting up the principle of "zero casualty warfare" in one's own camp as a moral norm. But why? It boils down to considering "that those lives are expendable and these are not." Therein lies the root of the scandal: by suggesting that the lives of the enemy are completely dispensable while ours are absolutely sacrosanct, one introduces a radical inequality in the value of lives, and this breaks with the inviolable principle of the equal dignity of all human lives.

It is remarkable that Walzer makes this reference to assassination the basis of a moral critique of the immunization of the national combatant in warfare at a distance. It is a thesis that relates to the ethics of assassination, not the ethics of warfare. On the surface, of course, what he is saying is that it is morally necessary to accept the principle of one's own exposure to risk in order to be able to kill in warfare. But, in truth, his palimpsest of citation is very well chosen, for how is it possible to justify homicide in a noncombat situation? For those seeking to justify such activities, there is only one place to turn: doctrines on political assassination. That is the lesson Walzer provides as a reader of Camus. On the surface, Walzer certainly does appear to make a clear misinterpretation, for he twists the nihilist maxim in order to turn it into the moral motto of classic warfare: "One cannot kill unless one is prepared to die." However, at an underlying level, he is making another point: the agents of "war without risks" are in reality like assassins who use bombs to kill their targets; unlike idealistic terrorists, they have chosen to be ready to kill only if they are certain that themselves will not die.

As Camus predicted, "Other men to come consumed with the same devouring faith as these, will find their methods sentimental and refuse to admit that any one life is the equivalent of another." And he went on to warn that then would come "the time of philosopher-executioners and of state terrorism."

The right to kill with impunity in war thus seems to be based upon a tacit structural premise: if one has the right to kill without crime, it is because that right is granted mutually. If I agree to confer upon another the right to kill me or my people with impunity, that is because I count on benefiting from the same exemption if I myself kill. The decriminalization of warrior homicide presupposes a structure of reciprocity. The killing is allowed only because it is a matter of killing each other.

But what happens to that right when there is no longer any effective possibility of reciprocation? In practice, “the basic equality of moral risk: kill or be killed” in traditional war was replaced by something akin to a “turkey shoot.” War degenerates into a putting-to-death. This is the situation introduced by the exclusive use of drones in asymmetrical warfare.

In the discourse of "applied military ethics," the whole discussion is reduced to the question of whether or not the use of armed drones can conform to the principles of the law of armed conflicts: is the use of this weapon potentially in conformity with the principles of distinction and proportionality? But what is forgotten is that because this weapon rules out combat and because it transforms war from being possibly asymmetrical into a unilateral relationship of death-dealing in which the enemy is deprived of the very possibility of fighting back, it surreptitiously slips out of the normative framework initially designed for armed conflicts. To apply norms designed for a conflict to slaughtering practices, and to be willing to pursue the discussion without questioning the presupposition that these practices still stem from within that normative framework, ratifies a fatal confusion of genres. As a result, the ethic of combat shifts and becomes an ethic of putting to death, a necroethic that makes use of the principles of jus in bello in order to convert them into criteria of acceptable murder: an ethic for butchers or executioners, but not for combatants.

The solution lies in nothing less than a forceful distortion of the law of warfare. To find a basis for a right to unilateral murder, there is only one theoretical possibility: to make the jus in bello follow from the jus ad bellum, making the former conditional upon the latter in a monopolistic way modeled upon an unconventional melding of police and penal practice, so as to reserve for the “just warrior” the right to kill without committing a crime.

This is what is proposed by Strawser and also by philosopher Jeff McMahan, both of whom totally reject the thesis of “the moral equality of the combatants” and replace it by a unilateral right to kill founded upon a concept of justa causa. “The warrior fighting for a just cause is morally justified to take the life of the enemy combatant,” writes Strawser, “whereas the unjust fighter is not justified, even if they follow the traditional principles of jus in bello [such as only targeting combatants and the like], to kill the justified fighter.” I have the right to kill, you do not. Why? Because I am just and you are unjust. I am good, you are bad, and only the good have the right to kill the bad. That, by and large, is what the puerile logic of this type of reasoning boils down to. The enemy’s reply to this will of course be that he is the good one, the opponent is the bad one, so it is in truth he who has the right to kill the other . . . and so it will continue until one of the two ends by winning the day, thereby providing, through force, irrefutable proof of his own right.

