Len Scott, Peter Jackson and R. Gerald Hughes (eds.), Exploring Intelligence Archives: Documents and the History of Intelligence and International Relations (Frank Cass).
Despite a thick London fog which delayed the arrival of many of the attending dignitaries, the London Naval Conference opened with great glitter and ceremony in the Royal Gallery of the House of Lords on 21 January 1930 with a speech of welcome from King George V. Attended by the world’s five major naval powers—Britain, the United States, Japan, France and Italy—it would meet continuously until late April in an attempt to extend the regime of naval limitation that had been agreed to eight years earlier at a conference in Washington. But there was more to these naval disarmament negotiations than met the eye. Conferences of this sort provided manifest opportunities for the intelligence services of the host nation, as control over the telegraphic facilities into and out of the country allowed for the easy interception of communications made by the delegations of visiting powers. It was an advantage which the Americans had seized at the Washington Naval Conference of 1921-22 to read the Japanese diplomatic traffic, giving them a significant advantage in negotiations.[1] At the London conference, one of the chief weapons in the British negotiating armoury was its similar ability to read most of the messages exchanged between the various delegations and their respective home governments. The British signals intelligence body, known as the Government Code and Cypher School (GCCS), ‘was one of the world’s largest code-breaking agencies [and] … possibly the best on earth between 1919 and 1935’.[2] In the course of each month during 1929–31, for example, it developed a file averaging about 300 decrypted diplomatic intercepts, ranging over American, Japanese, Italian, French, Greek, Persian, Spanish, Scandinavian and other minor states’ transmissions.[3]
The signals decrypted by the GCCS are now located in The National Archives, Kew, in series HW 12. The series comprises 338 volumes of intercepted diplomatic communications that stretch over the period 1919 to 1945; the earliest files began to be released for public inspection in 1996. The two documents presented here come from the special file opened by the GCCS to cover its work during the London Naval Conference: HW 12/126, ‘International Naval Conference for Fleet Reduction, January-April 1930’ (available for public consultation since 1997). It contains a total of 397 transmissions made by the attending parties that the GCCS succeeded in intercepting and deciphering. They had only partial success with American ciphers, leaving those decrypts with numerous gaps that could make interpretation difficult, while the French and Italian use of the diplomatic bag for many of their critical transmissions meant a less decisive intelligence yield. The Japanese traffic was a sparkling success, however, as their codes were almost completely penetrated, though there was sometimes a significant time lag in cracking particular messages. It is here, therefore, that the most significant influences on the London negotiations might be looked for.
What was at stake between Britain, the United States and Japan at the London conference was the global balance of naval power. The Washington Naval Treaty of 1922 had imposed a system of ratios to limit the total tonnage of capital ships (warships over 10,000 tons) among these three powers at 5 : 5 : 3, respectively.[4] This had successfully averted the threat of an Anglo-American naval race in battleships, yet other problems remained. The Washington conference left unsettled the limitation of warships displacing less than 10,000 tons, namely cruisers, destroyers and submarines, the main weapons for attacking and defending maritime lines of communication. In addition, Japanese navalists were left deeply unhappy at being forced to accept a 60% ratio in capital ships, rather than the 70% they saw as essential for national security and as an element of Japanese national prestige. More moderate naval voices in Tokyo had perceived the necessity for Japan to accept the proffered 60% ratio, both for political considerations in easing strained American-Japanese relations and to avoid a ruinously expensive naval race in the Pacific. The debate within Japanese naval circles over the question of relative strength compared to the United States would reverberate throughout the following decade. Following the failure of the first attempt to limit warships under 10,000 tons, at the Geneva Naval Conference of 1927, these two issues came to a head at the London conference in 1930, which again sought to tackle the problem. The Japanese were determined not to allow another denigration of their claims. All Japanese policymakers believed that any naval agreement should assure Japan’s predominance in the western Pacific. In particular, there was widespread consensus on the absolute need for a 70% ratio in heavy cruisers carrying 8-inch guns, which the Japanese admirals viewed as ‘semi-capital’ ships that could help offset the hated 60% ratio in battleships.[5] In contrast, the main goal of the British government at the London conference was a three-power agreement with the United States and Japan that would maintain the delicate Anglo-American-Japanese naval balance and preserve a naval regime that entrenched British superiority.[6] An Anglo-American compromise was successfully achieved in the conference’s opening weeks; the task of finding a formula that dealt with Japan’s demands for an increase in her tonnage ratio took nearly two months before the London Naval Treaty was finally signed on 22 April.[7]
The documents below come from key moments in this complex negotiation. The first document is a telegram sent from the Japanese Admiralty in Tokyo to the Japanese delegates in London during the opening weeks of the conference, demanding firmness in Japan’s negotiating position. The second document is a telegram from the delegation in London to the Japanese Foreign Minister in Tokyo near the end of the conference, indicating a willingness to accept a compromise solution. Both were intercepted and more or less completely decrypted by the GCCS. Some technical information will be helpful in reading and evaluating these two intercepts. The summary title at the top was provided by GCCS. In the top right corner is the GCCS intercept number and the date of decryption. Below on the left is the information on sender and recipient and the original despatch number and date of transmission. Items in square brackets are reprinted as they appeared in the original documents.
Document 1: ‘Naval Conference: Japanese Admiralty Views on American Proposals’, HW 12/126, N.C. 199.
Document 2: ‘Naval Conference: Japanese Summary of the Situation and Request for Instructions’, HW 12/126, N.C. 220.
MOST SECRET
Naval
Conference: Japanese Admiralty Views on American Proposals
No: N.C. 199
Date: 14th March, 1930.
