Chapter Thirteen

ADMIRAL MINEICHI KOGA, THE TACITURN COMMANDER IN CHIEF OF the Combined Fleet, had been a close personal friend of his predecessor Isoroku Yamamoto. But Koga lacked his late friend’s strategic and political audacity. He never dared to challenge the suzerainty of the Tokyo-based Naval General Staff (NGS) or the leadership’s single-minded fixation on the all-important “decisive fleet battle”—a concept that dominated Japanese naval planning and strategic thinking before and during the war. This climactic clash of fleets was to occur somewhere in the western Pacific. It would involve substantially all of the capital ships in both the American and the Japanese navy. It would occur (it was hoped) in tactical circumstances favoring the Japanese side. It might begin with punishing land- and carrier-based air attacks on the American fleet, perhaps while the Americans were tied down in support of some major amphibious operation. But the big guns of the Japanese battleships would deliver the coup de grace. The thrashing would be so complete, so shocking, and so devastating that the government of Franklin D. Roosevelt would be moved to ask for a truce. Diplomatic negotiations would follow, and Japan would secure a peace that preserved its sovereignty, its honor, and some portion of its empire.

It would be no exaggeration to say that bringing about a single, all-deciding naval battle amounted to an obsession among the Tokyo admirals. The idea had been inculcated into generations of students at the Japanese Naval Academy at Etajima through the writings of the American strategist Alfred Thayer Mahan. The annals of history provided many convincing examples of such battles, none better than Admiral Heihachiro Togo’s wipeout of a Russian fleet in Tsushima Strait in 1905. Intensive Naval Staff College study of the Anglo-German Battle of Jutland (1916) had further hammered the principle home—though in that instance, it was agreed, the British had grasped the chance of a decisive victory but let it slip through their fingers. Teikichi Hori, a “treaty faction” admiral purged from the Imperial Japanese Navy in 1933, observed that Japanese naval planning had ossified perilously in those years of ultranationalist ferment: “This kind of creeping formalism spread until it became a kind of strategic orthodoxy and [the navy] ended up as a smug little society which insisted that all ideas on strategy should conform to this orthodoxy.”¹

Literally from the first minutes of the Pacific War, events proved that airpower and submarine warfare had unseated the battle line as the ultimate arbiter of naval power. But the admirals had refused to relinquish their trust in the big guns. Though Isoroku Yamamoto had been one of the most air-minded officers to reach the top rungs of the Japanese navy, his move against Midway had been a bid to force a decisive battle early in the war. He had intended a critical role in that battle for the surface warships, including his flagship Yamato. Aviators were not fast-tracked to promotions, nor were they placed into important sea commands and planning jobs as they had been in the U.S. Navy. Not until 1944 was the carrier task force integrated into the heart of the Combined Fleet. War planning proceeded under the orthodox assumption that the battleships (especially the leviathans Yamato and Musashi) would play a leading part in the war’s final act.

So Koga never doubted that he must sooner or later hurl the Combined Fleet into the path of the advancing enemy. But when and where? Since 1942, the heavy ships (battleships and carriers) had been kept in reserve, well out of the enemy’s reach. The Japanese carriers had not come out to fight since the Battle of the Santa Cruz Islands in October 1942. After losing the battleships Hiei and Kirishima off Guadalcanal a month later, the navy had largely relied on its destroyers and land-based airpower to wage its campaign in the South Pacific. Koga would prefer to make his stand within range of friendly terrestrial air bases. He needed time to rebuild his carrier air groups, and as the fuel supply became critical, he had good reason to hope that the battle would be fought as near as possible to the oilfields of Borneo and Sumatra.

On the other side of the ledger, Koga had to acknowledge that time was not on his side. Month by month, the titanic output of the enemy’s industrial plant was arriving in the advanced war zones of the Pacific. American scientific and technical expertise was opening an ever-widening mismatch in air combat efficiency, especially in the vital categories of radar, radio communications, and antiaircraft defenses. The submarine campaign was obliterating 200,000 tons of Japanese shipping per month, and it would soon cripple the nation’s war industries. The enemy’s twin advances in the south and central Pacific were swallowing up strategically vital territories. All such considerations weighed in favor of committing the Combined Fleet as soon as tactical circumstances permitted.

Nor could any military commander ignore the increasingly strident demands emanating directly from the throne. The god-king Hirohito was pressuring his liaison conference to wage the war more aggressively, to confront the American fleet and crush it. As a young crown prince, Hirohito’s education had been largely entrusted to the leading army and navy heroes of the Russo-Japanese War, Admiral Togo and General Maresuke Nogi. The boy had been steeped in Mahanian doctrine, and apparently never doubted that the war must be won by a decisive fleet engagement in the pattern of Tsushima. He let it be known that he was sorely disappointed by the loss of Guadalcanal. After the fall of Attu in May 1943, he sternly rebuked his army and navy chiefs of staff. The emperor expressed anxiety over the waning prestige of Japanese military power and its consequences for the future of his Asian-Pacific empire. His queries became increasingly pointed, shrill, and even sarcastic. He demanded to be briefed in detail, whereas he had previously been satisfied with knowing only the broad strokes, and took a direct part in deciding major questions of strategy. On August 5, as Allied forces drove into the central Solomons, the emperor dressed down the chief of the Army General Staff, General Hajime Sugiyama:

If we continue fighting in this manner, it will be like Guadalcanal. It will only raise the fighting spirit of the enemy, and then the neutral countries will start to waver, China will get big-headed, and the impact on the countries in the Greater East Asian Co-Prosperity Sphere will be enormous. Can’t you somehow beat down the American forces head on at some front? All the battle fronts look bad. Can’t you give the American forces a walloping? If we continue to get pushed back steadily this way, it will have a significant impact on other countries, not just on the enemy. Now, just where are you going to show some success? Where are you going to stage a decisive battle?²

In mid-September, the emperor convened a series of meetings with military leaders at the Imperial Palace. On the agenda was a fundamental overhaul in war strategy. With the support of General Hideki Tojo, who simultaneously held the offices of prime minister and army minister, leaders at the liaison conference agreed to a “New Operational Policy” calling for a significant contraction in Japan’s western defense perimeter to islands far behind the front lines. The new “absolute defense line” would run from the Kurile Islands south through the Marianas, Truk, Palau, New Britain, western New Guinea, the East Indies, and Burma. Positions on that inner perimeter would be heavily reinforced with land-based air units and army troops to be transferred from Manchuria. Garrisons outside the perimeter, including those in the Gilbert and Marshall island groups, would receive no further reinforcement. If attacked, they must exact a bloody toll on the enemy before perishing in combat to the last man.

The buildup of strength on the inner perimeter would take time—time to move troops from the mainland, to erect new shore fortifications, to build aircraft, and to train new air groups—so the policy designated “mid-1944” as “our approximate target for full readiness.”³ But the new policy did not rule out an early confrontation with the American fleet, if opportunity offered: “Whenever the occasion presents, we shall capture and destroy the enemy’s offensive forces.”⁴

Koga was keen to fight his Tsushima sooner rather than later. Since May 1943, the commander in chief and his staff had been studying and revising plans for Operation Z, which envisioned a grand sortie of the Combined Fleet to confront the American fleet in the central Pacific, if possible while the enemy was pinned down in support of an amphibious invasion. Admiral Pownall’s carrier raids of September and October 1943 had twice coaxed a powerful Japanese fleet out of Truk. Koga personally commanded the second of these sorties, and anchored his ships in Eniwetok Atoll for four days in October. In both instances it was soon understood that the American movements had been hit-and-run carrier raids rather than sustained operations, and the fleet returned to Truk.

Relentless shipping losses and the MacArthur-Halsey drive on Rabaul prompted successive revisions to Operation Z. In late October it was decided that the fleet would not come out to fight in Micronesia or in defense of the Bismarcks. The decisive confrontation would be postponed until the enemy attempted to pierce the inner perimeter. Koga would make his stand in the Philippine Sea, in defense of either the Marianas or Palau, depending on the movements of the American fleet. Pursuant to those decisions, when the Fifth Fleet launched its offensive into the central Pacific, Japanese garrisons in the Gilberts and Marshalls received diffident air support and no naval support at all, except as provided by submarine patrols.

After suffering heavy air losses at the Battle of the Santa Cruz Islands (October 1942), the Japanese carrier task force had retired to the Inland Sea for repairs, for refitting, and for the rebuilding of its decimated air groups. Incumbent commander Chuichi Nagumo (who had led the carrier striking force in the attack on Pearl Harbor, the Battle of Midway, and the carrier duels off Guadalcanal) was relieved and sent back to the homeland to command the Sasebo Naval District. His successor was Vice Admiral Jisaburo Ozawa, a tall, gruff, unsightly man nicknamed “the Gargoyle.”^* Ozawa took the fleet carrier Zuikaku as his flagship. The force was reorganized and designated the Third Fleet, consisting of three carrier divisions, two battleships, six cruisers, and more than a dozen destroyers. Most of the Third Fleet’s air strength was concentrated in Carrier Division 1 (the veteran fleet carriers Zuikaku and Shokaku and the light carrier Zuiho) and Carrier Division 2 (the heavy sisters Hiyo and Junyo, both built on converted passenger-liner hulls and commissioned in mid-1942).

Throughout most of 1943, the Japanese carriers remained idle at Truk or in home waters. The same was not true of the air groups. Ozawa’s carrier planes were repeatedly sent south to bolster the deteriorating fortunes of the land-based naval air forces at Rabaul and satellite airfields in the Solomons and New Guinea. In July 1943, most of Carrier Division 2’s aircraft (about 150) flew into Rabaul to be integrated into the Twenty-Sixth Air Flotilla. The Zero pilots, trained at great expense to fly from aircraft carriers, suffered disastrous losses in pitched daily air battles against relentless waves of well-armed B-24s and B-25s.

In August and September, several Japanese fighter units were ordered to fly down to the primitive airstrip at Buin, on southern Bougainville, where conditions on the ground and in the air were even worse than at Rabaul. Halsey’s heavily reinforced Guadalcanal-based air forces (AIRSOLS) battered the hard-pressed dirt airstrip day by day, and many of the Japanese carrier pilots who survived longer than a week succumbed to exhaustion and disease. In October 1943, the flyable remnants of Japan’s badly mauled air forces at Buin began pulling back to Rabaul.

