Royal Navy Readiness for a War with Japan in Mid-1941: Intelligence and Capability I

The defence of the Indian Ocean became an increasingly important factor in British strategic calculations during the first half of 1941. They also showed how American determination to prioritise the Atlantic over the Pacific obliged the Royal Navy to resurrect the concept of an eastern fleet to ensure adequate security against Japanese intervention in the war. For a number of reasons, the mid-point of the year marked a watershed for both Britain and the United States in their perception of the risk from Japan, and the start of a series of moves on both sides that led inexorably to confrontation at the end of the year. The Royal Navy’s intelligence picture of the naval risk posed by Japan in mid-1941, as the prospects of her intervention increased, and as it became both possible and necessary to contemplate deploying significant naval reinforcements to the East. It also looks at the overall resources and capabilities available to the Royal Navy at this time to secure Britain’s vital interests across the wider eastern empire and to provide any necessary reinforcements.

The Royal Navy picture of the IJN in 1941

For most of the interwar period, the Royal Navy viewed the IJN as both its most likely opponent in a future war, and the only naval power that posed a significant threat. In terms of size, the IJN ranked third after the Royal Navy and US Navy, its fleet spanned the full range of naval capability, it had established a fine tradition, and was recognised as an innovator. It was by any measure a powerful force, and therefore a key standard against which the Royal Navy judged its strength and capability. The strength of the IJN, current and projected, was not only a key determinant of the Royal Navy rearmament programme from 1935 onward, but had also shaped Royal Navy programmes in the 1920s and early 1930s, especially for cruisers and submarines. The specifications of the County-class cruisers laid down from 1924 were largely dictated by IJN cruiser plans following the 1922 Washington Naval Treaty. The Southampton class laid down from 1934 were strongly influenced by the Mogami class. The ‘O’-, ‘P’- and ‘R’-class submarines, laid down between 1924–29, were specifically designed to hold off an IJN fleet in the South China Sea, pending the arrival of the main Royal Navy fleet. Until 1940, defence of the Far East was seen as pre-eminently a naval problem, and the Admiralty dominated Far East defence policy and planning. Before 1937, this policy and planning had a theoretical quality, but Japan’s invasion of China increased the risk of confrontation, and the prospect of an attack on British possessions. The 1937 imperial review and Far East appreciation represented a watershed in judging those risks. For all these reasons, the Royal Navy studied the IJN closely, drawing on both overt and secret intelligence sources.

The Royal Navy’s intelligence picture of the IJN has received contrasting reviews. One influential interpretation is that poor intelligence and wishful thinking allowed the Royal Navy to create an image of the IJN that made it less threatening, and hence easier to manage. As the Royal Navy increasingly struggled in the late 1930s to provide a fleet to match the IJN, so it sought convenient evidence that Japan could neither sustain nor make effective use of a superior fleet. This led it to set IJN efficiency at 80 per cent of the Royal Navy, to underrate Japanese technical and industrial capacity and to believe that the IJN would be a cautious opponent. These factors encouraged a disastrous reliance on inadequate forces, culminating in the Force Z deployment at the outset of war. This interpretation focuses too much on what was not known, rather than emphasising what was known and then concentrating on key gaps. It also fails adequately to distinguish between what was knowable and what was not knowable. It was not possible to predict the timing of a Japanese attack on Malaya with any certainty before autumn 1941, because no firm plan existed before then. Nor could the Royal Navy have predicted prior to 1941 that the IJN would group all its fleet carriers in a single task force to deliver strategic effect, because the IJN did not adopt this concept until early that year. Finally, it does not place intelligence within the wider context of political and military risk management. Perceived intelligence weakness must be weighed against the fact that British assessments of the forces the Japanese would deploy against Malaya were consistently good.

An alternative view of British intelligence performance in the run-up to war concedes that the Royal Navy’s assessment of the IJN was mediocre, but argues the failings were compensated by system and circumstances. Royal Navy strategy, based on ‘bean-counting’ of warships and a rejection of best-case planning, was not distorted by any underestimation of Japanese quality, nor was the latter responsible for the loss of Force Z. The Royal Navy may have failed sufficiently to recognise IJN technical achievements and innovation in tactical thought, especially in the use of naval air power, through the late 1930s. However, this was not in itself responsible for Britain’s inadequate dispositions in the Far East in December 1941, which were more the result of circumstances and the demands of grand strategy. Royal Navy intelligence failings would not, therefore, have been significant, had Britain been able to despatch sufficient naval and air forces to the East to implement a defensive strategy. Proponents of this view have also, rightly, stressed that in planning for war with Japan, the Admiralty always wanted its naval forces to be larger than they were, always preferred numerical superiority over the IJN, and calculated the balance of forces on the basis that one Japanese warship equalled one British warship of equivalent size.

In terms of unit numbers of IJN warships, Royal Navy ‘bean-counting’, or basic assessment of IJN order of battle, in mid-1941 was fairly accurate, though not perfect. An assessment prepared by the Naval Intelligence Division (NID) section responsible for Japan, NID 4, for the prime minister in late August 1941 gave an authoritative statement of overall IJN strength. This got most numbers correct. It overstated the number of operational destroyers (126 compared with an actual strength of 112) and submarines (eighty compared to sixty-five), but these errors were not significant and may be explained by the inclusion of units in reserve. However, the assessment also contained important mistakes in displacement and armament, which represented a more serious underestimate of actual IJN fighting capability, especially air power. It correctly identified eight aircraft carriers, but underestimated the displacement of the fleet carriers Hiryu and Soryu and the light carrier Ryujo by about 40 per cent. Overall aircraft capacity of the eight carriers was put at 362 aircraft, compared with an official IJN figure for embarked strength at this time of 473, an underestimate of about a quarter. Six cruisers, the four Mogami class and two Tone class, were mistakenly listed with 6in guns, whereas in reality they were all by this time armed with 8in. (The four Mogami class were initially fitted with fifteen 6in guns, since Japan had reached its London Naval Treaty limit on 8in cruisers, but they were designed to be upgraded, and this was done secretly in the late 1930s. Both the Royal Navy and US Navy missed this. The Tone class were built as 8in cruisers from the start, but NID 4 assumed they would repeat the 6in pattern of the Mogamis. NID 4 did not correct this error, shared by the Americans, until after the Battle of Midway when photographs of Mikuma and Mogami showed their main armament was ten 8in guns. NID 4 then concluded, rightly, that they had always been designed for upgrading to 8in, and assumed the two Tones would now be similarly armed. Finally, the assessment included a ‘heavy armoured cruiser’ of 14,000 tons, armed with six 12in guns, on trials which did not exist.

Estimates for the overall strength of the IJNAF produced during 1941 were also broadly accurate. In May that year the Joint Intelligence Committee assessed first-line IJNAF strength at 1496 aircraft, including 330 ship-borne. The figure for ship-borne air strength of 330 aligned closely with the subsequent NID 4 estimate for the prime minister in August. Given that the total included seaplanes and catapult reconnaissance aircraft, it underestimated carrier-borne strength, even allowing for the fact that there were only six carriers in commission at this time. (The large fleet carriers, Shokaku and Zuikaku, were still in build.) Meanwhile, the Joint Intelligence Committee assessed Imperial Japanese Army Air Force (IJAAF) strength at 886, giving a total Japanese frontline strength of 2384. NID subsequently circulated a figure of 1600 for IJNAF first-line strength in August and a later Air Ministry Weekly Intelligence Survey in December 1941 quoted a total IJNAF strength of 1561 with 547 ship-borne. The latter report set IJAAF strength at 1108, giving an overall Japanese frontline in December of 2669. By then there were ten IJN carriers in commission, so deployed carrier air strength had grown accordingly from the May estimates.

These estimates during 1941 compare with an official Japanese figure for frontline IJNAF strength on the outbreak of war of 831 shore-based and 646 ship-borne aircraft, a total of 1477. British estimates for ship-borne aircraft were fairly accurate for seaplanes and floatplanes, but underestimated the carrier total by about 25 per cent (338 in December, as opposed to a real total of 473). The estimates for IJNAF land-based bomber strength were also in the right order, but there was otherwise no detailed breakdown of either land-based or ship-borne strength into either category, ie fighters, torpedo bombers, heavy bombers, etc, or specific aircraft type such as A6M Zero, B5N2 torpedo bomber, or Type 1 bomber, etc. Alongside the above estimates of frontline strength, evidence from early 1942 suggests that British intelligence departments had a surprisingly accurate view on the size of the aircraft reserve held by the Japanese at the start of the war. Drawing on Joint Intelligence Committee advice, Ismay informed the prime minister in early March that the immediate reserve at the start of hostilities for the two Japanese air forces was 650 aircraft. Japanese figures show that the actual total for the IJNAF operational reserve was 309.20 It is unlikely the IJAAF reserve was larger, so a combined estimate of 650 was about right.

The Joint Intelligence Committee report of May 1941 did not expect Japanese air strength to grow significantly over the next two years, judging that production only balanced wastage and re-equipment. Wastage calculations here, of course, took account of ongoing operational losses in China. For the IJNAF, at least to end March 1942, this assessment was essentially correct. In the summer of 1941, production of naval aircraft of all types was only 162 aircraft per month and was only 179 per month over the first four months of war. Production was especially slow for the more modern combat aircraft, and this was compounded by the paucity of reserves noted above, which represented barely 25 per cent or less of frontline strength. There were only 415 Zero fighters available to the frontline at the outbreak of war and, as will be demonstrated later, carrier attack aircraft availability was depleting fast well before the Battle of Midway in June 1942. Overall, these estimates for immediate pre-war Japanese air strength, frontline, reserves and production capacity were significantly better than the comparable estimates for pre-war German air strength, and far superior to the estimates of German air strength and production during 1940 and the first half of 1941. In 1939 German air force frontline strength was overestimated by 18.5 per cent, while 1940 German aircraft production was overestimated by a minimum of 23 per cent, and at one point by about 75 per cent.

British military leaders have attracted harsh criticism for their lack of knowledge of Japanese aircraft performance. Although surviving records here are patchy, accurate performance tables for all Japanese aircraft in use at the outbreak of war had been circulated across the British intelligence community by mid-1941. The exceptional range of many Japanese naval aircraft, including the new Zero fighter, was identified in these tables. British intelligence also had a reasonably accurate picture of IJNAF aircraft armament. The main types of bomb carried were known, as were details of the Type 91 aerial torpedo. The capability of the Type 91 has often been exaggerated, compared to the standard Royal Navy 18in Mark XII and XIV Fleet Air Arm torpedoes. In fact, specifications, performance and reliability were close. The Type 91 could, in theory, be dropped from up to 500m (1640ft) altitude, but authoritative Japanese sources state that despite the optimistic claims of the ordnance department, only 10 per cent would function correctly at 200m (660ft) and 50 per cent at 100m. In an operational context, the preferred standard drop was from a range of 400–600m (1300–2000ft) at 160–170 knots at a height of 30–50m (100–164ft). This was little different from the operational performance envelope of the Fleet Air Arm Albacore, although the latter was somewhat slower. This intelligence on aircraft and armament had been absorbed within Far East air headquarters by early autumn 1941 at the latest.

As regards Japanese air deployment, an August NID report identified eight Japanese airfields in Indochina, along with details of the upgrades underway at Saigon and Bien Hoa, which would both be used by IJNAAF aircraft that December. Significantly, this report also included a chart showing the operational radius for heavy bombers deploying from Saigon and Soctrang, demonstrating that Singapore and the Malacca Strait would be in range of aircraft with a normal bomb load. A subsequent report provided details of the Type 96 IJNAF heavy bomber, correctly estimating it could carry 2200lbs (1000kg) to 950 miles. This report admittedly doubted Japan could afford to deploy more than 150 land-based aircraft for potential operations against Malaya and the Netherlands East Indies at this time, but this figure would be revised sharply upwards over the next three months. In judging operational effectiveness, Far East air headquarters drew a distinction between the IJNAF, whose quality was rated high, and the IJAAF, who were rated mediocre. This assessment would have been shared with the Far East Combined Bureau (FECB), whose role is described shortly, and the China Fleet.

The Royal Navy’s primary measure of its strength against the IJN was the number and quality of capital ships. Here it was certainly not complacent. In March 1940 Churchill, then First Lord, sent a paper comparing Royal Navy and IJN capital ships to the War Cabinet. It pressed the case to resume building capital ships, suspended at the start of the war, in order to ensure adequate capability against Japan from 1942 onward. However, the paper is important for what it says about the existing balance. It stated that the IJN battle-fleet, which had been completely modernised, had a speed at least equivalent to the Queen Elizabeth class and considerably greater than the ‘R’ class. Japanese battleships also outranged the unmodernised Royal Navy ships. The paper stated correctly that all ten IJN capital ships had received modern, more efficient boilers and engines, new main armament guns of increased elevation, probably to 40 degrees, additional armour, including horizontal protection, and new anti-aircraft armament. It would be unsatisfactory to send any Royal Navy ships to the East unless they matched IJN ships for both speed and range. Only seven of the existing fleet did so. These seven comprised the four Royal Navy reconstructions undertaken between 1934 and 1940, namely the three battleships, Warspite, Valiant and Queen Elizabeth, and the battlecruiser Renown, along with the two Nelsons and Hood. The four modernisations took broadly the same form as the IJN applied to all ten ships, although there were also two partial modernisations. The paper noted that although the new King George V class would broadly match anticipated new IJN ships in numbers, they would be out-gunned. In addition, Japan was believed to have four 12in gunned battlecruisers in build.

The September 1940 Future Strategy paper confirmed this assessment. It stated that Japan’s modernised ships were ‘superior to our own’, and that the new battlecruisers could only be countered by a King George V or Hood or Renown. NID 4 reminded Churchill of the IJN capital ship modernisation programme in August 1941, as he was beginning the debate with Pound over Far East naval reinforcement. In March 1942 the naval staff stated that there was no recent information on IJN gunnery and RDF (radar), but there was no reason why the IJN should not be ‘considerably superior in Long Range Direct Fire and Medium Range Blind Fire’. The IJN had, in fact, put much effort into long-range gunnery with significant results in trials, but this was based on traditional techniques of visual ranging. They never produced effective firecontrol radar, which the Royal Navy had in most capital ships by this time. Royal Navy anxiety over IJN capital ship superiority persisted through most of the war. As late as 1944, a paper by the Director of Naval Intelligence (DNI) stated that the four older IJN battleships of the Fuso and Ise classes were superior to the modernised Queen Elizabeths while the Kongo-class battlecruisers were superior to the modernised Renown, although it recognised Royal Navy radar could, in practice, make a decisive difference. This assessment was undoubtedly unfair to both the modernised Queen Elizabeths and Renown, but it reflected an enduring Admiralty view. As regards gunnery performance, it is unlikely there was much difference between the two navies in long-range fire under good daytime conditions, but radar gave the Royal Navy a major advantage at night, or in bad weather.

An effective assessment of the Japanese naval risk to British interests could not rest solely on bean-counting. It required knowledge of IJN organisation, the location and movements of its key forces, their fighting effectiveness and, above all, timely warning of hostile intent. Here, despite limited resources, the intelligence capability available to NID by mid-1941 was good enough to enable the Admiralty and local Royal Navy commanders to reach reasonable conclusions on the scale, quality and timing of an attack. To provide the Far East intelligence picture, NID relied primarily on the Far East Combined Bureau, a joint intelligence organisation set up in 1936, combining personnel from all three services to support Far East commanders. The FECB was a visionary concept for its day, and represented what in modern parlance would be described as an all-source ‘fusion centre’. Its primary roles were to give early warning of Japanese attack, build an accurate Japanese order of battle, naval, land and air, and provide operational intelligence once war had broken out. It also provided some set-piece assessments of Japanese capability, and occasional strategic overviews. Although the FECB was a tri-service organisation, because of the Royal Navy’s traditional pre-eminence in Far East intelligence, it was accountable to and administered by Commander-in-Chief China, not Commander-in-Chief Far East, and commanded by a Royal Navy captain. In the second half of 1941 this was Captain Kenneth Harkness. He was a gunnery officer, not an intelligence specialist, but capable, with a strong operational record and a good administrator. FECB, therefore, had a distinct naval bias. It devoted more effort to naval requirements and was more effective here than it was for the other two services.

Royal Navy Readiness for a War with Japan in Mid-1941: Intelligence and Capability II

Importantly, FECB collaborated closely with NID with inevitable advantages here for intelligence distribution and exploitation. Patrick Beesly, who worked for NID in the war, described FECB as a ‘well thought out and comprehensive organisation, the first example of an inter-service effort in the intelligence field, and for NID its most important outstation’. When it moved to Singapore from Hong Kong in 1939, FECB acquired its own Naval Operational Intelligence Centre (NOIC), modelled on that at the Admiralty, the Pacific Naval Intelligence Organisation, which provided a plot of all Axis naval activity from East Africa to the west coast of the Americas. In his 1939 command review, Commanderin-Chief China, Vice Admiral Sir Percy Noble, stated that FECB had ensured he knew with complete confidence the whereabouts of all elements of the Japanese combined fleet on the outbreak of the European war. It also located the German pocket battleship Graf Spee during her brief foray into the Indian Ocean through long-range direction-finding (D/F) fixes. In covering the IJN, FECB drew together intelligence from signals intelligence (SIGINT), human agents, aerial reconnaissance, submarine surveillance, diplomatic reporting and open sources.

