Hard Thoughts About Warfare

Total war, nuclear deterrence, terrorism, reprisals-drawing some moral lines.


Just and Unjust Wars, by Michael Walzer, New York: Basic Books. 1977. 361 pp. $15.

Warfare raises, in their most stark form, moral issues about aggression, initiation of force, rights of innocent third parties, responsibility for following orders, etc., and so should be of special interest to libertarians. The previous literature focusing on these moral issues of warfare is slight and insubstantial, however, so we should welcome Michael Walzer's sensitive, thoughtful, and illuminating study. Walzer's book is subtitled A Moral Argument with Historical Illustrations, and his method of taking up each of the issues as it applies in specific historical situations makes for especially interesting reading.

I shall selectively discuss the contents of the book, emphasizing those points I find especially illuminating and those with which a libertarian is especially likely to disagree. Walzer, a professor of government at Harvard University and an editor of the democratic socialist magazine Dissent, is a friend of mine, and so this will not be the first time he hears of the disagreement. I mention the friendship because as a student I used to be bemused when reading book reviews by one friend of another, where no mention was made of the personal tie. So, caveat lector (and friend, too!).


Walzer distinguishes the justice of war (is it defensive, aggressive?) from the legitimacy of means in pursuing it (is attack directed against innocent civilians? etc.). He holds that an ordinary soldier can be held responsible and punished only for using illegitimate means of warfare, even if he is fighting in the army of the aggressive nation that started the war. On this view, when that soldier kills a civilian on the other side by deliberately taking aim and firing, he is a murderer, but when he intentionally shoots an opposing soldier who is merely defending his nation against foreign attack he is not a murderer. Walzer holds that the leaders of the nation who decide to launch an aggressive war can be held responsible and punished for this, but ordinary soldiers cannot, provided they do not use unjust means of waging the war.

Libertarians will not agree that one is absolved of responsibility merely by being told (by the rulers that be) to use legitimate means, whatever the legitimacy of the end to which these are directed. So we face the interesting issue of precisely how to treat the captured soldiers of an aggressive army. (Are all to be tried for murder or attempted murder?)

Walzer correctly points out that the aggressor nation need not be the one who actually attacks first, for its threat of imminent aggressive warfare (not merely of hostile acts short of war) may be sufficient to justify a preemptive attack by the defending nation, which therefore turns out to be the first to fire.

I found the third part of the book the most interesting. In approximately 100 pages Walzer discusses with subtlety and good sense noncombatant immunity, sieges and blockades against places also containing civilians, guerrilla war, terrorism, and reprisals. Noncombatants may not be attacked, but suppose it is known that some will or may be harmed as a result of an attack? The traditional doctrine of double effect says that one may not directly aim at, intending to inflict death on, civilians; but if one knows that some civilians will be injured as a side effect of one's action (hence the term "double effect"—to distinguish two kinds of effects, directly intended and side), this is not strictly forbidden but counts as a moral cost of the action, which is to be weighed in a calculation against its gains.

In a paper about 10 years ago, I argued that the principle for such moral cost-benefit calculations is the following. If A is the act being contemplated, it is not enough that the moral benefit of A outweigh the moral cost. For A will be impermissible if there is available an alternative act B of lesser moral cost such that the extra benefit of A over B is outweighed by the extra cost of A over B.

Walzer nicely suggests adding to the traditional principle of double effect the additional condition that one also must intend to reduce the evil as far as possible, including by running some additional risk to oneself in order to reduce the (not directly intended) civilian harm. Walzer finds it hard to say how far you must go in accepting additional risks. I suggest we view this alternative risky act as the B in the principle I just stated; Walzer's emended principle of double effect, then, is simply an instance of the general principle, which principle then specifies how much additional risk one must run.

I don't know if Walzer would approve of this, since he tends not to sharpen a point or principle beyond what good sense can specify without formal apparatus. Often, however, it is useful to speculate on what the ideal precise rule would be, for that helps us to uncover all its components and think about how they fit together. Furthermore, it is unclear to me whether Walzer would agree to view it this way and count (even with discounting) the risk to one's own soldiers as a moral cost (or diminished moral benefit) in the same equation with injury to the noncombatant civilians. For he seems to hold (p. 157) that one must run increased risks in order to safeguard noncombatants all the way up to the point "where any further risk-taking would almost certainly doom the military venture or make it so costly that it could not be repeated."