This artistic blurring has puzzled many jurists. They tell politicians they must declare whether it is a matter of war or of legitimate self-defense; they must choose. However, the administration refuses to do that because the consequence of resolving ambiguities in the framework of reference would entail the prohibition of the present drone strikes or would drastically limit their lawfulness.

This citizenship of exposable lives has constituted an important—albeit not the sole—basis for a democratic critique of the power to wage war, not only in the institutional manner forecast by Kant, namely suffrage, but also as a vector of extraparliamentary mobilization in the antiwar movements of the twentieth century, However, the dronization of the armed forces, inasmuch as it reduces the military exposure of national lives to zero, tends to somewhat deactivate that critical attitude. It would nevertheless be a mistake to believe that an “aversion to losses” is the only possible motive and the cost/benefit calculation the only relevant rationality for the development of a critical voice in the face of state violence.

Every protective power needs those it protects to be vulnerable, even—as all racketeers realize—to the point of actively maintaining this state of vulnerability. But unlike the discourse of protective sovereignty, which begins by introducing the postulate of ontological vulnerability as the original state of its subjects, the starting point of the present critical discourse is political vulnerabilization, the authority's exposure of the lives of its subjects, which becomes the basis for the possibility of it becoming the object of a critique or some kind of limitation.

Whether the risks of war are transferred to natives or to machines, Hobson's views still hold. The dronization of the armed forces, just like any other procedure that externalizes those risks, alters the conditions of decision making in warfare. Because the threshold of recourse to violence is drastically lowered, violence tends to be seen as the default option for foreign policy.

The main effect is to introduce a massive bias into his decision. Any agent who can take action with fewer risks to himself or his camp is likely to adopt a riskier pattern of behavior—that is to say, riskier for others. Similarly, the drone is a classic “moral hazard”—a situation in which being able to act without bearing the costs of the consequences relieves agents of responsibility for their decisions.

More precisely, we are told that drones introduce a threefold reduction in the costs traditionally attached to the use of armed force: a reduction of the political costs associated with the loss of national lives, a reduction of the economic costs associated with armament, and a reduction of the ethical or reputational costs associated with the perceived effects of the violence that is committed.

But if one changes one's theoretical spectacles, swapping those of orthodox economics for an analysis in terms of class relations, the phenomenon takes on a different aspect. What is at stake in a tendency to substitute capital for military endeavor is not only an upset in the conditions affecting the political calculations of the democratic sovereign but also, and more fundamentally, an increased social and material autonomization of the state apparatus.

In fact, contrary to the optimistic claims of “democratic pacifists,” implicating the lives of the population in war decisions was far from an adequate guarantee that military butcheries would be prevented. But even if it failed to rule out catastrophes, its effects were by no means minimal. The state's dependence on the bodies of the lower classes to wage war was also one of the factors that made it possible for those classes to establish a durable bargaining power. The social state was in part a product of the world wars, the price paid for the cannon fodder, compensation for the blood tax levied by the struggle.

In fact, modern welfare states, inadequate as they may be, are in no small part the product of war—that is, of governments' attempts to appease soldiers and their families. In the U.S., for example, the Civil War led to the institution of widows' benefits, which were the predecessor of welfare in its Aid to Families with Dependent Children form. It was the bellicose German leader Otto von Bismarck who first instituted national health insurance.

On the face of it, the drone appears as the solution to the central contradiction of the discourse of protective sovereignty: wage war without endangering the lives of one's own subjects, preserve without loss, always protect. But here comes the bad news: the promise to preserve national lives goes hand in hand with the increased social vulnerability and precariousness of many of those lives.

I am my weapon: that is a counterintuitive thesis. It conflicts with concepts of instrumentalism, contradicting the idea that the nature of the subject is independent of the means of his actions. On the contrary, it affirms that the two are essentially identical. If that is the case, I cannot dissociate my intentions or ends—the things that make me what I am—from the means that I employ in order to achieve them. Ethically, what I am is expressed and defined by the nature of the weapons that I mobilize. The choice of weapons is important because it radically affects what we are, and at stake in that choice is the risk of losing our soul or essence.

