THE CODEBREAKERS

by DAVID KAHN

  This return to reality was also marked by an increasing maturity in security matters. The agency is less hysterical about petty details, suggesting that it is more confident in essential points. For several years, N.S.A. recruiting pamphlets did not so much as mention the words “code” and “cipher,” or any related words—presumably for the same reason that the Ming code was stamped “secret.” But in 1964, a pamphlet for prospective employees declared: “N.S.A., as the Agency of the government responsible for the security of all U.S. communications systems, has need to recruit and develop specialists in cryptography.” At about the same time, the agency declassified William F. Friedman’s 1937 War Department publication on the Zimmermann telegram. These incidents imply that the agency has closed the gap that the case of the road signs epitomized—the gap between what the agency’s preening self-esteem told it and what things were really like.

  All this points up the value of Congressional surveillance of intelligence agencies. Yet Congress has shown an odd reluctance to put into practice the lessons of this case history. Subcommittees of the House and the Senate Armed Services and Appropriations Committees undoubtedly oversee N.S.A.’s operations, as they do those of the C.I.A., but these shadowy groups—they are not even listed in the Congressional Directory—appear to be less than vigorous in discharging these duties.

  Though the Dunlap case came after reforms instituted as a result of the House Un-American Activities Committee investigation, both houses of Congress declined to look into it. “If a similar series of tragic blunders occurred in any ordinary agency of Government,” wrote a journalist who had studied the Dunlap case in detail, “an aroused public would insist that those responsible be officially censured, demoted, or fired.” Said Stewart Alsop: “The N.S.A. particularly could do with a bit of supervision. It has a horrible security record…. If the C.I. A. had been responsible for either case [Martin-Mitchell and Dunlap], there would have been a hullabaloo to make the Alger Hiss case seem tame.” (To put the N.S.A. episodes into perspective: C.I. A. has not yet suffered a single known defection or penetration.) Even the Un-American Activities Committee used kid gloves. “The sensitive nature of the operation of the National Security Agency was recognized and respected by the Committee on Un-American Activities during its investigation and hearings. The committee did not attempt to learn the details of the organizational structure or the products of the Agency, feeling it had no need for knowledge in these areas.” Yet greater knowledge may well have produced greater benefits.

  Indeed, Congress behaves toward N.S.A. as if it is trying to propitiate the sorcerers who control the dark powers of cryptology. In 1956, the director, Lieutenant General Ralph J. Canine, testified before a House committee in favor of a bill to increase the number of high-paying ($10,000 to $15,000) scientific jobs in the government, including N.S.A. Chairman Tom Murray later told the House that “The committee was so impressed with the need for adequately compensating people who have devoted a lifetime to this very important area, that at the request of General Canine we increased the amount from the original submission of 35 to 50 of these positions.” In 1959, Congress passed Public Law 36 to exempt the N.S.A. from the legal requirement binding all government bodies to file a full description of each job in the agency with the Civil Service Commission.

  In 1964, Congress gave the director of N.S.A. the power to fire at will any N.S.A. employee “whenever he considers that action to be in the interest of the United States.” An identical bill had died in the Senate Judiciary Committee in the previous Congress. Both had been introduced by the House Un-American Activities Committee as a result of its Martin-Mitchell investigation. They wrote into law some of the stricter employment practices that the committee and the agency had agreed upon. In both Congresses, the summary discharge power was bitterly attacked as a violation of the Bill of Rights principle that no one shall be “deprived of life, liberty, or property, without due process of law.” Liberals in the House quoted the Washington Post: “This [bill] is the very definition of arbitrariness. It means that an employee could be discharged and disgraced on the basis of anonymous allegations without the slightest opportunity to defend himself—without any hearing at all and without any administrative review or even any judicial review of the decision. This would put everyone working for N.S.A. at the mercy of any mischief-maker or malcontent or personal enemy who might call him a subversive or a homosexual or an alcoholic.” Though the bill’s sponsors did not rebut these arguments, the bill won an overwhelming 340 to 40 majority in the House and an easy voice passage in the Senate.

