The Second World War
Page 63
The ‘indirect’ offensive encouraged and sustained by the Allies against Hitler – military assistance to partisans, sabotage and subversion – must therefore be judged to have contributed materially little to his defeat. Among his army of 300 divisions deployed across Europe on 6 June 1944, the last moment in the war when Hitler exercised unchallenged control over the greater part of the territory he had conquered in 1939-41, fewer than twenty can be identified as committed to internal security duty. Outside central Yugoslavia, parts of western Russia far behind the Wehrmacht’s lines and tiny pockets of defiance in mountain Greece, Albania and southern France, all peripheral to his conduct of the larger war, occupied Europe lay inert under the jackboot. The ‘dawn of liberation’, so seductively promised by Churchill, Roosevelt and the governments in exile to the conquered populations, was signalled only by the flicker of gunfire at the military boundaries of the Wehrmacht’s zone of operations.
The bacillus of espionage
If the structure of Hitler’s empire was penetrated and fissured by clandestine activity, that took a form quite different from the ‘setting ablaze’ Churchill had so optimistically demanded in July 1940. Resistance may have been a gnat on the hide of the Wehrmacht; espionage was a bacillus debilitating its vital system. The real triumph of the Allies’ indirect campaign against Hitler between 1939 and 1945 was won not by the brave and often foolhardy saboteur or guerrilla warrior but by the anonymous spy and the chairborne cryptographer.
Of the two, spies were by far the less important. Popular imagination invests ‘Humint’ (human intelligence, in the jargon of the trade) with a significance even greater than it gives to the resistance fighter; certainly far greater than is given to ‘Sigint’ (signals intelligence). Moreover, governments to a considerable degree conform to popular estimations of the worth of a spy. The notion that the inner councils of the enemy are penetrated by an ‘agent in place’ who transmits their deliberations and decisions swiftly and directly to his master in the friendly camp is beguilingly attractive to any war leader; Churchill, Roosevelt, Stalin and Hitler were all beguiled in this way during the Second World War. Hitler, for example, was led to believe by the Wehrmacht intelligence service, the Abwehr, that it maintained an extensive network of agents in Britain who were able to report after June 1944 on the accuracy of pilotless-weapons strikes on London. Churchill and later Roosevelt were supplied through the Czech intelligence service with significant information of German capabilities and intentions by Agent A-54, probably the Abwehr officer Paul Thümmel. Stalin, who benefited from the dedication of the international communist movement’s members to the Soviet cause, drew on the Swiss-based ‘Lucy ring’ for information of German activities in occupied Europe, on Richard Sorge’s network in Japan for warning of German military intentions, on the ‘Red Orchestra’ for day-by-day intelligence of the German order of battle and, during 1941-2, on the Schulze-Boysen Luftwaffe network (a Red Orchestra component) for technical data; through the ‘Cambridge Comintern’, composed of British intelligence officers, Anthony Blunt, Guy Burgess and Kim Philby, Stalin also kept his finger intermittently on the pulse of Anglo-American higher strategy.
All these sources were, however, to some degree unsatisfactory or compromised. Agent A-54’s transmissions, for example, were too sporadic to provide a coherent picture of German strategy. The Red Orchestra was wrongly placed to monitor key German military activity, the Schulze-Boysen network was insecure and quickly penetrated (117 of its members were hanged), Sorge was too detached from the Comintern network to be always believed (though he undoubtedly influenced Stalin’s decision to transfer troops from Siberia to Moscow in the winter of 1941), while the Lucy ring was probably transmitting information tailored and supplemented by the Swiss intelligence service for its own purposes; another interpretation is that the Lucy ring derived its information from Bletchley, either directly or through Allied agents in Swiss intelligence. The Cambridge Comintern was perhaps the most influential of all Stalin’s networks, though in a reverse direction; Philby, by deliberately maligning his circle’s reliability, may have been decisive in dissuading the British government from lending support to Stauffenberg’s anti-Hitler conspiracy, which Stalin undoubtedly judged inimical to his long-term plans for the post-war control of Germany, since the conspirators were avowedly anti-communist. Least satisfactory of all the networks was the Abwehr’s in Britain; it was turned as early as 1939 by the capture of one of its spies, and after that all the information it transmitted to Germany was compiled by the British officers of the ‘Double Cross’ organisation which controlled subsequently captured agents. One of the controllers’ achievements was to persuade the German staff of the pilotless-weapons force progressively to shorten the range of their launched missiles, so that the majority fell south of London.
The role of Ultra
The contribution of ‘Humint’ to the direction of strategy in the Second World War looks all the more marginal and patchy when compared with that of ‘Sigint’. Signals intelligence is concerned with the interception, decryption and interpretation of the enemy’s secure messages, however transmitted, and the protection of one’s own from his interception services. In practice the majority of material so intercepted in the Second World War was radio traffic protected by elaborate mathematical ciphers, though some traffic was sent by the older system of codes, constructed from secret codebooks; much tactical intelligence, known to the British as ‘Y’, was gleaned from messages sent in ‘low-grade’ ciphers or even en clair (not in cipher) between units in the heat of the action.
