The Supreme Command

SHAEF Concentrates on OVERLORD

General Eisenhower arrived in London on 15 January 1944 to begin converting the COSSAC staff into the Supreme Headquarters, Allied Expeditionary Force (SHAEF). Planning would continue for months yet, but now with troops and a solid purpose. OVERLORD was coming to the fore. The hope of an early German collapse, strong in the fall, had dwindled with the arrival of winter. At the turn of the year COSSAC had put aside RANKIN C, and shifted to RANKIN B, a projected lodgment on the Continent in the event of a German withdrawal from France. COSSAC completed a directive for RANKIN B on 14 January. By then, however, the Germans were reinforcing France, and a voluntary withdrawal to the West Wall was becoming very unlikely. SHAEF's concern would be with OVERLORD, the invasion and drive into Germany, although the RANKIN conditions, a sudden partial or complete German collapse, could not be put entirely out of mind.1

In the RANKIN operations, civil affairs and military government would have been paramount from the first. Under OVERLORD they would lie subsidiary until the issue had been decided in the field, but ultimately no less essential. RANKIN would have resulted in an improvised occupation with limited forces probably not capable of reaching into Germany beyond the coastal cities and the line of the Rhine River. Even such a limited occupation, however, presupposed control with token forces and, hence, would have required a complete absence of German resistance and Soviet interference, especially the latter, since the Russians would probably have the overwhelming preponderance of strength on the scene. OVERLORD, on the other hand, would array fully operational U.S. and British forces on the Continent and, whether opposed or under so-called RANKIN conditions, would very likely culminate in a deep sweep into Germany. SHAEF would command the assault on Hitler's "Fortress Europe" and be the executive organ for establishing the occupation. Therefore, OVERLORD would require a military government able to operate in the wake of battle almost side by side with the frontline troops as well as to govern a defeated Germany from top to bottom. Either was a massive enough undertaking in its own right. Doing both required answers to questions that had so far barely been raised.

Civil Affairs Becomes G-5

Under COSSAC, civil affairs had found solid acceptance in principle but had remained an anomaly in the staff structure. COSSAC had used the American "G" system but had not assigned a G number to civil affairs. In October 1943, the country


GENERAL SMITH. (Photograph taken in 1946.)

GENERAL SMITH. (Photograph taken in 1946.)

houses had given way to the Civil Affairs Division, COSSAC, which consisted of a small planning staff to work on the RANKIN plans and six advisory branches. The branches, made up of the former country house personnel, were designated by function: legal, fiscal, supply, governmental affairs, economic affairs, and information. Their assignment had been to produce instructions for subordinate commands and to reconcile US and British civil affairs military government policies, that is, to write a combined manual-the Standard Policy and Procedure.2

In November, the War Department had authorized two general officers, ninety other officers, five warrant officers, and sixteen enlisted men for the Civil Affairs Division, COSSAC; but in the third week of the month only fifteen officers and one warrant officer had been assigned.3  In the branches, which numbered nearly two hundred officers, the Americans were only sparsely represented. Late in November, Brig. Gen. Frank J. McSherry, who had served in military government in Sicily and Italy, became the ranking US civil affairs officer in COSSAC and coequal chief of the Civil Affairs Division with General Lumley.

Although COSSAC had, in the Standard Policy and Procedure, stipulated a complete fusion of civil affairs-military government with the military command, it had in the Civil Affairs Division in fact maintained what amounted to a separate staff too diverse in its functions to be integrated with the military command. Eisenhower's chief of staff, Maj. Gen. Walter Bedell Smith, on his arrival in London early in January immediately recognized the latent similarity to AMGOT, toward which he had not developed the antipathy that prevailed in COSSAC and to some extent in the Washington Civil Affairs Division. Already thoroughly convinced by his Mediterranean experience of the value of civil affairs and in good part responsible himself for the AMGOT organization, Smith objected only to the "ponderous and unwieldy" dual British-American headship of the COSSAC Civil Affairs Division. He proposed to appoint a single head and, going even a step beyond the AMGOT analogy, asked for a civilian of subcabinet rank to fill the post, naming Assistant Secretary of War McCloy as his choice. General Hilldring, while replying that Secretary of War Stimson would not part with


