By June 1944, the Japanese Combined Fleet was betting everything on its next battle. A plan code-named Operation A-Go depended on roughly 500 land-based aircraft believed to be waiting in the Mariana Islands, ready to help tip a fight that carriers alone were not expected to win. Throughout that spring, aircraft factories in the Japanese home islands worked at full capacity to replace machines disappearing across the Pacific faster than anyone in Tokyo wanted to admit publicly. Workers built new torpedo bombers and shipped them south to island bases that stood, in theory, as a barrier between the American fleet and the home islands themselves.
Nobody in those factories tracked where any given machine ended up once it left the factory yard. One man, Vice Admiral Kakuta Kakuji, took on that task from a headquarters on the island of Tinian. He had to know how many had reached the fleet, how many were still flyable, and how many were already lost. By the time anyone else realized what he knew, the battle A-Go was meant to support had already slipped from their grasp.
For the last year of his life, Kakuta commanded Japan’s First Air Fleet, every land-based naval aircraft deployed across the Philippines and the central Pacific. Historians who later reviewed what remained of his reports to Tokyo tend to use the same word to describe them: dishonest. Kakuta had earned his position the traditional way, with carrier service dating back to raids on Ceylon and Dutch Harbor. He had been competent enough to be promoted to vice admiral in November 1942 after two years of commanding carrier divisions in battles that had mostly gone Japan’s way at the time.
Land-based air power was a different kind of command. It meant airfields spread across islands an admiral could not walk between. It meant relying entirely on what subordinate officers chose to report, and what Kakuta himself then decided to pass up the chain of command. In February 1944, that entire arrangement was tested for the first time, and it failed badly.
American carriers caught Kakuta’s newly arrived torpedo bomber units in the Mariana Islands on February 22 and 23. The aircraft had just completed the transit south from Japan and had not yet been given enough time to disperse or fortify. Vice Admiral Fukudome Shigeru, a colleague who wrote about the raid after the war, described the losses as amounting to complete annihilation. An exact count is impossible, because the unit filing the reports had just been destroyed, and destroyed units produce sparse paperwork in the days after an attack.
What is clear enough is that Kakuta spent the following months rebuilding a command that had, in practical terms, already been wiped off the map once. Much of what is known about how he managed that command comes from the kind of sources that survive from the paperwork of a defeated side: postwar interrogations of officers who served under him, speaking through interpreters to American interviewers mostly trying to reconstruct a force report that Tokyo never saw in its true form. It was assembled after the fact from men who had no particular reason to protect their old commander’s reputation by 1945. The picture that emerges is consistent enough to trust.
None of them describe him as a man deceived by his own optimism the way a junior officer might be. The portrait that appears is closer to someone who understood the math well and decided the fleet did not need to hear it. By June, replacements had filled the gaps enough that the Combined Fleet’s leadership relied on the Marianas garrison for that same number. The upcoming battle plan, code-named Operation A-Go, depended on Kakuta’s squadrons striking the American carriers first, weakening them before Vice Admiral Ozawa Jisaburo’s Japanese carrier fleet had to close the distance and risk a direct exchange.
It was Ozawa’s own staff who did the calculations, and they believed the number was accurate. Task Force 58 did not wait for that plan to begin. Starting June 11, American carrier aircraft struck Guam, Rota, Saipan, and Tinian day after day for a full week, destroying the same runways and parked aircraft before the two forces even came near each other. The aircraft that survived the February raid through dispersal fell this time on the ground in open revetments where crews had spread them out, after that earlier lesson had come too late.
By the time the invasion force reached Saipan on June 15, the airfields on Guam and Rota alone had lost nearly everything stationed there. Hangars and fuel depots burned, leaving only a fraction of what they had held two weeks earlier. The air fleet the entire opening of Ozawa’s operation depended on was destroyed before Ozawa’s carriers launched a single strike of their own. None of this was hidden from the men doing the flying.
Returning pilots knew exactly how few of their squadron mates had managed to take off with them in the first place. A base commander does not need a signal officer to tell him the runway has become a burning crater. He can count the wreckage from where he stands. What nobody above him knew was this: Kakuta himself knew all of it in detail, and what he sent to his own chain of command told a completely different story.