All five major combatants, Germany, Britain, the United States, the Soviet Union and Japan, devoted costly and extensive efforts to protecting their radio signals traffic and seeking to penetrate that of the enemy (telex, telegraph and telephone traffic sent along landlines – which carried, for example, 71 per cent of all German naval traffic in 1943 – proved generally impenetrable, but had limited use). Minor combatants, notably the Poles, French and Italians (whose codes and ciphers were exceptionally secure), also played a major role in the radio warfare. In particular it was the Poles’ early success in attacking Germany’s military machine cipher (Enigma), subsequently revealed to the French, which eventually allowed the British to break Enigma on a regular and rapid basis, thus laying the foundations for the Ultra organisation, which disclosed information of truly war-winning value to the Allies from late 1940 onwards.
The extent of the Ultra triumph is now so well known (the equally important American ‘Magic’ penetration of the Japanese naval and diplomatic ciphers less so) that consideration of both ought to wait upon an evaluation of the cryptographic successes of other combatants, which were more significant than received opinion admits. It is probable, for example, that the Russians had their own success against Enigma; so Professor Harvey Hinsley, the historian of Ultra, guardedly indicates. Russia’s own high-grade ciphers were certainly of the best quality, and this suggests a capacity to read others; they had not yielded to foreign intelligence services’ attack for some years before 1941 (Churchill forbade the British Government Code and Cipher School from making the attempt after 22 June 1941). Russian medium- and low-grade ciphers, however, were much read by the Germans in 1941-2 and perhaps later; the decrypts yielded valuable tactical intelligence. The Germans also succeeded in breaking the American military-attaché code in late 1941, and their decrypts of messages originating with the US liaison officer in Cairo supplied Rommel with important information about the Eighth Army in the desert.
The most important German success, however, was in breaking the British naval book ciphers, to which the Admiralty clung long after the army and air force had gone over to more secure ciphers. These ciphers (nos. 1, 2, 3 and 4) were more properly book codes: the letters of a message were translated into figures by reference to standard tables and then ‘super-enciphered’ by mathematical techniques, which were altered at regular intervals. The weakness of the system was that the book itself might be reco
nstructed if sufficient radio traffic was collected and analysed by the enemy, who could then apply normal calculations of mathematical probability to break the super-encipherment. This was exactly the achievement of the German navy’s Observation Service (Beobachtungs- or B-Dienst); by April 1940 it was reading 30-50 per cent of traffic encoded in naval cipher no. 1; when a change was made to cipher no. 2 it again broke this traffic on a large scale between September 1941 and January 1942; it had less success with the replacement, no. 4; but between February 1942 and June 1943, with short interruptions, it was reading as much as 80 per cent of no. 3, often in ‘real time’. ‘Real time’, a cryptographer’s term, means that messages sent are intercepted and broken by the enemy at the same speed as they are received and decrypted (or decoded) at the denominated station. What made the B-Dienst’s success against naval cipher no. 3 so disastrous for the Allies was that it was the code used to carry information between London and Washington about transatlantic convoys; as a result the B-Dienst ‘was sometimes obtaining decrypts about convoy movements between 10 and 20 hours in advance’ of their departure, according to Harold Hinsley. Such information was the key to the success of Dönitz’s U-boat wolf packs, which, alerted by such intelligence, could be deployed across the track of an east- or west-bound convoy in numbers that frequently overwhelmed their escorts. It was not until the Admiralty accepted a combined cipher machine (CCM), employed also by the American and Canadian navies and brought into general use in 1943, that the German navy’s penetration of convoy ciphers in the Battle of the Atlantic was ended. By then, however, the balance of advantage in the battle had been shifted to the Allies by conventional military means.
In the land and air war, the Allied armies and air forces conceded no such advantage to the enemy as the British Admiralty did by its arrogant persistence in the use of book codes; they used machine cipher systems from the outset and in consequence resisted German attack on their secure transmissions. The German services were also committed to a machine cipher system, but, having adopted theirs ten years before the British and Americans, found themselves – unwittingly – at the outbreak of the war equipped with a semi-obsolete system. Hence the success of the British Government Code and Cipher School (GCCS), located at Bletchley, between Oxford and Cambridge (from which it drew so much of its talent), at breaking into the transmissions encrypted on the Enigma machine the Germans used.
The Enigma machine outwardly resembled a portable typewriter but the depression of a key worked an internal system of gears which allotted any letter input an alternative letter not logically to be repeated before 200 trillion subsequent depressions. The Germans therefore understandably regarded Enigma transmissions as unbreakable in ‘real time’, indeed in any sort of human time whatsoever. Unfortunately for them they were deceived; because of the need to indicate to a receiving station the way in which the sending Enigma machine had been geared to transmit, the operator was obliged to preface each message with a repeated sequence of the same letters. This established a pattern which a trained mathematician could use as a ‘break’ into the message and so into the whole of its meaning. As the mathematicians recruited by Bletchley included Alan Turing, the author of universal computing theory, Gordon Welchman, a principal pioneer of operational analysis, and Max Norman and Thomas Flowers, designer and builder respectively of the first electronic computer (‘Colossus’, so called at Bletchley because of its enormous size), such ‘breaks’ were rapidly exploited to yield complete readings of German messages quite quickly after interception – and eventually in ‘real time’.
