The prolonged delay resulted primarily from the complexity of the new Japanese machine. Its impending introduction had been anticipated for several months in 1938, when distribution of the equipment was noted in intercepted messages. Only the more important Japanese embassies received the new machine, including Rome, London, Washington, and Berlin. Those which did not received the PURPLE equipment continued to use the Type A machine. in fact, when maintenance was required on the new equipment, a Japanese embassy frequently reverted to the Type A machine.58
Recovering from their initial confusion, U.S. cryptanalysts quickly began to exploit the new machine despite its complexities. By 10 April 1939, Frank B. Rowlett and Robert O. Ferner had produced partial texts based on similarities between the A and B systems.59 However, on 1 May 1939, apparently recognizing the vulnerability of their new system, the Japanese introduced significant complications to the recovery process. By 27 November 1940, however, U.S. Army analysts produced two translations which represented the first solutions to the B Machine.60 OP-20-G played a minor but
important role during this period. PURPLE analog machines, based on wiring designs developed by the Army, were made at the Navy Yard in Washington, D.C., and distributed to the War and Navy Departments, i.e., to SIS and to OP-20-G, and by mid-1941, to Admiral Thomas C. Hart, CINCAF. The British in London were given equipment originally intended for Hawaii. After the Army solved the system, Navy and Army cooperated in recovering the daily changing keys. Messages from both the PURPLE and Red machines were known as MAGIC.61 Once the PURPLE machine became readable and the need for translated current Japanese diplomatic messages became urgent, the War Department requested additional Navy assistance in the form of cryptanalysts and linguists.
Ishii Agreement of 1917, in fact, specifically recognized Japan's special position in Manchuria and on the Shantung Peninsula. moreover, until 1941 the U.S. consistently supplied Japan with the war materials necessary to undertake and sustain operations not only against China but against the Netherlands and France as well. At the same time, the United States maintained a naval rivalry with japan which, because of various factors, had already begun to tilt in Japan's favor following the end of World War I.
To the U.S. Navy these policies contained serious strategic implications. in the early 1920s the United States was faced with the unpleasant prospect not only of the continuation of a prewar Anglo-Japanese alliance with unfavorable balance of power implications, but with the equally distressing prospect of a superior Japanese fleet in the Pacific, occupying the German islands which lay astride U.S. lines of communication to Australia and making defense of the Philippines virtually impossible. Aided by Canada and Australia at the Washington Conference in late 1921, the U.S. succeeded in replacing the Anglo-Japanese alliance with a four-power treaty with Britain, France, and Japan. This treaty unfortunately limited U.S. and U.K. base building in the Pacific in return for reluctant Japanese acceptance of apparently unfavorable ratios in naval strength. Although not at first seen as an advantageous treaty for Japan, several factors conspired to make it so. Among these were an obsolescent British dreadnought fleet which effectively eliminated the British Asiatic Fleet as a force; a moratorium on battleship construction which saw the United States scrap twenty-eight vessels including eleven capital ships in various stages of completion;64 a U.S. commitment to a two-ocean navy which meant that not all new ships joined the Pacific Fleet; and the base-building restrictions of the four-power treaty. Collectively these measures left Japan in a position of local superiority and in a dominant position regarding the coast and approaches to China, the treaty notwithstanding.65
Forced by domestic economic considerations to cut back on military spending, the U.S. continued to adhere to arms limitations agreements and self-imposed building moratoriums well into the 1930s while the Axis powers skillfully circumvented them by modernization programs and new construction. By 1939 both the U.S. and British navies had fallen behind the Japanese Navy, not just in numbers of modern vessels but particularly in the technology of naval architecture and naval armaments, ship design, hull speeds, torpedoes, and the caliber of ships' guns. one bright spot during this period, as noted earlier, was that the U.S. was quick to react to the COMINT-derived information concerning battleship speed revealed in late 1936. Regrettably, there was no corresponding move to upgrade the Navy's COMINT program.
yeomen. Accordingly, since these students were not given career field designators appropriate to their unique role in naval communications (i.e., intercept operators), it was not unusual for almost twenty years to see correspondence from OP-20-G to naval district commanders rescuing these people from being assigned duties in their ostensible fields.
