In January 1942, at a shipyard on the Delaware River in Camden, New Jersey, an altered set of drawings changed the fate of Hull 59. The hull itself remained where it was, but everything meant to stand on top of it was rewritten. Started the previous May as a light cruiser named Amsterdam, the ship’s gun turrets were removed from the plans and replaced with a hangar and a 552-foot flight deck. Renamed Independence, it became the first of nine cruisers converted into aircraft carriers at that same yard before the summer ended.
Nothing below the waterline had to be discarded, which was the entire purpose of the conversion and the only reason the work could be finished in the time available. The project was not secret in any practical sense. The work took place in a commercial yard beside a New Jersey river, watched by tens of thousands of workers who went home each shift and talked about their day. It appeared in the trade press and in the shipping columns of ordinary newspapers.
What the Japanese Navy made of this, and of everything else the United States was doing between December 1941 and June 1942, is a question with an answer more awkward than simple ignorance. Intelligence for the Pacific War belonged to the Third Bureau of the Naval General Staff in Tokyo, organized by target. One desk watched America, another China, another Europe, and a fourth handled signals. Behind them stood a network of naval officers posted to embassies worldwide and a large listening organization at Owada that took down foreign radio traffic around the clock.
From spring 1942, the bureau was headed by Captain Kanji Ogawa, formerly the Navy’s man in the Washington embassy and chief of the America desk, the officer who had assembled the espionage that fed the Pearl Harbor plan. On paper, this was a serious outfit run by a serious professional. In practice, it sat several steps below the operations division, and everyone in the building understood that. The strongest officers went to operations, and intelligence took what was left, often on short postings that gave nobody time to become expert in anything.
By early 1942, the bureau’s view of America amounted to almost nothing. The human network the Navy had spent years and a great deal of money building inside the United States was gone before the shooting started, undone mostly by its own carelessness and shut down by American counterintelligence through the autumn of 1941. A parallel scheme to run radio interception teams out of Mexico collapsed within weeks when the Mexican government sided with Washington. Japanese scholarship on this period is blunt about the result: by the early weeks of the war, Tokyo had almost no dependable information from human or electronic sources about where American warships were or what they were doing.
What remained was a short list of sources. Radio traffic came in volume and told the listening stations something about enemy activity without telling them much about enemy strength. Printed matter reached Tokyo weeks or months late through neutral capitals. American commanders announced results after actions, which the Navy heard the same way anyone with a shortwave set heard it.
And there was one more source: whatever the United States government chose to say out loud, a great deal of it that winter, including President Roosevelt’s January 6 address to Congress naming production targets of 60,000 aircraft, 45,000 tanks, 20,000 anti-aircraft guns, and 8 million deadweight tons of merchant shipping. The first hard assessment the bureau had to make was about damage. Strike photographs from December 7 and submarine reconnaissance gave an account of the battleships at Pearl Harbor. That count was broadly right, and Tokyo built its picture of the Pacific Fleet on it for months afterward.
But the assessment counted ships only. The Pearl Harbor Navy Yard, its machine shops, its dry docks, and the fuel standing in tanks above the harbor went unassessed because none of them had been targets, and nothing in the assessment process obliged anyone to ask what a yard could put back together. A salvage division was stood up a week after the attack and given to Captain Homer Wallin at the start of January. By the end of February, the yard had returned the battleships Pennsylvania, Maryland, and Tennessee to service or made them fit to steam to the mainland, along with three cruisers, two destroyers, a seaplane tender, a repair ship, and a floating dry dock.
Nevada came off the bottom on February 12, went into dry dock six days later, and sailed for Puget Sound under her own power on April 22. Getting her up had taken more than 400 individual dives and over 1,500 hours in the water. California was afloat in March, and West Virginia came up in the late spring. Then the raids began, and for half a year they were the best information the Japanese Navy would get about its enemy.
Carriers hit the Marshalls and the Gilberts on February 1, Rabaul on the 20th, Wake on the 24th, Marcus on March 4, and Lae and Salamaua on the 10th. All were small, all ended inside a morning, and in every case the raiding force had gone before anything could reach it. Nobody held ground and nobody followed up. The Bureau drew a conclusion from the pattern, and the conclusion was correct: an enemy who raids and runs is an enemy who cannot yet do anything heavier.
Where the judgment failed was not in its reasoning but in its timing. It described the American fleet in February 1942 with real accuracy, but the state of that fleet in February was the least useful fact on offer. The conclusion reached paper fast. At a liaison conference on February 28, the government and the high command fixed the geographic limits of what Japan meant to hold.
