Storm Over Leyte

by John Prados

  On July 24 IGHQ informed the Southern Area Army, Japan’s operational command for the entire Pacific Ocean area, that it expected to direct a decisive battle against the Allies in the second half of 1944. For the Navy Admiral Shimada laid down broad outlines in one of his last directives as NGS chief. Admiral Oikawa Koshiro, who succeeded him at the head of the NGS, had yet to take up the reins. But Captain Yamamoto’s planners applied the standard to each direction the Allies might take. He had more than a dozen officers in the NGS operations section, so the work could be spread around enough to avoid overwhelming the staff. The project wound its way through the NGS while the politicos lined up to oust Tojo.

  It is striking how closely the Japanese timing matches President Roosevelt’s Pearl Harbor strategy deliberations. The Imperial Navy’s basic directives for the next campaign emerged from the warren of General Staff offices on July 26, the day before FDR convened his top commanders at Pearl Harbor. Some doubt Captain Yamamoto’s strategic good sense, but in fact he’d held this job for two years. For fifteen months before that he had been a section chief under Baron Tomioka Sadatoshi, a renowned staff officer under whom fools did not last, and Yamamoto had done a tour with the NGS operations bureau in the late 1930s. His reputation had been shining enough that, at a time when the Pearl Harbor attack plan remained a closely guarded secret, the air staff officer for Combined Fleet had felt comfortable telling him of the scheme. Yamamoto Chikao, then the skipper of a seaplane carrier, had noticed the absence of the fleet aircraft carriers from the southern invasion lists and had made inquiries. The Navy would promote Yamamoto to flag rank and send him on to lead a unit of aviation ships. Historian Samuel Eliot Morison claims the quality of NGS planning deteriorated under Captain Yamamoto, but the meticulous preparations for the upcoming campaign are evident.

  The Japanese christened their plan “Sho Go,” or “Victory Operation.” There were four Sho contingencies, numbered according to the geographic areas to which they would apply. Sho 1 would take effect if Allied forces invaded the Philippines. Sho 2 provided for the defense of Taiwan, the Ryukyu Islands, and southern Japan. Sho 3 would go into effect if the attack were to be against central Japan. Finally, if northern Japan became the Allies’ target, Sho 4 would be the response. Each plan provided specific countermeasures. The contingencies replicated the careful approach pioneered in Admiral Koga’s Z Plan.

  To grease the wheels of action before giving up his posts, Tojo sponsored the Supreme War Council. That group now decided to go all in on the next battle. As Combined Fleet commander Toyoda Soemu explained it to Allied interviewers after the war, Tokyo reviewed the full range of its resources along with its geographic position. If Japan were to be cut off from the southern area, it would no longer be able to find resources for the Inner Empire. The Supreme Council decided to apply 70 percent of Japan’s resources to the next battle and husband the rest for the struggle for the homeland. IGHQ now began fleshing out the concepts.

  Captain Yamamoto worked on the general schema. The NGS articulated the result in IGHQ Navy Directive No. 431, issued on July 21, as the Tojo government collapsed and the Americans mopped up on Tinian and Guam.

  Despite the Marianas disaster, Yamamoto and his colleagues still thought the operational concept perfectly sound: The Imperial Navy could use land bases to shuttle bomb the Allies and hold an umbrella that might cover its surface fleet. If there were a substantial carrier, or “mobile,” force, they would be even more potent. One reason the tactic had not worked so well at the Battle of the Philippine Sea, Yamamoto noticed, had been the restricted number of airfields on small islands, which made it practical for the Allies to neutralize the base network. With big landmasses like Taiwan, the Philippines, or Japan, that became impossible. There would always be alternate airfields. Under this concept the strength of land-based air forces became even more crucial to the overall mission.

