No review of the Navy's COMINT contribution to U.S. knowledge of Japanese pre-Pearl Harbor intentions would be complete without citing the benefits U.S. officials derived from the messages exchanged by Japanese diplomats in Washington and Tokyo. Although the credit for initial U.S. success against Japanese diplomatic machine systems must go to Army cryptanalysts, the Navy did play a significant role in providing collection and, after October 1940, in devoting the bulk of its cryptanalytic and linguistic resources to the exploitation effort. Unfortunately, as Safford had foreseen, the small Navy cryptanalysis effort in Washington was almost overwhelmed by the volume of messages from this source.133 Little time and fewer resources were left over to attack JN-25, the code which, if read, would have provided operational details concerning the Japanese Strike Force.
As soon as possible after the Purple machines became available to Army and Navy cryptanalysts, the English texts of all translated diplomatic messages were delivered to both the Office of Naval Intelligence and the Military Intelligence Service (MIS) each day. By agreement, OP-20-GZ was responsible for dissemination of these messages within the Navy (Secretary of the Navy Frank Knox; CNO Admiral Harold R. Stark; A/CNO Admiral Royal E. Ingersoll; ONI, Admiral Theodore S. Wilkinson; and Chief, War Plans Division) and to the White House for the President's Naval Aide, Captain John R. Beardall.134 Alwin D. Kramer and Arthur H. McCollum of the ONI Far East Desk decided what translations U.S. policymakers would see each day.135 This arrangement was consistent with the dissemination rules laid down in the 1937 Orange War Plan. Similarly, MIS was responsible for dissemination within the War Department (Secretary of War Henry Stimson; Chief of Staff George C. Marshall; and Chief, War Plans Division) and to the State Department (Secretary of State Cordell Hull).
In Hawaii, neither the Army nor the Navy commander had facilities for decoding Japanese diplomatic messages. Overall policy regarding dissemination of Japanese intercept by both G-2 and ONI dictated that MAGIC material based on diplomatic messages would not ordinarily be distributed to any commander outside Washington. The
primary reasons for this policy were to protect the source and to retain in Washington the evaluation of purely diplomatic material. There was, however, no rule in either the War or Navy departments which prevented dissemination of MAGIC information to theater commanders. Facilities for decoding Japanese diplomatic messages, including messages in the Purple system (MAGIC), were available to Station C in the Philippines. However, if any diplomatic messages were read and translated there, it is possible that in the Asiatic Fleet diplomatic messages were not considered by themselves to be a likely source of either strategic or tactical warning.
Warnings based, at least in part, on the contents of Japanese diplomatic messages were in fact sent to the Hawaiian and Philippine commands on at least three occasions, 24, 27 and 29 November 1941.136 It seems clear, however, that after July 1941, as a matter of policy and probably as a practical security precaution, no intelligence material directly from MAGIC was sent to Admiral Kimmel in Hawaii or to U.S. command officials in the Philippines.137
Receiving the actual Japanese diplomatic messages would have done neither Kimmel nor Hart any practical service, aside from their obvious value in pinpointing vital areas of Japanese policies and intentions. They contained no Japanese naval or military information. Messages between Tokyo and Washington largely concerned the ongoing negotiations between Secretary of State Cordell Hull, Japanese Ambassador Kichisaburo Nomura, Minister Reijiro Wakasugi and, later, Saburo Kurusu, Japanese Ambassador Extraordinary. Messages between Tokyo, Washington, and other diplomatic posts frequently concerned Japanese espionage tasks and the efforts of diplomats to obtain information concerning U.S. naval and air dispositions in Panama, Hawaii, Manila, and various locations on the U.S. West Coast. Collectively, these messages conveyed an alarming interest in major fleet activity and an unmistakably hostile intention toward the United States. Their tenor deteriorated sharply after 26 November when the U.S. delivered its ten-point response to the Japanese note of 20 November. They did not, however, disclose the movements of the Japanese fleet.138 Only the unread, untranslated Japanese naval messages held this vital information.139
Despite the fact that all messages in Japanese diplomatic channels were not available by 7 December and that the daily reports mailed from Hawaii and Corregidor were at least two weeks en route, by late November 1941 U.S. Navy officials in Washington, Pearl Harbor, and Manila well knew that war with Japan was imminent.140 Made aware of hostile Japanese intentions by a profusion of intelligence (most of it COMINT), Admiral Stark, CNO, after 23 November 1941, repeatedly warned his Pacific commanders of impending Japanese attacks, placed restrictions on ship movements, and probably approved DNC's orders to destroy codes. The weight of evidence overwhelmingly favored Japanese air and naval strikes against the Philippines, and this locale actually
appeared in the warning messages of 24 and 27 November as one of several likely Japanese objectives.141
Inexplicably, the warnings issued by Washington were virtually the only direct military actions taken which can be traced directly to COMINT despite the sense of urgency that COMINT reflected. Admiral Kimmel in Hawaii, though a recipient of ample warning on the approaching crisis, was not particularly alarmed by COM-14's reports. Lacking any information on the Japanese 1st Air Fleet, and except for the warning messages, unaware of the content of messages in Japanese diplomatic channels, his attention was focused on the western Pacific.