However, a few lines further on, Adorno introduces a decisive dialectical nuance into that original statement. Having emphasized that in this armed violence that involves no combat, the enemy is now relegated to the role of a “patient and [a] corpse” to which death is applied in the form of “technical and administrative measures,” he goes on to say: “Satanically, indeed, more initiative is in a sense demanded here than in old-style war: it seems to cost the subject his whole energy to achieve subjectlessness.”

Where is the subject that holds power? Against a background of neoliberalism and postmodernity, that question has become obsessive. Adorno’s words indicate a good way to resolve it: it is wherever it is working actively in order to make itself forgotten. The intense activity designed to efface it is what unfailingly singles it out: a great deal of subjective activity, involving huge efforts and enormous energy, designed to cover one’s tracks, efface evidence, and wipe out any trace of a subject involved in action.

Not only can there be no simple attribution of responsibility, but the description of that responsibility, diffracted amid this headless network of multiple agents, tends to become diluted. It changes from being intentional to being unintentional, from being a war crime to being a military-industrial accident. Rather, as in the case of the “junk bonds” skillfully elaborated by finance, it becomes very difficult to determine who is who or who has done what. This is a typical way of fabricating irresponsibility.

But to claim, as they do, that the law can be integrated into “weapon design” itself is a gross abuse of language. All that the roboethicists can do is integrate certain rules within the design of certain programs, which clearly can be uninstalled or reprogrammed at any time. If you are capable of doing this on your own computer, you can be sure that every army in the world is likewise capable of doing so. The discursive operation here consists in justifying the development of highly dangerous hardware by offering the option of virtuous software. Congratulations: in buying the car (or rather the robot tank), you have also won a magnificent key ring.

This is a typical “Trojan horse” type of procedure: in the name of the eventual prospect of ethical killer robots, you win the acceptance of the development of killer robots pure and simple, even though, as their supporters themselves know full well, public opinion still is massively opposed to them. By presenting the process of automatization as itself automatic and ineluctable, and by generously proposing in advance to moderate its excesses, Arkin and his associates are masking the fact that they themselves are the extremely active agents of that very process, effectively promoting it by providing the justifications that will be needed for it to prosper. The more widespread the legend of the ethical robot becomes, the faster the moral barriers to the deployment of killer robots give way.

Here, the military atrocity was, contrary to Arkin’s assumption, not the result of mistakes on the part of lower-rank soldiers confused by the “fog of war” or carried away by the passions of combat. There was nothing spectacular about the point of origin of this atrocity. It was simply a matter of fixing the threshold of a relevant variable. What would be the figure corresponding to the variable “minimum carnage”? We do not know. More than thirty citizens killed? Okay. But this little decision affecting a decision, effected by a single word or one tap on a keyboard, has multiple very concrete—all too concrete—results.

What is most surprising is that this could still come as a surprise. The greatest crime lies not in an open flouting of the law but in the recesses of its sovereign application. This ordinary military atrocity lies well within the law, and except under the most imperative conditions, it does not have to emerge. Most of the time it does not need to. Contemporary forms of atrocity are hugely legalistic. They are the rule rather than the exception, not so much because the law has been suspended, but rather because it has been spelled out in accordance with the interests involved to the point where it capitulates with scant resistance. This kind of atrocity is formalistic, cold, technically rational, and backed up by particular conditions—the same elements that are supposed to make the killer robots of the future eminently ethical.

The enigma of sovereignty and that of its constitution, as well as its possible dissolution, is resolved by the question of its material: what is the state made from? … After all, the master who oppresses you and “for whom you go bravely to war, for whose greatness you do not refuse to offer your own bodies unto death . . . where has he acquired enough eyes to spy upon you, if you do not provide them yourselves? How can he have so many arms to beat you with, if he does not borrow them from you?” Therein lay the fundamental material contradiction: if power is embodied solely by our bodies, we can always refuse to offer them.

Arendt explained that, owing to this fundamental corporeal dependence, the power of the state—even in the most authoritarian of regimes—must, despite everything, be power and not pure violence. There is no power without bodies. But, as she in a way acknowledged, the reciprocal statement is true: with no bodies mobilized, there is no longer any power.