McCloy, attempted to divert Smith from the idea of a civilian head by suggesting that he use the War Department Civil Affairs Division, which had a military chief, as his model. Anxious, however, to avoid seeming to interfere with a theater commander's right to organize his own staff, Hilldring assured Smith that he (Smith) and Eisenhower had a completely free hand in organizing civil affairs for SHAEF. He had asked for a high-powered civilian, Smith then explained, in order not only to have an expert administrator but to acquire a kind of lightning conductor as well. The idea of civil affairs entirely in professional military hands obviously made him somewhat uneasy. The Civil Affairs Division, he agreed, was military, but it had the Secretary of War behind it.4 No doubt, Smith was aware that the simplest solution for him and for Eisenhower would have been to turn civil affairs over to McCloy or to Hilldring, whom he also invited to assume the job, and thus draw the War Department into a kind of partnership within the theater.

By the end of the month, when Hilldring also declined the position, citing Chief of Staff Marshall's "forcible" opposition to his trying to "break out" of the Pentagon, Smith had become immersed in fundamental civil affairs problems that could not wait until a chief was found.5 Obviously the whole COSSAC Civil Affairs Division, a conglomerate of the former COSSAC civil affairs planning staff and the country house personnel to which a number of American officers were being added, could not be taken into the SHAEF staff. General McSherry advocated a virtually outright return to AMGOT, then beginning to be referred to more loftily as the Mediterranean system, with civil affairs subordinate to the Supreme Commander but otherwise practically autonomous.6 Smith also was still thinking in terms of the Mediterranean approach. Colonel Bendetsen as a proponent of integrating civil affairs into the military command stood alone among the Americans and would shortly be transferred out of SHAEF.

On 5 February, General Lumley, as the senior of the COSSAC Civil Affairs Division's dual heads, sent Smith an organization plan. Since SHAEF would be drawn together at one location in London within the next few weeks, Lumley believed it was time to fit civil affairs directly into the staff. To do this he proposed creating a G-5 division at SHAEF, initially with thirty-five or forty officers, later to be brought up to a strength of about sixty. G-5 would formulate policy and co-ordinate the work of a rear echelon to be composed of the existing Civil Affairs Division branches plus a German planning unit. The only completely separate function then would be training, which would be carried out at the American center in Shrivenham and at a British center in Eastbourne.7

In his plan, Lumley had solved the problem of dual heads and that of civil affairs' place in the command structure, neither, however, to Smith's satisfaction. The next day Smith told Hilldring that he had secured Eisenhower's approval for a plan of his own. SHAEF would retain a small civil affairs section, "possibly to be designated G-5." Lumley would head it but would have to lie supported "by a deputy . . . in whom both the Commander in Chief and


myself have complete confidence." For this assignment, Smith asked for Brig. Gen. Julius C. Holmes, who was then in Italy. Civil affairs training, organization, and detailed planning would be done under McSherry outside the general staff. 8 In his reply to Lumley on the 7th, Smith commented acidly that the general staff division would not need sixty officers. It should be a small, policy-making body. Lumley would be G-5, but McSherry would have to be "considerably more than a superintendent of training." 9

