The Ethics of Operation Cobra and the Normandy Breakout

James Jay Carafano
Lieutenant Colonel, USA
Executive Editor
Joint Force Quarterly
300 5th Avenue, Building 62
Fort Lesley J. McNair
Washington, DC 20319-5066

Phone (202) 685-4220
Fax (202) 685-4219
Email: <>


Like many, Michael Walzer’s Just and Unjust Wars was part of the foundation of my education in military ethics. In this paper I want to revisit Walzer’s use of history. Rather than derive theory from history, as his subtitle suggests Walzer uses historical examples to illustrate theory.He has little confidence in the study of facts. Rather history is subjective and vague – primarily a tool for illustrating the complexities of choice rather than a guide for action.

To discuss command responsibility Walzer draws on General Bradley’s decision to carpet bomb the area west St. Lo during the allied breakout from Normandy in July 1944.[1] During the bombing Allied planes killed and wounded an unknown number of French civilians, as well as hundreds of American soldiers. Walzer suggests unless Bradley had taken “due care,” made the minimum reasonable effort to minimize civilian casualties or consider reasonable alternatives he decision to bomb civilians in morally indefensible.Commanders had two responsibilities – they make positive steps to limit unintended civilian deaths and ensure subordinates are held to standards. 

This discussion is unsatisfying.First, Walzer failsto consider a third responsibility – the commander’s requirement to protect his own forces. Second, his analysis of Bradley’s decision is ambiguous. He admits he lacks the technical expertise and research to fairly judge the general’s choice and simply concludes “there is no sure rule against which to measure the conduct of General Bradley.” [2]

Walzer misses an opportunity to make an important point -- in modern warfare technical knowledge and expertise are central to the issue of determining command responsibility.As Charles Williams Maynes pointed out in a recent article, since World War II we have not become more moral, but we have become more accurate.The spot that required 9,000 bombs to hit accurately in World War II, required 200 in Vietnam and one in Kosovo.[3]Technical issues, accuracy among them, are an essential part of the calculus of determining morale responsibility – for a discussion of ethical conduct to contribute to understanding, the considerations of theory and practice must go hand in hand – practice is not simply a handmaiden for illuminating theory.Bradley’s deadly decision is a case in point.

Limits of Airpower

The most important constraints in safely employing airpower were well known to the army’s senior ground commanders. Even if the air forces attacked at the right time and place, there was still no guarantee they'd hit the enemy.Accuracy was problematic.For medium bombers to have a 95 percent chance of hitting a 6,000 square foot target, they had to drop 600 bombs (medium bombers were considered 2.5 times more accurate than Cobra's main punch, the heavy bombers).[4]On average, regardless of the type of bomb employed, only 90 percent fell within a thousand yards of the target.At best, even if the lead bombardier attacked with prefect accuracy, the odds were that at least 10 percent of the ordnance would fall over a thousand-yards from the target.As for the bombs that hit the target, they might land anywhere within that thousand-yard radius.This was the crucial limitation of aerial bombardment.The explosive effects of the bombs didn't extend much further than the crater. A direct hit devastated a target with explosive affect, shrapnel and concussion from the detonating bomb.On the other hand, all there was to show for a miss was a hole in the ground and a lot of dirt in the air.

Accuracy was a double-edged problem.Commanders also had to worry about protecting friendly forces.A “no bomb line” was used to mark the forward line of friendly troops and protect them from aerial fires.Ground commanders had limited flexibility in moving or marking the line.The line had to be placed on terrain recognizable from the air.In addition, to shift the line air units had to be alerted over the air support radio net one hour prior to the effective time to insure that the order reached all the air units, and this alert only took care of the fighter/bombers.Medium and heavy bombers couldn't communicate with ground forces.Messages had to be relayed to England and then to the planes.These requirements dictated that no bomb lines should be coordinated well ahead of time and placed on the most prominent terrain available.

