(Note: This is the write up of a paper I gave at a workshop in Berlin organised by Humboldt University’s Graduate School of Social Sciences and the Department of War Studies, King’s College London)
The drone war in the Pakistan borderlands cops a lot of flak. Some see it as illegal, others immoral, or it is a violation of state sovereignty or entirely counter-productive. It is a war within a war, and in blunt realist terms: “it works”, al-Qaeda’s number three position now comes with a short life-span as a condition of the job. Such blunt realism doesn’t lend itself kindly to ethical questions, and though there are many defenders of the drone strikes, I haven’t heard or read many people defending them on ethical grounds. The point is, however, that these strikes are legitimated as a temporary measure, something to be done until al-Qaeda is dead. This article is an argument against the “realist” legitimation, but argues that it is possible to construct a defence of sorts, a way of legitimising such methods in the long term. This defence, however, also argues that the way America conducts these strikes at present is mostly illegitimate. Legitimacy is a key aspect of contemporary war, at the very least from “our” perspective (Western states, their armed forces and their populations). We like to think that what we do is morally, and ethically, right, and we layer our warfighting methods with layers of legal accountability. For every accusation of war crimes committed by western forces, such as the notorious “kill team”, there are probably ten infantrymen bristling at “overly restrictive” rules of engagement. Furthermore, such gross breaches are considered exceptional: they lie “outside” the way we fight, and are therefore illegitimate.
My interest in all this came about because I’m currently wrestling with the idea of containing non-state actors. It is an argument that Patrick Porter advanced quite well in 2009, and others have also been advocating it in bits and pieces on and off over the last decade. In a nutshell, my thoughts are that, absent a method of eliminating al-Qaeda or criminal networks, states should be working to contain them. For al-Qaeda, this means something like disrupting the network to the point that they can’t organise another plot on the scale of 9/11. It means putting pressure on the network so that their ability to transfer the technical skills required to build bombs is restricted. It does, however, mean that one accepts that a 7/7 scale plot, another Madrid, or another Mumbai is likely to occur. In essence, states should seek to reduce the threat of terrorism to that of local, relatively untrained, amateurs. There are a host of reasons why I believe it is very difficult to reduce the threat below this level, without significant political accommodation, but most of these relate to my ideas of freedom and the relationship between the citizen and the state, arguing these points here would cause the entire article to sprawl! To this end, I’m currently pegging the operationalisation of such a strategy on the intellectual coat-hanger of “network management” – identifying the effects that operations have on the network structures of networked non-state actors, and whether they are suitable/sustainable for long term, open-ended application. A key component of this is legitimacy.
The focus of this piece is the effect that a switch from an end state of threat elimination to threat containment has on the legitimacy of the methods which are employed to target these networks. A great many actions were effectively legitimised, by both states and their populations, as well as the bureaucracies of states, in the aftermath of 9/11. The running thread of this legitimacy was the “state of exception” that existed after this event. In a nutshell: “We’re freedom loving, liberal democracies, we don’t want to do this, but we have to, but we promise to only do it for as long as we need to, and then get back to business as usual”. There were many arguments over the outputs of this exceptionality: Guantanamo, extra-ordinary rendition, waterboarding and torture, pre-emptive war and so on. Again, to include all the above would cause a sprawl, but they’re worth noting, particularly since some (torture) are things that I don’t consider could ever be justified. One type of action that I’m particularly interested in is drone strikes. In the state of exception, these work, while the people living on the other end might consider them illegitimate, particularly the civilians killed in the process. President Obama considers them okay, to the extent that he’s expanded their use considerably over the course of his presidency.
The problem with containment is that the actions undertaken because of such a strategy cannot be considered exceptional. Though it is possible to achieve some sort of distance from temporary activities, the timescale and permanence implied by containment means that activities conducted in its pursuit define a state and its society. They become normalised, and that normalisation becomes an identity. I think this is particularly important in international terms, because the US hasn’t quite got to grips with the implications of normalising lethal trans-border strikes by unmanned aerial vehicles. Remember that Israeli “targeted killing” was decried by the US before 9/11 as illegitimate. One wonders what the Middle East would look like if all the different states (and non state actors) therein got their mitts on rapidly proliferating drone technologies and considered it entirely normal and okay to start killing people across pesky international sovereign borders.
