(Corbis via Getty)

I imagine that the persistence of that question irritated Harry Truman above all other things. The atomic bombs that destroyed the cities of Hiroshima and Nagasaki fifty years ago were followed in a matter of days by the complete surrender of the Japanese empire and military forces, with only the barest fig leaf of a condition—an American promise not to molest the Emperor. What more could one ask from an act of war? But the two bombs each killed at least 50,000 people and perhaps as many as 100,000. Numerous attempts have been made to estimate the death toll, counting not only those who died on the first day and over the following week or two but also the thousands who died later of cancers thought to have been caused by radiation. The exact number of dead can never be known, because whole families—indeed, whole districts—were wiped out by the bombs; because the war had created a floating population of refugees throughout Japan; because certain categories of victims, such as conscript workers from Korea, were excluded from estimates by Japanese authorities; and because as time went by, it became harder to know which deaths had indeed been caused by the bombs. However many died, the victims were overwhelming civilians, primarily the old, the young, and women; and all the belligerents formally took the position that the killing of civilians violated both the laws of war and common precepts of humanity. Truman shared this reluctance to be thought a killer of civilians. Two weeks before Hiroshima he wrote of the bomb in his diary, "I have told [the Secretary of War] Mr. Stimson to use it so that military objectives and soldiers and sailors are the target and not women and children. … " The first reports on August 6, 1945, accordingly described Hiroshima as a Japanese army base.

This fiction could not stand for long. The huge death toll of ordinary Japanese citizens, combined with the horror of so many deaths by fire, eventually cast a moral shadow over the triumph of ending the war with two bombs. The horror soon began to weigh on the conscience of J. Robert Oppenheimer, the scientific director of the secret research project at Los Alamos, New Mexico, that designed and built the first bombs. Oppenheimer not only had threatened his health with three years of unremitting overwork to build the bombs but also had soberly advised Henry Stimson that no conceivable demonstration of the bomb could have the shattering psychological impact of its actual use. Oppenheimer himself gave an Army officer heading for the Hiroshima raid last minute instructions for proper delivery of the bomb.

Don't let them bomb through clouds or through an overcast. Got to see the target. No radar bombing; it must be dropped visually. ... Of course, they must not drop it in rain or fog. Don't let them detonate it too high. The figure fixed on is just right. Don't let it go up or the target won't get as much damage.

These detailed instructions were the result of careful committee work by Oppenheimer and his colleagues. Mist or rain would absorb the heat of the bomb blast and thereby limit the conflagration, which experiments with city bombing in both Germany and Japan had shown to be the principal agent of casualties and destruction. Much thought had also been given to finding the right city. It should be in a valley, to contain the blast; it should be relatively undamaged by conventional air raids, so that there would be no doubt of the bomb's destructive power; an educated citizenry was desired, so that it would understand the enormity of what had happened. The military director of the bomb project, General Leslie Groves, thought the ancient Japanese imperial capital of Kyoto would be ideal, but Stimson had spent a second honeymoon in Kyoto, and was afraid that the Japanese would never forgive or forget its wanton destruction; he flatly refused to leave the city on the target list. Hiroshima and Nagasaki were destroyed instead.

On the night of August 6 Oppenheimer was thrilled by the bomb's success. He told an auditorium filled with whistling, cheering, foot-stomping scientists and technicians that he was sorry only that the bomb had not been ready in time for use on Germany. The adrenaline of triumph drained away following the destruction of Nagasaki, on August 9. Oppenheimer, soon offered his resignation and by mid-October had severed his official ties. Some months later he told Truman in the White House, "Mr. President, I have blood on my hands."

Truman was disgusted by this cry-baby attitude. "I told him," Truman said later, "the blood was on my hands—let me worry about that."

Till the end of his life Truman insisted that he had suffered no agonies of regret over his decision to bomb Hiroshima and Nagasaki, and the pungency of his language suggests that he meant what he said. But it is also true that he ordered a halt to the atomic bombing on August 10, four days before the Japanese Emperor surrendered, and the reason, according to a Cabinet member present at the meeting, was that "he didn't like the idea of killing ... 'all those kids.' "

Was it right? Harry Truman isn't the only one to have disliked the question. Historians of the war, of the invention of the atomic bomb, and of its use on Japan have almost universally chosen to skirt the question of whether killing civilians can be morally justified. They ask instead, Was it necessary?

