Just War Theory
JUST WAR THEORY
In traditional just war theory there are two basic elements: an account of just cause and an account of just means. Just cause is usually specified as follows:
(1) There must be substantial aggression.
(2) Nonbelligerent correctives must be either hopeless or too costly.
(3) Belligerent correctives must be neither hopeless nor too costly.
Needless to say, the notion of substantial aggression is a bit fuzzy, but it is generally understood to be the type of aggression that violates people's most fundamental rights. To suggest some specific examples of what is and is not substantial aggression, usually the taking of hostages is regarded as substantial aggression while the nationalization of particular firms owned by foreigners is not so regarded. But even when substantial aggression occurs, frequently nonbelligerent correctives are neither hopeless nor too costly. And even when nonbelligerent correctives are either hopeless or too costly, in order for there to be a just cause, belligerent correctives must be neither hopeless nor too costly.
Traditional just war theory assumes, however, that there are just causes and goes on to specify just means as imposing two requirements:
(1) Harm to innocents should not be directly intended as an end or a means.
(2) The harm resulting from the belligerent means should not be disproportionate to the particular defensive objective to be attained.
While the just means conditions apply to each defensive action, the just cause conditions must be met by the conflict as a whole.
It is important to note that these requirements of just cause and just means are not necessarily about war at all. Essentially, they constitute a theory of just defense that can apply to war but can also apply to a wide range of defensive actions short of war.
The Intended/Foreseen Distinction
Just war theory presupposes that we can, in practice, distinguish between what is foreseen and what is intended, and some have challenged whether this can be done. So first one needs to address this challenge.
The practical test that is frequently appealed to in order to distinguish between foreseen and intended elements of an action is the Counterfactual Test, according to which two questions are relevant:
(1) Would you have performed the action if only the good consequences would have resulted and not the evil consequences?
(2) Would you have performed the action if only the evil consequences resulted and not the good consequences?
If an agent answers "yes" to the first question and "no" to the second, some would conclude that (1) the action is an intended means to the good consequences, (2) the good consequences are an intended end, and (3) the evil consequences are merely foreseen.
But how well does this Counterfactual Test work? Douglas P. Lackey argues that the test gives the wrong result in any case where the "act that produces an evil effect produces a larger good effect" (1987, p. 260). He cites the bombing of Hiroshima, Japan, as an example. That bombing is generally thought to have had two effects: the killing of Japanese civilians and the shortening of World War II. Now suppose we were to asked:
(1) Would Harry S. Truman have dropped the bomb if only the shortening of the war would have resulted but not the killing of the Japanese civilians?
(2) Would Truman have dropped the bomb if only the Japanese civilians would have been killed and the war not shortened?
And suppose that the answer to the first question is that Truman would have dropped the bomb if only the shortening of the war would have resulted but not the killing of Japanese civilians, and that the answer to the second question is that Truman would not have dropped the bomb if only the Japanese civilians would have been killed and the war not shortened. Lackey concludes from this that the killing of civilians at Hiroshima, self-evidently a means for shortening the war, is by the Counterfactual Test classified not as a means but as a mere foreseen consequence. On these grounds Lackey rejects the Counterfactual Test as an effective device for distinguishing between the foreseen and the intended consequences of an action.
Unfortunately, this is to reject the Counterfactual Test only because one expects too much from it. It is to expect the test to determine all the following:
(1) Whether the action is an intended means to the good consequences.
(2) Whether the good consequences are an intended end of the action.
(3) Whether the evil consequences are simply foreseen consequences.
In fact, this test is capable of meeting only the first two of these expectations. And the test clearly succeeds in doing this for Lackey's own example, where the test shows the bombing of Hiroshima to be an intended means to shortening the war, and shortening the war an intended consequence of the action.