The following month, Koga ordered nearly 200 planes of Carrier Division 1 into Rabaul from Truk (Operation RO). Among them were many of the Third Fleet’s most skilled and experienced airmen, including all of the squadron leaders. More than half of those planes were shot down within a week, and virtually none of the downed aviators were recovered. One of Admiral Ozawa’s staff officers recalled the scene at the Spring Island airfield in Truk as the survivors arrived on November 13:

I was astonished by the small number of planes that had returned. There were hardly any fighters to be seen. But it was worse with torpedo planes and fighter bombers. None of them had come back. . . . On the field Ozawa was giving the commander’s speech of instruction that customarily followed an operation. Afterward I heard from some fliers about the speech. Ozawa had climbed the platform to address them and, being overcome with dismay at how few of his men had survived, was unable to utter a word. He stood there on the platform in silence for a very long time, weeping bitterly.⁵

On January 25, 1944, as the combined AIRSOLS and Fifth Air Force campaign against Rabaul reached its overpowering culmination, Koga fed another 150 planes from Carrier Division 2 into the meat grinder. Few of the Zero pilots had flown in combat prior to the deployment. They were instructed to avoid engaging the American bombers and fighters except when “battle circumstances appear particularly favorable to you.”⁶ For the sake of morale more than anything else, Nakajima torpedo planes from Carrier Division 2 were dispatched on risky night missions against Halsey’s ships off Bougainville and MacArthur’s transports off New Guinea. Those flights scored no hits on American shipping, but dozens of the valuable bombers were lost to antiaircraft fire and operational accidents. Masatake Okumiya, an air staff officer, recalled the dreadful last days at Rabaul:

The days passed in a blur. Every day we sent the Zeros up on frantic interception flights. The young and inexperienced student pilots had become battle-hardened veterans, their faces showing the sudden realization of death all about them. Not for a moment did the Americans ease their relentless pressure. Day and night the bombers came to pound Rabaul, to smash at the airfield and shipping in the harbor, while the fighters screamed low on daring strafing passes, shooting up anything they considered a worth-while target. So intense were the enemy attacks that we were unable to find time to attack their bases. Our losses mounted steadily, and the list of dead and missing pilots grew visibly.⁷

The losses from Carrier Division 2 in this last operation at Rabaul amounted to about ninety planes. Among those few aviators who returned to Truk in late February, many had been laid low by malaria and other tropical diseases. They also brought with them the incubus of defeatism and despair, which inevitably spread through the remaining Third Fleet air groups.

IN JAPAN, NEWS OF THE FEBRUARY 1944 raid on Truk was reported with unusual candor. Some news reports falsely asserted that the Americans had attempted an amphibious landing and been repulsed, but on February 18 the Imperial General Headquarters released an accurate and unsparing account:

A powerful American task force suddenly advanced to our Caroline Islands Wednesday morning and repeatedly attacked our important strategic base, Truk, with a great number of ship-based planes. The enemy is constantly repeating powerfully persistent raids with several hundred fighters and bombers, attacking us intermittently. The war situation has increased with unprecedented seriousness—nay, furiousness. The tempo of enemy operations indicates that the attacking force is already pressing upon our mainland.⁸

As always, the dire tone of this release was intended to arouse the Japanese people to greater efforts and sacrifices. But the regime could not countenance any admission that Japan might be in danger of losing the war. The editors of the widely circulated Mainichi Shinbun must have failed to appreciate the subtle distinction. A front-page opinion piece, published on February 23, warned, “The decisive battles of offense and defense in the Pacific will not be carried out on the homeland shores of America and Japan. They will be fought out . . . on island bases several thousands of miles distant. If the point is reached when the enemy advances to the shores of our homeland, already there is nothing more that can be done.”⁹

The article included some pointed criticism of the training of civilians to fight with bamboo spears, and added that the only real hope of repelling the American advance was to build more warplanes and aircraft carriers. When the newspaper landed on the desk of Prime Minister Hideki Tojo, he blew his stack. Orders went out from the Information Board to suppress the article, but since more than 100,000 copies had already been distributed, the damage (such as it was) could not be undone. The Army Ministry issued a reprimand to the paper, forced the editor in chief to resign, and summoned other leading editors to be admonished against such commentary. The army drafted the journalist who had written the article, a common method of disciplining wayward writers.

Tojo was unpopular with the Japanese people and despised by most of the leading figures in the navy. He was pompous, shrill, and demonstrative, and tended to silence reasoned discussion with outlandish flights of rhetoric and sloganeering. He was always leaping to his feet, even in closed meetings of senior cabinet and military officers. In October 1943, when a minister referred to a disappointing harvest, Tojo stood and bellowed, “Even if we eat nothing, we members of the cabinet intend to give our lives for the nation!”¹⁰ His antics sometimes seemed to ape those of Hitler and Mussolini, but Japanese culture was unsympathetic to the notion of concentrating power in the hands of one man. The emperor was the nation’s singular figurehead, and his ministers were expected to govern with an attitude of humility and reticence.

The emperor was the ultimate underwriter of Tojo’s long ascendancy in the ruling circle. For nearly three years Hirohito trusted and listened to Tojo, and even after the war the emperor defended the general’s conduct and leadership. Immediately after the Truk raid, Tojo moved to consolidate his grip on the military planning and command functions of both the army and the navy. Since the Meiji era, authority in each service had been divided between a ministry and a general staff. But now, with the emperor’s support, Tojo forced the resignation of General Sugiyama, the army chief of staff, and named himself to that post while retaining his positions as army minister and prime minister. Navy chief of staff Admiral Nagano resigned and was replaced by the navy minister, Admiral Shigetaro Shimada, who also kept his existing job. The shake-up brought the planning staffs of both services under the control of the ministries. Because Tojo had always found Shimada a pliable colleague, it also brought the entire regime more firmly under the control of Tojo. Opposition within the general staffs melted away when the emperor let it be known that he had backed the plan.

BETWEEN NOVEMBER AND FEBRUARY, most of the aircraft carriers were again pulled back to Japan for repairs and to take on replacement airplanes and aircrews. The Zuikaku went into dry dock at Kure; the Shokaku, at Yokosuka in Tokyo Bay. The carriers were fitted with new radar systems and antiaircraft weaponry. New aircraft were embarked from Iwakuni Naval Air Station in Hiroshima Bay. Because the newly trained aircrews had not yet been cleared for carrier landings, it was decided to lift the planes aboard by crane rather than fly them aboard after the fleet had put to sea.

Six months was needed to train the new air groups. But the Americans could not be expected to oblige that timetable, so all understood that the Third Fleet might be thrust into battle before it was ready. Ozawa’s air staff fixed an optimistic deadline of April 1 to have the new carrier air force fully trained, equipped, and ready to confront the enemy. The Third Fleet was to have 500 aircraft on nine carriers (five heavy, four light). Another 400 to 500 land-based fighters and bombers would be positioned on island airfields within easy range of the Marianas and Palau.

Japanese war planners had hoped to produce 40,000 new military aircraft in 1944, but the production rate was barely half that level in the fall of 1943. Aviation plants were straining under the pressure of material shortages, maladroit logistics, and a paucity of trained machinists and engineers. Shipping losses bit deeply into deliveries of Malaysian and Jakartan bauxite, the industry’s chief source of aluminum alloys. The Mitsubishi complex in Nagoya had expanded steadily, employing 43,000 workers by the end of 1943, but it had turned out only 1,029 new Zeros in 1943, fewer than half the number demanded by the military services. The Japanese aircraft industry had relied to a disproportionate extent on a small, overworked coterie of talented craftsmen and technicians, and was never optimized for mass production. Belated efforts to introduce standard production-line techniques brought some improvement, but neither Mitsubishi nor the other major aircraft suppliers (Nakajima, Aichi, Kawasaki, Tachikawa, Yokosuka) managed to ramp up output fast enough to fill the military’s ballooning wartime orders.

When a government inspector passed through the Nagoya works in late 1943, he was surprised to learn that newly manufactured Zeros were still being hauled away from the plant by teams of oxen. There was no airfield adjoining the Mitsubishi plant. The new units had to be transported overland to Kagamigahara, twenty-four miles away, where the navy would accept delivery. The aircraft were too delicate to transport on trucks, and the railheads were not convenient. Twenty oxen had died, and the remaining thirty were verging on complete exhaustion. Feed had been obtained on the black market, but the supply was not reliable. Essential wartime deliveries of replacement aircraft thus hung on the fate of a diminishing herd of underfed beasts. Mitsubishi engineers at length discovered that Percheron horses could haul the aircraft to Kagamigahara faster and required less to eat. These ludicrous exertions, when compared at a glance to the arrangements at Boeing, Douglas, or Grumman, tell most of the story of Japan’s defeat.

Recent encounters with the new-generation American fighters—the Hellcat, the Corsair, and the Lightning—had settled any remaining doubts that the Zero was overmatched and obsolete. Jiro Horikoshi’s airplane had been a feat of pioneering ingenuity, and it will always be a milestone in the history of aviation. From the start of its remarkable career, however, the Zero had always embodied a set of design compromises. Horikoshi’s team had wrung every last ounce of surplus weight out of the aircraft, and given it a very large wingspan and control surfaces in proportion to its size. These traits gave the Zero long range, a fast climbing speed, and supreme agility, but they also made it sluggish at altitude, slow in a dive, and disastrously vulnerable to enemy fire. Mitsubishi had invested its limited design and development resources into two new interceptors, the J2M Raiden and the A7M Reppu. Both airplanes offered performance improvements over the Zero, but the early prototypes were plagued with mechanical failures and design flaws. Given enough time and engineering manpower, all of those problems might have been overcome (just as, for example, the assorted defects of the Curtiss SB2C and the B-29 Superfortress were corrected as those machines entered service). But time was short, and engineering manpower was sorely limited. Only small numbers of the Zero’s successors were placed into service, and they were never popular with the airmen or air staff.

Without a viable alternative, the Japanese navy instead committed itself to improving the Zero. Mitsubishi tinkered with successive alterations throughout the war, generally adding horsepower, firepower, and armor plating while reducing wingspan—the Model II, the Model 21, the Model 22, and the Model 52 “Hei,” of which more were built than any other version. Performance improvements were generally too slight to be noticed by the American pilots who engaged the planes. Lieutenant Commander Iyozo Fujita, one of the few veteran Zero pilots who survived the war, found that Model 52 lacked the speed and agility of its predecessors. It was “simply aggravating to fly,” he said. “The aircraft’s nose did not rise quickly enough, so it was hard to aim my machine guns well. Because of this, I felt that I lost a lot of opportunities to get hits on enemy aircraft.”¹¹ The navy also experimented with configuring the Zero as a fighter-bomber, fixing a 550-pound bomb under its fuselage, but an airplane powered by a 1,000-horsepower engine was not well suited to such a role. In all its versions, the Zero was still packaged in the same light airframe, which could not absorb much punishment without blowing apart or bursting into flame. Design alterations consumed an exorbitant amount of the Mitsubishi engineers’ time and energy, diverting their attention from the next-generation planes. They also clogged up the company’s manufacturing operations. Output at the Nagoya works fell short of the navy’s targets every year from 1940 through 1945.

In the 1930s, Japanese firms had imported American and European precision machine tools, needed to polish, grind, and mill high-performance metals. Prewar embargos had cut off those critical imports. By 1942, the plants were equipped with aging equipment that could not be replaced or upgraded. As a nation destitute of natural resources and mining deposits, Japan lacked access to the high-performance lightweight metals found in the 2,000-horsepower engines that powered the big American fighters. The Japanese aviation industry consistently struggled to produce reliable new aircraft engines that achieved high power ratings within desired weight limits. Atsushi Oi, an officer at the Naval Personnel Bureau, pointed to the small scale of Japan’s “so-called shadow industries such as the automobile industry which can be easily converted to produce aircraft engines.”¹²

Critical deficiencies at home were exacerbated by retrograde conditions in the advanced combat theaters. The Japanese had nothing to rival the civil engineering capability of the U.S. Seabees. Island airstrips were built with light construction equipment and backbreaking manual labor. In the early phase of the war, the Japanese had captured many of the best Allied airfields in the South Pacific and Southeast Asia. As they were wrested back by the enemy, the Japanese were forced to retreat to bases in which ground support facilities were inadequate and living conditions abysmal. Pilots and mechanics were quartered in tents, cooked over open fires, and bathed in fuel drums. Latrines were built over vile open-pit cesspools. Medical facilities were undermanned and undersupplied, and surgery was often performed without anesthesia. There was always a need for replacement airplanes and airmen, but that was only one facet of the logistics problem. Airfields in forward combat areas required a constant resupply of aviation fuel, spare parts, lubricants, ordnance, and ammunition. Fresh ground crews had to be flown in to replace those who succumbed to illness, injury, or death by bombing. Newly trained mechanics were less skilled than the veterans. Japanese air cargo transports were limited in number, and many were lost to operational accidents and enemy attacks. The throttling of Japanese maritime transportation was well underway and growing worse month by month.

