## Thursday, May 21, 2020

### More on Double Effect and statistical reasoning

Consider this case:

1. You are fighting a just war. There are 1000 people facing you and you have very good, but fallible, reason to think of each that they are an unjust aggressor that you are permitted to kill. At the same time, on statistical grounds you know one of the thousand is innocent. You kill the thousand for standard military reasons.

This is justifiable, assuming nothing defeats the standard military reasons.

1. You are fighting a just war. There are 1000 people facing you and you have very good, but fallible, reason to think of each that they are an unjust aggressor that you are permitted to kill. At the same time, on statistical grounds you know one of the thousand is innocent. Moreover, the enemy is superstitious and thinks the number 1000 is especially significant, so that if you kill 1000, they will instantly surrender.

Now this case is tricky. At first, it seems like it’s an easier case than (1). After all, you have two separate reasons: the usual military reasons for killing unjust aggressors and the fact that if you kill them all, the enemy will instantly surrender. But it’s trickier than that. The problem is that if you simply kill the thousand for the standard military reasons, then you can intend to kill each one qua aggressor—for you have good reason to think of each that they are an aggressor, even though you know you are mistaken about each one. But if you act on the enemy’s superstition, you are intending to kill each one simpliciter, not just qua aggressor, for all 1000 need to be dead for the plan to be fulfilled. In particular, the one who is innocent needs to be dead, too, in order for your plan to be fulfilled. But when you acted on the standard military reasons, you didn’t need the innocent one to be dead—as that one wasn’t a problem.

So, in case (2) you cannot legitimately act on the enemy’s superstition and reason: “I will kill these 1000 in order that there be 1000 dead, which will trigger surrender.” For then the success of your action plan depends on the death of the innocents among the 1000, and not just on the death of the guilty. (I am not worried here about the moral problem of exploiting the enemy’s superstition. If you are, you can modify the case.)

That doesn’t mean you can’t take that superstition into account in a way. For instance, while military motives might be primary, you might have a defeater for these motives, such as that the mission is really dangerous. But the fact that the mission would end the war could defeat the reasons coming from the danger. This would be a Kamm-style triple effect case. (A more difficult question: could the fact that the enemy will surrender, thereby saving much bloodshed, defeat the reason against the action coming from the death of the innocent? I suspect not, but it’s a tough question.)

The above case pushes me to the idea that killing is one of those acts that can only be permissibly be done for certain kinds of reasons.

#### 1 comment:

Philip Rand said...

The example reveals the anomaly in B-theory of time...