From: Head Military Affairs Bureau, Admiralty, TOKIO.
To: Japanese Naval Delegation, LONDON.
No: 39.
Date: 8th February, 1930
Plenipotentiaries’ telegram Nos. 105 and 106 [our N.C. 97 and N.C. 88]. Telegraphic instructions regarding the American draft proposals have already been sent, and whilst we believe the Plenipotentiaries will take the best action in the circumstances, the views here, which are as follows, are communicated for information:-
1. Although the principal points of our demands are already known to the other side, to give no consideration at all to them is not only unfair, but from another point of view must be considered an insult.
Thus if we make a numerical agreement in accordance with the earnest desires of the other side our faith will more or less come to nought.
2. According to the American draft plan not only would they abandon the claim for twenty-one cruisers but in certain circumstances, it appears, they would be satisfied with fifteen.
At the same time as they make this concession, they compare us with 60%, and would use this concession as a powerful argument in checking our claim for 70%. This is nothing more than a ‘bargain’.
Whichever way you take it the American concession in eight-inch cruisers shows that considerable concerted action has been set up between GREAT BRITAIN and the UNITED STATES. Thus, showing a combined front they plan to defeat first JAPAN and then FRANCE and ITALY in exactly the same way as at the Washington Conference.
It goes without saying that this is the parting of the ways so far as the success or failure of our demands is concerned.
3. According to our observations on the frequent telegrams from our Plenipotentiaries, a slight anxiety has been noticed in both ENGLAND and AMERICA. As the promoters of the Conference, that is their affair, and we may be composed and calm.
However, if they maintain their firm demands without conceding anything and proceed selfishly, ignoring us, then it will be our best plan to lead affairs into a situation where there is little hope of agreement between the three Powers.
4. Whilst GREAT BRITAIN and the UNITED STATES on one hand admonish [us] for leakage of the private negotiations, it is extremely difficult to understand how on the other hand they take advantage of a favourable opportunity and publish things at will. A typical example of this was the recent STIMSON statement, where he took formal steps to secure an understanding beforehand.
Thus it is most disadvantageous to us, as it is impossible for us to appeal to the fair judgment of public opinion.
We consider it important that we must of a necessity, as a counter-measure, secure a rigid understanding regarding the limits within which there shall be freedom to publish.
5. It goes without saying that whilst we adopt the above-stated attitude we must make preparations to draw up our concrete counter-proposals, and to present them at a favourable opportunity.
6. The suggestion in the American plan that we should each build one capital ship of 35,000 tons is contrary to the spirit of reduction which we and AMERICA have always advocated, and not only would it cause loss of confidence at home and abroad but as it would create a reason for making reduction in the future difficult, the Imperial Government could never agree.
Further the [? warship reduction proposals] in the American plan Paragraphs [? 14 and 15] would cause a change in the substance of the WASHINGTON Treaty:-
(a) By the postponement of replacements our object of reducing burdens can be easily achieved. If it is not, a reduction in ordinary expenditure can be made [? depending on the type of ship].
If you especially scrap, it can be understood that in [three groups] you would rather increase the burden; and finally you would not achieve the object. Therefore it is not agreeable to carry out the [? warship reduction] plan.
(b) As there will be
large differences in the effect on fleet strengths caused by crapping the KONGO
type with fourteen-inch guns and scrapping American vessels with twelve-inch
guns, it is considered that careful examination of this point and others should
be made before deciding yes or no.
MOST SECRET
Naval
Conference: Japanese Summary of the Situationi and Request for Instructions
No: N.C. 220
Date: 19th March, 1930.
From: The Japanese Delegation, LONDON.
To: The Foreign Minister, TOKIO.
No: 208.
Date: 14th March, 1930
[Cypher Class A].
[This message is referred to in LONDON to TOKIO telegram of 14th March, our N.C. 213].
Confidential:
During the past two months and more we have consistently maintained our claims with the ultimate result that the British and Americans have been driven to voice their dissatisfaction at our attitude as too unaccommodating, and as being one of obstinate adherence to our own country’s position and of failure to manifest the spirit of international cooperation. Nevertheless we have in wise abated our demands, but have patiently laboured to induce the British and Americans to approximate their views to ours, even hinting at a determination on our part not to shrink from breaking off negotiations, if they attempted to force upon us an unreasonably low rate.
Latterly however, as may be perceived from the WAKATSUKI-STIMSON interview of 12th instant following upon the MATSUDAIRA-REED conversations, the Americans have already practically recognised the principle of a collective seventy per cent ratio, there being in reality a disparity of something over two-hundredths. This seems to evidence anxiety on the part of the Americans to try to meet JAPAN’s wishes without laying themselves open to the criticism of an absolute surrender to Japanese demands. In regard to large cruisers they have not fallen in with our views, but actually we may take it that in the main up to the next Conference we retain the strength of seventy per cent or over. As regards submarines it is a subject for regret that the quota is less than we claimed, but we may regard it as a concession on the part of the Americans that they offered parity with us by a reduction of their own quota.
In our view, so long as a new situation does not develop, it would be difficult to induce the other side to make concessions in excess of the above. Whatever may be the case if a Five Power Agreement fails to materialise with the French question as the central factor, we should on all accounts be gravely affected if a breakdown of the present Conference should supervene as a consequences of JAPAN’s attitude. Careful reflection is therefore essential. Hereafter in the light of the Franco-Italian attitude and of other developments of the situation we shall of course exert our very best efforts to secure a successful realisation of our claims; but it is our hope that at this juncture the Government will consider the course of the above-mentioned negotiations and will in reply favour us with instructions.
