During WW1 the German ambassador in Washington, von Bernstorff, blamed their numerous transmissions of the same messages as the cause for the successful decoding done by the British. I don't see the reasoning unless the same messages were sent in different codes, but why would the Germans do that? Bernstorff said:

There is absolutely no cipher which they (the British) cannot decipher, provided they have before them a sufficient amount of telegrams. And this result, particularly in the case of the United States, was probably due to the fact that circumstances were such as to force us to make use of an extraordinarily large number of ciphered messages, and we often sent our reports and telegrams in double or triple form, in the hope that in some way they should reach Germany. Consequently, the British must have had an enormous amount of material in the way of cipher dispatches of ours, and in this way it was possible for them to break down our various ciphers.

  • $\begingroup$ Do you happen to know when von Bernstorff said that? $\endgroup$ – Patriot Mar 13 '20 at 15:07
  • $\begingroup$ Alphabetic frequency tables might be relevant here $\endgroup$ – Swashbuckler Mar 13 '20 at 15:08
  • $\begingroup$ Also it would appear that this answer is suggesting that "alphabetic frequency tables" which map characters to their occurance frequency is the answer here, but I think the question asks more along the line, how duplicate ciphertexts helped / whether duplicate ciphertexts were meant in the quote / context. $\endgroup$ – SEJPM Mar 13 '20 at 18:54
  • 1
    $\begingroup$ He said it in 1919 under the German hearings. I see the point with the Alphabetic frequency tables. "Das" or "Die" must have been very frequent in the German language and very frequent codegroups must thus correspond with certain words. The Germans however knew of this and applied severel codegroups to represent frequent words, making it harder for the British to decode but not impossible - if enough data of communications were collected to run these statistical analysis methods. Still duplicates of the same telegrams would't help decoders unless they were in different code, which I suspect. $\endgroup$ – Kristian Francisco Milla Niels Mar 14 '20 at 12:56

Dooley says that simple substitution or transposition were used by both sides from 1914 to mid-1916--for radio, telegraph, and phone messages. Radio became more important in the summer of 1916, and then cryptanalysis became the name of the game for determining order of battle.[1] Crystal set radio often required messages to be sent repeatedly due to signal attenuation, interference, operator error, etc. But this is probably not what he was referring to.

Von Bernstorff was likely referring to the debacle of the Zimmermann telegram, where the same message was sent using two different encryption systems from Berlin to Mexico City. It was a British intelligence coup. Both cipher systems were weak, but the older one was almost completely broken.

The Germans did that because their embassy in Mexico City did not have their government's more modern cipher book. Their people in Washington, D.C., fatefully relayed the Zimmermann telegram using an older cipher system.

Unsurprisingly, we have evidence that Von Bernstorff did employ one cipher which probably worked against the British, and that was the null cipher. David Kahn says how Bernstorff used null ciphers in press cables to get past British collection.[2]

[1] Dooley, John, F. (2016) Codes, Ciphers, and Spies: Tales of Military Intelligence in World War I (p. 49). New York, NY: Springer.

[2] Kahn, David (1967) The Codebreakers (pp. 521 and 1063). New York, NY: The Macmillan Company.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.