Top NSA Banner

Today, smooth cooperation among the military services is taken for granted. It was not always so.

As early as January 1932, the Army and Navy recognized the benefits of cooperation in cryptanalysis, particularly in avoiding the costs of duplication of effort. A draft agreement between them was drawn up, pledging the two services to share information and technical assistance. The draft agreement clearly stated, however, that either service was free to engage in whatever cryptanalytic work it felt was in its interest, regardless of what the other was doing.

The Services were rivals for appropriations in a tough budget climate, however, so even this limited cooperation was too much for the times, and it was not signed.

Cooperation was discussed in unofficial meetings between the cryptologic sections over the next several years, but no specific agreements resulted. The Navy, for its part, felt Army representatives had been "indiscreet" in discussing the intercept of diplomatic traffic with the State Department, and, according to a 1933 memo, was therefore chary about sharing its cryptologic secrets with the other service.

Some cooperation was achieved, largely by practice, and late in the decade the two services did sign an agreement confirming a process for cooperation in processing and disseminating decrypts.

In the absence of formal agreements, much of the early cooperation between the Army and the Navy in cryptanalytic work was due to the personal relationship between the SIS's William Friedman and OP-20-G's Laurance Safford. The brass on both sides was aware of their cooperation, and, as a history of OP-20-G dryly notes: "Some officers abetted, others acquiesced, and others forbade it, only to have it re-emerge after they left their posts for other duties."

Friedman and Safford were responsible for the character of the modern Communications Intelligence agencies in their respective services. Their friendship and cooperation also abetted the development of superior Communications Security equipment just before the war. Combining what they had learned about cryptography and cryptanalysis, they and their colleagues developed what the Army called the SIGABA and the Navy called the ECM-MK 1. By either name, it protected U.S. high-level communications, and was never solved by any of America's adversaries.

By mid-1940, the two services created a committee of their cryptologic experts to devise a plan for dividing work in decryption. It was clear each service would be responsible for the military traffic of its opposite numbers in Japan, Germany, and Italy, but diplomatic traffic remained a question. Originally, the cryptologic sections had agreed that the Navy would process Japanese diplomatic traffic, but the Army General Staff ordered the Signal Intelligence Service to continue processing German and Japanese diplomatic messages, despite the agreement.

In January 1941, the two services agreed to split the responsibility for distributing decrypts of Japanese diplomatic messages to the president and other executive consumers by month. The services would have the action in alternate months, with the Army starting that month.

By October 1941, because of a significant increase in the number of Japanese messages and the resulting difficulty in processing them in a timely way, the agreement was revised to make the Army responsible for decrypting and translating Japanese diplomatic messages on even numbered days and the Navy responsible for the odd.

Under the odd-even agreement, decrypts were disseminated by courier. OP-20-G and SIS would show the latest messages to a select group daily at 11 a.m., including the Director of Naval Communications, the Director of Naval Intelligence and their Army equivalents. A Navy officer would carry messages of interest to the Naval Aide to the President; an Army officer would take them to the Secretary of State. The recipients were not allowed to retain the copies.

There was an additional problem: the Army and the Navy did not have the only cryptologic organizations. The Office of Censorship and the Federal Communications Commission operated intercept stations and engaged in cryptanalysis. This heightened military fears of security compromises; one possible breach had been found in the State Department and the possibility of losses increased when there were uncontrolled or uncoordinated sources of decrypts.

Therefore, General George Marshall, Chief of Staff of the Army, and Admiral King, Commander in Chief of the U.S. Fleet, sent a memorandum to the president, asking that he order the other government agencies to discontinue their cryptologic operations. On July 8, 1942, Franklin Roosevelt issued a memorandum stating that he agreed with his service chiefs, and directing the FCC and the Office of Censorship to discontinue their cryptanalytic units.

There was some cooperation in the field. For example, in the Philippines, the Army and Navy had a local agreement to exchange both raw and analyzed COMINT. Personnel from the two cryptologic services would meet in Manila and let the other's representative read folders of intercepts. Nothing was copied or written out from these meetings. One of the participants remembered that the exchanges were carried out at Fort Santiago, Manila, in a former "Spanish dungeon."

The Army had a central cryptologic staff of only seven people for the years 1930-1936, and had expanded only to 19 by 1939. At the time of the attack on Pearl Harbor, the Army's cryptologic organization consisted of about 180 persons in Washington, and about 150 in the field. (By way of comparison, at the end of the war, the organization had just in excess of 10,000).

OP-20-G had a staff of 181 in Washington by December 1941, with about the same outside the U.S. The Navy organization built up to numbers equivalent to the Army by the end of the war.

The Army and the Navy achieved cooperation in some limited areas of endeavor in the period before the war, and liaison and cooperation were expanded during the war itself, but it was all done on a voluntary basis. Essentially, the U.S. Army and Navy each conducted their prewar and wartime cryptologic activities with little reference to the other.

SOURCES:

Howard W. Brown, The Reminiscences of Lieutenant Colonel Howard W. Brown, Signal Security Agency, 4 August 1945, SHR-045, CCH Files.

Captain J. S. Holtwick, Jr., USN (Ret), Naval Security Group History to World War II, June 1971, SHR-355, CCH Files.

Anon., The Achievements of the Signal Security Agency in World War II, SRH-349, CCH Files.

Anon., OP-20-G File on Army-Navy Collaboration, 1931-1945, SRH-200, CCH Files.