Those who say it was necessary argue that a conventional invasion of Japan, scheduled to begin on the southernmost island of Kyushu on November 1, 1945, would have cost the lives of large numbers of Americans and Japanese alike. Much ink has been spilled over just how large these numbers would have been. Truman in later life sometimes said that he had used the atomic bomb to save the lives of half a million or even a million American boys who might have died in an island-by-island battle to the bitter end for the conquest of Japan.

Where Truman got those numbers is hard to say. In the spring of 1945, when it was clear that the final stage of the war was at hand, Truman received a letter from former President Herbert Hoover urging him to negotiate an end to the war in order to save the "500,000 to 1 million American lives" that might be lost in an invasion. But the commander of the invasion force, General Douglas MacArthur, predicted nothing on that scale. In a paper prepared for a White House strategy meeting held on June 18, a month before the first atomic bomb was tested, MacArthur estimated that he would suffer about 95,000 casualties in the first ninety days—a third of them deaths. The conflict of estimates is best explained by the fact that they were being used at the time as weapons in a larger argument. Admirals William Leahy and Ernest J. King thought that Japan could be forced to surrender by a combination of bombing and naval blockade. Naturally they inflated the number of casualties that their strategy would avoid. MacArthur and other generals, convinced that the war would have to be won on the ground, may have deliberately guessed low to avoid frightening the President.

It was not easy to gauge how the battle would go. From any conventional military perspective, by the summer of 1945 Japan had already lost the war. The Japanese navy mainly rested on the bottom of the ocean; supply lines to the millions of Japanese soldiers in China and other occupied territories had been severed; the Japanese air force was helpless to prevent the almost nightly raids by fleets of B-29 bombers, which had been systematically burning Japanese cities since March; and Japanese petroleum stocks were close to gone. The battleship Yamato, dispatched on a desperate mission to Okinawa in April of 1945, set off without fuel enough to return.

But despite this hopeless situation the Japanese military was convinced that a "decisive battle" might inflict so many casualties on Americans coming ashore in Kyushu that Truman would back down and grant important concessions to end the fighting. Japan's hopes were pinned on "special attack forces," a euphemism for those engaged in suicide missions, such as kamikaze planes loaded with explosives plunging into American ships, as had been happening since 1944. During the spring and summer of 1945 about 8,000 aircraft, along with one-man submarines and "human torpedoes," had been prepared for suicide missions, and the entire Japanese population had been exhorted to fight, with bamboo spears if necessary, as "One Hundred Million Bullets of Fire." Military commanders were so strongly persuaded that honor and even victory might yet be achieved by the "homeland decisive battle" that the peace faction in the Japanese cabinet feared an order to surrender would be disobeyed. The real question is not whether an invasion would have been a ghastly human tragedy, to which the answer is surely yes, but whether Hoover, Leahy, King, and others were right when they said that bombing and blockade would end the war.

Here the historians are on firm ground. American cryptanalysts had been reading high-level Japanese diplomatic ciphers and knew that the government in Tokyo was eagerly pressing the Russians for help in obtaining a negotiated peace. The sticking point was narrow: the Allies insisted on unconditional surrender; the Japanese peace faction wanted assurances that the imperial dynasty would remain. Truman knew this at the time.

What Truman did not know, but what has been well established by historians since, is that the peace faction in the Japanese cabinet feared the utter physical destruction of the Japanese homeland, the forced removal of the imperial dynasty, and an end to the Japanese state. After the war it was also learned that Emperor Hirohito, a shy and unprepossessing man of forty-four whose first love was marine biology, felt pressed to intervene by his horror at the bombing of Japanese cities. The devastation of Tokyo left by a single night of firebomb raids on March 9–10, 1945, in which 100,000 civilians died, had been clearly visible from the palace grounds for months thereafter. It is further known that the intervention of the Emperor at a special meeting, or gozen kaigin, on the night of August 9–10 made it possible for the government to surrender.

