Monday, August 22, 2016

Max Hastings, The Secret War

Spies are liars. It is part of the job description, and for most of them it becomes a part of their nature that they don't shed when it comes time to write their memoirs. So if, like me, you have read a stack of books about spying and codebreaking during World War II, you have perhaps wondered how much of them was true. If so, read The Secret War. It's an excellent book and I highly recommend it. Max Hastings works judiciously through the claims that have been made over the years for spying, sabotage, and Ultra, sorting out what really happened and what effect if any it had on the outcome of the war.

Because frankly a lot of intelligence work made no difference whatsoever. All the nations spent huge sums on spies who hatched cloak and dagger schemes in various neutral countries, all to no effect. From 1943 on various German officers came up with scheme after scheme to contact British or American agents with proposals for a separate peace and a joint war against the Soviets, all of which was a waste because the British and Americans were not going to make a separate peace. Of all the combatants only the Soviets got much value from secret agents, because the worldwide appeal of communism gave them a large pool of potential traitors and their own expertise in conspiracy enabled them to make good use of these volunteers.

One of the striking things about the war is that even when the commanders had clear intelligence of their enemies' plans, they sometimes refused to believe it. Stalin's spies delivered to him detailed information about the Germans' coming invasion, but he never took their warnings seriously. Hastings regards the warnings the Americans received about the coming Pearl Harbor attack as definitive, and thinks that George Marshall among others should have been sacked for failing to take appropriate action. Before the 1944 "Market-Garden" airborne landings in the low countries, Ultra revealed that the 2nd SS Panzer Corps had been moved to Arnhem, but Montgomery asserted that this must be a German deception and ordered the operation to go ahead. (In the book and movie A Bridge too Far the discovery of panzers at Arnhem is attributed to aerial reconnaissance, but this seems to be a fabrication designed to protect the Ultra secret.) Japanese decrypts of American merchant ship codes made it clear that the US was about to attack the Marianas and Iwo Jima in 1944, but the high command ignored the intelligence and sent reinforcements to the Philippines. There are many such stories about all the combatants.

At other times commanders provided with good intelligence lost anyway. For example Ultra gave the British detailed plans for the 1941 German airborne assault on Crete days in advance, and the commanders took this seriously. But they were defeated despite this knowledge because the German paratroopers simply outfought British and Greek soldiers. No amount of intelligence could save an overmatched force from a determined assault. At other times perfect intelligence was rendered useless by changes on the battlefield; Ultra decrypts showed that in 1943 the Germans were planning to evacuate southern Italy, all the way past Rome, and this word was passed to the commanders planning the Allied invasion. That didn't happen, but not because the intelligence was wrong; Kesselring, the German commander, simply persuaded Hitler to change his mind. Mark Clark, commander of the US 5th Army, was so put off by this one failure that he never put much stock in Ultra again until near the end of the war.

Another problem with intelligence was that some leaders, especially Hitler and Stalin, ignored intelligence because they already knew what they wanted to do. There are several books about the massive effort the British mounted to deceive the Germans about the planned location of the D-Day invasion. Most of this effort was also wasted, because Hitler had already decided the landing was going to come at the Pas de Calais, and only the most nakedly obvious intelligence would have persuaded him to move troops to Normandy. Some American officers thought the British were wasting far too much effort trying to fool the Germans instead of just beating them. They had a point; Hastings notes that most spies and saboteurs could have served instead as infantry officers, of which the British in particular had a critically short supply.

The most important intelligence of war was signals work, the interception and (if necessary) decoding of enemy radio signals. World War II was the first radio war, in which land, sea and air units were in constant communication with each other via radio. Some communication, for example between pilots on missions or between tank commanders, was in plain speech, with just a few code words for objectives and units. All the belligerents employed teams of men to listen into this chatter and interpret it, and this was very useful. During their 1940 invasion of France the Germans had complete knowledge of French plans and movements from this one source. Rommel attributed much of his success in Africa to his excellent signal corps, which he said always identified the units he was fighting and much about their plans. Communications at a somewhat higher level were guarded by codes, sometimes generated by machines. These were not the most elaborate ciphers and all the combatants (even the Italians) regularly broke the low-level codes that brigade commanders used to message each other. At the highest level were top secret missives, for example from foreign ambassadors to their governments, or from central commands to army or theater commanders. These were protected by high-level codes that at the start of the war were widely considered unbreakable. They were not.

It is amazing, looking back, that governments and armies put so much reliance in their codes that they regularly broadcast their deepest secrets over the airwaves. The trust in these codes was so great that even when told by spies that their signals were being read, leaders refused to believe it. But none of the codes used during the war were entirely unbreakable and some part of the messages in all of them was read by enemies.