On March 7, they agreed on the outline that would govern the war for the next 18 months. Its first paragraph set the aim plainly: hold the gains, prepare for a long war, and work toward the end of American willingness to keep fighting. That document was about the enemy’s spirit, not his capacity, and the two are not the same thing. Twelve days later, the army chief of staff reported to the emperor on future operations, carrying a plan that cut the southern army from roughly 450,000 men to 250,000, down to seven divisions, 44 independent battalions, and about 50 air companies.
The reasoning behind the cut was that the southern campaign had finished, the war against the United States could now be handed to the navy, and the army should turn back to the continent and unfinished business in China. The plan provided for garrison duty and light policing in the occupied territories, on the assumption that nobody would be coming ashore against any of it. The reconnaissance in the structure shows the physical limits. On March 10, a Type 2 flying boat went out to look at Midway and did not come back.
It was the last aerial look at the atoll the Japanese Navy would get. A second flight was scheduled for May 31 and was called off. Whatever the bureau wrote in the interval about American strength in the Central Pacific, it wrote without a photograph taken in the previous 11 weeks. Here is the part that is genuinely hard to sit beside.
The Naval General Staff already owned a correct assessment of American industrial strength, and had owned one for years. It was the reason for the war. When Nagano Osami argued in 1941 that Japan had to fight and had to fight soon, his case rested on the American Naval Building Law of 1940 and what that law would do to the balance between the two fleets. The ratio was as good as it would ever be.
Within two years, on the Navy’s own arithmetic, it would be gone. This was not a finding turned out by an intelligence desk and waved away by operations. It was the operations argument itself, made at the very top of the Navy, and it was right. For six months after Pearl Harbor, almost nobody carried it forward.
The staff papers of that spring are about perimeters, about willingness, about where the next landing should go. The arithmetic that had justified starting the war is nowhere in any of them as a clock still running, and the second American Navy going up on the building ways at Camden, Newport News, and Quincy is not in them either. On April 18, 16 American bombers came over the home islands. They were army medium bombers that had come off a flight deck, and nothing in the Japanese picture of enemy strength had made room for such a thing.
The material damage was small, but the effect on the high command was not. Two answers were available. The navy could push its patrol line further out, or it could hunt down the ships that had launched the bombers. An operation already being argued over inside the combined fleet acquired the whole weight of that argument overnight.
The army, which had spent March planning to send men home, ordered its forces in China at the end of the month to take the airfields the raiders had been heading for. Three days earlier, the navy had put its second-stage plan on paper, and its aim was to bring the enemy’s main fleet to a battle it could not walk away from. Everything now hung on finding the American flight decks. At the Coral Sea in early May, the Fifth Carrier Division, which the navy regarded as its weakest, fought what Tokyo took to be the best the enemy had to a standstill.
The lesson drawn was that the first and second divisions would have no trouble whatever. Something further carried through the staff work that followed, which the Japanese war history series sets down without softening it: a view formed inside the navy that even with Midway taken, the American ships would simply not turn up. Japanese aircraft losses at the Coral Sea ran to three times what the public announcement admitted, and the Fifth Division did not go to Midway at all. Its absence was treated as a matter of no consequence.
On the atoll itself, the assessment gave 750 Marines and 500 laborers, six 5-inch guns, 12 3-inch anti-aircraft guns, and 96 machine guns across two calibers. The atoll in fact held about 3,000 men and 121 aircraft. The Japanese account of the battle, written afterward by the men who had fought it, puts the wider trouble in a single sentence: “We never knew to the end where or how many enemy carriers there were. ”
The files that would settle who wrote which of these judgments and who saw them largely do not exist.
Japanese intelligence organizations burned tens of thousands of documents in the last weeks of the war, and the Third Bureau’s own product from the first half of 1942 is among the missing. What survives is the trail of decisions taken on the strength of it, plus the questioning of men remembering across three and four years. Six days after Midway, on June 13, the Navy Section fixed its immediate guidance for operations. The paper was a light revision of the April plan.
It held to the offensive in the Pacific and put the earliest possible American counterattack in 1943, with none anticipated in what remained of the current year. In late July, the High Command issued its wider outlook on world affairs and moved the date again to the end of 1943 and after. On August 7, the First Marine Division landed on Guadalcanal, inside a fortnight of that outlook, on an island where Japanese labor troops had just finished an airstrip. When Fukudome Shigeru, who ran the first section of the Naval General Staff throughout the period, was asked in 1945 how the staff had judged enemy movements, he answered about the whole apparatus rather than any single failure of it.
“Those forecasts,” he said, “were worse than weather forecasts. ” Nothing that survives shows the question going back to the men whose job it had been, and no paper sets down what the Navy now believed instead, because from that point on, it was answering the Americans rather than assessing them.