But, though I fear I may get thrown out of polite society for saying so, I think that choice of terms should be resisted. The choice of words matters because "genocide" has come to serve as a conversation-stopper, shutting down reflection on the costs and benefits of various policy options in favor of vague-but-insistent injunctions to "do something," anything, to make it stop. The legal definition is a bit broad, and it's best to leave it up to the lawyers at the State Department and the United Nations to argue over whether it fits the case at hand. What comes to mind when non-lawyers hear the word "genocide," however, is something akin to the events of the Holocaust, where a regime pursues the destruction of an ethnic group as an end in and of itself. Without denying that monstrous things are being done in Darfur, I don't think that genocide -- in this sense -- is what's happening.

In defense of my apparently indefensible position on this, I'll say that several months ago when I was busy expressing an admittedly ignorant skepticism about the desirability of mounting a military intervention in Darfur, everyone recommended to me a book on the subject by Gerard Prunier. It's common enough in Washington to be recommended a book, then upon picking it up finding oneself uncertain as to whether the recommender had actually read it. In this case, though, I've been wondering if Prunier's fans even bothered to read the title -- Darfur: The Ambiguous Genocide.

In case there's any, um, ambiguity about this, Prunier is conveniently also the author of a book about Rwanda, titled The Rwanda Crisis: History of Genocide. Prunier's French, and the French title of his latest book is Le Darfour: Un genocide ambigu, while his Rwanda book was called Rwanda: le genocide. No special foreign language expertise should be needed to see where I'm going with this.

On some level, of course, debates about whether or not some given set of events involving the state-sponsored killing of a lot of people amount to "genocide" or not is purely semantic. The facts about Pol Pot's Cambodia aren't seriously in dispute, but some people call it "genocide" while others prefer not to, a disagreement about the meaning of the term rather than about what happened in Cambodia. Prunier chose to term what's been happening in Darfur as an "ambiguous genocide" which is a good choice in many ways. On another level, however, it's deeply problematic because in political terms the salient feature of genocide is that, according to an influential body of opinion, there's nothing ambiguous about it.

The significance of the terminology is brought home by the unsigned editorial that leads this week's New Republic special report on Darfur. The magazine, to no one's surprise, wants war. The editorial itself says curiously little about Sudan. It does, however, say a great deal about genocide, the point being that the monstrousness of the act should put devising a response beyond cost-benefit analysis and what the editorial itself terms "the foundations of diplomacy and also of common sense." We must violate these things because "if you are not willing to use force against genocide immediately, then you do not understand what genocide is. Genocide is not a crisis that escalates into evil. It is evil from its inception." This is not a line of thought that has much room for an ambiguous genocide, hence the oddity of a book review in the same issue that calls for war, cites Prunier as an authority, and refers to genocide simpliciter without reference to the ambiguous modifier.

If disambiguation is needed, I would suggest a schema laid out in Benjamin Valentino's excellent book Final Solutions: Mass Killing and Genocide in the Twentieth Century. Valentino divides mass killings into three major types. You have classic "ethnic genocides" such as perpetrated by Nazi Germany or in Rwanda. You have "Communist mass killings" perpetrated in the Soviet Union, China, and Cambodia. Last, you have "counter-guerilla mass killings" as seen during Guatemala's civil war or the Soviet occupation of Afghanistan in which atrocities are committed in an effort to break the will of armed resistance. Delicate souls may doubt, as Andrew Loewenstein writing in the same TNR issue apparently does, that the sort of brutal slaughter happening in Darfur could be "the hallmarks of counterinsurgency warfare" but the reality is that waging war on civilians has a long history as a military tactic.

Consider Saddam Hussein's Anfal campaign against Iraqi Kurdistan, another good candidate for the title of an ambiguous genocide. The Human Rights Watch report on the subject doesn't see the ambiguity, but acknowledges the basic reality that Saddam did not, in fact, kill all of Iraq's Kurds. What's more, the report states that "outside pressure . . . was not a significant contributing factor to the amnesty" that ended the slaughter. Rather, "it is clear from the Fifth Corps report on the Final Anfal that the decision to declare a general amnesty was made because Baghdad was convinced by September 6 that the peshmerga forces had been crushed." The inhumanity directed at civilians was, in other words, a military tactic, aimed at a military goal, the crushing of Kurdistan's armed guerillas who had risen in rebellion against the Iraqi state.

To cut a bit closer to home, in his biography British Air Marshall Arthur Harris justifies the bombing of Dresden in terms of "the effect on German morale, not only in Dresden but in far distant parts of the country" and notes that the operation "was at the time considered a military necessity by much more important people" than Harris. In light of their postwar conduct, it's perfectly clear that neither the British, American, nor even the Soviet government actually intended to exterminate the population of Germany in the way that the German government most certainly did want to exterminate the world's Jewish population. Nevertheless, as long as the war was on, violence against civilians would be conducted deliberately and without restraint.

The Japanese theater makes the utility of mass killing even more evident. With good reason, both the Roosevelt and Truman administrations dreaded the prospect of invading Japan. Instead, once America was in a place to do so it offered the Japanese an offer they couldn't refuse -- surrender or face extermination by a combination of starvation, incendiary bombing of civilian population centers, and eventually nuclear weapons. Japan chose to continue the war, and huge numbers of Japanese civilians were killed as a result. Once surrender was in hand, however, the Japanese were treated with a remarkable generosity -- mind-boggling brutal methods of warfare were no indication of exterminationist intent. Rather, given the weapons at America's disposal, they were judged the most efficient means of winning the war.

Obviously, the conflict in Sudan isn't the second World War, and Sudan's leaders aren't Harry Truman or Winston Churchill. Still, it remains the case that the leaders in Khartoum didn't wake up one morning and just decide to exterminate Darfur's inhabitants. The mass killing was adopted as a strategy in the midst of a war, and at the intersection of counter-guerilla mass killing and ethnic warfare lies the ambiguous genocide.

Does it matter? On one level, no. War crimes are war crimes, brutality is brutality, slaughter is slaughter, and we all have a duty to reduce its incidence. But once ambiguity re-enters the picture, so should common sense. Faced with counter-guerilla mass slaughter, you can't just stop the killing, any intervention necessarily entails taking a side on the basic question of the war. Advocates of intervening have a duty to explain what it is they intend to do -- create an independent Darfur? Controlled by whom? They also have a duty to answer, rather than simply dismiss, questions about the big picture of American foreign policy. How would attacking another Arab country affect America's larger security concerns? Would circumventing the UN merely provoke protests from China and like-minded human rights averse dictators, or will developing world democracies like India, South Africa, and Brazil see it as imperialism run amok? Is war with Sudan an efficient humanitarian measure when we could be curing measles without great controversy?

I'm prepared to be convinced of the case for coercive force, but so far its advocates have been long on table-pounding and short of persuasion. Faced with purest evil perhaps we really do have a duty not to think things through, but faced with ambiguity surely the obvious questions deserve answers before bombs start dropping.

Matthew Yglesias is a Prospect staff writer.