The Emperor's presence at a gozen kaigin is intended to encourage participants to put aside all petty considerations, but at such a meeting, according to tradition, the Emperor does not speak or express any opinion whatever. When the cabinet could not agree on whether to surrender or fight on, the Premier, Kantaro Suzuki, broke all precedent and left the military men speechless when he addressed Hirohito, and said, "With the greatest reverence I must now ask the Emperor to express his wishes."

Of course, this had been arranged by the two men beforehand. Hirohito cited the suffering of his people and concluded, "The time has come when we must bear the unbearable." After five days of further confusion, in which a military coup was barely averted, the Emperor broadcast a similar message to the nation at large in which he noted that "the enemy has begun to employ a new and most cruel bomb. ... "

Are those historians right who say that the Emperor would have submitted if the atomic bomb had merely been demonstrated in Tokyo Bay, or had never been used at all?

Questions employing the word "if" lack rigor, but it is very probable that the use of the atomic bomb only confirmed the Emperor in a decision he had already reached. What distressed him was the destruction of Japanese cities, and every night of good bombing weather brought the obliteration by fire of another city. Hiroshima, Nagasaki, and several other cities had been spared from B-29 raids and therefore offered good atomic-bomb targets. But Truman had no need to use the atomic bomb, and he did not have to invade. General Curtis LeMay had a firm timetable in mind for the 21st Bomber Command; he had told General H. H. ("Hap") Arnold, the commander in chief of the Army Air Corps, that he expected to destroy all Japanese cities before the end of the fall. Truman need only wait. Steady bombing, the disappearance of one city after another in fire storms, the death of another 100,000 Japanese civilians every week or ten days, would sooner or later have forced the cabinet, the army, and the Emperor to bear the unbearable.

Was it right? The bombing of cities in the Second World War was the result of several factors: the desire to strike enemies from afar and thereby avoid the awful trench-war slaughter of 1914–1918; the industrial capacity of the Allies to build great bomber fleets; the ability of German fighters and anti-aircraft to shoot down attacking aircraft that flew by daylight or at low altitudes; the inability of bombers to strike targets accurately from high altitudes; the difficulty of finding all but very large targets (that is, cities) at night; the desire of airmen to prove that air forces were an important military arm; the natural hardening of hearts in wartime; and the relative absence of people willing to ask publicly if bombing civilians was right.

"Strategic bombing" got its name between the wars, when it was the subject of much discussion. Stanley Baldwin made a deep impression in the British House of Commons in 1932 when he warned ordinary citizens that bombing would be a conspicuous feature of the next war and that "the bomber will always get through."

This proved to be true, although getting through was not always easy. The Germans soon demonstrated that they could shoot down daytime low-altitude "precision" bombers faster than Britain could build new planes and train new crews. By the second year of the war the British Bomber Command had faced the facts and was flying at night, at high altitudes, to carry out "area bombing." The second great discovery of the air war was that high-explosive bombs did not do as much damage as fire. Experiments in 1942 on medieval German cities on the Baltic showed that the right approach was high-explosive bombs first, to smash up houses for kindling and break windows for drafts, followed by incendiaries, to set the whole alight. If enough planes attacked a small enough area, they could create a fire storm—a conflagration so intense that it would begin to burn the oxygen in the air, creating hundred-mile-an-hour winds converging on the base of the fire. Hamburg was destroyed in the summer of 1943 in a single night of unspeakable horror that killed perhaps 45,000 Germans.

While the British Bomber Command methodically burned Germany under the command of Sir Arthur Harris (called Bomber Harris in the press but Butch—short for "Butcher"—by his own men), the Americans quietly insisted that they would have no part of this slaughter but would instead attack "precision" targets with "pinpoint" bombing. But American confidence was soon eroded by daylight disasters, including the mid-1943 raid on ball-bearing factories in Schweinfurt, in which sixty-three of 230 B-17s were destroyed for only paltry results on the ground. Some Americans continued to criticize British plans for colossal city-busting raids as "baby-killing schemes," but by the end of 1943, discouraged by runs of bad weather and anxious to keep planes in the air, the commander of the American Air Corps authorized bombing "by radar"—that is, attacks on cities, which radar could find through cloud cover.

