Army Ground Forces, Study No. 1



Responsibilities of GHQ for Air and Air-Ground Training

In its relation to the air forces GHQ passed through two periods divided by June 1941, when the Army Air Forces were recognized and established as an autonomous force within the Army.

In the first period GHQ was responsible for the training of air as well as ground elements of the Army. The GHQ Air Force, somewhat confusingly so named since it long antedated the activation of GHQ, comprised all combat aviation units in the continental United States. On 19 November 1940 this force was formally put under the command of the Commanding General of the Field Forces and under the "direct control" of GHQ.1 General McNair, as Chief of Staff, GHQ, therefore became responsible for the supervision of its training.

On 14 August, ten days after General McNair assumed his new duties, a comprehensive training directive was given to the Commanding General of the GHQ Air Force by the Chief of the Air Corps, Maj. Gen. H. H. Arnold. General McNair took this to mean that the actual supervision of GHQ over air training would be limited. In a memorandum to his Air Officer, Colonel Lynd, who had not yet reported at GHQ, he observed that General Arnold’s directive appeared “to constitute a radical change of policy. Apparently the action was a personal one by the Secretary of War to the Chief of the Air Corps.” At General McNair’s request, Colonel Lynd, shortly after arriving, prepared a comprehensive report on the organization, training, and combat readiness of the air forces. Notwithstanding the War Department directive of 19 November explicitly placing the Commanding General of the GHQ Air Force under GHQ in the chain of command, General McNair felt that he was called upon to do little beyond keeping himself informed regarding its training program.2

The second period opened with the reorganization of the air forces on 20 June 1941.3 This action regularized the increasing autonomy of the air arm.  The Army Air Forces were constituted directly under the War Department, with General Arnold as both Chief of the Army Air Forces and Deputy Chief of Staff for Air. A separate Air Staff, with the usual staff sections, soon developed. Within the Army Air Forces two major subdivisions were created: the Air Corps, charged with the control of fixed installations, individual training, supply functions, etc., and the Air Force Combat Command, which replaced the GHQ Air Force in controlling tactical aviation in the continental United States. These changes were recognized when GHQ was reorganized on the following 3 July.4 A distinction between air forces and ground forces was clearly drawn and GHQ was relieved of responsibility for air training. It was charged with the "supervision and coordination, as at present, of the training of all ground combat forces (except those assigned to air forces) and all combined air-ground training (except training for defense against air attack) in the continental United States." This division of authority remained in effect until the reorganization of 9 March 1942, which dissolved GHQ and established the Army Ground Forces and the Army Air Forces as separate and coordinate commands.

Throughout these years GHQ and the air forces shared many common problems in addition to the questions connected with combined air-ground training. In a general way most training activities of the ground forces were in score way related to the development of the air forces. Ground troops on maneuvers were frequently criticized by General McNair for insufficient precautions taken against the air threat. The training of


airborne troops, though a very small-scale matter in 1940-1941, directly involved the air forces. The establishment of defense commands, in which aviation was a principal component, created new organizational problems which were finally solved in accordance with the views of General McNair, published in the Field Manual on Air Defense. The assignment of antiaircraft artillery within a defense command to the interceptor commander was gradually accepted, owing to the influence of General McNair. Most of the relations between GHQ and the air forces, however, centered on the problem of the “combined air-ground training” over which GHQ received supervision on 3 July 1941. The problem of preparing for direct collaboration between air and ground forces in battle proved to be difficult to solve, and no final settlement had been agreed upon when the Army Ground Forces were established in 1942.

Basic Problems in Combined Air-Ground Training

In the blitzkrieg in which the German Army swept over Belgium and France in May-June 1940 no element appeared more successful than the close support given by aviation to ground troops in combat. Other armies had lagged in developing this type of cooperation. The United States was unprepared for such warfare both in equipment and in tactical doctrine. Before large-scale air-support training could begin it was necessary not only to procure equipment but to formulate, for the guidance both of production policies and of training programs, a new tactical doctrine for the close support of ground troops by aviation.

