uring the euphoria of the opening air campaign against Baghdad, commentary was filled with triumphal rhetoric about hitting ''legitimate military objectives'' while causing little or even no loss of civilian life. A week ago, the air war was pounding more than symbols of the regime, and the ground war had become a real war. Scarcely a speech, briefing or interview was being given that failed to mention the laws of war. The Iraqi regime, for its part, was broadcasting denunciations of American airstrikes replete with images of corpses and wounded civilians in hospitals; the United States responded that Iraq had systematically situated military targets and equipment in the midst of civilian areas. The sheer frequency of these references on all sides belied the ancient maxim inter arma silent leges -- in time of war law is silent.

People throughout the world obviously care about what is called jus in bello, law governing conduct during war. This is so even if they differ about jus ad bellum, law governing not the conduct of war but rather the resort to force itself. But even while there is agreement on the need for fundamental rules governing the conduct of war, there is profound disagreement over who has authority to declare, interpret and enforce those rules, as well as who -- and what developments in the so-called art of war -- will shape them now and into the future. In short, who ''owns'' the law of war?

Although most of the world's religious and ethical traditions, if they admit the moral possibility of war at all, say something about what conduct is permitted in war, modern law of war descended historically as a tenet of traditional Christian just-war theory. The practical expression of such law, however, began with the founding of the International Committee of the Red Cross in 1863. The moving force behind the establishment of the Red Cross was Henri Dunant, a Genevan who witnessed and later wrote a widely read account of the Battle of Solferino in the 1859 war between Austria and France. Dunant, together with inhabitants of a village near the battlefield, went about the appalling task of trying to tend to the thousands of wounded who had simply been left to die. Without bandages, stretchers, doctors or medicines, and above all without significant interest in the wounded by their governments, there was little to be done except offer water and prayer. In the aftermath, the Red Cross, organized to do what had not been done at Solferino, became the world's first secular international nongovernmental organization, the ur-N.G.O. One of its goals was self-contradictory and even ludicrous on its face -- to bring humanity to the battlefield -- but its idealism was expressive of a sweeping call for reform among the middle classes in late-19th-century Europe and America, where an indigenous humanitarian movement for the care of soldiers had taken root during the Civil War.

In Europe, the same spirit of reform touched governments and royal courts, as a series of diplomatic conferences set about codifying the centuries-long tradition of ''laws and customs of war.'' The most important result was the 1907 Hague Convention, which in 56 short articles covers vast legal terrain on the conduct of warfare -- including surrender and flags of truce, obligations to wear uniforms, treatment of prisoners of war, sieges and bombardments, protection of cultural property, prohibitions against pillage and terms of occupation.

Its general rules are still applicable, at least in principle, today. Indeed, most of the matters that coalition forces raise as violations of the laws of war by Iraqi forces -- perfidious surrender, fighting out of uniform or mistreatment of P.O.W.'s -- can be found somewhere in the Hague regulations.

Optimism was swept away a few years later, however, by the guns of August. During World War I, in which mustard gas, aerial warfare, tanks and, above all, the machine gun, were introduced, old rules were clearly no longer enough, and existing humanitarian organizations were simply unable to cope with suffering on a scale never before seen. Following the Great War, there was a resurgence of interest in the fields of jus ad bellum and jus in bello. The Red Cross built on its earlier work in fostering treaties for sick and wounded soldiers and moved directly into promoting humanitarianism in war through legal rules, convening meetings between states that eventually resulted in the Geneva Conventions of 1929. These conventions, along with the 1907 Hague Convention, were the primary codified laws of war in effect when World War II broke out. In addition, states negotiated the 1925 Geneva Protocol prohibiting gas weapons -- a ban that held with few exceptions, even during World War II, for 60 years, until Saddam Hussein broke it in the Iran-Iraq war and again in his genocidal 1988 Anfal campaign against the Kurds.