  N.S.A. conjures up its biggest prizes from the legislative pocketbook. For Congress, which can treat some supplicants for funds very harshly, smiles beneficently upon N.S.A. In 1962, the House Armed Services Committee named a special three-man subcommittee to look into N.S.A.’s request for $10,000,000 for its nine-story annex and for money to hire more people. Since this plea came only a few years after Congress had given the agency some $35,000,000 for an enormous brand-new home, one might have expected it to encounter a rather cool reception. Instead: “After an exhaustive briefing and tour, I personally was convinced and have withdrawn all objection,” said the financially conservative Republican member of the subcommittee, Durward G. Hall of Missouri. And of course N.S.A. got the money.

  What is the potent spell that N.S.A. casts over Congress? Why this amazing haste to grant what appears to be the slightest whim of the agency? Much, no doubt, comes from the simple fact that, by and large, the agency does a good job. But part also comes from some razzle-dazzle by the agency’s using the ultra secrets to which its work—unlike other agencies of the government—gives it access. Sometimes it lets key Congressmen take quick peeks at them, join the privileged fraternity of Those Who Know, and so champion the cause of their fraternity brother. ‘The members, and myself in particular,” said Hall of his tour, “of the Armed Services Committee have probably seen more classified equipment and been exposed to more classified construction—from communications through telemetry—than most Members [of Congress].” More often the agency enshrouds its secrets in fearful gloom, awing Congressmen with sacred mysteries that are no more to be uttered than is the tetragrammaton. “The Agency is faced with enormous security responsibilities. The missions assigned to the Agency seek to fulfill basic requirements of our national security. All activities conducted by N.S.A. to carry out these missions are highly classified. Disclosure of the nature of these activities or portions of them could seriously impair the success of the Agency’s efforts.” So intoned the counsel of the Defense Department, and so an almost trembling House committee printed it in support of the summary-discharge bill.

  This stratagem plays upon Congress’ fear and ignorance. Unfamiliar with the complexities of modern cryptology, the legislators worry that a single slip could betray what they usually refer to as “the” American code system. They do not realize that there is not one but dozens or hundreds of such systems, and that a full compromise of even one would involve a detailed description of a complex mechanism, lists of hundreds of rotor wirings, and long schedules of keying arrangements. They regard cryptology not rationally as what it is but superstitiously as a potent magic—and the non-rational view of things has hardly advanced civilization.

  N.S.A. exploits this attitude to withhold as much information as it can from Congress. Yet one may wonder whether the elected representatives of the American people may not be trusted with information handled daily by typists and technicians.

  Although these N.S.A. tactics are improper and shortsighted, responsibility for Congressional supervision rests ultimately upon Congress. It should exercise its most jealously guarded prerogative—investigation—as vigorously in the intelligence field as it does elsewhere. The President’s Foreign Intelligence Advisory Board cannot substitute for Congress, which is in a different branch of government from the intelligence agencies and which holds the pursestrings. Congressional surveillance would benefit both N.S.A. and th
e nation as a whole. It would, in the first place, help keep N.S.A. from reverting to its old, dangerous smugness; the Un-American Activities Committee investigation was the object lesson for this. It would, in the second place, make an essentially antidemocratic operation responsible to the processes of free men. The mail-opening activities of N.S.A. are repugnant to Americans, who tolerate them reluctantly only because of the Cold War. Its spyings can never be wholly reconciled with the ideals of a nation founded on a respect for the dignity of the individual. But they can be made accountable to those ideals, as embodied in the elected representatives of the people.

  Furthermore, to the extent that N.S.A. produces knowledge, it produces power, and, Thomas Jefferson said, “whatever power in any government is independent, is absolute also.” This problem is not as acute with N.S.A. as with C.I.A., because N.S.A. neither formulates nor executes policy, nor does it conduct actual operations, such as Cuban invasions. Nevertheless, N.S.A. should be energetically supervised by Congress to prevent abuse of power.

  All this is a nuisance. So is democracy. It is much easier to hire a dictator than to bother with elections and all the other details. It is much easier not to bother with checking up on N.S.A. But it must be done. Otherwise the nation jeopardizes some of the very freedom that N.S.A. exists to preserve.

