4.1. Captain Laurance F. Safford – In the Eye of the Controversy

At the center of the Winds message controversy – in fact its primary and almost exclusive source – was Captain Laurance Frye Safford. Safford had played a critical role in the founding and operation of modern American naval cryptology. He had been put in command of the fledgling cryptanalytic section of the U.S. Navy in 1924 – the (Cryptanalytic) Research Desk within the Code and Signal Section of the Navy's Division of Naval Communications (OP-20-G). He had overseen the recruitment, training, and formation of the corps of radio intercept operators who manned the Navy's monitoring sites around the world and in the United States. Safford played a role in the establishment of the navy's constellation of monitoring and direction finding (DF) sites in the Pacific region from the mid- to late 1930s. He had also recruited and staffed the research Desk of the Code and Signal Section with such notables of naval cryptanalysis as Agnes Meyer Driscoll, Joseph Rochefort, and Thomas Dyer. Safford had set up a program of training in cryptanalysis of selected naval and marine officers, rotating them into the Research Desk for periods of on-the-job training before they returned to positions in the fleet. Safford also had allowed, albeit reluctantly, the early experimental use of machine aids in cryptanalysis – among them early IBM punch card sorters to tabulate and inventory code groups and specialized typewriters modified to copy Japanese Kana characters sent via Morse code.^[]

Safford had nurtured OP-20-G through the hard and lean interwar years and, at the rank of Commander, was in charge of the entire section in late 1941. He was highly respected by other cryptologic and intelligence officers from both the navy and the army. From a technical standpoint, Safford was a talented officer, though his true ability lay in the collection, forwarding, and processing – the "front end" of cryptology – and not in the analysis of the intercept or dissemination of communications intelligence. Sometimes he simply misunderstood the analytic process, especially the technical background to major cryptanalytic breakthroughs and the fact that major systems were changed, or superseded, and required substantial efforts to recover them. This was illustrated in a short history of prewar communications intelligence he authored in late 1943, "The Undeclared War," in which he made two glaringly incorrect assertions. First, he claimed that the "Navy had solved the primary Japanese Fleet System (JN-25) to a partially readable extent."^[] This statement greatly overstated the actual progress that was limited to less than ten percent of the AN-1 codebook (later notated as JN-25B). In fact, he may have referred to the predecessor variant, AN (later notated as JN-25A), but it is unclear from his writing which system he meant – a vagueness that has confused some researchers in the decades since. Secondly, he attributed the S.I.S. solution of the Purple cipher machine to the fact that the "Army had acquired a model of the Japanese Diplomatic machine and the original set of cipher keys used with it." To this comment, William F. Friedman, in 1952, greatly objected and wrote an emphatic note in the margin of this section: "This is not true. Army acquired it the hard way – cryptanalytically!^[]

While Safford had fostered the development of OP-20-G and in 1936 had become its first permanently assigned commander, by the time of Pearl Harbor he probably had come to be overmatched by the enormous demands in time and resources made upon his organization.^[] The rapidly multiplying targets and the simultaneously growing workforce of OP-20-G – the worldwide mission included some 500 people – overwhelmed the prewar structure he had built. As mentioned earlier, the OP-20-G mission was stretched globally, with two centers of interest, the ongoing U-boat struggle in the Atlantic and the Pacific crisis that vied for the scarce resources of the section. The multiple demands may have simply outstripped Safford's ability to effectively manage OP-20-G.

Symptomatic of the problem was his approach to solve the German U-boat Enigma device in the eighteen months prior to Pearl Harbor. In 1940 Safford set up a small team dedicated to solving enciphered German U-boat traffic. At the time, OP-20-G did not know that this cipher traffic was generated by a more advanced naval version of the commercial Enigma cipher machine, a copy of which the navy possessed. The cryptanalytic effort was small, perhaps fewer than ten people, but it represented a diversion of scarce resources. In the months prior to Pearl Harbor, he resisted efforts to coordinate work with the British in OP-20-G's attack on the traffic. When he finally allowed cooperation, he often ignored their experienced technical advice in favor of that from his own analysts like Agnes Driscoll that proved ultimately to be an analytic dead end. The navy's attack on Enigma proved to be unproductive for the first two years.^[]

Safford's actions at OP-20-G in the months before Pearl Harbor were erratic: at times he controlled activities completely; at other times, as we shall see later in this section, he seemed to let parts of the mission slip. In testimony before various hearings that in the weeks leading up to the attack on Pearl Harbor, he stated he had worked longer than normal hours. He said that he was deeply involved in all aspects of the section's operations. And he was. In all probability, though, he tried to do everything, and as a result, many problems developed, especially in the administration of daily activities. In fact, many officers in naval intelligence and cryptology were working almost twelve- to fourteen-hour days. Yet, on 7 December, Safford stayed out of his office the entire day, only to return the following Monday morning.^[]

Curiously, in early 1942 Safford had suggested to the CNO staff that, as part of a recommended reorganization of OP-20, he be replaced as head of OP-20-G. By the spring of 1942, Safford was gone as the commander of the code-breaking element and was placed in charge of the office supervising the not unimportant job of developing and fielding of cryptographic systems for the U.S. Navy – OP-20-Q.^[]