The Doomsday Machine

by Daniel Ellsberg

  Hitler had regarded his attack on Warsaw as a demonstration of the fate of “defended” cities. Likewise, when the Germans bombed the center of Rotterdam on May 14, 1940, that city was under siege. Holland was still refusing to surrender, though negotiations were under way. The German ground commander, General Rudolf Schmidt, called for a bombing strike. But while the bombers were on their way, Schmidt tried to call them off because the garrison, he had learned, was on the verge of surrender. It was too late. Half the bombers did not get the message. Though Schmidt put up red flares as a warning that they should go back, the pilots didn’t understand and they destroyed the city center of Rotterdam. This occasioned a rare apology by the German military to the Dutch people.

  The initial word from the Dutch press was that thirty thousand people had died. In fact, it was just under a thousand. Nevertheless, the larger figure created an enormous sensation and Britain announced that they would not be bound by the promise that they had made to FDR, nor by the policy they had followed up to that point. The day after the bombing of Rotterdam, May 15, the British cabinet sent bombers into Germany for the first time against strategic targets in densely populated areas. The gloves were off.

  * * *

  I had long been following the phenomenon of strategic bombing, first horrified by the Nazi bombing of London, then following up when I had access to classified studies at the Air Force–sponsored RAND Corporation. One of the best accounts on the movement toward strategic bombing, which had a great influence on me when I read it at RAND, is The Road to Total War: Escalation in World War II, by Fred Sallagar; the unclassified RAND Report R-465-PR is dated 1969, but I first read it as an internal document about ten years before that.

  I had a number of talks about it with Fritz Sallagar, as we knew him. He was looking for lessons from World War II that might suggest how nuclear escalation could develop and proceed in a conventional war that turned nuclear. He was particularly concerned with the possibility of keeping such a war limited and under control. One of his themes was how often escalation developed in terms of misunderstandings, misinterpretations, and failures of command and control—as with the bombing of Rotterdam just described. That had served as the trigger and justification for Winston Churchill—who had come into office only four days earlier—to unleash the RAF for bombing civilian areas in Germany, something he had long believed in doing. As he said to the minister of aircraft production on July 8, 1940: “There is one thing that will bring [Hitler] back and bring him down, and that is an absolutely devastating, exterminating attack144 by very heavy bombers from this country upon the Nazi homeland. We must be able to overwhelm them by this means.”

  Nevertheless, it was crucial in the minds of the British public and many officials that this new British policy was introduced in the context of reciprocity. Said Churchill, “This is the way to pay them back; it’s legitimate for us to do so, and in fact it’s virtually obligatory for us to do so. If he’s starting this form of warfare, it’s necessary for us to do the same.”

  On the day the German attack on France and the Low Countries commenced, planes bombed the German university town of Freiburg. The Nazis denounced this as a violation of the Allies’ assurance they would not initiate bombing of undefended cities. The bombing was, in fact, a mistake by German Luftwaffe planes, which, after a navigational error, thought they had bombed a target in France. (It took forty years for the German commander in charge to acknowledge this mistake and false report in 1980.)

  The propensity for such costly errors was further demonstrated on the night of August 24, 1940, when German bombers drifted off course during a planned attack on oil refineries on the Thames and ended up bombing houses in London. Hitler, at this point still trying to avoid reprisal, had in fact issued the strictest orders that no bombs should fall on London, while reserving that as a possibility for later on. Nevertheless, this initial attack prompted the first British attacks on Berlin the next day, August 25, and then the day after that, followed by six attacks within the next ten days.

  After the fifth of those, Hitler was saying, “We will pay back a hundredfold145 if you continue this. If you do not stop this bombing, we will hit London.” Churchill kept up the attacks, and two weeks after that first attack, on September 7, the Blitz commenced—the first deliberate attacks on London. This was presented by Hitler as his response to British attacks on Berlin. The British attacks, in turn, were presented as a response to what was believed to be a deliberate German attack on London.

  In the beginning of the British strategic bombing, there was some debate between factions that believed in targeting civilian morale and a more dominant part of the Air Staff and even the Bomber Command that still adhered to what we could call the American doctrine. The latter was associated with General Billy Mitchell, who pushed the idea that industries were the things to hit, not people per se. The trouble was that as early as 1940, the British had discovered conclusively that Douhet’s notion that you could afford to ignore defenses, that the bomber would always get through, was wrong. They were losing so many planes to daylight raids that they had to switch to night bombing.

  The Germans initially had little capability for night interceptors; their fighters didn’t have the proper radar. As a result, British planes were fairly safe at night. The trouble was that in moving to night bombing, they rather quickly discovered not only that they could not identify or hit a factory at night, but also that they had great difficulty finding a small- or medium-size town. British nighttime navigation capability, even when there was a bright moon, proved to be much less reliable than they had imagined.

  Although these navigational methods would improve over time, there was an additional problem. Even if they found the right town, hitting something specific within that town, either finding it or managing to drop their bombs on it while taking evasive action in the face of antiaircraft flak, was impossible. Later photoreconnaissance showed that no more than one-third of their planes’ bombs were getting within five miles of their targets.