The ground commanders knew that marking the bombing limits was a problem.On July 17, General Eddy, the 9th Infantry Division Commander, and his air support officer went to the VII Corps headquarters to give General Collins a “first hand appreciation” of the problem.[5]They had good reason to be concerned.The most common method for marking the no bomb line was firing colored artillery smoke.This technique had severe limitations.The 4th Infantry Division, for example, in a report dated more that three weeks before Cobra observed, "smokes [sic] and dusts present on the field of battle may lead to confusion on the part of the pilot."[6]Smoke, the division concluded, should only be used if there was no alternative for marking a target.The air forces also knew that marking targets with smoke was a problem.Initially they did not want to use smoke to mark the northern boundary, but after conducting a reconnaissance they changed their mind and requested a series of smoke targets along the front at one miles intervals for ten minutes, starting five minutes before the initial attack.Even with this measure the air staffs remained deeply concerned about the danger of short bombings.[7]

In fact, there was no safe, effective means to mark the no bomb line -- and this wasn't news to anyone.The Infantry Board at Fort Benning, Georgia had experimented with the problem only a year before.They tried marking positions with smoke, phosphorescent paint, mirrors, colored lights, flares and even flags.No technique worked very well.The best pilots could do was identify targets at 2,000 feet (13,000 feet less than the planned attack altitude for the bomber attacks in Cobra).In addition, only direct radio communications proved effective in coordinating air and ground operations.[8]

During combat, even radio didn't completely solve the problem of air-ground coordination.In Normandy, where ground units had radio contact with tactical aircraft through their attached air support parties, short bombings or strafings were endemic.In the first week of the campaign there were nine official reports of accidental attacks.The actual number of incidents was probably much higher.Just about every combat soldier had a story to tell.As one veteran remembered:

I always knew our Air Force was pretty good, 

because twice we were strafed by our own planes. 

We tried to put identification panels out where our 

pilots could see them, but in the apple orchard country of Normandy that was kind of hard to do.[9]

During the assault on Cherbourg (the attack General Bradley said helped inspire the idea for the Cobra bombing), planes bombed one friendly position 2,000 yards behind the no bomb line (almost twice the safety distance planned for Cobra) three times.

Communications wasn't a panacea, because pilots found it difficult to fix their exact positions even under ideal conditions.Fighter pilots went into battle with their maps strapped to their knee -- reading coordinates, flying their aircraft, watching out for enemy planes, antiaircraft fire, friendly troops, civilian non-combatants and enemy targets all at the same time.Fighter pilots when asked how they decided when they were behind the enemy lines claimed:

It was quite simple. All they [pilots] did was watch 

the roads and when they saw a Frenchman flogging a horse and drawing a cart down the road they knew they were over German held territory. On our side of the line, trucks and transportation lined the road, generally bumper to bumper.[10]

As imprecise and haphazard as the task of locating and identifying the enemy lines was for the fighters, the difficulty increased exponentially for medium and heavy bombers flying at much higher altitudes. 

Despite such limitations General Bradley relied on the no bomb line as his primary tool for protecting the ground forces.The First U.S. Army no bomb line, marked A-B on the operations overlay, was easily identifiable because it followed parallel to the St. Lô to Périers road running straight east to west across the VII Corps’s front. 

General Bradley wanted the bombers to attack parallel to the road and use St. Lô to Périers road to mark the boundary between friend and foe.In addition, by taking a parallel approach the bombers would not fly over the VII Corps’s forces, reducing the chance that bombs released at the wrong point would land on the troops.To further identify the no bomb line, American units would mark their forward positions with florescent panels, while the artillery would fire red smoke along a 250-yard strip by along the road.Drivers were also told to paint white stars on the top of their vehicles to help low flying planes identify friendly units.

Fixing the distance between the forward line of troops and the no bomb line was another important factor in safeguarding the soldiers.If the ground force withdrew too far to avoid the bombing the enemy might recover from the affects of the attack before the friendly troops could assault.During Operation Charnwood the Allied forces withdrew 6,000 yards.By the time the attackers got to the front the German defenses had already recovered.

General Bradley elected to make a compromise between distance and caution.He wanted troops to withdraw only 800 yards, one hour before the attack.The air commanders wanted more than six times that distance, almost three miles.They then proposed 3,000 yards and finally compromised at 1,200 yards.For additional safety, only the smaller fighter/bombers would attack targets in the 250-yard strip directly adjacent to the road (making the bomber safety distance 1,500 yards).

Friendly troops were not the only concern on the battlefield.When days before Cobra, General Bradley briefed the press at the First U.S. Army press camp at Vouilly, reporters asked if the civilians had been warned.In response General Bradley remembered:

I shook my head as if to escape the necessity of saying no. If we were to tip our hands to the French, we would also show it to the Germans . . . . The success of Cobra hung upon surprise; it was essential we have surprise even if it meant the slaughter of innocents as well.[11]

Employing firepower incurred heavy moral responsibility.Though many civilians had evacuated the area, the target included a number of sizable villages.