With regard to the Philippines, the sense of alarm, at least in U.S. Navy circles, was paramount. By mid-September 1941, Admiral Thomas C. Hart, Commander in Chief Asiatic Fleet since July 1939,142 had become very concerned over the intelligence reports on Japanese naval activities supplied by COMINT and other sources. By November, Hart clearly saw, through his regular visits to Station C in the Malinta Tunnel on Corregidor, that his fleet would soon be at war and that "time was running out."143 Surviving records, however, do not provide a clear connection between a COMINT cause and an operational effect in the Asiatic Fleet.
Because of the general disagreement which prevailed among U.S. officials in Washington and the Pacific over U.S. objectives in the area, particularly with respect to British and Dutch possessions and the defense of the Philippines, Hart dispatched his fleet on a series of strategic deployments during September and October 1941 which first removed and then returned his forces to the Manila area. In September, convinced that the U.S. government would not defend the Philippines, Hart sent all his surface vessels to the south. On 7 October 1941, sensing a change in policy, Hart proposed to Washington that he join General Douglas MacArthur in defending the Philippines, and without first obtaining Washington's approval, he brought the Asiatic Fleet back to Manila. On 20 November 1941, Washington disapproved his plan. This forced Hart, at virtually the eleventh hour, to redeploy his surface vessels to southern Philippine and Netherlands East Indies ports. Patrol aircraft and submarines were retained in the Manila area. Under these circumstances, it is difficult to distinguish Hart's deployments in response to intelligence from those taken in response to Washington.
Lacking a declaration of war by the United States against Japan and keenly aware that the United States did not wish to appear to be the aggressor, Hart, since receiving the first war warning from OPNAV on 24 November 1941, had been "edging toward increasingly risky action."144 Based on "intelligence intercepts," Hart authorized air patrols over Japanese convoy movements along the China and French Indochina coasts. On 6 December 1941, after receiving confirmation that a Japanese amphibious force was steaming across the Gulf of Siam, Hart ordered one of his destroyer divisions to sail west
from Balikpapan, Borneo, to Singapore. There the division commander was to place his ships under the British Fleet commander.145
Admiral Hart in his personal diary for the period 1-30 November indicated a preoccupation with two major problems: coordination of his own plans for defense of naval shore facilities with the U.S. air forces in the Philippines and Japanese troop movements along the China and Indochina coast. He estimated that the information provided him represented primarily a threat to Thailand and to the British in Indochina. He believed he had taken all prudent measures in anticipation of an attack, although his diary referred to no specific actions taken in response to the warning messages he received on 24, 27, and 29 November. On 7 December he wrote in his diary, "Guess war is just around the corner, but I think I'll go to a movie." The entry for 8 December states, accordingly, "It [the attack] was no surprise by a matter of 18 hours."146
Despite these rather tenuous indications of Hart's responses, there is plenty of evidence from another source that Hart did in fact react aggressively to Japanese activities. Japanese radio intelligence messages from Taiwan between 17 November and 3 December 1941, which were not read and translated until after the war in 1945-46, contained many reports of U.S. Army air and U.S. Navy air and ship reconnaissance. In addition, the Japanese consulate at Manila was very active in reporting the arrivals and departures of submarines and surface vessels.147
The Japanese perception of Army B-17 and fighter activity in the vicinity of Manila was one of declining activity. On 2 December 1941, Taiwan reported U.S. Army planes as "extremely inactive recently," and on 3 December as "greatly reduced since 30 November. ... Prior to 30 November, 10 or more planes per day heard; on the 30th, 1; on the 31st, 2; none on the 3rd."148
U.S. Navy reconnaissance of the airspace around Luzon increased during November-December, according to Taiwan on 3 December 1941. Taiwan reported the area patrolled by aircraft on 2 December as "300 miles south and southeast of Manila and west of northern Luzon."149 In addition, the U.S. Navy was also active in surface patrols in the vicinity of Taiwan, according to the Japanese messages.150 In one of his many reports of visual observations, the Japanese consul in Manila, on 3 December 1941, reported departures from Manila of possibly seven submarines and, from Cavite, the departure of the light cruiser Houston, all to unknown destinations.