Most important of all, Japan lacked the merchant shipping resources necessary to sustain its economy during a war with the western powers and to reap the benefit of its conquests, especially in regard to oil. Furthermore, the IJN had neither thought about trade protection, nor devoted any effort to procuring the relevant capabilities, drawing on Royal Navy lessons in the Atlantic. When it began the war, the IJN had just four dedicated anti-submarine escorts, and had no underwater detection equipment available at sea until August 1942. At that time, the Royal Navy was deploying 2100 sonar equipped ships. IJN doctrine was narrowly focused on the concept of a decisive battle with the enemy fleet. Here it was ‘ingenious, imaginative and beautifully crafted’. But it neglected the wider aspects of maritime power on which Japan depended, and underestimated the limitations of Japan’s economic and industrial capacity which threatened rapid obsolescence.

Some qualifications are nevertheless required. Japan commissioned more carriers than either Britain or the United States in 1942, almost compensating for its heavy losses that year, and it out-built Britain in this category over the three fiscal years 1942–45. It also tripled aircraft production across these years and introduced some competitive new designs. In some respects its wartime aircraft industry outperformed that of the Soviet Union. During the war years from 1942 to 1945, Japan also produced 3,392,814 tons of merchant shipping, including 986,159 tons of oil tankers. This was very close to British output over this period, despite the huge disruption from bombing Japan suffered from mid-1944. Without bombing, Japan might have out-produced Britain in merchant shipping.

When due account is taken of all these factors, the evidence that the Royal Navy formally and consistently marked the IJN down on fighting efficiency is sparse. Three items are invariably highlighted: a 1935 report by the Tokyo naval attaché, Captain J P G Vivian; the 1939 Tientsin papers, which rated the fighting efficiency of the IJN at 80 per cent of the Royal Navy; and the joint planning staff and Joint Intelligence Committee assessments during 1941 that Japanese air forces should be rated on a par with the Italians. These items must be placed in context. The Vivian report attempted to identify traits in Japanese national characteristics and culture which might limit its efficiency. It undoubtedly attracted high-level interest in the naval staff when it appeared, but it has probably received more weight than it merits, simply because it has survived in the files, whereas alternative views have not. There is no evidence it had lasting impact, or ever represented a consensus. The 1937 Far East appreciation, two years later, assumed that, for planning purposes, the IJN had equal capability to the Royal Navy. Indeed, it specifically proposed the Royal Navy should aim for a capital ship advantage of twelve to nine.

The 1939 80 per cent rating appears to have persisted as a rough gauge of IJN quality over the next two years, but there is no sign this had any practical effect, either. It is also important to underline that the original reference was to maintenance, not to skill and commitment in battle. The IJN discovered serious design flaws in several classes of ships in the mid-1930s, notably the Mogami-and Takao-class cruisers and Fubuki-class destroyers. Correcting these involved major remedial work, with ships out of commission for long periods. It is likely the Royal Navy was aware of this and it may have contributed, therefore, to an 80 per cent availability factor.

Admiral Sir William Davis, a captain serving as Deputy Director Naval Operations (Foreign) in 1941, later said:

Many of us (on the Naval Staff) would not agree with the estimate that the Japanese were only 80 per cent efficient compared with ourselves. We thought they were tactically rather rigid and also behind us in anti-submarine tactics. But frankly we knew nothing else (about their efficiency) and many of us thought it best to over-estimate rather than underestimate.

That view rings true, and fits with the cautious approach the Royal Navy took on comparative battleship performance, as well as the experience gained in submarine surveillance operations. There was also some justification for questioning IJN fighting efficiency in terms of their inability to provide the logistic and personnel support necessary for a sustained war, and the Royal Navy may have acquired at least anecdotal evidence of this. In that specific sense, the concept of an efficiency correction was defensible.

Ranking the Japanese air forces with the Italians was potentially more serious, if it implied they were an inferior opponent. It was one excuse for the chiefs of staff limiting Far East air reinforcement in terms of both numbers and quality, although it is harder to argue it was a decisive factor, given the pressures to prioritise the Middle East. It is not clear from the joint planning staff and Joint Intelligence Committee papers where the analogy with the Italians originated. It seems likely it was a Royal Air Force, more than a Royal Navy, judgement and, if so, it is difficult to judge how much it influenced Royal Navy attitudes. It may at the least have encouraged the Royal Navy to be more relaxed than was justified about the air threat to operations in the South China Sea in late 1941, and to pose fewer questions about IJN carrier power. While these comparisons with the Italian air force are viewed as pejorative (and this was undoubtedly intended by the joint planning staff), the Royal Navy, certainly Cunningham and Somerville, had learned to treat the Italian air force with respect in the central Mediterranean.

By late 1941 the Italians had achieved significant successes. For example, they had torpedoed four heavy cruisers, with varying degrees of damage, in four separate incidents between September 1940 and July 1941. They had regularly attacked the fleet in Alexandria and conducted very disruptive mining operations in the Suez Canal. Finally, they had torpedoed and severely damaged the battleship Nelson during Operation Halberd, a convoy to resupply Malta run from Gibraltar in late September 1941, just two months before the Force Z operation. Ironically, Prince of Wales was part of the battleship escort, shooting down two Italian torpedo bombers. In his memoirs, Cunningham was unequivocal about the high quality of Italian air performance at sea during the first part of the war. He stated that they appeared to have some squadrons ‘specially trained for anti-ship work’; their reconnaissance was ‘highly efficient’ and ‘seldom failed to find and report our ships’; ‘bombers then invariably arrived in an hour or two’; ‘Italian high level bombing was the best I have ever seen, far better than the German’. This experience with Italian air performance at sea would surely have tempered any Royal Navy inclination to be complacent about IJNAF aircraft in Indochina if it had been recognised they were armed with torpedoes, and available intelligence on their range had been properly absorbed.

Overall, the Royal Navy’s picture of IJN capability, while it missed the impact of the ‘revolution’ in its approach to air power in 1941, was otherwise balanced and realistic. None of the key documents from 1937 onward show significant inclination to dismiss IJN fighting power, or to contemplate a major engagement except on ‘favourable terms’. One final issue deserves mention. Churchill told the First Lord, A V Alexander, in September 1940 that ‘The NID are very much inclined to exaggerate Japanese strength and efficiency’. This quote has been much repeated over the years and become part of the case against Churchill in assigning responsibility for the loss of Force Z. There are two points to make here. First, the quote is not consistent with the suggestion that the Royal Navy itself consistently underestimated the IJN. Secondly, it is almost certain Churchill was recalling arguments over comparative battleship strengths when he was First Lord earlier in the year. He was frustrated that he was being asked to justify investment in additional Royal Navy build on the basis of what he felt was little more than speculation about IJN strength. Not for the first time, he was subjecting his briefers to rigorous sceptical questioning.

How far the Royal Navy applied the intelligence picture of the IJN described in shaping and executing a strategy for dealing with the naval risk posed by Japan, from mid-1941 through to the outbreak of the Far East war. The resources available to the Royal Navy to defend the eastern theatre after mid-1940, and the Royal Navy policies and doctrine that influenced how those resources were used.

Royal Navy resources for an eastern war in 1941

Britain implemented a modified New Standard naval building programme over the three fiscal years 1937–39, designed to ensure adequate parity with the combined fleets of Germany and Japan by 1942. The outbreak of the European war in September 1939 triggered substantial changes to this programme. The plethora of new wartime shipping requirements, both new build and repair, and affecting both warship and merchant shipping needs, brought an immediate slowdown to previous plans. In May 1940, in order to meet the requirement for additional Atlantic escorts, merchant ships, and urgent army needs, pre-war orders were suspended for six capital ships, one aircraft carrier, eight cruisers and ten destroyers. In practice, most of the labour effort released here was diverted to the fleet destroyer and destroyer escort programmes, raising their manpower allocation by about 55 per cent. Not only were those vessels already under construction, including those temporarily suspended, all completed within the next two years, a further eighteen fleet destroyers and forty-two Hunt class were progressively laid down after 1 June 1940 and commissioned before the end of 1942. This enhanced destroyer output released older vessels for Atlantic escort, and was critical to supporting a new Eastern Fleet in 1942.

The suspended capital ships were Howe, the final King George V, the four Lion class and Vanguard. Howe was resumed in 1941 and completed in August 1942. Vanguard was laid down in October 1941, but not completed until 1946. The Lions were later cancelled and broken up on their slips. The aircraft carrier was Indefatigable, the last of the six Illustrious class. She was resumed in 1942 and completed in May 1944. Her sister Implacable was also much delayed, and completed even later in August that year. In September 1939 the naval staff had expected Implacable and Indefatigable to be commissioned in October 1941 and June 1942 respectively. Their availability in 1942 would have made a significant contribution to the Royal Navy’s global commitments, and made the Royal Navy the strongest carrier power until late 1943. The 1940 building programme agreed over the previous winter was also savagely pruned back. The impact of these suspensions and cuts, alongside war losses, on Royal Navy strength in 1941 and 1942 is illustrated in Table below.

Source: Future Strategy paper, September 1940, and H T Lenton.

Target strengths reflect planned production following May 1940 suspension and cuts, but do not allow for war losses. Actual strengths achieved, which take account of real building rate, overseas purchases and war losses are given in brackets. The figure of thirteen capital ships for August 1940 excludes Queen Elizabeth, undergoing modernisation until early 1941.

The message from these figures for Royal Navy resources available to counter the naval risk from Japan is stark. As a result of the demands of the European war, nominal Royal Navy strength in fleet units hardly changed from the time the Far East appreciation was issued in August 1940 and the outbreak of the Far East war in December 1941. In reality, the deficit against target strength was greater than displayed in the table, because the figures omit units under repair. Two fleet carriers and one battleship were out of action for between six and twelve months in August 1941 following war damage, and a further three capital ships were out of action for between six and eighteen months in January 1942. The number of modern or modernised capital ships able to engage Axis units on equal terms never exceeded eight in this period, and was often less.

The table also fails to bring out an acute shortage of modern fleet destroyers, since the destroyer figures in the table combine the categories of fleet destroyers and destroyer escorts. The Royal Navy entered the European war in 1939 with 103 modern fleet destroyers. By August 1940 war losses had reduced this figure to eighty-seven. There was a further net fall to eighty-three in August 1941, with a small increase to eighty-five by January 1942. However, during the second half of 1941, a further thirty destroyers were, on average, unavailable owing to damage or refit, so effective strength was rarely above fifty-five. This was a quite inadequate number to meet fleet needs in three separate theatres and would be an important constraining factor in building up an eastern fleet in 1941/42. The shortage was exacerbated by the fleet destroyer suspensions ordered in May 1940, which inevitably extended the period where losses outweighed gains from build. The balance in the destroyer total in the table comprised destroyer escorts which fell into three categories: older vessels completed before 1920 to First World War design; around thirty-five old American destroyers acquired in the destroyer for bases deal in August 1940; and the new Hunt class which began commissioning in mid-1940 and reached a figure of forty-three by January 1942. The United States actually provided fifty destroyers, but around fifteen were transferred to the Royal Canadian Navy and allies such as the Royal Netherlands Navy. These American vessels took around six months on average to refit for Royal Navy use and were then of only limited value – barely satisfactory as Atlantic convoy escorts. The Hunt class were originally designed for Atlantic escort, but their sea-keeping was a disappointment, and they were generally deployed in coastal escort and in the Mediterranean, where they often operated as fleet destroyers with reasonable success.

On the positive side of the balance, the Royal Navy inflicted proportionately greater damage on Germany and Italy in the period up to December 1941. While nominal Royal Navy frontline strength grew over this period, German and Italian strength in surface forces declined by about 40 per cent in each case. By end 1941, the Germans had lost one battleship, one pocket battleship, one heavy cruiser, two light cruisers and twelve fleet destroyers. They had also lost fifty-three U-boats to end August 1941, almost the total strength at the start of the war. The Italians had lost one battleship, three heavy cruisers, three light cruisers, twenty-four fleet destroyers and twenty escort destroyers. Italian submarine losses were also very heavy in this period, at thirty-eight against a total build of eighty-five between 1930 and 1942.

Even more important, by diverting building resources away from fleet units from late 1939 onward to meet the requirements of the Battle of the Atlantic with the U-boats, and judicious management of merchant ship resources, Britain substantially compensated, through a mix of new build, requisition and capture, for the 9.1 million tons of merchant shipping lost during the twenty-eight months to the end of 1941. Indeed, British-controlled shipping, comprising vessels over 1600 gross registered tons, actually rose from 17,784,000 tons at the outbreak of war in September 1939 to 20,693,000 tons at 31 December 1941. This even included a net gain in the crucial category of oil tankers of some 1.2 million tons. The Royal Navy commissioned about two hundred Atlantic escort vessels, primarily Flower-class corvettes, in addition to the destroyer escorts referred to above, across the two years 1940 and 1941, with about a third of these built in Canada. This provided a critical margin over U-boat build in this period. Britain may not have been winning the Battle of the Atlantic at the end of 1941, but it was in no immediate danger of losing. This was a far better outcome than anticipated by the chiefs of staff in the Future Strategy paper of September 1940 and their Strategic Survey prepared for the ABC-1 staff talks three months later. It demonstrated that Britain had got its immediate naval priorities right. The IJN would fail to do so. Finally, it is important to note that the Royal Navy would see a steady improvement in its overall destroyer situation in 1942. It would commission thirty-four fleet destroyers that year, more than double the numbers in 1941, and thirty-eight escort destroyers. Twenty-five fleet destroyers and eight destroyer escorts would be lost, but this still left a significant net gain in strength.

This resource picture explains why, as the Admiralty reviewed options for Far East reinforcement immediately following ABC-1, the only readily available force was Force H. This could be redeployed in emergency from Gibraltar, with the added prospect of the four ‘R’-class battleships, if and when the Americans released them from Atlantic convoy escort. Five months later, by end August, the outlook was better. Admiralty planners could anticipate a small increase in capital ship availability by the end of the year as three King George V class came available, and three Illustrious-class carriers arrived from new build and repair. So long as there were no further losses, this would allow a defensive force of modern units to be established in the Indian Ocean by spring 1942, even without full US Navy ‘Atlantic substitution’ and relief of Force H. This resource situation underpinned the dispositions which the First Sea Lord presented to the prime minister in their correspondence in the last week of August.

The result of neglect and ignorance!

Royal Navy Readiness for a War with Japan in Mid-1941: Intelligence and Capability III

Royal Navy air capability in 1941

The Royal Navy required a strategy for dealing with Japan that took realistic account of its available resources in 1941. However, it also needed to apply the right capabilities and fighting tactics. Here, understanding the new potential of air power at sea and applying the lessons of the war to date was particularly important. The Fleet Air Arm entered the European war with obsolescent aircraft, and also faced low priority for development and production. Once the war got underway, a combination of prioritisation by the Ministry of Aircraft Production on key Royal Air Force types, procurement mismanagement and shifting operational requirements caused development of the new generation fighter and strike aircraft ordered in 1939 and 1940 to move painfully slowly.

Of the three aircraft taken forward, none ultimately reached service until 1943. Only one then proved operationally adequate, albeit not in its intended role. They were the Fairey Barracuda torpedo bomber, the advanced next generation torpedo strike reconnaissance (TSR) aircraft to replace the Albacore and Swordfish; the Fairey Firefly specification N 8/39 two-seater escort fighter to replace the Fulmar; and the Blackburn Firebrand high-performance single seat fighter. The failure of the Barracuda was partly owing to the cancellation of the advanced Exe engine by the Ministry of Aircraft Production in 1940 to concentrate on other engine priorities. This was a decision beyond the Royal Navy’s control, but it led to substantial and unsatisfactory redesign. The development of the Firefly was inhibited by changing perceptions of fighter requirements and the best means of meeting these. It eventually proved a useful aircraft, but as a fighter-bomber and night-fighter rather than mainstream dayfighter, where it could not compete with contemporary US Navy fighters. It was not fully operational until 1944. The Firebrand was a high-performance single-seat fighter able to take on land-based aircraft, but was fatally compromised by hasty design and too many conflicting requirements.

The Norwegian campaign persuaded the Fleet Air Arm that its most critical need was a dedicated naval fighter with sufficient performance to counter the latest land-based strike aircraft and their escorting fighters. Experiments here with radar-assisted fighter direction emphasised the desirability of an agile aircraft able to gain height and distance quickly for interception. The first Royal Navy radar-controlled interception at sea occurred on 23 April 1940, using the converted anti-aircraft cruiser Curlew as a radar picket. The Norway experience also demonstrated the potential for integrating new technology: radar, lightweight VHF communications (which were far ahead of the US Navy) and radio homing. Essentially, the Fleet Air Arm was now reaching for a ‘Battle of Britain’ fighter defence concept, based on new perceptions of the threat and available technology. Two factors, therefore, came together here to drive a fundamental shift in fighter philosophy: the threat from high-speed aircraft and the technical means to counter this threat with carrier-based fighters, using the emerging Royal Air Force fighter direction techniques.

The result was the specification of the new Firebrand single-seat fighter in July 1940, and the decision to meet the gap before the Firebrand’s arrival by ordering the Grumman F4F Wildcat from the United States, which the Royal Navy named the Martlet. This followed an extensive debate over how to reconcile the traditional requirement for a long-range escort fighter to support carrier strike forces, which in the Fleet Air Arm view required a dedicated navigator as well as pilot, and the new need to defend the fleet against high performance attackers, where equivalent fighter performance was needed, implying single seat and acceptance of shorter range. The Admiralty continued to worry about the difficulties a single-seat fighter might face with navigation until well into 1941. The initial conclusion was that two aircraft were needed. By September 1941 it was already clear the Firebrand was well behind schedule and could not be operational before mid-1943.

