This paper investigates the question of legitimate targets in war and the traditional just in bello principle of discrimination, which is generally interpreted to mean that a bright line must be drawn between combatants and noncombatants, and that only the former may be attacked directly. Michael Walzer and John Rawls have proposed a "supreme emergency exemption" to this principle, which permits the targeting of innocent people in emergencies such as that of Britain in late 1940. Rejecting this, the paper offers as an alternative a principle of "graduated discrimination." This principle distinguishes three classes: innocents, combatants, and noncombatant belligerents (noncombatants are belligerent if they contribute directly to the enemy's war effort). It holds that the bright line must still be drawn, but between innocents and belligerents, and that, among the later, non-combatants may be attacked in severe conditions--even, in supreme emergencies, if their belligerent role is simply providing the regime with a popular mandate.
I agree with Toner that this "supreme emergency exemption" nonsense just won't do; an innocent person has done you no harm, so it makes no sense to say that it is just to harm, deliberately, an innocent. And, as Toner also notes, this "supreme emergency exemption" appears to stand in the way of the only licit end of warfare: return to a just peace. We cannot seriously allow such an exemption to exist.
I am not convinced by Toner's proposed alternative (which, it should be said, he is largely just airing and examining rather than actually using). I do like the recognition that the combatant/noncombatant distinction is not quite right; civilian commanders-in-chief, for instance, are not combatants, but qua commanders-in-chief are responsible for what the combatants do. However, Toner wishes to go farther and extend the category of noncombatant belligerent to anyone who is supporting the soldiers qua soldiers. So, while farmers wouldn't be fair game (they are supporting the soldiers qua human beings), munitions factory workers would be. More seriously, in a society in which the regime is supported by a 'mandate' from the people, the people would be complicitous, and therefore viable targets (under certain circumstances).
Now, I don't believe in any such superstitious nonsense as a 'mandate of the people'. Unless the decision to go to war were referred to the people in a referendum, and they specifically voted, "Yes, we will go to war", there is no mandate. There may well be a vague sense of support in the population, or even an enthusiastic sense of support; but there is no 'mandate'. Mandates are things made up by politicians as a ready-made justification for whatever policies they prefer; when, as far as anyone really knows, they may have actually been elected for entirely different policies, or for the way they look on TV. Even if there were such a mandate, the opposing army would have no serious way of distinguishing it from a mere appearance of the mandate, and no good way of discerning when the mandate changes, and so it seems to me that this whole line of thought is simply useless. Toner has a reply to this, but it is very weak; he responds that we don't demand certainty of our leaders, but just action on the best information they have. That's certainly true; but it seems to me to miss the point. There is no 'best information' here, and I don't think there is even in principle.
So I take a different stand on this sort of thing. When Britain terror-bombed German cities in World War II, Britain was morally in the wrong, however great the complicity of the German civilians in the Nazi regime. The Dresden bombing was a terrible thing that cannot be justified, either in terms of a "supreme emergency exemption" or in terms of a form of graduated discrimination. If the civilians are not actually engaged in military work (e.g., transporting weapons to military bases), it is utterly illegitimate to go around deliberately killing them. Strictly speaking one should even be doing what one can to limit the deaths of combatants, to the extent it is possible to do so while intending the just cause that is governing one's involvement in the war in the first place.