  Freeman Dyson was a physicist and later a nuclear bomb designer who in World War II was a young mathematician doing operations research on the British bombing campaign. He described one of the early results of a photoreconnaissance mission, which showed photographs of where the bomb damage actually was in relation to the target. The briefers had made a three-mile circle on the map around the targeted factory, in preparation for showing the results to higher command. He recalls somebody saying, “You know, there aren’t too many bombs within that circle146; maybe you’d better use a five-mile circle.”

  With high explosives, a 500- or 750-pound bomb, exploding even a hundred yards away, had essentially no effect on the target. So if they were hitting a mile or five miles away, people at the target wouldn’t even be aware that they were under attack. Analyses of the results were based on the bombing crews’ reports that they had annihilated this or that factory, or had just destroyed a particular wing of the factory. It wasn’t until separate missions were sent with Spitfires for photoreconnaissance—well into the war—when analysis revealed that they weren’t hitting anything they were aiming at, unless by accident.

  In the summer of 1941, with the United States not yet in the war but the Russians now under Nazi attack, the British really wanted very much to keep bombing Germany. In recognition of the impossibility of destroying individual factories at night, they moved to a different kind of target. Instead of worrying about whether they should hit oil refineries or ball bearings factories, the RAF shifted their focus to transportation targets.

  These targets had always been conceived as important from the earliest thinking about strategic bombing. But the reason they became primary targets at this particular time was that railheads, marshaling yards, and junction points for trains were in the middle of cities. If you took those as a target, you would not necessarily hit them but you would certainly hit something; the bombs would not land in a field, as most did when they were aimed at factories on
the edges of town. And there would be what they called a “bonus”: people would get killed—civilians, yes, but still enemies. Perhaps war workers, at least some of them.

  Among the decision makers and planners, some believed such enemy people should be the true targets anyway, but in 1941 that was still a minority attitude within the RAF.

  Sallagar presents an account in the air offensive history of the British based on official documents:

  If there was to be any strategic bombing at all,147 civilians would be killed—hospitals, churches and cultural monuments would be hit. The Air Staff, as represented by its Vice-Chief, Sir Richard Peirse, believed that what was inevitable was also desirable, but only insofar as it remained a byproduct of the primary intention to hit a military target in the sense of a power station, a marshaling yard or an oil plant.

  In short, it was all right or even good to kill civilians, but only if you didn’t “intend” to hit those people—you were aiming and intending only to hit the power plant. Meanwhile,

  Bomber Command, as represented by its Commander-in-Chief,148 Sir Charles Portal, already believed by September 1940 that this byproduct—human beings—should become the main or end product. He believed this had already been justified by previous German actions [the Blitz] and would be further justified as a strategy in the outcome.

  (Curiously, this belief was not tempered by the actual failure of the London Blitz to achieve any of its objectives with respect to British morale or production.)

  Sallagar cites a new British directive in late 1941 identifying objectives in large towns, with the primary aim of causing heavy material destruction. It instructed Bomber Command “to employ a high proportion of incendiaries149 and to focus their attacks to a large extent on the fires with a view to preventing the firefighting services from dealing with them, thereby giving the fires every opportunity to spread.” Sallagar comments, “If the Air Staff was still reluctant to come out openly in favor of attacking civilians, at least it was willing to adopt the German tactics that had proved so successful in killing civilians in British cities.”

  At the same time, Joseph Goebbels, the Nazi propaganda minister, was publicizing in enormous detail the horrific effect of what the Germans called terror raids. It was easy to discount his claims as enemy propaganda. But unimpeachable accounts were also coming from bishops in occupied territories or Germany itself that many civilians were dying under the bombs and incendiaries. On the basis of such accounts, which were in fact quite accurate, the American Jesuit John Ford and the British pacifist Vera Brittain150 strongly condemned what was happening. But their interpretation and critique of Allied bombing policies was not believed or accepted by most Americans or English citizens because it was invariably strongly denied by British and American authorities.

  To the end of the war, when questions about this policy were raised in Parliament and Congress, both the American and the British authorities on every occasion responded with some version of this formula: “Yes, some innocent people are being killed in warfare. That is the nature of war. It has always happened. Indeed, although it is unhappy and deplorable that these people are being killed, the fact is that the Germans did start this type of operation. They are fighting an aggressive war. They started it and they are getting back what they have given to us.”

  Of course, the civilians who were being killed were not precisely the ones who “gave it to us.” The fact that members of the German public were not exactly in democratic control of their country’s policy was glossed over. But they were seen as having been supportive of Hitler’s policy when he was winning, and that was largely true. So they deserved this regrettable but inevitable punishment. But “we are doing our very best, in view of our own basic values, to minimize civilian casualties, while we are hitting war factories, oil reserves, port facilities as accurately as we possibly can in the face of antiaircraft fire.”