We know for a fact that the Americans interviewed French civilians who had knowledge of what was going on beyond the German lines.The Americans knew many villages contained civilians -- that civilian casualties had been incurred from U.S. air attacks – the Americans also knew that the civilians did not have complete freedom to flee the combat area.First, moving on the roads would have put them in physical risk.Second, by moving they would become refugees – which incurred another whole set of risks. Its clear that the civilians did not have the freedom to leave the battle zone – and any assumption that civilians in a combat zone have freedom of choice would have been ethically flawed in itself.Bradley knew innocents were at grave risk in the attack. Noncombatant casualties were likely. 

Walzer offers-up several alternative courses of action that Bradley might have used to avoid the carpet bombing.In reality, a careful study shows none were really practical alternatives – and Bradley tactical scheme was not only appropriate, but the best option available.[12]

In addition, General Bradley’s decision to bomb without warning adhered to the laws of war with regards to protecting the innocents. Civilians whoremained in a known combat zone were considered adequately warned.While it was still unacceptable to intentionally target civilians, casualties might be incurred in pursuit of legitimate military objectives with a proportionally suitable amount of force.The law of war also recognized the necessity of secrecy and the inherent right of commanders to weigh the safety of their troops against the danger of revealing their intentions for an operation.Commanders could take prudent precautions, even if it meant noncombatants were put at risk.This didn't mean that civilian lives and property were forfeit.Commanders had a responsibility to take all practical measures to safeguard the innocents.In this case, General Bradley felt security concerns precluded any warnings to civilians.

All together, General Bradley thought this plan provided an adequate balance between the need for firepower, his moral obligations and the safety of his troops.On July 19, General Bradley and Major General Elwood R. “Pete” Quesada, Commander of the IX Tactical Air Command boarded a small two engine plane flew to Air Chief Marshal Leigh-Mallory’s headquarters in a ramshackle old mansion at Stanmore outside London to review the plan with the air commanders and staff.In addition to Leigh-Mallory, Air Chief Marshal Tedder, General Spaatz and General Brereton, a host of high-ranking air staff officers and commanders joined in the meeting which started about 3:30 p.m.Though both Bradley and Leigh-Mallory departed before the meeting ended, General Bradley left feeling upbeat, believing that potential problems had been addressed including the two most contentious issues, the distance the troops would withdraw from the no bomb line and the direction of flight for the attacking planes.General Bradley's aide Major Hansen recorded, "Conf[erence] went well, came out by five[p.m.] with Brad[ley] speaking amiably, laughing."[13]General Spaatz even ordered up a C-47 for General Bradley’s return flight to Normandy insisting that it wasn’t proper for the First U.S. Army Commander to be running around in a tiny two motor plane.As General Bradley returned to his headquarters in style, he assumed everything was set for the big attack.

Finally on July 24, despite continued poor weather, Air Chief Marshal Leigh-Mallory and General Bradley gave the go ahead to start Cobra, gambling that the weather would clear-up over the target area before the planes arrived.[14]By midmorning the weather had gotten worse and before noon the air chief marshal and General Bradley thought it best to recall the planes.Not all of the aircraft had received the message and some of the uninformed pilots completed their bomb runs with disastrous results.One pilot bombed an American ammunition dump, another an Allied airfield, other planes dropped their payloads north of the road on the 30th Infantry Division.[15]

Later that afternoon, Air Chief Marshal Leigh-Mallory met with General Bradley at the First U.S. Army field headquarters. General Bradley was furious.It wasn't just that the attack was postponed.It wasn't just that bombs fell on American soldiers. It was because General Bradley claimed he had learned for the first time that the heavy bombers had flown perpendicular to the target area, right over the American troops.This was not what he had agreed to.The air chief marshal claimed he did not know about the change in plans.His directive for the bombing, in fact, had not even designated the direction of attack.[16]He thought it had all been agreed to at the Stanmore meeting on July 19. 

