Islands of Destiny: The Solomons Campaign and the Eclipse of the Rising Sun

by John Prados

  The radio units produced three types of intelligence. The first consisted of decoded and translated decrypts of actual messages. The second kind of data embodied the results of pattern analysis—radio traffic analysis—wherein emitter locations and traffic volume profiles permitted conclusions on the identity, movements, and sometimes the intentions of forces. During the initial months of the Pacific war, when codebreaking had yet to hit its stride, and later when codes changed, radio traffic analysis would be the most valuable element of communications intelligence. The third type of information consisted of analysis applied to all this material.

  The British codebreaking operation formed part of the Government Code and Cipher School (GC&CS), and the Australian one a part of its Directorate of Naval Intelligence. Each had field units. The British, with commonwealth participation, had the Far East Combined Bureau, originally located in Hong Kong, then Singapore until the Japanese threatened that place, then Ceylon and East Africa. The Australians formed a small Signal Intelligence Bureau at Canberra. The Americans had Station Cast in the Philippines at Corregidor, subsequently evacuated to Australia as Station Belconnen, though during the prelude to the Port Moresby battle, that unit functioned in both places. Pearl Harbor had Station Hypo, while in Washington Op-20-G ran its own communications intelligence element, called Station Negat. The U.S. field units would later become Fleet Radio Unit Melbourne (FRUMEL) and Fleet Radio Unit Pacific (FRUPAC), FRUMEL being a joint Australian-American activity. All the stations and home offices had their own private, top-secret communications channel, or “circuit,” to exchange information using a machine-encryption device known as Copek. In the Allies’ drive to crack the Japanese codes, the first decrypts from JN-25, as the Allies called the Imperial Navy’s fleet code, came from messages sent in early March 1942. “Ultra” became the generic term for data derived from communications interception.

  Radio intelligence fed its information to high commands and fleet commands. In the U.S. system, Op-20-G reported to COMINCH, the commander in chief of the fleet and chief of naval operations, Admiral Ernest J. King, and through him to the Joint Chiefs of Staff. At Pearl Harbor, Admiral Chester Nimitz, leading the Pacific Command and the Pacific Fleet, received his data through his fleet intelligence officer, Captain Edwin T. Layton. Though this gets a little ahead of our story, Nimitz ordered the creation of “advanced intelligence centers,” starting with a Combat Intelligence Unit plus FRUPAC at Pearl Harbor, to meld data from all sources, including radio intelligence. The British Far East Combined Bureau, upon evacuation from Singapore, mainly served the chief of the British Eastern Fleet. In Australia the information came through their own units, which benefited from the findings of all the Allies. General Douglas A. MacArthur, after March 1942 the leader of the theater that included Australia and New Guinea, was served by his intelligence officer (G-2), Major General Charles A. Willoughby. All this apparatus came into play when the Japanese aimed at Port Moresby.

  Different codes had different degrees of complexity. Breaking into a code was affected not only by the number of messages sent and intercepted in that system, but also the amount of repetition that occurred. Standard reporting items generated repetition. Thus weather codes, constantly reporting the same basic data, were among the easiest to break. The first Allied indication of enemy interest in Port Moresby came after the fall of Rabaul, when monitors intercepted Japanese high-altitude weather reports for the Moresby area. On March 5, using the Copek circuit, FRUMEL suggested that a Japanese operation against Port Moresby impended. Another repetitive element in codes is employment of standard letter groups or words to designate places. These disguise locations and reduce transmission length, and the Imperial Navy made broad use of them. The tale of the Japanese use of “AF” to refer to Midway—and how verifying that helped Admiral Nimitz win that battle—is well-known. The Imperial Navy simplified Allied difficulties in identifying such indicators by their reliance on a method that utilized the same root to refer to related places. “R” in Japanese messages referred to Rabaul, “RZM” to Lae. Station Hypo recovered a March 25 message from Admiral Yamada instructing units on attacks against “RZP.” Two days earlier Station Cast (FRUMEL) had published a list of Japanese location designators that included RZP as Port Moresby. Commander Joseph J. Rochefort, the eccentric genius who ran Hypo, realized that Moresby was the only target Japanese naval air was attacking in the South Pacific, confirming RZP as that place.