On 20 August 1940 General Marshall directed his G-3, Gen. F. M. Andrews, an Air Corps officer, to initiate staff studies on this subject. The matter was turned over jointly to the Training Branch and the Miscellaneous Branch of G-3. Though Lt. Col. H. M. McClelland, an Air Corps officer in the Miscellaneous Branch, reported that his Branch had already taken sufficient action, the Training Branch nevertheless went ahead. Lieutenant Colonel Ramey of that Branch, after consultation with GHQ, produced on 26 September 1940 a memorandum for the Chief of Staff, signed by General Andrews.5

This memorandum distinguished five kinds of aviation support for ground troops:

(1) Close, direct-support fire missions on the mediate front of ground forces.

(2) Air defense of friendly ground forces and installations in the combat zone.

(3) Air attack against targets in hostile rear areas.

(4) Support of parachute troops and air infantry.

(5) Reconnaissance, liaison, and observation.

Of these items the first and second came to constitute the substance of the air-support problem. The third involved less coordination between air and ground forces in battle and was more a strategic than a tactical concern. The fourth has been described in Chapter VI. The fifth presented relatively few difficulties.

Observation Aviation

Aerial observation, according to General Andrews’ memorandum, was already well handled in the United States Army. Observation squadrons had been assigned in 1940 to various branches and echelons of the field forces. Their training came directly under


the authority of GHQ. They followed prescribed training programs in conjunction with their respective ground units but in May 1941 an inspection by the GHQ Air Section showed that the mere issuance of instructions had not been sufficient to guarantee results. It was found that the Air Officer of the Second Army had tested his observation squadrons very superficially. General McNair, privately referring to these tests as "classroom stuff," ordered further tests in the form of actual field exercises for all observation squadrons in the field forces. The tests were held in July and August 1941.6 At that time, following the reorganization of the Air Forces, the observation squadrons were transferred to new air support commands,7 and the question of aerial observation then merged into the larger question of air-ground cooperation.

The Problem of Air-Support Tests, 1940-1941

In the use of combat aviation for air-support, General Andrews’ memorandum of 26 September 1940 stated that the United States Army was inexperienced by European standards. General Andrews recommended that joint air-ground tests be held at once and that in future maneuvers whole armies, corps, and "large elements" of the GHQ Air Force should be trained to act together.

This recommendation brought about a struggle in the War Department between a group of officers favoring immediate air-support tests and another group who believed that the GHQ Air Force should expand its equipment, enlarge its personnel, and perfect its training in strictly air matters before participating in joint training with ground arms. The struggle was not simply between air and ground officers. General Andrews, an Air Officer, favored immediate tests, as did General McNair, who had concurred in the proposals on 24 September. General Arnold, Chief of the Air Corps, saw more value in air-ground combined training than did General Emmons, Commanding General of the GHQ Air Force. In a training directive for 1940-41, issued to the GHQ Air Force shortly before the Office of the Chief of Staff initiated the studies on air-ground relations, General Arnold had expressly said: "Every opportunity will be sought to engaged in field exercises with other arms." The most persistent opponent of the proposed tests was General Emmons. For a time he was supported by General Arnold, who, on 25 October, in a memorandum for G-3, urgently advised against immediate tests. He declared that equipment was lacking, stated that cooperative exercises with ground troops greatly delayed the "combat crew and unit training" of air personnel, and recommended that no air-support operations be introduced into large maneuvers until 1942. The Miscellaneous Branch of G-3 and the War Plans Division shared this point of view, though somewhat less positively.8

In November, however, the weight of authority turned in favor of the tests. General Arnold, made Deputy Chief of Staff for Air at this time, accepted the proposed tests with certain safeguards. He obtained from General Andrews on 12 November a statement that only one squadron of combat aviation, probably the one stationed at Ft. Benning, would be needed in the near future9 and then gave instructions for the issuance of a directive ordering the tests.10 He wrote also to General Emmons announcing his decision to concur in the G-3 proposal, and assured him that the directive, when issued, would contain a clause protecting the GHQ Air Force from the obligation of large-scale training with ground troops.11 On 20 November, Colonel Ramey, charged with preparing the directive, informally notified colonel Clark at GHQ that G-3 of the War Department favored naming GHQ as the coordinating agency for the various elements—army, corps, Armored Force, and GHQ Air Force—to be involved in the tests. G-3 seemed not to be thinking of small tests only.12