To determine whether the evil consequences are simply foreseen, however, an additional test is needed, which can be called the Nonexplanation Test. According to this test the relevant question is: Does the bringing about of the evil consequences help explain why the agent undertook the action as a means to the good consequences? If the answer is "no," that is, if the bringing about of the evil consequences does not help explain why the agent undertook the action as a means to the good consequences, the evil consequences are merely foreseen. But if the answer is "yes," the evil consequences are an intended means to the good consequences.
Of course, there is no guaranteed procedure for arriving at an answer to the Nonexplanation Test. Nevertheless, when we are in doubt concerning whether the evil consequences of an act are simply foreseen, seeking an answer to the Nonexplanation Test will tend to be the best way of reasonably resolving that doubt. For example, when applied to Lackey's example, the Nonexplanation Test comes up with a "yes," since the evil consequences in this example do help explain why the bombing was undertaken to shorten the war. For, according to the usual account, Truman ordered the bombing to bring about the civilian deaths, which by their impact on Japanese morale were expected to shorten the war. So, by the Nonexplanation Test, the civilian deaths were an intended means to the good consequences of shortening the war.
Just war theory has been challenged in various ways. Three of the most important are a conventionalist challenge to just means, a collectivist challenge to just means, and a feminist objection to just cause and just means.
A Conventionalist Challenge to Just Means
The criteria of just means have been incorporated to some degree into the military codes of different nations and adopted as international law. George Mavrodes (1984) contends that the criteria of just means ought to be met simply because they have been incorporated into military codes or adopted as international law. Mavrodes arrives at this conclusion largely because he finds the standard attempts to specify the convention-independent basis for condition (2) of just means to be so totally unsuccessful. All such attempts, Mavrodes claims, are based on an identification of innocents with noncombatants. But by any plausible standard of guilt and innocence with moral content, Mavrodes contends, noncombatants can be guilty and combatants innocent. For example, noncombatants who are doing everything in their power to support an unjust war financially would be morally guilty, and combatants who were forced into military service and intended never to fire their weapons at anyone would be morally innocent. Consequently, the guilt-innocence distinction will not support the combatant-noncombatant distinction.
Hoping to support the combatant-noncombatant distinction, Mavrodes suggests that the distinction might be grounded on a convention to observe it. This would mean that our moral obligation to abide by condition (2) of just means would be a convention-dependent obligation. Nevertheless, Mavrodes does not deny that we have some convention-independent obligations. Our obligation to refrain from wantonly murdering our neighbors is given as an example of a convention-independent obligation, as is our obligation to reduce the pain and death involved in combat. But to refrain from harming noncombatants when harming them would be the most effective way of pursuing a just cause is not included among our convention-independent obligations.
Still, Mavrodes does not claim that our obligation to refrain from harming noncombatants is purely convention-dependent. He allows that, in circumstances in which the convention of refraining from harming noncombatants does not exist, we might still have an obligation to unilaterally refrain from harming noncombatants, provided that our action will help give rise to a convention prohibiting such harm with its associated good consequences. According to Mavrodes, our primary obligation is to maximize good consequences, and this obligation requires that we refrain from harming noncombatants when that will help bring about a convention prohibiting such harm. By contrast, someone who held that our obligation to refrain from harming noncombatants was purely convention-dependent would never recognize an obligation to unilaterally refrain from harming noncombatants. On a purely convention-dependent account, obligations can only be derived from existing conventions; the expected consequences from establishing a particular convention could never ground a purely convention-dependent obligation. But while Mavrodes does not claim that our obligation to refrain from harming noncombatants is purely convention-dependent, he does claim that this obligation generally arises only when there exists a convention prohibiting such harm. According to Mavrodes, the reason for this is that, generally, only when there exists a convention prohibiting harm to noncombatants will our refraining from harming them, while pursuing a just cause, actually maximize good consequences.
