This is a really good item. No one ever really appreciated the true scale of the gathering operation itself and how effectiveness was still rather scattershot.
It is way too easy to talk about Alan Turing instead.
And like most Intel protocols, it is none of your business.
The rwal take home is that they captured and read massively which raised their confidence and now you understand why Stalin did not easily believe british Intel. nor should he, any more than you believe the CIA.
How did German commanders react to the fact that the British (and allies) long broke the Enigma codes? I’m especially interested in then living German General's opinion
The scale of British and Allied signals intelligence wasn’t fully revealed until the early 1980s. It’s possible we still don't know the full story. Most of the senior German Generals of WW2 didn’t live long enough to publish their reactions.
Heinz Guderian, who everyone thinks was an armour expert, was in fact a signals officer by trade, he built up German signals intercept units and pushed for the comprehensive use of radio/enigma to inform high command. He would have had an interesting perspective, but he died in 1954. Von Rundstedt died in 1953, Kesselring in 1960 Doenitz in 1980, by which time there were very few senior Nazi-era commanders left.
Commanders like Adolf Heusinger and Hans Speidel who later formed the Bundeswehr and became four star NATO generals, might have become aware of the scale of it as part of their post-war command appointments, but it would still be secret and so if they had reactions they wouldn't have published them.
Scarborough Y-Station, the signals intercept masts seen in shadow. One of thousands of stations across the globe.
It wasn’t so much that the codes were broken - any code can be broken - Enigma was constantly updated to keep ahead of penetration attempts, so they must have feared it was possible, and therefore probable.
It was the scale of the operation they never suspected. Or how deeply we studied their methods and systems, and therefore swiftly produced results.
If German command based their fear of Allied SIGINT on their own attempts - and Germany had good signals intercept systems and was routinely reading (some) Allied stuff - they would assume individual messages would come in as and when they were picked up, get read as and when they were intercepted, as and when code breakers were available to work on them, always in the hope the message contained something of merit.
But a huge mass of radio signals traffic is mundane time sensitive stuff* that is only really useful for that part of the day. Plenty is for training purposes, plenty was purposeful deception. Spending a day to decode a test signal, weather report or garbled nonsense of no value eats into the resources looking for the good stuff.
We used a range of coding systems, some of which were penetrated quickly. But every time we changed key the Germans had to start again.
They could be reading everything one day and nothing the next. It would only take a couple of hours, a day to a week to break the new key, but that meant gaps would appear in their picture, usually just before something nasty happened. They also had to contend with sifting through hundreds of thousands of messages a day.
They didn’t deploy the scale of SIGINT we did so their picture was incomplete. This harms confidence in the end product.
There is a big difference in intel picture between reading half a dozen unrelated signals a week, and reading almost the entirety of your opponents radio signals traffic within the hour of them being sent.
At it’s height the Allied SIGINT operation included 150,000 personnel, directly and indirectly. From Y-station listening posts in the UK and the global signals stations of the Foreign office, Post office, BBC, Royal Navy, RAF and Army feeding raw intercepts into the system.
Then you had people collating the information, analysing patterns, identifying operators and prioritising messages. Only then did we start breaking into the code using the Colossus computers and Bombe devices and transcribing the message. The electro-mechanical computing efforts allowed us such speed and efficiency, we could waste time on the mundane.
Colossus being operated by Wrens. No RAM memory and working at 12,500 bytes per 20ft of tape.
Even then it continued to pass through layers of evaluation and dissemination, covered by security offices whose job it was to keep the secret safe. By the time it ended up on a General’s desk it was often made to look like evidence of photo recce or captured PoWs.
Such was the scale of our operation we were able to sift the information in a way German decrypt could never imagine. Some of the most important stuff we regularly decoded was radio test messages and mundane information like weather patterns and training exercises.
This stuff wouldn't tell you grand strategy, but it did provide valuable traffic analysis patterns, user methods and operator details. We were recording so much we could identify individual German radio operators by their morse touch. We all give our most important jobs to our best people, so when the guys known to be good popped up on the key, we paid attention.
Even without decrypting, gathering evidence on the number of enigma stations in any one region, their operators skill and tempo of signal gave you huge insight into enemy movements and plans. These could then be targeted more effectively.
Such was the quantity of raw data collected we could build patterns of error, this made keeping ahead of new code developments easier. The only way we could do all this - waste time on unimportant messages - was because we built the resources to waste. It was a huge operation and it provided a huge picture of rich colour and deep layers. Therefore our confidence in it remained high.
I don't know of any senior German commander of WW2 commenting on the scale of Allied SIGINT in the war, but their casual attitude to sending vital information by radio and enigma and their straightforward attempts to maintain secrecy - focused on equipment not method - suggests they would be shocked at how big we went.
*Information may be mundane, but never worthless. All intelligence is vital - if you get it all - holes in intelligence pictures cause doubt - doubt reduces confidence in the product.
SIGINT like all intelligence is very user sensitive, it’s never offered as fact or absolute, and despite your best efforts to draw as rich and layered a picture as possible, with your most cleverest interpretations and suggestions, two analysts will commonly draw different conclusions from the same material, and commanders will always try to draw their own.
We tend to think of SIGINT as perfect intel - General A - HAS - told Colonel B to do C, that must be what will happen right? Ha, not in the army mate. All war is deception, that brings your first doubt to the authenticity of the message. Ambitious Colonels don't always do exactly as they’re told, martinets confuse themselves with additional angers, weak or weary Colonels can simply fail to galvanise their men to do what's asked.
Even if Colonel B is on top form and fully switched on frictions like reactions to enemy fire, complex timings, unfortunate emotional episodes, a lucky sniper or strafer and plain dumb luck get in the way all the time.
So the real intel value is a historical biography of your key players - down to battalion level. Reading all the mundane will give you a picture of the martinet, the weary or the first timer.
Germany especially relied on its junior officers to interpret orders and plan around them, so what was said over the radio is only an intention, not plan of action. Knowing the enemies gross intentions are great for four star level planning, but by the time its got to Corps level action it's of less use. What you want is a sense of an opposing units means and will to fight, and this can change by the hour.
Radio traffic such as: pre-planned fire missions; order of march; edges of air defended areas, orbats, locations and sentry weapon states can be vital if they get to the battalion commander facing that unit in time. A delay of a few hours for decoding and disseminating and its value has passed.
A lot of what you hear over the air is this kind of time sensitive detail. Great stuff to build a big picture from, but you don't know what is what when you start trying to penetrate the code. Sifting the chaff from the airwaves relies on you collecting and reading it all, or you don't know what’s chaff.
If you fail to pick up the message that switched your enemy from plan A to Plan B, or cancelled the operation entirely, your intelligence summary is now working against you. Orders are routinely re-interpreted and amended by face-to-face briefings and changed situations. If you miss these you're reacting to a shadow and can be in the wrong stance when your enemy gets active. These frictions are well known and understood on both sides of the line, so plenty of disinformation is deployed to help confusion along.
Decrypting is only the start of your problems.
No comments:
Post a Comment