There were important exceptions to Bletchley’s success. Luftwaffe ‘keys’ – the different methods of enciphering Enigma traffic used by separate German service branches – proved easier to break than army and naval keys, and some naval keys were never broken; significantly, nor was the Gestapo key, though it was not changed from 1939 until the end of the war. Enigma security depended heavily upon the experience and skill of the German senders; mistakes in procedure made by inexperienced, tired or lazy operators provided Bletchley with the majority of their ‘breaks’ into traffic. Gestapo operators were meticulous; so too were the naval officers who used the ‘Shark’ key controlling U-boats in the North Atlantic. For most of 1942, while the B-Dienst was regularly reading naval cipher no. 3 in real time, Shark resisted all Bletchley’s efforts; during those months (February-December) the Germans were masters of the radio war in the Battle of the Atlantic. In consequence hundreds of thousands of tons of Allied shipping were sunk.
The Shark episode was, however, an exception to the general rule that the British dominated radio warfare in the West, as the Americans did in the Pacific, where they matched Bletchley’s achievement by breaking both the Japanese naval (JN 25b) and diplomatic (Purple) machine ciphers before the outbreak of the war. The joint Allied triumph, successfully concealed from both enemies throughout the war, naturally prompts the query why, if the Allies enjoyed such direct access to the most secret internal messages sent by their opponents, they were nevertheless on occasion suprised by large-scale enemy initiatives – Pearl Harbor, Crete and the Ardennes offensive being the obvious examples. The answer is that there are limits to the usefulness of even the best intelligence system, and this is precisely demonstrated in different forms by each of these episodes. The Japanese, for example, disguised their intentions before Pearl Harbor by hiding their fleet in the remoteness of the Pacific and imposing absolute radio silence on its units, which moved to their attack positions by preordained plan; alert though the Americans were, they were thereby denied the intelligence which would have allowed them to anticipate attack. Before the Ardennes the Germans also imposed radio silence on their attack units. Nevertheless, primarily through troop movements, they unwittingly betrayed sufficient warning to the Allies of the attack they planned for a truly sensitive intelligence organisation to have detected the danger and alerted higher authority. Both the intelligence service and higher headquarters had, however, persuaded themselves that the Germans were too weak to launch an offensive on the Ardennes front during December 1944; accordingly they discounted the evidence to the contrary and so were discountenanced.
The case of Crete reveals a third and highly frustrating limitation in the use of intelligence: the inability to act on clear warning because of disabling weakness. Before the German parachute landings in May 1941, Ultra – the organisation which evaluated and distributed the raw decrypts produced by Bletchley – had identified from Enigma intercepts both the German order of battle and the German plan. However, Freyberg, the British commander on the island, lacked not only the troops but also (more ironically) the transport that would have enabled him to concentrate counter-attack forces swiftly against points of danger. As a result the Germans, despite heavy initial losses, seized a vital airfield which allowed them to fly in reinforcements and swamp the defences.
There is a fourth and universal limitation on the usefulness of intelligence: the need to protect a source. It has been widely alleged, for example, that Churchill ‘allowed’ Coventry to be bombed in November 1940 because to have taken extraordinary defensive measures against the attack would have revealed to the Germans the ‘Ultra secret’. It is now known that this interpretation is false; although Churchill did indeed have advance warning via Ultra of the Coventry raid, it was too short to enable defensive measures to be taken – which he would certainly have done, at whatever the risk of compromising Ultra, had time been available. A more telling accusation is that in the weeks before Barbarossa the British did not validate their warnings to the Russians of the imminence of the German attack by revealing the authenticity of the source. However, in view of Stalin’s wishful thinking to the contrary and his desire to placate Hitler at any cost, the betrayal to him of the Ultra secret would have plumbed the depths of insecurity. In this case, as in every other where such a calculation had to be made, Churchill was unquestionably right to put the long-term security of the source above current advantage.
Ultra’s greatest contribution to the war in the West, however, occurred during the Battle of Normandy, when Bletchley provided Montgomery with information of day-to-day German strengths at the battlefront, of the effect of Allied air-strikes, such as that which destroyed the headquarters of Panzer Group West on 10 June, and eventually of Hitler’s order to counter-attack at Mortain against the flank of Patton’s break-out into Brittany – a disclosure which led to the destruction of Army Group B’s armoured reserve and to the climactic encirclement of the Westheer in the Falaise pocket. The Mortain decrypts were certainly the most important which came to any general on any front throughout the course of the Second World War.