  This was false. Nevertheless, through 1941 many people at the top in the U.K. were still fooling themselves as to what they were actually doing and why they were doing it. But there came a time when they stopped deceiving themselves, though they continued to lie to the public for the rest of the war. The era of modern warfare in an important sense—the essential precursor, I believe, to the era of nuclear danger we still inhabit—began on February 14, 1942.

  It was not that city-bombing began on that date, as we saw in Shanghai, Guernica, and elsewhere. But deliberate bombing of urban populations as the principal way of fighting a war by a major industrial power can be said to have started on February 14, 1942, with a specific British directive I first encountered in Sallagar’s manuscript, which I read in my office at RAND in 1959.

  The document was an Air Staff directive later confirmed by the Joint Chiefs of Staff and the civilian Defense Committee:

  TO THE BOMBER COMMAND:

  The primary object of your operations151 should now be focused on the morale of the enemy’s civil population, and in particular of the industrial workers. With this aim in view, a list of selected area targets … is attached.

  The primary targets listed were four important cities152 in the Ruhr-Rhineland area. It was the start of the practice of naming cities as targets: not factories, not specific blocks, but cities. Of course in those days high explosives couldn’t destroy a whole city. It took hundreds of planes on many return missions to do that. Nuclear weapons made it possible to destroy whole cities with a single plane, and when nuclear weapons plans began to be written after the war, they designated only entire cities as targets. But that practice really started with this directive, and its handwritten addendum by the chief of the Air Staff, who “wanted no misunderstanding on whether the air offensive was to be directed against cities or against specific objectives.” He penciled an explanatory note for the guidance of the new chief of Bomber Command, General Arthur Harris, who was to take command the following week:

  Ref. the new bombing directive.153 I suppose it is clear that the aiming points are to be the built-up areas, not, for instance, the dockyards or aircraft factories.… This must be made clear if it is not already understood.

  Sallagar notes that “there was little danger that Air Marshal Harris would misread the intent of the directive, for it accorded with his own preference.” Bomber Harris, as he came to be known, had believed for years—in particular, since he studied with admiration the German attack on Coventry—that the notion of destroying a specific industry was not only infeasible but also would not have the desired effect. He believed that his bombers could only hit large areas, that this had a bigger effect on productivity than destroying individual factories, and this was the correct way to fight the war—to destroy as large a part of as many German cities as possible.

  [The air-war historians] Webster and Franklin refer to February 14, 1942 when this directive was issued as “a pregnant date in air history.”154 It was indeed, for it ushered in an onslaught on Germany that made the Luftwaffe attacks on London seem puny by comparison.

  For every ton of bombs dropped on England in the nine months of the Blitz, England and the United States, mainly England, eventually dropped a hundred tons of bombs on German cities. More than half a million Germans—civilians—were killed.

  For the first time, a bombing directive had singled out the parts of cities155 where civilians were housed most densely as the primary objective of individual attacks and of the overall campaign. Except for inescapable diversions [such as supporting the Normandy invasion,] it was to remain the primary objective for Bomber Command for the remaining years of the war.

  The largest part of the tonnage dropped by the British through the rest of the war was directed at the centers and most built-up parts of cities—not at factories or military installations, which tended to be on the outskirts—although high officials continued, falsely, to deny this to Parliament and the public every year of the war.

  When it came to killing civilians, practice preceded intention; but a change in intention did make quite a dif
ference. It was possible to kill more people from the air than the Germans had succeeded in doing in the London Blitz or the British had attempted to do by the end of 1941. The February 14, 1942 decision was the British authorization and directive to do just that.

  CHAPTER 15

  Burning Cities

  Early on in the course of trying to attack cities at night, the RAF discovered that high explosives did not get the desired effects, even when they were targeting housing. To begin, they chose the built-up portions of workers’ housing on the grounds that these houses were closer together so the fire would spread faster that way and the bonus damage would be greatest—a bomb missing one house would hit another. It wouldn’t fall in the yards that separated houses in middle-class or upper-class suburbs.

  They began to discover that fire, not high explosives, was a better way to destroy a city. In fact, delayed-action high-explosive bombs came to serve the purpose of discouraging firefighters from going after the incendiary bombs when they first hit the ground. By this time, the RAF was using magnesium-thermite bombs that couldn’t be put out with water. They had to be smothered with sand. Water would just intensify the flames or cause them to explode. But they could be extinguished if firefighters responded quickly with sand.

  In 1943 the RAF successfully tested a theory that had been conceived some time before: that the best way to destroy large parts of cities was to harness the forces of nature by appropriately designed technology and tactics. Specifically, it was hoped that a “firestorm” could be created, a kind of fire that would change the local winds—in effect, altering the area’s weather. If enough planes were sent in en masse to do patterned area bombing with incendiaries, a lot of little fires would start simultaneously throughout a large area. This would be helped by first dropping high-explosive bombs, which would break up the structures and make for better kindling, and also block fire trucks from the streets. The fire departments would be unable to deal with the many small fires, which would spread and join together until they became a mass fire. A large part of the city would burn uncontrollably.