Under such conditions, aircraft on the front lines fell into poor repair. Saburo Sakai, the fighter ace, told his superiors that every Zero fighter should receive a complete overhaul after 150 flight hours. In the South Pacific, however, most Zeros in service surpassed 200 hours and began running very rough.¹³ By 1944, most front-line fighters were shot down before they ever had a chance for a maintenance overhaul.

Writing years after the war, Jiro Horikoshi observed that his country could not draw from the deep wellsprings of engineering and technical expertise that existed in the United States. There was nothing in Japan to compare with America’s sprawling complex of universities, research laboratories, design firms, and heavy industries. Japan had a small circle of gifted engineers employed by the navy, the army, and about a dozen industrial firms. Owing to rivalries between the army and the navy and between rival companies and cartels (zaibatsu), much of their work was duplicative and wasteful. All too often their talents were squandered on impractical, profligate, stop-and-start projects that never got off the ground (in some cases, literally). They were resourceful and dedicated, but there were not enough of them. Horikoshi and his colleagues drove themselves to the verge of complete exhaustion and collapse, until the doctors and bosses ordered them to rest. “Such poor management of technical policy created the situation where we had no other choice but to rely on the Zeros from the beginning of the war until its end,” Horikoshi wrote, “and this, in turn accelerated Japan’s defeat.”¹⁴

THE NEWEST ADJUNCT TO OZAWA’S FORCE was the Taiho (”Great Phoenix”), a 29,300-ton fleet carrier with an armored flight deck designed to withstand the sort of dive-bombing attack that had destroyed four carriers at the Battle of Midway. This was the first Japanese carrier to have such armor. The heavy steel deck required costly trade-offs, with the result that the ship’s two hangars were small in proportion to her great size, and were serviced by just two elevators. She would carry only seventy-five aircraft, twenty fewer than the new American carriers. Taiho entered service in March 1944 and sailed that month to Singapore, where Ozawa took her as his flagship.

Two new carrier bombers were gradually replacing their obsolete predecessors. A new torpedo plane, the Nakajima B6N Tenzan (“Heavenly Mountain”), was faster and had a longer range than its predecessor, B5N (“Kate”). Its service introduction had been delayed by a balky in-house engine produced by Nakajima, and the navy had eventually insisted on a more reliable Mitsubishi power plant. The Tenzan, which the Allies designated the “Jill,” seated a three-man aircrew and could carry a torpedo or a 1,754-pound bomb load for a distance of 2,000 nautical miles. It had gone into production in 1943; by early 1944 Nakajima was turning out about a hundred new planes per month.

The long-serving Aichi D3A dive-bomber (“Val”) was giving way to the Yokosuka D4Y Suisei (“Comet”), a fast and maneuverable aircraft derived from a design by the German manufacturer Henkel, and powered by a 1,400-horsepower radial piston engine built by Aichi under license from Daimler-Benz. The Allies named her “Judy.” Like the Jill, the Judy’s service introduction had been delayed by various problems with early prototypes, notably a tendency for the wings to flutter while in a dive. The issue was eventually corrected with stronger wing spars and revamped dive brakes. With a maximum speed of 342 nautical miles per hour in level flight, the Suisei was the fastest carrier dive-bomber to be placed in general service by any combatant nation during the Second World War. It had been developed by engineers at the Yokosuka Naval Air Technical Arsenal, an Imperial Navy facility, but most of the units that went into service were built in Nagoya by Aichi. Its chief drawbacks were a lack of armor and self-sealing fuel tanks and its high speed in flight deck takeoffs and recoveries. Like all Japanese carrier planes, it was relatively easy to shoot down. The Suisei also lacked folding wings, a feature that had become de rigueur in the American carrier planes. Planes with fixed wings required more storage space, reducing the potential size of the air groups, and they required more time to cycle between the hangars and the flight decks. While the Essex carriers and their “airedales” had achieved quantum leaps in plane-handling efficiency, Ozawa’s crews were hard-pressed even to meet the standards set by Nagumo’s Kido Butai (carrier striking force) in 1941 and 1942.

If well-handled, the Tenzan and Suisei offered considerably better performance than their predecessor airplanes. They flew faster, higher, longer, and farther. Their introduction into the fleet had been rocky, but the same was true of carrier planes produced by other nations. If wartime Japan had possessed the resources to build a new fleet of larger and more sophisticated aircraft carriers to compare with the Essex class, and had simultaneously trained a new generation of pilots with the skill to measure up to 1942’s “first team,” the Tenzan and Suisei might have posed a deadly threat to the American fleet. But most of Ozawa’s aging Third Fleet carriers (the Taiho excepted) were too small to comfortably handle the hot new machines. The heavy sisters Hiyo and Junyo, whose top speed was about 24 knots, could not launch or recover them at all unless the wind was blowing hard. Because it needed a full deck run to achieve its takeoff speed, the Suisei had to be spotted well aft. That limited the number of bombers that could be launched in a single cycle. Air staffs experimented with new catapults and even fuselage-mounted rockets to get the new dive-bombers safely aloft.

Carrier recoveries were always a white-knuckle performance. The Suisei approached fast and low, often failing to snag the arresting cables. When the new dive-bombers entered the landing circle, recalled an air officer on the Zuikaku, “there was always great tension on the bridge. We had special respect for the crews of these planes.”¹⁵

The Japanese navy had been slow to acknowledge that its flight-training pipeline was inadequate. Most of the replacement aircrews who went aboard Ozawa’s carriers that spring of 1944 had spent fewer than 150 hours in the cockpit and had acquired only rudimentary flying skills. Virtually none had practiced gunnery or qualified for carrier landings. The projected flight-training shortfall had been anticipated and discussed among Japanese aviators prior to the war, but no concerted effort to expand the pool of qualified flyers had come until 1941, when it was too late. Ruinous air losses in the Rabaul campaign had left the Japanese with no choice but to rush their cadet pilots through truncated programs and send them out to the fleet. Ozawa now faced the daunting task of preparing those undertrained young men to fight the decisive naval battle that loomed in the immediate future.

The elite Japanese naval air corps that launched the Pacific War had been trained in a small, super-exclusive program at the Kasumigaura Naval Air Training Center, the “Japanese Pensacola,” near the city of Tsuchiura, about fifty miles north of Tokyo. The trainees had received two years of classroom instruction, flight training, and gunnery training, followed by an additional year of training in a forward operational unit. Recruits were selected from among recent graduates of Etajima (the naval academy), noncommissioned officers in the fleet, and civilian students between the ages of sixteen and nineteen. Screening criteria were extremely and even excessively rigorous. Even so, fewer than half those accepted survived to the end of the training program. When naval warrant officer Saburo Sakai applied for flight training in 1937, he was one of seventy chosen from a pool of 1,500 applicants. In primary flight training, each instructor was responsible for just three students. Lesser performers were pared from the program at every stage. Of the seventy men in Sakai’s class, forty-five washed out before receiving their wings.¹⁶ Most cadet pilots accumulated more than 500 cockpit hours before they were permitted to attempt a carrier landing.

The small and selective program produced some of the best aviators in the world, but it did not generate enough of them. In the mid-1930s, the Imperial Navy was producing only about a hundred new pilots per year. That figure grew after the China Incident (1937), but not rapidly enough to fulfill the navy’s internal goals. Even in the early stages of the Pacific War, when the Japanese naval air squadrons ruled the skies throughout the theater, senior aviation officers were concerned about the supply of replacement aircrews. The navy had about 3,500 trained pilots in front-line service at the start of the Pacific War. Reserves were thin, and attrition would immediately begin to strain the system. Three months before Pearl Harbor, the Imperial Navy optimistically adopted a plan to train 15,000 new pilots. Masatake Okumiya, serving as an air staff officer at Kasumigaura, pointed out that the navy did not have enough combat airplanes to continue training pilots like they did in peacetime—that is, by sending the Kasumigaura graduates to advanced units to perform an additional year of “on-the-job” training. Moreover, even if the funnel was opened to a much larger training pipeline, “We would not feel the effect of the mass-training program for at least another two to four years.”¹⁷

Japan’s flight-training system was simply not designed to expand. The small size of the aviation industry and the existing pool of flyers were inflexible constraints. Ninety percent of Japanese pilots were enlisted personnel or warrant officers. A navy-sponsored civilian flight-training program had been founded in 1934, drawing from the ranks of university students, but it was too small to matter until 1943. The sensitivities and prerogatives of the naval officer corps were a barrier to new recruitment avenues—the Etajima clique stalwartly resisted the influx of reservist pilots and treated them poorly when they were assigned to front-line units. The Yokaren program, which recruited adolescents directly out of secondary schools, was expanded rapidly, and new training centers established throughout the Japanese islands. But these young men often had little education in the sciences and no prior experience handling machinery, and they needed as much as two and a half years of ground instruction before starting primary flight training.

The dearth of qualified and experienced instructors was a bottleneck. The prewar student-instructor ratio of three to one was obviously unsustainable. In 1941, each instructor took on eight students; in 1942, twelve. More than twelve was deemed impracticable. The navy’s veteran aviators were desperately needed in front-line units, and they were rarely sent back to Japan for assignment as instructors. The Imperial Navy did not acknowledge the existence of such a syndrome as “pilot exhaustion,” preferring to believe that all such frailties could be solved by an appeal to the purified warrior spirit (Senshin), and never adopted a general policy of rotating pilots out of the combat zone. “They won’t let you go home unless you die” was the Japanese aviator’s mordant lament. But if veterans were not reassigned as flight instructors, men in the training pipeline could not benefit from their hard-won expertise.