World War II brought its own violations, partly owing to the new technology of air war against civilians. The carpet bombing of Dresden, for instance, inevitably swept civilians in with soldiers as targets. The conclusion of World War II brought about two signal developments in the laws of war. The first was the holding of criminal trials by the victors of those deemed to be chiefly responsible for the war. We tend to think of the Nuremberg trials as war-crimes trials, but in fact Nuremberg was principally about trying German leaders for the crime of aggressive war, for making war itself, crimes of jus ad bellum, rather than for the manner of its conduct. The chief American prosecutor, Robert Jackson, was content to leave what he regarded as the legally less-cutting-edge matters, the war's atrocities, to prosecutors from other countries.

Jackson's seeming triumph at Nuremberg, however -- establishing that international judges could try and convict a nation's leaders for the crime of aggression -- did not survive the establishment of the United Nations and the Security Council. The matter of determining aggression and maintaining international peace was stripped from international jurists and thrust into the hands of the permanent members of the Security Council, victors of World War II.

Of course, the Security Council, far from maintaining peace and security, has served as little more than a talking shop for nearly all of its history. The willingness of one of the permanent five members to brush off the Security Council when its core national interests are threatened -- as the United States and Britain just did with respect to Iraq -- is a norm of the Security Council, not a departure from it. Every permanent member has such interests, which are not necessarily even questions of national security but are frequently matters of ambition or even sentimental attachment. They include Tibet and Taiwan for the Chinese; Chechnya and, in 1999, Serbia for the Russians; and Francophone Africa for the French.

The enduring law established at Nuremberg has thus turned out not to be the ''crime of aggression'' but a reaffirmation of war crimes as traditionally understood -- with two important innovations made necessary by the Nazi death camps: genocide and crimes against humanity. Nuremberg also had serious gaps. Most significant, it failed to address the terror bombing of civilians and the deliberate consuming of whole cities (Dresden, Tokyo) by fire -- the most enthusiastic practitioners of the latter being the Allies.

The failure to prosecute the Allies for firebombing cities is one of the strongest arguments today for why war-crimes tribunals should not be conducted by the victors. Many regard this argument as so clinching, in fact, that the mere charge of ''victor's justice'' is enough to end debate. Yet it is far from obvious to me that ''impartial,'' uninvolved parties automatically carry greater moral authority than victors. Would it have been morally better, for example, for the victors of World War II, who had paid the price in blood, to have handed justice over to those countries that had remained neutral and refused to become involved? What matters far more than the supposedly virtuous impartiality that comes from neutrality is the quality of justice served in each case.

Another development in the laws of war following World War II was the four Geneva Conventions of 1949 -- the first covering wounded or sick soldiers; the second, shipwrecked sailors; the third, prisoners of war; and the fourth, civilians and occupation. The Geneva Conventions also introduced, for the first time in the canonical laws-of-war treaties, individual criminal liability and mandatory ''universal jurisdiction'' -- the ability (indeed, the obligation) of any and every state to try individuals (or turn them over to a state that would) whenever there was evidence of ''grave breaches'' of the Geneva Conventions.

The question of who is qualified to assert jurisdiction and then judge cases of war crimes is vital to determining who owns the laws of war. Universal jurisdiction says, in effect, lots of people own the laws -- but it leaves open the possibility of widely differing interpretations. A case in point is the continuing argument over whether the detainees at Guantanamo are indeed bona fide P.O.W.'s being mistreated under the Third Geneva Convention, as Amnesty International argues; or whether the Taliban among them are P.O.W.'s but some members of Al Qaeda are not, as Human Rights Watch says; or whether none of them are P.O.W.'s at all, as the Bush administration insists. Related to this is the question of whether detainees can be determined not to be P.O.W.'s by any means other than individual hearings. Many human rights advocates simply assume that the determination of whether a detainee is a P.O.W. must be reached by an individual hearing. This is a reading of the treaty with which even many conservatives agree. (Indeed, it is the reading presumed in 1997 United States military regulations that long precede Sept. 11.) But while it is surely the best interpretation of the Third Geneva Convention, it is not necessarily the way a literal reader would interpret phrases in Article 5's language like ''competent tribunal'' and ''should any doubt arise.'' It is (barely, but literally) available to the Bush administration to maintain that a) no doubt arises as to the status of the prisoners at Guantanamo and that b) the determination of whether any doubt has arisen does not necessarily require a hearing by a ''competent tribunal'' but merely a finding by the secretary of defense or the president.

