Station HYPO: The Codebreakers Behind Midway

Cdr. Joseph J. Rochefort, officer in charge of Station HYPO in 1941–42. U.S. Navy photograph, public domain, via Wikimedia Commons.

Special Units · The Battle of Midway, 1942

Station HYPO: The Codebreakers Behind Midway

A few dozen cryptanalysts, linguists and traffic analysts in a windowless basement gave Nimitz the objective, the timing and the approach of Operation MI.

Station HYPO was not a ship, a squadron or a battalion, and none of its people fired a shot at Midway. It was a radio intelligence office in the basement of the Fourteenth Naval District’s headquarters at Pearl Harbor, staffed in June 1941 by ten officers and thirteen enlisted men. Between March and the end of May 1942 its cryptanalysts, linguists and traffic analysts, working with the Navy’s other radio intelligence centers in Washington and Melbourne, gave Admiral Chester W. Nimitz an estimate of the Japanese objective, timing and approach that let him place his three carriers northeast of Midway before the Japanese arrived.

This page is about the unit itself: how it was organized, who did what, how it worked, what it could and could not deliver, and what became of it and its commander afterward. The codebreaking campaign of spring 1942, including the arithmetic of JN-25 and the full dispute with Washington, is told in Chapter 1. Here the emphasis is on the people and the organization, and on the uneven record they left. Much of the inside story rests on memoirs written decades later, chiefly W. J. Holmes’s Double-Edged Secrets (1979) and Edwin T. Layton’s And I Was There (1985); where this page relies on recollection rather than wartime documents, it says so.

The dungeon in Building 1

The Navy’s radio intelligence network in 1941 had three main stations, named from the phonetic alphabet: NEGAT in Washington, the headquarters element of the communications intelligence section OP-20-G; CAST on Corregidor in the Philippines; and HYPO in Hawaii. In the summer of 1941 the Hawaii unit was renamed the Combat Intelligence Unit of the Fourteenth Naval District. It worked in the basement of Building 1, the district’s administrative headquarters. Navy Region Hawaii’s history describes a windowless space reached by a single entrance guarded by an armed Marine. The Station HYPO historical society gives the main room as about 50 by 100 feet. The people who worked there called it the dungeon.

The unit’s roots went back to the 1930s. According to NSA’s biography of Thomas Dyer, he was transferred to Hawaii in July 1936 and in 1938 formed the Communications Intelligence Unit at Pearl Harbor, with intercept and direction-finding stations on Oahu. When Rochefort arrived as officer in charge on 2 June 1941, sent by Captain Laurance F. Safford of OP-20-G, Dyer stayed on as his assistant and chief cryptanalyst. Rochefort wanted, in the historical society’s summary, to turn the station from a research unit into an operating one: a unit whose product went to the fleet in time to be used.

The unit stayed in the basement until April 1943, when it moved to new quarters at Makalapa near Pacific Fleet headquarters, having by then been reorganized and expanded under the name Fleet Radio Unit, Pacific (FRUPAC). The dungeon was left vacant or used for storage and offices for decades afterward. Its history has since been rediscovered, and plaques now mark each stairway.

Who did what

HYPO’s strength was the combination of skills under one roof. Rochefort himself was, in NSA’s description, an expert Japanese linguist, an experienced intelligence analyst and a trained cryptanalyst, a rare mix in the prewar Navy. Below him the work divided into four streams. Cryptanalysis, led by Dyer with Lieutenant Commander Wesley A. “Ham” Wright and Lieutenant Commander Jack S. Holtwick, attacked the additive cipher that protected the Japanese fleet’s main operational code, JN-25. Traffic analysis, under Lieutenant Commanders Thomas A. Huckins and John A. Williams, studied call signs, routing and volume rather than content. Translation, led by Marine Major Alva B. Lasswell with Lieutenant Commander Joseph Finnegan, turned partly recovered messages into English. The plot, kept by Lieutenant W. J. “Jasper” Holmes, a retired submariner recalled to duty, with Yeoman William Dunbar, turned all of it into positions and courses.

The linguists were as important as the codebreakers, because JN-25 was never read whole. A recovered message might give a few code groups out of many, and someone had to judge what the fragments meant. A 2024 article in the Marine Corps University’s Marine Corps History describes Lasswell as HYPO’s chief translator and Finnegan as his partner in code recovery, the two splitting the translation work on a two-watch basis. Lasswell called Finnegan “my right arm”; Finnegan, in the same account, had the better ear for spoken Japanese and Lasswell the better eye for the written language.

After the attack of 7 December 1941, members of the band of the battleship California, sunk at her moorings, were assigned to the unit to run its punched-card machines. In the fall of 1941 five more linguists had joined. Shortages of people were a constant theme of the period; the historical society’s account notes that Safford at OP-20-G sent extra personnel when he could, and that links between stations were so unreliable that messages were often carried by jeep.

Machines, index cards and a daily brief

The unit’s method was cumulative. JN-25 was a codebook of five-digit groups protected by an additive cipher, and when the Japanese reused stretches of additives faster than new books could be distributed, messages could be stacked “in depth” and the additives stripped. That work depended on volume, and volume depended on machines. NSA credits Dyer, who had trained under Driscoll at OP-20-G from 1931, with developing the use of IBM tabulators to sort possible solutions, work that earned him the title “the father of machine cryptanalysis.” At HYPO the tabulators indexed enciphered groups, found the repeats that signaled messages in depth and tested additives in bulk, so that every recovered meaning made the next message faster to read.