Saburo Sakai, one of the few aces to be brought back to Japan as a wartime instructor (he had been wounded over Guadalcanal), was disgusted by the attitude of the young ensigns and lieutenants under his charge. Sakai and other noncommissioned officers were required to address the Etajima graduates as “Honorable Ensign” or “Honorable Lieutenant.” The officers behaved with self-aggrandizing disdain toward enlisted combat veterans, including wounded heroes and aces who had shot down dozens of enemy planes. “Hey, you!” they shouted rudely. “How’s my skill coming along?”¹⁸ Enlisted pilots were assigned menial tasks after hours, such as mess cooking and peeling potatoes, while the officers partied at restaurants and tea houses and got drunk on high-shelf liquor. Even in advanced combat squadrons, officers treated their more skilled and experienced enlisted wingmen like members of a lesser caste. The effect on unit cohesion was predictable. Long after the war, Sakai remained bitter:

I never learned anything from those officers who graduated from the Naval Academy about how to search out and spot the enemy, or how to outmaneuver and shoot them down. Instead, we learned from fellow noncoms, and the noncoms learned from the old noncoms, and the old noncoms learned from the warrant officers who were themselves learning from the special-service officers at the bottom of the officer heap. We became like brothers, looking after each other. Yet those officers, graduates of the Naval Academy, unskilled, lacking in any technique, they were officially our leaders. The nation does not know this. How many of their precious men were killed because of the misjudgments and lack of military acumen of those men! It was horrific, let me tell you!¹⁹

In 1943, as the Japanese naval air corps was macerated in the South Pacific, the Imperial Japanese Navy made a despairing bid to expand all its flight-training recruitment channels simultaneously. Selection standards for the Yokaren program were relaxed, and about a dozen new regional training centers were established. For example, Otsu Yokaren Class No. 19 admitted 1,500 recruits in December 1942. The following class, No. 20, admitted 2,951 young men in May 1943.²⁰ The brightest and most capable candidates were put into an accelerated track and rushed into primary flight training. Inevitably, the duration and quality of their training deteriorated. The amount of flight time required to graduate declined steadily. By early 1944, the average was fewer than 150 hours. Gunnery training and target shooting were dropped from the program; the cadet pilots would have to learn how to shoot after arriving in forward units. In retrospect, it was suddenly clear that the navy’s unduly restrictive prewar recruiting standards, and its practice of ousting so many apparently competent students from Kasumigaura, had been shortsighted and self-defeating. Sakai was undoubtedly right when he judged that the “forty-five pilots expelled from my own student class at Tsuchiura were superior to those men who completed wartime training.”²¹

The navy also began recruiting large numbers of university students into flight schools. The students were enthusiastic; the navy received 20,000 such applications by mid-1943. It seemed as if all the young men at the top universities (Tokyo, Keio, Waseda) were clamoring to become naval aviators. The scholar and diarist Kiyoshi Kiyosawa worried that the naval air war would devour the nation’s future leadership. “Japan was expecting great future contributions from these young people. These youths will be destroyed. . . . In the present war this is of the greatest concern.”²²

Kiyosawa was disturbed to hear that many of the young men were severely beaten and abused after enlisting in the navy. Inductees were forced to stand on tip-toe for an hour or more. If they failed, they were beaten savagely with bamboo rods, shoes, and leather sword straps. In May 1944, he recorded “reports that in the navy the thrashing of young recruits is widespread. It is particularly extreme toward student conscripts. I hear that they thrash them with clubs and a great many suffer broken hip bones.”²³ Later that year, he lamented the fate of new inductees at Yokosuka Naval Base: “According to friends, the recruits are cruelly beaten with clubs and other things, and because of this many become deformed. When one man said things in the wrong way, the entire squad was beaten and knocked down. Is there anywhere in the world a place as barbaric as this?”²⁴

AS A FORMER CHIEF OF NAVAL PERSONNEL, Admiral Nimitz held decided views about administration and staffing. Decades of experience had taught him that headquarters staffs, if allowed to drift unchecked, would grow inexorably until they were bloated, listless, and unwieldy. New branches and subdepartments would sprout like mushrooms after a hard rain. Growth in one department would elicit demands for more manpower from the others. New administrative strata would creep into the organizational charts. Departments would burst out of their appointed office spaces and surge into building annexes. Fiefdoms would arise and rivals would fall on one another in bitter internal disputes. Nimitz had seen it all before, and he was determined not to see it happen in Pearl Harbor. He let it be known that headcount was to be kept to a manageable minimum. He enforced the edict himself, if necessary, by hauling his department heads into his office and compelling them to explain and justify the addition of officers or enlisted men to the headquarters roster.

But Nimitz’s theater command was expanding rapidly as new ships, hardware, and personnel flooded into the theater from the mainland. A small staff, desirable as it was, could not handle the immense complexities of planning, command, and administration. A CINCPAC annex housed the staff functions for the management of logistics and the maintenance of the fleet. An ever-expanding array of subordinate commands built up their staffs. Whenever possible, those organizations took in members of all branches of the armed services.

Among Nimitz’s greatest priorities was to manage and contain the internal tensions and astringent personal rivalries inherent in the multiservice theater command. In this respect he was about as successful as any leader could have been. He frowned on open bickering between the services, and was scrupulously even-handed in hearing every man’s point of view. Even the stormiest discussion was guided to a constructive conclusion. No matter how virulently they argued, men left his office on cordial terms. Many who worked with Nimitz later recalled his shrewd use of the well-applied joke. While winding down a testy planning session in the spring of 1944, the admiral said he was reminded of history’s “first amphibious operation,” conducted by Noah. “When they were unloading from the Ark, he saw a pair of cats come out followed by six kittens. ‘What’s this?’ he asked. ‘Ha, ha,’ said the tabby cat, ‘and all the time you thought we were fighting.’ ”²⁵

The antagonisms could be soothed, but they could not be banished altogether. Several individual relationships were strained to the breaking point. Kelly Turner and Holland Smith occasionally behaved like the best of friends, but when they were forced to commit themselves to the details of planning an operation, there was always the risk of a new eruption. In February 1944, Turner was promoted to the three-star rank of vice admiral, and Raymond Spruance was promoted to full admiral with four stars, but Holland Smith was left behind in the two-star rank of major general. They had all worked together in the successful Gilbert and Marshalls operations. Smith had suspected that the navy would take care of its own while keeping the marines in their place, and here was stark confirmation. “You people ought to be proud of making stars,” he spat at Turner. “You people going around and getting a star. . . . Every time the marines go into battle, you’re liable to get a star.” Turner replied, “Yes, and I’m proud to get stars for my command of the marines.”²⁶ In Washington, General Vandegrift pushed Admiral King to approve the well-deserved promotion. Smith received his third star in May 1944.

The army continued to chafe under the setup that placed Smith in command of the Fifth Amphibious Corps (VAC). Before Operation GALVANIC, General Robert C. Richardson had tried and failed to extract his amphibious 27th Infantry Division from Holland Smith’s command. Now he returned to the theme with redoubled determination. Saipan and Guam, next stop on the road to Tokyo, were not mere ribbons of coral sand, like the central Pacific atolls—they were large and ruggedly mountainous islands. The marines, according to the army view, were specialized assault troops trained to seize a beachhead. Warfare on larger landmasses was the army’s area of special expertise. Thus an army general should command the VAC, or at least the army troops therein. Nimitz gave Richardson a respectful hearing, but then he rejected his proposals on the grounds that one does not fix what is not broken. But the CINCPAC had to keep Admiral King closely advised of these disputes, because there was always the possibility that they would be appealed to General Marshall, Secretary of War Henry Stimson, or even President Roosevelt.

Nor was the Army Air Forces pleased with its lot in Nimitz’s theater, which was to be stuck fast under Rear Admiral John H. Hoover, Nimitz’s commander of shore-based aircraft. Hoover was at loggerheads with his principal USAAF subordinate, General Willis Hale, over various issues related to procedures on airbases with mixed army, navy, and marine air units. Richardson had inserted himself into that dispute as well, and the reverberations had reached the Joint Chiefs of Staff through the Army Air Forces chief Hap Arnold. Hoover and Hale were said to cordially hate one another, though both denied it when queried by Nimitz. Hale had even failed to call on Hoover when passing through Funafuti, where Hoover had his headquarters—a serious breach of decorum. General Arnold would later demand local command of the islands serving chiefly as USAAF bases, and Nimitz would refuse. Arnold felt strongly enough to take the issue up with his fellow joint chiefs, who had little inclination to intervene in a local issue of command relationships. Stimson bluntly advised Arnold to stop rocking the boat.

The relationship between Nimitz and his deputy, the aviator John Towers, remained strained. Towers and the brownshoes had continued to press their case for greater recognition of the ascendancy of carrier aviation. Towers wanted action to rectify the disparity in decorations awarded to army and navy pilots, he wanted more real power for aviators on all planning staffs, and he wanted to permit more press coverage of the American carrier raids deep into enemy waters. War correspondents in Pearl Harbor did what they could to stoke the flames. Wartime censorship limited what they could actually write about the discord, but they liked to goad the aviators into complaining about the surface officers. When Admiral Nimitz learned that one of his staff officers had been asked, “What camp do you belong to?” he realized that the aviators had been airing the navy’s dirty laundry in the presence of newsmen. Not for the first time, or the last, the CINCPAC read Towers the riot act.

But the issues raised by Towers were substantive, and they demanded to be addressed. Admiral King (himself an aviator) had asked retired Vice Admiral Harry E. Yarnell, something of a legend in the navy, to conduct a thorough survey and submit recommendations for future personnel policies. Yarnell had solicited the written and oral testimony of most of the navy’s major commanders in the Pacific, and produced a report, which was delivered to King in the fall of 1943. His chief recommendation, that the commander in chief of the Pacific Fleet should be an aviator, was deemed unacceptable inasmuch as it would require relieving Nimitz of command. But the collected views of the admirals did lend strength to the aviators’ case. A few ambitious brownshoes demanded sweeping all of the non-aviators, including Nimitz and Spruance, out of their jobs. But most leading figures (including Halsey, Spruance, Kinkaid, Turner, Calhoun, and Fitch) stressed that aviation must be welded into an effective component of the fleet and naval commands. Aviators should be strongly represented on all planning staffs and in every major command, and take a leading role in planning deployment of the carrier task forces. But the situation in the Pacific did not call for a general purge of blackshoe officers. Indeed, some senior line officers worried that the navy had been too hasty in moving young aviators up the career ladder. They could not suddenly acquire the skills and experience needed to command a big ship at sea. Time and again, remarked the blackshoe Carl Moore, aviators had been rushed through seagoing commands with the sole purpose of making them eligible to skipper a carrier. “So they put them in a tanker for a year and then they give them command of a 50,000-ton carrier, and they terrify everybody else in the same ocean.”²⁷

On April 28, 1944, Navy Secretary Frank Knox died of a heart attack. He was succeeded by his undersecretary, James Forrestal, the former bond salesman who had been recruited out of the Wall Street firm Dillon Read in 1940. Forrestal had championed the cause of the aviators. He had often run afoul of Ernest King, who resented the civilian’s aggressive interest in policy-making. The two men did not get along at all. But before Knox’s death, Forrestal and King had agreed on a significant new personnel policy. All non-aviation officers in important commands must take an aviator as chief of staff, and all aviators in important commands must take a blackshoe as chief of staff. It was a compromise intended to spread the aviators’ views and expertise into every command, but without forgoing the contrasting perspective of surface officers.

The announcement brought protests from many who were content with the status quo. Among the two most prominent duos broken up by the new policy were Raymond Spruance and Carl Moore (Fifth Fleet), and Marc Mitscher and Truman Hedding (Task Force 58). Neither Spruance nor Mitscher wanted to lose their incumbent chiefs of staff, and both asked that an exception be granted. Both were refused, though Moore was not forced out until the summer of 1944. Mitscher’s new chief of staff was Arleigh Burke, who made his name commanding a destroyer squadron in the Solomons. Burke reported aboard Mitscher’s flagship, the Lexington, in March 1944.

