Japanese Reaction to Yardley's The American Black Chamber

On 1 June 1931, Herbert O. Yardley's The American Black Chamber stunned the world by exposing American codebreaking activities. Among the diplomatic messages uncovered by his Cipher Bureau, "the most important and far-reaching telegram" as Yardley put it was an instruction to the Japanese plenipotentiary sent to the Washington Naval Conference in 1921 (see another article).

While there have been statements that the book caused Japan and other nations to revise their codes at once, Kahn (2004) undermined such a myth by showing the number of solutions in the British archives of secret messages of Japan and Germany did not decrease after the publication of the book (Kahn p.131-136) (A skeptical view on such a traditional "offhand statements" is also presented by the editor of ASA, Wayne G. Barker (ed.), The History of Codes and Ciphers in the United States during the Period between the World Wars, Part I. 1919-1929 p.137). Materials in Japanese archives appear to support Kahn's statements that Japan had been updating its diplomatic cryptosystems every couple of years and did not immediately change to new ones.

Table of Contents

First Reception

Sensation in Japan

Code Security Measures Before and After the Publication of The American Black Chamber

Proposal of Introduction of Cipher Machines

Preparation for Facing the Parliament

Did Yardley Sell Secrets of Codebreaking to Japanese Embassy?

Yardley Continues Disclosure

Japanese Reaction to Code Security Incidents


First Reception

The first notice of the publication of The American Black Chamber came in on 6 June in a telegram from the Japanese ambassador in Washington, Debuchi Katsuji, to the Foreign Minister, Shidehara Kijuro, who had been ambassador to the US during the Washington Naval Conference back in 1921-1922. It reported that the American Black Chamber had been reading foreign diplomatic secret telegrams and that the book stressed that any code could be broken with scientific methods. With regard to Japan, the book not only revealed that telegrams to and from the Japanese plenipotentiary during the Washington Naval Conference had been read but also published many telegrams translated into English (Yardley file (1) p.38).

The ambassador sent two copies of the book (Yardley file (1) p.51) and related newspaper articles (Yardley file (1) p.54).

Sakuma, Chief of the Telegraph Section of the Ministry of Foreign Affairs, wrote an extensive memorandum on the matter as early as 10 June (Yardley file (2) p.40-48). It shows that the codebreaking per se was no surprise and that the Foreign Ministry had already paying efforts to improve its cryptosystems (Sakuma understood that any code might be broken some time but still believed in cipher machines). Rather, Sakuma discusses more at length the disclosure of telegrams and inquires into ways to charge Yardley somehow (Section (2)). It also shows a bureaucratic reality that hampered the efforts to improve cryptographic security (Section (3)).

Regarding the Book by Yardley, Head of Codebreaking in US Military Intelligence Section
10 June 1931
Chief of Telegraph Section

(1) June 1931. According to a telegram report from Ambassador Debuchi in the US, Herbert O. Yardley, who was head of codebreaking in a US military intelligence section, recently published a book titled The American Black Chamber, which disclosed that he decoded more than forty thousand secret telegrams of Japan, Britain, Germany, etc. from 1917 to 1929 to provide the most sensitive diplomatic information to the Department of State etc.; that he broke secret communications in "invisible ink" used by German spies during the Great War and by secret emissaries of the Soviet Union after the war to divulge their secret schemes; and, stressing that any code can be broken with systematic studies by scientific methods, states that after much toil and study, he could broke every one of Japanese telegrams in code and read telegrams between the Japanese plenipotentiary and the government as in the time of the Washington Conference; and prints telegrams at the time translated into English. Since this would be soon reported in newspapers and magazines in Japan and may be taken up in the parliament etc., it is necessary to make preparations now.

(2) Whether There Is a Way to Charge Yardley
(A) The above-mentioned Yardley is considered to be the same person as the Yardley who offered to sell to the imperial embassy in the US secret deciphering of Japanese government codes, to which we ended up in giving 7000 US dollars to receive many copies of deciphered coded telegrams and records of codebreaking methods etc. If so, it is a breach of trust that he described how he deciphered Japanese encoded telegrams in the present book. When paying the 7000 dollars, we emphasized that these secrets should not be leaked and our instructions to the ambassador to the US asked for securing a guarantee to that effect (our instructions stipulated "upon buying these, sufficient guarantee should be secured that their contents would not leak in future by making him surrender these telegrams or destroy them etc.") (see Telegram No. 105 to the ambassador to the US from the [Foreign] Minister sent in June 1930). Although the reports from the ambassador did not include any mentioning in particular of securing such a guarantee other than receiving copies of decipherments of encoded Japanese telegrams and other documents, it may be presumed there was some understanding with Y. explicitly or implicitly. Therefore, we might cause the embassy in the US to accuse his breach of trust by saying that leaking the secret as above after our giving him 7000 dollars is outrageous to the extreme. However, it should have been imaginable from the first that a person like Y., who sought to sell secrets of the government of his own country to a foreign embassy, might commit such a breach of trust. Further, even if Y. could be caused by our accusation to cancel the portions in the book concerning us, it would little help eliminate the facts once divulged and it cannot be denied it would rather arouse further suspicions. Further still, if Y. takes offense at our accusation and discloses the fact that he received 7000 dollars from the imperial embassy in the US, we might end up in stirring up a hornet's nest. Thus, it is considered that regarding Y.'s book, it is proper to proceed by ignoring it without taking any measures.

(B) Next, Article 2 of the International Telegraph Convention stipulates "The High Contracting Parties undertake to adopt all necessary measures to ensure the secrecy and prompt despatch of messages." and Article 73 of the annexed International Telegraph Service Regulation (Article 70 before 1925; "photographs" was added in 1925) stipulates that originals or copies of telegrams may be shown only to the sender or the addressee, after verification of their identity, or to the authorised representative of one of them and that the sender and and the addressee of a telegram or their authorised representatives have the right to obtain certified copies or photographs. Further, Paragraph (1) of Article 91 of the Regulation stipulates that private telegraph enterprises working within the frontiers of one or more Contracting States , and participating in the international service, are regarded, from the standpoint of this service, as forming an integral part of the telegraph system of these States. Therefore, the government of a contracting state would have to act quite covertly to obtain from telegraph stations copies of telegrams to and from foreign ambassadors, ministers, etc. to study codes of foreign governments. Although the United States is not a contracting party to the above-mentioned Convention, since telegraph companies of the United States conduct telegraphic communications with the contracting states, it should be said, at least with respect to amity, they are obliged not to collect copies of encrypted telegrams of a foreign government from telegraph stations in the United States for the purpose of studying codes of the foreign government. In practice, too, Ambassador Debuchi reports Y. tells to the effect that it is rather difficult at the US cipher institute to obtain from telegraph companies copies of encrypted telegrams and copies are obtained by bribing a clerk etc. From this, it appears that telegraph companies in the US would not freely deliver copies of telegrams handled by them to the US government authorities. Therefore, it would not be impossible to point out to the US government that Y. disclosed by his book that the US cipher institute is obtaining copies of telegrams of the Japanese government and demand some appropriate measure be taken. Further, the US authorities may take some disciplinary action on Y., who leaked secrets of the government, such as confiscation of the books, a ban on its sales, or punishment, etc. However, such a representation from us would not be appropriate, either, because it might end up in stirring up a hornet's nest should our active representation results in Y. confessing the fact that he got 7000 dollars from us.