A conference in the Civil Affairs Division three days later predictably culminated in a victory for the advocates of the Mediterranean system.  Lumley and McSherry agreed that Lumley would become Assistant Chief of Staff, G-5, with Holmes as his deputy, but G-5 would have no more than thirty-five officers and would lie limited to policy-making, advisory, and review functions. McSherry would be appointed Deputy Chief Civil Affairs Officer (DCCAO) and head a civil affairs special staff removed organizationally and physically from SHAEF G-5, its seat being at Shrivenham. The DCCAO would direct training and prepare detailed plans and instructions. To accomplish the latter function, country missions, a revival of the old country houses including a German section, would be established. After the invasion, the country missions would "sit alongside" restored governments in liberated countries, with the German section eventually becoming the US element in the Allied control organization for Germany. SHAEF G-5 would review the DCCAO's plans and "exercise general supervision over their execution," but only one of the G-5's six sections would lie concerned directly with civil affairs operations. The rest would handle fiscal, supply, legal, economic, and staff duties within SHAEF.10

In Staff Memorandum No. 2, 15 February 1944, Smith established SHAEF G-5 and the Special Staff effective on this date, and confirmed Lumley and McSherry in their appointments.11 Ironically, on this day the members of the Civil Affairs Division were drawn together in one place for the first time. The division had begun moving the day before from scattered locations in London, at Cadogan Square and Norfolk House, to Prince's Gardens. By the 16th, when the move was completed, the reorganization had also been accomplished and the staffs, though in physical proximity, were organizationally separate. Even this condition would not last long. The Special Staff was to move to Shrivenham before the end of the month and the G-5 section would go in early March to the SHAEF headquarters compound, WIDEWING, in Bushy Park on the outskirts of London. 12


In the conversion of Civil Affairs Division, COSSAC, into G--5 and the Special Staff, SHAEF, the Mediterranean concept seemed to have won out. G-5 had been practically shorn of operational control, the means for exercising it being concentrated


in the Special Staff. The proponents of the Mediterranean system, however, had not yet scored a complete victory. Staff Memorandum No. 2 had in no way rescinded the Standard Policy and Procedure, which had been the fundamental guidance for the army groups and armies since December 1943. In fact, on 12 February Eisenhower had approved 21 Army Group's first joint air and ground force plan for the invasion, and General Sir Bernard L. Montgomery's staff, following the Standard Policy and Procedure, had assumed civil affairs to be an integral function of the tactical commands from the army group on down.13 On 19 February, General Smith revealed that he had in fact not made a definitive choice between the proposed COSSAC and the Mediterranean systems; he approved a G-5 directive to McSherry, confirming the missions assigned to the DCCAO four days earlier but only during the planning and preparation for OVERLORD. The real issue, how civil affairs in northwestern Europe (military government in Germany) would be conducted during and after the invasion, then, remained undecided.14

After Holmes arrived in London, his and Lumley's views proved so divergent as to impel them, on 10 March, to appeal jointly to Smith for a decision.15 On the surface, Lumley appeared to have entered an unequal contest; he had been a reluctant choice for G-5. Smith had asked for Holmes. Holmes had two proposals to make: the first, to cancel the Standard Policy and Procedure; the second, to issue a new directive to McSherry expanding his authority and extending it into the period following the invasion. Standard Policy and Procedure, Holmes argued, was inconsistent with Staff Memorandum No. 2. The memorandum placed civil affairs planning and operations on a countrywide basis, while the Standard Policy and Procedure put all the authority in the hands of the individual military commanders. Furthermore, the Standard Policy and Procedure made no provision for the so-called hiatus areas, parts of a liberated country which might not actually be occupied by SHAEF troops. This omission and the statements suggesting that SHAEF and its subordinate commands might not set up military governments everywhere they went were, Holmes insisted, in violation of both international law and declared United Nations objectives. Moreover, Holmes said, to claim, as some were doing, that the Standard Policy and Procedure had to be continued in force because it constituted the only guide for planning by subordinate echelons was fallacious. FM 27-5 and the British War Manual provided adequate general policy statements.16

As his second proposal, Holmes wanted to delegate the Supreme Commander's legal authority to conduct military government to McSherry as DCCAO, making him the commander of all civil affairs organizations in northwest Europe. He would appoint chiefs for each country, and the staffs at Army group and lower formations would be special staffs receiving their orders through civil affairs channels, not from the military commands. McSherry would control all of the civil affairs detachments working directly under SHAEF,


issue technical instructions to any other detachments, and dispatch orders to the army group and lesser commands through the appropriate G sections of SHAEF.17