Until late 1941, the number of intercept operators in the Pacific was never very high, thus making their daily availability a matter of some concern to the resident officer-in-charge (OIC). In the Philippines in 1933, for example, it reached an unusually large total of eighteen men whose orders of assignment to the 16th Naval District carried the caveat "only for intercept or RI [radio intelligence] research work." Ordinarily, from 1930 to 1936, when the first DF site opened at Sangley Point on Luzon, the average assigned strength was only nine. Lack of numbers, however, did not reduce either the amount or nature of the work required. in 1937, a fifteen-man work force at Station C was divided into four three-man watches. "Other duties" for this group included electronic and typewrite maintenance, translation, and traffic indexing.66
The number of officers involved in radio intelligence in the Philippines was even smaller. For the eight years between 1934 and 1941, it was typically limited to two Washington-trained cryptanalysts. To extend their presence in the fleet for as long as possible, they would usually serve two tours: the first as OIC of Station C, followed by the job of radio intelligence officer on the staff of the Asiatic Fleet. In the staff assignment the RI officer worked with an ONI officer trained in the Japanese language who was usually senior in grade, a situation which, under certain operational circumstances, could prove awkward. To forestall any rank-generated problems, an agreement was struck in the Asiatic Fleet which placed final COMINT responsibility on the OIC of Station C.67
|Name||OIC Station C||CINCAF RI Officer|
|Lieutenant Thomas B. Birtly||July 31-Oct 31|
|Lieutenant Bern Anderson||Apr 32-Oct 32|
|Lieutenant Joseph N. Wenger||Oct 32-Jun 34|
|Lieutenant JG Thomas A. Huckins||Aug 34-Nov 35|
|Lieutenant JG E.S.L. Goodwin||Jun 34-Nov 35||Nov 35-Mar 37|
|Lieutenant JG Roy S. Lamb||Nov 35-unk||unk-Dec 37|
|Lieutenant Jack S. Holtwick||unk-Nov 37||Dec 37-Mar 39|
|Lieutenant J.A. Williams||Nov 37-Feb 39||Mar 39-Jan 40|
|Lieutenant Jefferson R. Dennis||Feb 39-Jan 40||Jan 40-unk|
|Lieutenant Bernard F. Roeder||Jan 40-unk||unk-Oct 41|
|Lieutenant Rudolph J. Fabian||unk-Feb 42|
|Lieutenant J.M. Leitwiler||Feb 42-Apr 42|
The method of selecting an officer as a trainee in cryptanalysis was slightly different though colored with the same sort of influences and priorities found among enlisted men. Between 1920 and 1940, a career as a naval line officer (e.g., gunnery officer) in the fleet was the prime pathway to success for Academy graduates. naturally, the fleet had first priority on any and every line officer. Normally, and Academy graduate spent his first seven years at sea, two years ashore, then three years at sea.68 Tours in gun turrets on battleships and cruisers usually led to more responsible positions on board ship (such as navigator) and eventually a command, particularly after a two-year shore assignment at the postgraduate school in Annapolis. on the other hand, tours in intelligence or in radio intelligence, often given to people who had failed to obtain the assignment of choice, were viewed as dead-end assignments leading to poor efficiency reports upon reassignment to the fleet.69 This situation had a noticeably chilling effect on career decisions for officers, and by 1936 had come to the attention of Admiral William H. Standley, CNO. Standley, one of the few officers of his time aware of the importance of radio intelligence, advised his personnel chief that some action must be taken to eliminate the stigma of such assignments.70 Nevertheless, it is clear that, with few exceptions, both officers and enlisted men preferred almost any other assignment when faced with the prospect of an assignment in radio intelligence.71 Although communications officers did enter the COMINT filed, and some, such as Wenger and John R. Redman, did eventually become admirals, most of the officers who led the Navy's COMINT effort in 1941 were either reservists or line officers who had willingly or unwillingly given up their chance for flag rank to serve in the obscurity of radio intelligence assignments.72 Moreover, until Safford's appointment as head of OP-20-G in 1936, the Code and Signal Section had not had a permanent, full-time chief through there is little doubt that Safford retained control while on detached sea duty. in an attempt to fill the officer-cryptanalyst quotas spelled out in war plans at theater level, the Navy conducted an elementary cryptanalysis training program for reservists in several naval districts between 1934 and 1`939. The numbers over the five-year period, however, amounted to less that 119 throughout the entire Navy.73
Board. From its inception the board concerned itself with prospects of war with Japan, particularly after Japan emerged victorious from the Russo-Japanese war in 1904-05. A fundamental assumption by the board was that the Philippines would always be Japan's first wartime objective.