There is little doubt on balance that COMINT from Station C contributed to the Hart decision-making process. Certainly COMINT can claim a major share of the credit for the fact that on 8 December 1941 Asiatic Fleet losses were minimal, two amphibious patrol aircraft (PBY) and the gunboat Wake. Little if any of the COMINT provided by Station C came from cryptanalysis. Because Washington could not supply current code group meanings, Station C was not able to read messages in the Fleet General-Purpose System, JN-25, or in several of the minor naval codes.151
COMINT After the Opening Attack
Following the attack on Pearl Harbor, COMINT resources in Hawaii, after a two-day pause probably caused by the loss of contact with Heeia, resumed publishing a daily COMINT summary. This summary continued to follow the activities of the Japanese Fleet. Corregidor concentrated its efforts on supporting local Navy and Army commanders by providing warning of incoming air strikes as well as supplying both CINCAF and CINCPAC information on the whereabouts of the Japanese Fleet.
At Pearl Harbor cryptanalytic emphasis shifted by mid-December from the Japanese Flag Officers Code and the various shipping codes to recovering and exploiting the Japanese Fleet General-Purpose System. This was part of Safford's regrouping of tasks and responsibilities between Washington and Hawaii after the events of 7 December revealed with painful clarity the type of information probably contained in the Japanese Navy's messages and after the volume of diplomatic material declined. OP-20-G also recognized the tenuousness of its position in the Philippines and quickly put in motion plans to salvage Station C's manpower. Some of the senior officers in the communication directorate took this opportunity to centralize control of the entire COMINT operation in Washington.152
Throughout its relatively short life, OP-20-G, both in Washington and in the Pacific, had suffered a lack of manpower. In the final months of 1941, the lack of overall manpower resources combined with the disposition of the available cryptanalysts, resulted in the failure to read the critical messages of the Japanese Strike Force targeted for Pearl Harbor. Briefly recapped, two thirds (fifty-three) of the officer cryptanalysts were in Washington where, if they were assigned to technical positions, they were exploiting Japanese diplomatic messages, operating a twenty-four-hour watch and performing coderoom tasks which included running the Atlantic DF network, and conducting research on Japanese Navy cryptographic systems, e.g., JN-25. Less six officers in transit, the remainder were assigned in unequal proportions to Hawaii (twelve) and Corregidor (nine) where, in both stations, some were diverted to traffic analysis and machine room responsibilities. It may be argued that a more or less even distribution of collectors and DF operators between East Coast and Pacific stations was also a misalignment of critical resources, but it is clear that the placement and occupation of cryptanalytic personnel penalized Japanese Navy targets.
It is clear too that, between September and mid-November 1941, the activities of the Japanese Navy, as it prepared for war, were also overlooked by official Washington while it followed every fluctuation in the diplomatic negotiations between Ambassador Nomura and Secretary of State Hull. Judging from the conflicting guidance given to CINC Asiatic Fleet, abundant warning information produced by Hawaii and Corregidor from their analysis of the Japanese Navy's communications system and the activity it reflected was
apparently either ignored altogether or treated as unsubstantiated rumor lacking any supporting evidence from readable messages. It is certain, however, that COMINT based on Japanese Navy communications available to these officials did not indicate that the Japanese intended to attack Pearl Harbor.
In an attempt to place some restraints in the path of the Japanese government, U.S. military leaders, with the approval of President Roosevelt, agreed with the British to establish separate commands for the Philippines and the Southwest Pacific; the former under Admiral Thomas Hart, CINC Asiatic Fleet, the latter under the direction of the senior British officer. The principal U.S. goals in the western Pacific at this time were to avoid being drawn into the British plans to defend Singapore and to avoid antagonizing the Japanese government. The U.S. claimed that these arrangements, along with the presence of the U.S. Pacific Fleet at Pearl Harbor, were sufficient deterrence but without representing a belligerent act.