(C) Next, while it might be possible to confer with other nations involved such as Britain, Germany, etc. about measures to be taken, the nature of the issue is quite delicate and it would not be an appropriate option.

(D) Since any of the above measures would not be appropriate, after all, it would be sufficient to consider domestic measures, that is, how to respond when the issue is taken up by the parliament or domestic newspapers.

(3) State of Our Cryptology and Degree of Solution at the US Cipher Institute
While secrecy of code is linked with the complexity of the code system and the frequency of updating the code, it cannot be denied that there have been vulnerable points in the way codes have been compiled and updated at the Telegraph Section of the Foreign Ministry. The attached table [not found in the archives] lists the kind of codes from 1917 up to now, their period of use, the number of entries, etc. Since the Telegraph Section had no Code Research Group and the telegraph staff compiled codes in their spare time, it is natural that the degree of complexity as well as the frequency of updating have been insufficient. Compilation and printing/binding of one code take several months and it would take six months to one year to begin its use after distributing copies to all the overseas establishments. Thus, as long as codes are compiled only in spare time, not only is it difficult to sufficiently consider the code system to make it complex, but also frequent updating of code is impossible. Now, however, the Telegraph Section has a Code Research Group with Captain Inoue of the [Reserve] Navy as its head to study codes of foreign countries. (It may be presumed that their achievements are comparable to code study institutions of the United States and other countries. Of the foreign codes attacked so far, the only code that could not be solved is one of the Soviet Union, which we set on studying not long ago.) Since codes are now compiled and distributed as a routine business by using the knowledge learned from this, our code is being improved steadily. Despite such compilation of codes as regular business, however, the number of codes that can be completed and distributed during a year is relatively small. Although the expenditure required to reinforce the Code Research Group was included in the budget of the fiscal year 1929 and approved by the parliament, austerity measures resulted in its omission in the working budget. Thereafter, the above expenditure has been requested year after year but has not been appropriated. It is considered that reinforcement of the Code Research Group need to be achieved as soon as possible.
As long as a code system is used, any code would be broken with scientific methods if substantial materials and time are allowed, as actually experimented at the Code Research Group of the Telegraph Section. With the present personnel at the overseas establishment of the [Foreign] Ministry, there is naturally some limit in practice in increasing the complexity of the code, while the codes currently in use have already nearly reached the limit. On the other hand, even if the personnel of the Code Research Group of the Telegraph Section is to be reinforced from the present staff, there is also some limit on the frequency of updating codes. Thus, while it should be possible to prevent a code from being broken during its use, it is hardly possible to ensure that a code is not broken several years after it comes into disuse. This is the case for any state and cannot be avoided.
In order to entirely prevent codebreaking with scientific methods, there is no option but to use a perfect cipher machine having almost infinite variations. While we did use cipher machines made by the Navy during the London Naval Conference [in 1930], not only were their conditions very unsatisfactory but also they were very expensive. The Code Research Group of the Telegraph Section has purchased a latest German cipher machine, the manual-type Kryha, and is currently studying its utility. It has a similar structure to the naval machine and some improvement might put it in a condition for practical use. However, using the machine in all the overseas establishments would require a huge expenditure at a time, which should be said to be unlikely to be attained under the current austerity measures. However, since it is clear that the code system will be at a dead end in future and the Code Research Group of the Telegraph Section has started studying use of machines for encrypting as noted above, it would be better to purchase an electric automatic-type Kryha machine for the above-mentioned study.
As detailed above, the best efforts are now being made in the [Foreign] Ministry for protection of code security. While Yardley boats as if he has broken every one of our old codes (far more rudimentary as compared to the current codes), materials purchased from him (there are many of them, which are now under study) appear to indicate that codes of relatively infrequent use have not been broken and that those completely broken are ones that were used frequently and for a relatively long time as well as that have a significantly simpler organization as compared to codes currently in use (see the attached table [not found]). There is still some doubt as to whether the codebreaking was achieved during the lifetime of the code and whether it was done only through scientific methods.

(4) Domestic Measures That Should be Taken with Respect to Yardley's Book
From the foregoing, if issues arise in the parliament etc. and the [Foreign] Ministry is required to express some view, it is considered appropriate to respond in the following tenor:
"While there is no means to ascertain to what degree Yardley's book tells the truth, any code may be broken with scientific methods if substantial time and materials are allowed. Further, it may be presumed that every nation is making considerable efforts in codebreaking against other nations. Thus, there is hardly a way to prevent that our code is broken some day. Notwithstanding, the Foreign Ministry is taking the best measures for the protection of security of our codes."
Further, if Yardley confesses to the fact that he sold secrets to us at 7000 dollars and we are put in a situation that makes it impossible to deny it any way, we would have no option but to respond in the following tenor:
"Yardley was a former head of codebreaking in a US military intelligence section. Thus, as long as he offered to sell secrets of breaking of our codes to the imperial embassy in the US, in view of his position, it would be hard to dismiss it and we gave him the above amount in order to ascertain whether our codes were indeed broken."

Sensation in Japan

Tokyo Nichi Nichi newspaper (20 July - )

The reaction of the Foreign Ministry was slow. After the above memorandum was written, little appears to have been done until on 20 July 1931, the Tokyo Nichi Nichi newspaper carried a long article about The American Black Chamber, headed "What an Idiot." It retranslated into Japanese some of the secret telegrams disclosed by Yardley and supported them by referring to corresponding contemporary articles on the paper.

Various reactions to the revelation were carried on 22 July (also Kahn (1967) p.363). An informed source of the House of Peers accused Ambassador Shidehara, now Foreign Minister, by pointing out only Japanese telegrams, not those to the embassy in Britain but only those addressed to the embassy in the US, were read during the Washington conference and the Ambassador was unaware of it. He also questioned the attitude of the Foreign Ministry as if it were simply a matter of disgrace to the US. Baron Ikeda of the House of Peers deplored that plenipotentiaries could not request instructions but only send reports and asked for measures for the Geneva Conference scheduled next year. An anonymous naval officer directed his criticism to the breech of privacy of communication in the US. An authority of the Minstry of War said he knew of the incident by the newpaper coverage; pointed out naïveté of Foreign Ministry codes; expressed the Army would take every necessary measure for the League of Nations General Disarmament Conference (Wikipedia) scheduled next year, which would discuss reduction and limitation of the army; and sarcastically concluded he might give advice to the Foreign Ministry from good will.