This apparently esoteric legal issue, seemingly nothing but sheer technicality, is important. Why? It is not because it prevents someone from being charged with war crimes -- anyone can be charged with war crimes -- but because those determined not to be bona fide P.O.W.'s have far fewer procedural rights at trial. Bona fide P.O.W.'s (even those charged with war crimes) must be given the same legal protections at trial that an American soldier, accused of the same crime, would be given (and these are considerable), whereas an ''unprivileged combatant'' receives only minimal due process protections. The issue is about to arise again, this time in connection with Iraq's fedayeen militia fighters, who wear no uniforms and therefore might not qualify as bona fide P.O.W.'s, some of whom may be transferred shortly to Guantanamo.

This dispute points, in addition, to an unsettling feature about the struggle over who ''owns'' the law of war. The various constituencies that believe it belongs to them -- in this case, the United States government on the one side and human rights organizations on the other -- feel little obligation to acknowledge frankly the arguments made against their legal interpretations.
 

Despite these struggles over authority to interpret the 1949 Geneva Conventions, they are accepted universally as binding law on the treatment of people under a particular status -- soldier, sailor, P.O.W., civilian. The fundamental moral and legal principles are plain. First, noncombatants may never be made the object or target of attack, nor may noncombatants ever be used by defenders to shield legitimate military objectives from attack. Second, military operations, whether by attackers or defenders, must be undertaken with care to distinguish between noncombatants and combatants. Moreover, military officials must refrain from operations likely to produce collateral civilian damage that is excessive relative to the military advantages. So, for example, if the Iraqi Republican Guard has based itself in a crowded Baghdad neighborhood -- and even if it has done so illegally, by refusing to order or even allow civilians to evacuate, in order to use civilians as a shield -- the United States must still weigh the military advantages of attacking against the probable cost of civilian deaths and injuries.

The first principle is categorical, admitting of no exceptions. It leads in turn to a cluster of specific, categorical requirements aimed at ensuring that noncombatants or soldiers who are no longer in combat, like P.O.W.'s, are not attacked -- for example, requirements that combatants wear uniforms or other identifying marks and carry weapons openly. Although some of these categorical obligations have ambiguities -- what qualifies as a uniform? -- and although evidence, interpretation and intent might complicate matters, in principle they are either adhered to or not.

The second principle -- refraining from causing excessive collateral civilian damage -- involves, by contrast, a weighing of costs and benefits and, making things still more difficult, those costs and benefits as they might accrue in the future. Every day, every night, Air Force lawyers and planners must consider possible targets and weigh what they think the military value might be, in the future course of war, against the best intelligence data on how many civilians might be killed or injured, or how much civilian property destroyed. It is a thankless game of guesswork. By their nature, such judgments involve factual evaluations and guesses that cannot be legally challenged, unless something approximating willful, intentional gross negligence can be shown. Gross negligence has to consist of more than a lot of collateral damage, including gruesome civilian death and injury, that might be the result simply of a cruise missile aimed in good faith but gone astray. Legal culpability cannot be determined simply by looking at the level of damage and the death and injury caused. There is no moral equivalence between stray missiles aimed in good faith, using the best technology available, and deliberate violation of the categorical rules of war, like using human shields, shelling civilians to prevent them from fleeing Basra and rape or summary execution of prisoners. There can be no element of judgment, or weighing of costs and benefits, in deciding whether or not to target civilians or take them hostage; it is always wrong.

And yet. The reality remains for many that this moral distinction is sophistry. How can it not be, when we see every day on television and in the newspapers (especially non-American ones) so much death and injury to civilians? To speak of ''aiming'' at one thing while simultaneously foreseeing that, in a sizable percentage of cases, you would ''accidentally'' hit another -- if this is what ''rules'' of war consist of, then they are no more than artificial salves on the consciences of combatants.