Lumley defended the Standard Policy and Procedure, not in its details but in what he called its basic conception, namely, that the first purpose of civil affairs was to further military operations. Full-scale military government, he therefore maintained, ought to be confined to the areas where military necessity was paramount, the zones of operations and communications. In hiatus areas of liberated countries the indigenous governments should be responsible, under just enough surveillance by SHAEF missions to make certain that they did not prejudice military operations by their actions or through failure to establish their authority. In the agreements with the exile governments, Lumley pointed out, promises had been given not to interfere in civil matters any more than was essential to military operations and to restore authority to the national governments as soon as the military situation permitted. Germany, he conceded, was a somewhat special case since the whole country would be placed under military government, and a country headquarters would be required because SHAEF would not relinquish control to a German government but rather to an Allied agency of some kind. To these arguments Lumley added one other: canceling the Standard Policy and Procedure would reverse and reject COSSAC's policy after it had been in effect for several months. "The British Army particularly" would not be happy, and much good will built up for civil affairs in the military staffs might be lost by so complete a turn about. The best course, he suggested, would be to keep the Standard Policy and Procedure in force as a general guide, supplementing it as needed with additional plans and directives while holding to the basic premise of civil affairs integration into the staffs at all levels.18 The Supreme Commander, Lumley insisted, should delegate his military government authority not to a DCCAO but to the army group commanders, who could redelegate it as needed to their subordinate commanders. Plans drafted in the Special Staff would then be approved by G-5 and issued to the army groups for them to convert into operational directives.

Smith's reply read like a death sentence for the Mediterranean system. He admonished Lumley and Holmes, though Lumley scarcely needed it, to bear two points in mind: first, that the Mediterranean organization, although it did a good job, had many defects; and second, that conditions in northwest Europe were different from those in the Mediterranean. Therefore, he continued, civil affairs staffs would be closely integrated with normal staffs throughout the chain of command; civil affairs headquarters would not be established unrelated to military headquarters; and the AMGOT approach would he avoided. That AMGOT would not be duplicated in northwest Europe, he stated, had been directed in "the latest paper from the U.S. Chiefs of Staff." 19 Continuing, he described the Standard Policy and Procedure as a sound document in which revisions could be made as long as they did


not end in scrapping the basic principles. He rejected Holmes's proposals on organization completely except for a civil affairs technical communications channel, but even it would run through G-5 SHAEF to the appropriate sections in lower headquarters, not through the Special Staff. Although he had put operating personnel in the Special Staff, Smith insisted, he had not meant thereby to enhance the DCCAO's role. The command and staff channel would run from SHAEF and G-5. G-5 should add one or two officers to its Operations Branch for each of the national areas. In the operations zone, SHAEF would relieve the army groups of civil affairs duties outside combat areas as soon as it could, and would assume the responsibility itself.20

In the reply to Lumley and Holmes, Smith made the fundamental decision on civil affairs organization for the SHAEF period. He now accepted Standard Policy and Procedure, which he had undercut in Staff Memorandum No. 2, almost intact along with Lumley's defense of it. Lumley's success, however, unfortunately for him, was going to shorten his tenure as G-5 rather than prolong it.