Meanwhile, deploying the interim Martlet from Royal Navy carriers posed a problem. Without folding wings, none of the lifts in Ark Royal or the first three Illustrious class were large enough to accommodate them. The solution was to ask Grumman to produce a folding wing variant and eventually to adopt the US Navy practice of deck parking. Because of these limitations, the first Wildcats delivered were used in a land-based role for the defence of Scapa Flow. Designing and producing a folding wing took much longer than hoped, and the Admiralty briefly considered undertaking the work in the United Kingdom. In early 1941 it expected the folding wing variant to begin arriving in the middle of the year, but this proved wildly optimistic. Nevertheless, at the end of 1941, the Royal Navy still assessed the Wildcat the best naval fighter in the world. Trials had by now shown that it was more manoeuvrable than a Hurricane I, was faster at all heights up to 15,000ft, and climbed better too. The Royal Navy ordered 100 aircraft in July 1940, 150 more in December 1940 and 200 in October 1941. It appears a further eighty-one were taken over from French orders in mid-1940. Grumman had originally promised delivery of twenty aircraft per month from late 1940, sufficient to equip the Fleet Air Arm frontline fighter force by the following autumn. This would have given Royal Navy carriers broadly the same fighter capability as the US Navy in confronting the IJN in late 1941 and in 1942. The US Navy had initially selected the Brewster Buffalo as its preferred naval fighter, but switched to the Wildcat in mid-1941, partly on the basis of British experience. The Wildcat then remained its primary naval fighter throughout 1942. The Buffalo, as we have seen, became the mainstay of British fighter defence in Malaya.

However, by the end of 1941 Grumman had delivered less than half the numbers contracted and paid for and, crucially, none had the folding wings, essential to the Royal Navy, which Grumman had also promised. Churchill, who monitored the Fleet Air Arm fighter problem throughout 1941, described this as ‘a melancholy story’. In September that year, in typical fashion, he cut to the heart of the issue. ‘All this year it has been apparent that the power to launch the highest class fighters from aircraft carriers may re-open to the Fleet great strategic doors which have been closed against them.’ He instructed that – ‘the aircraft carrier should have priority in the quality and character of suitable (aircraft) types’. The latter was, of course, easy to say, but much harder to deliver. The US Navy, by then also acquiring Wildcats, blamed the delays on acute shortages of materials across the US aircraft industry, owing to concentration on the four-engine bomber programme.

The end of 1941, therefore, found the Fleet Air Arm badly let down by American industry and desperately trying to plug its fighter gap through stopgap adaptations of the Hurricane and Spitfire, neither of which were really suitable for carrier operation. Once it became clear that Wildcat deliveries would be delayed, 270 Hurricanes were released to the Fleet Air Arm between February and May 1941. Inevitably, some of these Royal Air Force releases were of poor quality. Alexander again summarised the status of Fleet Air Arm procurement for the prime minister in early December. He explained why the Hurricane and Spitfire, although necessary as stopgaps, were inherently unsuitable for carrier operation. It remained essential to maximise Wildcat supplies until new British naval fighters were available, but supplies could only be assured by asking the United States to prioritise naval fighter over heavy bomber production. Despite Alexander’s insistence that the Spitfire was unsuitable, it appears that the possibility of converting Spitfires for carrier use, including the provision of folding wings, was considered as early as February 1940, but was ruled out at this time because of the impact on Fighter Command requirements. The idea was then resurrected in September 1941, technical issues were resolved by January 1942, and an initial order placed for 250 aircraft, by now designated the Seafire. The modification of the Spitfire for naval use inevitably still took time, and the first twenty-eight aircraft were not delivered until July 1942. At that point, deliveries were expected to run at twenty-five per month and by late August the initial order for 250 was expanded to 452 for the end of 1943. This reflected the view by then that the Firebrand would have to be abandoned.

Given the delay to the Barracuda, the Fleet Air Arm also considered the merits of purchasing strike aircraft from the United States. The current US Navy torpedo bomber, the Douglas Devastator, offered no significant performance advantage over the Swordfish or its Albacore successor, which began coming into service in 1940. The US Navy did have an excellent dive-bomber in the Douglas Dauntless SBD2, but the Royal Navy had long chosen to prioritise torpedo attack over dive-bombing, and its carriers had insufficient capacity to carry reasonable numbers of both types. There is a compelling argument here that the US Navy ended up prioritising the dive-bomber because its aerial torpedoes did not work satisfactorily. By contrast, the Royal Navy had excellent torpedoes. The Fleet Air Arm was, however, impressed with the forthcoming Grumman TBF Avenger, which would prove the outstanding naval strike aircraft of the war. Two hundred were ordered in late 1941 under lend-lease, but did not reach the Royal Navy until 1943.

This background meant that until well into 1942, the Fleet Air Arm was largely dependent on obsolete biplane strike aircraft dating from the mid-1930s and a more modern fighter of interim design, the Fulmar, introduced in 1940. Indomitable, completed in autumn 1941 with larger lifts and extra hangar space, was equipped with Hurricanes from the start, while modifications to Illustrious and Formidable in the United States in late 1941 enabled them to deploy Wildcats when they rejoined the fleet at the end of the year. While they were not competitive with the latest IJN and US Navy carrier aircraft, the limitations of the standard British Fleet Air Arm types can be overstated. The Fulmar, as primary fighter from 1940, could not match a Zero for speed or manoeuvrability, but it was still quite fast enough to catch the most modern loaded strike aircraft, and had other advantages both as an interceptor and reconnaissance aircraft. It was more robust than a Zero, could dive at 400mph (faster than almost all contemporaries and a speed which would cause a Zero to break up), had a four-hour endurance, and was an excellent gun platform with large ammunition capacity (which was a Zero weakness), all of which, when allied to radar control, made it a useful defensive fighter.

The Swordfish and Albacore torpedo bombers had reasonable range, a good communications fit, were married to an excellent torpedo, and their biplane manoeuvrability was an advantage in marginal flying weather. The Fleet Air Arm torpedo here was the 18in Mark XII introduced in 1940. Contrary to the impression often given that the Royal Navy was backward in aerial torpedo capability, the Mark XII could be dropped at up to 200ft (61m) and 150 knots, making it easily comparable to the IJN Type 91. It also had two speed options, an excellent warhead and the option of an advanced detonating pistol, the duplex magnetic proximity fuse. In addition, it had gyro-angling so the attacking aircraft could offset, rather than flying direct at the target.

The Fleet Air Arm’s Swordfish and Albacore torpedo bombers could not match their IJN counterparts for speed and range, but compensated through their ability to operate at night or in bad weather, and they had high reliability. It is doubtful either IJN or US Navy aircraft could have taken off in the conditions prevailing in Ark Royal on 26 May for the crucial attack on the Bismarck. In addition, from early 1941 significant numbers were fitted with ASV air surface search radar, giving them a night and bad weather search and attack capability which neither the IJN nor US Navy could match. The first airborne radar sets, known as ASV 1, were deployed in Fleet Air Arm aircraft in late 1939, but were fragile and unreliable. The much improved ASV 2, with an effective range of about 15 miles to detect a medium-sized warship, began to be deployed in early 1941. The 825 Squadron embarked in the new carrier Victorious were entirely ASV 2-equipped, and this facilitated their night attack on Bismarck on 24 May. Some aircraft in Ark Royal were also equipped, enabling them to operate against Bismarck in the atrocious weather prevailing two days later. Judged, therefore, as an overall weapon system, the Fleet Air Arm torpedo bombers were effective and competitive.

By the autumn of 1941, despite the limitations of its aircraft both in numbers and quality, the Royal Navy carrier fleet had achieved notable operational successes across different theatres and under different commanders, amply demonstrating that the Royal Navy was definitely not tied to a conservative battleship mentality. One measure here is that the Royal Navy had sunk or disabled five modern or modernised capital ships through aerial torpedo attack by end May 1941. At Taranto in November 1940, just eleven Fleet Air Arm Swordfish sank or disabled three modern or modernised Italian battleships, while eighty-nine IJN B5N2s were required to sink or disable five older US Navy battleships in the first-wave attack at Pearl Harbor. While the IJN is invariably credited with sinking the first capital units at sea (Prince of Wales and Repulse), the Fleet Air Arm had previously disabled a modern German battleship at sea (Bismarck) and come close to disabling another, the modern Italian battleship Vittorio Veneto, at Matapan. They had also severely disabled the new French battleship Richelieu at Dakar on 8 July 1940.

The Royal Navy had also completed more modern carriers than either the IJN or US Navy, even if, for all the reasons discussed, it badly lagged both in the number and quality of frontline aircraft, together with trained and experienced crews. By autumn 1941 the Royal Navy had completed five large modern fleet carriers, against four by the IJN, and three by the US Navy. The IJN and the US Navy had also each completed a single light carrier earlier in the 1930s. Despite all its procurement problems and heavy wartime attrition, frontline air strength had almost doubled in size after two years of war, and it would continue grow steadily into 1942. Relevant figures here are as follows:

Aircraft reserves by autumn 1941 were also substantial compared to those of the IJNAF. At the end of September 1941, alongside the frontline strength recorded above, there were 141 strike aircraft and seventy-two fighters allocated to training, a further 540 and 335 respectively in general reserve and fifty-seven and fifty-five in transit.

The Royal Navy nevertheless had to divide its carrier fleet across four theatres, Atlantic, the two ends of the Mediterranean, and the Indian Ocean, and the initiation of the Russian convoys from late 1941 effectively added a fifth theatre, the Arctic. This made it impossible further to explore and train for the multi-carrier operations and mass strike tactics which the Royal Navy had investigated in the 1930s, and which the IJN developed so assiduously during 1941. In theory, this left it poorly placed to handle a mass carrier engagement of the Coral Sea or Midway type, which the IJN and US Navy would experience in mid-1942. However, by mid-1941 the Royal Navy had made innovations which partly compensated for lack of numbers and quality of aircraft. Both the US Navy and IJN had embraced an operational philosophy based on the single massive strike, which made it difficult to manage the more flexible flying cycle required to handle the competing tasks and demands of a multi-threat environment. The Royal Navy, by contrast, emphasised operational flexibility with rapid switches of aircraft tasking. The evolution of this sophisticated multi-role flying cycle is well illustrated in some of the reports on Force H operations by Vice Admiral Somerville as early as autumn 1940. These show the execution of techniques and tactics, including radar pickets, which would still be in use in the 1960s, and were quite beyond US Navy and IJN capability at that time.

Royal Navy Readiness for a War with Japan in Mid-1941: Intelligence and Capability IV

The latest Royal Navy carriers had the most advanced radar of the day. Aircraft communications systems and direction-finding homing were steadily improved. The Royal Navy continued developing and refining the techniques for long-range enemy raid detection and fighter direction that were first tried in Norway and employed during subsequent operations in the Mediterranean. Impartial testament to the quality of Fleet Air Arm operations at this time came from an American observer, Commander E G Taylor (US Navy), who served in the Fleet Air Arm in 1940 as a Royal Navy sub lieutenant. He informed the US Navy that the Fleet Air Arm in the spring of 1940 was in a desperate condition as regards pilots, aircraft numbers and quality, but did an excellent job with what it had through skill, innovation and new technology. Taylor described Fleet Air Arm fighter control techniques as very effective and far in advance of the US Navy which, in his view, was still not tackling the issue seriously at the end of 1941. Taylor subsequently played a part in Royal Navy procurement of the Wildcat, drawing on his war experience. The US Navy was later impressed with, and subsequently largely adopted, Royal Navy air defence tactics demonstrated by the Royal Navy carrier Victorious when on loan to the US Navy Pacific Fleet in mid-1943.

By mid-1941 the Royal Navy also had the basis of the action information organisation (AIO) that would revolutionise the way naval commanders used available information to drive operations. It would be followed with only minor incremental changes in the Royal Navy and US Navy through to the 1960s. The IJN had no radar at all before late 1942, totally inadequate airborne communications, and no real concept of fighter direction. IJN combat air patrols essentially ran themselves. This meant that IJN air defence was inefficient and susceptible to surprise.

The Royal Navy had been experimenting with night flying operations, including night torpedo attack, since the early 1930s. The scope for night attack pre-war was, however, limited because of the difficulty of conducting night searches before the advent of radar. It depended on night shadowing of an enemy already located at dusk, at which the Fleet Air Arm practised and became adept. However, during 1941, about half of the Royal Navy’s strike aircraft were fitted with ASV, providing a genuine night search capability for the first time. This enabled the Royal Navy carriers to undertake attacks at night or in bad weather, conditions impossible for the IJN or US Navy. Examples of night/bad weather operations were the attack on Taranto in November 1940, a night attack at a range of around 180 miles from Illustrious, and the previously mentioned bad weather attacks on Bismarck from Victorious and Ark Royal on 24 May and in the afternoon and evening of 26 May 1941. ASV was arguably crucial to the success of the Bismarck attacks, as emphasised by the commanding officer of 825 Squadron in Victorious. ‘ASV proved to be of assistance beyond all measure. While remaining hidden from the enemy by cloud, my observer was able to give me a clear picture of own and enemy forces up to the very last moment of breaking cloud for attack’.

By mid-1941 the Royal Navy had, therefore, acquired a much wider base of wartime experience in carrier operations than is often suggested. Some of this reflected the application and elaboration of pre-war thinking. The ‘find, fix and strike’ role on behalf of the battle-fleet evident at Matapan and the Bismarck operations fall into this category, and even the Taranto operation drew heavily on pre-war plans. The Royal Navy had formed a special committee to look at air attack on enemy bases in 1929, and investigated the merits of specialised weapons for this. Mediterranean Fleet orders in 1936 included instructions for shallow-water torpedo settings for an attack on Taranto. However, in air defence the Royal Navy was breaking new ground with capabilities, not only radar but also VHF communications and ‘identification friend or foe’ (IFF), which were not available before the war. It was also becoming adept at exploiting the full range of carrier capabilities: search, strike against land and sea targets, air defence, and anti-submarine, which all featured in complex multi-threat scenarios such as Mediterranean convoy operations.

All of this had demanded much tactical innovation and a new approach to force composition. It was the Royal Navy which consolidated the concept of the fast multi-role carrier task force in this period, of which Force H was the supreme example. This model would not be embraced by the US Navy until much later, because until its new build battleships (North Carolina and Indiana classes) became available for the Pacific in late 1942, it had no capital ships fast enough to operate with its carriers, and it also lacked the oil tankers to support them until well into 1943. By contrast, the Royal Navy not only had its battlecruisers, but also widely used the modernised Queen Elizabeth class, significantly faster than their US Navy equivalents, alongside its carriers in both the Mediterranean and Indian Ocean. This not only ensured support against surprise surface attack, but also provided substantial anti-aircraft firepower. Finally, as proof that it could conduct the most advanced multicarrier operations, the Royal Navy undertook a four-carrier operation against the most challenging opposition, air, surface and submarine, in Operation Pedestal, the convoy to Malta, in August 1942. Both the IJN and US Navy would have struggled to conduct a comparable operation against an equivalent level of air attack at this time. Overall therefore, while the Royal Navy could not hope to compete with the IJN in late 1941 in either carrier numbers or quantity and quality of embarked aircraft, it would be quite wrong to view it as left behind in the exercise of air power at sea. The Royal Navy carrier force had significant strengths and experience to draw on.

Royal Air Force maritime strike

The 1937 agreement gave the Royal Navy full control of its carrier air arm, but left land-based maritime attack with the Royal Air Force. For the Royal Air Force, this role inevitably took lower priority in the first phase of the war to the air defence of the British homeland, to strategic bombing, and even within the maritime sphere to the anti-submarine effort. The emphasis on strategic bombing reflected long-standing service doctrine and it was the only immediate way of directly attacking Germany.

By late 1940 the Royal Air Force had nevertheless deployed a new torpedo bomber, the Beaufort, for maritime strike. It was broadly equivalent in capability to the IJN land-based torpedo attack bombers, albeit with less range. However, a British equivalent of the IJN 11th Air Fleet would have required a significant diversion of resources from Bomber Command to Coastal Command that was never forthcoming, even in the face of the acute threats in the Battle of the Atlantic and in the Indian Ocean in 1942. The United Kingdom-based Beaufort torpedo force struggled, therefore, to reach four operational squadrons by the end of 1941.226 The only overseas torpedo force at that time comprised the two Malaya squadrons equipped with obsolete Vildebeest biplanes. The parlous state of Royal Air Force torpedo strike in early 1942 was summarised in a report for the Assistant Chief of Air Staff (Operations). This stated that although torpedo attack was more effective against ships than bombing, the Royal Air Force had neglected it for both technical and tactical reasons. It highlighted lack of high-level commitment, shortage of torpedoes and associated maintenance and support facilities, and inadequate training for this specialised role, given constant diversion of aircraft and crews to other tasks. This presented major difficulties in expanding the force.

The lack of high-level support for torpedo attack partly reflected the scarcity of high-value warship targets accessible to the United Kingdom force, since bombing was seen as adequate for interdiction of commercial traffic. Bombing was also the preferred option for attacking naval targets in port or dockyard. However, it mainly reflected the entrenched belief across most of the Royal Air Force leadership that the strategic bombing of Germany must take primacy and that everything else, with few exceptions, represented a diversion from winning the war. Statistics did not help the Coastal Command case for enhanced surface strike resources. They demonstrated that surface strike was very expensive for only limited benefit. Between April 1940 and March 1943, Coastal Command conducted 3700 attacks which sank 107 vessels (almost all merchant) for the loss of 648 aircraft. Minelaying, primarily by Bomber Command, sank three times the number of ships (369), for about half the losses (329).