THE MARIANAS HAD NOT FIGURED PROMINENTLY in prewar American planning for a war against Japan (War Plans Orange and Rainbow), chiefly because the islands lay well north of the direct sea route between Hawaii and the Philippines. But Admiral King was convinced that the Marianas represented the war’s strategic keystone because they were well situated on the sea routes linking Japan to its resource territories in the South Pacific and its fleet base at Truk. They also would offer a satisfactory base of operations for the Pacific submarine fleet, as they were located on the threshold of the fleet’s hunting grounds. And the island of Tinian offered suitable territory for airfields that would allow the new B-29 bomber to reach the Japanese home islands.

In successive Allied conferences in Washington (May 1943) and Quebec (August 1943), the Combined Chiefs of Staff had deliberately chosen to defer final decisions concerning Pacific operations in 1944, with the understanding that the direction and timing of the next thrusts would be guided by events and opportunities. King had managed to insert references to the Marianas in Allied conference documents, but the Combined Chiefs did not formally order the capture of these islands until the Cairo conference in November 1943. Cairo’s final conference documents endorsed a series of “Specific Operations for the Defeat of Japan, 1944,” including the stipulation that “a strategic bombing force will be established in Guam, Tinian and Saipan for strategic bombing of Japan proper.”²⁸

For all the agitation between the navy and the USAAF, King and Arnold were now wary allies. Arnold believed that airfields on these islands offered the best hope of fulfilling the promise of the B-29 Superfortress. This huge airplane could carry twice the bomb load of the B-17 to a range of some 1,600 miles. Its pressurized cabin allowed it to operate at high altitudes, safely above the reach of enemy fighters. The B-29’s introduction into service had been repeatedly delayed by defects large and small, most notably in its 2,200-horsepower Wright Duplex Cyclone engines. Arnold had been a voluble champion of the B-29, repeatedly assuring Marshall, Stimson, Congress, and the president that it would be ready for service by the beginning of 1944, and that several hundred of the giant machines could begin regular bombing runs over Japan from airfields in China. The Congress had kept faith with the B-29, pouring more than $1 billion into it by the end of 1943. But only ninety-seven B-29s had been produced, and only sixteen of those could be flown. Arnold told the president, “I regret exceedingly to have to inform you that there has been a holdup in production of engines. It looks now as if it will be impossible to get the required number of B-29s together in China to start bombing before the first of March, and with the possibility of not getting them there before the first of April. At this writing I expect to have 150 B-29s in China by March 1st, of which 100 can be used against Japan.”²⁹

FDR had personally assured the Chinese government that the B-29s were coming and would bomb Japan from Chinese airfields. It was a mainstay in his strategy to keep China in the war against Japan. He feared the consequences if the Army Air Forces was unable to make good on the repeated pledges. Meanwhile, Arnold was growing increasingly skeptical that China offered the best location for B-29 airbases. Chiang Kai-Shek had agreed to convert and enlarge existing airfields to receive the big bombers, but construction had lagged behind schedule, and there was no assurance that the Chinese could protect the bases against the ever-enterprising Japanese army. In early 1944, it appeared that the B-29s would have to begin operations from bases deep in the country’s interior, at Chengtu and Chungking. Smaller airfields nearer the coast might be used for refueling stops, but they were not secure enough to serve as major operating bases. In April 1944, the first B-29s were flown into China from India, with the pilots taking them “over the hump”—a route that was both dangerous and resource-inefficient because of the immense amount of fuel required.

With all of those considerations in mind, USAAF planners had taken an increasing interest in the Marianas. Guam, Saipan, and Tinian offered the best prospects of big airfields within range of Japan. They were large enough to accommodate a fleet of 1,000 B-29s. The bomber’s effective radius of 1,600 miles would bring Tokyo and the industrial regions of the Kanto Plain within reach. Once captured, bases on the Marianas would be secure against the enemy and could be amply supplied by sea. In October 1943, USAAF planners formally recommended that the Fifth Fleet and the amphibious forces of the central Pacific extend their offensive to the Marianas. That recommendation was accepted and adopted by the Combined Chiefs of Staff in Cairo two months later.

With its breathtaking range and tremendous weight-carrying capacity, the B-29 was a weapon without precedent, and it was coveted by commanders all over the Pacific. MacArthur expected a major allocation of the new planes to General Kenney’s Fifth Air Force. Nimitz and his subordinates were interested in their capabilities for long-range antisubmarine patrols and mine-laying missions. But Hap Arnold had long preached the gospel of concentrated airpower and had often expressed his grief and indignation at the tendency to disperse bombers all over the globe. He was determined to deploy virtually all available B-29s in bombing raids against Japan and to keep them under his direct control. With Marshall’s support, he convinced the Joint Chiefs of Staff to create the Twentieth Air Force, comprising the 20th and 21st Bomber Commands, to be controlled directly by the JCS. This was presented as a fait accompli to MacArthur and Kenney, who raged against the decision but could do nothing to reverse it.

Once the Japanese were ousted from the Marshalls, MacArthur was determined to bend Nimitz’s line of advance southward. Indeed, MacArthur hoped to consolidate all available forces for his planned invasion of Mindanao, the southernmost major island of the Philippines. The Pacific Fleet might cover his northern flank by seizing Truk and Palau. Even better, the Allies might seize this opportunity to exorcise the demon of a divided command, and anoint MacArthur the Pacific Supremo. The Pacific Fleet and the formidable Fifth Amphibious Corps would then come under his direct command. An invasion of the Marianas did not fit into this master design.

MacArthur’s chief of staff, Major General Richard Sutherland, formally presented the case against the Marianas in Washington. He observed that 1,000 miles of ocean lay between the Marshalls and the Marianas. To mount a major amphibious invasion across such a distance would drain Allied shipping resources and require many months of preparation. The invasion could not be supported by any land-based aircraft, and the carriers of Task Force 58 would not be able to establish complete dominion in the skies above the landing beaches while simultaneously fighting off the Japanese fleet, should it come out to fight.

Nor did MacArthur hesitate to bypass the Joint Chiefs of Staff by appealing directly to Secretary Stimson and even to Roosevelt. He predicted that the bloodbath at Tarawa would be repeated, but on a far greater scale. MacArthur would contain American casualties by bypassing strong positions and moving quickly to seize territory where the enemy was not strongly entrenched. This last argument, as MacArthur and his acolytes knew well, would be endorsed and trumpeted by the general’s supporters in Congress and in the Hearst press. In mid-January 1944, MacArthur sent a statement of his views on the subject to Stimson (via Brigadier General Frederick H. Osborn):

These frontal attacks by the Navy, as at Tarawa, are tragic and unnecessary massacres of American lives. . . . The Navy fails to understand the strategy of the Pacific, fails to recognize that the first phase is an Army phase to establish land-based air protection so the Navy can move in. . . . Mr. Roosevelt is Navy-minded. Mr. Stimson must speak to him, must persuade him. Give me central direction of the war in the Pacific, and I will be in the Philippines in ten months. . . . Don’t let the Navy’s pride of position and ignorance continue this great tragedy to our country.³⁰

While the relatively bloodless conquest of the Marshalls still lay in the future, trepidation about the potential costs and risks of taking the Marianas found a foothold even in Nimitz’s CINCPAC headquarters. On January 27, 1944, a conference of senior officers from throughout the Pacific convened in Pearl Harbor. MacArthur was represented by Sutherland and his air and naval chiefs, General Kenney and Admiral Kinkaid. Halsey remained in the South Pacific but sent his chief of staff, Admiral Carney. All leading players on Nimitz’s staff were there, as were the naval air chief John Towers, the logistics chief William L. Calhoun, and General Richardson of the army.

From the outset, it was apparent that few were inclined to invade the Marianas. According to Kenney, the sole voice in favor of the operation was that of “Soc” McMorris, Nimitz’s chief of staff. The inevitable absence of land-based air support for the fleet and amphibious forces was deemed a serious obstacle. Many were wary of infantry combat losses on a scale much greater than those of Tarawa. Towers was concerned that the Japanese would stage punishing airstrikes through bases on Iwo Jima and Chichi Jima. Kinkaid pointed to a lack of good natural harbors in the Marianas, remarking that “any talk of the Marianas for a base leaves me entirely cold.”³¹ Calhoun agreed that it would be simpler to move the supporting train of auxiliaries and support vessels along the northern coast of New Guinea. Kenney, with Towers’s emphatic support, cast doubt on the entire enterprise of using B-29s against Japan from bases in the Marianas. Fighters would lack the range to accompany the bombers on the long missions, and without fighter support the Superfortresses would be obliged to bomb from a very high altitude, where accuracy would be poor. Such bombing raids, he said, would be “a stunt” and nothing more. Their consensus recommendation against invading the Marianas was recorded in a memorandum, which Nimitz endorsed and sent on to King. General Kenney later wrote, “The meeting finished with everyone feeling good and ready to work together to get the war over.”³² At a closing luncheon on the twenty-eighth, Nimitz told the assembled group that victory was in sight.

Informed by Sutherland of the result, MacArthur assumed all was settled. All opposition to his proposed line of attack had apparently evaporated, and Nimitz had at last agreed to commit his powerful fleet to the thrust toward the Philippines. MacArthur contacted Marshall and asked for follow-up directives from the Joint Chiefs. His proposed strategy significantly exceeded what the conferees in Pearl Harbor had discussed. He should have the bulk of all Pacific naval forces placed under his command, with Halsey as his new fleet chief, and he wanted the B-29s.

Startled by the sudden uprising, King moved quickly to stamp it out. He wrote Nimitz: “I read your conference notes with much interest, and I must add with indignant dismay.”³³ He rebutted the major arguments against the Marianas and closed with the blunt observation that the proposal to join forces with MacArthur in the south “is not in accordance with the decisions of the Joint Chiefs of Staff.”³⁴ Marshall and Arnold were similarly unmoved.

In any case, subsequent events in the Pacific undermined the case against invading the Marianas. Prewar ideas concerning major amphibious operations had been rendered moot by the immense power of the Pacific Fleet to sweep away opposition in the air and on the beaches. In Operations FLINTLOCK, CATCHPOLE, and HAILSTONE, Task Force 58 began to reveal the scale of its overwhelming strength. In the four months to come, Mitscher’s rampages deep into enemy waters would demonstrate that land-based air cover was no longer the sine qua non of amphibious invasions and other major naval operations. Truk was neutralized and could be safely bypassed. A precipitous decline in the quantity and quality of Japanese airpower permitted more ambitious offensives across the Pacific. Many American officers felt a special obligation to liberate the people of Guam, a U.S. territory under occupation since the start of the war. New doubts were raised against MacArthur’s proposed alternative in the south. The fast carriers could not be deployed to their best advantage in the confined waters between the Dutch East Indies and the southern Philippines. Asked to review the competing views emerging from the Pacific, the Joint Chiefs of Staff planners came down squarely on the side of continuing two parallel offensives, and stressed that the northern line of attack offered the shorter route to Tokyo. Nimitz and his officers reconsidered their objections and finally accepted that the Marianas operation was feasible. On February 29, the operation was designated FORAGER, and planning got underway at CINCPAC headquarters.

