It should surprise no one that evolving military practice raises novel legal questions. It may surprise many that the increasing use of “shift cold” techniques by advanced militaries, including our own, raises subtle and complex legal questions, implicating the most basic concepts and rules of the law of armed conflict. As expertly described by Mike Schmitt and Lt Col Matthew King,
A “shift cold” occurs when an operator … redirects a guided munition, such as a missile or guided bomb, away from its initially-intended point of impact to another location while the munition is in flight (that is, post-launch or release). This is generally done to avoid harm to civilians or to friendly forces in the target area who, at the time of weapon launch or release, were not expected to be there.
Most notably, if information obtained after the missile is launched makes it apparent that the original target is in fact civilian rather than military, or that the attack will cause disproportionate harm to nearby civilians, then the missile might be redirected to a point of impact where it will harm fewer civilians or no civilians. According to Schmitt and King, if “there is an identified point of impact with no or significantly less expected civilian harm, it is difficult to fashion an argument that there would not be an obligation to do so.”
Unfortunately, things are not so simple. In fact, shift cold techniques may be used in so many different circumstances, raising so many different legal and moral issues, that it is difficult to address any of them in isolation, let alone in the space allowed here. But I will try to offer some preliminary thoughts.
First, as Schmitt and King mention in passing, under international law, attacking forces “shall direct their operations only against military objectives.” Conversely, “[c]ivilian objects shall not be the object of attack.” Importantly, civilian objects are negatively defined as all objects which are not military objectives. Directing attacks against civilian objects is categorically prohibited, irrespective of the value of the civilian object in question or the reason for making it the object of attack. It seems to follow that it would be unlawful to redirect the missile away from its original target toward any civilian object, even an empty parking lot.
Second, international law also prohibits attacks which may be expected to cause incidental harm to civilians or civilian objects which would be excessive in relation to the concrete and direct military advantage anticipated. Since no military advantage is anticipated from redirecting the missile away from the original target, any expected incidental harm to civilians or civilian objects might seem excessive. Importantly, under international law, the proportionality rule does not call for a general balancing of the goods and evils likely to result from military action, but instead calls for a narrow balancing of civilian harm and military advantage. If no military advantage is anticipated from an attack, then under international law there is nothing to balance against the civilian harm expected.
The development of shift cold techniques therefore seems to trigger a direct conflict of legal obligations. As Schmitt and King discuss, attackers have a legal obligation to cancel or suspend attacks if it becomes apparent that these attacks are in fact directed at civilians or civilian objects or may be expected to cause disproportionate harm to civilians. On the other hand, attackers have legal obligations not to direct attacks at civilian objects or to carry out attacks they expect will harm civilians or civilian objects and produce no military advantage. In the cases under discussion, it seems that the only way to comply with the former obligation would be to violate the latter obligations.
Understandably, Schmitt and King seek to defend the legality of shift cold techniques that avoid or minimize harm to civilians. While I share their aim, I’m not persuaded by their argument, which seems both too quick and too strong. According to Schmitt and King,
Redirecting a munition into an area in which civilians are present is not a deliberate attack on a civilian target; it is a continuation of an ongoing attack rather than a separate attack. Same aircraft, same weapon, and same personnel in control of the weapon, and the act can best be described as guiding the weapon away from a point.
Similarly, they later write that “it is our view that the shift cold is not a new attack. Instead, the attack has unfolded in a way that was not foreseeable when planned and executed.” Since the shift cold is not a new attack but a continuation of an attack, the principles of distinction and proportionality do not apply as one might otherwise expect.
In my view, Schmitt and King offer the wrong explanation for the right result. The basic problem with their view is that, in general, the legality of redirecting a guided munition clearly turns on the legal status of its new target and the new balance of anticipated military advantage and expected civilian harm. Here are two admittedly extreme cases that illustrate the conceptual point:
Distinction: An operator launches a guided missile at a legitimate military target with no civilians nearby. While the missile is in the air, the operator learns that a beloved cultural leader [or property] is within the missile’s range, whose death [or destruction] will terrorize or demoralize the civilian population and undermine their support for opposing forces. The operator redirects the missile toward the cultural leader [or property], knowing that the latter is a civilian person [or object].