Adopting a series of colors to identify its plans, the board developed the first Japanese war plans (ORANGE) in 1904-05. The usual pattern was for the joint plan to be augmented by individual service plans which were constantly reviewed and refined each year depending on military necessity, the moods of Congress, and the international situation. As the plans grew in complexity, the service plans themselves were augmented by individual service plans such as Naval Communications and Naval Intelligence.
The version of the Navy's ORANGE War Plan which was current in 1941 actually had its inception in May 1929 as WPL13. Changed eight times in ten years, ORANGE number six in May 1937 brought the Navy's plan into line with the Joint Army-Navy Basic War Plan ORANGE. From the point of view of OP-20-G, this change was unique since, for the first time in section 7, chapter II, part I (The Strategic Plan), ONI was tasked to plan for the "collection, evaluation, and dissemination of all information of military and economic value." This language appears to have inspired extensive revisions to the Communications Service plan, Appendix 4 of which pertained to COMINT.
Addressing issues such as wartime organization and subordination of communications intelligence within the naval establishment, the responsibility for dissemination of COMINT, the subordination of translators attached to COMINT activities, and U.S.-U.K. relationships, the DNC plan now provided for COMINT as an integrated service under the Chief of Naval Operations. The provisions of Appendix 4 also provided for dissemination of COMINT by the Naval Intelligence Service and, not surprisingly, stated that the COMINT organization "worked for the DNI under the DNC."74 Organizationally, the 1937 plan more than reinforced the decentralized wartime operation visualized originally by DNC planners. Appendix 4 was revised again in 1939, and the COMINT function was somewhat streamlined by the proposal to drop the COMSEC function of OP-20-G into another appendix. Accordingly, a revised Appendix 3 for COMSEC was completed in 1939, although it would be 1942 before OP-20-G was actually relieved of COMSEC responsibilities.
After the war plan review of 1937, OP-20-G commissioned a study of its current posture.75 This study was probably intended to measure the current COMINT organization against the needs of the war plan. The results reflected a work force quite inadequate to the tasks as outlined in the war plan. Instead of a required seven "intercept nets," openly five were found to exist and a total of only eighty-seven radiomen served the intercept and direction finding function of these nets.76 To be faced with a totally inadequate collection structure was one thing, but OP-20-G found it necessary immediately to enhance its commitment to a research effort. Characterizing it as the "nucleus" or a wartime organization, OP-20-G proposed to enlarge its current research manpower (i.e., purely cryptanalyst) authorizations worldwide to a total of forty-three.77
The resource review acknowledged that the foremost operational problems facing OP-20-G were manpower shortages, an expanding and increasingly complex Japanese cryptographic environment, and the resulting cryptanalytic backlogs which continued to engulf the small work force. All of these problems focused on research manpower. Having recognized the susceptibility of Japanese cryptography to machine exploitation, an immediate solution was to recommend the installation of IBM tabulating equipment in all research units as rapidly as funds would permit. The idea was that this equipment would enable fewer people to do more work. By equipping these units with the latest IBM equipment in peacetime and developing other machines to meet improvements in Japanese cryptography, OP-20-G believed that necessary cryptanalytic techniques could be developed and the people properly trained before hostilities began. Theoretically, no delays would occur after war began in exploiting the Japanese cryptographic systems, at least not through lack of equipment.
It is both interesting and instructive to follow the vicissitudes of War Plan ORANGE (WPL13) from 1937 to 1941 because of the bearing they may have had on the resource decisions made concurrently in DNC and OP-20-G and because they will provide a revealing insight into the events at Pearl harbor on 7 December 1941.
A familiar and fundamental feature of WPL13 in 1937 was a U.S. Navy offensive into the western Pacific from Pearl Harbor. The initial objective of this operations was to either relieve its defenders or recapture Manila Bay. Although the Army thought the offensive aspects of this ORANGE plan in 1937 were "an act of madness," they could not argue that Manila Bay was the best and possibly the only base from which to conduct future offensive operations in support of other U.S. interests in the Far East. Here was an obvious area for future compromises.