Apparently, without clearly relaying their final decision to General Bradley, the air commanders decided the bombers needed to approach perpendicular to the target, rather than fly a course parallel to the American positions.The perpendicular course they concluded would minimize the time aircraft would be over enemy lines, exposed to antiaircraft fire.More important, the target box was narrower on the parallel axis.On a parallel approach fewer planes could fit over the target area.It was simple geometry.It would take much longer to complete the bombing on a parallel course.A parallel bombing run would be more like a shifting steady rain, rather than the single powerful strike General Bradley needed.To maximize shock affect and deliver the most bombs in the shortest time, the bombers needed to attack perpendicular to the target.In addition, a perpendicular course would be safer for the aircraft, they would spend less time over the target exposed to anti-aircraft fire.

The reason that the air commanders had not pressed their case on General Bradley more clearly at the conference at Stanmore on July 19, or in the days following the meeting is far from apparent – and remains one of the unresolved controversies of the campaign.There is evidence to suggest that some senior air staff officers raised the issue at the Stanmore meeting.Air force notes made after the conference even stated that they thought General Bradley had understood the risks of short bombings and that he said he was “prepared to accept such casualties no matter which way the planes approached.”On the other hand, at the same meeting officers recalled that Air Chief Marshal Leigh-Mallory quickly rejected the proposal of bombing on a parallel course.In short, while it is clear that the air commanders decided to conduct the perpendicular bombing at the Stanmore meeting, General Bradley and Air Chief Marshal Leigh-Mallory were probably not present when the final decision was made.

Confusion persisted, after the meeting Leigh Mallory’s deputy Major General Hoyt Vandenberg claims in his dairy to have contacted the air chief marshal to confirm that General Bradley would accept a parallel bombing.After the aborted attack on July 24 a team of air staff officers again met with General Bradley to impress on him the imperative of a parallel bombing, but there appears to be no record of senior air commanders ever taking up the issue personally with the First U.S. Army commander.In fact, there appears to be no joint Army/Air Force records of the precise coordination agreed to during the period of 19-24 July.Either some one lied or, if both sides are to be believed, each left the July 19 meeting believing exactly opposite thing.Indeed, misperception on both parts may be the answer.The bulk of the evidence suggests that the confluence of mismatched personalities, the convoluted relationship between air and ground commanders and the lack of a sound doctrine created a situation in which, at the end of the day, commanders simply failed to understand what each other were doing.[17]

Now that the damage was already done, General Bradley and Air Chief Marshal Leigh-Mallory had to decide what to do next. They had a simple choice.They could repeat the bomb run, bomb on a parallel course that would not give General Bradley the quick, powerful strike he wanted to lead off the attack or call off the aerial bombardment.It was a difficult choice.General Bradley had already seen the tremendous disruption that could be caused by a short bombing.If he had any hopes that the other safety measures he had put in place would be sufficient, he now had incontrovertible evidence that they were inadequate.If the planes attacked the same way again, Americans would most likely die from American bombs. 

Bradley decided to repeat the bombing the next day and again planes hit friendly troops.The total loses in the VII Corps were approximately 108 dead and 472 injured.[18]Despite the fact that less than 3 percent of the bombs landed on the American lines, the attacks created a near crisis in the VII Corps.In addition, to the outrage and frustration at the tragic killing and wounding of soldiers, the disaster threatened the whole offensive.The first effect was the obvious physical damage and the disruption of attack plans.Just as serious was the psychological injuries.Friendly fire undercut unit morale, creating a heightened sense of fear and hopelessness in soldiers and a mistrust and apprehension over again employing friendly fire support systems. In some cases the results were severe.The 30th Infantry Division, for example, reported an additional 164 cases of combat exhaustion.All total, the short bombings were a disastrous start for the beginning of the breakthrough.[19]

Evaluating Bradley’s Decision

Half a century later, the debate over the Cobra disaster has lost none of its vitality.Today, typing on a keyboard far from the terror of the hedgerows, it is a simple task to point out the shortfalls of the bombing plan.On that July day, the issue seemed very different.An officer at the front concluded:

I believe every man in the company will agree that if 

we have such an attack again they would want the bombing just where it was, right to our lines.We would rather take the ones that fall on us to get the effect on the Germans in front of us.[20]

The issue, however, was not whether or not the bombing should have been conducted, but if the commanders had done everything possible to mitigate the risks.

