Description
We talk about war in terms of neat categories (combatant, non-combatant, civilian, etc), yet targeting decisions in war are inherently imperfect. There is therefore a dissonance between just war theory and the practice of war. This is important because the standard of certainty used by combatants has important moral consequences. Too low a standard would lead to widespread unjustifiable killing, too high a standard would severely inhibit combatants from successfully waging war, which could also have moral consequences if their cause is just. This paper introduces the suspected combatant problem as a means of thinking through and evaluating moral arguments about identification practices and standards of certainty in war. Taking "suspected combatants" as the standard case forces us to reconsider our basic assumptions about ethical standards for the use of force in war. The suspected combatant problem highlights the moral consequences of uncertainty inherent in targeting decisions in war, as well as the considerable variance in uncertainty due to differences between armed groups, objects, and domains of warfare. This paper highlights the importance of this issue by examining critiques of artificial intelligence use in war that rest upon assumptions of perfect identification as a possibility, and baseline standard in war.