The poor Coastal Command surface attack record was painfully underlined by the woeful performance of the Beaufort force during the escape of the German battlecruisers through the Channel in February 1942. By contrast, both Scharnhorst and Gneisenau were badly damaged by mines dropped along their predicted track, and Gneisenau was wrecked by Bomber Command in dock at Kiel a fortnight later. Nevertheless, despite limited resources, the Beaufort force had some notable successes. An attack on 6 April 1941 crippled Gneisenau alongside in Brest, and a night operation on 13 June seriously damaged the pocket battleship Lützow. This attack on Lützow off the Norwegian port of Egersund took place at a range of about 400 miles, only slightly less than the later IJN attack on Force Z. It was the first successful attack by a land-based torpedo force on a capital ship at sea and a harbinger, therefore, for the attack on Force Z. The Lützow operation obviously contrasted with the poor performance by all three available United Kingdom Beaufort squadrons against the German battlecruisers six months later, but this reflected incompetent communication, command and control, rather than any deficiency in aircraft performance.

Small size and lack of dedicated expertise in the United Kingdom Beaufort force constrained the development of a torpedo strike force overseas. However, two squadrons were deployed to Egypt via Malta in early 1942 to compensate for the planned departure of the Alexandria battle-fleet. This was destined for the Indian Ocean following the Japanese attack, although the ships were badly damaged by Italian frogmen before they could be released. This planned substitution of Beaufort torpedo bombers for battleships was an almost exact analogy of Yamamoto’s use of the 22nd Air Flotilla to counter Force Z. It demonstrated that the Royal Navy leadership recognised the strategic impact that an appropriate land-based strike force could achieve. However, the Beaufort force was too small, and its crews insufficiently trained, to stand real comparison with its IJN counterpart. The force took some while to achieve operational efficiency in the theatre and there were insufficient resources for an additional force in the Indian Ocean, although one of the Egypt squadrons would be deployed onward to Ceylon at the end of April. The Chief of Air Staff accepted that torpedo bombers were the most pressing need in the Far East, but the only way of providing these was to replace the remaining two Beaufort squadrons in UK with obsolete Hampdens. The remaining Beauforts would then be deployed to the Middle East, releasing the two squadrons already despatched there to go on to India. Clearly, the impact of just two squadrons would have been limited, although still a significant contribution to the defence of Ceylon.

The Far East also suffered problems with the planned production of Beauforts in Australia. Australian Beauforts were to replace the Vildebeests in Malaya by mid-1941 as part of the ‘336’ reinforcement plan. When Brooke-Popham took stock of planned reinforcements following his arrival as Commander-in-Chief Far East at the beginning of 1941, he expected to have two squadrons of Beauforts and reserves, comprising fifty aircraft by the end of the year. However, he was aware, following a visit to Australia, that this reflected an estimated output in Australia of twenty aircraft by October and seventy at the end of the year. Given the embryonic state of the Australian aircraft industry, he should perhaps have anticipated that the targets would slip, even without problems in the American supply chain. Shortages of key parts from the United States meant that only seven aircraft were completed by the end of 1941, when production halted for six months. Even if the original target had been met, the new Beaufort squadrons could not have achieved operational efficiency until well into 1942, not least due to a shortage of suitable torpedoes. The modernisation of the Malaya strike force on which Far East commanders placed much emphasis was, therefore, always compromised although neither the chiefs of staff nor Far East commanders seem to have known this. Following the Japanese attack, they repeatedly pressed the Australian government to expedite the provision of Beauforts to Malaya.

There is no doubt that British maritime air power was quite inadequate to take on the IJN in December 1941. However, this situation cannot be reduced to simple explanations. The Royal Navy’s overall record and experience in 1940 and 1941 demonstrates a good grasp of the risks and opportunities presented by modern air power at sea, and often showed remarkable innovation with limited resources. In rebalancing its naval programme in 1940, the Royal Navy cut battleships, not carriers. The Royal Navy knew its Fleet Air Arm aircraft were inadequate. It made reasonable and timely decisions to address this, but fell victim to production priorities in the United Kingdom and United States it could not foresee. The Royal Air Force’s Beaufort force had similar potential capability to the IJNAF, but suffered the limitations of small size, multiple roles and inadequate training for the specialised maritime environment. The Royal Air Force could have invested more in the Beaufort force at the expense of bombing. For decision-makers in 1941 that trade-off was hard to justify, but it made it difficult to build up the force when it was urgently needed overseas in 1942. Matters did begin to change from mid-1942 with the decision to develop a variant of the well-proven Beaufighter as a replacement for the Beaufort. This was the Beaufighter TFX, known as the Torbeau, which became a very successful maritime strike aircraft, with the first squadron going operational in the United Kingdom in November that year. The Royal Air Force should, however, have been quicker to spot the problems with the prospective Australian Beaufort force, given its importance to Far East reinforcement.

The lack of adequate maritime air resources to meet the IJN is invariably presented as a consequence of over-stretch. However, the problem was less one of overall capacity than strategic choices. Britain chose to prioritise its air resources in 1941 on Fighter and Bomber Commands at home and in the Middle East, and deliberately to carry risk in the Far East. For a whole series of reasons outside the scope of this book, which included steady attrition, production failings, and the need to re-equip the whole frontline, Bomber Command took a long time to reach a size with which it could have any hope of strategic effect. Air Chief Marshal Sir Arthur Harris stated that, when he took over the command in February 1942, when prospects in the Far East were most grim, immediate strength was 378 aircraft of which only sixty-nine were heavy bombers and fifty were light bombers. His primary force was, therefore, 250 medium bombers, about the same as those in the Middle East and those planned for India by the autumn. Against this background, in the spring of 1942 Pound, with partial support from the new Chief of the Imperial General Staff, Field Marshal Sir Alan Brooke, argued forcefully for air resources to be diverted from bombing to protection of sea communications. Pound’s primary concern here was the Atlantic, but he also emphasised the need to improve the defences of the Indian Ocean. He met vigorous opposition from Portal as Chief of Air Staff, and the Harris numbers help to explain why.

The subsequent argument has been described as one of Whitehall’s most notorious battles of the war. The debate ran for three months before a compromise position was reached in July, which provided Pound with only a limited part of what he had sought. Four factors have been suggested to explain why Pound substantially conceded:

– The attitude of the prime minister, which was broadly behind Bomber Command;

– Production limitations, which made it impossible for a diversion to maritime air to achieve results in a meaningful timescale;

– Shifting strategic perceptions in favour of bombing U-boat bases and focusing effort on the Bay of Biscay;

– Improved efficiency in Coastal Command by mid-1942.

Portal also had sound reasons for rejecting demands from Attlee, Beaverbrook and Wavell, as well as Pound, for more bombers to be sent overseas at this time beyond those currently planned. These included the obvious constraints in time and shipping space inherent in any move of large air forces; the technical teething problems with the latest aircraft which would raise immense serviceability issues outside United Kingdom infrastructure; the ‘strategic immobility’ caused by the need to replace Bomber Command’s frontline; and the impossibility of adapting heavy bombers to a torpedo strike role. These constraints did not reflect well on previous Royal Air Force planning and procurement, but as ‘we are where we are’ arguments they were powerful. Also important was the impact of Midway, and the drastic reduction in the threat to the Indian Ocean. In effect, a slow or even minimal air build-up in the East was now tolerable.

The striking point is how little was really changed by this debate. Minimal Admiralty requirements in the Atlantic area were met by the end of 1942, and arguably in the Indian Ocean too, but more from the impact of American production than any shift in British priorities. Figures drawn from the British Official History put the debate in context and underline the limited scope for more diversion from Bomber Command. In September 1942 Royal Air Force strength was allocated as follows:

At this time the US Army Air Force also had two bomber groups with ninety-six heavy bombers in the United Kingdom, three squadrons in the Middle East and two in India.

A final perspective is provided in post-war remarks from Dr Noble Frankland, the official historian of Bomber Command, to Admiral Sir Algernon Willis, Somerville’s second in command in the Indian Ocean during 1942.245 Frankland described Bomber Command during the first part of the war until mid-1943 as a small, undernourished force, which did not grow at all in 1942, partly due to re-equipping the frontline, but also substantial diversions to Coastal Command and overseas. Any further diversion might have led to the collapse of the offensive. Frankland did not believe either Portal or Churchill harboured illusions that bombing alone would win the war, but they did doubt it could be won without it, and they may have been right. Harris may have had illusions that bombing alone would do it, but much of his bombast was also aimed at maximising resource share. Frankland also stressed the need to understand that the strategic air offensive embraced many aims and achievements beyond the general attack on German industry. Initially, reducing the scale of German air attack on the United Kingdom was a priority, while in 1941–43 a large part of its effort went on the Battle of the Atlantic, directly or indirectly. If the results were disappointing, the lesson was not that the strategy was wrong, but that the force was operationally inadequate. That reflected poor Royal Air Force equipment and training decisions before the war. Frankland accepted that by switching more resources to Coastal Command (and by implication wider strategic communications), success at sea might have come a little earlier. But the resulting excess resources could not easily have been switched back later to an effective strategic air campaign. The Battle of the Atlantic had to be won, but winning this was no good on its own.

German Advisors at Shanghai 1937

Seeckt served as a member of parliament from 1930 to 1932. From 1933 to 1935 he was repeatedly in China as a military consultant to Chiang Kai-shek in his war against the Chinese Communists and was directly responsible for devising the Encirclement Campaigns, that resulted in a string of victories against the Chinese Red Army and forced Mao Zedong into a 9,000 km retreat, also known as the Long March.

Operation Iron Fist was the main German contribution in the initial stages of the Shanghai campaign, but it was far from the only one. German advisors were present both on the staffs and at the frontline. Their pivotal role was no secret, and even the newspapers regularly reported about them. Wearing the uniforms of Chiang Kai-shek’s army, the German advisors not only provided tactical input, but gave the Chinese troops an invaluable morale boost, showing them that they were not on their own in the struggle against the mighty and ruthless Japanese Empire. The “German War” was the name that some Japanese gave to the battle of Shanghai, and for good reason.

When war with Japan broke out in the summer of 1937, the German advisory corps consisted of nearly 70 officers, ranging from newly graduated second-lieutenants to five full generals. It was a major asset for the Chinese, and one that they were free to exploit. Even though most of the Germans were in China on short-term contracts and could have left once the shooting started, they felt an obligation to stay at a key moment when their host nation’s survival was at stake. “We all agreed that as private citizens in Chinese employment there could be no question of our leaving our Chinese friends to their fate,” Alexander von Falkenhausen, the top advisor, wrote later. “Therefore I assigned the German advisors wherever they were needed, and that was often in the frontlines.”

Alexander Ernst Alfred Hermann Freiherr von Falkenhausen (29 October 1878 – 31 July 1966) was a German General and military advisor to Chiang Kai-shek. Some 80,000 Chinese troops, in eight divisions, were trained and formed the elite of Chiang’s army. However, China was not ready to face Japan on equal terms, and Chiang’s decision to pit all of his new divisions in the Battle of Shanghai, despite objections from his both staff officers and von Falkenhausen, would cost him one-third of his best troops. Chiang switched his strategy to preserve strength for the eventual civil war.

The situation was the culmination of a relationship that had evolved over a period of several years. Germany had started playing a role in China’s military modernization in the late 1920s, with initial contacts facilitated by Chiang Kai-shek’s admiration for German efficiency. The German government’s decision to abandon all extraterritorial privileges in 1921, followed seven years later by the diplomatic recognition of Chiang’s government, also created a benevolent atmosphere. In addition, as a result of its defeat in the Great War, Germany was a relatively safe bet for China. It was, in the 1920s and early 1930s at least, the only major power unable to resume its imperialist policies of the years prior to 1914. Germany and China were in fact in similar situations, Chiang once mused. “They were oppressed by foreign powers,” he said, “and had to free themselves from those chains.”

Yet another factor behind the expanding Sino-German military ties was the lack of suitable employment for officers in Weimar Germany, whose military, the Reichswehr, was severely curtailed by the demands of the post-war Versailles Treaty. The shadow existence they led at home contrasted starkly with the prestige they enjoyed in China. By the mid-1930s, the Germans had a status among the Chinese that no other westerners had ever experienced. When Chiang met with his generals, his chief German advisor at the time, Hans von Seeckt, would sit at his desk, giving the signal that the foreign officer’s place in the hierarchy, while informal, was near the top. When Seeckt had to go by train to a north Chinese sea resort for health reasons, he traveled in Chiang’s personal saloon carriage and was saluted at every station by an honorary formation.

Seeckt visited China the first time in 1933, and immediately set about salvaging bilateral ties strained by German condescension towards the Chinese. As the host nation and employer, China was to be shown respect, was his order to the German officers stationed in the country, and being a traditional German, he expected to be obeyed. When he arrived in China for his second tour the year after, he was accompanied by Falkenhausen. No novice to Asia, Falkenhausen hit it off with Chiang Kai-shek almost immediately. It helped that both knew Japanese, the language of their soon-to-be enemy, and could converse freely without having to go through aninterpreter. It was an additional advantage that Falkenhausen’s wife was on superb terms with Madame Chiang. Falkenhausen’s break came when Seekt, suffering from poor health, returned to Germany in early 1935. From then on, he was the top German officer inside China.

It is likely that Falkenhausen felt a deep sense of relief to be posted abroad. His mission removed any immediate obligation to return to Germany and work with the Nazis. “In the 30s we could have in good conscience stayed in China,” one of Falkenhausen’s subordinates later rationalized. “China was in much greater danger than Germany.” Falkenhausen had a very personal reason to adopt that rationale. His younger brother, Hans Joachim von Falkenhausen, a war veteran and a member of the Nazi Party’s paramilitary Sturm-Abteilung, was executed in a bloody showdown among rival factions inside the party’s ranks in the summer of 1934. He was 36 when he died.

Falkenhausen’s unhappy relationship with Berlin’s new rulers put him on one side of a political generation gap that divided most of the German advisors in China. Among conservative officers of his age and background, feelings about Hitler, a mere corporal in the Great War, ranged from skepticism to adoration; in between was quiet acceptance of an overlap of interests with Germany’s new Nazi rulers, who wanted rapid rearmament and the creation of a vast new army. The younger German officers serving in China were far less ambivalent. They were often ardent Nazis. The racist ideology the young Germans brought with them from home may have contributed to lingering tension with the Chinese. Since most of them expected to leave within no more than a few years, virtually none bothered to change their lifestyles in order to fit into their new surroundings. Rather, in the traditional way of Europeans in Asia, they lived in their own enclave in Nanjing, a small piece of Germany in the heart of China. If they paid any attention to local mores, it was with a shrug of the shoulder. Brought up on austere Prussian ideals, they considered, for example, the Chinese habit of elaborate banquets a costly waste of time and resources.

The Chinese, too, looked at the foreign advisors in mild bewilderment. The German habit of wearing monocles was a cause of wonder and led them to ask why so many were near-sighted on only one eye. A few Chinese did not just puzzle at the behaviour of the strange foreigners, but had attitudes bordering on hostile. Zhang Fakui, for one, appears to have had a particularly delicate relationship with the German advisors. He did not trust them, did not share any secrets with them, and did not take any advice from them. “I had always had a bad impression of the Germans,” he told an interviewer decades later.

Falkenhausen’s own outlook underwent profound change. At the time of his arrival, he had been somewhat indifferent to China, but he gradually grew fonder of the country, and in the end he was very close to accepting an offer of Chinese citizenship from Chiang. As time passed, he even showed signs of divided loyalties between his old and new masters, ignoring pleas from Germany to favor its weapon producers when carrying out arms procurements abroad. Instead, he bought the arms he thought would serve China best, regardless of where they had been manufactured. Finally, he developed a high degree of resentment of the Japanese foe. “It is sheer mockery to see this bestial machine pose as the vanguard of anti-Communism,” he wrote in a report to Oskar Trautmann, the German ambassador in Nanjing.

Once war broke out, Falkenhausen was in favor of an aggressive and all-encompassing strategy against the enemy. He advised that the Japanese garrison in Shanghai be attacked and wiped out, regardless of the fact that it was located inside the International Settlement. He even urged air attacks on western Korea and sabotage on the Japanese home islands. These steps went much further than almost any of his Chinese hosts was prepared to go. Perhaps they feared setting a task for themselves that they could not handle. Falkenhausen, on the other hand, never seemed to have harbored any serious doubts about China’s military prowess. Rather, its army’s willingness to make sacrifices appealed to his special German passion for absolutes. “The morale of the Chinese Army is high. It will fight back stubbornly,” he said. “It will be a struggle to the last extreme.”


Baba Toraji, a 21-year-old employee of the exclusive department store Mag-asin Franco-Japonais, was growing more nervous for every minute that passed on the morning of August 18. A younger colleague of his, fellow Japanese Sakanichi Takaichi, had left earlier to buy bread for his colleagues, and he had not returned. In the end, Baba decided to go looking himself. It did not take long before he found Sakanichi, caught up in a Chinese crowd that had identified him as Japanese. Both men were mauled severely and left on the street. Baba was pronounced dead by the time medical personnel arrived. His younger colleague was sent to hospital with serious injuries.

Earlier in the month, a group of eight Japanese had unwisely shown up on the Bund, trying to push their way through a dense crowd. Jeers started. Someone picked up a discarded shoe and threw it at them. The Japanese broke into a run, and seven managed to escape. A huge brick went sailing through the air and hit the eighth in the back. He fell to the ground, and the mob was upon him. “Men could be seen jumping in the air to land with both feet on the unfortunate man’s body,” the North China Daily News reported. “Others, with stick and bricks that seemed to come from nowhere, belabored him from head to foot.” He was eventually rescued and hospitalized in a critical condition.