  Admiral Shimada, not yet supplanted from his NGS post, signed the orders. Directive No. 431 provided that the air fleets would be reorganized and massed in the homeland, from where they could sally. The object “of attacking and destroying the enemy fleet and advancing forces” became the goal. Surface naval forces would concentrate their main strength in the southern area, close to Japan’s oil sources. Mobile forces—by which the directive meant carrier units—were lumped together with the surface fleet, but the inclusion of language providing that some units would be stationed in Empire waters indicates that the NGS had already begun grappling with the treatment of aircraft carriers—something that emerged as another headache. The fleet should give special attention to “raiding operations”—mounting surprise attacks—including ones designed to catch the enemy in their advanced bases. The NGS envisioned many kinds of raiding operations. The overall concept, like that of the Z Plan, remained one of “intercepting and destroying the enemy within the sphere of the [land-based] air forces.”

  On July 26 IGHQ Navy Directive No. 435 followed, establishing Sho contingency areas and fleshing out arrangements for air operations. Following the new combined action model (which had not worked in the Marianas, but perhaps had simply been too novel), there would be a unified command—the air fleets to include both Imperial Navy and Japanese Army units. Navy officers would direct when enemy fleets were the targets. Directive No. 435 specified which Army air units would be controlled by Navy air fleets. Captain Yamamoto’s planners had been thorough. The order went so far as to list aircraft types and designate what kinds of enemy forces they should destroy.

  Prior to battle, air strength should be dispersed as much as possible to avoid incidental losses from Allied attacks. When the moment came to hurl the air forces, one element would blast Allied aircraft carriers, while the major strength made both day and night attacks against enemy transport convoys and—again—carriers.

  Laying down the framework accelerated. On August 1, Admiral Toyoda Soemu issued orders providing for cooperation with the Army “to intercept and destroy the invading enemy at sea in a Decisive Battle.” This directive, so far as the Philippines were concerned, envisioned rapid preparation of air bases to accommodate the planes of a re-formed First Air Fleet as well as a newly created Second Air Fleet. Allied carriers were to be sunk by air attack, while warships and planes combined to smash transport convoys. If the Allies’ invasion succeeded, their amphibious shipping would then become the primary target. The surface fleet would sortie within two days of any invasion, and air strikes would begin at least forty-eight hours before warships reached the scene. A separate top secret directive specified a new fleet organization for the battle.

  By late July, Imperial Army commanders were starting to line up behind the option for putting nearly all remaining resources behind a single roll of the dice. At senior levels of Imperial General Headquarters the discussions climaxed the week of August 9, when planners held talks every day.

  Finally everyone agreed, and on August 18 the services jointly sent an IGHQ petition to Emperor Hirohito. An audience with him took place the next day. Apparently unaware of the petition, privy seal Marquis Kido viewed the meeting’s purpose as somewhat vague, with a roundtable discussion and the emperor encouraged to pose questions.

  After the war Hirohito recalled, “I agreed to the showdown battle . . . thinking that if we attacked . . . and America flinched, then we would probably be able to find room to negotiate.” Historian Herbert Bix, a notable Western expert on Hirohito, found this to be an example of the emperor’s destructive influence on military operations, but the truth seems more nuanced. The Japanese Army did not change its plans for defending the Philippines merely because of the decision in Tokyo. Rather, its commitment to send additional forces to the Philippines made it possible for area commanders to expand their inadequate defenses beyond the island of Luzon.

  THE WATCHERS

  The Marianas campaign became an intelligence bonanza for the Allies. For example, with the loss of Sixth Fleet headquarters, the
Imperial Navy’s top submarine command, the Americans captured thousands of documents. By itself Saipan contained the command posts of Nagumo’s Central Pacific Area Fleet, of naval base forces, of the 31st Army, and of the 1st Combined Communications Unit of the Imperial Navy’s radio intercept service. Tinian also contained the nerve center for Admiral Kakuta’s First Air Fleet. All their records were vulnerable. Not only did the fighting go quickly enough that the enemy had very limited time to destroy papers, but Allied intelligence outfits had gained expertise and their commanders had put new mechanisms in place. Intelligence teams accompanied the combat troops and did triage—making on-the-spot assessments of critical documents for immediate translation. The Marianas battles became the first instance where the take of captured documents could be measured by the ton, and the first time captured Japanese soldiers numbered in the thousands.