Moreover, it does seem to millions of people worldwide that there is indeed a moral equivalence between the tactics of the Americans -- hitting targets from the air and pleading collateral damage as a defense against responsibility -- and the tactics of the Iraqis, who, lacking other means to attack, use their own civilians as a material and moral resource, no matter what laws of war it might violate. This was the attitude, it should be said, held by Churchill, who intended a scorched-earth defense of Britain (including the use of poison gas) without much regard for the lives of British civilians, should the invader ever arrive.

There is, I think, only one way to evaluate these conflicting claims. The idea of ''acceptable'' collateral damage is firmly embedded in Western legal and moral thought, but in fact it is the product of a far more particular Christian strand of moral thought than many of us, accustomed to the tradition, readily admit. The ''doctrine of the double-effect,'' turning on a supposed moral difference between intended and unintended but foreseeable consequences, is not morally obvious. It can be defended, however, as a moral doctrine if we consider the alternatives. To deny the distinction means that you either accept that virtual nonviolence is the only tenable position or that you are indifferent to the lives of civilians, since you are guilty of anything that happens anyway -- and in that case, anything becomes a target. The justification for the principle of the double-effect is that it appears to be the only principled way of steering between a pacifism that few of us, in real life, would accept, and a brutal realism that denies the moral necessity of even trying to distinguish between combatants and noncombatants.

Even if you accept the principle of this distinction, however, it must be with a knowledge that it is a compromise affair. It therefore puts a great moral burden on those who fight to find better ways to separate civilians from fighters and to improve the ability, through technology or other means, to aim and hit solely military targets. And if war is, as the poet Rene Char wrote, ''this time of damned algebra,'' a matter of endless calculation and recalculation of effects, then the law of war must take that into account. It consists on the one hand of both categorical demands and prohibitions, and on the other of calculations of cost and benefit, civilian loss and military advantage -- and these calculations are always in flux.
 

The attempt to address these complexities and make law of them was undertaken by United Nations negotiations in the 1970's. The 1977 Additional Protocol I grew out of these negotiations. Although now ratified by some 160 nations with varying ''reservations'' (statements as to certain treaty articles a country does not accept as binding), the United States has never ratified it (nor, it should be noted, has Iraq). Yet without accepting the treaty as such, United States officials over the past 20 years have indicated that the United States accepts various parts of the treaty without accepting the whole. It remains a disappointment and a puzzle that the Department of Defense has never been willing to state publicly and definitively which parts it accepts and which parts it does not and why.

But the American problems with Protocol I generally fall into three main categories. First, certain provisions are unacceptably political in nature. Jus in bello has always insisted on exactly the same treatment for all sides in combat, the same rules whether for the Allies or for the Germans, communists or capitalists. Protocol I, however, grants combatants rights, including the vital right to be treated as a P.O.W., on the basis of certain motives for fighting, referring specifically to those who fight against ''racist regimes'' (as in South Africa under apartheid) or ''alien occupation'' (as in Israel).

Second, certain provisions appear to the United States to restrict methods and means of warfare that it believes are legitimate. For example, Protocol I contains no exceptions in its rules for nuclear weapons, while at the same time it categorically prohibits reprisals against civilians, including the use of nuclear weapons in reprisal for a nuclear attack, which is the basis for nuclear deterrence.

The third category of America's objections concerns rules in Protocol I that are aimed at accommodating guerrillas and irregular fighters, as during the Vietnam War or in Iraq. Unquestionably, these rules make life legally easier for irregular fighters, and some would see this as making the rules of war more ''fair.'' Yet the rules also create new risks for civilians. For example, the protocol grants legal combatant status even to guerrillas who conceal themselves and their weapons among the civilian population, as long as the fighters reveal themselves to the adversary ''preceding the launching of an attack'' -- which is to say, often shortly before attacking from among the civilians who will, inevitably, be caught in the crossfire. It is unfortunate for Saddam's irregulars that Iraq and the United States have not ratified the protocol, as it would have provided legal protection for many of those fighters' attacks, if surely not the civilians providing them with cover.