Comments from an authority of the Foreign Ministry were as follows.

Regarding Japanese telegrams in code at the time of the Washington Conference having been secretively read by the US, while our failure to change codes frequently was indeed a fault, that the United States, a major power hosting the Washington Conference, was secretively reading secret telegrams in code, which participants of international negotiations were entitled to, is a disgrace and a dishonor to such a major powr as the US, which has now been revealed to the world by Mr Yardley's publication. The author is a person with such a nerve as to have visited the Imperial Embassy in the US at the time of the Washington Conference and said, "Japanese codes are all broken. I may sell them." thus offering a deal to the very target of his codebreaking. He may have published the book for the publicity of his own work, while he was disgruntled with the US government because he had been discharged.

The newspaper on the same day carried an advertisement of imported books from a bookseller Maruzen, which highlighted The American Black Chamber (7 yen), which was to arrive in early August.

Next day, on 23 July, a correspondent reported words from Yardley, whom he met in Yardley's hometime in Indiana where he then lived. Yardley pointed out that every state other than the US had a black chamber and there was no option for the US but to establish one or conclude a treaty to make other states abolish it. He said he offered to the US government a new measure for undecipherable code by means of a "code machine", for which he had not received any response. He declared that there was no conventional code/cipher that could not be broken.

The Tokyo Nichi Nichi daily carried partial translation of The American Black Chamber from 24 July. (The whole book would be published by its parent company, Osaka Mainichi Shimbunsha, in August.)

General Reception

Apparently, the Tokyo Nichi Nichi's scoop did not arouse a nationwide sensation. For example, the Asahi newspaper does not appear to have carried a follow-up article. Still, an American in Japan felt a "wave of anti-American feeling [Yardley's book] stirred up among all classes of the Japanese people" (ASA op.cit., p.136).

English-language newspapers referred to "mild sensation" or "serious sensation." At least, there was a persistent and considerable interest in the issue (Kahn (1967), p.363-364).

Nippon newspaper (15 August - )

Moreover, it did add fuel to partisan criticism of the Foreign Minister Shidehara. The Nippon newspaper carried a series of articles from 15 August titled "Revealed History of Failures of Shidehara Diplomacy", which accused diplomacy of the Foreign Minister Shidehara as complaisant. The Yardley files include clips of these articles. .

The second article on 16 August mentioned Yardley's book. The article did not make much of codebreaking itself. Instead, it criticized breaking into the Japanese General Cosulate in New York to get access to a codebook and suspected other dubious means for breaking Japanese code. The third article on 17 August insinuated a possibility that Shidehara's treatment of the staff at the embassy might have produced cooperators to the Black Chamber.

The article reported comments of Shiratori (Wikipedia (in Japanese)), Head of Intelligence Department of the Foreign Ministry: "Secretive reading of messages in code is a problem during pendency at the telegraph office, not after their reception at the embassy. Moreover, there was once an offer of a deal from the author and others but we turned them down." The reporter added he heard there had been an offer as early as at the time of the Washington Conference of a deal about reading of secret telegrams but the Foreign Ministry also took no notice of it, which the reporter criticized by saying at least the veracity of the reading should have been checked.

Foreign Ministry Aroused from Inactivity

The newspaper coverage appears to have roused those concerned from inactivity. On 28 July, Sakuma drafted replies to possible questions that may be brought up in the parliament about the issue and on 31 July further inquiries were made to the ambassador in the US. On 29 July, Ito Risaburo, a naval officer, renewed his proposal for the Foreign Ministry to use cipher machines. On 15 August, instructions about security of code were drafted and sent to overseas establishments on 17 September. These will be described in the following.

Code Security Measures Before and After the Publication of The American Black Chamber

9 September 1930

In a document titled "Regarding Protection of Security of Telegraphic Code" dated 9 September 1930, the Foreign Minister alerted overseas establishments about storage of codes as well as telegrams in code. It even asked for caution when quoting a secret telegram so as not to give any clue to codebreakers. (JACAR B12080889500 p.9)

In this fall, the Foreign Ministry assigned personnel to check outgoing telegrams, one of whose duty was to make sure that the code was fully used. There was tendency that only relatively simple parts of the code were used. Failure to use the full range of the code results in not only high telegraph costs due to increase in the number of words but also vulnerability of even a relatively complex code. (JACAR B12080889500 p.15-16)

1 April 1931

Sakuma, Chief of the Telegraph Section, drafted a document again titled "Regarding Protection of Security of Telegraphic Code" dated 1 April 1931 (Yardley file (3), p.32-35).

Based on reports from the overseas establishments solicited in the above document of 9 September 1930, it focused on handling and storage of not only codes as well as drafts/translations of telegrams in code and required use of a steel cabinet with a combination lock etc.

It is silent about cryptologic security as discussed in the memorandum of 10 June 1931 after the publication of Yardley's The American Black Chamber. But then, it was only natural because preparation and distribution of codes are the responsibilities of the Foreign Ministry, not the embassies or consulates.

Code Research Group

Sakuma relied on expertise of Captain Inoue of the Reserve Navy, who worked for the Telegraph Section since 1929 as a temporary member and would join the regular staff in 1934 (Shiraishi p.103, n.(9)). Inoue headed a Code Research Group in the Telegraph Section. The group had been established in 1921 in the Telegraph Section of the Foreign Ministry by the ministries of War, Navy, Foreign Affairs, and Communication and had broken English and American codes (Hori Eizo, Daihonei Sanbo no Joho Senki; Kahn (1967) p.495). The group took over compilation of codes from the telegraph staff in the Telegraph Section. Sakuma describes as follows in a memorandum dated 9 October 1933 (JACAR Ref. B12080889500).

