The infinite disvalue strategy for modeling deontological constraints

An standard way to handle deontological constraints is to simply specify an infinite disvalue for breaking the constraints. Thus, you shouldn’t kill an innocent person to save ten innocents, because the disvalue of your murdering the one is infinitely greater than the finite value of the lives of the ten.

A standard response to this is to imagine cases where one deontological violation prevents multiple similar deontological violations. That cannot be handled by disvalue, since the multiple violations should have greater cumulative disvalue than the single violation. However, such cases may seem contrived.

But I just realized recently that they need not be contrived. In fact, the standard strategic bombing of civilian targets cases may be like that. These cases—which were probably exemplified by the bombings of Hiroshima, Nagasaki and Dresden—are usually described as cases where it is expected (or at least hoped) that the deaths of the innocents will persuade the enemy to surrender and stop a greater number of deaths.

However, in cases—like the World War II cases—where one is fighting an unjust aggressor, killings performed by the enemy (whether the victims are civilians or military personnel) are typically murders. Thus such cases may very well be cases where a smaller number of murders—committed by means of bombing—by one’s side prevents a greater number of murders committed by the other side. Thus, we have historical cases, or at least cases very close to historical cases, where a smaller number of immoral acts is thought to prevent a greater number of immoral acts of the same kind. And hence we have uncontrived cases where the disvalue strategy for modeling deontological constraints fails.