His reports to Ozawa described nearly 300 land-based aircraft still in the fight, inflicting damage on Task Force 58 enough to bring the two sides closer to even. None of that was true, and Kakuta had far better information than anyone reading his reports in the fleet. Ozawa had no way to verify the claim from his position hundreds of miles away while his carriers approached the Marianas. He built his opening moves on June 19 on the basis of support that had already burned a week before he committed to battle.
A man who has just received the number he was hoping for does not stop to ask who calculated it. Ozawa did not. On the morning of the 19th, American radar picked up Japanese aircraft heading toward Guam, and Admiral Raymond Spruance sent fighters to meet them, treating the contact as a real threat from a base that, by any honest accounting, should have had almost nothing left to send. It sent something anyway, whatever the garrison could get into the air.
None of that changed what happened over the water that day, as American pilots met wave after wave of Japanese carrier aircraft in an engagement that earned a nickname before the sun went down. When the day ended, Ozawa had lost the vast majority of the aircraft his carriers had launched that morning, along with two of his carriers to submarine torpedoes the same day. He still did not stop planning the attack even after that. Ozawa assessed the situation that evening and found roughly 150 aircraft still in his possession, a fraction of what he had started the day with, and by any reasonable reading of that number, the battle was already lost.
He continued anyway, convinced there were still hundreds more on Guam and Rota, ready to be called upon the next time he needed them. He did not learn until long after the battle was decided that the land-based support his entire plan depended on had ceased to exist in any meaningful sense a full week before he issued the order to attack, let alone on its second day. What happened to Kakuta took a little longer to unfold, and it happened near where the lie began. He returned to Tinian with what remained of his command, and when the American invasion of the island began in late July, he was the senior naval officer present, though not the one in charge of the land defenses.
That responsibility belonged to an army colonel running a garrison of roughly 4,500 men. No surviving message records Kakuta explaining to himself or anyone else the exact moment his optimistic reports shifted from mere hope to something closer to decision. That dividing line appears in no document that survived him, and it likely never existed as a single sentence anyone could have pointed to even when it mattered. He tried more than once to leave the island by submarine at pre-arranged rendezvous points with the naval forces he had spent his career serving.
More than once, the attempt failed. What remains of his final transmission is short and reads as dry as last messages tend to be once they are finished and there is nothing more to add. “I will make an attack with all forces under my command,” it says. “All secret documents have been destroyed.
This may be my last message. ” After that, Kakuta and his staff hid in a cave on the eastern coast of Tinian, and no one saw them alive again. American forces secured the island within days. His men buried him somewhere along that coast at a location no one recorded for anyone to find later.
No American report or postwar Japanese account identified the exact spot. What is missing from the records is not the fact of his death, which no one disputes, but the one detail that would allow anyone to stand where he was buried. Nobody wrote it down, or if someone did, that page did not survive the rest of that summer. Within a week of the island’s fall, Navy construction battalions were already working on the runway his aircraft had once used, repairing the very strip from which his air fleet had operated during the year it was a fighting force worthy of the name.
Fifteen hundred of them landed with the invasion forces and started work before the fighting had even finished clearing the island. They continued building through the rest of that year and into the next, moving nearly a million cubic yards of earth and coral by hand and bulldozer, adding runway after runway to the one that existed when they arrived. The airfield engineers planned the entire base on a street grid, complete with named streets the construction crews designed as a joke after Manhattan. By the following spring, that same stretch of coral and cleared cane fields held four runways, each nearly two miles long, feeding hardstands built for hundreds of heavy bombers.
It had become by far the largest airfield anywhere on Earth. In the final months of the war, bombers took off from it against the home islands in numbers that would have meant nothing to the man who once commanded the ground beneath it, because he was no longer alive to compare it to anything. Just after midnight on August 6, 1945, searchlights illuminated that same runway as a single aircraft moved into position and lifted into the air, carrying a weapon that would help end the war Kakuta’s command had spent the previous summer failing to hold back. Ground crews had spent the night before loading it into the bomb bay from a pit built for that very purpose, on an island Kakuta had once commanded and been trusted to defend.
The date fell almost exactly one year since his last signal had gone silent. He was buried somewhere along that coast in ground no one had ever marked. The runway he never had the chance to defend outlived every man who ever flew from it on either side of that summer’s battles.