Being Japanese in Shanghai in August 1937 was dangerous. By contrast, Shanghai’s western residents only came into contact with the horrors around them in an indirect fashion. They watched the dense black smoke rising over Hongkou, and they saw the flotsam drifting down Suzhou Creek—cows, buffaloes, and a steady stream of uniformed corpses. The debris of war served as a warning that the battle was escalating and could soon engulf the foreign enclaves. It was time for the women and children to leave. A total of 1,300 British and American evacuees departed from Shanghai on August 17. The British left for Hong Kong on the Rajputana, while the Americans boarded the President Jefferson for Manila. On August 19, 1,400 more British citizens, mostly women and children, sailed on destroyers to board the Empress of Asia at Wusong.50 This was part of a scheme to evacuate a total of 3,000 British nationals, including 85 percent of the women and children in the city.

Staying on the fringe of a great battle, as the foreigners did, made life more dangerous. Even so, they were not deliberately targeted, and that made them the envy of the Chinese population. Shanghai shops saw brisk sales of the national flags of major non-belligerent nations, as Chinese residents hung them at their doorways in the hope that the sight of a Union Jack or the Stars and Stripes would ward off enemy fire, very much in the fashion that the images of guardian deities kept traditional Chinese homes safe from evil spirits. However, few had faith that anything they could do would make a difference, except running away. Desperate crowds, many uprooted from their homes in the north of the city, gathered in the International Settlement, clamoring for food. Looting soon became widespread. Crowds attacked trucks transporting rice, or smashed their way to shop supplies. The authorities were merciless in tackling the problem. On at least one occasion, French police opened fire on a crowd that had attacked a food hawker. Law enforcers in the International Settlement handed over dozens of looters to the Chinese police, knowing perfectly well they would be shot within hours.

Violence in many forms, often lethal, was meted out in liberal doses among the Chinese. An atmosphere of intense suspicion permeated the city, and everyone was a potential traitor. On the first day of fighting, six Chinese nationals were executed. All were sentenced to death for spying on behalf of the Japanese or for carrying out acts of sabotage in Zhabei and other areas under the control of the Shanghai municipal government. On another occasion, two women and seven men were decapitated for working for the Japanese. Their heads were placed on top of poles and put on display in the market square, as thousands of men, women and children watched with glee.

Following rumors published in the local press that the Japanese had bribed collaborators to poison the water supply, gangs armed with clubs and other primitive weapons raged through the streets, stopping suspicious-looking individuals. Anyone caught with a powdery substance, even medicine, was severely beaten. Fifteen innocent Chinese were killed and 40 injured that way, according to police. Even having the wrong appearance could be deadly. On the morning of August 17, an unregistered Portuguese man was beaten to death by a mob because he was thought to look Japanese. A Sikh police officer who came to his rescue was in turn badly mangled by the crowd.

One group of Shanghai residents was particularly unfortunate and unable to go anywhere, despite being directly in the middle of some of the worst fighting. They were the inmates of Ward Road Jail, Shanghai’s largest prison, located in Yangshupu. Thousands of them, along with their wardens, were trapped when the battle started. On the morning of August 17, a shell struck the prison, killing ten people and causing extensive damage to both the cells and the prison staff’s quarters. In the days that followed, the prison suffered several direct hits when Chinese artillery in Pudong or at the North Railway Station misfired.

By August 20, the penal authorities began evacuating the prisoners, starting with the criminally insane, who would pose the greatest danger if a chance grenade were to make escape possible. On August 22, a more comprehensive evacuation was planned to take place, but buses meant to bring 150 juvenile criminals to the Chinese district via the International Settlement were stopped by Japanese guards at the Garden Bridge. The juveniles were young and could be recruited for the Chinese war effort and they were returned to their prison. From then on, the evacuation drive nearly stopped, and weeks later, the Ward Road facility was still brimming with inmates, exposed to the deadly fire from both sides.


The Japanese marine units dispatched from Manchuria on August 16, the day of crisis for their compatriots in Shanghai, arrived in the city during the morning of August 18 and were immediately thrown into the battle. A few hours later, the Japanese Cabinet announced the formal end of a policy of non-expansion in China, which by that time had been a hollow shell for several weeks anyway. “The empire, having reached the limit of its patience, has been forced to take resolute measures,” it said. “Henceforth it will punish the outrages of the Chinese Army, and thus spur the (Chinese) government to self-reflect.”

On the same day, the British charge d’affaires in Tokyo, James Dodds, suggested a peace proposal to Japanese Deputy Foreign Affairs Minister Horinouchi Kensuke. The proposal, drafted two days earlier by the British, American and French ambassadors to Nanjing, called for the transformation of Shanghai into a neutral zone based on a commitment by both China and Japan to withdraw their forces from the city. Japan was not excited about the idea, and on August 19, Horinouchi presented the British diplomat with his government’s official refusal, stating that China would have to retreat to the boundaries outlined in the truce that ended hostilities in 1932. Japan was gaining confidence.

Meanwhile, there was a growing feeling on the Chinese side that important opportunities had been missed. On August 18, Chiang Kai-shek dispatched Deputy War Minister Chen Cheng, one of his main military aides, to the Shanghai front in order to confer with Zhang Zhizhong about how to carry the battle forward. The two generals reached the conclusion that rather than focusing the attacks on the heavily fortified Hongkou area, they should turn their attention to the Yangshupu district, seeking to push through to the Huangpu River and cut the Japanese forces in two. This was the decision the German advisers and the frontline commanders had been waiting for. The gloves had come off, and the self-defeating reluctance to attack Japanese troops inside the settlement borders was gone.

As the forces that had been in Shanghai since the start of hostilities were beginning to show signs of attrition, the generals decided to place the main responsibility for the attack with the 36th Infantry Division, which had only just arrived, and was being moved to the eastern side of the Hongkou salient. It was an obvious choice, as its soldiers were from the same German-trained elite as those of the 87th and 88th Divisions. Two of the division’s four regiments were ordered to attack straight south in the direction of the Huangpu, down streets running perpendicular to the river. In order to reach the wharf area, the soldiers would have to pass five heavily defended intersections. Severe casualties were expected.

The two regiments launched the attack almost immediately, moving out in the early hours of August 19. Sabotage and incendiary bombs resulted in a number of large fires that helped improve visibility during the night fighting. However, the intersections proved a problem. The Chinese soldiers, most of whom were seeing battle for the first time, became defenseless prey to Japanese infantry posted on the rooftops or in windows on the upper floors of buildings along their route. In the absence of any other cover, they often had to duck behind the bodies of those already killed. Even so, for a brief period of time, the Chinese believed they had finally managed to break the back of the hated Japanese. “I thought we could push the enemy into the river and chase them out of Shanghai,” said Zhang Fakui, watching the battle from the other side of the Huangpu.

Once they had reached Broadway, the last street running parallel with the Huangpu River, they faced the most formidable obstacle of them all. The Japanese defenders had taken up positions on top of high walls protecting the wharfs. Dislodging them was akin to storming a medieval castle. A large steel gate formed an entrance into the wharves, but it yielded to no weapon that the Chinese had brought; even the 150mm howitzers could not destroy it. Officers and soldiers tried to scale the gate, but were mowed down by enfilading Japanese machine gun fire. Also located near the river were Japanese-owned factories, many of which had been turned into veritable fortresses. One example was the Gong Da Cotton Mill at the eastern edge of the International Settlement. Again, the Chinese attackers did not possess weaponry powerful enough to penetrate the Japanese defenses there.

While the Chinese were short of large-caliber guns, the Japanese had plenty aboard the Third Fleet anchored in the Huangpu. The 36th Infantry Division was subjected to merciless bombardment, which threw several of its units into disarray. The following night, between August 19 and 20, the 88th Infantry Division for the first time showed that its ability to wage war had been so severely compromised it was, temporarily at least, unable to carry out meaningful offensive action. When ordered to attack, it moved in a belated and reluctant fashion, and got nowhere. While the Chinese were getting weaker, the Japanese were growing stronger. The marines dispatched from Sasebo arrived in Shanghai on that same night, boosting the number of marines inside the garrison to 6,300 well-armed men.

Despite a propensity to husband expensive equipment, the Chinese decided at this point to throw major parts of their new tank force into the battle. As was the case with the German-trained divisions and the air force, this was another key asset that had taken years to build up. Following the 1932 incident, when Japan had used its armor to some effect, the Nationalist government had decided to acquire its own tank arm, purchasing tanks from a variety of European nations, including Germany, Britain and Italy. As a result of these efforts, by the outbreak of hostilities in 1937, China was able to deploy the British-built, 6-ton, single-turret Vickers model in Shanghai.

The 87th Infantry Division was given disposal of two armored companies, and it lost everything. Some of the tanks had just arrived from Nanjing, and their crews had not had any time to undertake training in coordinated attacks, or even simply to establish rapport with the local troops. As a result, the tank companies were mostly left to their own devices without infantry support. The Chinese also often neglected to seal off adjacent streets when deploying their tanks, allowing Japanese armor to outflank them and knock them out. To be sure, the Japanese, too, lacked experience in coordination between armor and infantry and frequently saw their tanks annihilated by Chinese anti-tank weapons.

On August 20, Zhang Zhizhong was inspecting the Yangshupu front when he met one of his former students from the Central Military Academy, who was in charge of a tank company that was about to attack the wharves. Some of the tanks under his command had been under repair and hastily pulled out of the workshop. “The vehicles are no good,” the young officer complained. “The enemy fire is fierce, and our infantry will have trouble keeping up.” Zhang was relentless, telling the young officer that the attack had to be carried through to the end nonetheless. A few moments later the tank company started its assault. The young officer and his entire unit were wiped out in a hail of shells, many of them fired from vessels anchored in the Huangpu River. “It saddens me even today when I think about it,” Zhang wrote many years later in his memoirs.

In this battle, modern tank warfare mixed with scenes more reminiscent of earlier centuries. Wu Yujun, an officer of the Peace Preservation Corps, was manning a position in the streets of Yangshupu on the morning of August 18 when a detachment of Japanese cavalry attacked. The raid was over almost instantly, and left numerous dead and injured Chinese in its wake. The Japanese repeated the assault two more times. The third time, Wu Yujun prepared an elaborate ambush, posting machine guns on both sides of the street. As the riders galloped past, they and their horses were chopped to pieces. Apart from four prisoners, all Japanese lost their lives. The 20th century had met the 19th century on the battlefield, and won. It was a typical incident, and yet in one respect also very atypical. In the streets of Shanghai in August 1937, Chinese soldiers were far likelier to confront a technologically superior enemy than the other way around.

Many of the Chinese units arriving in Shanghai had never tasted battle before, and in the first crucial days of fighting, their lack of experience proved costly. Fang Jing, a brigade commander of the 98th Division, one of the units to arrive early in Shanghai, noticed how his soldiers often set up inadequate fortifications that were no match for the artillery rolled out by the Japanese. “Often, the positions they built were too weak and couldn’t withstand the enemy’s 150mm howitzers,” he said. “The upshot was that men and materiel were buried inside the positions they had built for themselves.”


No one was surprised that the Japanese soldiers put up a determined fight in Shanghai. Since their 1904–1905 triumph over the Russian Empire, the legend of the “brave little Jap” had become firmly established in the mind of the global public. So widespread was this view that if Japanese soldiers did not fight to the death, it was a source of genuine surprise. However, at moments of absolute frankness, the Japanese themselves could feel a need to add nuance to foreign stereotypes about their countrymen’s behavior in battle. “Our soldiers would prefer death to surrender,” a Japanese diplomat was quoted as saying, “but the majority secretly hope that they will return honorably to their own country, either wounded or unscathed.”

Foreign journalists noticed to their astonishment that there seemed to be little in the Japanese code of honor that prevented them from fleeing from a hopeless situation. One of them remembered seeing a number of Japanese soldiers run back from a failed attack during the battle of Shanghai, with the Chinese in hot pursuit. There were even rare instances of Japanese soldiers raising the white flag. The same correspondent witnessed a party of about 50 Japanese motorcyclists who had become bogged down in a rice paddy near the city and were surrounded by Chinese. They surrendered immediately without making any effort to resist.

These were minority cases. Most Japanese soldiers lived up to the high expectations placed on their shoulders at home and abroad. Physically, they tended to be short by western standards, but they were strong and capable of enduring immense hardship. This was as a result of rigorous training combined with draconian discipline, underpinned by the threat and liberal use of corporal punishment. The training was so efficient that a Japanese soldier entering the reserve never ceased to be a soldier again. In the early months of the war, American correspondent John Goette met a Japanese private in his late 30s who had just been called up from his civilian occupation as a dentist. “Hundreds of thousands like him had made a swift change from civilian life to the handling of a rifle on foreign soil,” he wrote. “Twenty years after his conscript training, this dentist was again a soldier.”

An added element in the training of Japanese soldiers was indoctrination, which came in the form of repetition of the virtues—self-sacrifice, obedience and loyalty to the emperor—which the soldiers had learned since childhood. The result was mechanic obedience on the battlefield. “Even though his officers appear to have an ardor which might be called fanaticism,” a U.S. military handbook remarked later in the war, “the private soldier is characterized more by blind and unquestioning subservience to authority.” The downside was that soldiers and junior officers were not encouraged to think independently or take the initiative themselves. They expected to be issued detailed orders and would follow them slavishly. When the situation changed in ways that had not been foreseen by their commanders—which was the norm rather than the exception in battle—they were often left perplexed and unable to act.

It could be argued that the Japanese military had few other options than to train its soldiers in this way, since to a large extent it drew its recruits from agricultural areas where there was limited access to education. It was said that for every 100 men in a Japanese unit, 80 were farm boys, ten were clerks, five factory workers, and five students. Nevertheless, reading was a favorite pastime among Japanese soldiers. Military trains were littered with books and magazines, mostly simple pulp fiction. When the trains stopped at stations, even the locomotive’s engineer could be observed reading behind the throttle. Some of them were prolific writers, too. A large number of Japanese in the Shanghai area had brought diaries and wrote down their impressions with great perception and eloquence. Some officers even composed poems in the notoriously difficult classical style.

Many Japanese soldiers grew large beards while in China, but in a twist that was not easy to understand for foreigners, they could sometimes mix a fierce martial exterior with an almost feminine inner appreciation of natural beauty. Trainloads of Japanese soldiers would flock to the windows to admire a particularly striking sunset. It was not unusual to see a Japanese soldier holding his rifle and bayonet in one hand, and a single white daisy in the other. “Missionaries have found,” wrote U.S. correspondent Haldore Hanson, “that when bloodstained Japanese soldiers break into their compounds during a ‘mopping up’ campaign, the easiest way to pacify them is to present each man with a flower.”

Many Japanese soldiers also carried cameras into battle, and as was the case with the Germans on the Eastern Front, their snapshots came to constitute a comprehensive photographic record of their own war crimes. Journalist John Powell remembered his revulsion when he saw a photo of two Japanese soldiers standing next to the body of a Chinese woman they had just raped. He had obtained the image from a Korean photo shop in Shanghai where it had been handed in to be developed. “The soldiers apparently wanted the prints to send to their friends at home in Japan,” he wrote. “Japanese soldiers seemingly had no feelings whatsoever that their inhuman actions transgressed the tenets of modern warfare or common everyday morals.”


On August 20, five Chinese aircraft were returning after another fruitless attack on the Izumo, which was still moored in the middle of the Huangpu, when they encountered two Japanese seaplanes over western Zhabei. A Chinese plane broke formation, went into a steep dive and fired a short machine gun salvo at one of the Japanese. It did not have a chance. It burst into flames and plunged to the ground. The other Japanese plane disappeared in the clouds. The entire encounter had only taken a few seconds. It was one of a series of hits that the Chinese Air Force scored during a brief period in August before it was completely subdued by its Japanese adversary.

Mitsubishi G3M medium bomber

In particular, it posed a threat to Japanese bombers, such as the highly flammable Mitsubishi G3M medium bomber aircraft assigned to striking targets in Shanghai and other cities in central China. Japan’s First Combined Air Group lost half of its medium attack planes in the first three days of the battle for Shanghai, some missing, some confirmed shot down and others heavily damaged. Their crews were particularly vulnerable, since they did not bring parachutes on their missions. From late August, the air group’s bombers were escorted by Type 95 Nakajima A4N fighter biplanes. This action amounted to a humiliating admission that China’s nascent air force was a force to be reckoned with.

Nakajima A4N

“In view of the pressing situation in the Shanghai area,” said the First Combined Air Group’s commander, “our air raids reminded me of that famous, costly assault against the 203-Meter Hill.” The battle for 203-Meter Hill had been one of the bloodiest episodes of the entire Russo-Japanese War, claiming thousands of casualties on both sides. The Chinese performance was significant enough that even foreign military observers paid attention. British intelligence, in a report summarizing military events in the middle of August, noted Chinese claims of having downed 32 Japanese aircraft. “This statement appears well-founded,” the report’s writer added.

Even so, the Chinese airmen had been mostly untested and only partly trained when the war started. Their inferiority, especially against Japanese fighters, began to tell, and they gradually disappeared from the skies over Shanghai. Their compatriots on the ground expressed frustration over the lack of air cover. “We occasionally spotted two or three of our own airplanes, but the moment they encountered enemy anti-aircraft fire, they disappeared,” said Fang Jing, a regimental commander of the 98th Infantry Division. “They were no use at all. After August 20, I never saw our planes again.”