A different age produces a different image. In 1924, a popularizing scientific magazine announced a new invention: a radio-commanded policing automaton. The robocop of the twenties was to be equipped with projective eyes, caterpillar tracks, and, to serve as fists, rotating blow-dealing truncheons inspired by the weapons of the Middle Ages. On its lower belly, a small metal penis allowed it to spray tear gas at unruly parades of human protesters. It had an exhaust outlet for an anus. This ridiculous robot that pissed tear gas and farted black smoke provides a perfect illustration of an ideal of a drone state.

The difference between these two vignettes heralds what is at stake politically in the dronization and robotization of the armed branches of the state. The dream is to construct a bodiless force, a political body without human organs, replacing the old regimented bodies of subjects by mechanical instruments that would, if possible, become its sole agents. … All the same, once this stage was reached, it is also possible that its increasingly evident destiny would be to be dumped in a junkyard like any other piece of scrap metal.

Once the state apparatus thus becomes, in effect, simply an apparatus, it would finally have acquired a body that corresponded to its very essence: the cold body of a cold monster. It would at last technically fulfill its fundamental tendency. As Engels wrote, “This power, arisen out of society, but placing itself above it and increasingly alienating itself from it, is the state.”

The separation of illusion and reality vanishes for the television warriors. Alienation and sterilization approach perfection. After kissing their wives goodbye and battling the rush hour traffic to work, the television warriors will settle down to a day of watching TV at the Ministry of Peace.

What is more likely is that the successive levels of war technology will coexist side by side, much as the missiles and the bombers do. Then there is the question of invincibility, the superhuman precision, the omniscience of sensors loaded on pilotless RPVs hooked to computer networks . . . etc. For those who are impressed by these claims we recommend paying attention to similar claims made in the past. There exists a vast difference between the results obtained under controlled conditions and the actual battle conditions. . . . The image of pinpoint destruction of individual resistors is a false one.

But neither is co-presence simply coexistence. For two entities to be said to be coexistent, it is enough that they exist at the same time, simultaneously. But co-presence assumes more than that, namely the possibility for one term to affect the other or to be affected by it (a relation of causality). In other words, co-presence is defined by an instantaneous but not necessarily actualized possibility of a real relationship. Yet another way of putting this is to say that co-presence is defined by the accessibility of one term to the other. To be co-present they must be within reach of each other. Two armies are in the presence of each other when they are within range, and that is the case even when not a shot has been fired. It is this notion of range that distinguishes co-presence, as a pragmatic concept, from mere co-existence.

The prey and its predator lurking in the shadows are co-present even if the former is not yet aware of this (or even if the prey cannot yet see the predator). For there to be co-presence, all that is necessary is that one of the terms involved should be included in at least one field in range of the other. There are paradoxical forms of unilateral co-presence in which entity A can act upon or be affected by entity B, while entity B is not in a reciprocal position. Here, the prefix “co-” indicates no reciprocity in the relationship but simply a common inclusion.

The body is active, perceiving, perceptible and vital and all those aspects are fixed in one and the same place, which is itself what we call the body. “The body” is the name of this immediate and seemingly indissoluble synthesis of those four dimensions or aspects of experience, which call out to one another and which, quite literally, march together. It is this immediate unity that tele-technologies radically undo, appending to the immediate synthesis another one, a technical synthesis that radically reconstructs the relations between those four aspects. They used to be linked, but now some have become independent, dissociated. The body has become dismantled and, in a partial reproduction of an organic body, it is possible to dissociate elements that the body, in its immediate unity, used to associate.

Tele-communication devices generally adopt structures of a reciprocal nature, but that is by no means a necessity. Again, it is a matter of a decision on the design of the tele-technological device. The structural choices for the devices used by drones are, on the contrary, of a nonreciprocal nature.

If our ears are no longer prepared for such praise, now that the drone paradoxically reembodies that representation, that is no doubt because the twentieth century has taught us to recognize its dark face. Where the Berlin philosopher saw a triumphant teleology pointing the way to a rationalization of state violence, the twentieth century, horrified, discovered something quite different. When John Ulric Nef took up this theme at the end of World War II, his tone was far less confident: “Progress has purged actual warfare of much of the emotional anger which accompanied the struggles of olden times. Except among the infantry, killing has become so impersonal that the killer resembles a boy with a toy pistol, or a man in a bathroom stepping on cockroaches.”