On 2 December Colonel Ramey finished his draft of the directive, which took the form of a War Department letter to the Chief of Staff, GHQ, ordering him to conduct tests and specifying their content. The draft reached the Office of the Chief of Staff


on 4 December. It was there amended by General Arnold, Deputy Chief, who now appeared convinced of the value of the experiment. His amendments doubled the number of questions on which tests should be held. Three days later the amended draft, fully approved, left the Office of the Chief of Staff, but on 12 December Colonel Howard, an Air Corps officer, Chief of the Miscellaneous Branch of G-3, vehemently objected to General Arnold’s extension of the scope of the tests which, he said, the entire GHQ Air Force could not execute in the time allowed. Colonel Howard disclaimed responsibility and predicted a violent protest from General Emmons. These objections, however, were overruled by Colonel Twaddle, acting G-3, who observed that the Chief of Staff had already approved the directive and on 13 December he transmitted the amended draft to the Adjutant General.13

The directive issued to General McNair on 17 December 1940 laid the basis for tests which lasted well into the following summer, and the results of which were eventually incorporated in Basic Field Manual 31-35, Aviation in Support of Ground Forces, published on 9 April 1942.14

Within ten days of receiving the directive General McNair had assembled comments on it from his staff sections, transmitted it to the Third Army with instructions for its execution, requested the Chief of Infantry to furnish a battalion of parachute troops, and asked the Chief of the Armored Force and the Commanding General of the GHQ Air Force to state what forces they could make available.15 Then began a period of difficulty and delay. General McNair himself, despite his prompt action, was not yet convinced that the tests should be hurried. He had been told that the air forces were not prepared, and his principle that fundamental training should take priority over training in specialized operations seemed to run contrary to the scope of the proposed tests. He therefore wrote to General Marshall on 16 January 1941 mildly criticizing "the present test at Fort Benning of air-ground cooperation, as being premature."16

In January 1941 new difficulties developed regarding the size of the tests. The War Department directive, following General Arnold’s assurances to General Emmons, stipulated that aviation used in the tests should "be restricted to the bombardment squadron now stationed at Fort Benning, Georgia, and such additional units of other type aircraft as the Commanding General, General Headquarters Air Force, may make available without undue interference with the unit training and expansion program of the Air Corps." In order to use ground troops in the neighborhood of Ft. Benning the Third Army assigned the tests to the IV Army Corps. General Benedict, Commanding General of the IV Corps, soon found that the number of planes in the squadron at Ft. Benning was too small to permit worthwhile tests.

To iron out this difficulty, a special conference was called at the War Department on 17 January 1941. Two changes in the program, somewhat in the nature of a trade, were approved. First, General Arnold stated that by early April a whole group of light bombardment planes might become available. Second, it was agreed that responsibility for matters primarily of aviation technique should be transferred from the Third Army to the GHQ Air Force. The date for completion of the tests was deferred to 1 August 1941.17

Execution of Air Support Tests, Spring 1941

The tests began on 11 February and lasted until 17 June.18 Eighteen tests were held, becoming increasingly large as additional units of aviation were made available. Command channels were complex. The GHQ Air Force delegated its responsibilities to the Southeast Air District—renamed the Third Air Force while the tests were in progress—which in turn delegated them to the 17th Bombardment Wing. Under the 17th Wing was first the 3d Light Bombardment Group and second the 15th Light Bombardment Squadron,


which with certain observation and pursuit units performed the exercises prescribed. The ground troops employed were the 4th Motorized Division, the 2d Armored Division, and the 501st Parachute Battalion, together with small units of other arms, all attached to the IV Corps, which itself was responsible to the Third Army and GHQ. Command over the tests was not wholly unified, since General Benedict of the IV Corps and General Brereton of the 17th Wing occupied in some respects coordinate positions. Some misunderstanding resulted,19 but the tests were nevertheless found by Colonel Lynd, the GHQ Air officer and observer, to be well managed.20

Aviation worked alternately with the 4th Motorized and 2d Armored Division in various tactical combinations involving both support and attack of ground troops. No live bombs were dropped. The Army possessed no dive bombers, and plans to employ Navy dive bombers failed to materialize. A severe shortage of radio equipment limited communications, and lack of air transport made it impossible to experiment profitably with airborne troops. All units participated at less than authorized strength.

