by DAVID KAHN
The Soviet military establishment seems to have had neither the tradition, the work force, nor the success in cryptanalysis that the secret police enjoyed. The inclusion of a military unit within the Spets-Otdel in 1933 testifies to this subordination. In any event, much less has been heard of it. Probably this is because each branch of the Soviet armed forces restricts itself to the crypto-systems of its opposite numbers—the Red Army to the secret communications of the German, Japanese, British, American, and other armies, and similarly for the Red Navy and Air Force. Cryptanalysis naturally constitutes a part of intelligence, and in 1941, the G.R.U., the Army’s Chief Intelligence Directorate, had its cryptanalytic group as Section 8 of the eight sections of its operations branch: (1) Europe, (2) Near East, (3) Western Hemisphere and India, (4) advanced technical information, (5) terroristic activities, (6) production of false documents, (7) frontiers, (8) cryptology. What these had in common was the production of raw military intelligence by both clandestine and open methods. (In addition to operations, the G.R.U. had three other branches: information, which evaluated and disseminated operations’ output; training; and auxiliary, which handled the housekeeping duties.)
In 1943, when Soviet military intelligence was reorganized into strategic and tactical branches, it expanded into a block-square baroque palace at 19 Znamensky Street, while retaining its former quarters in a building on the square at the Kropotkin Gate. There were also several auxiliary buildings, including a factory for photographic paper outside of Moscow, nearly all of whose output was consumed by a white two-story building in a yard of the intelligence complex. This was the photographic laboratory that produced and developed the films used by Soviet military intelligence for much of its mail communication with its agents abroad. What appeared to be a gold research institute on the Vorobiovy Gory was actually the Osobyi Radio Divizion (“Special Radio Division”), by which the G.R.U. maintained radio contact with its secret agents around the world. When Soviet spies tapped out their messages to the impersonal “Centre” in Moscow, an O.R.D. radioman received and acknowledged them. And when the agents received their orders, curtly signed “Director” (of Soviet military intelligence), the O.R.D. transmitted them. It was their only link with the agency for which they risked their lives. The O.R.D. was also staffed with radio technicians, who assigned wavelengths and schedules that would provide optimum reception from various points on the globe, and with clerks who allocated call-signs.
Cryptography was handled in a separate G.R.U. branch headed by Lieutenant Colonel Kravchenko. Among its clerks was another man not unknown to future fame—Igor Gouzenko. “I well remember the first telegram I was given to decipher at Intelligence Headquarters,” he wrote. “It came from Harbin [in Manchuria] …. The telegram sounded like a page from a novel, giving minute details of the hiding place [of an agent’s radio] in the vicinity of the Governor-General’s palace, and described habits of the people living in that district…. The follow-up telegram was then given me for coding. It gave instructions for meeting the pick-up man in Harbin. The actual street corner and the alternative street corner were named together with the time, the day, tokens of recognition and passwords to be used.” Gouzenko and his fellow clerks, among them his friend Lieutenant Burukin, sometimes followed agents day after day, living in imagination all the thrills and danger of a spy’s life as they routinely deciphered his messages.
Soviet cipher clerks learned their trade at a number of schools. Gouzenko took the basic course at the Kuibishev Military Engineering Academy, where the political director was an owl-eyed former cipher clerk named Maslennikov, nicknamed Kriptus; despite his twisted body and rather pathetic personality, he was a good teacher and knew a lot about cryptography. Gouzenko got advanced training at the Higher Red Army School, generally known as the Intelligence Academy. Cryptography was among the courses taught at the Red Navy’s Electric Mine School at the big base at Kronstadt; here, Petrov, as a young recruit, first studied it two winters and two springs, going to sea during the summers and autumns. The course included some cryptanalysis. Afterwards, he served for two years as senior cipher rating aboard the destroyer Volodarsky, where he lived and worked in a small cabin under the bridge. His military service completed, he left the Navy and joined the O.G.P.U. as a cipher clerk. Cryptography may also have been taught in the Red Army’s Military School for Signal Communication in Leningrad; this would be for use with the troops. The Army also ran a research institute for communications at Sokolniti in Moscow, which, since it encompassed such exotic studies as cosmic rays, almost certainly included cryptography. In 1937 it merged with another institute into a central scientific research institute for the Red Army.