Fumbling through all the above, and thinking of all the implications of just war theory, I stumbled across “signature strikes”, which are, prima facie, almost indefensible on moral grounds. Essentially, bombs are dropped on the heads of people who aren’t known to be terrorists, or militants, but who act like them. Given the principles of non-combatant immunity, this couldn’t possibly be defended in the just war tradition, right? There are obvious problems with the application of just war theory to the drone strikes in places such as Pakistan, Yemen and further afield. The first being the question of whether it even counts as a war. Personally, I’m inclined to see drone strikes as an act of war, though not one directed at a particular state. Clausewitz might not have had al-Qaeda in mind, but his work still applies. The terms used to describe these strikes, such as “targeted killing” and so on are not warlike, but these are acts of war. They are organised acts of violence intended to compel an opponent to fulfil America’s political will. It may not be 19th century warfare, but it is still an act of war.
So what charges might we level at the perpetrators of these drone strikes? For the purposes of this piece, I am going to focus on signature strikes. This might seem unfair. America does target and kill known terrorists. But it seems to me that the murkier moral waters of signature strikes are more important. If these can be justified, then the relatively “safe” concept of dropping a bomb on a known terrorist’s head is probably defensible by extension. Conversely, a just war defence of strikes on known terrorists would not necessarily be applicable to signature strikes. Again, this is the examination of these strikes in the just war tradition, as interesting as the legal quandaries Obama entangled himself in by killing Anwar al-Alwaki (as well as the ones that he skipped out on by not capturing him) are, and as important as they are, for the sake of brevity, I’m restricting myself to just war theory.
Therefore, in the just war tradition:
The case for the prosecution
Signature strikes violate both traditions of just wars, and are indefensible except by recourse to arguments of pure power. Though international society lacks the means and will to prevent America from conducting these strikes, that we must accept them does not justify them. In order to construct this argument, we will demonstrate that they are a completely illegitimate tool, and therefore indefensible according to the principles of jus in bello (for the non-just war theorists, justice in war, principles of the tradition governing the conduct of war).
One of the key principles in just warfare is the division between combatants and non-combatants. In particular, non-combatants have immunity from conflict. We will admit that there are a number of instances in which belligerents can kill civilians, and it not constitute a crime. The primary one of these is double-effect. We understand that war can no longer be completely separated from the civilian world, and that citizens may be situated next to military targets. In this instance we accept that some civilians may die in a legitimate strike on a military target, as long as those deaths are proportional. We would not accept the destruction of a civilian hospital because it happens to be situated next to an army checkpoint, but we may be persuaded of the legitimacy of civilian deaths that occur from an attack on a military airfield. We will even grant the defence the idea that factory workers, while at work in a munitions factory, are legitimate targets, though not while they sleep in their beds at night. Therefore people aiding and abetting terrorists and militants could be legitimately killed while helping them, but at the same time, the execution of those running Hawala networks which allow terrorists to finance their operations are illegitimate. Indeed, we will even gift the defence the fact that terrorists and militants may be considered legitimate military targets, because this (debated) point will not help them in the slightest.
Given the very generous ethical berth we have given the defence, it is clear that signature strikes are still materially unjust. The principle of targeting is that a military unit must be targeted. Whereas we might accept that an identified terrorist can be targeted, and those nearby may be killed in the explosion, we cannot accept the same with signature strikes. By definition, these strikes target those that the Americans think display the same pattern of action as terrorists. We’re going to put our “acceptance” cap on and accept that they are telling us the truth, because the precise nature of these “signatures” is classified. Even so, they do not know who they are targeting. They may be able to examine the debris after the fact, or monitor conversations between terrorists in the aftermath, but at the exact point of pressing the button or pulling the trigger, they do not know who they are killing. And I am afraid our “acceptance” cap does not extend to the figures provided for the low number of civilian casualties, which have been derided. Post facto justification is morally indefensible, particularly where civilians may die. The just war tradition demands (yes, demands), that acts which may result in civilian death are proportional. In order to gain the protection of double-effect, the target must be identified, and must be thought to be worth the deaths of the proportional number of civilians. In the absence of this identification, the death of a single civilian would constitute a war crime.
