Prof. John FERRIS: Polish Intelligence and the Road to Ultra, 1925-45

Polish Intelligence and the Road to Ultra, 1925-45

Photo of Prof. John FERRIS

Prof. John FERRIS

Professor of History at the University of Calgary, an Honorary Professor at the Department of International Politics of the University of Aberystwyth, and the Department of Law and Politics, Brunel University. Expert in diplomatic, intelligence, imperial, military and strategic history. His latest book is Behind the Enigma: The Authorised History of GCHQ, Britain’s Secret Cyber Intelligence Agency.

other articles by this author

The common view that Ultra shaved two years off the war is hard to sustain, but it did save the allies months in time and hundreds of thousands of lives. Britons and Poles can take pride in that achievement. Neither needs to exaggerate it.

.During the interwar years, Poland was a regional state which its leaders thought was a Great Power. It had remarkable weaknesses and strengths, like a tiny industrial sector, and good military intelligence. Poland’s intelligence triumph was against the German cryptographic machine, “Enigma”. This work was hard, innovative, essential for the success of “Ultra” during the Second World War, and valuable to the Allies, but it did little direct good for Poland. This story usually is told in anecdotes, which come from nationalist narratives, whether British and Polish, and structured as myth. In order to be understood as history, that story needs context.

During 1919, the new state of Poland developed a signals intelligence agency based around amateur cryptologists, mathematicians and veterans of an excellent codebreaking agency, the Austro-Hungarian Chiffregruppe. They broke Soviet codes and enabled Poland’s partial victory in the Russo-Polish war of 1919-21. In codebreaking, Poland was a Great Power, though so were other medium sized states of the time, like Finland and Sweden. Far more than most cryptanalytic agencies of the era, Poles focused on military rather than diplomatic systems. Poland, caught between two hostile neighbours, Germany and the USSR, looked for every means to boost its strength against them, including signals intelligence. Polish codebreakers split their resources between these two targets. We know little about their successes against Soviet systems, because Poles destroyed these records in 1939, but Polish work against Germany is well known. 

During the 1920s, many states considered adopting cryptographic machines so to bolster the security of their signals, but Germany alone did so. Its military services purchased the rights to a commercial machine, Enigma, which was mechanically reliable but had poor security, with the aim of improving it. Over coming years, German military forces put versions of Enigma into service, slowly boosted its security and developed the best military cryptographic system of the interwar period. Though Enigma, an electro-mechanical system based on rotors, is best known because it was broken, it was a good machine, which Germans often used badly. Enigma could not be solved by the techniques used against codebooks. Success required acquiring a copy of the machine, determining how its electro-mechanical components worked, including how rotors were wired, and then attacking with a combination of mathematics, to discover fractures in the system, which the brute force of handheld devices or machines then shattered. 

Other states knew that Germany had adopted Enigma, but did not attack it. Until 1937, the British Government Code& Cypher School ( GC&CS)  did not attack any German systems, although it assessed the commercial version of Enigma, and understood its weaknesses. Instead, it launched a failed campaign against Soviet cryptosystems, which were increasingly powerful. French military intelligence merely monitored German developments. Only Poland struck Enigma, through the greatest technical achievement in codebreaking of the interwar years. It began when German signals were easy to collect, but technical improvements in Enigma were weak. From the start, Polish intelligence created a special unit to conduct the attack: employing mathematicians as codebreakers. GC&CS did so only in 1939, just before war began. 

In 1928, Polish military intelligence acquired, accidently, a copy of the commercial Enigma machine, and then began to intercept German military signals sent by their version of Enigma. Poles chose to attack these signals, and through an unusual means. Though Poland had codebreakers experienced against book systems, it assaulted Enigma through mathematicians, who had given good service during the Russo-Polish war.  In 1929, Polish intelligence created a university course for mathematicians and others, focused on radio and cryptanalysis. In 1932, Polish intelligence hired these mathematicians, Marion Rejewski, Jerzy Rozyki and Henryk Zygalski, to attack Enigma. They had backgrounds in “group theory”, the study of mathematical permutations. That expertise was valuable against Enigma at that stage, though it became less useful as the machine acquired stronger security. This team developed a mathematical analysis of Enigma rotors which, however, was bitty in many places. This problem was overcome when it received manuals, operating procedures and rotor key settings for military Enigma, which French army intelligence acquired through a well-placed agent, and provided to Poland. Rejewski and his colleagues quickly reconstructed the working of the German military Enigma of the day, which enabled large scale and real-time solution of messages. Over the next six years, Germans frequently changed details of their system, but still the Poles kept pace, in particular by using handheld devices to speed the attack, and ultimately through an electro-mechanical system ( nicknamed “bombes”) through which they could test rapidly possible solutions to Enigma keys.    