But is there no other way to support our obligation to refrain from harming noncombatants? Mavrodes would deny that there is. Consider, however, Mavrodes's own example of the convention-independent obligation not to wantonly kill our neighbors. There are at least two ways to understand how this obligation is supported. Some would claim that we ought not to wantonly kill our neighbors because this would not maximize good consequences. This appears to be Mavrodes's view. Others would claim that we ought not to wantonly kill our neighbors, even if doing so would maximize good consequences, simply because it is not reasonable to believe that our neighbors are engaged in an attempt on our lives. Both of these ways of understanding how the obligation is supported account for the convention-independent character of the obligation, but the second approach can also be used to show how our obligation to refrain from harming noncombatants is convention-independent. According to this approach, since it is not reasonable to believe that noncombatants are engaged in an attempt on our lives, we have an obligation to refrain from harming them. So interpreted, our obligation to refrain from harming noncombatants is itself convention-independent, although it will give rise to conventions.
Of course, some may argue that whenever it is not reasonable to believe that persons are engaged in an attempt on our lives, an obligation to refrain from harming such persons will also be supported by the maximization of good consequences. Still, even if this were true, which seems doubtful, all it would show is that there exists a utilitarian or forward-looking justification for a convention-independent obligation to refrain from harming noncombatants; it would not show that such an obligation is a convention-dependent obligation, as Mavrodes claims.
A Collectivist Challenge to Just Means
According to a collectivist challenge to just means, more people should be included under the category of combatants than the standard interpretation of just means allows. Just means, as noted earlier, imposes two requirements:
(1) Harm to innocents should not be directly intended as an end or a means.
(2) The harm resulting from the belligerent means should not be disproportionate to the particular defensive objective to be attained.
According to advocates of this challenge to just means, the problem is that the standard interpretation of (1) does not assume that the members of a society are collectively responsible for the actions of their leaders unless they have taken radical steps to oppose or disassociate themselves from those actions, for example, by engaging in civil disobedience or by emigrating. Of course, those who are unable to take such steps, particularly children, would not be responsible in any case; but for the rest, advocates of this collectivist challenge contend that failure to take the necessary radical steps, when one's leaders are acting aggressively, has the consequence that one is no longer entitled to full protection as a noncombatant. Some of those who press this objection against the just means criteria of just war theory, like Gregory Kavka (1985), contend that the members of a society can be directly threatened with nuclear attack to secure deterrence, and so reject noncombatant immunity, but then deny that carrying out such an attack could ever be morally justified. Others, like James W. Child (1986), reject both noncombatant immunity and proportionality by contending that the members of a society who fail to take the necessary radical steps can be both indirectly threatened and indirectly attacked with what would otherwise be a disproportionate attack.
In response to this collectivist challenge the first thing to note is that people are more responsible for disassociating themselves from the unjust acts of their leaders than they are for opposing those same acts. For there is no general obligation to oppose all unjust acts, even all unjust acts of one's leaders, because this would impose an unreasonable demand on individuals, and we are not morally required to be saints. Nevertheless, there is a general obligation to disassociate oneself from unjust acts and to minimize one's contribution to them, because this is not an unreasonable demand to impose on each of us. Of course, how far one is required to disassociate oneself from the unjust acts of one's leaders depends on how much one is contributing to those actions. If one's contribution is insignificant, as presumably a farmer's or a teacher's would be, only a minimal effort to disassociate oneself would be required, unless one's action could somehow be reasonably expected, in cooperation with the actions of others, to put a stop to the unjust actions of one's leaders. However, if one's contribution is significant, as presumably a soldier's or a munitions worker's would be, a maximal effort at disassociating oneself would be required immediately, unless by delaying one could reasonably expect to put a stop to the unjust actions of one's leaders.
In support of this collectivist challenge to just war theory Child (1986) offers the example of a member of a board of directors of a company that is engaging in the immoral and illegal activity of pouring large quantities of arsenic into the public water supply as a matter of ongoing operations. When the policy is before the board, she votes no but does nothing else. Later, when sued in tort (or charged in crime) with these transgressions of duty, she pleads that she voted no. Child argues that mere formal dissent in this case does almost nothing to relieve her liability, legal or moral.