Ever since I took office as chief of the Telegraph Section in July 1930 (at the time, there was the Yardley issue), I recognized grave defects in the code system of the Foreign Ministry and first of all set about improvement of telegraphic codes. Compilation of codes, which had been done by the telegraph staff in their spare time, has been transferred to the Foreign Code Research Group headed by Captain Inoue of the [Reserve] Navy. The group was to be invariably engaged in studying, compiling, and printing of new codes as part of its routine business. (Since the telegraph staff are fully occupied with almost no spare time, it appears, previously, they often set about updating a code only when pressing need arose for some reason. Further, while the telegraph staff are to use codes, the Study Group is researching into codes/ciphers in a scientific manner. So it is unquestionable that the group is more suited for compilation of codes than the telegraph staff.)
However, even if conducted as a routine business, the number of codes that can be compiled in a year is naturally limited for various reasons such as considerable complexity of such codes as currently used requiring considerable time for investigation, the varying load of the printing office, and considerable time needed for proofreading. (Previously, compilation of one code took three to six months.) Moreover, distribution of the completed code to all the overseas establishments requires more than a full year. (Use of different codes in different regions is highly necessary. However, we often need relaying of telegrams and transcoding every time the same telegram into various codes requires considerable increase of the staff. Thus, naturally, it has not been achieved up to now.)
For these reasons, although I have been paying much attention to updating of codes ever since I took office as chief of the Telegraph Section, the results are relatively limited and the frequency of updating the codes of the Foreign Ministry is, naturally, not making much progress.
According to the observations of Captain Inoue, as noted before, with a code with up to around one thousand entries, whatever variation in structure may be made would be naturally limited and even if code is very frequently updated, it would not be immune to codebreaking in a scientific method. Eventually, he says, there would be a case in which one long telegram allows breaking.
After I took office, recognizing the need of reinforcing the Code Research Group, I applied for a budget that would further extend the previous plan, which passed the parliament as the budget for the previous fiscal year [1932]. However, since the ministry reform proposal to implement it was merged with a proposal to set up an Investigation Division [to handle the situation after the Mukden Incident in the Foreign Ministry], it was trapped in the Privy Council and although the budget has been appropriated, it has not been realized. When the budget is executed, the Study Group will see a considerable reinforcement and thus will also serve for compilation of codes more than it does today.

17 September 1931 -- After the Publication of The American Black Chamber

After the publication of Yardley's The American Black Chamber in June 1931 and its newspaper coverage in late July, needs for additional measures for code security were felt. On 15 August 1931, instructions about proper use of telegraph codes were drafted (Yardley file (3) p.75).

In particular, it warns against using a code to encode a report known to another country and stipulates use of a proper code among various codes provided, depending on the nature of the particular message to be encoded.

The draft was revised and sent out to overseas establishments on 17 September 1931 (its text is at JACAR B12080889500 p.10).

As reported in the morning issue on the 20 July 1931 of the Tokyo Nichi Nichi newspaper etc., an American named Yardley caused a sensation by publishing a book titled The American Black Chamber, which describes how he decoded our telegrams in code etc., as you will be aware. It may be presumed that any major nation has a unit for studying foreign telegrams in code and is successful in codebreaking to some extent. Thus, Yardley's case happens to have become such an issue only because he disclosed affairs inside the US cipher institute from his discontent towards the US authority etc. Today, techniques of codebreaking have advanced such that it cannot be admitted but that any code could be broken by a talented codebreaker with scientific methods if substantial time and materials are allowed. Thus, in order to make it difficult break a code, the organization of the code must be more complex and the code must be frequently updated. However, in view of the current number of the telegraph staff at the Foreign Ministry and overseas establishments as well as the amount of telegrams, there is some limit in increasing the complexity of the code and because compilation and distribution of code take considerable time, the frequency of updating is also subject to some limit. As a result, our code is in more danger of being broken. Since it is absolutely necessary to prevent encoded telegrams seriously involved in our interests from being decoded at the time of transmission, the Foreign Ministry is making the best efforts such that our code is not broken during its use.
However, when one looks at how codes are used at overseas establishments so far, when a telegram to be encoded is drafted, the director or other person in charge specifies it as "secret" [gokuhi], "treat as secret" [gokuhi-atsukai], "encrypt" [an], "abbreviate" [ryaku], etc. depending on the degree of necessity of keeping secrecy of the telegram and a telegraph clerk, apparently without considering how to prevent the risk of the code from being broken, that is, without considering protection of secrets of the code, mechanically uses various codes according to the label of "secret", "treat as secret", etc. However, the degree of necessity to keep secrecy of the content of a telegram and the degree with regard to protection of secret of code are not always the same. For example, regardless of the importance of secrecy of the content of a conference with a foreign minister of the other country, the content is known to the other party. Thus, an encoded telegram reporting this would be relatively easy to decode for a codebreaking unit of the other country and, even if it is, there would be no great disadvantage to us. Thus, in this case, "secret" does not refer to secrecy to the government of the other country but secrecy to a third party in general.... To the contrary, instructions describing bargaining tactics concerning an important negotiation must be absolutely prevented from being known to the other country or a third party in general, whether by codebreaking or otherwise. If such a "secret" telegram is encoded with the same code as that for the above-mentioned "secret" telegram, there is a risk of the latter being used as materials to break the former. Further, since the more the same code is used, the more risk there is for the code to be broken. Thus, the code to be used for telegrams that require secrecy in every regard, such as the above-mentioned instructions describing bargaining tactics in a negotiation, must be limited in its use as much as possible in order to prevent codebreaking.
In this regard, telegrams are broadly classified as follows.
[The rest is omitted here.]

The above is followed by an explanation of proper use of various codes depending on the nature of the message to be encrypted (see another article). In addition, it was stressed that the full range of a code should be used so as to avoid spelling a word letter by letter (i.e., kana by kana) and that alternative groups should be used when the same word/phrase is frequently used in a telegram.

Delivery of the Japanese Translation of The American Black Chamber

The Japanese translation of The American Black Chamber was published in a book form in August 1931 (priced at one yen per copy). The publisher, Osaka Mainichi Shimbunsha, was the parent company of the Tokyo Nichi Nichi newspaper that covered the book in July.

On 22 August, Sakuma, Chief of the Telegraph Section, drafted a proposal on behalf of the Foreign Ministry asking for Cabinet approval to purchase 138 copies for distribution to overseas establishments, explaining that the book "describes in detail how the US codebreaking unit tried every vicious means to break codes of Japan and other countries and it would be very beneficial for protection of security of our code to let the staff of the overseas establishments, in particular telegraph staff, be informed of its contents." (Yardley file (1) p.72) (The "vicious means" appears to refer to breaking into consulates to get access to a code (cf. JACAR Ref. B12080889500 p.14) or an attempt to cause specific information to be telegraphed in order to get a crib (ibid. p.13), the kind of means also used by the Japanese.) The books were despatched as of 18 September (Yardley file (1) p.70).

August 1931 -- Observations of the Army

The Army also drew some lesson from The American Black Chamber. In August 1931, the "7th Section" of the Army drafted a memorandum titled "Measures for Communication Interception from the Viewpoint of Military Communication." Its introduction observes that although Yardley's book mainly deals with peacetime diplomacy, experience from the First World War necessitates measures against communication interception in peacetime as well as in wartime. The overall structure is as follows.

Introduction (C14010456900)

1. Communication Medium for Intelligence and Communication Policy (C14010457000) ... As evident from The American Black Chamber, as long as a foreign communication network is used, the foreign government may obtain a copy of a telegram. Short-wave radio may be used for intelligence agents but it involves a high risk of being discovered.

2. Choice of Communication Medium (C14010457100) ... Wireless media is prone to interception and wired media should mainly be used.

3. Communication Instruments (C14010457200) ... In amphibious operations and air-ground communication, wireless communication is prerequisite. Development of various technologies should be conducted.