Despite these difficulties a long list of matters was investigated during these months:

(1) The minimum distance from friendly troops at which aviation might safely bomb.

(2) The minimum altitude at which support aviation might safely operate.

(3) Methods of communication between ground and air.

(4) Methods of notifying friendly ground troops when supporting air action was terminated.

(5) Methods by which ground commanders might call for air support and designate targets for bombardment.

(6) Methods by which fliers might distinguish friend from foe on the ground, and ground troops distinguish between friendly and hostile aircraft.

(7) Methods to secure timing of air attack in coordination with ground action.

(8) Proper kinds of targets (near or distant, stationary or moving, transitory or permanent) for aviation in close support of bombardment.

(9) The lapse of time between request for support and delivery of bombardment.

(10) Methods for control of aircraft, whether by attachment to ground troops or otherwise.

On these and other matters General Benedict on 19 July 1941 presented a thorough report, concurred in by General Brereton, who also forwarded a report for the 17th Bombardment Wing. Colonel Lynd found both reports excellent.21 At the same time a draft for a training circular on air support was forwarded by General Benedict.

The Air Support Commands and the 1941 Maneuvers

These reports were eagerly seized upon by both GHQ and the Army Air Forces. The Air Forces, recently constituted as an autonomous body, were engaged in reorganizing air-ground relations. A directive of 25 July 1941 created an Air Support Section in the staff of the Air Force Command, providing, however, that for liaison purposes this


section should be initially located at GHQ.22 Colonel Lyrd became chief of this section. The same directive created five air support commands to include observation and light bombardment planes formerly allotted to ground units. One air support command was to be included in each of the four armies. The fifth air support command was reserved for the Armored Force. To determine the functions of the new commands an Air Support Board met on 28 July.23 It carefully considered the reports of the IV Corps and the 17th Wing and recommended material for inclusion in the forthcoming training circular.

General McNair had long urged speedy completion of reports on the air-support tests held in the spring of 1941, since conclusions were needed in time to formulate instructions for the fall maneuvers. While recently relieved of his supervisory authority over air training, General McNair was still officially responsible for supervision and coordination of combined air-ground training.24 He had feared that instructions for air-ground action in the coming maneuvers might be delayed by the “upheaval” going on in the Air Forces. The placing of all support aviation it air support commands, he wrote, "is one more step in the separation of the air from the rest of the Army. What may be the result is hard to predict, but it seems quite unlikely that it will facilitate the interworking of air and ground." He was not satisfied with the results shown in General Benedict’s report, which he praised as accurate and thorough. "Frankly," he wrote to General Benedict on 26 July, "I am disappointed in the capabilities of air support as indicated by your tests. It seems that aviation may intervene once in a battle—possibly at the time and place needed, possibly not. It requires a great stretch of the imagination to visualize such action as even remotely decisive, if indeed it is felt at all by the ground troops. I hope that the maneuvers may develop something more impressive in the way of speed and ferocity of air action."25

It was clear by July 1941 that the fall maneuvers would involve large-scale air support operations, such as General Andrews had envisaged and General Emmons had objected to almost a year before. Though General Emmons,26 now Commander of the Air Force Combat Command, had not changed his opinion, General Arnold ordered eight groups to take part in both the Louisiana and the Carolina maneuvers.27 The Navy agreed to supply dive bombers.28

The eight-group program called for speedy preparations. Further air-ground tests were scheduled for Ft. Knox in August.29 Starting at the end of July Colonel Lynd as well as General McNair and General Clark worked on the Air Corps draft of an Aviation Supplement to the GHQ Umpire Manual. Much reduced in bulk by its passage through GHQ, with sentences shortened and expression generally clarified and with the number of air umpires cut about 30 percent, the Aviation Supplement was published on 21 August.30 Meanwhile, Colonel Lynd substantially modified the draft training circular submitted by General Benedict, taking account of the findings of the Air Support Board. The War Department published Colonel Lynd’s draft, with a number of changes, as Training Circular No. 52, 29 August 1941, "Employment of Aviation in Close Support of Ground Troops."31

To improve the doctrine set forth in this training circular and to obtain guidance for the preparation of a field manual, General McNair on 8 September requested the principal field commanders to submit reports on air support in the light of the coming maneuvers. On 15 September GHQ was formally instructed by the War Department to prepare a field manual.32 The stage was set for a thorough test of air-ground cooperation.

