Explaining partial diffusion during military assistance

In essence, military diffusion is about the transfer of military technology, tactics, and organisational forms from one place to another. Diffusion therefore implies a demonstrable path of causality between originator and subsequent implementation in a new context. Although the diffusion of military technology has attracted particular attention, material artefacts rarely exist in a social vacuum, and so the successful diffusion of military technology is typically accompanied by the ideas and knowledge that animates its effective application.Footnote ⁴ Importantly, disparities in either technology or praxis can confer dramatic, if fleeting, advantages in war.Footnote ⁵ Consequently, neorealism expects states to rapidly import new military practices in order to survive. According to Kenneth Waltz, ‘states imitate the military innovations contrived by the country of greatest capability and ingenuity … so the weapons of the major contenders, and even their strategies, begin to look much the same’.Footnote ⁶ Yet, in reality, technology and praxis often diffuse slowly and separately, producing novel ‘hybrids and creole artefacts’ in what Michael Horowitz describes as a process of ‘mutation’.Footnote ⁷ Moreover, while these partial or hybrid military forms can sometimes be more potent in their local context than the original model, they often prove to be significantly less capable.Footnote ⁸

Traditionally, this apparent contradiction has been explained through system-level factors that moderate the imperative for change. Alliances and external balancing can provide alternatives to profound military change, while offensive and defensive technologies and doctrines may appeal to states differently, depending on the peculiarities of local geography.Footnote ⁹ Equally, where diffusion is undertaken for reasons other than functional performance, the propensity for selective, partial, or dysfunctional importation may increase. Foreign military models are sometimes imposed through third-party pressure, as with the Warsaw Pact armies that practiced ‘Potemkin drills in Potemkin skills’ for Moscow's benefit.Footnote ¹⁰ Similarly, a lack of clarity over military best practice can lead states to emulate leading countries as a form of hedging, while normative perceptions can encourage diffusion for the legitimacy (rather than the security) such practices confer.Footnote ¹¹ However, systemic imperatives are typically instrumentalised through national and subnational processes, focusing attention on the local ‘blockages to technological assimilation’ that moderate diffusion, or ‘the political economy of technology’.Footnote ¹² Emily Goldman, for instance, found that national military cultures can be more or less tolerant of external ideas, affecting their receptiveness to foreign military innovations, while the structure of domestic civil-military relations can likewise impede the importation of foreign military innovations.Footnote ¹³ Consequently, Horowitz has proposed an ‘adoption-capacity’ model, in which the expected benefits of military diffusion are weighed against the expected costs of implementation, providing an explanation for both state-level and systemic responses to military change embedded in the practical obstacles individual states face in importing particular military models.Footnote ¹⁴ In so doing, these scholars draw attention to the importance of implementation mechanisms, as opposed to motives, in understanding patterns of military diffusion.

Practically, though, diffusion occurs through a variety of mechanisms, defined according to the level of external involvement the importation of military change. For Horowitz, this range is largely subsumed within the broader calculus of the financial and organisational capital required to implement military change, yet the nature and degree of foreign involvement can itself affect the desirability, feasibility, and quality of diffusion processes. At one end of this spectrum, imitation or emulation need not involve any direct foreign assistance, and battlefield learning can sometimes even unwittingly enable diffusion between enemies.Footnote ¹⁵ However, diffusion often benefits from external assistance to some degree. This may be informal, as with the professional discourse between international defence attachés, exchange officers, and technical experts.Footnote ¹⁶ At the other end of the spectrum, foreign support may be formal and overt. Weapons sales typically include contracts for training and technology transfer, while states may also receive military-to-military aid, known as military assistance. This typically takes the form of seconded training missions or contracted advisors, though foreign military officers sometimes assume command of local forces in order to better direct change.Footnote ¹⁷ In principle, military assistance can be expected to produce the greatest fidelity in military diffusion, inasmuch as it provides direct access to the expertise (and often also equipment) that less formal mechanisms struggle to acquire. Certainly, attempts at emulation are often implicitly limited by the established organisational priorities and cultural perceptions that shape institutional understandings of foreign military change, as Thomas Mahnken has shown.Footnote ¹⁸ Yet, in practice, military assistance often struggles to overcome conflicting political, cultural or bureaucratic agendas, even in the face of clear imperatives – and despite active recipient engagement.Footnote ¹⁹