And the defence:
The core of the prosecution’s case lies in certainty, that we can be certain that an individual is a combatant or a civilian. This is, however, false. In war, one can never be certain, but we allow for combatants to take actions in times of war that we do not accept outside it. Since the prosecution raises this issue specifically, let’s look at possibly the simplest case of killing, that of a soldier armed with a rifle. That soldier, when looking down the sights, sees a target that he (because, statistically speaking, it probably is a “he”) considers to be an enemy combatant. He pulls the trigger, and the target drops. Irrespective of whether that target is found to be a combatant or a civilian, he has not committed a war crime. True, if it turns out that he has killed a civilian, he has done something wrong, but this is war, we legitimise him pulling the trigger if he honestly believes that he is shooting at an enemy combatant. But the soldier can never truly know, particularly at the fringes of his ability to acquire information about a target, that the person he is shooting is, or is not, an enemy combatant. In contemporary warfare, we can only recognise combatants by their patterns of action. If they happen to be holding a rifle, or sitting on a mortar, this is a very good identification, however, we perceive this because we’re not used to armed civilian populations. In the countries in which we fight, possession of rifles might not indicate that the possessor is a combatant. What we rely on is the judgement of our soldiers. We train them excessively in the use of force, we restrict their capability to exercise such judgement with rules of engagement so as to eliminate the “grey areas” as best as possible in which they may end up killing civilians. But all this cannot entirely eliminate the possibility of such deaths occurring. Our moral legitimation of these people is that they are trained to use force, they are restricted in the use of force and they are responsible for the use of force. In all cases, the judgement over the use of force, the decision, is either the individual soldier’s, or his commanding officer. Therefore, in war, we cannot absolutely know whether our target is military or civilian, but we tend towards restricting ourselves if the target could be civilian. But this ambiguity is not present. We do this repeatedly. We know it works, we have killed two thirds the people we’re after in this way.
Identity, knowledge and warfare
The basic problem in this is identity, moreover, knowledge of identity. How does one know who is who? Returning to Clausewitz, he said you can’t. Well, more specifically, he said that uncertainty was an integral part of war. It was, however, easier when people were bunched up into formations and prodded towards a field of battle in something resembling a uniform to identify who was who. Even in this relatively simple case, we’re relying on a number of key identifying markers to discriminate soldiers from civilians: location, uniform, weapon and so on. Such markers no longer exist, or are no longer indicative of an active combatant. So when we consider it thus, the emphasis in justice and legitimation relies upon intent. It relies upon a question asked of the combatant at the point of the trigger pull: Do you honestly believe that you are going to kill an enemy combatant, and that any civilian casualties incurred as a result of your actions are going to be proportional?
If the question is framed in this manner, then I believe there is a case for the legitimacy of these strikes. The problem is that any evidence that could conclusively prove or disprove this is likely to remain in a black box for quite a while. We can’t know what they use as criteria for the “signatures” that induce attacks. We can hope that it is something more than “driving an SUV around North Waziristan”, but there is no way of knowing. But I’d argue that this isn’t as important. There are things that we do know about the campaign, firstly that it is a repeated process. We know that they find out afterwards, which means that they have the means to monitor the effects of what they’re doing. I like to think that means that they are also improving their targeting methods, but that’s just me.
The big problem I have, the one that I consider de-legitimises the whole endeavour, as currently employed, is the lack of responsibility and accountability for when things go wrong. Because they will go wrong. Despite the US Govt claiming drone strikes now almost civilian casualty-free, when we go to war, innocent people will die. That’s one of the things about war which should serve as a moral tripwire to the decision: are you willing to kill innocent people over this? In our societies, we are. We’re willing to send young people off around the world armed to the teeth with instructions not to kill civilians, but understanding that they probably will by accident at some point along the line. Drones strikes appear to be no different in this regard. But we have ways of dealing with that. Specifically, the institution that we charge with the application of legitimate force on our behalf has means and methods of dealing with civilian casualties. After action reporting, military police, courts martial, chain of command, and so on. The example of the soldier above does not, in our way of war, stop at the trigger pull. If that civilian is dead, there will probably be some form of reparations. There will be the soldier’s report going up the chain of command. If the soldier was felt to have acted irresponsibly, he might be penalised. Importantly, if he did so, his commanding officer might also be penalised, and so on. We consider the way we fight to be legitimate because it incorporates layer upon layer of responsibility and accountability. Of course, there are detractors, there are people who point out that the armed services have an innate wish to close ranks, to protect their own and give them the benefit of the doubt. Personally I’m minded to think that it would be very hard to form an armed force of any decent capability with the proviso that if any member ever commits an error, they will be thrown to the wolves, or the Guardian’s comment section, whichever is worse.
However at the end of the day, all the above discussion on responsibility stems from a simple action in warfare, a soldier pulling a trigger. Contemporary warfare doesn’t work like that. As a thought experiment, consider the role of information transfer and decision in forming responsibility among the following for a kill: a sniper team, an artillery piece and its spotter, a helicopter gunner being fed information from networked sources.