Peter Mauch (Ritsumeikan
University), ‘The Japanese Navy and the London Naval Conference’
The London Naval Conference brought together delegates from the United States, Great Britain, Japan, France, and Italy for the purpose of extending and strengthening the era of naval limitation that had been ushered in by the earlier Washington conference. The Washington negotiators had not, however, agreed to limit the building of smaller vessels—including cruisers, destroyers, and submarines—and thereby unwittingly sparked a new race in cruiser construction. After a failed attempt to limit these vessels at Geneva in 1927, the chief naval powers concluded the London conference by signing a treaty which allowed Japan a ratio of 10 : 10 : 6.975 in auxiliary categories.
The United States and Britain entered the London conference convinced that the existing 60% ratio in capital ship strength was equally applicable to cruisers and auxiliaries. The negotiating position of Japan’s delegates, including Navy Minister Admiral Takarabe Takeshi, called instead for a 70% ratio in both cruisers and auxiliary vessels, and 78,000 tons in submarines. The 70% ratio vis-à-vis the United States Navy—the Japanese navy’s chief hypothetical enemy—had long been a cardinal objective of Japan’s sailors. Anything less, they reasoned, jettisoned Japan’s chances of defeating the United States on the expanses of the Pacific. This much virtually all Japanese naval officers agreed on. What remained in hot dispute throughout the London conference was the question as to whether the 70% ratio represented Japan’s final position, or whether it was reducible in the course of diplomatic negotiations. At issue was the very rationale of the Japanese navy. Did it exist to fight its American counterpart, or was it a tool of deterrence against American attack?
Throughout the London conference, a small cadre of Japanese officers—often referred to as the ‘treaty faction’—adhered firmly to the latter position. This group included the Navy Ministry ‘trio’ of Vice Navy Minister and Vice Admiral Yamanashi Katsunoshin, chief of the Naval Affairs Bureau, Rear Admiral Hori Teikichi, and Navy Ministry Adjutant Captain Koga Mineichi; Admirals Okada Keisuke, Suzuki Kantarō, and Saitō Makoto; and Vice Admirals Nomura Kichisaburō and Kobayashi Seizō. These men actively identified themselves with the legacy set by Admiral Katō Tomosaburō at the earlier Washington conference. They argued that the urban and industrial complexes which sustained America’s maritime might were such that even a 70% ratio for Japan would not bring victory in the Pacific. They were also convinced that an unbridled naval arms race could result in a ratio far less favourable for Japan than might be secured by international agreement. This shared weltanschauung goes a long way toward explaining Admiral Suzuki’s tart comment: ‘Only the mediocre could clamour for 70 percent or nothing’.[8]
On the other side of the divide stood those officers—often referred to as the ‘fleet faction’—who maintained that, because the navy existed to fight its American counterpart, the Washington conference’s 60% ratio in capital ship strength was a mistake that should not be repeated at London. According to this reasoning, Japan’s insistence on a 70% ratio was non-negotiable. Throughout the London conference, the chief proponents of this view were Chief of the Navy General Staff Admiral Katō Kanji and his immediate subordinate, Vice Admiral Suetsugu Nobumasa. These officers and their large band of followers maintained that should the London conference break down as a result of an Anglo-American refusal to meet Japan’s demand, the Japanese navy would still attain the vital 70% ratio, in the context not of international cooperation but of competition.
The arguments of these hawkish elements were compelling for a very good reason: the Navy General Staff, which was responsible for command and operations, maintained virtual independence from the Navy Ministry, which was responsible for the administration, maintenance and overall control of the service. This goes a long way toward explaining the first document under review (N.C. 199). Authored by Rear Admiral Hori, this document dates to a time when US-Japanese negotiations in London had reached an impasse. The Japanese delegates maintained an insistent demand for a 70% ratio, while their American counterparts were equally insistent upon a 60% ratio. Against this backdrop, Hori’s telegram—his widely recognized position as an advocate of naval limitation notwithstanding—gave voice to an exceedingly hard-line stance. In what amounted to an ‘insult’, the American delegates had given ‘no consideration’ to Japan’s negotiating position. The United States and Britain, moreover, were working together to ‘defeat’ Japan. Should this situation continue, it was best for Japan to ‘lead affairs into a situation where there is little hope of agreement between the three Powers’. What might be gleaned from this document is the treaty faction’s need for concessions from the United States and Great Britain if it were to successfully override the fleet faction’s ‘70 percent or nothing’ arguments.