On March 11, the Joint Chiefs published a new schedule of operations against Japan. MacArthur’s long jump to Hollandia, on the north central coast of New Guinea, would be completed on April 15. Truk would be kept under pressure by regular air raids and bypassed. The Marianas would be taken by June 15, Palau by September 15. The deadline for the invasion of Mindanao was set for November 15, though that date was merely a placeholder. The return to the Philippines might become feasible earlier, depending on the course of events. That took care of the entire program against Japan for the calendar year 1944.

Nimitz and MacArthur had not yet met in a face-to-face command summit. The latter had routinely declined every invitation to travel outside his theater. The admiral, sensitive to protocols, could not invite himself to Australia. MacArthur was senior to him by a long stretch. (The general had received his first star when Nimitz was a commander.) When the CINCPAC asked for permission to tour naval commands in the Southwest Pacific area, the request was quickly granted, and Nimitz was invited to visit Brisbane. He arrived by seaplane on March 25 and found MacArthur and a retinue of staff officers waiting at the dock.

Nimitz and his flag lieutenant, Arthur Lamar, socialized pleasantly with the MacArthur family. They brought candy and orchids from Pearl Harbor and a silk outfit for the general’s six-year-old son. Their meetings the next day, Admiral Kinkaid recalled, were stiffly formal but largely constructive. There was only one strained moment, when Nimitz called MacArthur’s attention to a clause in the latest directive from the Joint Chiefs. The two theater commanders were to consider a contingency plan, to be activated in case Japanese resistance suddenly weakened, to move directly toward Formosa and the China coast. MacArthur, who never tolerated any talk of circumventing the Philippines, let loose a gale of indignant rhetoric. In a private report to King, Nimitz recapitulated the exchange that followed:

Then he blew up and made an oration of some length on the impossibility of bypassing the Philippines, his sacred obligations there—redemption of the 17 million people—blood on his soul—deserted by American people—etc. etc.—and then a criticism of “those gentlemen in Washington, who, far from the scene, and having never heard the whistle of pellets, etc., endeavor to set the strategy of the Pacific war”—etc. When I could break in I replied that, while I believed I understood his point of view, I could not go along with him, and then—believe it or not—I launched forth in a defense of “those gentlemen in Washington” and told him that the JCS were people like himself and myself, who, with more information, were trying to do their best for the country, and, to my mind, were succeeding admirably.³⁵

MANY ORDINARY JAPANESE BEGAN to suspect that they had been misled, and that the tide of war had turned against their forces in the Pacific. The pattern of official news releases defied logic and common sense. Imperial announcements referred confusingly to glorious victories and strategic withdrawals. New “defense lines,” always closer to the homeland, must be held at all costs. Military experts, writing in the newspapers or speaking on the radio, soberly explained that Japan’s grand strategy was to “draw the enemy in”—to allow him to come closer to the homeland, where at last he could be crushed with one blow. Courageous Japanese troops defended advanced positions to the last man, perishing all together like gyokusai—“smashed jewels.” Whenever a new position was captured by the enemy, it was confidently explained that defeat had been expected, and preparations had been laid in place for it. Official reporting about the war situation, wrote a Japanese American woman who spent the war years in Japan, “involved such obvious contradictions that even the more simple-minded listeners became doubtful.”

Everyone who could think at all realized that the country was in a more and more desperate state, its back to the wall. When it became impossible to hide the truth longer, the broadcasters would announce a battle or an island lost, and each time they did so the program was ended with music. It was always the same—the sad, sweet strains of Umi Yukaba, a well-loved old song. All over the nation people would bow their heads while someone quietly turned off the radio. The conviction of ultimate defeat had become widespread but everyone was careful not to speak his opinion; each carried on silently lest his doubts prevent another from doing his best.³⁶

But they wondered, and worried. Aiko Takahashi, a young woman living in Tokyo, mused in her diary, “What exactly is happening? It’s like picking colors in the dark, and for better or worse, there is no criticism of the government. It simply makes us uneasy.”³⁷ The regime generally permitted accurate reporting of the progress of the war in Europe, and it was evident that Japan’s Axis partners were not faring well against the Allies. The surrender of a quarter of a million German troops at Stalingrad was reported in Tokyo the same week that the Japanese people learned that their army had abandoned Guadalcanal. The straightforward accounts from Europe were contrasted with the vague, shifting, and often contradictory narrative about the war in the Pacific. In September 1943 came news that the Italian government of General Pietro Badoglio had capitulated to the Allies. Since the previous July, when the dictator Benito Mussolini had been deposed and arrested, press reports had assured the Japanese people that the new government would carry on fighting for the Axis. Now Allied forces were pouring into southern Italy, and the German army was in retreat. At a newsstand in downtown Tokyo, observed the diarist Kiyoshi Kiyosawa, “people are standing in long, long lines, and they are eager to buy newspapers. It appears they received a considerable shock.”³⁸ Japanese military authorities took to the airwaves and editorial pages and confidently asserted that the fall of Italy was a piece of good fortune, for the Third Reich had finally “thrown off the burden called Italy and an indestructible resistance would now be possible.”³⁹ Kiyosawa lamented the stupidity and shortsightedness of official propaganda. The state-controlled news media was playing fast and loose with whatever remained of its authority and credibility. “It appears that the Japanese newspapers do not even have the common sense and logic of elementary school students.”⁴⁰

Regime-sanctioned slogans and themes extolled the indomitable power of Yamato Damashii—the collective Japanese fighting spirit. Even if the enemy possessed material superiority—and in 1943, hints of America’s enormous industrial output were seeping into news accounts—Japanese workers and warriors were capable of arousing themselves to new peaks of devotion and self-sacrifice. In the end, no matter what the cost, the transcendent spirit of the Japanese must prevail. When American forces stormed the Aleutian island of Attu in May 1943, a Japanese garrison of 2,650 soldiers fought a savage battle and died almost to the last man. Their cause had been hopeless—General John L. DeWitt’s 7th Infantry Division assault troops had outnumbered the defenders by five to one—but they chose collective death over the ignominy of surrender. In Japan, the annihilation of the Attu garrison was inscribed as an important moral victory. It had proved that the Japanese warrior was willing to give everything for his country. The term gyokusai, first coined to describe the mass combat deaths and suicides at Attu, became increasingly familiar as one Japanese island garrison after another perished in similar fashion. Leading Japanese newspapers reported that “the heroic spirits of Attu” literally rose from the dead, took corporeal form, and renewed the fight against American forces when they landed on Kiska island three months later. “Foreign reports reveal that the American forces fought intensely and bitterly against this army of spirits over a period of three weeks,” the Japan Times and Advertiser reported on August 24, 1943. “In the South Pacific sector, too, spirits of the Japanese troops have tangled with the enemy, causing many of them mental derangements and others to kill themselves as a result of nervous breakdown and morbid fear.”⁴¹

By contrast, noted the Japanese commentators, the individualistic and luxury-loving Americans could never hope to emulate Japan’s feverish devotion to victory in the Pacific. They wanted only to survive and go home alive. Long before they reached Japanese shores, the Americans would tire of the war and ask for terms of peace. So it was said. The chief of the Board of Information made the case at a speech in Yokohama in May 1943: “Soldiers are not tools but spirit! They are souls! American soldiers are crudely made and over-produced.”⁴² When General George S. Patton slapped an American soldier in Sicily in August 1943, the incident prompted excited commentary in the Japanese press. Here was incontrovertible proof that the enemy’s fighting spirit was flagging! Since the United States was a democracy, its people would sooner or later bring pressure to bear on their leaders to seek a peaceful accommodation with Japan. Time was on Japan’s side, because “conditions within America do not allow for a long drawn-out war.”⁴³

Homefront civilians were forever being exhorted to arouse themselves to greater efforts, to unite and summon their collective spiritual strength for an impending decisive confrontation with the enemy. “The present situation is truly grave,” the Showa emperor told them, in an Imperial Rescript opening a special session of the Diet on October 26, 1943. “The Japanese people must fully display their total strength and thereby destroy the evil ambition of the enemy nations.”⁴⁴ The people replied with exceptional fervor. That fall, apparently without a word of complaint, the Japanese government abolished holidays and weekends. There would be no more days of rest; the people would “return their holidays to the emperor.”⁴⁵ Labor drafts were steadily expanded into new categories. Unmarried women and university students were conscripted into civil defense work and war industries. Progressively younger children were drafted into the workforce: in September 1943, all girls over the age of fourteen; the following April, all children of both genders over the age of ten. Although the American air raids were still more than a year away, measures to evacuate children from the cities were initiated in October 1943. Military conscription was expanded. In the fall of 1943, college students aged nineteen or older (except those majoring in science or engineering) were called to service. Middle-aged men as old as forty-five were inducted. The send-off ceremonies were every bit as elaborate and well attended as those of 1941, but the mood had turned perceptibly darker. People had seen too many wooden boxes wrapped in white cloth, containing the cremated remains of dead soldiers and sailors repatriated from the war zones. Aiko Takahashi, the young woman in Tokyo, told her diary that the send-off processions had become “rather pathetic affairs. . . . As I look at them, I have the wrenching thought that today once again, a funeral of living people is passing and youngsters with their sleeves rolled up are being sent off to die.”⁴⁶

ALMOST EXACTLY A YEAR after Isoroku Yamamoto’s aircraft was shot down over Bougainville, his successor as commander in chief of the Combined Fleet met a similar fate. Wary of Mitscher’s rampaging carrier task force, Admiral Koga had made up his mind to transfer his headquarters ashore, from the battleship Musashi (then at Palau) to the town of Davao on Mindanao. On the night of March 31, Koga and most of his senior staff officers and cryptographers boarded three Kawanishi flying boats off Babelthuap and took off for the 650-mile flight. The big four-engine planes flew into a tropical witches’ brew off Cebu, and two of the three went down. Koga’s plane simply disappeared, and none of the aircrew or passengers was ever seen again. A plane carrying Vice Admiral Shigeru Fukudome, Koga’s chief of staff, ditched at sea about six miles off the coast. The vice admiral survived and managed to get ashore, where he was promptly taken prisoner by Filipino guerrillas. His captors found a cache of highly classified documents on his person. These included a signal book, an updated codebook, and a copy of “Plan Z,” the Combined Fleet’s master operational plan for a fleet action in the western Pacific. Fukudome was returned unharmed to local Japanese army forces under circumstances that remain obscure, but the captured documents were handed over to the Americans. A submarine spirited the precious intelligence to MacArthur’s general headquarters in Brisbane.

Realizing that the crown jewels had fallen into the enemy’s hands, the Imperial Japanese Navy moved quickly to promulgate a new edition of the codebook and to write a new operational plan for the looming battle. Koga’s death was hushed up until May, when a new commander in chief was appointed. This was Admiral Soemu Toyoda, a torpedo specialist who had graduated from Etajima in 1905. Toyoda established his headquarters in an anchored flagship (the light cruiser Oyodo) in Tokyo Bay. A new plan of battle, designated “A-Go,” was prepared in Tokyo by the Navy General Staff and distributed the same week.