If Smith had not changed his mind during the month after G-5 and the Special Staff were created, he had certainly developed a much firmer attitude toward the two systems than could have been deduced from Staff Memorandum No. 2. The reasons are not hard to find. He had no doubt concluded, as Lumley also suggested, that the Standard Policy and Procedure, even though it was sired by Colonel Bendetsen, an American, had been an agreed document, one which could not be discarded simply because Americans coming from the Mediterranean liked the system they had used there better. Moreover, he had had time to become aware, if he had not been before, of the Quebec decision in August 1943 not to establish national military governments in liberated countries.21 Most importantly, in the document to which Smith had referred in his reply to Lumley and Holmes, the JCS, while recognizing as Lumley had in his memorandum the probable need for some kind of co-ordinated military government, had described a fully combined U.S.-British military government on the pattern used in Italy as apparently not feasible for Germany.22

After Smith made his decision, although the controversy by no means ended, the case for the Standard Policy and Procedure rapidly became stronger. On 13 March Smith had asked how G-5 could be reorganized to permit it to supervise detailed planning and later on to co-ordinate operations on the Continent without country headquarters. In its answer, G-5 undertook to delineate the respective advantages of a functional organization (for example, supply, public health, labor, and law, without regard to national boundaries) and a regional organization. The functional organization as recommended in the Standard Policy and Procedure, G-5 maintained, would permit the Supreme Commander to govern with an even hand throughout northwest Europe and not have to contend with national staffs each devising its own policy. In any event most civil affairs problems would not fall within specific national boundaries. A regional organization, on the other hand, would ease the way for restora-


tion of national governments and would take account of differences in laws, languages, and customs. In general, the advantages of the one were the disadvantages of the other. Nevertheless, the weight of advantage lay on the functional side because Eisenhower's responsibility would be for the whole continent and many problems could be dealt with functionally without reference to boundaries, but few could be handled on a purely national basis. In its proposal, then, G-5 attempted to give predominance to the functional organization without entirely excluding the regional. G-5 would revise its functional branches to enable them also to cope with problems which required a national approach, and it would set up nuclei of a few officers around which national staffs could be built if they were required after operations began. 23

Subsequently, the weight of advantage shifted even more decisively to the functional organization when Smith showed himself to be receptive. The Mediterranean system's advocates lost ground when they could not prove that because the system had been tested in action in Sicily and Italy it would succeed under different conditions in northwestern Europe. The proponents of the Standard Policy and Procedure, on the other hand, who had argued, while they were on the defensive, mainly for the basic concept, could also cite practical military advantages: the elimination of a separate civil affairs command channel, the direct tailoring of the civil affairs plan to the military plan, and the integration of civil affairs and military supplies. This last point was particularly telling. A frequent complaint of the civil affairs operation in the Mediterranean had been that it had last call on shipping space for relief supplies and even when it was able to obtain ships could not get them loaded and unloaded. Under the Standard Policy and Procedure the military commanders, being responsible for civil affairs, would have to combine civil affairs supplies with their military supplies. One balance sheet drafted in G-5, probably in late March or early April, enumerated eleven points in favor of the functional organization as opposed to four for the Mediterranean System. 24

Before the end of March the functional system clearly predominated. On the 24th, G-5 announced a forthcoming reorganization designed to eliminate the last serious doubts on this score and at the same time strike enough of a compromise to avoid an absolute rejection of the Mediterranean approach. G-5 would be divided into two parts, Policy and Operations. Policy, subdivided into functional branches, would come directly under the Assistant Chief of Staff, G-5. McSherry would head Operations, bringing enough officers with him from Shrivenham to establish country branches. He would retain the functions he had as DCCA0 without the title, which in civil affairs usage implied direct subordination to the Supreme Commander who was also the Chief Civil Affairs Officer. He would gain one function, policy supervision of army group plans and their execution, lout without direct command authority. He would also continue to control, through a deputy, the Special Staff which, made up of the country sections less the officers drawn into the Operations side of G-5, would stay at Shrivenham. Although McSherry if anything seemed to have gained


somewhat, his and the Special Staff's loss of autonomy was a crippling blow for the Mediterranean system, and it was underscored by a provision that hard-and-fast lines would not be drawn to separate the competences of Operations and Policy.25