  More RZP messages appeared over succeeding days, along with mention in JN-25 of an “RZP campaign,” as well as ones linking the Japanese aircraft carrier Kaga with that operation. On April 9 a JN-25 decrypt revealed the plan’s architecture—it would include a striking force of carriers, an attack force, occupation force, and a support force. The existence of occupation forces for other places, one of them Tulagi, also stood revealed. At Pearl Harbor two days later, Captain Layton issued a fleet intelligence summary warning of a South Pacific offensive. On April 13 the Americans circulated a nearly complete decrypt of a message from Rear Admiral Hara Chuichi, commanding Carrier Division 5 with the Shokaku and Zuikaku, advising Vice Admiral Inouye of his expected arrival at Truk in two weeks. Already shown by an earlier intercept was that the Kaga had been damaged running aground. Hara’s ships replaced her. The British contributed a report on April 14 noting the withdrawal of the Nagumo force from the Indian Ocean, speculating that part or all of it might attack Port Moresby. Later data indicated that light carrier Shoho would participate, as well as the four heavy ships of Cruiser Division 6. Traffic analysis followed the movements of many Japanese fleet units as they gathered. Australian intelligence noted increased air patrols over Moresby, the Coral Sea, the waters bounded by New Guinea, the island chains, and the Australian mainland. Intelligence then discovered the actual code name “MO Operation.” On May 1, Commander Rochefort was able to report that MO had begun.

  Admiral Nimitz responded to the threat even before he had a complete picture. At Pearl Harbor on April 17, while the famous Doolittle raid was being launched against Japan, Nimitz brought together his experts to examine the existing reports. Fleet intelligence officer Layton predicted the Japanese would divide into an assortment of forces and conduct parallel activities. With a few more days’ intelligence, Nimitz confirmed his decision, more confident that two U.S. aircraft carriers plus Australia-based airplanes could face down the Imperial Navy. The Pacific fleet commander had kept the Lexington task force in the South Pacific; now he ordered it to rendezvous with another built around carrier Yorktown. The Royal Australian Navy also participated, with an Allied cruiser-destroyer group led by British admiral John G. Crace. The united force would inhabit waters the Japanese had to pass. Nimitz made a quick trip to the continental U.S. to brief Admiral Ernest J. King, who met him in San Francisco. Upon returning to Pearl, Nimitz formalized his instructions in an April 28 directive. He foresaw a battle in the Coral Sea focusing on countering the Japanese.

  The Imperial Navy’s first move came on April 28, when an aviation tender advanced to Shortland islet off Bougainville to act as a forward base for air patrols. Then Rear Admiral Shima Kiyohide steamed toward Tulagi, where he appeared on May 3. There the Japanese intended to install a major seaplane facility. At the moment Shima began landing SNLF infantry and aviation specialists, Admiral Inouye’s invasion armada still rode at anchor in Rabaul. Carriers Shokaku and Zuikaku, with Rear Admiral Takagi Takeo’s striking force, were en route. Covering and support forces sailed west of the Solomons, about a day from Tulagi. Also west of the Solomons were Rear Admiral Frank J. Fletcher’s U.S. carriers, which joined on the first of May. It was Fletcher who opened the battle, with an air strike against the Japanese at Tulagi. The Americans did little damage, but they stunned the enemy.

  In Imperial Navy practice, in which the battleship held sway even though the aircraft carrier proved itself more and more the decisive weapon, heavy ship leaders exercised command over carriers when they acted together, except where special pr
ovisions were made—as in the creation of the First Air Fleet, the Kido Butai. Thus the Japanese carrier group sailed under Admiral Takagi, the cruiser leader, not Hara Chuichi, the carrier boss. Until Coral Sea there had never been a naval battle in which aircraft carriers duked it out with aircraft, and surface ships never fired their guns except at planes. Takagi had no basis for judgment. He himself was a veteran of the Dutch East Indies fighting—and victor of the Battle of the Java Sea—but Japanese forces there had not been seriously threatened by aircraft. To complicate matters, no current Imperial Navy carrier formation leader was a pilot—not Hara, not Nagumo, not anyone else. This reflected the gradual acceptance of carriers in the Japanese Navy.