Recently, principal-agent theory has been advanced as a framework to understand these dynamics. According to principal-agent theory, partial change can be understood as the product of a divergence of interests between recipient ‘agent’ armies, who seek military assistance to achieve particular goals, and foreign ‘principals’, who provide it to further their own national aims.Footnote ²⁰ This interest asymmetry encourages the agent to ‘shirk’ particular changes desired by the principal, but which diverge from their own preferences – something exacerbated by the normative prisms through which principals perceive (or misperceive) their own and their agent's interests.Footnote ²¹ In theory, principals can use conditionality in the provision of military assistance to mitigate shirking, using punishment and reward in carrot and stick fashion to structure the interests of their agent, thereby improving the fidelity of military diffusion. During the Korean War, for example, US conditionality successfully overcame South Korean objections to military professionalisation, while the application of conditionality to US military assistance in Iraq produced an immediate, if modest, reduction in shirking.Footnote ²² Yet, even proponents of greater conditionality in military assistance have been highly circumspect about its likely impact. Stephen Biddle, for example, has argued that military assistance will generally improve recipient capabilities somewhat, if only because more training is better than less. Nevertheless, interest asymmetry is still held to create ‘a ceiling on real effectiveness’, in part because it concomitantly reduces principal's access to reliable information on agent behaviour necessary for enforcing conditionality.Footnote ²³

Typically, though, principal-agent theory has been used to describe the bilateral relationships between one military patron and their local partner, often from the perspective of the donor nation. Yet, in practice, principal-agent interactions in rarely exist in isolation. Just as principals retain the option to seek alternative agents, so agents may court multiple patrons – with potentially profound implications for the utility of conditionality during military assistance. With other types of international aid, recipient states can use these pluralistic relationships to bypass principals’ conditionality, despite an enduring dependence on foreign help.Footnote ²⁴ Indeed, US conditionality in military assistance to Iraq appears to have been implicitly limited by the Maliki regime's simultaneous ability to seek rival sources of military patronage from neighbouring Iran, rendering conditionality a ‘least-bad’ policy option rather than a proverbial silver bullet.Footnote ²⁵ Yet, little concerted attention has so far been paid to these dynamics in the literature on military assistance. How, then, do these parallel relationships affect the principal-agent politics of military assistance? And what impact do they have on the quality of military diffusion?

Conceptualising principals and agents in military assistance

In a principal-agent relationship, the principal contracts an agent to conduct activity on their behalf, in furtherance of a desired goal. Principals typically employ agents because they lack the appetite or attributes necessary to directly achieve the aim themselves. However, this compact need not be formal, and as Eli Berman et al. note, agents frequently do not recognise themselves as such. Instead, principal-agent relationships are often tacit, defined by the overall pattern of power relations between each state, and the provision of rewards in exchange for desired activity.Footnote ²⁶ Consequently, the function of military assistance in principal-agent interactions will vary, depending on each party's goals and interests. When military diffusion is actively desired by the client, assistance may form the currency of principal-agent interactions, representing part of the reward offered by the principal. When the principal desires the suppression of some (likely shared) threat, recipient military change may simultaneously form part of the activity expected of the agent, insofar as it facilitates the achievement of the principal's goal. In some cases, military diffusion may actually represent the substance of a principal's goal, perhaps in order to deter or constrain a third-party.