Viewed in retrospect, these circumstances very strongly suggest that both OP-20-G and the Chief of Naval Operations had been swept along by the same overpowering pull of events in Europe and the Atlantic and were confused by conflicting American and European objectives in the Far East. Because of the unexpected German successes in 1940 and early 1941, the entire U.S. military establishment was confronted by an abrupt shift in both political and military priorities which, in January 1941, and became partially institutionalized by the first American-British Conference (ABC-1). This conference established the primacy of the Atlantic theater over all war planning. The decision to put the defeat of Germany in first priority ultimately led in June 1941 to cancellation of all war planning of any consequence in the western Pacific and placed naval emphasis on the western Atlantic and in the Pacific east of 180 degrees.
We will probably never know precisely why OP-20-G arranged its manpower resources in Washington and the Pacific as they were when the Japanese attacked Pearl Harbor and Manila. Undoubtedly, the new policies had left OP-20-G in an awkward position, creating new problems and aggravating old ones. Unwilling and perhaps unable to dismantle the COMINT edifice in the Pacific it had worked for twenty years to build, OP-20-G throughout 1941 let Hawaii and Corregidor perform the functions for which they had been prepared and trained while the work force in Washington did its best to provide support in both theaters as well as to abide by its odd-even agreement with the Army concerning Japanese diplomatic messages. The historical manpower problems generated at least in part by a lack of respect for the intelligence profession were probably felt most keenly in Washington, where a decentralized management philosophy had been unable to prevent a concentration of more work load than the work force could possibly achieve. The Washington center's limited cryptanalytic resources were not, moreover, focused on the Japanese Navy where they belonged. As a result, they could not read any of the Japanese Navy cryptographic systems, and they became preoccupied with the Japanese diplomatic targets, which were providing unprecedented exposure for Army and Navy COMINT centers at the highest levels of government. Without minimizing the influence of strong interservice rivalry, the fact that OP-20-G was not concentrating its resources on Japanese
Navy targets may suggest instead that a transition in emphasis toward the Atlantic by assigning more and more people to the twenty-four-hour watch had in fact already begun when the Japanese attack came.
132. See Appendix C, where all the OPNAV warning messages of November 1941 are shown in the specific context of the COMINT summaries from Hawaii and indirectly from Corregidor. See also PHA, part 10, 4834, in which Layton asserts his conviction that the two messages in fact stimulated the war warning from CNO on 29 November.
133. See Appendix B. In preparing Appendix B, I have selected certain Japanese diplomatic messages intercepted and translated between July and 6 December 1941. The messages were usually seen by authorized recipients on the day they were translated. Army and Navy translations are shown in juxtaposition with the warning messages issued by the Chief of Naval Operations using the heading OPNAV.
134. Roberta Wohlstetter, Pearl Harbor, Warning and Decision (Stanford, CA: Stanford University Press, 1962).
135. Wohlstetter, 176-86.
138. Connorton, Vol. I.
140. James Leutze, A Different Kind of Victory (Annapolis, MD: Naval Institute Press, 1981), 198-230.
142. Morison, Vol. III, 151.
143. Leutze, A Different Kind of Victory, 226.
144. Thomas C. Hart, Narrative of Events, Asiatic Fleet, a personal diary (Washington, D.C.: Naval Historical Archives).
145. Leutze, A Different Kind of Victory, 226.
146. Hart, Narrative of Events, Asiatic Fleet.
147. SRN-117284, 116772, 117290, RG 457, NA.
148. SRN-116645, 116647, 116729/116730, RG 457, NA.
149. Leutze, A Different Kind of Victory, 222, and Commander Walter Karig and Lieutenant Welbourne Kelley, Battle Report, Pearl Harbor to Coral Sea, (New York: Farra and Rinehart, Inc., 1944), 129-130.
150. SRN-117284, 116772, 117290, RG 457, NA.
151. Tentative chronology/memorandum for the record dated 3 August 1944 from OP-20-G, NMCG, subject: "Assignment of Washington on Various Naval Systems during the Year 1941," found in Rochefort papers, Series IV.W.X.1.a, CCH History Collection (classified).
152. OP-20-G was completely reorganized in February 1942, though not along the lines recommended by Safford. On 5 March 1942, COM-16 was directed by CINC to "evacuate personnel of Radio Intelligence unit as soon as possible."