  Slightly ameliorating this situation was the fact that Takagi and Hara were academy classmates and friends, disposed to cooperate, and that their senior staff officers had somewhat more knowledge of airpower, one a former manager of aircraft development, the other a torpedo specialist whose staff work in armaments production had included monitoring aircraft manufacture. At Truk before this sortie, the staff men Yamaoka Mineo and Nagasawa Ko had speculated on the possibility of an aeronaval battle. Though Japanese intelligence knew of no U.S. carriers in the Coral Sea, they expected to find them. Once Fletcher hit Tulagi, Yamaoka and Nagasawa knew they were right.

  Preparations cost the Japanese a full day. Because of losses among Yamada’s land-based groups, Rabaul needed more Zero fighters to cover the Port Moresby invasion force. A canceled ferry mission led to the need for Admiral Hara’s carrier pilots to fly replacements into Rabaul, returning aboard aircraft sent to retrieve them. Meanwhile the American task force took position, and the intelligence network produced critical decrypts. On May 3, Station Hypo relayed details of Fourth Fleet’s orders to Takagi, plus a Rabaul status report. Allied codebreakers continued to stream position, heading, and intentions data for the Moresby invasion flotilla.

  Nimitz sent Fletcher two dispatches on May 5. The first outlined the Japanese strike force plan; the second specified the probable date for Japan’s invasion and alerted the admiral that it would open with a carrier strike on Moresby likely to be launched from the Coral Sea. Fletcher had admirable intelligence if he chose to use it. A set of air searches predicated on this data might have revealed Hara’s carriers. Fletcher does not seem to have laid on any such scouting pattern. He relied upon standard searches.

  Admiral Fletcher also had the blessing of a mobile radio unit, a communications intercept detachment embarked on the Yorktown. Led by Lieutenant Forrest R. “Tex” Baird, this small unit, complemented by another under Lieutenant Ransom Fullinwider in the Lexington, afforded real-time ability to listen in on Japanese transmissions. Tex Baird is truly scathing on the subject of Frank Fletcher, whom he saw as having squandered this advantage through indecisive leadership, mistrust, and poor coordination. Baird and Fullinwider were among a small cadre of “language officers” the U.S. Navy had trained in the tongue and customs of the adversary as well as codebreaking techniques. Where the math and cipher experts were masters on Japanese codes, the language officers formed the core of U.S. ability to interpret the results, as well as the interface between codebreakers and line officers.

  Station Hypo created these detachments to accompany task forces on their missions. One day Commander Rochefort summoned Tex Baird and another language officer, Lieutenant Gilven R. Slonim, to his office and flipped a coin, asking Baird to choose heads or tails. He lost the toss. This led to his assignment to Fletcher (Slonim went with “Bull” Halsey aboard Enterprise). Baird served Fletcher as conduit for secrets and supplied the results of his own monitoring. Before the battle Tex Baird aroused Fletcher’s ire by refusing to describe the top-secret U.S. codebreaking to the admiral’s entire assembled staff. They had another run-in when Rear Admiral Alva Fitch told Fletcher his Lexington had intercepted a sighting report from an enemy submarine—Baird found it not credible. Fletcher and Fullinwider had previously served together, which inclined the admiral to believe the Lexington’s claim. At Coral Sea, Tex Baird’s unit recorded a message revealing Hara’s position. Fletcher preferred Fullinwider’s erroneous interpretation, preventing a U.S. attack. On the first day of battle, when Fletcher had already ceded tactical control to Admiral Fitch—here Baird disputes Fletcher’s after-action report, which he views as falsified—the task force commander refused to act on the intelligence.

  At Coral Sea both sides experienced what U.S. officers would later term “makee learnee,” educating themselves by trial and error. Fletcher sent his carriers in succession to refuel, removing them from the battle area. Hara’s ships abandoned their communications security with first contact, enabling the U.S. mobile radio units to intercept uncoded messages. The Americans struck at Japanese light carrier Shoho with all their aircraft very early one morning. Though Fletcher and Fitch knew a large Japanese task force was somewhere nearby, they held nothing back in case it was sighted. Aircraft continued hitting Shoho once she was visibly sinking, forgoing the chance to cripple other warships. The Japanese misidentified an American oiler and destroyer as a carrier and cruiser—corrected as a transport and cruiser, and wrong both times—then waited too long to recall attack groups, leading to a wasted strike. They then misidentified Admiral Crace’s force, detached to block the final passage to Port Moresby, as a unit of battleships and cruisers. Japanese planes returning from their attack—with no bombs or torpedoes—saw the U.S. carriers late in the day. A desperate attempt at an evening attack led to Japanese planes in the air after nightfall. Some of them even tried to land on the U.S. ships.