Aviation in the 1941 Maneuvers

The September maneuvers gave the most spectacular exhibition of air power ever seen in the United States. The eight Army Air groups took part as well as seven


squadrons of Navy and Marine aviation. Even so, in proportion to the 350,000 troops engaged, air strength was below the normal requirements of modern war.

In his comments General McNair confined himself to brief statements. He noted in his critique that the troops failed to respond adequately to the air threat. He pointed out that columns of men and vehicles in close order on the roads would suffer disastrously from real air attack and that for observation planes to fly for two hours at low altitudes over enemy territory was highly unrealistic. General Arnold also noted these weaknesses as well as others: poor use of radio; excessive dependence on telephones; scattering of bombers on small missions; ignoring by aircraft of danger from antiaircraft artillery; and undue length of communication channels between the ground commander’s request for support and its delivery by the air unit. General Arnold found the air support command organization vindicated in principle by the maneuvers, but requiring development in detai1.33

In the Carolina maneuvers in November the First Army was pitted against the IV Army Corps, most of whose elements had participated in the air-support tests earlier in the year. The eight groups of Army aviation were again engaged and, theoretically, dropped fourteen thousand bombs.34 Parachute troops also were employed.

General McNair still thought the ground troops careless in the face of the air threat. He found bombardment aviation used aggressively and effectively by the IV Corps Commander, pursuit planes employed normally but too often wasted on attacks against ground objectives, observation planes still too inclined to long leisurely flights over enemy positions, and the capabilities of aerial photography neglected. A report by an A-2 Air Staff observer dealt chiefly with matters of special interest to the Air Forces.35

Disagreements over Air-Ground Command Relations

After the September maneuvers attention again turned to the precise means by which air and ground units should be administratively related.

The Chief of Field Artillery, not content with the new air support commands, recommended on 8 October that at least seven observation planes be organically included in the artillery component of each division and corps. GHQ concurred with G-3 of the War Department in disapproving this proposal.36 "I favor exhausting the possibilities of the new air-support organization," wrote General McNair on 21 October 1941, "since it gives promise of effecting a great improvement. There is grave question in my mind whether it is feasible or desirable that a ground arm attempt to operate aviation. The ground arms can and must learn to cooperate with aviation, and the process may as well begin with observation." These words mark a change, perhaps brought on by the maneuvers, from his distrust of the air support commands at the time when they were created. He now believed that observation planes could survive only where general air superiority was maintained and that they would be wastefully used if decentralized in division commands. The recommendation of the Chief of Field Artillery had no immediate effect in this connection. A War Department order of 27 October prescribed that observation units of air support commands should be attached to ground units as required, in peacetime by agreement between GHQ and the Air Force Combat Command—with the War Department as arbiter when agreement was impossible—and in wartime by decision of the theater ccmmander.37

The Armored Force also expressed dissatisfaction with air-ground command arrangements. The Commanding General of the I Armored Corps wrote on 20 October that, when a ground commander did not control his supporting air unit, he could not be certain of what support he could draw on and therefore often gave less prominence to aviation in


his plans than it deserved. General Devers, Chief of the Armored Force, accordingly recommended to GHQ that the 5th Air Support Command be attached to "participating elements" of the Armored Force in future maneuvers.38

When the air support commands were created in the preceding July, the 5th Air Support Command had been specifically designed for attachment to the Armored Force and each of the other four for attachment to one of the armies. For the duration of such attachment the army or Armored Force commander had authority over his air support command. The present difficulty involved relations with the subordinate echelons; corps, division, or combat team under the army or Armored Force; observation, bomber, or pursuit units under the air support command. These forward units of air and ground forces, which did the actual fighting, stood in a cooperative relation to each other. They were links in separate chains of command which converged only at the top. In the daily and hourly realities of warfare command was divided. The ground commander of a corps or lesser unit could request, but could not order, the corresponding air support officer to give support. The Air Forces insisted on maintaining these arrangements. General Emmons said: "Coordination is primarily the responsibility of the commander of the troops supported. To his reasonable needs and requests the air commander will conform."39