The second document under review (N.C. 220), authored by Japan’s delegates to the London conference, came against the backdrop of a Japanese-American compromise plan produced on 13 March, the Reed-Matsudaira compromise, under which Japan in fact gained many of its claims. The Japanese navy would accept a 60% ratio in heavy (8-inch gun) cruisers, but only in exchange for allowing Japan a 70% ratio in light (6-inch gun) cruisers and destroyers, and parity in submarines at the level of 52,500 tons. Rather than having to return to Tokyo without a treaty and be held responsible for the failure of the conference, the Japanese envoys, including Wakatsuki Reijiro, a former prime minister, and Matsudaira Tsumeo, Japan’s ambassador in London, recommended accepting this formula. As the telegram made clear, the terms of the compromise plan ‘practically recognized’ the 70% ratio. In terms of submarine tonnage, the plan fell somewhat short of the ‘quota [Japan] called for’, although it did offer US-Japanese ‘parity’. Implicitly asking for the government’s consent to the compromise plan, the delegates warned that it ‘would be difficult to induce the other side to make concessions in excess of the above’. In Tokyo, the cabinet of Prime Minister Hamaguchi Osachi—which included the foremost advocate of cooperation with the United States, Foreign Minister Shidehara Kijūrō—was receptive to this plan. They believed that the negotiators in London had accomplished as much as could be expected. Japan’s limited financial capabilities ruled out a naval race and they were loathe to appear responsible for breaking up the conference. The Navy Ministry also accepted the plan; though they were not entirely happy with the formula, the diplomatic, political and financial considerations meant the imperfect bargain had to be accepted. For the hard-liners of the fleet faction, however, the compromise contained both the symbolism and substance of defeat. After lengthy debate, the government finally insisted that the compromise had to be approved for the sake of the nation’s foreign policy, even though the naval authorities considered it unsatisfactory from the defence standpoint. Prime Minister Hamaguchi on 1 April cabled his government’s acceptance of the plan to the delegates, and some three weeks later the London Naval Treaty was signed. The aftermath of the conference was, however, messy and complicated. The Navy General Staff registered its unalterable opposition to the treaty, and in the process charged the government with having infringed its right of command.[9]
These documents illustrate an important but overlooked matter. Diplomatic intelligence illuminates the attitudes of states and their factions. It shows concealed levers and hidden hands and means to manipulate them: how to act or to signal. It gives good news and shows the limits to bad, increasing one’s certainty that unknown and unpleasant developments are not happening. Minor comments in a good source can determine the plausibility of major statements in an uncertain one. The reports of one party can reveal the intentions of another. So, too, knowledge may be useless. The point of diplomatic intelligence is to aid action, by shaping one’s policy or another’s. This is not easy. More often than not, intelligence provides first-rate information on third-rate issues, or knowledge which one cannot apply to policy. Actions, usually words of influence or threat, may be hard to deliver as desired or with effect, and have unintended or counter-productive consequences. Intelligence has greater value in cases of bargaining, when every party must act, and on the same issues. Then, the power of sources like journalists or officials rises, even more the ability to steal papers, bug offices or intercept signals. Meanwhile, information on the bargaining strategy and tactics of other players can help one take tricks, though their value depends on the stakes. Thus, at the Washington Conference of 1921-22, American code-breakers solved secret Japanese telegrams which outlined the lowest ratio it would accept in the strength of battleships and aircraft carriers compared to the American and British navies. This knowledge helped American negotiators force Japan to a 5 : 5 : 3 ratio, rather than the 5 : 5 : 3.5 ratio it wanted. Sailors on both sides thought that difference important in case of war. Yet the 5 : 5 : 3 ratio might have emerged anyway, as the United States wanted it and had a strong bargaining position, while Japan’s break point was guessed by The New York Times.[10]
A decade later, Britain had an analogous opportunity. As the London Naval Conference occurred at its capital, it had unmatched opportunities to pump the rumour mill, and to intercept traffic to and from the American, French, Italian and Japanese delegations. The GCCS, the best code-breaking agency of that era, mastered these messages, and the aims and means of other players. Britain could monitor any diplomacy between other parties by cross-checking reports from several sources, including the codes of one, usually more, participants. Yet intelligence matters only insofar as it aids action. Here, as often, it did not affect the formulation of policy but did guide the execution. The Prime Minister, Ramsay MacDonald, dominated British policy. He called the conference primarily to spur world disarmament, rather than to further narrow interests. When bargaining over the tonnage of warships, MacDonald aimed less to strengthen Britain’s position against the United States, than to change American attitudes, and gain their support for further moves toward liberal internationalism. To impress Washington, he accepted many of its demands, abandoning earlier British positions. In the long term, this crippled British seapower; immediately, it weakened Britain on disputes with the United States, and forced it into intricate negotiations over naval strength with all participants. Here, intelligence mattered. It was avidly used by the Admiralty, Foreign Office and Labour politicians. Whitehall and Washington hoped everyone would reason together, but also believed a naval treaty could be reached even if France and Italy refused to sign (as eventually happened). That was not true of Japan. Neither the United States nor Britain would accept a formal change in the 5 : 5 : 3 ratio, although American unwillingness to finance its navy really had reduced it almost to 5 : 4.5 : 3. Japan, however, demanded a ratio of 5 : 5 : 3.5 for cruisers and destroyers and large strength in submarines under any new treaty. Unless that conflict could be resolved, the conference would fail, making international relations worse than they would have been without it. Equally, compromise was possible. Britain worked for one, using the clash between Japan and the United States to play each against the other, and reduce both their demands.
Intelligence is about secrets. As the point was bargaining, everyone had to signal its position, while the Japanese, American and British civilian delegations were open with each other, in the freemasonry of liberalism. Yet each had a bottom line to hide and Britain had an edge. It had means to discover secrets, they did not. Before the conference, the GCCS provided the first, though predictable, news that Japan would demand a 70% ratio. It showed policy formation in Tokyo and Washington and the abortive discussions between them; it let politicians ensure British sailors were not working with Japanese ones to thwart arms control, although Labour distorted this danger. When the conference began, the GCCS did better than American code-breakers had done at Washington. It showed, usually in real time, all reports of the Japanese delegation, including material its naval and civilian sides hid from each other.[11] Britain knew their divisions as well as did their members. Against this, Japanese civilians did not hide the fact of these splits, which improved their bargaining position against other countries—thus, code-breaking reinforced Britain’s will to compromise. By 13 March 1930, after a month of secret talks, the British, American and Japanese delegations accepted a complex deal which let all claim victory. Delaying the construction of three out of 18 large American cruisers gave Japan almost a 70% ratio during the tenure of the treaty in practice, but not in theory, while containing construction on both sides. Britain’s chief negotiator, Robert Craigie, originated this alchemy, and secretly suggested the Japanese delegation fall back on it in case of deadlock with the Americans. This action stemmed from his knowledge of the issues, and intelligence on them. Press reports, statements by American and Japanese delegates, and especially solutions of their traffic, showed the details of their discussions, the internal politics of their policy, and that their governments wanted a deal. Above all, politicians in Tokyo told the civilian delegation that when they made the best settlement they could, the government would sell it at home, and make the admirals accept it. Craigie pressed the other sides to negotiate, fed them a solution which met their bottom lines while pushing their programmes down as Whitehall wanted, and rejected any alternatives that Britain disliked. This triumph of policy and intelligence overcame Britain’s bargaining weakness and met its aims over cruisers and world disarmament.