4. Countermeasures against Enemy Communication Intelligence on the Battlefield (C14010457300) ... Countermeasures for wireless and wired communications are discussed. Among the quotations of the measures adoped by the British Navy is "7. Different codebooks should be used for different theatres of operations and should be frequently changed."

5. Communication Tactics (C14010457400) ... Active measures to deceive the enemy, rather than passive countermeasures, are discussed.

6. Cryptologic Service (C14010457500) ... Importance of codes/ciphers and prevention of codebreaking are discussed in view of The American Black Chamber. The following observation is made regarding preparation of new codes/ciphers.

The book [The American Black Chamber] has a passage "[the Japanese] had employed a Polish cipher expert to revise their code and cipher systems. Theoretically the Japanese codes were now more scientifically constructed; practically they were easier to solve than the first code" [p.279]. This should be taken into consideration by those who prepare codes/ciphers in future. That is, codes/ciphers should get rid of conventional practice and should be creative and appropriate in view of the character of the national language as well as suitable for the situations of the unit that would use the code/cipher, the capacity of the user, etc.

7. Format of Communication Message (C14010457600) ... many and long telegrams, repetition of common phrases, regular construction of telegrams, disclosure of a material to be compared with the communication content, etc. would help codebreaking. In particular, the descriptions of Chapter 15 of The American Black Chamber (p.281) are quoted to warn against routine telegram construction.

8. Detection of Enemy Communication Intelligence (C14010457700)

9. Defense agianst Espionage (C14010457800)

Comments of the Ministers of the Navy and the Army

Comments of the Ministers of the Navy and the Army appeared in the Osaka Mainichi newspaper on 31 August 1931.

This is Not Wartime -- Abo, Minister of the Navy
The publication of the Black Chamber is a big warning to the nations in the world. At wartime, it may be expected that there should be issues of secretive reading of coded messages. Indeed, there have been many cases in which a major war was influenced by code breaking. However, it is quite surprising that such an act occurred during an international disarmament conference, which should have been a gateway for peace. While I refrain from commenting on its contents because it has grave implications to international relations, it was very opportune that the Osaka Mainichi and Tokyo Nichi-Nichi newspapers published a translation of the book so soon, which was very beneficial in alerting the nation.

Important Lesson -- Minami, Minister of the War
Upon reading it, I did not feel like censuring the United States for such an act but drew a very important lesson that we should be on the alert in future. The next year will see a general disarmament conference [in London] and international relations will be more and more complicated, which would call for utmost caution on our side. In the military in particular, it would be disastrous if operational secrets were to be stolen by the enemy and we must buckle down for utmost security.

Proposal of Introduction of Cipher Machines

Even before the publication of The American Black Chamber, vulnerability of code was recognized and during the London Naval Conference of 1930, cipher machines made by the Navy were used by the Foreign Ministry. There was a proposal from the Navy after the conference that a cipher machine should be consistently used for all the important telegrams but it was only after the publication of The American Black Chamber that the proposal was renewed and adopted.

On 9 July 1930, Ito Risaburo, Chief of the Telegraph Section of the Naval Ministry, sent a document titled "Observations on Code/Cipher for Naval Conference" to several persons in the Foreign Ministry on behalf of the Navy (JACAR Ref. B04122581000). It explained techniques of codebreaking and superiority of cipher machines, which had been used to the great advantage during the London Naval Conference in January to April 1930. Ito recommended use of cipher machines for all the telegrams during the next major international conference because concurrent use of a weaker code might compromise the security of the machine cipher.

The proposal was forgotten until after the publication of The American Black Chamber in June 1931 and its newspaper coverage in late July. It may be because the new gadget had been unsatisfactory for the Foreign Ministry. Not that cryptologic security was questioned but there were several problems in practice. The Navy improved the design and Ito renewed the proposal by telephone to Sakuma on 29 July 1931 in view of the situation now that "the Yardley incident is much clamored".

Now Sakuma took up the proposal seriously and visited Ito. Satisfied with his explanation of improvements, he drafted a proposal on behalf of the Foreign Ministry, dated 6 August 1931, asking for Cabinet approval of introducing the new cipher machine of the Navy for use during the Geneva Naval Conference (Wikipedia) scheduled in 1932 (Yardley file (3), p.36-39). He pointed out the machine cipher would be more secure than the conventional code system and the defects of the previous model used during the London Naval Conference in 1930 were to be largely corrected.

This cipher machine, introduced in 1931 (or the Japanese Imperial Year 2591), would be known as RED as designated by the US codebreakers (Wikipedia).

(See another article for development of the RED machine.)

Preparation for Facing the Parliament

Q & A

On 28 July 1931, Chief Sakuma drafted replies to possible questions about the Yardley issue that may be brought up in the next session of the parliament (Yardley file (1) p.1-13, (2) p.51-66, (3) p.40-53). The first question concerns the responsibility of the Foreign Ministry and is adapted from Section (4) of the above memorandum of 10 June 1931 drafted shortly after the publication of The American Black Chamber.

(1) Question 1
(A) Yardley's The American Black Chamber states he solved Japanese government's telegrams and contains many English translations of what is alleged to be Japanese government's telegrams and a photograph of a decipherment in roman letters. Are they authentic or not? If authentic, it is not a fault on the part of Foreign Ministry?
(B) Reply
Solutions contained in Y.'s book are accurate to some extent. In the first place, unless an ideal cipher machine is used, any complicated code could be broken with scientific methods if substantial time and materials are allowed. Further, it may be presumed that every nation is making considerable efforts in codebreaking against other nations. (Of course, there is no confirmation; Y.'s book describes Britain and France have codebreaking institutes.) Thus, there is hardly a way to prevent that our code is broken some day. Notwithstanding, the Foreign Ministry is always taking the best measures for the protection of security of our codes. The methods of codebreaking described in Y.'s book are common methods known to code experts in general. Further, Y.'s breaking our codes by employing substantial expenditure and organization is nothing remarkable. It happened to have become an issue only because it was disclosed in a published book and therefore there is nothing to blame in the Foreign Ministry.

(2) Question 2
(A) Then, do we study foreign encrypted telegrams and succeed in breaking them in Japan?
(B) Reply
A very delicate issue is involved in replying to this question. Various options as follows may be possible.
Draft 1
"It belongs to the secrets of the government and no particular replies could be given."
Draft 2
"Y.'s book describes that the US Secretary of State has an opinion that foreign governments' encrypted telegrams are inalienable. We feel the same way and the Foreign Ministry does not conduct breaking of foreign codes etc."
Draft 3
Add the following to Draft 2.
"Needless to say, various studies of code which would be necessary to protect secrets of our codes are conducted. But nothing is done that might infringe on privacy of foreign governments' communication."

The other questions are as follows. None of these items concerns cryptologic security of diplomatic codes. (Their answers, occasionally bureaucratic, are of some interest but are off the main topic here.)

