Jus in bello prohibits deliberate killing of non-combatants. But who is combatant? Plainly, a uniform a combatant does not make. If a dictator decreed that all toddlers are to wear military uniforms, that would not make them into combatants. Nor would they be combatants if he issued them with guns which they never fired. (The question of what should be done if they did shoot is a more involved one.)
I find particularly challenging the case of people pressed into military service who have the intention to refrain from violent acts. Under compulsion, they wear a uniform and carry a gun, but are no more combative than an unarmed toddler. In World War II, only 15-20% of American soldiers themselves along the line of fire would fire at the enemy in any given battle, even if the engagement lasted two or three days (I got this from Grossman's book on killing; Grossman doesn't say if the same 80-85% of soldiers were refraining from shooting in different engagements).
Suppose, then, that you are fighting a just war, and know (e.g., from intelligence reports) that 80% of enemy servicemen have a personal commitment either never to shoot or to shoot only in the air, and are wearing the uniform and carrying weapons under compulsion. You have an armed enemy serviceman in your gunsights. You do not have time to observe him long enough to figure out if he is one a uniformed pacifist or a soldier. Is it licit to shoot him?
If not, then justly waging wars against enemies whose soldiers are likely to be like that seems nigh impossible. On the other hand, how can it be licit to shoot someone who is more likely than not as innocent of bellicose activity as any conscientious objector? If you are a police sniper who sees in the distance five people, one of whom you know is a terrorist about to trigger a bomb via remote control and the other four are innocent bystanders, surely you are not permitted to pick off all five when you can't tell which one is the terrorist.
I see five solutions to this problem:
- Drop the moral prohibition against killing innocent people who aren't any danger to anybody. This seems clearly wrong.
- Prohibit lethal engagement in such situations. This makes much of what most people would consider the just conduct of war impossible. But perhaps double effect would still allow use of weapons not targeted at particular individuals, with the intention of killing the guilty.
- Argue that by wearing uniform and carrying a gun on the side of injustice in a war, one has made oneself a part of an unjust war effort, and that in and of itself makes one a combatant. After all, even if one is not shooting oneself, one is boosting the war effort by providing a certain amount of cover for those who are shooting, etc. And anybody who is boosting the war effort on the side of injustice may be fairly shot. This seems mistaken. If our own POWs were forced to wear enemy uniform, and at gunpoint mixed with enemy troops, we would not consider them combatants on the enemy side--i.e., traitors--as long as they refrained from shooting at us. And the nationality of compelled uniformed people should make little moral difference here.
- Follow Germain Grisez in saying that even in wartime, it is always wrong to intentionally kill anybody, guilty or innocent. But one can use guns and bombs and the like to intentionally make enemy soldiers incapable of harming us, and we can do this by the Principle of Double Effect (PDE) without intending that the enemy soldiers die. Their death is not a means to our goal of self-protection. All we need for self-protection is that they be out of commission for the course of the war. If this is correct, then it does not matter if 4/5 of the enemy is uniformed and pacifists, since their deaths are "collateral damage", just like the deaths of civilians standing near the enemy HQ when the HQ is bombed. But is this plausible? While some weapons can be thought of as disabling the enemy, with the enemy's death an unintended side-effect (e.g., if I hit an attacker over the head with a club, then it is plausible to suppose that I do not intend to kill him, but merely put him out of commission; if he dies, that's an unfortunate side-effect). But it feels like a stretch to use this kind of justification for all weapons. Suppose a sniper shoots an enemy soldier in the head. Let us say, with Grisez, that this is to disable the enemy soldier. But how is the sniper disabling the soldier? By destroying the brain. And how does destroying the brain disable the soldier? By killing him, it seems. And so the killing is intended, as a means, contrary to Grisez. Now maybe one can argue that one is merely trying to destroy the parts of the brain involved in fighting, and the fact that one destroys the whole brain is a mere unintended side-effect. This feels sophistical. And, anyway, if it is wrong to kill the innocent, it seems to be also wrong to intentionally destroy healthy portions of their brains.
- Use another Double Effect line of justification. One aims at the enemy serviceman's heart and fires, say. But one isn't intending that the enemy serviceman be dead, or even that his heart be unable to oxygenate his body sufficiently for fighting (as per the previous suggestion). Rather, one is intending a conditional effect: One is intending that this serviceman be dead, or at least that this heart be unable to oxygenate his body sufficiently for fighting, if he is genuinely a combatant. One doesn't know that he is a uniformed pacifist (if one knew that, shooting him would be plainly wrong), and one's intended goal can be conditional. This solution strikes me as the least unsatisfactory of the five, but is not that satisfactory. It would allow the police sniper to shoot the five people one of whom is a terrorist. On the other hand, this solution coheres with the following intuition: It makes relatively little moral difference whether you throw a grenade at five people, killing them at once, or shoot each one individually. (Though the latter may be much more traumatic.) But it could be licit to throw a grenade at five people, four of whom were innocent and one of whom was a terrorist about to kill many people.
Here is an interesting and, I think, important conclusion. When deciding whether the proportionality condition in jus ad bellum holds--whether the war would eliminate more evils than it would cause--one needs to count among the evils the many non-bellicose enemy soldiers who would die as a result of the war. This can have real consequences. Suppose that one estimates that if an invading force is unopposed, they will murder 500,000 of one's people, some soldiers and some non-soldiers, but will cause no other evils. Suppose, further, that by opposing them, one will be able to reduce the death-toll on one's own side to 100,000, but one will need to kill a million enemy soldiers to do so. In killing a million enemy soldiers, one might be killing 800,000 completely innocent and non-bellicose people. Here, the proportionality condition would not seem to be met--one kills 800,000 innocent people to save 400,000. One should instead surrender. Unless, of course, one can argue that the enemy will not stop at killing the 500,000, which in practice is likely. But in any case, one must take into account the innocent death toll among enemy soldiers into account when figuring out if a war is licit.