Proportionality: An operator launches a guided missile at a legitimate military target with no civilians nearby. While the missile is in the air, the operator learns that a more valuable military target, with many civilians nearby, is within the missile’s range. The operator redirects the missile toward the more valuable military target, knowing that the civilian harm she expects far exceeds the military advantage she anticipates.
In both cases, the redirection of the missile is clearly illegal. This shows that the legality of shift cold techniques cannot rest on the broad claim that redirecting a missile is per se not a new attack but instead a continuation of an attack. That would prove far too much. In the cases above, redirecting the missile is a new attack, with a new target, to which the principles of distinction and proportionality apply afresh. Alternatively, if redirecting the missile is a continuation of an ongoing attack, as Schmitt and King propose, then the principles of distinction and proportionality continue to apply as well, until the final target is selected and the final balance is struck between anticipated military advantage and expected civilian harm. Either way, the attack–new or ongoing–may not be directed at civilian objects or be expected to cause civilian harm and produce no military advantage.
Similarly, the legality of shift cold techniques cannot turn on the notion that they “can best be described as guiding the weapon away from a point.” Obviously, the illegal attacks in Distinction and Proportionality could not be defended on such grounds. On the contrary, shift cold techniques seem justifiable precisely insofar as they guide the weapon toward a point, specifically a point where the weapon will cause the least civilian harm.
A more promising argument might be that redirecting a missile away from its original target is not the continuation of an ongoing attack but instead the cancellation or suspension of an attack. Redirecting a missile also initiates a new attack in cases like Distinction and Proportionality, but not when the operator seeks only to avoid or minimize harm to civilians. This is because international law defines “attacks” as “acts of violence against the adversary, whether in offence or in defence.” If a missile is redirected at a military target, as in Proportionality, then there is a new attack because there is a new act of violence against the adversary. If a missile is redirected at a civilian person or object, as in Distinction, then there is a new attack if and only if the act of violence is ultimately directed against the adversary, its military operations, or its general war effort. However, if an operator redirects a missile solely to avoid or minimize harm to civilians, then this act, violent though it may be, is neither proximately not ultimately directed against the adversary. This explains why redirecting a missile toward a civilian object such as an empty parking lot need not constitute an attack on that civilian object.
By way of analogy, suppose that a fighter jet is critically damaged on its way to its intended target. The pilot turns away from the target, cancelling or suspending the initial attack. If the pilot redirects the plane toward another military target, or toward a civilian target for the purpose of indirectly harming the adversary, then there is a new attack. However, if the pilot seeks only to crash the plane in the least populated area that she can find, then there is no new attack.
Even this argument may prove too much. After all, the law of armed conflict undeniably prohibits gratuitous acts of violence against civilians, motivated by sheer cruelty or hatred, intended neither to harm the adversary nor benefit the attacking force. Perhaps such gratuitous acts of violence are not prohibited attacks as such, but instead violations of fundamental guarantees against murder, cruel treatment, collective punishment, and so forth. For this approach to succeed, we would at least need to show that the civilians in these cases are “in the power” of the attacking force, or that such fundamental guarantees otherwise apply to the use of standoff weapons.
Another concern arises from a different scenario discussed by Schmitt and King, in which an operator redirects a missile away from friendly forces and toward a civilian area. As Schmitt and King suggest, it seems that the proportionality rule must apply in such cases. Otherwise, it would be lawful to redirect a missile away from very few friendly forces on a routine mission toward very many civilians. Yet the proportionality rule only applies to attacks and, on the view we are considering, the initial attack is over, and no new attack has begun. How, then, could the proportionality rule apply to the redirection of the missile?
One answer might be that the proportionality rule applies to the preparation, execution, and termination of the initial attack. Put another way, the initial attack may not be cancelled or suspended in a way that may be expected to cause harm to civilians that would be excessive in relation to the military advantage anticipated from cancelling or suspending the attack (in this case, avoiding harm to friendly forces). Similarly, the initial attack must be cancelled or suspended in a way that avoids or minimizes harm to civilians.