By 1939, however, handheld devices could no longer generate the brute force needed to break Enigma: only data processing machines could do so, and only if well used and in strength. Anyone could lease IBM data processing machines, but these were expensive. Before 1939, the only signals intelligence service which regularly used data processing machines was the United States Navy’s Op 20 G, which had much money but few codebreakers. However, when war began in 1939, other countries immediately discovered the power of data processing machines against all forms of codes, and acquired them in numbers ranging from dozens to hundreds. Meanwhile, IBM itself hampered the use of machines for cryptanalysis, because it would not let consumers improve how they worked, with one exception: British Tabulating Machines ( BTM). By 1939, BTM had one of the world’s best data processing machines, and personnel skilled in designing and producing them. Poland lacked the resources to continue its successes against Enigma, while Britain had the capability but did not know how to use it. By mid 1939, British research against German military Enigma, conducted by one of its best codebreakers, Dillwyn “Dilly”  Knox, had gone some way, but suffered from one elementary flaw: an inability to determine the first connections ( or the “entry plate”) between the Enigma keyboard, and the rotors, which worked through simple alphabetical order. British codebreakers could not imagine that German cryptographers would adopt so vulnerable a practice. The British were two steps from success—understanding how the system worked, and the need to attack it with machines. GC&CS was close, but that counts in the game of horseshoes, not in cryptology.    

Poles broke this stalemate. During the 1930s, British, French and Polish codebreakers rarely cooperated. French and British codebreakers liaised loosely against Soviet systems, as did French and Polish intelligence against Enigma. Poland perhaps cooperated more with Japan ( against their common Soviet problem) than any other country, while British authorities underestimated and mistrusted Polish intelligence. In July 1939, Polish codebreakers called their British and French counterparts to a meeting outside Warsaw, and revealed the history and techniques of their work against Enigma. This meeting built trust between British and Polish codebreakers. Rejewsky was amazed by how quickly Knox understood Polish techniques, and the British were staggered by Polish success at a task which now was their highest priority. Poland’s continuous work against Enigma overcame prior British and French failures, and enabled Ultra. The British and French teams rapidly reported this news home, and the three allies began to cooperate. When Poland was overrun, many of its codebreakers escaped through Rumania, and continued their work in Britain and France. 

During the first nine months of the war, there was a cross-pollination between British, French and Polish codebreakers. This relationship was as intimate as that which later emerged between American and British codebreakers, and would have stayed that way had France not fallen. The mathematicians whom GC&CS recruited, including the founder of computing science, Alan Turing, were more sophisticated than their Polish counterparts, and approached the attack on Enigma with fresh and powerful techniques. Yet Turing frequently visited Polish codebreakers, speaking their only common language, German, and adopted and adapted their best procedures. In particular, with the assistance of Gordon Welchman, he devised a far more powerful “bombe”, which BTM was able to produce with remarkable speed.  The manufacture of the first two bombes let Britain combine mathematics and machines in attack, but not until 1942 did it have enough bombes to rely on brute force assaults. Instead, during 1940-41, British attacks still rested on Polish procedures, and handheld devices. Without Poland, there could have been no Ultra, but Poles had no direct influence on the development or deployment of that brute force attack.     

This story involves a competition between two myths. Poles think that Britons underestimate their contribution to Ultra, with some truth. However, this error stemmed from secrecy, rather than snobbishness. Turing knew of the Polish contribution, but his colleague and near equal as a cryptanalyst, Welchman, did not. When Welchman learned of this contribution, he praised it. So too, The Official History of British Intelligence during the Second World War initially underestimated the Polish contribution, because it lacked records about the matter, but later offered apologies and praise. John Tiltman, the best classical British codebreaker of the age, later spoke almost tearfully of the debt owed to Polish siginters, who remained silent even when sentenced to Auschwitz. Contemporary British people know of the Polish contribution to Ultra, and value it. 

With even more truth, Poles complain that Britain exploited Polish work, and gave little in exchange. Bletchley needed Polish help to start its work, but not to finish it. Britain dominated the new technical work against Enigma, which it kept hidden from its allies, after Germany overran both of them, not surprisingly, given the risks to security. No signals intelligence service would have behaved differently. Britain also happily received and used the freely given product from Poland’s excellent human intelligence service. These were major contributions to British intelligence and to the war, of which Poles can be proud. Poles complain that Britain did not back them enough against Soviet imperialism during the war. In fact, Britain did some things for Poland. Britain could have done more than it did, but such actions would have failed, given the scale of Soviet power. Poland was a Great Power in intelligence, but not military strength, in a tough neighbourhood. Having entered a war against greater states, and being crushed by them, Poland could gain little direct help from its intelligence services. Britain did not create this tragedy, but it ensured that Polish codebreaking helped to defeat the most dangerous of Poland’s enemies.    

.Both British and Polish national myths overestimate the impact of Ultra on the war, and thus of their own place in it. The story is told uncritically, from the perspective of Allied sword against Axis shield, at Ultra’s peak. Intelligence, however, shapes actions, but does not make them. It affects events in complex ways. The intelligence war was a competition, involving Axis successes and Allied failures. Given German gains from attacking American, British and above all, Soviet, codes, Germany did well in the signals intelligence war. Ultra did not create allied victory, though it sped that outcome. The common view that Ultra shaved two years off the war is hard to sustain, but it did save the allies months in time and hundreds of thousands of lives. Britons and Poles can take pride in that achievement. Neither needs to exaggerate it.      

John Ferris

This content is protected by copyright. Any further distribution without the authors permission is forbidden. 25/11/2022
Fot. Bletchley Park Trust / SSPL / Forum