Typically, though, the aims of the agent will diverge somewhat from those of the principal, while the act of delegation invariably means that the principal will lack complete oversight of the agent's behaviour. Hence, neither party can be entirely certain of the other's commitment or intentions. This information asymmetry creates powerful incentives for the agent to ‘foot-drag’, delaying or shirking activity to secure (more) reward from the principal, even when their goals closely align. Moreover, the more the interest asymmetry, the more the agent will benefit from diverting or misappropriating the principal's resources in furtherance of their own divergent aims, known as agency loss.Footnote ²⁷ The Afghan National Army, for example, readily accepted US training and equipment, but actively avoided politically unpalatable defence reforms, preferring instead to rely on deployed US troops to bail them out.Footnote ²⁸ In principle, the best way to reduce agency loss is to select agents whose goals closely match the principal's own, but this is not always possible in practice. Principals can mitigate agency loss through close supervision of the agent, but this reduces the cost effectiveness of delegation, and risks sucking the principal into the activity itself. Instead, conditionality provides a means to manage interest asymmetry, by using contingent rewards and penalties ‘to manipulate allies’ incentive structures in ways that encourage them to work and not shirk’ – often by coercively reducing or withholding support.Footnote ²⁹

To be effective, conditionality relies on a significant power differential between principal and agent. In Berman et al.'s framework, principals must by definition possess the ability to apply reward and punishment to their agent, replace them altogether, or forsake the goal entirely. Hence, in this understanding, the US could not use principal-agent incentives against a state of equal standing, like the Soviet Union, or a state possessing reciprocal leverage, as with Pakistan's recent custodianship of supply lines into Afghanistan.Footnote ³⁰ Indeed, the recipient state's relative dependence on their patron is seen as a defining feature of principal-agent relationships, and is presumed to limit the recipient's strategic options while facilitating those of the patron. As Susan Shapiro explains, whereas ‘principals are risk neutral (they have diversified …), agents are risk averse, because they have placed all their eggs in this one basket’.Footnote ³¹ This is what Berman and Lake call ‘subordination: the requirement that the principal be able to influence the proxy more than the proxy can influence the principal’, such that ‘the only way the proxy can affect the principal is by choosing effort or shirking’.Footnote ³²

Yet, in reality, an agent is rarely confined only to working or shirking. Historically, some states have attempted to use mercenaries alongside formal military assistance as means to hedge against the influence of would-be principals, while those that come to depend on a single foreign patron for security risk outsourcing survival to the whims of another.Footnote ³³ Importantly, recourse to multiple sources of military assistance might allow an agent to bypass conditionality, by playing rival principals off against each other. In economics, where agents frequently engage in multiple simultaneous principal-agent relationships, this is known as a ‘common agency’ problem.Footnote ³⁴ Logically, though, this possibility depends on a series of conditions being met. Firstly, the agent must have access to more than one source of military assistance. States with limited options must take what help they can, and accept the strings that come with it. Secondly, the assistance in question must be essentially fungible between suppliers. Significant disparities in the quality, quantity, or sincerity of assistance available are unlikely to provide the agent with credible leverage against the dominant supplier – especially if the inferior offer does not meet that agent's needs. Thirdly, and critically, these principals must prefer to compete against each other rather than cooperate for shared control. In other contexts, this is known as a principal-principal conflict.Footnote ³⁵ Here, the agent's ability to play rival principals off against each other will be assisted by variation in the conditionality each principal applies to its assistance, or else by uncertainty over the scope and conditions of a rival offer.

When principals share common goals, coordination provides clear benefits, effectively transforming multiple principals into a single actor.Footnote ³⁶ This should increase the utility of conditionality, though at the risk of potential free-riding among principals. Equally, cooperation may be beneficial even when principals’ goals somewhat diverge. This is especially likely when a lesser principal lacks the resources or will to compete with more hegemonic patrons. Here, bandwagoning might provide the best route to a minimal level of goal accomplishment. Conversely, when states have similar abilities to coerce and reward an agent, but possess mutually exclusive or zero-sum goals, competition may appear the best option.Footnote ³⁷ In Iraq, for example, Biddle has argued that US conditionality would have been more effective had it been able to coordinate with Iran. Failing that, the US might have negated Iranian influence by ensuring its carrots and sticks outstripped Iran's. Certainly, limited assistance combined with unwanted conditionality is a recipe for ‘encouraging would-be partners to look elsewhere for assistance’.Footnote ³⁸ Clearly, though, competition between principals does not automatically mean that an agent will be able to exploit this rivalry, and principals may sometimes prefer to call the agent's bluff than concede conditions. This seems especially likely when reducing strings would increase one-sided risks for the principal to such a degree that the goal is no longer worth pursuing.