General Eisenhower, for one, was concerned about the implications of assigning blame for the disaster.He wrote to General Marshal:

Complete investigations are underway....In themeantime I am Anxious that treatment in the newspapers, if the matter become one of public discussion, be moderate and sensible So that we do not get a ground versus air war started that is completely senseless and harmful.[21]

At the time no one was anxious to fix responsibility to any senior leader.In fact, Lieutenant Colonel Alvin B. Welch from the First U.S. Army’s Inspector General’s office conducted an official investigation of the incident, finding “no delinquencies on the part of the ground units.”His report specifically did not address at all the activities and decisions by senior air and ground commanders.[22]

Official reports offered no criticisms of senior commanders. Historian Martin Blumenson also absolved the First U.S. Army commander of any responsibility for the short bombings.More recent studies by historians including John Sullivan and Richard G. Davis are more critical of both Bradley and Leigh-Mallory, stressing their lack of experience in planning air-ground operations and inadequate coordination with the air commanders.[23]

These historians, however, do not go far enough.While Bradley cannot be faulted for the civilian casualties, a thorough review of the evidence suggests that not only must General Bradley bear major responsibility for the friendly casualties, but that his failure was more than simply inexperience or an inadvertent lack of proper oversight.

Why did things go so terribly wrong?The issue has too frequently focused on the simple issue of parallel versus perpendicular attack.General Hobbs, the 30th Infantry Division Commander attributed blame to the route taken by the heavy planes.He claimed, "bombers that flew parallel to our front lines dropped their bombs on the target.Many of the bombers who flew across our line dropped their bombs short and it was these that caused our casualties."[24]But, the general’s assessment and the historians who generally followed his reasoning are incorrect.

A parallel attack was no “magic bullet” for solving the challenge of bombing in close proximity to ground troops.In fact, with the battlefield totally obscured by smoke, dust and fire there was no guarantee that a parallel approach would have been any more successful in preventing short bombings.In addition, even on a parallel course there was a risk of hitting American troops.If the bombers overshot or undershot the target area, bombs might have landed on U.S. forces holding positions either northeast or northwest of the target area.All the previous Allied experience in air-ground coordination suggested that a parallel approach was not a sufficient precaution to preclude a serious threat from short bombings.

While the seeds of disaster were sown with a lack of clear understanding between the ground and air commanders over the direction of the aerial attack, there is much more to the story. Some of the most tragic short bombings were made by the supposedly more accurate fighter/bombers and some of these planes were flying parallel, not perpendicular to the target and at an altitude of only 2,000 feet!In short, even planes flying a parallel course at low altitudes attacked American troops.

In fact, neither the parallel bombing nor any of the other precautions General Bradley ordered had proved sufficient.The smoke markings actually worsened situation.When the corps artillery fired red smoke to mark the no bomb line, the smoke clouds drifted north obscuring the line.After the first bombs hit their targets, dust from the explosions mixed with the smoke, further exacerbating the problem.The panel markers placed in front of the friendly positions and white stars painted on the vehicles were equally useless.In fact, all the marking and coordination techniques employed by the Americans were inadequate, even for marking targets for the fighter/bombers.This was not a new discovery.These techniques had consistently failed to prove themselves earlier in the campaign. 

Another problem during Cobra was that commanders hadn't ensured that the ground troops were adequately prepared.Even though troops were supposed to withdraw 1,200 yards from the no bomb line, some units were positioned as close as 800 yards or less.Others pulled back shortly before the air strike, but were not told to dig-in.The 2d Battalion, 120th Infantry Regiment, for example, which was hit hard on July 24, was deployed in battle formation preparing to attack, not dug-in and protected. Lieutenant Chester H. Jordan in the 3rd Battalion, 47th Infantry Regiment recalled, “I reasoned the Germans would be in no shape to throw anything at us so why dig a hole.”When the attack started his platoon “cheered the bomber on like kids at a football game,” until the first short bombs began to land.[25]The failure to take protective cover was invitation to disaster.Neither the First U.S. Army or the VII Corps ordered additional precautions on July 25 or warned units that heavy bombers were flying a perpendicular route and that there was an increased likelihood of short bombings.In fact, the VII Corps issued a message on July 25 at 1:55 a.m. reassuring commanders that there would no bombings north of the road.Some units, on their own discretion, ordered troops to dig-in.Foxholes protected soldiers from the fragmentation effect of exploding bombs, but even then they were not completely safe.As the death of General McNair illustrated, entrenchments were not sufficient protection from direct blast effects. 