HYPO’s product reached the fleet through one man. Lieutenant Commander Edwin T. Layton, Nimitz’s fleet intelligence officer, had been Rochefort’s friend since both went to Japan as language officers in 1929, and Rochefort briefed him daily. Formally, Rochefort’s chain of command ran to OP-20-G in Washington; in practice he worked for Nimitz. That arrangement produced the best intelligence of the Pacific war’s first year and also the friction that later cost Rochefort his job.

What HYPO delivered for Midway

The unit’s contribution to Midway came in three parts: the objective, the timing and the approach. The designator “AF” had been correlated with Midway by the Corregidor unit on 13 March 1942, and HYPO and Melbourne agreed by April. What changed in May was the stakes, as traffic built up around an operation larger than anything the Japanese navy had mounted, and senior officers in Washington weighed Samoa, the West Coast, Hawaii and the Aleutians as alternatives. On 20 May, according to the Marine Corps History article, Lasswell began work on material that amounted in practical terms to a large part of the Japanese operation order for Midway. Asked by Nimitz how sure he was of the target, he is recorded as answering “100 percent certain.”

The water ruse was the unit’s means of persuading Washington. Midway was instructed by undersea cable to report a breakdown of its fresh-water distillation plant by radio in plain language; Japanese radio intelligence picked it up, and a Japanese report that “AF” was short of water was intercepted in turn. The Melbourne unit issued its translation on 22 May, and after it the debate about AF disappears from the record. Credit for the idea is shared in the sources. Donald “Mac” Showers, a junior HYPO officer, recalled Holmes proposing it; Finnegan is also credited, and the Lasswell article states that Lasswell did the cryptanalysis and translation of the Japanese reply at HYPO.

Timing was harder, because the Japanese protected dates with a separate cipher. Accounts based on Layton’s memoir credit Finnegan with discovering how it worked, and a sustained effort by Finnegan, Wright and others in the last week of May produced an estimate that the Aleutians would be struck on about 3 June and Midway on about 4 June. The window was closing as they worked: the replacement code system, JN-25c, took effect on 28 May after two postponements. From Layton’s estimate came the forecast that the Japanese carriers would be sighted at 325 degrees, 175 miles from Midway, at about 0600 on 4 June. A Midway PBY found them close to that point.

It is worth being exact about the limits. HYPO never held a complete Japanese operation order, could say little about Admiral Isoroku Yamamoto’s Main Body, and could not tell Nimitz where the carriers would turn or when they would rearm. Once JN-25c came into force the flow of fresh decrypts slowed, and during the battle itself the Americans relied on search planes, traffic analysis and tactical intercepts. Samuel J. Cox, the Navy’s director of naval history, calls codebreaking “only a part, albeit significant,” of the intelligence Nimitz used. The unit’s achievement was a structured estimate good enough to act on, delivered in time.

After Midway: credit, removal and FRUPAC

Nimitz recommended Rochefort for the Navy Distinguished Service Medal, and he recommended Lasswell for the same award. Neither received it at the time. Admiral Ernest J. King denied Rochefort’s medal, on the advice, according to Rochefort’s biographer Elliot Carlson, of staff officers who disliked him, among them the Redman brothers who ran naval communications and OP-20-G. Nimitz gave Rochefort, Dyer, Wright and others letters of commendation instead. Layton received the Distinguished Service Medal.

In October 1942 OP-20-G centralized control of Pacific intercept and codebreaking, and Rochefort was ordered to Washington on what was presented as temporary duty. He did not return. He was assigned to a floating dry dock under construction on the West Coast, and later headed the Pacific Strategic Intelligence Group in Washington. The account of his removal comes largely from an unsigned 1944 memorandum written by a supporter and from later biographies, and it should be read as partisan; defenders of Washington point to the large share of recoveries made by NEGAT and Melbourne and to the reasonable case for centralizing a growing effort.

The unit itself kept working. After Rochefort left, Lasswell became in effect the senior language analyst, and every translation passed through him. The organization grew, took the name Fleet Radio Unit, Pacific, and moved to Makalapa in 1943. In April 1943 it was Lasswell who decrypted and translated the message giving Yamamoto’s inspection itinerary in the Solomons, which led to the interception and death of the Japanese commander on 18 April; Dyer is also credited by NSA with a part in that work, and with breaking the Japanese merchant shipping code and the main Japanese weather code. Dyer stayed in the Pacific until 1945.

Recognition, decades late

Secrecy kept the unit’s story out of the published histories for a generation. The role of the codebreakers became widely known only with declassification and the memoirs of Holmes (1979) and Layton (1985). Rochefort died on 20 July 1976, before most of that material appeared. A campaign led in the 1980s by Showers, then working for Director of Central Intelligence William Casey, succeeded: on 30 May 1986 President Ronald Reagan presented the Distinguished Service Medal to Rochefort’s family in the White House. NSA inducted Rochefort into its Hall of Honor in 2000, Dyer in 2002 and Lasswell in 2019. Lasswell received the Legion of Merit in 1946 for his wartime service; the Midway Distinguished Service Medal for which Nimitz had recommended him never came.

The Naval History and Heritage Command’s profile of Rochefort also states that he received the Presidential Medal of Freedom posthumously. This collection has not found a primary record of that award, and NSA’s biography does not mention it, so it is not listed in the honors above. What can be documented is enough: a unit of a few dozen people, working in a basement with index cards and tabulators, gave a fleet commander the information he needed to accept a calculated risk. Nimitz’s own summary of the battle, as quoted by NHHC, was that it was “essentially a victory of intelligence.”