The strides that the Russian Army had made in cryptography after the traumatic experiences of World War I were dramatized by an interchange of messages between incredulous Russian units at the very start of the Russo-German War. Moments after the Nazis launched their sneak attack at 3:30 a.m. June 22, 1941, a Red outpost wirelessed frantically, “We are being fired on. What shall we do?” Back came the stern reply, “You must be insane. And why is your signal not in code?”
Red Army cryptography rested in World War II upon the enciphered code. The system appeared in four series: 5-digit codes for strategic messages, 4-digit for high-level tactical communications, perhaps of the rank of army headquarters, 3-digit for medium-level tactical, as of brigade rank, and 2-digit for the front. The Soviets replaced their tactical codes frequently, although sometimes a code that had been used in one sector of their thousand-mile front reappeared later in another. The 4-digit codes were enciphered by 10 × 10 tables, one table for the first 2 digits and another for the second pair. The 5-digit codes were enciphered by additive tables of 300 groups changed daily. The Army and Navy shared the 5-digit strategic system; border patrol and N.K.V.D. units had their own systems, usually 4-digit. In addition, the Soviets got some Hagelin M-209s in Lend Lease, which they apparently used as models for their own constructions, though it is not known where these were used.
With enough traffic, enciphered code can of course be read. One of the first to do so in the case of the Russian military was the Swedish expert Arne Beurling. During the bitter struggle of Finland against Russian aggressors in the Winter War of 1939-1940, Sweden fed intelligence based on cryptanalysis to her neighbor. Beurling attacked the top system, the 5-digit strategic, which was actually a 4-digit code with an extra digit added as some form of check. In several of the codes, the page digit—the second—was repeated, so that the groups would look like 52217, 88824, and so on. In others, the fifth digit gave the unit total of the preceding four digits, so that 6432 would have a check digit of 5, making the codegroup 64325. Beurling wrote the cryptograms on a sheet of graph paper with five-millimeter squares that was so large—about 3 × 4 feet—that he had to order it specially. He would make his runs, and, if messages overlapped, he would be on his way.
Soviet strategy against Finland called for a five-pronged invasion along their north-south border. The middle force drove for the tiny village of Suomussalmi to cut Finland at her waist; the force just north of that one rolled on another little village, Salla, in a secondary cutoff. But intelligence developed in the Swedish cryptanalytic office helped the Finns to repulse both Russian attacks.
Crucial to Marshal Mannerheim’s victory at Suomussalmi was the information that the Russian 44th Division, a crack motorized outfit from Moscow, was advancing from Raate. He immediately sent reinforcements to Suomussalmi. Two days after his five battalions reached there, the Finns, dressed in white and moving like the ghosts of the north, attacked the Russian forces in the village, broke their resistance, and forced them to flee across frozen Lake Kiantajärvi. The skiing wraiths then cut off the retreat of the 44th Division, severed its column and destroyed it section by section in fighting that continued into the first week of 1940. Large quantities of stores were captured, but, Mannerheim wrote, “The enemy’s casualties could not be estimated, as great snowdrifts over a
large area covered the fallen as well as the wounded who were frozen to death.”
Temperatures during the battle dropped to 56 degrees below zero, and it was under such conditions that the Swedes solved some pitiful messages from isolated Russian units. One encircled group radioed that they were burning their papers and were going to shoot their last horse for food, and that this was their final message. Silence followed, and soon thereafter the Swedish cryptanalysts learned that the Finns had crushed them. Another Russian battalion sent a coded message that they were desperately short of supplies and would build three fires in a triangle to show the Red Air Force where to parachute desperately needed food and ammunition. The Swedes solved it and gave it to the Finns, who built a triangle of fires and watched with bitter satisfaction as the packages floated down into it.