(3) As a result of our important instruction telegrams during the Washington Naval Conference being broken by the US codebreaking unit, did we not suffer a major disadvantage during the negotiation?
(4) While Y. describes in his book that he obtained materials for codebreaking by various means such as bribing embassy staff or photographing a codebook by opening a safe etc., were there such incidents in our establishments in the US, in particular the embassy in Washington and the consulate general in New York?
(5) Regarding Y.'s book, has any remonstrance or representation been made to the US government?
(6) What sort of person is Yardley and what were the organization and activities of the US codebreaking unit?
(7) Is there still a codebreaking unit in the US?
(8) Do any states other than the US have a codebreaking unit and are engaged in breaking foreign codes?
(9) During the Washington Naval Conference, were telegrams other than those of the Japanese government broken?
(10) While Y.'s book describes breaking of codes of Spain and Mexico more or less in detail and the fact of breaking codes of the Japanese government in greatest detail, there are little description for other states such as Britain or France. What is the reason for this?
(11) What about making an international treaty stipulating that diplomatic communications are inviolable?
(12) Reportedly, there are deplorable incidents of slack discipline such as delivery men openly entering the Telegraph Section of the Foreign Ministry (Nippon newspaper, 1 August 1931) or a diplomat lacking care when carrying a code; or failing to lock the safe storing codes in establishments overseas (Nippon newspaper, 1 August). Are these true?
(13) Since radiotelegraph may be intercepted, wire telegraph would be safer for keeping secrets of code. Which of radiotelegraph and wire telegraph does the government mainly use? [Answer: Radio is cheaper. Even if wire telegraph is used, a copy may be obtained at the destination country. So "with regard to protection of code security, there is no point in distinguishing wire and radio. A code that may be broken when sent by radio would also be broken if sent by wire. Therefore, although the Foreign Ministry uses radio relatively frequently, wire telegraph is also used in some cases."]
(14) How is Y.'s book received in the US.
(15) Apart from the above, someone who has some outside information about our suspected code leak incidents may ask questions. Relatively recent ones of such incidents include the following.
(A) Incident of Clerk Yatabe during the Genoa Conference about Seamen (ILO) (July 1920, Telegram No. 1087 from Embassy in France etc.)
(B) Incident of Policeman Sagawa at Consulate in Manchuria (February 1923, Telegram No. 54 from Consulate General at Harbin etc.)
(C) Incident of Interpreter Trainee Nakamura at German-Swiss Border (June 1923, Telegram No. 142 from Embassy in Germany etc.)
(D) Incident of Employee at Embassy in Turkey Revealed by Agabekov (Wikipedia) (Secret Telegram No. 30 from Ambassador in Turkey dated 30 January 1931 etc.) (cf. JACAR Ref. B04013190800, B04013191100).
All these incidents have been reported in detail and they either involved no actual damage or were ungrounded. (It should be remembered that secret Japanese telegrams disclosed by Russia in 1933 were alleged by the Japanese government to be sham, but there is evidence that the denial was false; cf. Shiraishi p.97)

Privacy of Communication

Regarding (5), Sakuma noted reports were expected from the Ambassador to the US and further information would be available by the parliament session.

Indeed, on 31 July 1931, the Ambassador was instructed to inquire into whether Yardley's disclosure was a breach of federal law or state law by infringing on privacy of communication. The Ambassador reported on 11 August after an interview with an official of the Department of State that, while one should wait and see the proceedings of the investigation by the Department of Justice, the only legal problem if any was how he could obtain the telegrams.

The Ambassador was further instructed to collect information about US federal and state laws as well as rules of telegraph companies about censorship and privacy of communication. Exchanges went through September, November (Yardley file (1) p.75-78, (2) p.7-11), and December (Yardley file (2) p.18-36).

Sakuma compiled information from the Ambassador in a memorandum dated 21 March 1932 about privacy of communication in the US (Yardley file (1) p.27, 39, 40, (3) p.1). Sakuma used this as a basis for a supplement of the possible questions for the parliament dated 15 May 1932 (Yardley file (1) p.15, (2) p.66, (3) p.54). Further inquiries were made on 20 May 1932 (Yardley file (2) p.49) and, with reports in response, Sakuma prepared a supplement dated 10 August 1932 for the memorandum of 21 March 1932 (Yardley file (1) p.34, (3) p.8).

These efforts did not find any politically practical means to charge Yardley or telegraph companies. Nor did Sakuma intended that. As a bureaucrat, he was doing his best to discharge accountability. His very first memorandum of 10 June 1931, drafted shortly after the publication of The American Black Chamber, states "Since any of the above measures would not be appropriate, after all, it would be sufficient to consider domestic measures, that is, how to respond when the issue is taken up by the parliament or domestic newspapers."

Parliamentary Session

In the mean time, the session of the 60th Imperial Parliament this winter had been from 26 December 1931 to 21 January 1932. As far as can be known from the proceedings of the Parliament (NDL), the plenary session under a new cabinet inaugurated in December 1931 was dominated by eagerness to approve the marching of the army in Manchuria since the Mukden Incident in September 1931.

Did Yardley Sell Secrets of Codebreaking to Japanese Embassy?

Sakuma's memorandum dated 10 June 1931, translated above, mentions more than once a deal whereby the Japanese embassy paid 7000 dollars to Yardley for copies of decoded telegrams etc. This memorandum is mentioned on p.273 of Kahn (2004) (David Kahn Official Website) as a source of two American authors reporting this story of Yardley's betrayal.

Kahn doubts the story and believes that it was fabricated to denigrate Yardley and save Japanese face. This sounds right as for the alleged offer of a deal during the Washington Conference, which was mentioned in the Foreign Ministry's comments on the 22 July issue of the Tokyo Nichi Nichi and which the reporter of the Nippon said he had heard.

However, Sakuma's memorandum refers to June 1930 and the references sound like more than a mere fabrication. Rather than trying to denigrate Yardley to save face of Japan or the Foreign Ministry, the author seems to be really concerned that Yardley might disclose his selling the documents.

While Kahn mentions "great motivation" for the allegation, Yardley also had his own motivation. In October 1929, when Yardley's Cipher Bureau was abolished, the Great Depression plunged the world into a devastating economic crisis. Yardley gave up his fancy apartment in New York and removed to his hometown in Indiana. (Kahn (2004), p.104) Report has it from his friends that Yardley offered his documents to the US government when he left the service but the government refused them (Yardley file (3) p.14, citing New York Times, 21 February 1933). In April 1930, his application to the navy for a position of an instructor in cryptanalysis was turned down. So was his proposal in the spring to a publisher to write about his activities, after which he "did nothing for a few months", until he found a contact which led to the publication of The American Black Chamber (Kahn (2004), p.104-105). The date "June 1930" of the Telegram No. 105, albeit not discovered, coincides with the blank in Yardley's career during the nadir of his fortune.

What happened in Japanese code security measures after June 1930?