The Navy Base War Plan ORANGE for 1938 contained three new assumptions inspired by extensive Army revisions to the Joint Plan, which eliminated all references to offensive warfare: (1) outbreak of war would be preceded by a period of strained relations; (2) ORANGE would attack without warning; and 93) a superior U.S. fleet would operate west of Hawaii.
The eighth and final change to WPL13 was made in March 1939. This change reflected the initial shift in U.S. strategic thinking from the Pacific to events in Europe and the Atlantic Ocean, away from offensive operations toward a concept of defensive operations and readiness. At the same time a new planning system replaced the colors adopted over thirty years before with the RAINBOW Plans described briefly as follows:
[HyperWar editor's note: RAINBOW did not "replace" the color plans, so much as merge their one-on-one scenarios into a more realistic set of plans for coalition warfare -- the U.S. with allies against the various colors representing the Axis powers; hence the "rainbow" reference. The portion of RAINBOW 5 (which went into effect with the U.S. entry into the war after Pearl Harbor) dealing with the war against Japan adopted ORANGE virtually intact.]
RAINBOW 1 (WPL42): Limited action in order to prevent a violation of the Monroe Doctrine as far south as 10 degrees south latitude. This plan was approved by the Secretaries of War and Navy on 14 August 1939.
RAINBOW 2: RAINBOW 1 in first priority followed by concerted action by the United States, Great Britain, and France against the Fascist powers. U.S. forces responsible solely for the Pacific.
RAINBOW 3 (WPL44): RAINBOW 1 in first priority followed by projecting American forces into the western Pacific.
RAINBOW 5 (WPL46): RAINBOW 1 in first priority followed by U.S. armed forces into east Atlantic or Europe and Africa in concert with Great Britain and France. (Modified to conform to the course of the war in Europe during 1940 until December 1941.)
Planning for WPL13 appears to have continued during 1919 and into early 1940. Attempting to add realism, planners in September 1939 assumed that Japan would dominate the Asian coast and adjacent waters as far south as Indochina. They rejected a hypothesis that Japan already controlled the Netherlands East Indies and was poised to take over Singapore and the Philippines. The planners also considered a third alternative -- than Japan had not yet moved southward from Formosa -- since the central issue was at what point the United States would intervene. The planners rejected this alternative because they could not decide whether it would necessitate intervention and were not certain that the American people would support such preventive measures as early movement of the U.S. Fleet to the Philippines, to the East Indies, or to Singapore.
RAINBOW 2 was the final Pacific-first strategic plan. It was never adopted by the Joint Board or published. Beginning in early 1940, the entire focus of American strategy changed following Germany's easy victories in Norway and Denmark. The shift in focus was signaled by a letter from the joint Planning Committee to the Joint Board on 9 April 1940, recommending that planning begin immediately under RAINBOW 5, leaving RAINBOW 3 and 4 in skeletal form. With this letter, Pacific-first strategic thought and planning was virtually at an end. The fall of France in June 1940 and the subsequent Battle of Britain raised serious questions about the security of the United States itself, whether or not the British Isles would fall as France had, and the fate of the British Navy. Suddenly, the fate of England and control of the Atlantic Ocean were the most vital planning issues in American policy.
The brief but interesting evolution of RAINBOW 5 from being one among equals to the preeminent U.S. war plan is also instructive. it not only involves the final stages of the other four plans, but its details, too, lend insight to the events of December 1941.
After the fall of France in June 1940, General George C. Marshall, Army Chief of Staff, and Admiral Harold R. Stark, Chief of Naval Operations, submitted to President Roosevelt a draft entitled "Bases for Immediate Decisions Concerning National Defense.: As amended after the president's views were obtained, it became on 27 June 1940 a plan for national defense. Its six provisions were as follows:
Although planning for war with Japan was not extinct, the end was now near. on 25 September 1940, a memorandum prepared by Army planners for their boss, Major General George V. Strong, examined U.S. prospects in the event of a British defeat in the Atlantic in the context of the American commitment in the Pacific (i.e., RAINBOW 3 vs RAINBOW 4) and concluded that they were incompatible policies. Army planners went one step further and warned against a more active policy of pressure toward Japan. They recommended rapid U.S. rearmament, aid to Great Britain, refraining from antagonizing Japan, remaining on defensive in the Pacific, and finally, moving to ensure the security of the western Atlantic.