A month earlier, Tokyo had ordered another major fleet reorganization, the fourth since the attack on Pearl Harbor. The main striking force of the Japanese navy was redesignated the “First Mobile Fleet” (Dai Ichi Kido-Kantai). Its two principal elements were the Third Fleet, which included nine aircraft carriers arrayed in three task groups; and the Second Fleet, consisting of most of the navy’s major surface warships, including the battleships and heavy cruisers. Vice Admiral Ozawa was “fleeted up” to command the entire force, while simultaneously retaining direct command of the Third Fleet, with Taiho as his flagship. The Second Fleet, commanded by Vice Admiral Takeo Kurita, was to function as a subordinate screening force, although Ozawa might send it to engage the enemy in a conventional naval gunnery battle if and when opportunity offered. The Combined Fleet had been reshuffled several times before, but the creation of the First Mobile Fleet took the unprecedented step of placing the carriers at the nucleus of the force, with the battleships relegated to a subsidiary role. For the first time in the history of the Imperial Navy, a carrier admiral would exercise tactical command over the battleships.

Traditions, dogmas, and cultural norms had long stood in the way of this necessary reform. Apart from the fixed ideas of the battleship adherents, it was simply not the way of the Imperial Navy to place a junior admiral over a senior admiral in the command chain. Before August 1943, when Kurita relieved Vice Admiral Nobutake Kondo in command of the Second Fleet, Ozawa (Etajima class of 1909) could not be placed above Kondo (class of 1907) without causing severe loss to the latter’s “face.” It was necessary that Kondo be retired or transferred to other duties before Ozawa took command of the mixed force. Kondo was given command of the China Area Fleet, and he held that position until May 1945.

Throughout the spring of 1944, Ozawa’s fleet moved between Singapore and anchorages in the southwestern part of the Sulu Sea (between the Philippines and Borneo). His movements were largely dictated by three considerations, all markers of the Japanese navy’s deteriorating fortunes. First, he needed to keep his flight decks in close proximity to terrestrial airfields, because his inexperienced pilots could not always be trusted to land safely on the carriers. Second, the fuel shortage, which had been exacerbated by the sinking of tankers, forced the fleet to operate near its sources of oil in Borneo and Sumatra. Third, the escalating threat posed by Pacific Fleet submarines kept Ozawa’s fleet besieged in shallow and enclosed waters.

In early March, the Japanese carriers anchored off Singapore and their air groups dispersed to the region’s various airfields for training exercises. The early results were not encouraging. Many of the new pilots had first climbed into the cockpit of a trainer aircraft less than six months earlier. The powerful new Suisei and Tenzan bombers were a bit too hot for many of the green pilots to handle, even from terrestrial airfields. They had received little or no gunnery or navigational training, and were obligated to follow a squadron or division leader who knew where he was going. “The training was intense, and almost every day there was an accident,” said maintenance officer Hiroshi Suzuki of the Zuikaku. “There were a lot of crashes because the pilots and mechanics were mostly rookies, and they were flying new aircraft. Besides that, the weather in Singapore was usually very bad, and we had a lot of rain.”⁴⁷

In the second half of March, Carrier Division 1 and two divisions of the Second Fleet moved about a hundred miles south to Lingga Roads, off the coast of Sumatra. This anchorage, near the Palembang oilfields, was spacious enough for the carriers to conduct basic flight operations. Its narrow entrance channels could be policed against submarine infiltration. Here the rookie pilots would practice their carrier landings, but when the planes moved from airfields to flight decks, operational losses rose to fearsome proportions. Minoru Nomura, air officer on the Zuikaku, recalled watching as a Tenzan dropped a dummy torpedo and roared low over the carrier’s bow, just clearing it. As the crew heaved a sigh of relief, “the pilot began to bank for a right turn. At that instant his right wingtip made contact with the water. Plane and crew disintegrated instantly. It was over in a moment.”⁴⁸ The powerful new torpedo bombers and dive-bombers were unforgiving when flown at low speeds. Many dropped into the sea astern of the carriers as they made the final turn in the landing approach. Planes failed to snag an arresting wire and went careening into the island or over the side. Midair collisions occurred directly overhead. “This is self-destruction air warfare,” a Junyo pilot ruefully commented.⁴⁹

Admiral Shimada issued the revised Plan A (“A-Go”) in Directive No. 373 on May 3. The plan envisioned a fleet battle in one of two “decisive battle” zones—in waters off Palau or the western Caroline Islands. The First Mobile Fleet would be supported by nearby elements of the shore-based First Air Fleet. If the Americans attacked the Marianas, Ozawa would remain in local waters in the hope of “luring” the enemy fleet south. Thus the plan depended on the Americans’ accepting the proffered bait. “The decisive battle,” Shimada wrote, “will be fought as close as possible to the forward base of our mobile fleet.”⁵⁰

The battle must be fought in those southern waters because the fuel situation required Ozawa to stay within easy reach of Dutch East Indian oilfields. The relentless attrition of oil tankers, chiefly credited to the American submarines and their now-reliable torpedoes, had chained the fleet to the refineries at Tarakan, Balikpapan, and Palembang. The immediate fueling situation could be alleviated by allowing the ships to take on unrefined Borneo petroleum, which was pure enough to drive the engines, but volatile and dirty. Ship’s engineers detested the stuff because it left layers of filthy sediment in the boilers, and it greatly increased the risk of explosions and fires.

Until mid-May 1944, therefore, orders issued by Admiral Toyoda pursuant to A-Go did not admit the possibility of a fleet battle off the Marianas. If the Americans attacked Saipan and its neighbors, land-based naval air forces on the islands must attack and destroy the enemy fleet. These air units could be reinforced from the homeland, by sending planes to stage through the “Jimas” (the Bonin Islands, including Iwo and Chichi). Meanwhile, ground forces on Saipan were being strengthened and reinforced, and the local army commander, Lieutenant General Yoshitsugu Saito, had confidently pledged to repel any amphibious invasion. Still, uneasy questions remained: What if the Fifth Fleet declined to chase the enemy into the southwest? Could Japan risk the loss of territory so near the homeland without committing its fleet to make a stand?

The answer was given on May 11, when the Naval General Staff rescinded its prohibition against pumping crude oil directly into Japanese ships. In the third week of May, most of the First Mobile Fleet anchored at Tawi Tawi in the Sulu Archipelago, between Mindanao and Borneo, and shuttle tankers from Tarakan filled every ship to capacity. But Tawi Tawi was no good for flight training because there were no suitable airfields in the area. Ozawa, appalled by the deadly losses incurred during carrier flight operations, had decided not to continue such training exercises: “I cannot bear to lose the lives of any more of my men in these accidents.”⁵¹ Though urgently in need of more preparation, the air groups were largely idled in the last month before the Japanese fleet was compelled to give battle. According to Lieutenant Commander Zenji Abe, air group leader on the Junyo, the prevailing breeze was not strong enough to get the Suisei carrier bombers aloft even at the ship’s maximum speed. “So we couldn’t take off,” said Abe, “and had no training after we left Japan from early May until 19 June, when the big American task force came over to the Marianas to attack Guam and Saipan. . . . So for forty or fifty days, we had no training flights at all.”⁵²

Admiral Matome Ugaki, the erstwhile Combined Fleet chief of staff, who had functioned as Yamamoto’s right hand and had survived the aerial ambush that killed the former commander in chief in April 1943, had recovered from his wounds and was back in the fleet. He had been appointed commander of the First Battleship Division of the Second Fleet, a command that included both the Yamato and the Musashi. He was glad to be back at sea but was discouraged by the state of training, both in his own battleships and throughout the fleet. In the privacy of his diary, Ugaki wavered between despondency and mystic appeals to bushido, the warrior spirit. He regarded the strategic plans to be basically sound, but doubted that the First Mobile Fleet possessed the means to execute them. Meeting the enemy at sea seemed a daunting prospect because “the only possible consequence is to become easy prey to enemy planes and submarines. The further we venture out, the more we shall be beaten. It’s like one can’t help getting soaked if one goes out in the rain without a coat or umbrella.”⁵³ But the fleet could not afford to procrastinate, he observed a few days later, because the Americans were gaining strength all the time.⁵⁴ On April 27, he went aboard Ozawa’s flagship, the Taiho, to participate in tabletop war-gaming exercises. In Ugaki’s judgment, the game supervisors allowed assumptions that were unrealistically favorable to the Japanese side. He judged that the navy’s fixation on a decisive battle amounted to an irrational obsession, and wondered why “they don’t give enough consideration to attacking enemy elements easy to destroy.”⁵⁵

Ugaki’s entries in April and May 1944 are filled with conflicting judgments and sentiments. The tone leaves no doubt that he knows his country is defeated, but he will not say so outright, even in his diary. On April 24: “The enemy’s present strength is just like a raging fire, so irresistible that a small amount of water can hardly put it out.”⁵⁶ His pessimistic insights ring true, while his efforts to arouse his own spirits are feeble and half-hearted: “However, at the same time the enemy, too, may not be as good as we think. It may turn out all right, if and when we fight with them.”⁵⁷

If the decisive battle was to be fought off the Marianas, a scenario made feasible by the Tarakan crude, Ozawa would launch his strike at very long range, 300 miles or more to the westward of the enemy. The Japanese planes would attack the American fleet and then land on airfields in Guam. There they would replenish fuel and ammunition and continue flying sorties until the enemy fleet was annihilated. This use of “shuttle-bombing” from shore bases was at the heart of A-Go. Successive aerial hammer blows would fall on the American fleet from Ozawa’s carriers and the Marianas airfields. Ozawa would then unleash all his forces, including the battleships and other surface warships, to pursue the cripples and wipe them out. Shimada’s directive put across Tokyo’s absolute conviction in the plan: “Complete success is anticipated.”⁵⁸

Keeping the First Mobile Fleet out of the enemy’s reach appealed to the logic of “outranging,” a guiding tactical principle in Japanese planning and weapons systems. Outranging was a theme found in the design of the Imperial Navy’s many long-ranged aircraft, including the Zero and the G3M and G4M medium bombers. For all the problems and challenges of the Japanese aircraft industry in wartime, the new-generation carrier dive-bombers and torpedo bombers had significantly greater range than their American counterparts. (The Suisei flew so far that it was often configured as a reconnaissance plane, and had been employed on ultra-long-range missions since 1942.) Outranging had been employed to good effect in surface naval actions by the Type 93 (Long Lance) torpedo. The superbattleships, with their 18.1-inch guns, were designed to strike an enemy from beyond his effective range. Outranging and shuttle-bombing offered some theoretical promise of rectifying the disparity in force between the two opposing fleets. The Japanese were well aware that they were outnumbered. A May 9 intelligence estimate distributed by the Naval General Staff predicted that the U.S. fleet would sail with sixteen battleships, including eight of the fast new types. As for carriers, the estimate was eight large fleet types, ten light carriers, and approximately twenty jeep carriers.⁵⁹ Those figures were close enough to the mark.

The shuttle-bombing tactic embodied in A-Go was sound in theory, but it was an expediency chosen to spare the pilots the ordeal of landing on the carrier flight decks. It was a tactical concept selected, in part, to mitigate the deficiencies in the carrier pilots’ skills, and therefore pointed to the larger problem that would inevitably decide the outcome of the impending battle. The Japanese aviators were simply not ready to meet their adversaries on anything approaching equal terms.