Where these three circumstances prevail, however, a recipient state might play multiple principals off against each other to avoid conditionality, but at the price of less coherent diffusion. Not only would this allow recipient states to cherry-pick aspects of military diffusion without having to import politically unpalatable (but objectively necessary) components, it might also result in a composite hybrid of various different practices and techniques, as multiple sources of foreign expertise are used to bypass the political obligations attached to each. Such a circumstance would account for the practical limitations on conditionality in military assistance, while simultaneously explaining how states seek to reconcile imperatives for diffusion with national and subnational preferences and priorities. This article now turns to explore the impact of principal-principal conflict in one prominent case of hybrid military diffusion: the military modernisation of nineteenth-century Japan.

Nineteenth-century Japan as a case study in hybrid military diffusion

The modernisation of nineteenth-century Japan represents one of the most remarkable instances of military diffusion in modern history. When Commodore Perry's ‘black ships’ arrived in 1853, Japan had been a closed country for over two centuries. Under the Tokugawa Shogunate, Japan was divided among rival clan-based domains, or han, ruled by the samurai warrior class. The Shogun's Bakufu government presided on behalf of the titular emperor, maintaining a fragile balance of power through feudal privileges, draconian punishment, and international seclusion. Perry's arrival exposed Japan's relative military inferiority, compelling the Bakufu to accept series of unequal Ansei Treaties that created extra-territorial treaty ports, regulated tariffs, and conferred Western citizens with extra-judicial status.Footnote ³⁹ It also precipitated a period of rapid military diffusion, in which military assistance played a central role. Until its overthrow in 1868, the Bakufu received a succession of Dutch, French, and British missions to modernise its army and navy. The Emperor Meiji's successor government likewise employed British and French help to develop its navy, while French and Prussian advisors trained its army. Both regimes also purchased weapons and warships from various Western nations, alongside diverse forms of less formal assistance.Footnote ⁴⁰ By the 1890s, Japan had begun to export indigenously produced weapons to Western states, and went on to defeat Imperial China in the Sino-Japanese War of 1895. Japan subsequently suppressed the Boxer Rebellion alongside the Western powers; a volte face confirmed by Japanese victory over the Russian Empire in 1905.Footnote ⁴¹ Japan finally renegotiated the unequal treaties in 1894, leading to the end of the treaty port system in 1899 and the return of Japanese tariff autonomy in 1911.Footnote ⁴²

Consequently, Meiji Japan is sometimes presented as the archetypal ‘rational shopper’, selectively cherry-picking the most appropriate military technology and organisation in a ‘broad, deep, and faithful diffusion’ of Western military practice.Footnote ⁴³ The Meiji military certainly sought to expunge Japan's indigenous martial heritage, adopting European sabres over samurai swords, and importing so much leather for Western-style uniforms that one wag suggested a switch to elephant hide.Footnote ⁴⁴ In Ernst Presseisen's words, Meiji Japan ‘adopted Europe's military methods and technology without question … The reason lay in a technical superiority for which the Japanese found no equivalent in Oriental knowledge.’Footnote ⁴⁵ Yet, this veneer of coherence belied significant martial hybridity. As David Wittner notes, the ‘Guards at the Imperial Palace resembled the Zouaves at the Tuileries and marched to French light infantry bugle calls, while marines went on parade to the sound of fife and drum in the tradition of their British counterparts’.Footnote ⁴⁶ Such outward contrasts reflected a deeper amalgam of various Western practices, created as ‘Meiji officials adopted technologies whose functionality did not match stated expectations’.Footnote ⁴⁷ Indeed, the fledgling army and navy were even unable to agree on a standard gauge for screws; the army advocating (continental) metric sizes while the navy favoured the (British) imperial system.Footnote ⁴⁸ While a degree of local adaptation was accepted, embodied in the slogan wakon yōsai or ‘Japanese spirit, Western technology’, Meiji officials nonetheless downplayed this hybridity as a transient product of Japanese backwardness.Footnote ⁴⁹ As a French legal advisor wrote, ‘there is not a word about the special condition of Japan … it is a question, in their opinion, not of knowing themselves, but of transforming themselves’.Footnote ⁵⁰ One French diplomat likewise observed that, by 1875, Japan displayed ‘a hybrid state of things as displeasing to the Japanese as to foreigner himself’.Footnote ⁵¹