General Bradley appreciated the necessity of closely integrating aerial fires with the other elements of the plan, but on this occasion he failed.All the American experience with employing close air support in World War II up to that point had indicated that the safety measures adopted would be inadequate.Nevertheless, official inquiries into the bombing absolved senior air and ground commanders of any personal culpability.The reports cataloged twelve separate incidents (not including short bombing and strafings by the fighter/bombers), determining that the causes of the short bombings were human error or confusion and disorientation of pilots and bombardiers.[26]

The preconditions for disaster were the cumulative result of three critical failures; inadequate coordination with the supporting air forces, inability to mark the forward positions of ground forces, and the lack of warning and protection of the troops.But, what the reports fail to fully emphasize is that none of these shortfalls were beyond General Bradley's control.In particular, he made the most important and tragic decisions in regards to the bombing; how far the troops would withdraw from the no bomb line and what precautions had to be taken.Considering the Allies’ history of problems in short bombings and strafings, particularly in Normandy, Bradley’s failure not to pay more attention to the issue and work with the air commanders until the issue was fully resolved was inexcusable. The overall fire support plan was his initiative and it was seriously flawed. 

In his memoirs A Soldier’s Story, Bradley goes to great lengths to acquit himself.For insistence, in the text his preference for having the bombers fly a parallel course is italicized, suggesting that if the air commanders had just followed his wishes there would have been no casualties – a dubious proposition.General Bradley also included curious statements in his account of the pre-Cobra planning, such as a remark that General Eddy, the cautious 9th Infantry Division commander, had:

balked initially at this order to withdraw.After having fought hard for that mile he disliked giving it up with the prospect of having to fight once more to regain it.But I [Bradley] was unwilling to chance a bombing any closer to our lines.[27]

The recounting of this conversation fails to note that the actual safety separation General Bradley had decided on was 1,200 yards, (1,500 for the heavy bombers) a distance well short of a mile.Bradley also does not reveal that he himself had initially argued for only an 800 yards separation.In addition, he does not point out that if the American troops had withdrawn the 3,000 yards suggested by the air commanders there would have been minimal frontline casualties from the short bombings.

Also if you look at the time it took the forward troops to recover from the attack and the significant delay the bombing caused in delaying German counterattacks it is arguable that adding an extra 1,000 yards to the safety zone would not have significantly affected the outcome of the day’s battle.In short, having troops so close to the no bomb line was not a military necessity.

Finally, General Bradley’s description of General Eddy’s concerns hardly squares with Eddy’s complaints on July 17 about the danger of short bombings.[28]Nor does General Bradley mention that Brigadier General F. H. Smith, the Director of Operations for the Allied Expeditionary Air Force at the July 19 meeting at Stanmore had briefed Bradley that there might be “gross bombing errors causing troop casualties.”[29]In short, General Bradley’s own accounts of the Cobra bombing did much to obfuscate his personal responsibility.The evidence suggests he knew the front lines troops would incur casualties.He was willing to take that risk. He was not, however, prepared to acknowledge this truth publicly. His explanations are a disingenuous moment in what is regarded as an otherwise honorable and distinguished career.

Why General Bradley made such poor choices with regard to this particular aspect of the campaign cannot be confidently fixed from the available evidence.However, it is clear that whether due to a serious tactical misjudgment, intentionally assuming the terrible risk of killing and maiming his own troops, or unrealistic wishful thinking on his part, General Bradley failed to adequately address the issue of protection for his ground forces.[30]


History has more to offer than Walzer suggests. In fact, we can judge the appropriateness of Bradley’s actions – and we can extract useful ethical guidelines from a detailed study of the facts surrounding his decision.Walzer’s mistrust of history suggest that knowing the difference between right and wrong is sufficient. Knowing standards, however, is wholly insufficient for making correct moral decisions in war.Commanders must have the appropriate technical expertise to implement decisions.History offers not only an illustration of ethical dilemmas, but a practical laboratory for understanding the intellectual – as well as the moral tools needed to make the right decisions in battle.
Note: Complete bibliographic citations for the endnotes can be found at Additional resources for the study of the breakout can found at