Beneath the “progress” of long-distance violence, we can now detect the barbarity of aseptic homicide. By now doubt is certainly cast upon the notion that mechanized murder and murder from behind a desk is less terrible than blood shed in passion.

This involves a way of forgetting the mediation, but that forgetting is merely pragmatic: there is no need to think about it in order to take action. So one is unaware of it, not in the sense of being incapable of recognizing its role or existence (epistemic ignorance), but only in the sense of being able to disregard it in order to take action (pragmatic forgetting). This pragmatic forgetting about the mediation by no means implies a failing in the perceiving subject or, on his part, any inability to perceive or recognize his role; it is, on the contrary, the product of long efforts to appropriate the instrument and make it part of oneself so as not to have to think about it. That momentary forgetfulness of the instrumental mediation is a state that one needs to succeed in attaining. It is not an epistemic failure but, on the contrary, a pragmatic success.

What he suggests here is not so much a right to equal combat, which would involve insisting on a struggle with equal weapons, but rather an equal right to fight. The fact that the law should not, by the norms that it imposes, make it "impossible from the start for one side to hope for victory" implies, not turning the war into a tournament—with pistol against pistol or sword against sword but, on the contrary, taking into consideration the disparity of the forces confronting each other and taking care not to increase it by imposing short-sighted laws that would confer to one side greater advantages while depriving the other side of even the possibility of fighting.

For Chaumont, it is a matter, not of bringing about an anachronistic return to a chivalric ideal but, on the contrary, of an attempt to integrate the parameters of contemporary asymmetrical conflicts within the law on armed conflicts, and to do so in a realistic way. His central preoccupation is the following: combatants deprived by law of any legitimate possibility of fighting no longer have any pressing reason to conform with the principles of a law that itself excepts them from any role except that of targets to be hit.

Some jurists detect a dangerous indistinction. The United States, writes Laurie Blank, relies "on both armed conflict and self-defense as legal justifications for targeted strikes outside of the zone of active combat. . . . It is the United States' insistence on using reference to both paradigms as justification for individual attacks and the broader program of targeted strikes that raises significant concerns for the use of international law and the protection of individuals by blurring the lines between the key parameters of the two paradigms."

For even though the law on armed conflicts defines the principle of proportionality as a good relation between estimated collateral damages and the scale of the military advantage expected, it provides no estimated scale nor, clearly, any unit of measurement. As Sharkey points out: “There is no known metric to objectively measure needless, superfluous, or disproportionate suffering. It requires human judgment. No clear objective means are given in any of the laws of war for how to calculate what is proportionate” (Sharkey, “Killing Made Easy”). To convert the principle of proportionality into a calculation would be like adding together apples and pears without any method of recognizing them. What is the commensurability between a particular number of lost civilian lives and an expected military advantage? What could the common unit of measurement be? However, as Eyal Weizman has shown, this kind of calculation, as necessary as it is impossible, serves solely to legitimate the deaths that result, and serves this function simply by being done.

As Sharkey points out: "There is no known metric to objectively measure needless, superfluous, or disproportionate suffering. It requires human judgment. No clear objective means are given in any of the laws of war for how to calculate what is proportionate" (Sharkey, "Killing Made Easy"). To convert the principle of proportionality into a calculation would be like adding together apples and pears without any method of recognizing them. What is the commensurability between a particular number of lost civilian lives and an expected military advantage? What could the common unit of measurement be? However, as Eyal Weizman has shown, this kind of calculation, as necessary as it is impossible, serves solely to legitimate the deaths that result, and serves this function simply by being done.

By eliminating human involvement in the decision to use lethal force in armed conflict, fully autonomous weapons would undermine other, non-legal protections for civilians. First, robots would not be restrained by human emotions and the capacity for compassion. . . . Emotionless robots could, therefore, serve as tools of repressive dictators seeking to crack down on their own people without fear their troops would turn on them. . . . Emotions do not always lead to irrational killing.