The officers and pilots told one another that A-Go was a good plan that would deliver a badly needed victory for Japan. By some reports, the Japanese officers and crews were confident in the week before sailing for battle. High hopes were invested in the land-based First Air Fleet, also called the Base Air Force, commanded by Vice Admiral Kakuji Kakuta and headquartered on Tinian. Shore-based naval air had always been closely integrated into the Imperial Navy’s fleet command structure. Naval leaders had retained a lingering belief, despite plenty of contrary evidence, that their land-based bombers could deal devastating blows to enemy fleets at sea.

On paper, the First Air Fleet had a huge complement of about 1,750 planes, but that figure overstated Kakuta’s actual strength in the Marianas by more than threefold. His command suffered from poor ground support facilities, a shortage of qualified mechanics and aircrews, and scarcities of spare parts, ammunition, and fuel. The submarine threat in waters between the home islands and the Marianas had rendered it impractical to crate the planes and send them by sea. Most were flown in from Japan by unseasoned and undertrained pilots through airfields in the Bonin and Volcano archipelagoes. Operational accidents, including crashes and navigation failures, accounted for crippling losses of both aircraft and aviators. On the eve of the American attacks, Kakuta was thought to have about 435 aircraft in flyable condition.

Quite apart from the amphibious campaign and the ubiquitous depredations of the fast carriers, American submarines were cutting the internal tendons of the Japanese war effort. The undersea campaign had grown steadily more potent since mid-1943. On June 7, the Naval General Staff announced that 210,000 tons of shipping had been lost in May. Since the beginning of April, the Japanese had lost five destroyers, six tankers, four troopships, and fourteen freighters. The Japanese fleet was constantly stalked, reconnoitered, harassed, and hunted by submarines operating from Pearl Harbor and Australia. Each of Japan’s superbattleships had caught a torpedo fired from an American submarine. The Yamato had taken a hit from the Skate on Christmas Day, 1943; the Musashi was torpedoed by the Tunny while departing Palau on March 29. Though the huge ships could easily withstand single torpedo hits, the damage had to be repaired in Japanese home waters, requiring long voyages and absences that consumed scarce fuel and deranged the formations and plans of the Japanese commanders.

In April and May 1944, every major movement of Ozawa’s fleet was tracked and reported by submarines. American boats skulked off Lingga Roads near Singapore, observing and reporting as Japanese ships entered and exited. They assailed tankers shuttling between Tarakan and Tawi Tawi, off the northeast coast of Borneo. Freighters and troop transports carrying reinforcements from China went down between Shanghai and Manila. The veteran light cruiser Yubari was destroyed in an attack by the Bluegill off Halmahera on April 26. On May 6, the Crevalle sank the big tanker Nisshin Maru off Borneo.

As the First Mobile Fleet gathered in Tawi Tawi in mid-May, it was reconnoitered by the submarines Lapon and Bonefish. On May 15, the Bonefish (Commander Thomas W. Hogan) damaged a tanker and sank one of its escorting destroyers, the Inazuma. His sighting reports confirmed that a major fleet rendezvous was in progress at Tawi Tawi, and two more submarines, the Puffer and Bluefish, shadowed the approaches during the following week. The Puffer’s attack on the seaplane tender Chitose on May 22 resulted in two hits, both apparent duds. On June 5, she sank two valuable cargo ships, the Takasaki and the Ashizuri. In a three-day rampage off Tawi Tawi between June 6 and 9, the submarine Harder (Commander Samuel D. Dealey) ran up one of the most extraordinary scores in the history of submarine warfare. On the sixth, just south of Tawi Tawi, the Harder sank a charging destroyer, the Minazuki, with a down-the-throat shot. She repeated the performance the following day with another down-the-throat shot, sinking the destroyer Hayanami. On the eighth, Dealey found the veteran destroyer Tanikaze in his sights and sank her with a four-torpedo salvo.

Destroyers, having been conceived and designed as submarine hunters, were the submariner’s natural enemy. The destruction of so many such vessels was a source of singular satisfaction to the American submarine crews. It also weakened Ozawa’s antisubmarine screening force, a factor that would prove decisive in the impending carrier battle.

While fretting over the location and intentions of the Fifth Fleet, the Japanese army and navy were also forced to react to MacArthur’s offensive up the northern coast of Dutch New Guinea. In a finely choreographed series of sealifts, Admiral Kinkaid’s naval forces had put army amphibious troops ashore on Hollandia (April 21) and Wakde Island (May 17). In each case the invaders had secured airstrips in a matter of days, and USAAF planes had begun exerting pressure on points farther west. In May, MacArthur was driving toward the “Bird’s Head” peninsula of northwestern New Guinea. On May 20, naval and amphibious forces began maneuvering into position to attack Biak Island. About 10,000 Japanese troops held the island, including an elite 222nd Regiment that had been battle-tested in China. The initial landings caught the defenders by surprise, but in the drive to capture Biak’s three airfields, the fighting bogged down and both sides suffered heavy losses.

The Japanese deemed Biak a critical outpost in the defense of the southern inner perimeter. The island’s airfields were an important node in the forthcoming battle envisioned in A-Go. If they were lost, they would be turned against the Japanese. USAAF B-24 bombers would launch a lethal bombing campaign against western New Guinea, Palau, and the southern Philippines. They would command the sea-lanes east of Mindanao. Therefore, Admiral Ugaki wrote in his diary, “Biak Island is the most critical crossroad of the war.”⁶⁰ He suspected that the Japanese army was less than sanguine about its prospects on Biak and therefore was inclined to abandon the garrison as it had on so many other islands. But the navy could not countenance such a retreat, and Ugaki urged Ozawa to pressure Tokyo to send all available air and troop reinforcements into the area.

On May 29, Admiral Toyoda of the Combined Fleet ordered “Operation KON” to commence. KON involved the massing of land-based air units and troops behind the lines, and plans to move troops quickly into disputed areas to respond at the point of attack. It was thought that the U.S. Fifth Fleet might even be lured south to intervene in the campaign, an event that would activate A-Go and perhaps lead to a war-altering decisive naval victory. Kakuta was ordered to fly some 480 planes into airfields at Sorong and Halmahera island, from which they could attack the American amphibious fleet off Biak. Encouraged by reports from the island, where the Japanese garrison had repeatedly driven American forces back in counterattacks, the Imperial General Headquarters resolved to send in land reinforcements. Short of transports, the navy resorted to the old tactic of using destroyers and cruisers as makeshift troopships, which staged at Sorong. The flotilla, committed by Vice Admiral Naomasa Sakonju, would fetch 2,500 troops of the Second Amphibious Brigade from Zamboanga, Mindanao, on May 31 and land them on Biak three days later. Much of their weaponry and supplies would be towed in by barges.

Sakonju’s approach was detected on June 3 by Allied reconnaissance planes, and the flotilla was intercepted by a cruiser-destroyer force under Admiral Victor A. C. Crutchley. Heavy air attacks on June 8 sank one of Sakonju’s ships and damaged another. The Japanese warships, crowded with troops and in no condition to fight, withdrew in a running battle.

The failure of the first reinforcement mission only redoubled Ozawa’s determination to hold Biak. At Tawi Tawi, where divisions of ships were arriving from Lingga Roads, the staff pored over maps of Biak and western New Guinea and contemplated their options. Their prevailing view remained that the U.S. fleet would sail through the Palau islands to force a diversionary fight. Ugaki was keen to bring his heavy ships into action but reluctant to risk them to force the troop reinforcements through to Biak. On June 10, Admiral Toyoda ordered the battleships Yamato and Musashi, with the light cruiser Noshiro and six destroyers, to attempt another troop landing. Admiral Ugaki, who had admitted in his diary the previous morning that “I thought we had no choice but to give up Biak,” would command the combined force. He sailed from Tawi Tawi that afternoon and rendezvoused with additional forces in Batjan on June 11.⁶¹

With nothing to compare to the Americans’ sophisticated amphibious equipment and landing craft, the Japanese warships were forced to the expedient of carrying army landing craft on their decks—one each on the destroyers, two each on the cruiser and battleships. With the decks awkwardly burdened in this way, the ships could not come to action or use their guns to bombard shore positions. The troops would have to be landed before the squadron went into action; if the Americans interfered, this reinforcement mission was likely to end in another failure. En route, Ugaki admitted in his diary that he was “worried” about his anemic air cover, and considered that his ships were likely to come under punishing air attack before they could get troops ashore or bring their naval guns to bear on any enemy target: “I think that to reach the front line itself without air cover is awfully difficult, but I won’t grumble about it now. I shall do all that is humanly possible.”⁶²

The Biak operation was complicated by a series of indications that the Americans were staging a major operation north of the equator, probably aimed at the Marianas. Long-range scouting flights launched from Nauru had photographed a powerful American fleet anchored at Kwajalein and Majuro. A few days later, on June 9, another flight confirmed that the carriers and battleships seen earlier were gone. Where were they? Where were they going? On Toyoda’s flagship, still anchored in Tokyo Bay, the staff officers continued to believe that the Pacific Fleet’s next thrust would be aimed toward Palau and the southern Philippines. Only the staff intelligence officer, Commander Chikataka Nakajima, predicted that the enemy was headed to Saipan. On June 10, American carrier planes were sighted about 300 miles east of Guam. The next day came a powerful fighter sweep over Saipan and Guam. On the night of June 12, Kakuta’s headquarters radioed its best estimates of the situation off the Marianas, judging that the enemy force consisted of eleven carriers.⁶³

Kakuta’s early reports on the air fight over the Marianas were largely upbeat. Though he was losing a great many planes, he believed (or said he believed) his forces were inflicting proportionate damage on the American carrier forces. Air scouts had not yet detected Turner’s amphibious fleet, so the Japanese surmised that the action in the Marianas might be nothing more than a raid or a feint. Other air scouts had seen MacArthur’s transports passing through the Admiralties, indicating that a major landing might be at hand in the south.

At this moment, the strategic merit of the Americans’ double-headed offensive was plainly in view. The Japanese high command found itself frozen in place, awaiting clarity on the location and objective of the Fifth Fleet.

On June 13, with minesweepers working off Saipan’s leeward beaches and battleships bombarding the coast, it seemed clear that an invasion of that island was imminent. Without waiting for orders, Ugaki suspended the Biak operation and began clearing his decks of the landing craft. Impatient for orders to fall in with Ozawa, he weighed the pros and cons of rushing north on his own initiative. Two hours later, Admiral Toyoda suspended Operation KON and gave new orders: “Set Op A in motion for the decisive battle.”⁶⁴ The First Mobile Fleet filed out of the channel at Tawi Tawi. Ozawa set course at high speed for the Guimaras Strait in the Western Visayas region of the Philippines, where the fleet would take on 10,800 tons of fuel before sailing for the San Bernardino Strait. Ugaki’s force of battleships, cruisers, and destroyers broke off the attack on Biak and sped north at 20 knots, keeping Mindanao to port, intending to rendezvous with Ozawa in the Philippine Sea. The Biak garrison was left to make the now-customary last stand, and die to the last man.

^* The nickname was Oni Gawara in Japanese. Ozawa had suffered a facial injury while serving on a destroyer early in his career. The injury had immobilized his facial muscles, making it impossible for him to smile or show other emotions through facial expressions.