Now, the sniper team is possibly the second simplest case of responsibility. In a nutshell, a two-man team share any responsibility for the trigger pull. While the spotter might not actively kill the target, he is responsible for transferring information to his rifle-holding team member, without which the kill would not take place. Either can decide that the shot is no good. In this case even though the spotter does not pull the trigger, he is held equally responsible. This contrasts with the artillery unit and spotter. In this instance, the artillery unit does not have a decision. Unless the officer in command of the artillery unit knows that they will be firing on civilians, they are essentially unable to know what it is that they’re firing at. In this instance, the person calling in the artillery strike bears the responsibility for its effects. Of course, if the artillery shells fall short and so on, then such responsibility blurs, but for the purposes of this thought experiment, consider that they don’t. Moving to our last thought experiment, consider the position of a helicopter gunner who is told via multiple sources that he is firing upon enemy combatants. This is closer to how we fight, particularly in networked warfare. This gunner assimilates the identity of his target from a number of different sources. If a single one of these is wrong, if they perhaps place the target on the wrong street, or if he is hearing “They are five blocks north of us” from a unit who subsequently mis-communicates their location, or other SNAFU, then the gunner might end up targeting and killing the wrong person, maybe even a civilian. Here, we encounter a problem, in that it is hard to define responsibility. Does the responsibility lie with the gunner? Well, no, since he is ultimately unable to identify ground units, particularly un-uniformed insurgents and so on. The gunner’s world view is constructed from the information given to him, and if that information is wrong, he cannot be held to account. Unlike the example of the single soldier, the gunner did not discern this information himself, it was provided to him. So the responsibility lies with the people providing the information. Here is where we enter very dangerous territory, because the information provided to the gunner comes from multiple sources, it might not even be directly provided to the gunner for the purposes of targeting insurgents. Therefore, we cannot assign direct responsibility to people putting information into the system. This creates a curious twilight world of responsibility in which the people responsible for both the information transfer, and the decision to pull the trigger, cannot be held directly responsible for its outcome. If the whole networked warfare trend continues (it will), this will likely happen more often.
Thankfully, hard-wired into the military structure is a solution to this particular Gordian knot, namely the chain of command. If we can’t find someone to blame, we can button hole the officer in command. Not exactly fair, but then that’s the responsibility of command: you have a duty to ensure that your subordinates don’t kill civilians, to the best of your ability. So, in the instance of the helicopter gunner, the officer in command, the one that told them to fly out and provide air support, is responsible. And herein lies the problem with signature strikes, and lots of other drone strikes: they’re done by civilians, not the military. I know practically nothing about the CIA’s operational methods, for a reason, because they are a civilian intelligence organisation. I cannot, and an American citizen could not, argue that the CIA’s chain of command works in the same way as the military’s. Intelligence organisations are one big black box that we are prevented from peering into. There is no way for us to know, one way or the other, where responsibility for killings lies within their organisational structure, nor is there any way for us to know how they handle things when operations go bad. I’m not arguing that this is particularly wrong, simply that we don’t know, and we don’t know for a good reason: it would jeopardise the purpose of the organisation itself. The military contains similar black boxes: special forces and intelligence. However, soldiers working in those fields are still soldiers. They might be very, very good soldiers, but they are still soldiers. Those military black boxes are integrated within a military command structure that takes responsibility seriously, and we, as casual observers, can see that. So even though we can’t “know” what goes on within, we can be pretty sure that soldiers don’t enter these zones of military operation and immediately cease to be, and act like, soldiers. Furthermore, since we liberal democracies are inherently suspicious of military forces, we retain all sorts of controls over their actions, one of which usually entails hauling senior officers in front of politicians and cameras and holding them to account. We can’t say the same about intelligence organisations. We can say that there are controls, the UK has the Joint Intelligence Committee and so on, but those underlying visible principles are not present. Though intelligence agencies go to great lengths to inform us of their legitimacy and accountability, the nature of their work is such that they could never prove it, whereas the military are visible enough that they can, to a certain degree. In this, I’m not arguing that intelligence agencies are unscrupulous or whatnot, simply that there is a limit to our knowledge: we cannot know, for certain, one way or the other. Having said that, investigative journalists have a good sideline in pulling them up on complicity in torture and so on. For that reason, that lack of public assurance of responsibility for the effects of killing, I’d consider the CIA pretty much indefensible and illegitimate as an organisation to conduct a widespread drone strike program.