Then came the two telegrams under review. In one (N.C. 220), solved quickly, the delegation told Tokyo this deal was the best possible. The other message (N.C. 199), as often happened with telegrams from the Japanese Admiralty, the Kaigunsho, was read only a month after being dispatched. Had it been solved when sent, on 8 February, Whitehall would have been concerned and might have changed its actions. This message (like some others solved between 13-18 March, but even more so) showed the Kaigunsho insisted on the 70% ratio. Rather than accept anything less, it was preparing the ground to wreck the entire conference by working with another party. Whitehall feared France might wish to do the same. The GCCS closely followed all French and Japanese reports of their discussions, which showed they had common ground, while the French minister to Tokyo pressed Japan toward positions on submarines which might sink the conference. British and American diplomats soon queried Japanese and French officials on these matters, aiming to destroy any intrigue by showing it was known, though these actions posed small danger.
This material illuminated a major issue and in time to act, though that would have happened in any case. The Kaigunsho’s position was public knowledge four days after the GCCS detected it. The Foreign Office saw ‘a strong possibility of the naval element in the Japanese delegation overwhelming their political colleagues and preventing an agreement’, while ‘a most determined effort is being made by Japanese naval authorities to reject this compromise and nothing should be left undone to prevent such a disaster occurring’.[12] In hindsight, this assessment was alarmist; in practice, little could be done about it. As Britain opposed further concessions to Japan, it could neither appease sailors nor strengthen civilians. It could gain nothing by working on, or through, the delegation; useful action could be taken only at Tokyo. The Foreign Office thought the Kaigunsho had little power, while the Cabinet was strong and wise enough to accept the compromise and not subvert the conference. Still, the Foreign Office immediately reported the facts (though not the source) to American authorities, including the Secretary of State, Henry Stimson, and to its own ambassador in Tokyo, John Tilley. He should ‘keep in close touch’ with the Minister for Foreign Affairs, Shidehara Kijuro, and ‘use all your influence with the Japanese Government in favour of a reasonable settlement of this particular question at the earliest possible moment’. Tilley should act with the American ambassador, William Castle, while avoiding anything which looked like ‘joint or concerted pressure’. Notably, the men who decided how to act did not know first-hand the intelligence that inspired them, nor did one of their two governments. Tilley chose to work ‘almost entirely’ through Shidehara, refusing to contact the Prime Minister, Hamaguchi Osachi, as conversation with him ‘would be most unlikely to elicit any statement of opinion and would attract great and probably unfavourable notice in the press’. Nor would he use the naval attaché to contact the Kaigunsho, as ‘the real problem is a political one’ while the position of Shidehara, ‘who was disposed to defend the compromise even at the risk of having to resign, might be weakened if it were known, or believed, that a violent Anglo-American campaign was in progress and because alternately his own sympathy might be diminished if he learned that we were working by other means than through himself’. Tilley carefully avoided ‘too much trace of concerted action’ with Castle, but they compared notes and moved in parallel. Both contacted the Emperor’s political advisor, Saionji Kimmochi; both pressed Shidehara and asked him to deliver messages to Hamaguchi from MacDonald and Stimson in praise of the 13 March compromise; neither told the full story to his superior.[13] According to Japanese records, once he realized Shidehara’s resolve, Tilley told him to ‘just pigeon-hole the message’, as did Castle. They discarded what their superiors intended to be trump cards, so to show Shidehara how much they trusted him. He picked these cards up and played them in a game of his own, as he publicised that situation among leading liberals in Japan, presumably to show how he mastered problems with gaijin and Japanese.[14]
This intelligence affected British actions, but not events in Japan. Britain and the United States would have done much the same without it. Though it affected the form and timing of their approaches to Shidehara and Saionji, the latter would have acted as they did anyway. British actions merely reinforced Japanese decisions, nor were the results simple. In part because the role of code-breaking at the Washington conference became public knowledge, delegitimized that agreement, and left the competence and patriotism of Japanese liberals in doubt, the London Naval Treaty set the stage for a political explosion which started Japan down the road to the Pacific War.[15]
Andrew Webster (Murdoch University), ‘Conclusion’
As the commentaries by John Ferris and Peter Mauch make clear, these two documents represent useful examples of how signals intelligence has been intertwined with the events of a major international conference. Yet perhaps what is most significant about these documents is, ironically, the lack of attention that has yet been paid to them—and to signals intelligence in general. As Ferris notes, the American success at cracking Japanese codes during the Washington Naval Conference soon became public knowledge, with the publication in 1931 of a book by the chief American code-breaker, Herbert Yardley.[16] In addition, the several post-1945 investigations into the December 1941 catastrophe at Pearl Harbor produced widespread awareness of the success achieved by the United States in reading Japanese codes before and during the Second World War. Yet most histories of the Washington conference up to the present have either omitted any reference to the role played by signals intelligence in the conference’s outcome or accorded it only passing mention.[17] It is rather only the more specialised literature on intelligence that has given due consideration to the success of the American cryptographers.[18] Similarly, there has been almost no mention of the role played by the GCCS in any of the historical studies analysing the London Naval Conference. Those studies published before any documents pertaining to the work of the GCCS became publicly available are perhaps more understandably bare of consideration of signals intelligence. Nonetheless, parallels with the American experience at the Washington conference were there for those historians who wanted to look for them. That none did shows up the same analytical blind-spot towards signals intelligence as that contained in all studies of the Second World War before the revelation in 1974 of the Allied ability to read the German wartime codes, the so-called ‘Ultra secret’.