1. Walzer, Just and Unjust Wars, pp. 316-319.
[2]Ibid, p. 321.
[3] Maynes, “Principled’Hegemony.” 
[4]. Gooderson, “Heavy and Medium Bombers,” p. 372.
[5]. Activities of General Eddy, July 17, 1944.
[6]. Headquarters, 4th Infantry Division, U.S. Army, Letter to G-3 Air, VII Corps “Air in Support of Infantry,” July 1, 1944.
[7]. The Battle for Normandy, File 8-3.1 AF, p. 170, CMH.
[8]. Office of the Infantry Board, Report, “Artillery-Infantry Test,” p. 27, USAFAS.
[9]. Dorst, My Times, p. 56, MHI. There were probably many times that number of unreported attacks by friendly aircraft.The Military History Institute at Carlisle Barracks for a number of years has been collecting surveys of World War II veterans.In a random search of 40 surveys I found that 35 of the respondents claimed to have been attacked by friendly aircraft. See World War II Survey, MHI. 
[10]. Charles F. Howard, AGF Observer Report, July 27,1944, pp. 6-7, CARL.
[11]. Bradley, Soldier's Story, pp. 157-8. Colonel Dickson remembered the briefing in his memoirs.He recalled General Bradley was “calling his shots” with the press like a pool player. See Benjamin A. Dickson, Algiers to the Elbe, Dickson Papers, p. 139, MHI.In another encounter with the press General Patton’s staff inadvertently leaked news of the big attack before General Bradley had authorized discussion with the press.See Blumenson, The Patton Papers, pp. 482-4.
[12].Walzer, p. 317.For an analysis of Bradley’s tactical decisions see James Jay Carafano, After D-Day, Operation Cobra and the Normandy Breakout.
[13]. Hansen, War Diary, July 19, 1944, MHI.See also Bradley, Oral History Interview, Tape 7, revised transcript, pp. 10-11, MHI. Davis, Carl A. Spaatz, p. 720 n. 70, questions whether General Spaatz was in attendance. General Bradley’s recollection claims he was (see Bradley, Soldier’s Story, p. 340) as does General Quesada (see Elwood R. Quesada, Interviews of LT. Gen. Elwood R. Quesada USAF Oral History Interview, 12-13 May, 1975, AFHRA). General Spaatz also visited General Bradley’s headquarters on July 20, 1944, see Sylvan, War Diary, July 20, 1944, MHI.There is no record he discussed bombing tactics for Cobra.
14. For General Spaatz’s criticism of the decision see Memorandum foe General Spaatz’s Diary, Subject: Operation “Cobra,” 24 July 1944, Box 168, Carl Spaatz Papers, LC.
15. It was a B-24 Liberator that bombed a Ninth Air Force airfield at Chippelle when the bombardier accidentally hit the release toggle switch. The bombs destroyed two Thunderbolts parked on the ground, killed four and wounded fourteen.
16. Headquarters, Allied Expeditionary Force, Subject: Operation “Cobra,”20th July, 1944, microfilm, Roll A5129, AFHSO.
[17]. The account of General Bradley’s coordination with the air staff was reconstructed from Bradley, Oral History Interview, Tape 7, revised transcript, pp. 10-11; Quesada, USAF Oral History Interview, AFHRA; Hoyt S. Vandenberg, Papers of Hoyt S. Vandenberg, Dairy, Box 1, entries July 19-24, 1944; Memo for General Spaatz’s Diary, subject “Operation Cobra,” dated July 24, 1944, Carl Spaatz Papers, LC; Blumenson, Breakout and Pursuit, pp. 220-21, 231-33; Davis, Carl A. Spaatz, pp. 465-71; Hughes, Overlord, pp. 199-201.In addition, Hansen’s War Diary contains a memo titled subj; ”Combined Air and Ground Operations west of St. Lô on Tuesday 25 July 1944.” The memo lays out General Bradley’s interpretation of the coordination.Davis, Spaatz, p. 720 n.60 suggests it was composed sometime after the event (though the memo is dated July 25, 1944). He has no evidence to support this assertion. In fact, it would seem to be exactly the kind of document General Bradley would have wanted to put together in light of the investigation that was sure to follow. Official Eighth U.S. Air Force memos on the coordination also appear to have been written after the event.See also Blumenson, Battle of the Generals, pp. 136-7.