  On May 8 came the climactic day. Hara’s carriers resumed radio discipline and changed their call signs, confusing U.S. monitors at a key moment. Each side sighted the enemy’s carriers within moments of the other. Both sent out the biggest strikes they could muster. The Japanese scored many hits on the Lexington, and one crucial bomb fell on Yorktown, until then lucky and skillful at evading many torpedoes. The bomb missed Baird’s radio room by just thirty feet. When the Americans hit the Japanese, carrier Zuikaku happened to be hidden beneath a squall and the attack went against Shokaku, disabling her flight deck. The Imperial Navy retreated. Admiral Inouye had already recalled the Moresby invasion force. On the American side, it seemed at first that the Lexington might survive, but a terrific internal explosion rocked the ship, fires spread, and she had to be abandoned.

  The Battle of the Coral Sea is usually accounted a draw. Its most important impact lay in educating the warfighters in aeronaval tactics, though to judge from Nagumo’s performance at Midway, the Americans learned better. But it also confirmed the value of codebreaking, especially in conjunction with aerial reconnaissance, as pillars of intelligence. Though Tex Baird left the Yorktown—to be replaced by Ransom Fullinwider—mobile radio units became firmly established as elements of American admirals’ staffs. Midway, of course, prized communications intelligence in unmistakable fashion. The codebreakers and the aerial photographers, along with the coastwatchers, would watch intently as the Japanese continued their war.

  Around the time of Coral Sea, Japanese naval commentator Ito Masanori wrote a series of articles for a Nagoya newspaper. He warned against overextending Japan’s perimeter. Ito had covered naval affairs for years, and even with Japan’s tightly restricted public debate he got the sense the high command was advancing mindlessly. Now Ito advocated a halt, a terminal point for the offensive. Later a friend, an Army officer, warned him against publicizing such views—they might lead military hotheads to act against him. Nothing would dissuade the Imperial Navy.

  The next stage began innocently enough. Japanese naval troops from Tulagi crossed to Guadalcanal, where they shot a half dozen cows, carrying them off for food. Coastwatchers on the big island reported that. A scout plane spotted a JNAF aircraft photographing Guadalcanal. Almost a month later a Japanese survey party came from Tulagi and began examining a site at Lunga Point. They stayed. Several barges of men arrived on Guadalcanal after June 19. On Ju
ne 25, Army G-2 reported the SNLF soldiers burning off grass, putting up tents, and starting work on a boat dock. Two water tanks were erected. Japanese patrols sought the coastwatchers, asking around the villages about whites. In early July the two vessels of Admiral Matsuyama’s Cruiser Division 18 dropped anchor off Lunga and landed construction troops and naval infantry. The Japanese soldiers dug trenches and set up machine guns. Several days later Matsuyama returned to deliver the balance of the 11th and 13th Construction Units and more soldiers of the 81st and 84th Naval Guard Units. Captain Monzen of the engineer force became island commander. Most of these developments appeared in one or another Allied intelligence report. On July 5 the codebreakers identified the construction units and inferred their purpose—building an airfield.

  By this time the Allies also knew of Captain Kanazawa’s 8th Base Force at Rabaul, the 14th Base Force at Kavieng, the 1st Base Force at Buin, and the 7th Base Force at Lae, all from radio traffic. Photo reconnaissance permitted strength estimates: 21,000 around Rabaul, 1,500 at Kavieng, some 13,000 on Bougainville, and 2,000 at Lae. At Emirau, a small island off Kavieng, a Japanese patrol captured coastwatcher Cornelius L. Page. He was held for a couple of weeks, then executed as a spy. General MacArthur sent Admiral Nimitz a dispatch on June 27 that evaluated Japanese air strength at Rabaul as having returned to pre–Coral Sea levels, except in fighter aircraft, of which only about two dozen were estimated. MacArthur noted shipping activity exceeding normal requirements, plus signs of new shore installations and airfields.