[19] What is more noticeable is the lack of any analysis in recent studies of the London conference, since evidence has become available. Initially this evidence came in limited and indirect fashion. In 1986, for example, the very first issue of the journal Intelligence and National Security reprinted in full a highly revealing memoir of the work of GCCS between the wars, written in 1944 by Alastair Denniston, who had been the operational head of GCCS.[20] In it, he noted many of the code-breaking triumphs of the service, particularly those against Japan, and recalled in one particularly revealing line that ‘throughout the period down to 1931 no big conference was held in Washington, London or Geneva in which [GCCS] did not contribute all the views of the Japanese government and of their too verbose representatives’.[21] Such a pointed reference to the cities that hosted the three great naval conferences of the interwar period can hardly have been accidental. Specialised studies of British intelligence have indeed noted the successes of the GCCS, including those against Japanese codes during the 1920s and early 1930s. These have been particularly explicit and detailed since the late 1990s, when the material in the HW 12 series first became open for consultation.[22] Overall such revelations, it might be thought, would have prompted historians subsequently to have at least raised the question of the possible influence of signals intelligence on British conference diplomacy in London and elsewhere, but this has not been the case.[23]
How then should historians evaluate the role of British code-breaking in its policy-making, and in particular the successes against the Japanese? The great mass of intercepted and decrypted foreign diplomatic telegrams now available in the HW 12 series are unquestionably of great importance. Antony Best, for example, has recently demonstrated the fruitful results possible from a careful use of the GCCS intercepts, concluding that ‘in the field of British policy towards Japan in the interwar period it is not an exaggeration to state that no authoritative military, diplomatic or even economic history can be written in future without extensive use of the HW 12 files’.[24] But it is not a simple matter to put these files to use. The HW12 series is in many ways so large, and so undefined, that it is difficult to make effective use of it. The data is in the rawest of forms: the intercepts alone, organised solely by date. There is no indication of how they were interpreted, for the comments made on the documents by officials in the Foreign Office have been destroyed, apparently incinerated as a matter of security procedure soon after being compiled.[25] Nor is it clear who even saw this intelligence. Some of the intercepts from the naval conference carry indications as to their distribution—to the director of GCCS (Denniston), the Foreign Office (five copies) and the Admiralty—but such a list of recipients is neither surprising nor illuminating. Scattered evidence from the private diaries of some politicians and officials, and the occasional cryptic reference to secret sources in the official files, does however make it clear that during the time of the London conference the GCCS intercepts were given at least a limited circulated within Whitehall.[26] Stripped of their context, the problem is how to fit these intercepts into the web of other available documents on British policy-making. Consequently, a detailed knowledge of the diplomatic context is essential to making use of them in any analysis of British policy. They are not able to provide some kind of new ‘complete’ picture of events, but they can suggest new ways to view events and can illuminate some obscure references in private papers. MacDonald’s well-known fulminations in his diary against the French during the London naval conference—‘They have been conspiring against us!’—only become fully comprehensible when it is realised that GCCS intercepts had shown him the extent to which the French delegation were secretly attempting to press Japan to refuse any compromise deal that heavily restricted submarines.[27] John Ferris has also warned elsewhere against the dangers of taking the GCCS decrypts during the conference strictly at face value: for example, out of his key concern to strengthen Anglo-American relations, MacDonald warned a member of the American delegation not to report anything significant in State Department codes.[28]
At first sight, it would appear that the
two documents printed here exposed to the British, first of all, the firmness
of the early Japanese negotiating position (N.C. 199), and then later the point
at which the negotiators believed the time had come to back down and accept the
deal on offer (N.C. 220). However, as Mauch and Ferris point out, things were
not so simple. In the first place, the Japanese policymaking establishment was
not monolithic, and it is important to distinguish between documents sent to or
received from the Admiralty and the Foreign Ministry. Secondly, the gap between
a telegram’s transmission and its decryption was critical: until a document can
be read, it cannot be put to use. Thus, N.C. 220 from the Japanese
delegation in London to the Foreign Minister in Tokyo was sent using ‘Cypher
Class A’ and took only five days for GCCS to decrypt. In contrast, it took GCCS
some five weeks to decrypt N.C. 199 from the Japanese Admiralty to the
delegation in London, indicating the use of a much stronger form of cipher, and
meaning a quite different negotiating situation had developed by the time it
arrived on the desks of British policy-makers. Third, it seems clear that the British naval negotiations, and the
wider diplomacy surrounding them, in many ways proceeded independently of the
input from signals intelligence. What this signals intelligence
provided, in the context of the ‘conference diplomacy’ of the London naval
conference, was a degree of confidence and confirmation for the British
representatives in their ability to secure a deal with the Japanese. While one
must avoid automatically equating ‘most secret’ documents with being ‘most
important’, documents such as these should, at the least, require histories of the
naval conferences to take into account the ability of some host nations to play
‘dirty poker’ in the complex game of diplomatic negotiation.