  Since its organization in 1949, A.F.S.A.-N.S.A. has always been headed by a general or an admiral. The three services rotate the command. Terms have ranged from 18 months to more than four years. Six men have directed this silent agency: Rear Admiral Earl Everett Stone, U.S.N., July, 1949, to August, 1951; Lieutenant General Ralph Julian Canine, U.S.A., to November, 1956; Lieutenant General John Alexander Samford, U.S.A.F., to November, 1960; Rear Admiral Laurence Hugh Frost, U.S.N., to May, 1962; Lieutenant General Gordon Aylesworth Blake, U.S.A.F., to June, 1965; and Lieutenant General Marshall Sylvester Carter, U.S.A. The only thing they have in common seems to be their conspicuous absence from the public eye.

  Stone, 53 when he assumed the directorship, had been in communications for virtually all his Navy career except sea tours. He holds a master of science in communications engineering from Harvard. His shore duty was entirely in naval communications, and as assistant director of naval communications from 1942 to 1944 he commanded the Navy’s communications-intelligence unit. He served as director of naval communications from 1946 until his appointment to head the newly formed A.F.S.A. After his two-year tour, he commanded Cruiser Division 1 during the Korean War, bombarding shore installations, and then held two high naval training posts before retirement in 1958.

  Canine was the only N.S.A. director not to have graduated from a service academy, and his cryptologic experience was limited to that obtained during a year’s duty as part-time communications officer from 1919 to 1920; he also had very broad military experience. He was called to active duty as a second lieutenant of field artillery in 1917 after having graduated the previous year from Northwestern University. He served with the A.E.F., and in June, 1919, became communications officer and adjutant for the 7th Artillery Brigade at Camp Funston, Kansas, presumably handling some codes and ciphers. During the 1920s and 1930s, he taught military science at Purdue University, studied at the Field Artillery School, and served as a regimental supply and liaison officer, a trial judge advocate for courts-martial, a post exchange officer, a plans and training officer, student at the Command and General Staff School, professor of military science at Ohio State University, and as commander of the 99th Field Artillery Battalion. In August, 1942, he became an assistant chief of staff and then chief of staff of the XII Corps, the post in which he served through Normandy, the Battle of the Bulge, and contact with the Russians. After several command posts, he was appointed deputy assistant chief of staff for intelligence at Army headquarters in September, 1950, and ten months later, at age 55, took over A.F.S.A., which became N.S.A. during his four-year term. He was perhaps the best liked of N.S.A. directors.

  Samford learned to fly at Kelly Field after his 1928 graduation from West Point, and spent the prewar years in routine duties at hot and dusty flying fields in Texas, Illinois, the Canal Zone, Virginia, and Florida, plus a four-year stint as a flying instructor at Kelly Field. He spent most of World War II in England as deputy chief of staff and then chief of staff of the 8th Air Force, whose Flying Fortresses pounded Germany. For two years, starting in 1944, he served as deputy assistant chief of staff for intelligence at Air Force headquarters in the Pentagon, and, after tours as commander of the Antilles Air Division in Puerto Rico and of the Air Command and Staff School, he became director of intelligence for the Air Force in October, 1951. In July, 1956, he became deputy director of N.S.A., serving for four months until he was named director at age 51. The end of his four-year term was marred by the Martin-Mitchell scandal.

  Frost, who took a two-year postgraduate course in line and applied communications at Annapolis from 1933 to 1935, spent much of his naval service in communications. He handled presidential messages as communications officer of the U.S.S. Indianapolis when that cruiser took Franklin D. Roosevelt on a goodwill tour to Argentina in 1936. He was decorated for his command of the destroyer Waller in the Pacific in 1943, and then served in communications in the Solomons and in Washington to the end of the war. From 1945 to 1950, except for a year, he worked in intelligence, and after a year’s study at the National War College and two years on sea duty, including command in Korean waters, he was assigned as N.S.A. chief of staff from 1953 to 1955. Following a year of sea duty he was named director of naval intelligence and then N.S.A. director, with temporary rank of vice admiral. He was 58. After a short 18-month term, during which he bore the brunt of the House Un-American Activities Committee investigation, he was named commandant of the Potomac River Naval Command.