Swarms of Russian Air Force cryptograms were downed by the Swedish codebreakers. Many were orders to bomb Helsinki, and often these were solved before the bombers took off from airfields in Latvia and Estonia for the 20-minute flight to their target. Finnish authorities thus had ample time to sound air-raid alerts; as a result the capital suffered exceptionally light civilian casualties considering the number of bombs dropped.
But little Finland was no match for the colossal U.S.S.R. despite her cryptologic advantages, and in March she signed a peace treaty. When the Germans invaded Russia a year later, Finland declared war against her harasser and later exchanged cryptographic intercepts with her new ally.
German radio intelligence against the Soviet Union appears to have been characterized by a severe split. Strategically it enjoyed no success at all. The Germans did not solve the cryptosystems of the top Soviet military commands—primarily the 5-digit codes. Perhaps by 1941 the Russians had corrected their cryptographic technique enough to keep the Germans from repeating the 1939 Swedish successes. Whatever the reason, cryptanalysis contributed little to O.K.W.’s overall picture of Russian strategy.
Tactically, however, the Germans reaped rich harvests of intelligence. In mid-1940, when Hitler first decided to attack the Soviet Union, Germany had no radio-intelligence service of any kind in the East; a year later, when Hitler struck, the new intercept service had already provided him with good information on Russian order of battle. In July, a captured Russian Air Force captain betrayed one of the Air Force systems. This intelligence windfall helped the Luftwaffe destroy hundreds of Soviet airplanes on the ground and another hundred in a great air battle over Minsk.
The resultant air superiority, plus surprise, momentum, armor, and speed, carried the Wehrmacht forward in a surge of victories. In 1941 and again in 1942 Germany mounted massive offensives and overran vast areas of Russia. But in the winter of 1942-43, Stalingrad held and the German 6th Army capitulated; at the same time, Germany lifted the two-year siege of Leningrad. By next summer, it had become evident that Nazidom could not win its great victory over Bolshevism, but the troops hoped at least for a stalemate that would stabilize their conquests. The high command decided on some limited attacks to cripple Soviet offensive power. With the waning of Luftwaffe air mastery, Nazi intelligence had to depend less upon aerial reconnaissance and more upon wireless surveillance. In tactical operations during the Battle of the Dnieper in October, 1943, the chief of staff of the 48th Panzer Corps declared, “The best and most reliable source of intelligence was our Wireless Intercept Service.”
A few months later, that corps participated in one of the attacks that Army Group South, one of the three major German groupings on the Eastern Front, mounted to flatten the Kiev salient and further forestall Soviet offensive propensities. The 48th Panzer Corps had as its objective the disruption of the Russian 60th Army. Air reconnaissance produced no information, and the corps decided not to send out ground scouts for fear of alerting the Russians. The attack at 6 a.m. December 6 completely surprised the Russians, who recoiled in confusion.
In those days [wrote the corps’ chief of staff, Colonel F. W. von Mellenthin] we were really good at intercepting Russian wireless traffic; enemy messages were promptly deciphered and passed to Corps in time to act on them. We were kept well informed of Russian reactions to our movements, and the measures they proposed to take, and we modified our own plans accordingly. At first the Russians underestimated the importance of the German thrust. Later a few antitank guns were thrown into the fray. Then slowly the Russian Command got worried. Wireless calls became frantic. “Report at once where the enemy comes from. Your message is unbelievable.” Reply: “Ask the Devil’s grandmother; how should I know where the enemy comes from?” (Whenever the Devil and his near relations are mentioned in Russian signals one can assume that a crack-up is at hand.) Towards noon the Russian 60th Army went off the air, and soon afterwards our tanks overran the army headquarters.
By that evening the Germans had rolled up the Russian front for 20 miles, and by the night of December 9 the Soviets’ projected offensive was jolted thoroughly off balance. In the next few days additional blows punished them further. “The Russians were certainly flabbergasted by these ghost-like thrusts, which seemed to come from nowhere, and their wireless traffic provided abundant evidence of their bewilderment and anxiety,” Mellenthin wrote.