That may have been hyperbole, but it was undeniable that the evolving Japanese air superiority proved a major handicap for the Chinese. The Chinese commanders soon realized that they had to carry out major troop movements under the cover of darkness. Japan’s domination of the skies affected everything the Chinese soldiers did and even determined when they could get food. “We didn’t eat until at night,” said Fang Zhendong, a soldier of the 36th Infantry Division. “That was the only time we could get anything. In the daytime, it was impossible to transport provisions to the frontline.”

Without fighter protection the troops on the ground were dangerously exposed. They had very little in the way of anti-aircraft weaponry, mostly 20mm Solothurn guns produced in Switzerland. However, even these weapons made next to no difference as they were primarily deployed against enemy infantry. Also, the Chinese officers were reluctant to use their anti-aircraft guns lest they reveal their positions to the Japanese aircraft. In late August when Japanese Admiral Hasegawa was asked by a Reuters journalist visiting his flagship if he was in control of the air, his reply was prompt: “Yes,” he said. “I believe so.”

Civil War and the New Imperial Army I

The Toba-Fushimi fighting marked the opening battle of the Boshin (dragon) Civil War, named after the Chinese zodiacal cyclic character that designated the year 1868. The new army fought under makeshift arrangements with unclear channels of command and control and no reliable recruiting base. Samurai and kiheitai units paradoxically were fighting in the name of the throne, but they did not belong to the throne. To correct this anomaly and defend the court, which was in open rebellion against the shogunate, in early March 1868 the newly proclaimed imperial government created various administrative offices, including a military branch. The next month it organized an imperial bodyguard, about 400 or 500 warriors, composed of Satsuma and Chōshū units augmented by veterans of the Toba-Fushimi battles, yeomen, and masterless warriors from various domains, who reported directly to the court. The imperial court next notified domains to restrict the size of their local armies and contribute to the expenses of a national officers’ training school in Kyoto.

Within a few months, however, authorities disbanded the ineffective military branch and the imperial bodyguard, which lacked modern equipment and weapons. To replace them, in April authorities established the military affairs directorate, composed of two bureaus: one for the army, one for the navy. The directorate drafted an army organization act based on manpower contributions from each domain proportional to its respective annual rice production. This conscript army (chōheigun) integrated samurai and commoners from the various domains into its ranks.

As the Boshin Civil War continued, the newly formed military affairs directorate had expected to raise troops from the wealthier domains. In June 1868 it fixed the organization of the army by making each fief responsible, at least in theory, for sending to Kyoto ten men per each 10,000 koku of rice the domain produced. The policy put the government in competition with the domains to recruit troops, a contradiction not remedied until April 1869 when it banned domains from enlisting soldiers. The quota system to recruit government troops, however, never worked as intended, and the authorities abolished it the following year.

Meanwhile, in mid-March 1868 Prince Arisugawa took command of the Eastern Expeditionary Force as loyalist columns pushed along three main highways toward the shogun’s capital at Edo (present-day Tokyo). Skirmishes involving a few hundred warriors on either side brushed aside bakufu resistance, and the columns swiftly converged on the capital. The advancing army continually proclaimed its close bond with the imperial court, first to legitimize its cause; second, to brand enemies of the government as enemies of the court and therefore traitors; and third, to gain popular support.

For food, supplies, horses, and weapons, the government army established a series of logistics relay stations along the three major thoroughfares. These small depots stocked material supplied by local pro-government domains or confiscated from bakufu agencies, senior retainers of the old regime, and anyone opposing the government. The army routinely impressed local villagers as porters or teamsters to move supplies between the depots and the frontline units. Japan’s largest merchant families also contributed money and supplies to the new army. The Mitsui branch directors in Edo, for example, donated more than 25,000 ryō (US$25,000) as insurance to protect their storehouses from pro-bakufu arsonists and probably government troops as well.

Government propaganda teams accompanied the army to extol the new imperial government’s virtues and to attract adherents by offering an immediate halving of taxes on rice harvests. To complement the effort, the army issued regulations governing conduct. All ranks would share the same food, accommodations, and work details; troops would immediately report anyone spreading rumors that might lower morale; quarreling and fighting in camp were forbidden; attacks against foreigners were strictly prohibited; and commanders were supposed to prevent arson, plunder, and rape from tarnishing the new government’s image. The results were mixed. If the populace cooperated, they were treated fairly, which meant the men might be persuaded to work as military porters or laborers, the villages to donate food, and promises made to cooperate with the government army. But the standard tactic to combat the roving bands of Tokugawa supporters who harassed government columns with hit-and-run attacks was to burn nearby homes to deny the guerrillas shelter. Suspected collaborators were summarily executed.

As the government troops pushed into bakufu strongholds, coercion replaced persuasion. Soldiers requisitioned food; confiscated weapons, valuables, and cash; and impressed villagers for labor details. Although government orders prohibited arson, it was an effective tactic during battle and for pacification purposes. Uncooperative villages risked being burned to the ground, likely because soldiers understood that inhabitants feared arson above all other forms of retribution. In extreme cases, such as the final northeastern campaign, government troops torched more than one-third of the homes in the Akita domain. To avoid that fate, villagers along the army’s route-of-march provided commanders with food, supplies, and intelligence. But this cooperation was based on little more than extortion and did not indicate a sudden shift in allegiance to the new government.

By the time the main loyalist forces reached Edo in early May, Saigō Takamori had already negotiated a peaceful surrender of the city with bakufu agents, the shogunate being divided internally between hard-liners and those favoring an accommodation with the new government. The 41-year-old Saigō stood almost six feet tall and weighed about 250 pounds, making him a giant by Japanese standards. After a decade as a minor provincial official, in 1854 Saigō moved to Edo to promote Satsuma policies. Four years later reactionary shogunate officials forced him to flee to Satsuma, but he was then exiled. After his pardon in 1864, Satsuma officials sent Saigō to Kyoto to handle the domain’s national affairs.

There was no denying Saigō’s ability, but the man was an enigma, given to lengthy silences that could be interpreted as contemplative wisdom or hopeless stupidity. His indifference to awards, honors, or material trappings, complemented by his dynamic charisma and humanism, made Saigō the most respected personality in early Meiji Japan. His deal with the bakufu, however, had enabled more than 2,000 warriors loyal to the shogun to escape from Edo, and the guerrilla war these reactionaries were waging against the loyalists was ravaging the nearby countryside.

North of Edo pro-bakufu forces held Utsunomiya castle; in early June, government forces defeated the shogun’s troops in a series of minor engagements between Utsunomiya and Edo. They withstood a subsequent bakufu counteroffensive and, strengthened with reinforcements, occupied the castle to secure the northern approaches to Edo. The vanquished bakufu units fled to northern Japan.

Meanwhile, loyalist troops had garrisoned the shogun’s capital without opposition, but their efficiency and morale slowly disintegrated as the provincial troops settled in to the comforts and fleshpots of the big city. As martial skills eroded, Saigō worried about the diehard pro-shogun radicals who retained de facto control of the city. The most powerful of these bands was the Shōgitai (League to Demonstrate Righteousness), formed in February 1868, which eventually enrolled about 2,000 warriors, each sworn to kill a Satsuma “traitor. ”

Operating from its headquarters in Edo’s Ueno district, Shōgitai units selectively cracked down on the roving criminal gangs that had turned the nighttime capital into a place of robbery, murder, and extortion. Besides punishing these outlaws they also ferreted out Satsuma informers and pro-court spies, and while the government army slipped into idleness, Shōgitai units busily constructed strongholds around Ueno Hill. The ongoing fighting in the north and the deteriorating conditions in Edo created the impression that the new government was unable to control the strategically vital Edo region. When reports of the impasse in Edo reached the Kyoto government, leaders Ōkubo Toshimichi (with Saigō a central leader of Satsuma since 1864), Kido Koin (leader of Chōshū with Takasugi from 1865), and Iwakura Tomomi (a high-ranking court noble) sent Ōmura Masujirō to restore government control in the city and eliminate the Shōgitai influence.

Compared to Saigō, Ōmura seemed physically clumsy; and, unlike the gregarious Saigō, his introverted personality and perpetually sour countenance attracted few friends, much less casual admirers. Saigō’s patient wait-and-see style and measured diplomacy left Ōmura seething at the sight of a seemingly powerless government army standing idly by while the Shōgitai incited antigovernment violence in Edo. Ōmura demanded action, but Saigō’s disinclination to turn the city into a battleground was a major reason for his lenient terms with the shogun’s representatives.

Assured from Kyoto that reinforcements and money were on the way, Ōmura and his lieutenant Eto Shimpei decided to attack. They believed that the Shōgitai were few in number and ignored Satsuma commanders’ arguments that more loyalist troops were needed to stabilize the city. Moreover, Ōmura was certain that his artillery would rout the pro-shogunate forces. His only concession was to the weather; he postponed the offensive until the arrival of the rainy season rather than bombard the city during the dry weather, when Edo’s wooden buildings would burn like tinder.

On July 4 Ōmura summoned Saigō and ordered him to attack the Shōgitai’s Ueno strongpoint immediately, overriding Saigō’s objections with a casual wave of his fan. As Saigō and his captains had predicted, the attack ran headlong into a ready-and-waiting enemy, the Shōgitai having been alerted to the impending assault by Satsuma deserters. Although outnumbered two to one, the approximately 1,000 Shōgitai troops fought from behind well-prepared defenses that were anchored by a small lake on one flank and thick woods on the other and could only be taken by a frontal assault.

Ōmura expected that artillery fire would drive the Shōgitai from their shelters, leaving them vulnerable to Saigō’s infantry. From Edo castle, Ōmura watched the billowing smoke and heard the loud explosions and thought that his plan had succeeded. But the artillery guns soon malfunctioned, and the few that did fire were wildly inaccurate, producing noise and fireworks but few enemy casualties. Saigō’s vanguard, led by Kirino Toshiaki, charged directly into the Shōgitai’s barricades, losing at least 120 men killed or wounded. About twice that number of Shōgitai supporters were slain, although many more were apprehended fleeing from Ueno.

The victory secured Edo for the new government, crushed one of the largest and most violent antigovernment units, restored the momentum of the imperial forces by releasing them to move north, and, by showing that the pro-Tokugawa forces could not defeat the emperor’s army, calmed fears of a prolonged civil war. Despite the artillery fiasco and the heavy government casualties, Ōmura emerged as a hero, acclaimed for his grasp of modern military science, much to Saigō’s chagrin, who had lost face over the incident. Formidable resistance continued in northern Japan, but by controlling Edo the army had turned a corner. To signify the bakufu’s demise, the next month the capital moved from Kyoto to Edo, which had been the de facto political center of Japan for 250 years. On September 3 Edo was officially renamed Tokyo, and the next month Emperor Mutsuhito adopted the reign name of Meiji and henceforth was identified as the Emperor Meiji.

The Northern Campaigns

On June 10, 1868, two Aizu samurai attacked a senior Chōshū officer in a Fukushima brothel. Desperate to escape, he jumped through a second-floor window only to land face-first on a stone walkway, where he was seized by his assailants and later executed. Among the murdered official’s personal effects were confidential government plans to subjugate the northern domains, including the Tokugawa stronghold of Aizu. Reacting to these threats, twenty-five pro-bakufu vassals in northern Honshū formed a confederation to resist the Satsuma-Chōshū government, as distinct from the imperial government. To crush the league, Ōmura devised a complicated strategy to capture the city of Sendai on the Pacific (east) coast and then send converging columns southward to attack the Aizu rebel stronghold at Wakamatsu castle from the rear. Landings in Echigo Province on the Japan Sea (west) coast with probing attacks against the major passes would fix the defenders in place.

Yamagata Aritomo commanded the 12,000-man-strong government army contingent that landed along the west coast in late July and quickly captured Nagaoka castle. Then he hesitated, even though his veteran forces (leavened with about 1,000 kiheitai and Satsuma warriors) outnumbered the rebels three to one. Yamagata spread the entire army along a 50-mile-long defensive front, leaving him no reserve. He parceled out artillery, giving each unit a few guns but not enough firepower to blast through league fortifications. When he tried to evict rebel defenders from strategic mountain passes, the army suffered successive reverses. Thereafter the rainy season made streams and rivers impassable and restricted campaigning. By that time, Yamagata was barely on speaking terms with his second-in-command, the veteran Satsuma commander Kuroda Kiyotaka, who spent much of the campaign sulking in his tent.

The opposing armies faced each other about 6 miles north of Nagaoka, with their flanks bounded on the west by a river and on the east by a large trackless swamp. Stuck in their cold, wet trenches, morale among government troops plummeted, abetted by rebels chanting Buddhist funeral sutras throughout the night. Meanwhile, battle-hardened, aggressive pro-Tokugawa commanders executed an active defense, probing and raiding government outposts and disrupting their rear supply areas.

On the east coast, operations initially went smoother. The main government army moved overland from Edo on Shirokawa castle, which dominated a strategic mountain pass leading to Aizu. By mid-June government troops had taken the castle in a frontal attack coordinated with columns enveloping the town. Government artillery destroyed the prepared defenses and routed the 2,500 defenders. Several smaller domains in the league promptly capitulated. The hardcore Aizu defenders withdrew into their main defenses, which were constructed to blend into the rugged mountainous terrain. With Yamagata’s western campaign stalled, there was little pressure on the Aizu rear, enabling the rebels to concentrate their forces against the eastern prong of the government offensive. The Kyoto high command sent reinforcements under Saigō’s leadership to reinvigorate Yamagata’s operations.

Unable to break into the southern approaches to Aizu–Wakamatsu castle, government troops commanded by Itagaki Taisuke and Ichiji Masaharu maneuvered north along the main highway and in mid-October captured the northern outposts protecting Aizu’s rear areas. They then pivoted west through the mountains and defeated successive Aizu contingents by concentrating artillery fire to pin down defenders while small bands of riflemen turned their flanks. When the rebels shifted troops to protect their vulnerable flanks, loyalist forces smashed through the weakened ridgeline defenses. By late October Itagaki and Ichiji were within five miles of Wakamatsu castle.

Around the same time, Saigō seized Niigata city, a major port on the Sea of Japan about forty miles north of Nagaoka castle. Capturing the port cut off rebel access to imported foreign-manufactured weapons and interdicted a major supply route running from the city to Aizu-Wakamatsu. But the day Niigata fell, rebel troops farther south capitalized on their knowledge of the local terrain to cross the supposedly impassable swamp, outflank the main government lines, and then overwhelm the small government garrison at Nagaoka castle. With government forces divided and their line of communication threatened, Yamagata fled for his life, allegedly discarding his sword and equipment in his haste. Saigō countermarched south and retook the castle five days later. Yamagata, however, failed to cut off the withdrawing rebel forces, and the western campaign stalled again.

These setbacks and the slow progress on the eastern front sowed doubt among government leaders in Kyoto that the army could finish the campaign before the region’s heavy winter snows made further campaigning impossible. Satsuma troops, which formed the backbone of the army, came from southern Japan and were neither acclimated nor equipped for winter operations. Rather than risk the imperial army’s carefully crafted reputation for invincibility and its best troops, Ōmura suspended eastern operations against Aizu until the following spring.

Itagaki and Ijichi ignored Kyoto’s orders and unilaterally assaulted the main Aizu stronghold, forcing the heavily outnumbered Aizu samurai to commit their entire reserve. One such unit, the White Tiger Brigade, consisted of a few hundred 16- and 17-year-olds and suffered severe losses. Retreating toward Wakamatsu castle, sixteen survivors mistakenly assumed that the thick black smoke and red flames rising from the adjacent castle town meant the castle had fallen. With all hope apparently lost, they committed suicide, an act of sincerity that redeemed their treason and apotheosized the White Tiger Brigade as a symbol of loyalty and selfless courage that resonated powerfully among the public.

Once the main Aizu defenses collapsed, the army overran minor strongholds in rapid succession and rebel survivors fled to Wakamatsu castle. Although the government artillery could not penetrate the thick castle walls, the defenders were unprepared for a winter siege and surrendered in early November. Losses of pro-Tokugawa forces for the entire campaign were around 2,700 killed. Sporadic fighting continued in northern Japan until mid-December, when the league finally capitulated and its leaders took responsibility for defeat by committing ritual suicide.

During the Tokugawa period suicide to atone for mistakes or defeat was an accepted cultural norm among the warrior class. During the Boshin Civil War warrior customs regarding fallen enemies or prisoners likely encouraged battlefield suicides. Both armies routinely cut off the heads of dead or wounded enemies for purposes of identification or morale-enhancing displays. In one case a bakufu leader’s severed head was brought to Kyoto for public display. The gruesome practice was widespread, and western doctors working in Japan reported that they rarely saw wounded enemy soldiers, apparently because of the criteria for head-taking and surrender.

Surrender was recognized, provided both sides agreed on terms for capitulation. During the Boshin War’s early stages, potential prisoners were usually executed for suspected cowardice for surrendering (the idea of the shame of surrender) or because their wounds testified that they had resisted government troops. Suspected spies or bakufu agents were summarily executed. Many Shōgitai prisoners captured at Ueno fit all categories and were promptly beheaded. In the later stages of the war, the new army accepted surrenders of pro-bakufu troops as the government realized that reconciliation was necessary to unify the nation.

This conciliatory attitude did not carry over to the disposition of the dead. The government honored its war dead with special ceremonies that Ōmura later institutionalized by establishing the Shōkonsha (shrine for inviting spirits) in June 1869 as the official shrine to commemorate government war dead. He hoped that official memorialization of the war dead would stimulate a popular national consciousness by enshrining the concept of an official death for the sake of the nation, not merely a private death in a meaningless vendetta.