The ferocity of the air war eventually adopted by the United States against Germany was redoubled against Japan, which was even better siuited for fire raids, because so much of the housing was of paper and wood, and worse suited for "precision" bombing, because of its awful weather and unpredictable winds at high altitudes. On the night of March 9–10, 1945, General LeMay made a bold experiment: he stripped his B-29s of armament to increase bomb load and flew at low altitudes. As already described, the experiment was a brilliant success. By the time of Hiroshima more than sixty of Japan's largest cities had been burned, with a death toll in the hundreds of thousands.

No nation could long resist destruction on such a scale—a conclusion formally reached by the United States Strategic Bombing Survey in its Summary Report (Pacific War): "Japan would have surrendered [by late 1945] even if the atomic bombs had not been dropped, even if Russia had not entered the war [on August 8], and even if no invasion had been planned or contemplated."

Was it right? There is an awkward, evasive cast to the internal official documents of the British and American air war of 1939–1945 that record the shift in targets from factories and power plants and the like toward people in cities. Nowhere was the belief ever baldly confessed that if we killed enough people, they would give up; but that is what was meant by the phrase "morale bombing," and in the case of Japan it worked. The mayor of Nagasaki recently compared the crime of the destruction of his city to the genocide of the Holocaust, but whereas comparisons—and especially this one—are invidious, how could the killing of 100,000 civilians in a day for a political purpose ever be considered anything but a crime?

Fifty years of argument over the crime against Hiroshima and Nagasaki has disguised the fact that the American war against Japan was ended by a larger crime in which the atomic bombings were only a late innovation—the killing of so many civilians that the Emperor and his cabinet eventually found the courage to give up. Americans are still painfully divided over the right words to describe the brutal campaign of terror that ended the war, but it is instructive that those who criticize the atomic bombings most severely have never gone on to condemn all the bombing. In effect, they give themselves permission to condemn one crime (Hiroshima) while enjoying the benefits of another (the conventional bombing that ended the war).

Ending the war was not the only result of the bombing. The scale of the attacks and the suffering and destruction they caused also broke the warrior spirit of Japan, bringing to a close a century of uncontrolled militarism. The undisguisable horror of the bombing must also be given credit for the following fifty years in which no atomic bombs were used, and in which there was no major war between great powers. It is this combination of horror and good results that accounts for the American ambivalence about Hiroshima. It is part of the American national gospel that the end never justifies the means, and yet it is undeniable that the end—stopping the war with Japan—was the immediate result of brutal means.

Was it right? When I started to write this article, I thought it would be easy enough to find a few suitable sentences for the final paragraph when the time came, but in fact it is not. What I think and what I feel are not quite in harmony. It was the horror of Hiroshima and fear of its repetition on a vastly greater scale which alarmed me when I first began to write about nuclear weapons (often in these pages), fifteen years ago. Now I find I have completed some kind of ghastly circle.

Several things explain this. One of them is my inability to see any significant distinction between the destruction of Tokyo and the destruction of Hiroshima. If either is a crime, then surely both are. I was scornful once of Truman's refusal to admit fully what he was doing; calling Hiroshima an army base seemed a cruel joke. Now I confess sympathy for the man—responsible for the Americans who would have to invade; conscious as well of the Japanese who would die in a battle for the home islands; wielding a weapon of vast power; knowing that Japan had already been brought to the brink of surrender. It was the weapon he had. He did what he thought was right, and the war ended, the killing stopped, Japan was transformed and redeemed, fifty years followed in which this kind of killing was never repeated. It is sadness, not scorn, that I feel now when I think of Truman's telling himself he was not "killing … 'all those kids.' " The bombing was cruel, but it ended a greater, longer cruelty.

They say that the fiftieth anniversaries of great events are the last. Soon after that the people who took part in them are all dead, and the young have their own history to think about, and the old questions become academic. It will be a relief to move on.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.