18.Casualty figures for the attack vary.The figures given here are from the official Inspector General investigation. See Headquarters, First U.S. Army, subject: Investigation of Bombing of Ground Troops by Friendly Planes on 24 and 25 July 1944, 16 August 1944, Adjutant General Section, General Correspondence 1940-47, box 119, RG 338, NARA, p. 1. There are no authoritative casualty figures for civilians. Françoise Dutour, The Liberation of Calvados 6th June 1944 – 31st December 1944, provides a summary of archival records on civilian casualties, but no specific figures for Operation Cobra.
[19]. For a useful introduction into the problems of fratricide on the modern battlefield see Charles R. Shrader, ”Friendly Fire: The Inevitable Price,” Parameters XXII (Autumn 1992): 29-44. 
[20].4th Infantry Division Combat Journal, “Interview with Lt Gude.”
21. Chandler, Dwight David Eisenhower Papers, p. 2032.
22. Bombing of Ground Troops by Friendly Planes on 24 and 25 July 1944, 16 August 1944, Adjutant General Section, General Correspondence 1940-47, box 119, RG 338, NARA. 
23. The official First U.S. Army after action air support report called the Cobra carpet bombing “a complete success” and did not even mention the short bombings. See Headquarters, First U.S. Air Support report, 6 August 1944, Section V. Operation Cobra, microfilm, Roll C5127.Air Chief Marshal Leigh Mallory also rated the mission a success, arguing it proved the value of strategic bomber support for ground operations.See Despatch by Air Chief Marshal Sir Trafford Leigh-Mallory, box 104, Carl Spaatz Papers, LC, pp. 57-8.The Commanding General of the Army Air Forces report did not mention the short bombing in its narrative of Cobra.See Second Report of the Commanding General of the Army Air Forces to the Secretary of War, pp. 11, 14. For a photographic analysis of the bombing coverage and the pattern of the short bombings see U.S. Strategic Bombing Survey, Europe War, G-2 Target File, 2478, Box 140, RG 243, NARA.Experiences late in the war demonstrated that strategic bombing could support tactical operations without fratricide.The 8th U.S. Air Force eventually developed an extensive air-ground marker and communications system that was used to support Operation Queen in November 1944. See Report of the General Board, Air Power in the European Theater of Operations, Study No. 56. p. 17, CARL; Gooderson, ”Heavy and Medium Bombers,” pp. 390-1. Blumenson absolves General Bradley of responsibility for the short bombings. His assessment can be found in Blumenson, Battle of the Generals, p. 131.For another assessment of the responsibility for the failure to adequately coordinate air support for operation Cobra see, John J. Sullivan, “The Botched Air Support for Operation Cobra,” Parameters XVIII (March 1988): 97-110. 
24.AGF Immediate Report No. 16., CARL.
25. Chester H. Jordan, Bull Sessions World War II Company K, 47th Infantry Regiment, 9th Infantry Division from Normandy to Remagen, p. 42. Instructions given to soldiers across the front were not uniform.Some were told to dig-in and did so.See, for example, Robert C. Baldridge, Victory Road, p. 110.
26. See Harold W. Ohlke, Report of Investigation of Bombing, July 24-25, Box 168, Spaatz Papers, LC; Headquarters Eighth Air Force, Special Report of Operations 24 and 25 July, 1944, 11/9/44, Box 168, Carl Spaatz Papers, LC, p. 7.In fact, post-strike air force analysis rated the overall bombing as technically accurate with less bombing errors than anticipated.In total 42 planes (35 bombers) dropped bombs on friendly positions. See Eighth Air Force, History of Operations Analysis Section, October 1942-June 1945, AFHRA.See also Walter E. Todd, Report of Investigation of Bombing July 24-25, Box 71, RG 243, NARA.
27. Bradley, Soldier’s Story, p. 341.The italicized passage on parallel bombing is on p. 330. 
28. Activities of General Eddy, July 17, 1944. 
29.Headquarters Eighth Air Force, Special Report of Operations 24 and 25 July, 1944, 11/9/44, Box 168, Carl Spaatz Papers, LC, p. 1.
30.There is one other potential source of evidence that may provide insight on the controversy.John L. Walsh, General Collins’ aide de camp wrote his wife shortly before and after the bombing.The letters have not been released to the public.They may offer insight further insight into General Collins’ knowledge concerning Bradley’s actions. See John L. Walsh, letters for July 24, 26 and 27, 1944, DDE.