But while one might agree with Child that in this case the member of the board of directors had at least the responsibility to disassociate herself from the actions of the board by resigning, this does not show that farmers and teachers are similarly responsible for disassociating themselves from the unjust action of their leaders either by engaging in civil disobedience or by emigrating. This is because neither their contributions to the unjust actions of their leaders nor the effect of their disassociation on those unjust actions would typically be significant enough to require such a response.
This is not to deny that some other response (e.g., political protest or remunerations at the end of the war) would not be morally required. However, to meet this collectivist challenge, it suffices to show that not just any contribution to the unjust actions of one's leaders renders the contributor subject to attack or threat of attack; one's contribution must be significant enough to morally justify such a response.
A Feminist Challenge to Just Cause and Just Means
A formidable challenge to both the just cause and just means criteria of just war theory comes from feminism. According to the feminist challenge to just war theory, sexism and militarism are inextricably linked in society. They are linked, according to Betty Reardon (1985), because sexism is essentially a prejudice against all manifestations of the feminine, and militarism is a policy of excessive military preparedness and eagerness to go to war that is rooted in a view of human nature as limited to masculine characteristics. Seen from a militarist perspective, other nations are competitive, aggressive, and averse to cooperation, the same traits that tend to be fostered primarily in men in a sexist society. By contrast, the traits of openness, cooperativeness, and nurturance that promote peaceful solutions to conflicts tend to be fostered primarily in women, who are then effectively excluded from positions of power and decision making in a sexist society. Consequently, if people are to rid society of militarism, Reardon argues, they need to rid society of sexism as well.
But even granting that sexism and militarism are inextricably linked in society in just the way Reardon maintains, how does this effect the validity of just war theory? As just war theory expresses the values of proportionality and respect for the rights of innocents, how can it be linked to militarism and sexism? The answer is that the linkage is practical rather than theoretical. If the leaders in militarist-sexist society have been socialized to be competitive, aggressive, and averse to cooperation, then they will tend to misapply just war theory when making military decisions. This represents an important practical challenge to just war theory. And the only way of meeting this challenge seems to be to rid society of its sexist and militarist attitudes and practices so as to increase the chances that just war theory will be correctly applied in the future.
Of course, still other challenges could be raised to just war theory but, in large measure, just war theory has stood the test of time. Moreover, if the theory can be reconciled with the most morally defensible form of pacifism, such that the only wars and large-scale conflicts that definitely satisfy the requirements of just war theory are the only wars and large-scale conflicts to which antiwar pacifists cannot reasonably object, then it is really hard to see how the theory could be displaced.
See also Civil Disobedience; Cosmopolitanism; Multiculturalism; Postcolonialism; Republicanism.
Bibliography
Cady, Duane L. From Warism to Pacifism: A Moral Continuum. Philadelphia: Temple University Press, 1989.
Child, James W. Nuclear War: The Moral Dimension. New Brunswick, NJ: Transaction, 1986.
Holmes, Robert L. On War and Morality. Princeton, NJ: Princeton University Press, 1989.
Kavka, Gregory. "Nuclear Deterrence: Some Moral Perplexities." In The Ethics of War and Nuclear Deterrence, edited by James P. Sterba, 127–138. Belmont, CA: Wadsworth, 1985.
Lackey, Douglas P. "The Moral Irrelevance of the Counterforce/Countervalue Distinction." The Monist 70 (3) (July 1987): 255–276.
Mavrodes, George. "Conventions and the Morality of War." In Morality in Practice, edited by James P. Sterba, 302–310. Belmont, CA: Wadsworth, 1984.
Reardon, Betty. Sexism and the War System. New York: Teachers College Press, 1985.
Sterba, James P. "Reconciling Pacifists and Just War Theorists." Social Theory and Practice 18 (1) (Spring 1992): 21–38.
Walzer, Michael. Just and Unjust Wars: A Moral Argument with Historical Illustrations. 2nd ed. New York: Basic, 1992.
James Sterba (2005)