This German victory at the Battle of Radomyshl delayed but did not prevent the Russian offense. At Christmas, Army Group South began its retreat from the Ukraine. Several months later the Russians had driven the Germans back 650 miles from their farthest penetration.
Mellenthin has remarked that “The Red Army of World War II was vastly different from the Imperial Russian Army of 1914-17, but in two important respects the Russians have not changed. They are still addicted to mass attacks, and they still show an extraordinary indifference to wireless security.” This comment seems to hold true only in a tactical sense, and the adjective “extraordinary” is probably justified only under conditions of retreat and its accompanying confusion.
Army Group North, for example, read 5-digit code messages very rarely. Of the intercepts in 2-, 3-, and 4-digit codes, it read 28.7 per cent—13,312 messages out of 46,342 from the beginning of May, 1943, to the end of May, 1944, a year in which the Russians pushed back the northern sector of the front slightly, though not nearly as much as the southern. A month-by-month and system-by-system breakdown of the cryptanalytic success of Army Group North (excluding 5-digit codes) is shown in the table on p. 648.
As might be expected, the 2-digit systems, being the simplest, succumbed the oftenest. However, fewer 3-digit than the presumably more difficult 4-digit enciphered codes were solved, even though more 3-digit messages were picked up. The reasons for this seem to lie partly in the probable concentration on the information-rich 4-digit messages, partly in the many more 3-digit systems in use and the consequent difficulty of finding overlaps to strip off the additive and of getting sufficient text for solutions. This multiplicity of 3-digit systems can be seen in the number of new 3-digit systems reported solved each month by the cryptanalysts, which is invariably greater than the number of new 4-digit systems. In November, 1943, for example, Army Group North solved 15 new 3-digit systems as compared to one 4-digit; in December, the figures were 8 and 4, in January, 1944, 15 and 8. The cryptanalysts do not give the number of new systems introduced by the Soviets that the Germans did not solve.
Solved messages, said the cryptanalysts’ report for February, 1944, “contain operational combat reports, statements about assembly areas, command posts, loss and replacement reports, reports about chain of command and positions prepared for the attack (e.g., messages of the 122nd Armored Brigade on February 14 and 17). Besides these reports, the plaintexts of the messages made possible the identification of seven armored units, including their numerical designations, as well as confirmation of twelve armored units. With few exceptions the material could be worked up in good time and put to use.”
A Russian World War II military message partially solved by German cryptanalysts
These tactical solutions could, at best, produce local su
ccesses. The apparent failure of German cryptanalysts to solve Russia’s strategic cryptosystems, with the valuable information that they concealed, led one German cryptanalyst to adjudge that Russia lost World War I in the ether and won World War II there.
A truth he never suspected may lurk in his apothegm. For the Russians may have done as well in solving German cryptograms as in protecting their own. By 1942 they had cracked messages in the Enigma, a rotor machine. And the Germans themselves paid a left-handed tribute to Soviet cryptanalytic perspicacity when a 1943 conference of signal officers ruefully ordered: “It is forbidden to mark the Führer’s radio messages in any special way.”
At the same time, the Soviet Union guarded her diplomatic flanks by the one-time pad, a practice she had begun in 1930. Consequently her crucial Foreign Office messages were read by neither foes, nor neutrals, nor allies. Any schemes that she may have instigated against those who, at the end of the war, were to become either her puppets or her adversaries remained among the most inviolate of her secrets.
During World War II, the secret prospectors of the G.R.U. and the N.K.V.D. drilled for information in scores of places all over the world. Three of the spy crews struck gushers of it. The fabulous “Lucy” network in Switzerland, the Rote Kapelle in Germany, and the Sorge ring in Japan pumped a continuous stream of the most detailed and precise intelligence into the Kremlin. And this they did through a pipeline that, despite the most strenuous bangings and poundings of counterintelligence, remained hermetically sealed against cryptanalysis. All three rings employed the then-standard Soviet espionage cipher. It achieved a triumph of encipherment, for it is a system that the spymasters of the Soviet Union rightly regarded as unbreakable.
It brought the old Nihilist substitution to a peak of perfection. It merged the straddling checkerboard with the one-time key.