In a similar study two month later, Navy war planners under Captain Richmond k. Turner discovered that realistic Pacific operations under RAINBOW 3 would be impossible if the naval detachment required under RAINBOW 4 were transferred to the Atlantic. With the forces available, they reported, the U.S. Navy could operate in only one theater. This discovery led Admiral Stark to write his famous "Plan Dog" memorandum to Secretary Knox on 12 November 1940. The ideas contained in his memorandum had not changed significantly between June and November, although they did reflect some of General Strong's thoughts from September. his conclusion, however, was remarkable: the United States might "do little more in the Pacific than remain on a strict defensive." Clearly, the first U.S. priority was to the British war effort and to prevent the war in Europe from spreading to the Western Hemisphere. Still it is startling to see the Chief of Naval Operations, in the fall of 1940, advocating a policy of avoiding even a limited war with Japan after over thirty years of planning for an unlimited offensive war. The only concession Stark would make was to leave the U.S. fleet at Pearl Harbor because of the U.S. diplomatic commitments in the Far East. His firmness in this purpose was to cost Admiral James O. Richardson, Commander in Chief, U.S. Fleet, his job.79
Cause and effect relationships are often difficult to establish, particularly in resource decisions in Naval Communications and OP-20-G. Frequently one must work backwards from the end result. Using this method, it is clear that manpower resources in OP-20-G were adversely affected by the CNO shift in policy and planning from Pacific-first to Germany-first. Chart B graphically displays the fact that in December 1941 over 60 percent of all COMINT manpower had been concentrated in Washington where the only current mission was Japanese diplomatic and Atlantic DF. It is also clear in Chart B that two thirds of the officer cryptanalysts available to OP-20-G were also assigned to discussions of Japanese naval systems, it was noted that the year before only from two to five people could be spared to work on JN-25. There is moreover, ample reason to believe that emphasis had not changed materially by December 1941. Given the new Germany[-first policy, how interested were naval decision makers in OP-20-G, War Plans, and Operations in the Pacific-related intelligence being pumped out by Pearl Harbor, Corregidor, and OP-20-G? These units were individually and collectively flooding the desks and in-baskets of Navy officials in Washington with alarming reports of Japanese war preparations, some of which must have read like cribs from a RAINBOW 2 planner's wastebasket. yet, as became painfully clear throughout 1941, only readable messages were bankable, and only Japanese diplomatic messages were being deposited. On 11 June 10941, Admiral Stark formally canceled WPL13.79
|Distribution of Navy Comint Personnel|
|Intercept Stations/DF Control||178||72||42||--||292|
|Outlying DF Stations||60||84||8||--||152|
War games were introduced into the U.S. Navy in the late nineteenth century by Lieutenant William McCarty Little, a member of the staff of the Naval War College. As a result, early in the twentieth century, the War College virtually backed into a war planning relationship with the Navy's General Board. The board was to designate a country for which a Naval War Portfolio was to be prepared; ONI would provide the necessary information, and the board would prepare a plan aided by the staff of the War College. In fact, the College did most of the work.
Before World War I, students at the War College raised the issue of planning for wars without guidance from the political establishment. They felt that trying to plan without the answers to such questions as, What are the intentions of the United States in China, Japan, the Monroe Doctrine? was an exercise in futility. Their questions found willing listeners on the board, at ONI, and even in the Secretary of the Navy. The Navy proposed various remedies to both the legislative and executive branches based on the premise that "plans not based on the interrelation of the enemy's and our own motives are of little value" -- but to no avail.
58. Interview Frank B. Rowlett, 26 November 1974, by Vincent Wilson, Henry Schorreck, David Goodman and Earl Coates, OH-01-74, NSA (classified).
60. Friedman monograph Preliminary Historical Report on Solution of the B Machine, 14 October 1940, Series IV.I.2.28, CCH History Collection (classified).
61. SRH-305, RG 457, NA.
62. Part of OP-20-G's workload is described in SRH-149, RG 457, NA and SRH-305, RG 457, NA. See also the history of OP-20-GY, Series IV.W.I.5.10, CCH History Collection (classified.)