Two of the most important intelligence coups of the war were the American cracking of the Japanese naval code and the British breaking of the German Enigma. It was intelligence from code-breaking that made possible the American victory at Midway, which Hastings says is the clearest occasion in the whole war when an intelligence coup led to a military victory.

The breaking of Enigma is of course the most famous intelligence story of the war, how Alan Turing and a band of other geniuses cracked the "unbreakable" German code, built one of the world's first computers, and generally used brain power to somewhat even the odds on the battlefield. This is a hugely complex story, because Enigma was not just one system – the naval version was always harder to crack than the army or especially the Luftwaffe codes – and even at their best the codebreakers at Bletchley Park could only read a portion of intercepts fast enough to do any good. But cracking the Enigma really was a masterstroke; Hastings says that from 1942 on Ultra, as the British called their breakthrough, was their biggest contribution to the allied war effort.

Here's something I did not know. The most important use the British made of Ultra in 1941 was to fight German U-boats in the Atlantic. Using decrypted information about the locations of submarines and their planned movements, they were able to route convoys around ambush points, greatly reducing losses. But for most of 1941 the Germans were also reading the British merchant ship code. This summons up a vision of a weird situation in which the British get information about U-boat locations and re-route their convoys, and then the Germans get information about convoy re-routes and move their submarines accordingly, and so on. Almost incredibly, it took both sides nearly a year of this to realize that their messages were being read. The British eventually changed their code to one the Germans never reliably cracked, and the German navy responded by adding a fifth coding wheel to their Enigmas. This refinement called forth Turing's greatest display of genius, as he and his team broke the new code after nine months of disturbing silence.

As for spies of the regular sort, Hastings asserts that only the Soviets got any real intelligence from secret agents. (Nazi spying, he says, was comically bad.) Communist sympathizers in the US and Britain passed mountains of data to their bosses, so much that less than half of it was ever translated. Philby, Burgess, and the rest of the Cambridge spies kept Stalin well informed about the thinking of the British government and general staff, including much of the intelligence that the British were getting from Ultra. At one crucial point in the war, this actually hurt them; Hastings explains that one reason Stalin did not believe the Germans would attack the Soviet Union in 1941 was that British general staff did not think Hitler would do it, and Stalin trusted the judgment of the British high command more than that of his own agents. On the other hand communist sympathizers kept the Soviets well informed about the British and American nuclear weapons programs, eventually delivering detailed plans of not just the Nagasaki bomb but centrifuges and other devices for enriching uranium.

Intelligence can on occasion be vital, but it has a big problem. The more you have, the more likely it is to contradict itself. It hardly ever happens that all sources tell the same story, and leaders always have to decide what to believe. Historian Hugh Trevor-Roper spent the war as a British intelligence analyst, and Hastings relies heavily on the long-classified report that Trevor-Roper wrote in 1945 summing up the intelligence efforts of all parties in the war. Trevor-Roper noted that in the three weeks before the allies landed on Sicily in 1943 German intelligence received at least 75 reports of planned landing sites for the armada that was obviously being assembled, which he cataloged as Norway 3, Channel Coast 4, Azores 1, Spanish Morocco 1, Southern France 6, Italy 8, Corsica 7, Sardinia 4, Sicily 6, Dalmatia 9, Greece 7, Crete 8, Dodecanese 8, Cyclades 1, Romania 2. Such "information" can be worse than useless. It requires good judgment to make sense of the flow of intelligence, and then decisive leadership to take advantage of the insights gained. The mere accumulation of secret information does no good at all.


szopen said...

British have not break the "unbreakable" Enigma. Polish mathematicians did that, and then they transferred the details of the method and plans of the prototypes of "bombes", mechanical devices, to the British.

John said...

Polish mathematicians first cracked early Enigmas in 1932, and they shared what they knew with the British and French. Which was no doubt a huge help. But the Enigmas of 1942 and later were vastly more complex, with millions of times more possible combinations, requiring large leaps of sophistication and raw computer power to break. So there was still a great deal for Turing and company to do. As for the bombes, perhaps the Poles made some, but everybody was making calculating machines in the 1930s; the Americans broke Japanese codes using IBM punchcard machines. Which is why I wave off all arguments about who made the "first computer." Colossus, made to Turing's design, was a great leap ahead in computing and if the Brtish had not been so determined to keep it secret (for no good reason, really) other earl computers like von Neumann's at Princeton would be less famous.