[1] Malcolm H. Murfett,
‘Look back in anger: The western powers and the Washington Conference of
1921-1922’, in B.J.C. McKercher (ed.), Arms
Limitation and Disarmament: Restraints on War, 1899-1939 (Westport and
London: Praeger, 1992), p. 90.
[2] John Ferris, ‘The road
to Bletchley Park: The British experience with signals intelligence,
1892-1945’, Intelligence and National
Security, 17, 1 (Spring 2002), p. 67. On the GCCS, see also: Christopher
Andrew, Secret Service: The making of the
British intelligence community (London: Heineman, 1985), pp. 375-376,
499-501; Christopher Andrew, ‘Secret intelligence and British foreign policy,
1900-1939’, in Christopher Andrew and Jeremy Noakes (eds.), Intelligence and International Relations,
1900-1945 (Exeter: University of Exeter, 1987), pp. 16-22; F.H. Hinsley, British Intelligence in the Second World War
(abridged edition, London: HMSO, 1993), pp. 5-7; John Ferris, ‘Whitehall’s
Black Chamber: British cryptology and the Government Code and Cypher School,
1919-1929’, Intelligence and National
Security, 2, 1 (January 1987), pp. 54-91.
[3] The National Archives
(TNA), Kew, HW 12/114-150.
[4] For the text of the
Washington Naval Treaty (6 February 1922), see: Foreign Relations of the United States, 1922 (Washington:
Government Printing Office, 1938), vol. 1, pp. 247-266. France and Italy had
each been accorded ratios of 1.75.
[5] On the internal debates
surrounding Japanese naval policy, see in particular the studies by Sadao
Asada: ‘The Japanese navy and the United States’, in Dorothy Borg and Shumpei
Okamoto (eds.), Pearl Harbor as History:
Japanese-American Relations, 1931–1941 (New York and London: Columbia
University Press, 1973), pp. 225–59; ‘Japanese admirals and the politics of
naval limitation: Kato Tamosaburo and Kato Kanji’, in Gerald Jordan (ed.), Naval Warfare in the Twentieth Century,
1900–1945: Essays in honour of Arthur Marder (London: Croom Helm, 1977),
pp. 141–166; ‘From Washington to London: The Imperial Japanese Navy and the
politics of naval limitation, 1921–1930’, in Erik Goldstein and John Maurer
(eds.), The Washington Conference,
1921–22: Naval Rivalry, East Asian Stability and the Road to Pearl Harbor
(Ilford: Frank Cass, 1994), pp. 147–191.
[6] B.J.C. McKercher, Transition of Power: Britain’s Loss of
Global Pre-eminence to the United States, 1930-1945 (Cambridge University
Press: Cambridge, 1999), pp. 50-58.
[7] For the text of the
London Naval Treaty (22 April 1930), see: Documents
on British Foreign Policy, 1919-1939, Second Series, vol. I (London: HMSO,
1946), appendix I.
[8] Grand Chamberlain
Admiral Suzuki Kantarō, quoted in diary entry for 28 March 1930, Thomas
Francis Mayer-Oakes (ed.), Fragile
Victory: Prince Saionji and the 1930 London Treaty Issue (Detroit: Wayne
State University Press, 1968), p. 103.
[9] On these events, see: Sadao
Asada, From Mahan to Pearl Harbor: The
Imperial Japanese Navy and the United States (forthcoming); Asada, ‘From
Washington to London’, pp. 172–82; David C. Evans and Mark R. Peattie, Kaigun: Strategy, Tactics, and Technology in
the Imperial Japanese Navy, 1887-1941 (Annapolis: Naval Institute Press,
1997); Takashi Itō, Shōwa Shoki
Seijishi Kenkyū: Rondon Kaigun Gunshuku Mondai o Meguru Shoseiji Shūdan
no Taikō to Teikei, (Tokyo: Tokyo Daigaku Shuppankai, 1969); Tatsuo Kobayashi,
‘The London Naval Treaty’, in James William Morley (ed.), Japan Erupts: The London Naval Conference and the Manchurian Incident,
1928-1932, selected translations from Taiheiyō Sensō e no Michi,
(New York: Columbia University Press, 1984); Minoru Nomura, Nihon Kaigun no Rekishi, (Tokyo:
Yoshikawa Kōbunkan, 2002).
[10] An excellent account is
in David Kahn, The Reader of Gentlemen’s
Mail: Herbert O. Yardley and the Birth of American Codebreaking (New Haven:
Yale University Press, 2004)
[11] Sadao Asada discusses
in some detail the telegraphic confusion within the Japanese delegation, as its
civilian and naval members attempted to keep their secrets from each other.
Asada, ‘From Washington to London’, p. 177.
[12] Minute by Thompson, 25
April 1930, and Tokyo embassy to Foreign Office, dispatch no. 145, 25 March
1930, FO 371/14263, A2796; Documents on
British Foreign Policy, 1919-1939, Second Series, vol. I (London: HMSO,
1946), pp. 249-51.
[13] Documents on British Foreign Policy, 1919-1939, Second Series, vol.
I (London: HMSO, 1946), pp. 249-66; Foreign
Relations of the United States, 1930, vol. I (Washington: Government
Printing Office, 1945), pp. 66-79.
[14] Mayer-Oakes, Fragile Victory, p. 100.
[15] For these events as a
whole, see also: Ian Gow, Military
Intervention in Prewar Japanese Politics: Admiral Kato Kanji and the
‘Washington System’ (London: Routledge, 2004); Morley, Japan Erupts; John Ferris, ‘“It is our business in the navy to
command the seas”: The last decade of British maritime supremacy, 1919-1929’,
in Greg Kennedy and Keith Neilson (eds.), Far
Flung Lines: Essays on imperial defence in honour of Donald Mackenzie Schurman
(London: Frank Cass, 1997). Solutions referred to in this piece are in TNA, HW
12/119-125.