  A fleet or task force commander—say, Admiral Spruance with the Fifth Fleet in the Marianas, or William F. Halsey, who followed him, in charge of the Third Fleet—had a fleet intelligence officer on his permanent staff. More reporting came directly from Pearl Harbor, or from several sources in Washington, DC, including Army, Navy, and Marine. General MacArthur had his own intelligence organization in Australia and New Guinea, which included radio and photo intelligence shops, its own collection units, and its own interpretation offices, for ground, naval, and air forces. The Australians had a parallel apparatus. The British—in Burma and India—had yet another autonomous spy agency. And the British had a Far East Combined Bureau (FECB) at Colombo as a kind of fusion center. Subordinate units had less elaborate staffs to do the same. In short, a robust intelligence network supported the Allied offensive.

  All agree the most important entity was the Joint Intelligence Center Pacific Ocean Area (JICPOA) at Pearl Harbor. It was a fusion center like FECB, using all-source intelligence, working directly for Admiral Nimitz, commander in chief Pacific Ocean area (CINCPAC). Something like 1,100 people worked for JICPOA. Wags called it “the Zoo.” JICPOA sent the intelligence teams that accompanied invasions, and its cohort, the Fleet Radio Unit Pacific (FRUPAC), supplied mobile radio detachments to commanders, to whom they supplied urgent news. The ensemble had just gone through the bedlam of its latest expansion.

  Descended from the now-famous Station Hypo—the radio spooks at Pearl Harbor who had divined the Japanese plan to attack Midway Island, leading to a remarkable victory—JICPOA had mushroomed, adding sections for estimates, photographic interpretation, document translation, prisoner interrogation, mapping, terrain modeling, and psychological warfare. They worked to intercept actual Imperial Navy messages, in a code the Allies knew as JN-25, which were decrypted to reveal the thinking of the enemy. The intelligence product was called “Ultra.”

  The center outgrew its original Hypo basement offices, moved to a large frame building constructed on Makalapa Hill, then outgrew that too. The Navy had begun the war with a small cadre of radio intelligence specialists—code breakers working the intercepted messages—and radiomen (called the “On the Roof Gang” for the erstwhile location of their radio shack atop Navy Department headquarters in Washington) trained to record the letter-characters the Japanese used in transmission. A little over a month before the Philippine Sea battle, JICPOA moved to a matching headquarters just finished next door, while FRUPAC took over the remainder of the first building save for one suite occupied by the Estimates Section. The radio unit had swollen to roughly the size of the rest of JICPOA, about 30 percent of everyone in Navy radio intelligence at the time. (OP-20-G, the Navy’s code-breaking organization, had 3,722 civilians, officers, and enlisted sailors in February 1944.)

  Messages broken by the code breakers were interpreted by “Japanese language officers,” men carefully trained as Japanese linguists. At the beginning there were just a few dozen of these men, with knowledge of Japanese culture and, usually, experience as naval attachés. War brought huge pressures for more language expertise. The Navy started programs to train language officers quickly at Harvard, Berkeley, and the University of Colorado at Boulder. British authorities did the same thing at the School of Oriental and African Studies in London. In the United States the group trained at Boulder would be the largest by far, and “Boulder Boys” became something of a slang reference to the freshly minted language officers. Beginning in 1943 the stream of Boulder Boys arriving at Pearl Harbor became a steady flow. A number of them roomed in apartments rented from the Honolulu madame Jean O’Hara. By day they translated intercepted radio messages or captured documents, or interpreted during interrogations of Japanese prisoners of war. At night, who knows?

  All streams of intelligence fed into the JICPOA Estimates Section, under Commander Wilfred J. “Jasper” Holmes. In many ways Jasper Holmes was the glue that held together the entire outfit. Invalided out of submarine service for arthritis of the spine, Holmes had worked in naval intelligence since the Hypo days. He had been the apostle for operational intelligence, creating a “combat intelligence unit” that earned kudos in the Solomons campaign, and had then spun it off into the Estimates Section, a unit that supplied weekly and monthly appreciations of Japanese strength on the land, at sea, and in the air. It maintained an all-source running plot of the locations of Japanese warships, and a watch section responsible for urgent warnings of enemy action.