Walzer's discussion of sieges and blockades is very fine, and his condemnation of terrorism directed against innocent civilians is surely right. Walzer applies this point to the FLN in Algeria, also condemning the elitism of terrorists, points he made publicly in the late 1960s in Cambridge before intensely disapproving young leftists.

Reprisals as a means of moral enforcement raise interesting issues, for though directed to enforce moral norms another is violating, they need not be done directly to the violator. Walzer discusses and condemns the reprisals by Free French forces who executed 80 captured German soldiers in response to German summary executions of 80 militarily dressed captured Free French forces. His discussion of Israeli reprisals against sporadic attacks is judicious and discriminating, and he defends the Israeli reprisal in 1968 against property (Lebanese aircraft) in response to PLO attacks (based in Beirut) against an El Al plane in Athens. Not any reprisal goes, however, as he argued also in a recent New Republic discussion of the insufficiently limited and pinpointed Israeli reprisal in Lebanon. But if another nation will not or is not able to control action against another from within its borders, then the attacked party may act in self-defense.

I found that Walzer's discussion of guerrilla warfare pays insufficient attention to the way guerrillas exploit the morality of those they attack. If it is important that the noncombatants be immune from attack, it is important that the combatants be distinguishable when encountered; such is the function of military uniforms. (So perhaps we should not scorn those who delight in wearing military uniforms, for they groove on the garb that demarcates them as legitimate targets.) What then shall we say of those combatants who refuse such garb and who presumably would imitate any special markings civilians adopted to designate themselves? When civilians are harmed, does all responsibility for this lie with those who act upon them, or does some rest on those who by their actions make the civilians so vulnerable to being confused with combatants? And when guerrillas operate in an area, living in the midst of civilians whose presence is a moral shield for the guerrillas and an extra moral restraint on the means those attacking the guerrillas may use, do the civilians have any responsibility to get out of the way?

If someone shoots at passersby from the window of my office and the police arrive to cope with him and, while not being kept a prisoner there, I insist on staying, shouting out, "I'm not violating anyone's rights, and so I will stay here at my desk working; it's your job to make sure not to injure innocent people, including me," then if I do get hurt as the gunman is overpowered, is none of that my fault? Is it only the police who are to be condemned for being insufficiently discriminating? The questions also are of interest to libertarians who hold that anyone has a perfect right to do anything provided it does not violate the rights of others. Does my staying in my office violate anyone's rights? Am I under an obligation not to impede by my presence the apprehension of the gunman? (Must others induce me to get out of the way, or may they demand I do so, with the consequences of my staying resting on my head?) And if I get shot in the middle of the confused melee, have my rights been violated by the police; am I owed the same compensation as if I had been held hostage there?

Walzer assumes that if a guerrilla movement cannot be dislodged from the population who point them out, then it has the effective democratic support of the population. This takes insufficiently seriously the ways the guerrillas retaliate against those who turn them in. (The only case I know where this did not happen is that of the Irgun, led by Menachem Begin, which did not attack the Hagganah people who helped the British apprehend them during what was known as the season.) Undislodgeable is not the same as popularly supported, given the fear or passivity of the others, as might be shown if ever a guerrilla group that came to power would permit (within Walzer's guidelines) its opposition to wage guerrilla warfare against it.


Part 4 of Walzer's book contains an interesting discussion of the rights of neutral nations; an argument that in supreme emergency of national existence, the moral limits are relaxed and may permissibly be violated (which you will have to read and ponder); and also an unsatisfactory discussion of nuclear deterrence. I do not say this as a special criticism, for I am not familiar with (and don't have up my sleeve) a satisfactory moral position on nuclear weapons, on a deterrence defense posture that threatens great destruction of civilian lives. (Must one just renounce such weapons and succumb to the brandisher of them, who is willing to use them to crush one's non-nuclear might and civilian nonviolent resistance strategy?) We need some libertarian thinking on these issues.

I have touched only on some of the topics in this rich and stimulating book—those that sprang to mind as I looked at the table of contents six months after reading the book, when asked to review it here. You will learn a lot from this book and be led to new thoughts (sometimes different from Walzer's). If not, go ask Walzer for your money back. It's inappropriate, anyway, for an editor of the socialist magazine Dissent to receive and keep too many royalties. (Do you hear that, Irving Howe?)

Robert Nozick teaches philosophy at Harvard and is the author of Anarchy, State, and Utopia. Copyright 1978 by Robert Nozick