Enemy dead were regarded as traitors and ineligible for enshrinement in government shrines. Their beheaded corpses were often left where they fell, and local villagers would furtively bury the remains. Partly because the Aizu warriors had so fiercely resisted the government army and partly owing to Aizu’s long-standing enmity with Chōshū, the Meiji leaders forbade even the burial of Aizu corpses and ordered them left to rot in open fields.

At the end of 1868, Tokugawa supporters still controlled Ezochi (Hokkaidō). In October 1868, bakufu ground and naval units from Sendai had landed on Hokkaidō’s south coast and overwhelmed the undermanned army garrison. The remaining government troops abandoned the island, fleeing to the safety of nearby Honshū. In the spring of 1869, reconstituted and much-strengthened government ground and naval forces returned and converged on the major rebel stronghold at Hakodate. The outnumbered pro-shogunate rebels fell back and by early May prepared for a final stand from Hakodate’s pentagon-shaped fortress.

Constructed by the bakufu between 1857 and 1864 to protect Ezochi from the Russian threat, the first western-style fortress in Japan defended Hakodate’s harbor, where the three remaining bakufu warships were sheltered. During naval battles on May 11, one government ship sank after a shell exploded its powder magazine, but two bakufu warships ran aground, and the third was previously damaged. The subsequent fighting for the fortress produced more sound and flash than bloodshed. Thousands of government artillery shells fell on the fortress, causing three casualties. Government losses were two dead and twenty-one wounded. The siege did reduce the rebels to near starvation, forcing them to surrender on May 25.

Civil War and the New Imperial Army II

After the fall of the fortress, one company of government troops remained in Hokkaidō, but it did little more than police Hakodate city. In 1870 the colonization office responded to increasing friction with Russia over Sakhalin Island by relocating samurai families from Tokyo to Hokkaidō to create self-sufficient military communities organized around units composed of farmersoldiers. Saigō Takamori enthusiastically backed the plan and in 1873 moved more displaced samurai to Hokkaidō to reinforce the garrison, provide jobs for otherwise unemployed warriors, and open the northern island to development. An Ainu (aboriginal) uprising spurred cries for a permanent garrison, but the sparse Japanese population made it impossible to conscript enough soldiers locally to meet troop personnel requirements.

Authorities relied on the tondenhei system promoted by Lt. Gen. Kuroda, concurrently vice director of the colonization office. Beginning in May 1875, the government gave settlers (ex-samurai and army reservists) a small parcel of land, which became their property if they cultivated it for three years. Most of the thirty-seven tondenhei settlements were located along a 120-mile trace running north from Sapporo and protecting the northwest side of the island. Despite Kuroda’s notions about an ideal warrior-farmer ready to take up arms to defend his home and family, Japanese warriors were not farmers, and few wanted to emigrate to faraway Hokkaidō. As late as 1905 Hokkaidō had only about 4,000 regular troops and about 1,000 reservists.

An Assessment

Throughout the northern campaigns of 1868–1869, the government army outnumbered rebel forces and enjoyed overwhelming material and technological superiority. Cannons and rifles, not samurai swords and spears, decided the outcome of the Boshin Civil War. Firepower may have relegated traditional samurai weapons to the scrapheap, but paradoxically the Tokugawa forces consistently displayed superior élan and fighting spirit and atoned for their defeats with collective suicides. In short, antigovernment forces exhibited the type of battlefield behavior and morale that the government rarely saw in the new army’s conscript soldiers.

The campaigns highlighted major differences between field commanders and central headquarters. Army headquarters in faraway Kyoto often proposed plans at odds with the local conditions, and as tensions increased, line officers ignored central direction in favor of unilateral action. Lacking a strong central staff capable of enforcing orders, the army was at the mercy of individual commanders for leadership and direction. Similarly, the absence of unified tactical doctrine and disputes over appropriate tactics and doctrine between line and staff left units to fight according to the tactics favored by their respective commanders. Resentment flared because the nobility and Chōshū and Satsuma samurai monopolized senior army posts. Within the ranks the composition of the new government army, especially its use of commoners, created resentment among soldiers of samurai lineage. Although the restoration wars would later take on a romantic hue filled with swashbuckling samurai, the new government’s military success in the Boshin Civil War left a residue of disgruntled warriors, marginalized commoners, and a torn social fabric.

Reorganizing the Army

Saigō, Ōmura, Ōkubo Toshimichi, and other military leaders of the restoration were divided over the new army’s organization. Saigō was enigmatic, leading the new army yet retaining strong military ties to his regional base in southern Japan. Ōmura sought a strong central government at the expense of the regional domains and recommended a national conscription system to build a standing army under the new government’s direct control. He also proposed European-style reforms for the new army and the abolition of traditional warrior-class privileges such as carrying swords. Ōkubo, one of the most powerful leaders of the new regime, wanted a samurai army, and proposals that filled the national army with commoners and peasants while they eliminated samurai privileges repelled him.

Ōmura and Ōkubo did share a deep-rooted fear about the survival of the imperial government that they had created. Danger seemed everywhere as the new, unstable society seethed with sedition and treachery. In early January 1869, for example, six sword-wielding anti-foreign assassins murdered a senior councilor in broad daylight on a major Kyoto thoroughfare. Riots and peasant uprisings underscored rural instability, disgruntled samurai were suspect, and foreigners might overwhelm Japan as they had China. The authorities first sought to control the reactionary warriors, who represented the immediate armed threat to their new government, by counterposing a military organization that relied on conscription and indoctrination to inculcate soldiers’ loyalty to the government and emperor.

In order to educate junior officers, in 1868 the court and council of state (dajōkan) established a school of military science in Kyoto on the site of the former French training ground. Enrollment was initially restricted to sons of the nobility and government officials. The following year the government converted the former shogunate’s Yokohama Foreign Language School into a French-style military academy to train samurai from Chōshū and elsewhere for an invasion of Korea. In July 1869, however, the government reorganized the military directorate into the military affairs ministry under the titular command of Prince Yoshiaki. Ōmura served as vice minister of military affairs with responsibility for training and organization.

Ōmura refused to perpetuate the samurai monopoly on warfare. He first incorporated the school of military science into the military affairs ministry and then in September transferred the Kyoto facility to Osaka, the country’s maritime and overland transportation hub. Osaka’s strategic central location would allow government troops to move rapidly in any direction to suppress antigovernment uprisings. The decision also shifted the military locus of power westward to check Satsuma’s growing influence. Ōmura next constructed a French-style maneuver area and barracks cantonment for a battalion of Chōshū troops as well as a major arsenal. His ambitious five-year plan anticipated a revised conscription system, standardized equipment, military academies to train noncommissioned officers and educate professional officers, and a new army force structure.

With institutional army reforms in progress, in September 1869 Ōmura traveled to Kyoto to inspect the new school and stayed at a nearby inn. In Kyoto’s late summer twilight, assassins stormed into the two-story latticed wooden inn and mortally wounded Ōmura in a wild sword-swinging melee. His two lieutenants jumped through a paper lattice window into the alley, only to be cut down by waiting gang members. Believing that one of their bloodied victims was Ōmura, the murderers fled. Ōmura succumbed to his wounds in an Osaka hospital in early November.

Twenty-four years later the army officially dedicated a statue to Ōmura during ceremonies held at the Yasukuni Shrine in 1893. Ōmura’s statue, the first western-style bronze work commissioned in Japan, still rises over the entrance to Yasukuni, garbed in samurai finery and bearing two swords. The period representation, however, misrepresents the man who was determined to build an army of commoners and peasants and was a driving force in the abolition of samurai privileges, including their right to bear swords. It is faithful in one respect. Ōmura looks out at Ueno Park, scene of his greatest military victory, and one that ensured the Meiji ascendancy.

An Army in Turmoil

The attack on Ōmura shocked the fledgling Meiji government. Acting on a nationwide alert, police quickly apprehended the killers, who turned out to be reactionary samurai from northern Japan and disgruntled former kiheitai members. Ōmura’s death threw the army’s leadership into turmoil with major implications for the institution’s future. Maebara Issei succeeded Ōmura but soon resigned after quarreling with Kido’s decision to forcibly suppress rebellious kiheitai units (and subsequently quit the government altogether in September 1870 for reasons described below). Itagaki Taisuke was nominated to fill the vacancy, but he did not get along with Ōkubo, who vetoed his selection. Kuroda Kiyotaka, Yamagata’s nemesis during the Boshin War, was unavailable, having departed the army in 1870 to oversee Hokkaidō’s development. This left Yamagata Aritomo, who had promptly returned to Japan from his European inspection tour after learning of the attack on Ōmura. In August 1870, the council of state appointed Yamagata minister of the military department, a post that had been vacant for almost one year.

Yamagata’s remaining rival within the army was Maj. Gen. Yamada Akiyoshi, the deputy minister of the military department, known as the Little Napoleon because of his planning skill as the Eastern Expeditionary army’s chief of staff during the Boshin War. Critics, however, dismissed him as a poseur. Shortly after Yamagata took control of the army, Yamada joined the Iwakura Mission that departed Japan in November 1871 for an extended inspection tour through the United States and Europe with a delegation of Japan’s most prominent statesmen. Yamagata’s last serious rivals, Itagaki Taisuke and Saigō Takamori, quit the government in 1873 over the Korea controversy. Itagaki became involved with the liberty and people’s rights movement, and Saigō, an army general, commander of the Imperial Guard, and concurrently a senior councilor, became a magnet attracting disaffected samurai.

With his major competitors gone, Yamagata would dominate the army and cultivate a clique of Chōshū officers to consolidate his power base. Due to his patronage, by 1888 sixteen of forty-two general officers were of Chōshū lineage. With the exception of imperial family members, men from Satsuma and Chōshū would monopolize the top positions in the army until 1907. Protégés such as Katsura Tarō, Kawakami Sōroku, Kodama Gentarō, and Terauchi Masatake, among others, would lead the army’s next generation and also rise to the highest civilian cabinet positions. Yamagata eradicated localized loyalties that might threaten central authority while he created his personality-based regional power base.

As Yamagata consolidated his position, the same regional and individual loyalties that had provided the glue that held the government army together during the restoration wars became suspect as impediments to the new government’s goal of a unified Japan. Central authorities rapidly disbanded the independent military organizations of the various domains in favor of a national army, but regionalism and factionalism dominated the selection and training of the army’s future leaders.

In 1869, ostensibly for financial reasons, the government reduced the size of the quota-based conscript army that it had assembled from various domains. Government authorities questioned the conscripts’ reliability, and domain leaders disliked their troops serving as long-term conscripts in a military force outside their control. That December the government eliminated the Chōshū kiheitai and similar units as part of an army-wide reorganization.

Younger men replaced the older veterans when the army dismissed soldiers over 40 years of age or in poor physical condition. Army authorities raised officers’ pay but lowered it for the more numerous rank-and-file to save money. The discarded veterans were the foot soldiers who had borne the brunt of the fighting in the restoration wars. In return they received miserly pensions and no rewards for their service. Kiheitai veterans, especially peasants and townsmen, vigorously protested plans to cut the 4,000-man force by half and the associated reforms.

Reactionaries played on these grievances to foment armed revolt in Chōshū. Insurgents opposed the dissolution of volunteer units, resisted westernization of weapons and equipment, and refused to obey their officers’ commands. In January 1870 about 1,800 disaffected former kiheitai members, more than 70 percent of them peasants, attacked Yamaguchi castle, surrounded the government’s administrative offices, and temporarily seized power. Peasants and merchants, aggrieved by poor harvests and runaway inflation, joined former government soldiers to raid government offices and stockpiles. Peasant revolts also broke out in nearby areas, forcing Kido, one of the government’s three senior councilors (Saigō and Ōkubo were the others) to mobilize loyal units in Chōshū and elsewhere to suppress the uprising by early February 1870. About 200 men on both sides were killed or wounded in the fighting, but the government executed an additional 130 rebels. The harsh sentences were aimed less at disaffected samurai than at cowing a possible peasant uprising that Kido feared would quickly sweep across Japan.

Building the New Army

Most of the new government’s leaders favored some form of national conscription to create a centralized military force to maintain domestic order, but there was a fundamental question about who was qualified to serve in the army. Yamada had originally allied himself with the late Ōmura’s Chōshū followers in the military department to deflect the attempts of Ōkubo, Kuroda, and other Satsuma leaders to create a professional military caste. In March 1870 the government instituted a revamped conscription system based on rice production, requiring each domain to maintain 60 troops per 10,000 koku of arable land production. The abolition of domains in late August 1871 simultaneously swept away the administrative infrastructure underpinning the conscription system, although a quota of five men from any class between the ages of 20 and 30 for every 10,000 koku of the new prefecture’s harvested rice stayed in effect.

Meanwhile, in March 1871,Yamagata, Iwakura, Saigō, and Kido had organized samurai from the three most powerful loyalist domains—Chōshū, Satsuma, and Tosa,—into an Imperial Guard to protect the throne and replace the diminished national army. The emperor donated 100,000 ryō to underwrite the new unit, which was directly subordinate to the court. The 6,200-man imperial bodyguard consisted of infantry and artillery units with a few cavalry squadrons and doubled as a national army, marking the beginning of modern Japan’s military institution.

More institutional changes soon followed. The military department was reorganized in July 1871, and on August 29, simultaneous with the decree abolishing the domains, the council of state ordered local lords to disband their private armies and turn over their weapons to the central government army. Though Yamagata played on the foreign threat, especially Russia’s southward expansion, to justify a national army, the government’s immediate perceived danger was domestic insurrection. Consequently, on August 31, with the court’s approval, the military department accordingly divided the country into four military districts, each with its own garrison or chindai, to deal with peasant uprisings or disaffected samurai insurrections. The Imperial Guard formed the Tokyo garrison whereas conscripts filled the ranks at the Osaka, Kumamoto, and Sendai garrisons. The four garrisons marshaled about 8,000 troops—mostly infantry, but also a few hundred artillerymen and engineers. Smaller detachments guarded outposts at Kagoshima, Fushimi, Nagoya, Hiroshima, and elsewhere. By late December 1871 the military department set army modernization and coastal defense as priorities. Yamagata and his deputies devised long-range plans for an army to maintain internal security, defend strategic coastal areas, train and educate military and naval officers, and build arsenals as well as supply depots. Despite Yamagata’s previous heated rhetoric about the foreign menace, little serious planning was directed against Russia.

In February 1872 the military department was abolished and separate army and navy departments were established. The Imperial Guard also underwent several reorganizations, and in January 1874 Yamagata became the Guard commander and concurrently vice army minister. Although created to defend the home islands against foreign aggression, the elite unit’s principal mission was to protect the throne by suppressing domestic samurai revolts, peasant uprisings, and antigovernment demonstrations.

No one, however, held high hopes for a quota-based, conscript-manned garrison system. By the summer of 1871, the army had inducted about 1,500 conscripts but found 25 percent of them to be physically unfit for military service. The remaining 1,100 or so were divided into various branches—infantry, cavalry, artillery, construction, and a bugle corps—collectively known as the Osaka unit.

The conscripts found themselves in a new, much more westernized world. They were forbidden to wear Japanese-style swords, although they could attach western-style bayonets to their uniform belts. New diets introduced the conscripts to meat, a change from the traditional rice and vegetable staples and one that caused indigestion and worse. In theory the conscripts received standardized uniforms, but a contemporary newspaper account described a mixed bag of conscripts, some wearing wooden clogs, others straw sandals, and a potpourri of tunics and jackets. Even more distressing, some were talking or reading books while on duty. One senior officer later recalled inspecting otherwise resplendent but bare-footed conscripts. In short, garrisons suffered from lack of weapons, money, authority, and personnel recruiting.

Military Training and Education

In early 1874, under French direction, the army relocated its small-arms firing ranges to the estate of a former shogunate official. The new home of the Toyama Infantry School (founded in 1873) taught minor tactics, marksmanship, bayonet practice, and physical education designed to prepare junior officers for command assignments. The eight-month course eventually standardized infantry doctrine and officer training throughout the army. The army concurrently established specialized technical schools in Tokyo for ordnance, military construction, and high explosives under the arsenal bureau’s direction as well as veterinary science, farriery, and equestrian schools. Each had around sixty students, and by 1875 numerous French-designed training facilities, rifle ranges, and specialized schools were active in Tokyo.

The army invested heavily in officer and noncommissioned officer (NCO) education. The reorganized school of military science opened in January 1870 with fifty-seven students enrolled in junior officer or NCO courses. The same year the government ordered each prefecture to send cadets in numbers proportionate to their respective rice production, and thereafter classes annually averaged about 100 students. The Yokohama facility also relocated to Osaka in 1870.

The Osaka school was renamed the Army Military Science School in November 1871 and moved to Tokyo in early 1872, where the army reorganized it into three provisional sections: a preparatory school with a middle school curriculum plus instruction in western languages, in effect a junior cadet school; a military academy that taught branch technical skills; and an NCO academy, which was removed from the cadet school the following year. In October 1874 the officers school formally became the military academy, and the following May the preparatory school was designated the cadet academy. The military science school was abolished. In January 1877 the cadet school was amalgamated into the military academy.