[16] Herbert Yardley, The American Black Chamber
(Indianapolis, Bobbs-Merrill, 1931).
[17] Gerald E. Wheeler, Prelude to Pearl Harbor: The United States
Navy and the Far East, 1921–1931 (Columbia: University of Missouri Press,
1963), pp. 53–58; Stephen Roskill, Naval
Policy Between the Wars. Vol. I: The Period of Anglo-American Antagonism,
1919–1929 (London: Collins, 1968), pp. 300–330; Thomas H. Buckley, The United States and the Washington
Conference, 1921–1922 (Knoxville: University of Tennessee Press, 1970), p.
196; William R. Braisted, The United
States Navy in the Pacific, 1909–1922 (Austin and London: University of
Texas Press, 1971), pp. 599–602; Roger Dingman, Power in the Pacific: The Origins of Naval Arms Limitation, 1914–1922
(Chicago and London: University of Chicago Press, 1976), p. 203; Christopher
Hall, Britain, America and Arms Control,
1921–1937 (Basingstoke: Macmillan, 1987), pp. 24–35; Robert G. Kaufman, Arms Control during the Pre-Nuclear Era: The
United States and Naval Limitation between the Two World Wars (New York:
Columbia University Press, 1990), pp. 43–72; Emily O. Goldman, Sunken Treaties: Naval Arms Control between
the Wars (University Park: Pennsylvania State University Press, 1994), pp.
121–130; Richard W. Fanning, Peace and
Disarmament: Naval Rivalry and Arms Control, 1922-1933 (Lexington:
University of Kentucky Press, 1995), pp. 4-8; Phillips P. O’Brien, British and American Naval Power: Politics
and Policy, 1900-1936 (Westport and London: Praeger, 1998), pp. 166-169.
[18] In particular, the
path-breaking study by David Kahn, The
Code-breakers: The Story of Secret Writing (1967; revised edition, New
York: Scribner, 1996), pp. 356-359, specifically highlighted this aspect of the
Washington conference. See also: David Alvarez, Secret Messages: Codebreaking and American Diplomacy, 1930-1945
(Lawrence: University Press of Kansas, 2000), pp. 8-16.
[19] On the
historiographical revolution prompted by the revelation of Ultra, see: Russell
M. Anderson, ‘The public disclosure of Anglo-American signals intelligence
since the Second World War, with particular reference to Ultra and Magic’,
unpublished Ph.D. dissertation (Cambridge, 2004). For examples of this same
blind-spot towards signals intelligence in studies of the post-1945 period,
see: Christopher Andrew, ‘Intelligence and international relations in the early
Cold War’, Review of International
Studies, 24, 3 (July 1998), pp. 321–330.
[20] A.G. Denniston, ‘The
Government Code and Cypher School between the wars’, Intelligence and National Security, 1, 1 (January 1986), pp. ??. The original copy of this
memoir, dated 2 December 1944, is in the Denniston papers, Churchill College,
Cambridge, DENN 1/4.
[21] Denniston, ‘Government
Code and Cypher School’, p. ?.
[22] Michael Smith, The Emperor’s Codes: Bletchley Park and the
Breaking of Japan’s Secret Ciphers (Bantam Press: London, 2000), pp. 4-5,
17-18, 20-41. Also see works cited in introduction (note 2) above.
[23] Raymond G. O’Connor, Perilous Equilibrium: The United States and
the London Naval Conference of 1930 (Lawrence: University of Kansas Press,
1962), pp. 62–108; Stephen Roskill, Naval
Policy Between the Wars. Vol. II: The Period of Reluctant Rearmament, 1930–1939
(London: Collins, 1976), pp. 37–70; Hall, Britain,
America and Arms Control, pp. 88–109; Kaufman, Arms Control in the Pre-Nuclear Era, pp. 129–138; Goldman, Sunken Treaties, pp. 14, 148, 151,
190–1, 204–9; Fanning, Peace and Disarmament,
pp. 126-128; O’Brien, British and
American Naval Power, pp. 210-218; McKercher, Transition of Power, pp. 50-58.
[24] Antony Best, ‘Intelligence,
diplomacy and the Japanese threat to British interests, 1914-1941’, Intelligence and National Security, 17,
1 (spring 2002), p. 98. See in particular his discussion of the Leith-Ross
mission to the Far East in 1935 (pp. 89-92).
[25] Best, ‘Intelligence,
diplomacy and the Japanese threat’, p. 89.
[26] MacDonald diaries, TNA,
PRO 30/69/1753/1, entries for 18, 24 and 25 March 1930; Ben Pimlott (ed.), The Political Diary of Hugh Dalton, 1918-40,
1945-60 (London: Jonathan Cape, 1986), pp. 69-72, 80-81. See also: John
Ferris, “‘Indulged in all too little’?: Vansittart, intelligence and
appeasement,” Diplomacy and Statecraft,
6, 1 (March 1995), pp. 124-128.
[27] MacDonald diary, 18
March 1930, TNA, PRO 30/69/1753/1; Dobler (Tokyo) to Quai, 17 March 1930,
Ministère des Affaires Etrangères, series SDN, vol. 1118, fos. 41–4; intercepts
in TNA, HW 12/126. A complete summary of all the French manoeuvres, based upon
‘secret sources’, is in Foreign Office memo, 24 March 1930, TNA, FO 371/14261,
A2166.
[28] Ferris, ‘Road to Bletchley
Park’, p. 68.