  After the attack on Pearl Harbor, then-Station Hypo had soon identified exactly which Japanese warships had done the dastardly deed. Holmes and others made a huge poster adorned with silhouettes of each archenemy vessel. The poster moved with him to the wall of his new Estimates Section, where it appeared like the FBI “Most Wanted” list. Holmes confessed to “unprofessional vindictive satisfaction” each time he crossed out one of the culprits, and he had a standing offer of a bottle of Scotch for any submarine skipper who sank a Pearl Harbor enemy. He had made good with every sinking until the Marianas—where Commander Herman J. Kossler and his Cavalla dispatched the Shokaku, the fifth of six Imperial Navy aircraft carriers that had participated. The sub skipper’s superior never presented Kossler with his reward, and years later it still pained Holmes that he had not made good on his promise, especially since it turned out Shokaku would be the only one sunk by submarine.

  As a trained engineer, Jasper Holmes’s solutions were practical ones, and despite the exponential growth of U.S. intelligence in the Pacific he kept the Estimates Section on an even keel. They learned from Japanese prisoners, scanned captured documents for important details, and incorporated photo analysis, not to mention the decrypted text of messages supplied by FRUPAC. They added sections when work in subject areas became extensive—an air section, an “enemy land” section, another for “enemy bases,” a mine warfare section, and so on. He corresponded with counterparts elsewhere, cheerfully chiding a colleague at Navy headquarters that the Estimates Section would be happy to copy any great idea Washington came up with, or to host visiting staff for temporary duty at its office.

  Captain Edwin T. Layton, the redoubtable fleet intelligence officer for CINCPAC Nimitz, had maintained a file of biographic information on senior Japanese naval officers. Pressed with other business, he’d neglected it, and finally turned the file over to Lieutenant John Harrison, one of the language officers in the Estimates Section. Harrison soon endowed JICPOA with the best collection of Japanese naval biography outside Tokyo. For example, the Japanese command turned over again several months before the Marianas battle when Combined Fleet leaders fled Palau ahead of a U.S. carrier raid but the planes carrying the officers crashed amid storms. Fleet commander Koga Mineichi perished. His successor became Admiral Toyoda Soemu. That puzzled a lot of Americans, since Toyoda seemed virtually unknown. Looking into his files, Lieutenant Harrison found a photo of the Japanese admiral and could produce a short biography, so JICPOA put out an intelligence bulletin on changes in the enemy high command. This became the first time most Americans had heard of Toyoda.

  Jasper Holmes, soon promote
d captain, wore two hats. He worked as deputy to General Joseph J. Twitty, an Army brigadier who had been among that service’s Japanese language officers, but also as officer in charge of the Estimates Section, which swelled to thirty-two officers and a dozen enlisted personnel, including three women. Their reports went to Admiral Nimitz at CINCPAC and to other commands. Eddie Layton functioned as Nimitz’s indispensable aide on all things intelligence. When Holmes was away from the Estimates Section, Commander Donald M. “Mac” Showers ably seconded him. Another key weapon in the JICPOA arsenal, Mac Showers had played central roles in both the 1942 breakthrough that had identified Midway as a Japanese target, enabling the United States to lay a trap there, and the decrypts that led to the aerial ambush where perished Japanese fleet commander Yamamoto Isoroku a year later. Those were among the most important achievements of the operational intelligence Holmes championed.

  In addition to operational intelligence, biographical reporting, and estimates of opposing strength, JICPOA did translations from the Japanese. It published special studies that surveyed islands, regions, or practical subjects based on captured documents. The translation unit started in the fall of 1943 and at first had little to do. But document hauls at Tarawa, Kwajalein, and Eniwetok had grown steadily. From early 1944 on it became impossible to translate every paper that crossed the transom. Soon the section morphed into as many as fifteen units, each of which handled documents in a particular subject area.

  The fall of the Marianas proved grist for the mill. On Saipan alone, JICPOA took control of fifty tons of documents, including information about the defenses on Tinian and Guam. There were codebooks, radio monitoring schedules, exemplars of Japanese radio direction-finding equipment, and more. The take included actual diagrams of every major aviation facility in the Home Islands, a digest of Japanese naval air bases, and a complete set of Imperial Navy administrative orders. Translators spoke of laboring in the salt mines.