Chief Sakuma of the Telegraph Section replaced his predecessor Sawada in July 1930. He wrote in a later memorandum dated 9 October 1933 (see above) that at the time he took office "there was the Yardley issue" (JACAR Ref. B12080889500 p.20) but the expression "at the time" may broadly refer to "ever since he took office." (Considering he was concerned that Yardley might disclose the deal, it is unlikely he casually mentioned it even in a memorandum stamped "top secret".)

On 9 July 1930, Ito submitted his first proposal of introduction of cipher machines. The Foreign Ministry did have used cipher machines of the Navy during the London Naval Conference in January to April 1930 but Ito thought it fit to make a proposal at this timing. However, Ito's first proposal is silent about Yardley, while it does mention older code security incidents (see another article). Thus, it appears Ito did not know the deal.

At the end of 1930, a naval code was updated from Red to Blue (dubbed by the US Navy codebreakers) (NSA; Budiansky, Battle of Wits, p.83; "Red" should not be confused with the RED cipher machine mentioned above). But of course, the Navy no more needed "the Yardley issue" than the Foreign Ministry to be motivated to revise its code.

Thus, no more convincing evidence for the deal was found than the really concerned tone of Sakuma's memorandum of 10 June 1931 and the coincidence of the timing mentioned. Although I am inclined to think the offer in June 1930 was actually made, I would then have to explain why Shiratori, Head of Intelligence Department, mentioned it to the media, when Sakuma, Chief of Telegraph Section, feared its disclosure.

At least, the question should be left open until someone finds further evidence, either positive or negative. Needless to say, even if there was the deal, it does not have to detract from Yardley's achievement in codebreaking.

(Notes added in November 2014) A recollection by a witness is recorded in Kosaka Masataka and Nagata Junko, "Nichi-bei Ango Senso 2 -- Kimitsu Ango Purple" in Rekishi-heno Shotai 27, NHK, Japan (1983), p.151-152, based on a TV program in 1982 (NHK On-Demand). Kase Toshikazu (1903-2004) (National Diet Library (in Japanese), Wikipedia (in Japanese)), assistant secretary in his late twenties to the Japanese embassy in Washington at the time, says he was present when secretary Togo Shigenori (Wikipedia) received Yardley, who brought many deciphered papers (but Kase does not mention any monetary request). Deciphering was one of assistant secretary's jobs and Kase confirmed to Togo that Yardley's deciphering was correct.

KASE: One day, a suspicious-looking man appeared at the embassy and demanded an interview with the Ambassador. Asked at the reception about an appointment, he said he had none. Even though he was told he couldn't meet the Ambassador without an appointment, he wouldn't leave. Then, someone sent word to the Secretary [to the Embassy]. It was decided then to ask Secretary Togo (Togo Shigenori, later Foreign Minister) to receive him. It was decided then that, as an assistant, Kase should be present, too, and Togo and I met that suspicious man. He was a hard-featured man and had a gloomy face.
When asked of his business, he said outright all our codes were read by the Americans. It was too much of a shock. Rather, we did not believe him, neither Togo nor I. Then, he put a heavy bunch of papers on the table, saying those were the deciphering of Japanese codes. Secretary Togo was not familiar with code. I was an assistant secretry. Deciphering of codes was one of my jobs and I knew some simple codes by heart. I took up some and felt uh-oh. About eighty percent was deciphered. Of course I couldn't tell it before Yardley. So I signaled the Secretary with my eyes to the corridor and said, yes, they did read. It was Secretary Togo's turn to be surprised and ask whether it's true. Then, it was reported to the Ambassador. The Ambassador was surprised and petrified. A telegram was sent to the home office, where people were further surprised.

Kase was also a prolific writer and he provides more details in his memoirs (Kase Toshikazu Kaisoroku (1), 1986, p.20-21). The memoirs date Yardley's production of his deciphering on the day after his first visit. He further identifies the code he recognized as a decipherment table "P go" ("the Issue P"). (Of course, he could not have recognized a code used back in 1921-1922, since he was in the US as assistant only from June 1928.)

Further according to Kase's memoirs, Yardley not only asked the embassy to buy the deciphering with 10,000 dollars but also offered to teach the art of deciphering. The home ministry decided to accept the offer and Kase was chosen as the pupil because he had already met Yardley. But Yardley disappeared without giving his lessons.

That Togo's memoirs and biographies do not tell the episode may be natural because he must have been more concerned about what was told in telegrams rather than the ciphering of the telegrams. The problem of Kase's testimony, however, is that he identifies the date as "the spring of 1929". He appears to have relied on the time of the inauguration of President Hoover and Secretary of State Stimson (March 1929) but it is highly improbable that Yardley ever thought of selling the secret of the Cipher Bureau so early, considering that Yardley seems to have had no concern as of the end of May (Kahn (2004), p.267, the sixth note; p.97). (Moreover, the Cipher Bureau was paid until the end of October 1929 (Kahn (2004), p.97-99).) One is tempted to consider it was an error for "the spring of 1930", which would match Sakuma's memorandum and the nadir of Yardley's career. However, Togo left Washington in May 1929 (Kase helped his packing) (Sofu Togo Shigenori no Shogai [Life of My Father, Togo Shigenori], p.111). Is it possible that Kase mistook who was his superior secretary at the time?

(Notes added in February 2018) I now find the issue is thoroughly reviewed by John Dooley (2011), "Was Herbert O. Yardley a Traitor?", Cryptologia 35(1):1-15 (pdf). He agrees with Kahn that "Yardley was a rotter, not a traitor" because no clear evidence has been found. Although he does not seem to be aware of Kase's testimony above, he examined many materials and previous publications, which would be a starting point for further research.

Yardley's alleged deal with Japan became widely known by Ladislas Farago (1967), The Broken Seal: The Story of "Operation Magic" and the Pearl Harbor Disaster but Dooley points out that none of the sources given by Farago (except for one memo from 1925, which does not mention Yardley but William Friedman) are before the publication of The American Black Chamber. Of the materials examined by Dooley, "The Sakuma Memorandam" dated June 10, 1931 [21] is translated above. "Draft of Answers to Hypothetical Questions in the Diet about Yardley's Black Chamber prepared by Chief of Cable Section" by Shin Sakuma July 28, 1931 [8] is given above as "Q & A" in excerpts. The "Circular" of September 18, 1931 not found in JMFA seems to be the one in Yardley File (1), p.70, mentioned above that notified despatch of copies of Yardley's book.

Farago dated the alleged treason in the summer of 1928 but other researchers revised it to the summer of 1930 (p.2). The latter is consistent with the Sakuma Memorandum. It is wondered whether Farago's "1928" was simply an error.