The second reason as to why the military should be in control of the drone strike program is that of last resort. Going back to the black box of the CIA, whatever is written or fictionalised about them, we don’t know what other tools they have in the cupboard. With the military, we do. We know that the US military has at its fingertips special forces which it uses for capture/kill raids, like the one that got Osama Bin Laden earlier this year. While US strategy undoubtedly integrates the two within Obama’s “shadow wars”, parallel programs don’t exactly instil confidence that drone strikes are a last resort, much less signature strikes. The application of violence needs to be proportional in order to be justified, and, as argued above, the targeting needs to reduce the chances of civilian casualties to the lowest possible. Night raids on houses might be a considerable source of tension, as are cross-border insertions of troops, but they are preferable to dropping high explosives on a compound. However good drone cameras are right now, they’re no match for teams of highly trained soldiers. Maybe future advances in drones and camera technology will take us to some “uncanny valley”, but that point isn’t now, and it is in the present day that people are doing this.
Let us make no bones about it: a drone strike is the easy option. With a drone strike you don’t risk a single soldier’s life. Compared to inserting a team of special forces operatives via helicopter, a drone strike is relatively risk-free. Furthermore, if a drone does crash, then the political fallout is relatively limited. There’s no chance of someone from the US military being hauled up in front of cameras for the world to see. But legitimising them, indeed, choosing them, for this is utterly reprehensible, morally. In short, I’d call it a coward’s way out. If the people targeted in this manner are so important that they require targeting in this (dubious) manner, then they are also so important that they are worth risking the lives of American service personnel. If they’re not, then they’re not, and America has no business in dropping bombs on their heads. Of course, there are situations in which a drone strike is feasible, and a raid is not. With the American military, we can at least be sure that within a single command structure, a commander has the two options available, and if he takes the drone strike option, then he’s responsible. With the CIA, we can’t be sure. Furthermore, the very existence of a parallel program of lethal targeting presents the executive with a political dilemma in that a covert, civilian, drone strike program is seductive in its simplicity and limited liability. It is neater to short-circuit the social limitations on the application of force.
To return to the normalisation of violence, I wonder if America has truly thought through the ramifications of what it is normalising. There is considerable debate as to whether civilians are legally entitled to participate in such programs. But in keeping with this post, I’ll keep to the ethical side. In America, federal agents are legitimated in using violence, the DEA sends armed agents abroad to work alongside state forces. Making them active combatants, well, now that opens up a barrel of worms, doesn’t it? Since they’re killing people, they can only be classed as combatants under the laws of armed conflict etc. That makes them a legitimate military target. Since we’re dealing with bureaucracies here, that makes every CIA station and agent, worldwide, a legitimate military target. Since all those people are collecting information for an organisation that retains a lethal capability, any state could quite happily classify them as a legitimate military target. It would, after all, be hard to argue against. We like to think of ourselves as legitimate, and the other side as illegitimate (I’m hard pressed to think of a war in which both sides considered the other side legitimate). We have a way of warfighting which defines a division between combatant and civilian. As far as identifying those divisions, contemporary war presents problems, as far as assigning a person or organisation to those divisions, it doesn’t. Either you are a combatant, or legitimate military target, or you’re not. In the popular imagination, James Bond and other fictional spies who kill are fun, but institutionalising the capability to lethally target people has ramifications, and I don’t think America, or anyone else is prepared to accept those. Does America really want to herald a world order in which states are okay with civilian agencies deploying military vehicles and lethal force against non-state actors or one another? Is it prepared to accept its civilian agencies as legitimate military targets? This is important. Even though death is inevitably viewed in an asymmetric way (them killing us is illegitimate, us killing them is legitimate and totally justified), at the core of our ethical and legal principles of war is a certain equality: if you join the military and die in an act of war, you’re no different than your counterpart across the border. Even though it is not a nice thing to think about, in a war military bases are legitimate targets, the Pentagon is a legitimate target, and so on. If you sign up to work as a secretary at the Pentagon and a war kicks off, sorry, but you signed up. Did the secretaries and janitors at Langley sign up for the same exposure? If the intelligence network of the CIA becomes a legitimate military target, what about the people working at CIA offices and fronts worldwide?
To conclude, Obama’s shadow wars might work. They might be killing al-Qaeda faster than they can organise. I doubt they’d truly eliminate al-Qaeda since repeated drone strikes offer no political solution to the conflict. In normalising these activities, they become an identity. I’m not actually arguing one way or the other on this point. Maybe people are okay with perpetual targeted/signature killings on their behalf, provided it is at a low level, at the fringes of the state system, and there’s no other option (some aren’t). I know some who would probably accept that, as distasteful as it is, over sending hundreds of thousands of troops abroad to refashion entire states. But as the era of counter-insurgency draws down to mixed results, and the sound of bombs in Baghdad, America will need to take a cold look at itself if this is the way it will be dealing with al-Qaeda in future.