War Department G-2 had received reports from the Middle East, which seemed to mean that the Royal Air Force had conspicuously failed to support ground troops. The British had been disastrously defeated by Rommel in the spring of 1941, and one cause of their weakness was held to be the separation, both in training and in combat, between the Royal Air Force and the Army. This separation was reported to have been bridged over only by the presence in the theater of a personal envoy of the Prime Minister, Sir Oliver Lyttleton, Minister of State. Partly in view of these reports, Colonel Kingman, Armored Force officer at GHQ, favored General Devers' recommendation for the attachment of air support to subordinate ground units. He stated that the question involved was the unity of command in a task force and that General Devers proposal followed the German system, which had proved repeatedly successful and which gave control over aviation to subordinate field commanders within an army or theater.40 General McNair, however, took no action on Colonel Kingman's recommendation.

The principles of centralization and decentralization of air support strength were in conflict. Centralization would require the attachment of air forces to the highest Army commands, allotting these forces only temporarily to lower units as occasion required. This method preserved the fluidity and mobility of support aviation and made possible the assembling of mass striking power against the most important objectives. This was the principle urged by General McNair for the organization of tank destroyer units. Decentralization, through attachment of air support to lower commands, would speed up the local delivery of support. The bad feature of decentralization was to immobilize air strength in places where it might not be needed or to fritter it away on local and insignificant missions. The bad feature of centralization was to set up long command and liaison channels and to slow down the process of getting air assistance to ground troops.

FM 31-35 "Aviation in Support of Ground Forces."

This issue, along with other air-ground problems, was covered in the manual, FM 31-35, Aviation in Support of Ground Forces, which was published by the War Department on 9 April 1942. From November 1940 to January 1941 GHQ had received the reports on air-ground operations in the fall maneuvers requested by General McNair on 8 September. Colonel Lynd and other officers, working with these reports, had produced a draft field manual, which was submitted by GHQ to the War Department on 31 January 1942,41 Except for a few minor changes in wording, and with no changes in the attached organization charts, this draft was accepted by the War Department.


On air-ground command relations the doctrine was flexible. "An Air Support Command," said the manual, "is habitually attached to or supports an army in the theater." Normally the air support commander was to function under the army, theater, or task force commander. He would allocate, and in exceptional cases might attach, aviation units to subordinate ground units, but it was emphasized that the air support commander was to control all participating aviation. In general, the principle adopted was that of centralization.42

Various means were prescribed for achieving the advantages of decentralization. Air observation units would normally be allocated so as to permit corps and division commanders to plan their use and to call on them directly for missions. They were allotted on the basis of current T/O’s. Combat aviation, the manual stated, might be attached to subordinate ground units in rare and exceptional cases, when effective control of such units could not be retained by the air support command. Normally, combat air support would be arranged at subordinate levels between air and ground officers by liaison methods. Each corps headquarters, and on occasion division headquarters, was to include an “air support control,” a group of officers in direct communication both with airdromes and with the air support command. Lower headquarters, down to any level required by the tactical situation, might include an "air support party," defined as "a highly mobile group composed of one or more air support officers... to transmit air support requests to air support control..." Within an army corps air support parties would rarely be detailed to a unit headquarters below that of infantry division. On the other hand, in armored forces and cavalry divisions, they would frequently be detailed to headquarters below the divisional level to meet the requirements of rapid movement. They could transmit only requests approved by the ground unit commander and only to an air support control.

The manual emphasized that aviation called for by ground commanders and obtained through air support controls was not subordinate to the supported commander, but remained under the control of the air support command. It was hoped that decentralization of liaison and communications would provide promptness and accuracy in the delivery of air support, in spite of this rigid centralization of air command. Provision for unified command was made only in the loose statement that the air support commander "normally functions under the army, theater, or task force commander."

In other words, the manual did not decide the basic problem of centralization or decentralization. The advocates of either principle had strong arguments to support their views, but a final and realistic decision could only be reached on the basis of active combat experience of our own forces. This opportunity was not offered until the Battle of Tunisia in the spring of 1943, when the Army Ground Forces, instead of GHQ, had to apply the lessons learned.



Go to:

Previous Chapter

Next Chapter

Return to Table of Contents

Search CMH Online
Last updated 17 February 2005