Yardley Continues Disclosure

Even after the publication of The American Black Chamber, Yardley did not stop taking advantage of the materials he retained after the Cipher Bureau was abolished. They were reportedly kept in safe deposit boxes at various places. Though Kahn considers it unlikely, at least when Yardley needed them in late 1931, Yardley asked D. Laurans Chambers, the editor of Bobbs-Merrill, the publisher of The American Black Chamber, for "the Japanese telegrams that are in your safe" (Kahn (2004) p.278, p.158, Yardley file (3) p.14, citing New York Times, 21 February 1933).

On 11 June 1931, Yardley revealed in Chicago Daily News what is alleged to be a Japanese diplomatic message sent in 1921 containing instructions for propaganda for the coming Washington Naval Conference (Yardley file (1) p.50).

In the issue of 10 October 1931 of Liberty, Yardley further disclosed more recent Japanese telegrams: one dated 14 May 1927 from Tokyo to London and another dated 17 May 1927 from London to Tokyo (Yardley file (1) p.79-86). The Yardley files also include the original Japanese messages (Yardley file (2) p.1-7).

Yardley's activities could not escape attention of the Japanese embassy. On 2 June 1932, the Ambassador in the US sent two copies of Yardleygrams (Yardley file (2) p.51), which presented codebreaking puzzles in story form for the general public.

More alarming was a report from New York in the spring of 1933 that Yardley's manuscripts titled Japanese Diplomatic Secrets were seized by the authority (Yardley file (3) p.12). Written by a ghostwriter, it was a bore, interspersing the intercepted dispatches (hundreds of dispatches during the Washington Naval Conference) with scraps of dry, impersonal text (Kahn (2004) p.159).

The renewed attempt to publish intercepted telegrams might have deteriorated foreign relations for the US. Thus it led to an introduction in the US Congress a bill "For the Protection of Government Records" on 27 March 1933, though Yardley was never mentioned when it was first proposed (Kahn (2004) p.162, 164).

The Foreign Ministry of Japan followed these issues intently (Yardley file (3) p.15-29). They detected that the bill was motivated by Yardley's seized manuscript (Yardley file (3) p.23, 18, 20). They were alarmed in particular because there was information that the seized manuscripts contained Japanese telegrams related to the Mukden Incident in 1931 (Yardley file (3) p.18). The Embassy in the US contacted the publisher, Macmillan, through an agent and, relievingly, had an impression that Yardley's manuscript was a recompilation of old materials, not a product of new materials after the Mukden Incident (Yardley file (3) p.21).

In the US, the bill drew protests as infringing upon freedom of speech. The closed hearings in the Senate committee limited the bill to "code", thus effectively directly aiming at Yardley (Kahn (2004) p.165, Yardley file (3) p.25-27). When the bill was discussed in the Senate on 10 May, the advocates of the bill denied it was designed to punish a particular person but Yardley's name came up in the debate (Kahn (2004) p.165-166). One senator was convinced that "there are other motives behind this proposed legislation, and entirely unrelated to Mr. Yardley" by calling the bill a "gag law." (Kahn (2004) p.168-169) But the bill became law in June (Kahn (2004) p.169-171; US Code 952 of Title 18).

Japanese Reaction to Code Security Incidents

Upon the publication of The American Black Chamber, the efforts of the Foreign Ministry of Japan were directed to legal issues concerning privacy of communication rather than replacement of codes. Apparently, it was because the very code used during the Washington Naval Conference was no longer used and the code currently in use as of 1931 was a more complex one, albeit the same hybrid two-letter/four-letter system was employed (see another article).

The reason that the same system as the one broken by Yardley was still in use is explained in Sakuma's memorandum dated 9 October 1933 mentioned above (JACAR Ref. B12080889500).

A relatively simple one of the hybrid two-letter/four-letter codes (about 800 entries) is said by Yardley in his book to have been broken. Notwithstanding its security being insufficient, the staff of the Foreign Ministry are most accustomed to it and it is advantageous in cost reduction and is convenient to interpret mutilation of telegrams. Thus, it is the most frequently used one, being a code between the most secure code and mere abbreviating code. (It is used for messages that should be kept secret to some extent and that would not cause great harm even if scientifically broken.) However, the current NI Code is not very easy in its use. With the present amount of telegrams, two or three times that before the [Mukden] Incident everyday, the current staff of the Foreign Ministry is almost fully occupied even if the NI Code is the only one used.

It was recognized that any code could be broken and the codebreaking per se was no surprise. As long as the current code was not compromised, it was no serious concern.

Such an attitude may be contrasted with the reaction to other code security incidents. When there was an incident of telegram theft at the embassy in Turkey in June 1932, the RO Code was replaced by the HA Code, which had already been in use in the embassy in Russia. When it was found that the HA Code was known to Russia in 1933, it was discontinued for establishments in Russia and superencryption with a memorable key was introduced (see another article). At the latter occasion, detailed inquiries were made as to whether the leak was due to codebreaking or theft by burglary (JACAR Ref. B12080889300, esp. B12080889500).

Yardley's publication did not cause replacement of Japanese diplomatic code outright but it was mentioned when there was a renewed proposal from the Navy for use of a cipher machine, which had been tentatively used during the London Naval Conference in 1930. While it was recognized that conventional code with superencryption with a memorable key was not absolutely secure, a cipher machine was believed to provide the maximum security.


David Kahn (2004), The Reader of Gentlemen's Mail, Herbert O. Yardley and the Birth of American Codebreaking

David Kahn (1967), The Codebreakers

Japan Center for Asian Historical Records (JACAR)

Yardley-related materials are under Reference Codes B13080930400, B13080930500, and B13080930600 (referred to as Yardley file (1), (2), and (3) herein; the page numbers are those of digital images). The following lists some related documents.

Navy's proposal of introduction of cipher machine (9 July 1930) by Chief Ito Risaburo of the Telegraph Section of the Naval Ministry (B04122581000)

Memorandum shortly after the publication of The American Black Chamber (10 June 1931) by Chief Sakuma of the Telegraph Section of the Foreign Ministry (Yardley file (2) p.40)

Anticipated Questions and Proposed Answers of 28 July 1931 by Sakuma (Yardley file (1) p.1-13; Yardley file (2) p.51-66, Yardley file (3) p.40-53)

Proposal of Introduction of Cipher Machine of 6 August 1931 by Sakum (Yardley file (3) p.36-39)

Memorandum on a code security incident (9 October 1933) by Sakuma (B12080889500 p.20-36), printed in Shiraishi Masaaki (1998), "Iwayuru 'Kaibunsho Jiken' ni kansuru Sakuma Denshin Kacho Kiso Chosho ni tsuite", Gaikoshiryokanpo, June 1998, pp.97-104 (白石仁章「いわゆる"怪文書事件"に関する佐久間電信課長起草調書について」,『外交史料館報』,1998.6,pp.97-104)

©2014 S.Tomokiyo
First posted on 14 March 2014. Last modified on 23 February 2018.
Articles on Historical Cryptography
inserted by FC2 system