Once the Allies started breaking the codes sent by the Germans, did they find that the communications made perfect sense (to a German speaker)? Or were the words themselves further coded, the way the Navajo words and phrases the US Codetalkers were real Navajo words, but they didn’t make any sense to someone who didn’t know how they were being used?
I remember reading about a Navajo who was captured by the Japanese, and they tortured him until he agreed to translate the Navajo code. And when he told them that it was real Navajo words they were using, but they didn’t make any sense and he couldn’t understand, they tortured him even more brutally.
KRKR ALLE XX FOLGENDES IST SOFORT BEKANNTZUGEBEN XX ICH HABE FOLGELNBE BEFEHL ERHALTEN XX J ANSTERLE DES BISHERIGXN REICHSMARSCHALLS J GOERING J SETZT DER FUEHRER SIE Y HVRR GRZSSADMIRAL Y ALS SEINEN NACHFOLGER EIN X SCHRIFTLSCHE VOLLMACHT UNTERWEGS X ABSOFORT SOLLEN SIE SAEMTLICHE MASSNAHMEN VERFUEGEN Y DIE SICH AUS DER GEGENWAERTIGEN LAGE ERGEBEN X GEZ X REICHSLEITEI KK TULPE KK J BORMANN J OB.D.MMM DURNH FKST.KOM.ADM.UUU BOOIE.KP
KRIEGSNOTMELDUNG [An] Alle: Folgendes ist sofort bekanntzugeben: Ich habe folgende Befehl erhalten: 'Anstelle des bisherigen Reichsmarschalls ‘Göring’ setzt der Führer Sie, Herr Großadmiral, als seinen Nachfolger ein. Schriftlische Vollmacht unterwegs. Ab sofort sollen Sie sämtliche Maßnahmen verfügen, die sich aus die gegenwärtigen Lage ergeben. Gez. Reichsleiter (Tulpe) ‘Bormann’ [Von] Oberbefehlshaber der Marine, durch Funkstelle der Kommandierender Admiral der Unterseeboote.
As a general point relying on jargon, slang, etc as a level of security is not a good move. You should assume the person trying to decipher your message is just as well versed in those as you (that was certainly the case at Bletchly park)
I don’t know much about the coding, but I remember from a documentary about the code breaking of the Enigma that recurring words like “Wetter” (“weather”) were crucial for deciphering the code.
The military Enigma had 29 keys. Adding another 10 keys would have substantially increased the size, complexity and cost of the machine. The designers probably figured it wasn’t worth it since numbers are redundant. It’s the same reason there aren’t punctuation keys.
If I’m understanding the OP correctly, he’s asking about the use of code words not slang.
So the Allies might decipher a German enigma message and have something like “ATTACK TARGET ALPHA AT SIX AM ON MONDAY”. So they’d know an attack was coming and when - but they wouldn’t know where without knowing what Alpha was code for.
An example of this was with the breaking of Japanese codes. The Americans had intercepted a message about an impending invasion of “AF”. The problem was they didn’t know what AF was code for. They suspected it might be Midway. So they had Midway broadcast a signal that one of their desalination plants had broken and they were short of fresh water. Then they intercepted another Japanese message, which when decoded said that AF was short of fresh water. So now they knew where the invasion was coming.
Not really a blatant flaw of the machine itself. The weakness was more due to the way it was used. It would have been easy to use simple substitutions instead of spelling out every number and digits. You could, for example, just use the first one or two letters of the name of each digit, preventing attackers from having easy guesses for the words in your message.
Additional compexity would be practically zero. Just more keys, lamps, spring loaded contacts. Size might be an about 40% greater footprint. I doubt cost would have played a role. Machines with number keys had been commercially available for some time.
Given that, it is unclear to me why there were no number keys. If you are trying to save space, why have 29 keys and not 26? You could save even more by omitting rarely used letters and substituting them with some letter combination.
It does seem like a weird, artificial limitation where 29 chars is small enough to successfully implement (design, manufacture, train & distribute) but nine or ten more is too much, especially given how much numeric data would have needed encoding. I feel like I’d rather have a 5 than a G during my turn at the radio desk.
Coding messages on top of encrypting them would have added points of weakness.
Since a relatively large number of people would be expected to be able to receive communications, each would need a corresponding codebook. For a limited number of places/concepts, memorization might work, if also only a very limited number of people would need to know the codes. But the more units that need to know the code and the more complex the messages and concepts to send/receive would require increasing large dictionaries which themselves would be subject to capture. No free lunches in data security.
But for individual intelligence operatives working alone and only communicating with a limited (perhaps only 1) handler, thus needing no explicit codebook to carry? Sure, it can add another layer of obfuscation and security to both encode and encrypt messages at the cost of additional time to decode and decrypt those messages and additional training time.
This, all by itself, is a security flaw. It’s the typical convenience vs security trade-off. Secrets should be limited to as small a number of people as practical. Better for each channel to have its own codebook.
The flaw isn’t so much in the machine, but how it’s operated. It’s always the humans’ fault.
(On preview, I see you’ve added much the same point.)