Military academy classes originally enrolled two types of cadets, regular (graduates of preparatory academies or those qualified by examination) and provisional (graduates of the NCO academy or selected NCO volunteers under 27 years of age). The former had a three-year course, the latter about half that, although both received regular commissions. The curriculum covered ordnance and weapons, equestrian skills, marksmanship and physical training, foreign language (French), and records administration. Almost all of the original students were of samurai stock, with just 13 commoners versus 719 warriors enrolled in the 1872 classes. One of Yamagata’s first decisions after taking over the military department, however, was to open the school of military science to commoners, and by 1881 the numbers had tilted in favor of commoners, whose 258 cadets outnumbered the 158 from warrior families.

Former samurai had as much trouble with the new military discipline as peasant conscripts. When the school of military science prohibited officer candidates from wearing swords and encouraged cadets, who were overwhelmingly from the samurai class, to cut their traditional topknot, many quit. Those who stayed had to adjust to wearing western-style clothing, which was as uncomfortable for them as it was for conscripts. Pants chafed, and boots and shoes caused blisters and were generally uncomfortable.

Even time changed because the military academy used the western measurements of minutes and hours to govern curriculum and schedules. Freshly commissioned officer graduates in turn drilled their conscript soldiers with the new way to measure time, and the draftees brought the concept back to their villages, factories, or offices. By the early 1870s senior officers carried pocket watches, an acknowledgment that coordination of troop movements, training schedules, and timetables made the clock a technological reality of military life.

Civil War and the New Imperial Army III

The military academy aimed to educate officers in western military science and tactics, but Japan lacked schools that taught foreign languages, western pedagogy, or natural sciences. Educational reforms of 1872 established four years of primary schooling and three of middle school, but attendance was not compulsory and most students did not graduate from primary school. Until the military preparatory schools produced their first graduates, the military academy operated on a transitional basis. During the three-year interval the provisional academy trained former samurai in French doctrine, including battalion echelon maneuver, drill, squad regulations, and the basics of field fortification. In the best of circumstances, which these were not, it would take a decade to educate youths for specialized military training.

In January 1875 the relocated military academy opened at Ichigaya (a former daimyō residence in Tokyo) under the direct supervision of the army ministry. Altogether 158 students enrolled in either a two-year course (for the infantry and cavalry branches) or a three-year curriculum (for the more technically complex artillery and engineer branches). Upon graduation, cadets were commissioned as second lieutenants. Under the tutelage of thirteen French instructors (including eleven military officers), cadets studied French, tactics, military organization, and debate. The army adopted French army doctrine and tactics because of the availability of French military instructors in Japan (who had been employed by the shogun), Japanese officers’ acquaintance with the French language through the Yokohama language school, and their familiarity with the French military system. Until 1877 the army relied exclusively on French military advisers, whose numbers reached a high of forty-three personnel in 1875.

Students who enrolled in the NCO academy’s twelve-month (later fifteen-month) course graduated as corporals eligible for competitive examination for entrance to the military academy. Initial confusion about the difference between the NCO academy and the cadet school led to cases of students entering both or dropping out of the NCO academy to enter the cadet school. Many selected that option, including three who later rose to the rank of full general. Others accepted NCO rank and with their academy colleagues would become the backbone and small-unit leaders of the new national army.

The reorganized NCO academy offered a two-year course of instruction exclusively for noncommissioned officer candidates. Conscripts volunteered for the course and required their regimental commander’s recommendation to attend. After graduation they were promoted to NCO rank. The army closed the academy in 1889 because it was attracting few volunteers, was soliciting men who had not been conscripted, and was producing lackluster graduates. The other source of NCOs was conscripts who volunteered for additional service after completing three years of active duty. Their respective regimental commanders had to endorse the application. Between 1889 and reforms in 1927, potential infantry branch NCOs learned their craft within the regiment through on-the-job training and by observing and learning from the regiment’s senior NCOs. Other branches had technical schools where their NCOs continued to go for training and instruction.

French instructors at the military academy relied on detailed explanations in exercises and drills, not innovation, to solve set-piece tactical problems. The order of battle rigidly divided formations into skirmishers, main force, and reserves. Officers learned their specific responsibilities within each echelon. The battalion (about 800 men) was the tactical unit for purposes of instruction, so little attention was given to large-unit operations. By stressing technical proficiency, the French advisers taught the Japanese to organize, train, and command military units from company to brigade echelon. By confining instruction to minor tactics and excluding strategy, however, they underestimated their students.

The military academy organized its core curriculum around military science, mathematics, and natural sciences. Subjects included tactics, military organization, weaponry, military geography, and engineering. French was the only foreign language taught until 1883 when the curriculum introduced German, followed by Chinese the next year and English in 1894. After 1877 the government retained a handful of French advisers as language instructors and ordnance technicians.

Military texts imported from the United States were popular among officers because commentators then hailed the American Civil War as the first modern war. Aspiring Japanese officers also read the French strategist Antoine Henri de Jomini, finding that his formulaic principals of warfare followed a rational progression they easily understood. Karl von Clausewitz was translated into Japanese during the late 1850s, but the Prussian strategist’s theory was deemed overly complex.

Under French direction the Toyama Infantry School emerged as the center for musketry, particularly after mid-1876 when Maj. Murata Tsuneyoshi, a Boshin War veteran and a Guard battalion commander, returned from his tour of Prussian and French ordnance and arsenal facilities. Murata quickly put his observations to work by designing extended firing ranges with moving and pop-up surprise targets. He also became the army’s leading designer of small arms. By 1880 Murata had produced his initial rifle at the Tokyo arsenal. After redesigns that included an eight-round magazine that made loading easier and increased firepower, in 1889 the Murata rifle was issued to all active duty soldiers. It remained the standard army rifle until 1910. Reservists continued to use the older Snider rifles.

The modern artillery guns used during the Franco-Prussian War (1870) impressed military observer Maj. Gen. Ōyama Iwao, who returned from Europe convinced that without an independent arms industry Japan would never be a truly sovereign nation. Acting on his advice, in July 1871 the military affairs ministry established the army artillery bureau to manage ordnance matters, including production of small arms, artillery, and ammunition. The same year an arsenal began producing obsolete French bronze mountain artillery guns to equip each garrison. In the mid-1870s the Guard imported the superior German-manufactured steel artillery piece from the Krupp factories, a practice that continued because Japan lacked iron ore and steel plants for casting. In 1879 the Osaka arsenal began manufacturing artillery guns using Italian techniques and fielded these bronze cannon army-wide in 1885. Besides the arsenals, the government retained control of strategic industries such as dockyards, machine shops, and woolen mills (which produced military uniforms). The nascent strategic factories were centers for the absorption and dissemination of modern manufacturing techniques and skills that contributed to capital formation.

The New Conscription Ordinance

The government’s two previous attempts to introduce national conscription had failed. Its third was more ambitious, more radical, and more controversial, particularly about the size of reserve forces. No unanimity existed about a conscription system. Yamada’s opinion, for example, had changed after his return from abroad, and he became an advocate of the militia system similar to those he had observed in Switzerland and the United States. Other officials wanted warriors, not commoners, to fill the ranks. Yamagata favored a Prussian conscription model. Others, including the French-trained Maj. Gen. Miura Gorō, the Tokyo garrison commander, wanted a French-style system.

Yamagata, however, understood the broader implications of national conscription. He distrusted the warrior class, several members of which he regarded as clear dangers to the state, making it foolish to rely solely on samurai to defend the government. Thus, Yamagata used conscription to get rid of the samurai volunteer army, curtail the warrior class, and simultaneously inculcate a mentality of national service for the sake of the emperor and the state. To eliminate preferential treatment for the warrior volunteers, Yamagata revived ancient imperial myths and largely fanciful traditions of military service to the imperial household to promote loyalty to the emperor while curtailing samurai independence. Conscription would likewise break down the old order’s feudal customs, promote the restoration’s goals, and create a future pool of trained soldiers available in times of foreign crisis and able to protect their homes in times of internal disorder. Furthermore, army indoctrination could translate the conscripts’ regional loyalties into national allegiance and send them home as veterans to proselytize army virtues, modernization, and proto-nationalism to their communities. This in brief was the notion of “good soldiers—good citizens.”

The January 10, 1873, conscription ordinance was based on the French model and provided for seven years of military service: three on active duty and the remainder in the reserves. There were liberal exemptions, including the obvious—criminals, hardship cases, and the physically unqualified—and the less obvious—heads of households or heirs, students, government bureaucrats, and teachers. A conscript could also purchase a substitute for 270 yen, an enormous sum that restricted this privilege to the wealthy. Less apparent but more enduring was the emphasis on primogeniture, which though eventually abolished would exert a lasting influence on the enlisted ranks. Under the new ordinance the 1873 conscript army was composed mainly of second and third sons of impoverished farmers who manned the regional garrisons while former samurai controlled the Guard and the Tokyo garrison. The long-term effect was the army’s custom of discouraging first-born male conscripts from making the service a career.

Because of the army’s small size and numerous exemptions, relatively few young men were actually conscripted for a three-year term on active duty. In 1873 the army numbered approximately 17,900 (from a population of 35 million); it doubled to about 33,000 in 1875.62 The 1876 cohort of 20-year-old males numbered about 300,000 candidates, of whom the army evaluated 53,226 as suitable for active duty. Assuming a one-third turnover as three-year enlistments expired, the army needed about 13,000 draftees annually, about one-quarter of those who passed the rigorous physical examination—or approximately 4 percent of the entire cohort. Until the conscription reforms in the 1880s, the percentage of conscripts fluctuated between 3 and 6 percent of the eligible cohort, making military service highly selective (see Table).

One’s chances of being drafted also varied considerably by region. In 1876 the physical examination disqualified more than 85 percent of potential inductees in the Tokyo and Osaka districts. In northeastern Japan, 70 percent of inductees failed, and even among those judged physically eligible for military service, not all were called to active duty.

Regardless of how few men were actually drafted, conscription irked all classes. Former samurai opposed the draft as a class-leveling device that cost them their privileges as the armed guards of the feudal elite. Peasants equated military service with the corvée system imposed by the Tokugawa regime and resisted this latest manpower levy. Others saw it as an unwanted intrusion into their lives, and although the conscription regulations contained numerous exemptions, these applied mainly to wealthier households. Some draft-eligible young men evaded conscription by self-imposed exile, moving from their home district to faraway remote Hokkaidō or distant Okinawa (which was not subject to conscription until 1898). Publishers offered manuals on how to beat the draft. Second or third sons adopted by other families as heirs and heads of households became known as “adopted conscripts.” But these measures were benign compared to the large-scale violent protests against conscription that took place in 1873.

Anti-conscription riots that year joined massive popular disturbances in Fukuoka Province over inflationary rice prices and demonstrations against new excise taxes on seafood and the abolition of the ban on Christianity. In May 1873 tens of thousands of peasants rioted in Okayama Prefecture, burning the homes of the wealthy as well as torching schools and burakumin (the untouchable class) residences, and murdering government officials and school teachers. Rioters destroyed about 400 homes and other buildings and killed or injured more than a dozen people. Among their demands were the elimination of conscription, compulsory education, and forced hair cutting. The Osaka garrison had neither the troops nor the equipment to deal with the mass demonstrations, leaving the government to hastily enlist 300 ex-warriors to suppress the uprising.

Despite severe penalties, including the beheading of more than a dozen demonstrators and the imprisoning of several hundred others, demonstrations quickly spread to neighboring Tottori prefecture, where in mid-June 22,000 people joined five days of protests demanding an end to conscription, compulsory schooling, and the newly adopted western calendar. By the end of the month tens of thousands of peasants rioted in Kagawa prefecture, where the authorities eventually executed several ringleaders and fined 17,000 people for joining in the disorders. Again the army needed outside help, which came this time in the form of fifty former warriors, to put down the demonstrations.

Although the demonstrators opposed the “blood tax,” a term to describe the conscription system that many peasants took literally, having heard wild rumors that the army would drain conscripts’ blood to make wine for the westerners, at root the mass protests demanded a reduction in soaring rice prices caused by rampant inflation as well as abolition of primary schools and the western calendar. Responding to the widespread unrest, Yamada Akiyoshi pronounced national conscription premature and proposed to postpone its introduction for about a decade. During that interval the government would educate the people in the values of the new nation, which in turn would make conscription acceptable to them. Meanwhile, a citizen militia led by professional officers could protect the nation. Soon afterward Yamada left the army to pursue a career in the law; he later would become justice minister. His opinions about the value of a militia system and a small standing army, however, continued to resonate with influential officers for the next two decades.

Army and Emperor

The new Meiji leaders relied heavily on the imperial institution to connect past to present when formulating national values. The army in particular stressed its links to the emperor to inculcate loyalty in the ranks. An imperial memorial issued on February 3, 1870, proclaimed the emperor a living embodiment of godhood and his throne a holy office established by the Sun Goddess and handed down in unbroken succession to the present. In 1872 the military department promulgated an imperial rescript (tokuhō) of eight articles that set standards of conduct for the new army by enunciating a soldier’s duties based on bedrock principles of loyalty to the throne, obedience to orders, courtesy and respect for superiors, and the prohibition of various types of disruptive conduct. Thereafter, when recruits entered the barracks for the first time they were welcomed with a ceremony that included their officers reading the rescript aloud to them, a practice that continued until 1934.

Besides issuing memorials, the emperor presided over military ceremonies, beginning in February 1870 when he reviewed about 4,000 troops on the imperial grounds in conjunction with the establishment of the ordnance bureau. In October 1871, Emperor Meiji observed his newly formed imperial bodyguard conduct field training exercises in a driving rainstorm. The following year he presented the new Guard infantry regiments their unit colors, inaugurating a tradition that continued until the end of World War II. In April 1873 he attended military exercises at Tokyo and led Imperial Guard cavalry troops through field maneuvers.

Emperor Meiji also presided over the military academy’s first commencement, in July 1878, publicly affirming the bond between the army’s officer corps and the imperial institution. To further strengthen identification with the throne, imperial family members routinely served in the military, although the court had not produced active warriors since the wars of the fourteenth century. They were exempted from conscription, preinduction physicals, and written qualification tests and were directly commissioned by an imperial order. In 1885, the year after the creation of aristocratic ranks and titles, young men from twenty-three newly ennobled families were transferred from the Peers School to the military academy, where they formed about 12 percent of the 108-man class.69 The emperor also promoted social changes. In 1873, January 4 became Tenchōsetsu (the emperor’s birthday), a national holiday. In March of that year he cut his hair in the western style, and in June he appeared in a western-style army uniform during a military exercise.

Tactics for the New Army

Until 1870, “modern” infantry tactics were found in an 1829 translation of a Dutch tactics manual. Concurrent with the new conscription system, the army adopted French tactical military doctrine, its 1873 infantry manual being a translation of the 1869 French edition. Training was conducted in squad-size groupings commanded by a corporal. Recruits received six months of basic instruction that included physical conditioning, platoon and company formations, dispersed movement, marksmanship, bayonet drill, basic sanitation, and military customs and ceremonies. Field training consisted of mandatory drills, execution of the manual of arms, rote memorization of set-piece tactical problems, and small-unit command and control. Instructors relied on strictly choreographed drill regimens, endless repetition, and iron discipline to get results. The army penal code issued in April 1873 prescribed severe punishment for disobeying superiors or conspiring against the government. In 1874 seventy soldiers were tried under the penal code; 530 were tried the following year.

In October 1874 the army created the corporal’s group, a squad-size unit of conscripts organized for drill and training purposes and overseen by a noncommissioned officer or junior officer. The system would eventually evolve into the squad section that regulated training, discipline, and conscripts’ lives in the barracks. Squad leaders initially handled only training, but revised regulations in 1880 assigned NCOs greater administrative responsibilities for the unit’s daily activities. Administration and management, however, were weak. Training regimens varied according to garrisons because of an absence of standardized training programs, frequent changes to drill manuals, irregular tables of organization, and the personalities and interests of commanders and NCOs. The army’s first standardized table of organization appeared in 1877 and established the strength of an infantry company at 160 enlisted and eight NCOs, each responsible for a twenty-man section.

Monday through Friday the conscripts awakened at 5 a. m. for morning drill followed by afternoon training sessions. After Saturday morning drill, they spent the day preparing their barracks for a 4 p. m. inspection. A Saturday-night bath followed inspection, and Sunday was a free day. Although the army conscripted from all classes of society, its officers and NCOs were drawn mainly from former samurai, and the traditional hierarchical relationship of samurai and peasant endured. Former warriors who became NCOs routinely beat or physically punished the farmer conscripts to instill discipline and ensure compliance with orders. Under such conditions, desertion flourished; of fifty conscripts sent by one domain, half deserted, one of them on four different occasions. It is true that the Imperial Guard received better training and equipment, but it too often relied on uncritical and uncomprehending imitation of western-style tactical drill because the army lacked a professional officer corps sophisticated enough to devise more appropriate methods.

Five years after the restoration, the new government had created a national army from scratch, fought a civil war, and imposed domestic order. It had established conscription, military schools and training facilities, and begun to standardize equipment and training. The army was small, depending on conscription as much for political reasons (to eliminate warrior influence) as social (as a leveling device) and economic (it was affordable). Significant differences about the nature of the armed institution remained unresolved; its deployment capability was haphazard at best, its personnel unproven, and its leadership divided. Standardized training and equipment were still lacking, and the effectiveness of military schools was still undetermined. The army had no popular base of support, conscription was greatly resented, and samurai bands were leading armed rebellions. On the positive side, the Meiji leaders had erected the framework for a national army that relied on the imperial symbolism of the emperor as the military leader, although they protected the army against direct imperial intervention in military affairs by establishing a system in which the emperor could rarely decide anything by himself. From these shaky foundations, the new army ventured into an uncertain and dangerous future.