The parameters of the discussion about nuclear weapons are well known and appear to be relatively fixed. It seems as if there has been little on that front in forty years. Most civilian scholars have lost interest in nuclear weapons and moved on to other topics. But it is the habit of the military mind to learn from the past; even today there are lessons to be learned from Cannae, Waterloo, and Vicksburg. It will not surprise thoughtful military officers to find that the past has something important and interesting to tell us about nuclear weapons. The conventional wisdom is that nuclear weapons are horrible, probably immoral, but necessary. We keep them because they have a unique ability to coerce and deter. There are psychological characteristics to the weapons--as Secretary of War Henry Stimson pointed out in the first semiofficial discussion of them in 1947--that make them unlike other weapons. (1) Now evidence is throwing doubt on these decades-old conclusions. Actually not new evidence, but additional evidence culled from a careful study of the past. Hiroshimas The first and most important revision to history has to do with the efficacy of bombing Hiroshima and Nagasaki. (2) This evidence, however, has nothing to do with the school of Hiroshima history. The revisionist school ascended in 1964 with the publication by Gar Alperovitz of a book arguing that bombing Hiroshima and Nagasaki was unnecessary--the Japanese would have surrendered anyway. (3) This debate has caused controversy and aroused passions for almost fifty years. But it is not really about nuclear weapons. The revisionists argue that the bombings were horrible and, since they weren't necessary to win the war, they were immoral. The counterrevisionists argue that the bombings were required and were, therefore, moral. But this is a debate about whether the United States acted morally, not about whether nuclear weapons work. New evidence seems to suggest that while the bombs destroyed the cities, they didn't play much of a role (or perhaps any) in convincing Japan's leaders to surrender. Over the last twenty years, increasing access to records in Japan, Russia, and the United States has revealed that in the three days following the bombing of Hiroshima Japan's leaders had little idea that they had to surrender as a result of the bombings. (4) Meeting notes, diary entries, and the actions that various actors took during this period show that while Japan's leaders knew Hiroshima had been destroyed by a nuclear weapon, they saw this as another problem in an already difficult war, not a war-ending crisis. The Foreign Minister, Togo Shigenori, actually suggested convening the Supreme Council two days after the bombing of Hiroshima to discuss it and found he could not generate enough interest on the subject to get it on the agenda. When the Soviet Union, which had signed a five-year neutrality pact with Japan in 1941, broke that agreement and joined the war at midnight on 8-9 August, however, it touched off a crisis. Within hours of the news reaching Tokyo, the Supreme Council met to discuss unconditional surrender. It is clear from all the evidence now available that Japan's leaders surrendered because of the Soviet entry into the war and not because of the nuclear bombings. There are reasons to doubt the traditional story that the Emperor was horrified by the bombing of Hiroshima. The documentary evidence is thin, (5) and if the Emperor was so moved, it begs the question: why was he moved by secondhand reports of a city destroyed in August when he was not moved by driving through Tokyo and personally witnessing the devastation of that city in March? Would it not be sensible to expect that firsthand experience would have a stronger emotional impact than a secondhand report? (6) In some ways, this conclusion about Hiroshima makes sense. In order to believe that Hiroshima was the cause of Japan's surrender, it was necessary to believe that Japan's military men didn't know their business. …