63. History of Signal Security Agency, Vol. III (classified), contains a copy of appropriate portions of the act, Series III.hh. See also W.J. Holmes, Double-Edged Secrets (New York; Berkley Books, 1981), 13-14, which states that this act prohibited a commercial radio company in Honolulu in early 1941 from sharing daily Japanese ship position data inherent in commercial weather messages with COM-14's intelligence collection effort.
64. Prange, Verdict of History, 7.
65. H.P. Willmott, Empires in the Balance (Annapolis, MD: Naval Institute Press, 1982), 30-51. The actual ratios: battleships and heavy cruisers, 5.5.3; light cruisers and destroyers, 10.10.7; and submarines, 1.1.1. See also Prange, Verdict of History, 7. By standing aloof from the League of nations, the U.S. had rejected collective security, yet it had refused to provide for itself an adequate unilateral national defense system. Prange also asserts a weak argument supported by several quotations that blame for Pearl harbor actually lay with the American people who, because of complacency, idealism and a false sense of security based on racial arguments, failed to insist that Congress provide for adequate military strength after World War I.
66. "U.S. Naval pre-World War II Radio Intelligence Activities in the Philippine Islands, 1931-1942," SRH-180, RG 457, NA.
67. SRH-355, 421, and Appendix VII, 289-90. The linguist in at least one instance was also head of the Diplomatic section in Station C.
68. Ibid., 311. RG 457, NA.
69. Interview, Thomas H. Dyer, Captain USN (Ret) January-March 1982, by Robert Farley, OH-01-82, CCH, and NSA interview Rudolph T. Fabian, Captain, USN (Ret) 4 May 1983 by Robert Farley OH-09-83, CCH (classified). Hereafter Dyer interview, Fabian interview.
70. SRH-355, 167-168, RG 457, NA.
71. Dyer interview.
72. Dyer interview; interview Wesley A. Wright, Captain, USN(Ret), 24 May 1982, by Robert Farley and Henry Schorreck, OH-11-82, CCH; interview Edwin T. Layton, February 1983, by Robert Farley, OH-02-83, CCH.
73. SRH-355, 330-331, RG 457, NA.
74. Ibid., 218-222, RG 457, NA.
75. Military Study, SRH-151, RG 457, NA.
76. Asiatic (1), Mid Pacific (2), West Coast (3), Cavite-18*, Heeia-17*, Astoria-7, Guam-10*, Shanghai-7, Augusta-1, (*: includes 2 for HFDF). East Coast-U.S. Fleet 94), East-West Coast DF (5), Winter Harbor-7, CINCUS-10, Navy Dept.-2, Students-8, Total 87. "A Military Study of the Radio Intelligence Organization," June 1937, Series IV.W.X.11, CCH History Collection (classified).
77. Cavite, a minor research center, 6; Flagship Asiatic Fleet, an advanced mobile unit, 3; Flagship U.S. Fleet, a mobile unit, 4; Pearl Harbor, a major research nit, 5; and Washington, a major research unit, 25, including administrative personnel. Ibid.
78. On 3 July 1940, the COMINT unit aboard the USS Trenton was transferred to the USS Omaha in Squadron 40T and began covering Italian naval circuits in the Mediterranean. Naval battles between Italian and British forces during July and August and the attack on Rhodes in September 1940 were reported by this unit. SRH-355, 372, RG 457, NA.
79. The material in this section concerning U.S. strategic planning is based on several sources: "War Plan Orange; Evolution of Strategy," Lewis Morton, World Politics, January 1959, 221-250; "Naval Contingency War Plans 1891-1945," Scholarly Resources, Wilmington Delaware, 1979; "The U.S. Navy: Strategy, Defense and Foreign Policy 1932-1941," Michael Kedian Doyle, Ph.D. dissertation, University of Washington, 1977, University Microfilm International, Ann Arbor, MI, and London; "The U.S. Navy and War Plan Orange 1933-1940: Making Necessity a Virtue," Michael K. Doyle, Naval War College Review, May-June 1980; and "Strategic Planning for Coalition Warfare 1941-1942", Maurice Matloff and Edwin M. Snell, Office of the Chief of Military History, Dept. of the Army, Washington, D.C., 1953.
80. "Professors of War: The Naval War College and the Development of the Naval professions," Ronald Spector (Newport, RI: Naval War College Press, 1977), 71-2, 108-111.