In addition, sections of Protocol I, while dealing with indiscriminate attacks -- military operations that fail to distinguish between combatants and noncombatants -- in detail, mention the obligations of defenders far more briefly. This is despite the fact that the level of collateral damage incurred in military operations is often determined by where the defender chooses to locate its military assets.

And so Amnesty International released a report last year on the actions of the Israeli Defense Forces in Jenin that signally failed even to mention the legal obligations of the Palestinian forces toward their own civilians. Human Rights Watch also issued an entire report on Jenin that raised only in a few sentences the fact that Palestinian fighters had situated themselves among civilians. (To Human Rights Watch's credit, and perhaps in response to criticism, it has begun taking careful note of the obligations of forces on the defensive as well as those of attackers.) And the International Red Cross, in a message from its president at the commencement of the Iraq war, called in general terms upon parties to observe the laws of war but dwelt mainly upon attackers, the United States and Britain, neglecting to say anything specific of the Iraqi defenders. But the inescapable fact is that the structure of Protocol I practically invites such neglect.
 

The fact remains that every war is a petri dish for the next round of the laws of war. And while the war in Iraq is principally about well-established legal principles, and their violation, it, too, will end with a reconsideration of the laws dictating how war should be waged.

For the past 20 years, the center of gravity in establishing, interpreting and shaping the law of war has gradually shifted away from the military establishments of leading states and their ''state practice.'' It has even shifted away from the International Red Cross (invested by the Geneva Conventions with special authority) and toward more activist and publicly aggressive N.G.O.'s -- including the ad hoc coalitions that produced the Ottawa Treaty, banning land mines, and the new International Criminal Court. These N.G.O.'s are indispensable in advancing the cause of humanitarianism in war. But the pendulum shift toward them has gone further than is useful, and the ownership of the laws of war needs to give much greater weight to the state practices of leading countries. This does not mean that state practice is all that matters, nor does it mean that all state practice matters -- Iraq, after all, is a state, and it is fighting, too -- but it does mean that the state practice of democratic sovereigns that actually fight wars should be ascendant in shaping the law. And this includes raising the standards of the laws of war to reflect, for example, advances in technology and precision weapons, standards that should become the norm for leading militaries, first for NATO and then beyond.

N.G.O.'s are also wedded far too much to a procedural preference for the international over the national. But that agenda increasingly amounts to internationalism for its own sake, and its specific purpose is to constrain American sovereignty. It thus promotes, embedded in an agenda of human rights and the laws of war, the ceding of sovereignty, even democratic sovereignty, as the most virtuous act that a state can perform on behalf of its citizens. This agenda of privileging internationalism, unfortunately, is even sometimes allowed to override obvious steps backward in the laws of war, like privileging guerrilla combatants over the civilians in their midst. For this reason, one consequence of the Iraq war for the future of the laws of war will have to be an understanding that the solicitude of Protocol I for irregular fighters hiding among civilians is wrong and that the United States was right to have rejected it.

More broadly in recent years, the N.G.O.'s have been promoting an ever more utopian law of war, in keeping with absolutist human rights ideology. In practice, alas, this utopianism is aimed only at one side in conflicts -- the side that in fact tries to obey the law. And so a second consequence of the war in Iraq for the future of the law of war will have to be a halt to raising the standards ever higher for protecting the civilian population when that burden effectively falls only on attacking forces, unreciprocated. The status quo has the effect of rewarding defending forces for recognizing that war crimes against their own civilians are the best strategy against a powerful but scrupulous enemy. It risks in the end creating a law of war that assumes, for all practical purposes, that the burden is all on one side, the side with the more advanced technology and the less desperate military. After the last cruise missile has been launched and the last irregular fighter silenced, we will look back on the war that was wrought. What we will find is that the meaning of ''asymmetric'' warfare is not what we thought. The issue is not so much disparities in technology. Instead, a form of warfare has re-emerged that tacitly assumes, indeed permits, that the weaker side must fight by using systematic violations of the law and its method. This is unsustainable as a basis for the law of war. Reciprocity matters.
 

Kenneth Anderson is a law professor at American University and a research fellow at the Hoover Institution, Stanford University. He is the legal editor of the book ''Crimes of War.''