Friday, March 4, 2016

Did the German codebreakers solve the Japanese Purple cipher machine? – Information from mr Otto Leiberich

In the late 1930’s the Japanese Foreign Ministry distributed the Purple cipher machine to its most important embassies and it was used to encipher high level messages to and from Tokyo. Unfortunately for the Japanese the introduction of this new cipher machine wasn’t able to secure their diplomatic communications.

The codebreakers of the US Signal Intelligence Service were able to solve this device in 1940 and according to Russian historians the codebreakers of the Soviet Union, led by Sergei Tolstoy, also solved it.

The British codebreakers were not able to solve this system on their own but they received information and a copy of the device from their American allies in 1941.

Speculation on the German effort versus the Purple cipher machine

US reports based on the interrogation of German cryptanalysts claim that the Germans made an effort to solve the Purple cipher machine but were not succesful. However the German historian Jürgen Rohwer mentioned in his book ‘Stalin's ocean-going fleet’ that in 1996 he received a letter from a mr Cort Rave, who claimed that the Germans were in fact able to solve the Purple machine and decode the Japanese messages.

According to US reports Kurt Rave was an OKW/Chi employee detached to the Foreign Office Cryptanalytic Section (Pers Z S), so his testimony has to be taken seriously.



I have looked into this case and presented all the available information in German success with Purple?.

Unfortunately I haven’t been able to get a copy of the letter written by Rave. When I emailed mr Rohwer in 2011 he said he would look into it but he never got back to me.

Another person who made similar statements regarding the German solution of the Purple machine was Otto Leiberich, chief cryptologist of the German cipher department in the period 1972-1990.


Zwei Erfolge verdienen eine besondere Würdigung: die Entzifferung des Purple-Verfahrens der Japaner und die Entzifferung der amerikanischen Chiffriermaschine M 209.

Während des Krieges hatten die Japaner eine Chiffriermaschine entwickelt und zum Einsatz gebracht, die der amerikanischen Aufklärung größte Probleme bereitete. Da gelang es einer amerikanischen Gruppe um den Kryptologen William Friedman, diese Maschine, die als purple machine bezeichnet wurde, zu rekonstruieren und zu entziffern.

Dies gilt seither in Amerika als der größte Erfolg in der Kryptologie-Geschichte. Angeregt durch eine kürzlich ausgestrahlte Fernsehsendung fragte ich bei einem ehemaligen Kollegen nach, der während des Krieges auf diesem Gebiet tätig gewesen war, und erhielt bestätigt, woran ich bis dahin nur eine ungefähre Erinnerung hatte: Auch die Deutschen hatten die Sendungen der verbündeten Japaner bearbeitet, insbesondere die Meldungen, die der japanische Botschafter Oshima aus Berlin nach Tokio sandte. Einer Gruppe von Kryptologen und Technikern der Chiffrierabteilung des Oberkommandos der Wehrmacht (OKW) unter der Leitung des Mathematikers Erich Hüttenhain war die Entzifferung ebenso gelungen wie den Amerikanern. Hin und wieder war ein Bericht schon entziffert und weitergeleitet, wenn Tokio wegen Übermittlungsfehlern um nochmalige Übersendung bitten mußte. Wenn also die Entzifferung der purple machine der größte Entzifferungserfolg während des Zweiten Weltkrieges gewesen wäre (er war es nicht!), so hätten ihn Hüttenhain und sein Team ebenfalls errungen. Leider existieren in Deutschland hierzu keine Unterlagen mehr
.’

Since I was interested in this case I managed to contact mr Leiberich in 2013 and we had a short telephone conversation. Since his article was written a long time ago he didn’t remember all the details but he was able to give me the following information:

1). He did not have any documentary evidence and his statements were based on discussions with his colleagues.

2). Rave said it was a cipher machine and he remembered specific messages.

3). Regarding the years of solution he said probably 1943-44.

4). He had asked Rave why he hadn’t mentioned this success to the American interrogators and he responded that he feared it would be dangerous for him to say too much.

The last statement can be confirmed, in part, by the ‘Final report of TICOM team 3’, p12 which shows that some of the captured Pers Z higher-ups wanted to give away as little as possible and tried to influence their coworkers to do the same.


2 comments:

  1. Solution to that "mystery" is provided in the quotation from Mr. Leiberich. Note the fragment:

    Hin und wieder war ein Bericht schon entziffert und weitergeleitet, wenn Tokio wegen Übermittlungsfehlern um nochmalige Übersendung bitten mußte.

    Approximate translation: We were occassionaly able to decrypt and pass on the message, whenever Tokio had to request the retransmission due to the transmission errors.

    It is clear that the Germans were only able to break Purple messages whenever the Japanese provided them with the depth: two or more nearly identical messages enciphered with the same key. It is very far from the regular decryption operation, including reconstruction of the ciphering machine.

    Note that according to the revelations underlying your text the breaks were to take place in 1943-1944. In the same period Oshima's messages decrypted by the Allies were providing them (among the others) with crucial intelligence regarding German preparations in France. If the Germans could read Oshima's messages on a regular basis, they would have to assume that they are not the only ones to do so and they would probably take a far more cautious approach in diclosing to the Japanes their military secrets.

    ReplyDelete
    Replies
    1. Yes, I’ve made the same argument back in 2011 in ‘German success with Purple ?’ when I said:

      ‘Taken together all this information points to German success with the Purple machine.The extent of this success however is a mystery.
      Did they solve only messages that were sent several times to ensure reception in Japan or could they read all the traffic? Did they only decode messages in the Berlin-Tokyo link or of other embassies too?’

      You also said: ‘Note that according to the revelations underlying your text the breaks were to take place in 1943-1944. In the same period Oshima's messages decrypted by the Allies were providing them (among the others) with crucial intelligence regarding German preparations in France’

      Crucial ? No. Occasionally useful? Yes. That's my opinion.

      Regarding Japanese ciphers the Germans already knew they were weak and they chose to exploit that instead of warning them. Check:

      http://chris-intel-corner.blogspot.gr/2014/05/the-japanese-j-19-fuji-code_1.html

      Delete