Of course, one might object that the initial attack may not be cancelled or suspended in a way that makes a civilian object into the object of attack. But this would be a mistake. A terminated attack has no new object, though it may have new effects. The proportionality rule and the precautions rule constrain how attacks are terminated precisely because these rules concern the effects of attacks, which may occur after these attacks have come to an end.
Finally, recall that the proportionality rule prohibits attacks that may be expected to cause civilian harm but produce no military advantage. Does this entail, paradoxically, that it is unlawful to perform a shift cold for humanitarian reasons rather than military reasons if any harm to civilian persons or objects is expected? No. As we have seen, attacks by definition seek some kind of advantage over the adversary. The proportionality rule constrains the pursuit of such advantage. If the only advantage anticipated is not military but, say, political or economic, then any expected civilian harm would be disproportionate. Similarly, if the only advantage anticipated is not concrete and direct but instead speculative and remote, then any expected civilian harm would be disproportionate. In contrast, if no advantage of any kind is sought, then the proportionality rule has nothing to constrain and simply does not apply. Accordingly, if an attack is terminated in order to spare civilians and not to gain advantage, then any expected harm to civilians must be avoided or minimized but cannot be disproportionate.
Let me close by acknowledging that things may be as they first appeared. New technology may create conflicts between existing legal norms. Importantly, attempts to avoid such conflicts by narrowly interpreting the relevant norms may do more harm than good, creating gaps in the general legal protection enjoyed by civilians and civilian objects in order to avoid a specific legal problem. Rather than avoid such conflicts by defining them out of existence, we may have to resolve such conflicts. More precisely, we may have to specify the conditions under which one obligation outweighs or takes priority over competing obligations. In doing so, we may have to look to the purposes and values underlying those obligations–to their underlying moral justification–to determine which obligation should prevail in the specific circumstances. But I will leave that more difficult but often necessary task for another day.
A Note on the Ethics of Shifting Cold
As many readers will have noticed, shifting cold to avoid or minimize harm to civilians shares many of the morally relevant features of the original Trolley Problem all-too familiar to undergraduate philosophy students. If an instrument under your control will kill many innocent people unless you redirect it toward significantly fewer innocent people, then it seems morally permissible and probably morally obligatory to do so. This is because allowing an instrument which you set in motion and which remains under your control to harm others is morally comparable to harming them yourself, as opposed to allowing them to be harmed by some independent agent. Faced with the choice of harming more people or fewer people, you ought to harm fewer people.
The morality of shifting cold is more straightforward than the legality of shifting cold, for several reasons. First, there is no categorical moral prohibition on directing attacks at civilian objects (remember the empty parking lot), which do not enjoy the same moral inviolability as civilian persons. Second, the moral rule of proportionality weighs incidental harm to civilians against the moral good anticipated from military action, including avoiding harm to other innocent people, whether or not that moral good constitutes a military advantage. While these divergences between law and morality may prove justified, they certainly call for such justification, lest the law fail to provide combatants with sufficient reason to follow the rules.
In the typical case, shifting cold has at least two other morally relevant features worth mentioning. The first feature is that, assuming the original target is a military objective, if the missile strikes the original target then there will be some military advantage to offset the harm to nearby civilians. In contrast, if the secondary target is a civilian object, then there will be no military advantage to offset the harm to nearby civilians. Nevertheless, if redirecting the missile will harm many fewer civilians, or no civilians, then the difference in civilian harm will typically morally justify a shift cold.
The second feature is that, if the operator performs a shift cold, then the attacking force may look for another opportunity to attack the original target in the future. If that future attack also harms civilians, then the total harm inflicted by the redirected attack and the future attack might exceed the military advantage anticipated from successfully attacking the target. Again, this is not a general moral objection to shifting cold, only a reminder that the iterative nature of warfare raises moral issues absent from one-off emergency situations in ordinary life (or philosophical thought experiments).