The watchword for SHAEF reorganizations in those early days was "no sooner said than done," but this one took over a month to accomplish. In part the delay can probably be attributed to the Mediterranean forces fighting a rear guard action, the rest, no doubt, to Smith's determination not to have Lumley as chief of the reorganized G-5. Smith had not been happy with Lumley in January and did not want him at all as Assistant Chief of Staff of a much enlarged G-5. Even though civil affairs would not have a separate role, he still wanted an officer with more rank and military stature than Lumley possessed. Lumley, like some other ranking British civil affairs officers, was not a professional soldier but a former colonial administrator, most recently governor of Bombay.26 Smith, who by no means objected in principle to a uniformed civilian, did not see the same potentials in one from the upper reaches of the British civil service that he saw in a deft Pentagon and Washington hand like McCloy. Whether the G-5 chief was a civilian in uniform or career military, Smith's apprehension of the inherent touchiness of civil affairs made him prefer an American. But the position in the end fell to British Lt. Gen. Sir A. E. Grasett, until then chief of SHAEF's European Allied Contact Section. Although Smith proposed several American officers, Eisenhower preferred to leave the post to the British to avoid friction in an area where England had strong traditional interests.27 Grasett took over as Assistant Chief of Staff, G-5, in mid-April, receiving his official appointment on the 22d. Lumley was relieved, and Holmes remained as deputy.

Among Grasett's first actions as G-5 was an attempt to lay to rest the Standard Policy and Procedure-Mediterranean system controversy still smoldering in the background. On 19 April he issued a policy statement for general distribution to serve "as a guide to all future planning and preparations." In the introductory paragraph he stated: "It has become apparent to me that some confusion of thought exists on the method by which civil affairs will be conducted once operations start. Time is short and cannot be wasted on fruitless discussion." SHAEF, he said, would control civil affairs operations directly, as would the subordinate commands. There would be no intermediate staffs between SHAEF and the forces in the field. How the country sections would be employed, if at all, had not been determined; but no matter what the decision, they would work under SHAEF, not independently of it, in the liberated countries and in Germany as well.28

The fight took its toll. Lt. Col. James H. Shoemaker, of the Provost Marshal General's Civil Affairs Division, in England at the time on a two-week tour of duty, found morale sagging. Most civil affairs officers, as far as he could discover, tended to favor the Standard Policy and Procedure system. General Holmes, however, still wanted a substantial role for the country sections, and many officers not familiar


with the whole picture saw only serious disorganization.29

On the 28th, Grasett announced the final form of the reorganization, to be effective on 1 May. G-5, Policy, would form six branches: legal, fiscal, supply, public health, displaced persons, and economics. Operations was to have a plans branch and six country sections. Through the Special Staff, it would supervise the old country sections at Shrivenham as units and the training sections at Eastbourne and Shrivenham.30 Though this was not by any means to be the last reorganization, it did establish an important principle: as long as SHAEF existed, civil affairs and military government were to be a direct responsibility of the military commanders. The principle was underscored on 1 May in the publication of a revised Standard Policy and Procedure in which all of COSSAC Civil Affairs Division's basic assumptions were retained. Two concessions to the Mediterranean concept were added in the form of a provision for a direct civil affairs channel of communications and an authorization to establish civil affairs in and undertake rehabilitation of a whole country even though SHAEF forces occupied only a part of it.31 The second point was in part also a response to the President's transfer of relief operations in Europe to the Army in November 1943.

The principle was established but the uncertainty did not diminish. Three weeks later, on 25 May, Brig. E. A. L. Gueterbock, McSherry's deputy in charge of the Special Staff, attempted to explain the status of the German Country Unit to its members. It was curious, he said. They were definitely not part of SHAEF. They were a planning unit but not a control mission in embryo, though they might someday in some fashion become one. He was sorry that it all sounded somewhat "woolly," but the decisions at the highest level had not yet been made. 32




Previous Chapter        Next Chapter

Return to the Table of Contents

Page Last Updated 16 April 2003

Return to CMH Online