  Blake likewise spent nearly all his service career in communications, beginning with study at the Signal School at Fort Monmouth from 1933 to 1934. He held other communications posts in the 1930s, and during World War II commanded the Army Airways Communications System in the Pacific. He attended the Air War College and then headed research and development work at Wright-Patterson Air Force Base, Ohio, for four years. He was appointed Air Force Director of Communications in 1953 and, three years later, assistant deputy chief of staff for operations. From 1957 to 1959 he commanded the U.S.A.F. Security Service. After serving as chief of staff, Pacific Air Forces, and commander, Continental Air Command, he took over N.S.A. at age 51.

  Carter, who had been deputy director of the C.I.A. for three years, had to be moved from that post when President Johnson assigned Admiral William Raborn as Director of Central Intelligence. The National Security Act of 1947 prohibits both top C.I.A. positions from being held simultaneously by military men. When Johnson moved Carter, 53, to N.S.A., the general remarked: “I’ve had some beauts, but this beats them all.” Though he has had no specifically cryptologic experience, Carter is probably the best prepared of all N.S.A. directors in view of the great pervasiveness and importance of modern communications intelligence. A 1931 graduate of West Point, he spent his first ten years with various antiaircraft artillery units and teaching in West Point’s Department of Natural and Experimental Philosophy. He spent most of World War II in the logistics group, Operations Division, War Department General Staff, and after a brief tour in China was named special representative in Washington for General Marshall, then on his China mission. When Marshall became Secretary of State, Carter became his special assistant in January of 1947, serving for two years and undoubtedly getting a good picture of American foreign policy. Between 1943 and 1949, he attended six international conferences, including the Big Four at Cairo and two U.N. General Assemblies. After brief tours in the American embassy at London, as a student at the National War College, and as commander of an antiaircraft group in Japan, he served from 1950 to 1952 as director of the executive office of the Secretary of Defense under General Marshall and his successor, here gaining an overall view of American defense. From November 1952
until his appointment to the C.I.A. ten years later, Carter held various command posts in infantry, antiaircraft, and air-defense units. His three years as No. 2 man in the C.I.A. must have given him vast experience in seeing where communications intelligence fits into the general intelligence pattern and, perhaps, many ideas for helping N.S.A. better fulfill its mission.

  Such are the chieftains of history’s largest cryptologic unit. Though their power dwarfs that of England’s 18th-century Decyphering Branch, even on a relative basis, the directors probably do not possess even a tenth of Bishop Willes’s cryptanalytic ability. Nor do they need it. They have cryptanalysts to handle that particular specialty among the dozens of specialized functions that modern cryptology entails. Their own task is directed outward, at Congressional committees, at the United States Intelligence Board, at the Secretary of Defense, at the heads of the service cryptologic agencies. For internal administration, they appear to lean on their deputy directors.

  These deputy directors have come from backgrounds even more varied than their chiefs’. The first, Captain Joseph N. Wenger, U.S.N., who was named A.F.S.A. deputy director on July 15, 1949, had spent most of his naval career after graduation from Annapolis in 1923 in communications. He headed the Navy’s cryptanalytic agency at Nebraska Avenue during most of World War II, rising to deputy director of naval communications. When N.S.A. was created, he became vice director, a post that seems to be no longer in existence; he left N.S.A. in August, 1953.

  Deputy director for about four years, ending in 1956, was Joseph H. Ream, a lawyer who had worked his way up the corporate ladder of the Columbia Broadcasting System to become secretary, director, and executive vice president before going to N.S.A. Afterwards he returned to C.B.S., becoming vice president of its television network in 1959. His N.S.A. successor was the late Howard T. Engstrom, holder of a Yale Ph.D. in mathematics who had been vice president of Remington Rand, Inc., for three years when he was tapped for N.S.A. He had taught mathematics at Yale from 1926 to 1941, when he may have become a Navy cryptanalyst. After leaving N.S.A. in 1958, he became vice president of Sperry Rand, Inc.