Importantly, Japanese military diffusion took place against a backdrop of significant domestic insecurity and foreign coercion. The Bakufu faced a series of internal threats, eventually leading to its defeat in the Boshin War, while the victorious Imperial government likewise faced a succession of internal revolts and a further civil war in 1877. These pressures were simultaneously exacerbated by the Western treaty powers, who jockeyed to secure ever greater trading privileges in Japan. Indeed, the Ansei Treaties typify what John Gallagher and Ronald Robinson have described as the ‘imperialism of free trade’ through which ‘informal empire’ was conducted.Footnote ⁵² Externally controlled tariffs, for example, imposed fixed exchange rates that resulted in capital flight, currency debasement, and 200 per cent price rises in Edo. Most-favoured nation clauses likewise extended the benefits of one foreign power to all, while denying reciprocal rights to Japan.Footnote ⁵³ Diplomatic historians have emphasised the dualistic character of nineteenth-century international relations, in which European powers justified their imperial activities as part of a ‘civilizing’ process, creating a two-tier international system in which all states were notionally equal, yet ‘lesser’ (that is, militarily inferior) non-Western nations were simultaneously denied full sovereignty on the basis of their lack of Westernisation.Footnote ⁵⁴ Moreover, the treaty powers repeatedly used or threatened force to assert their interests in Japan, making negotiation Japan's ‘best weapon to defend themselves from the West’ as an ersatz ‘form of resistance’.Footnote ⁵⁵ Japan's lack of indigenous military capability thus rendered it the archetypical agent state.

As a result, the modernisation of nineteenth-century Japan provides an ideal case study to examine the impact of principal-principal rivalry on the fidelity of military diffusion. Successive Japanese regimes were faced with a series of potent domestic threats – ultimately leading to the downfall of the Shogun – while simultaneously attempting to reclaim national sovereignty from the very powers on which they were dependent for military aid. Moreover, Japan not only employed multiple sources of foreign military assistance, but the ensuing diffusion displayed exactly the lack of fidelity that principal-principal conflict might be expected to produce. In the existing literature, however, functional inconsistencies in Japan's martial modernisation have typically been explained through reference to normative, cultural processes. Goldman, for example, has described Japanese military modernisation as a ‘quest for legitimacy and acceptance into the community of modern nations’, articulating Japan's cultural receptivity to modernisation in stark contrast to the contemporary Ottoman Empire's comparative cultural resistance and martial ossification.Footnote ⁵⁶ Eleanor Westney, meanwhile, observed a path-dependent ‘contagion effect’ in Meiji diffusion, whereby a lack of reliable information led Japan to privilege existing partners over unknown but objectively better ones.Footnote ⁵⁷ Equally, Tessa Morris-Suzuki has attributed uneven modernisation to the labyrinthine social networks through which Western ideas, technology, and knowledge flowed into Japan.Footnote ⁵⁸ Collectively, these approaches emphasise the importance of culturally contingent socialisation processes in Meiji military diffusion.

However, while these perspectives collectively provide a compelling explanation for the fusion of local and imported traditions, they struggle to account for the simultaneous amalgam of different Western models in Japanese diffusion. Indeed, arguably, path-dependent social networks and contagion effects alone should have led to a smaller number of hegemonic patrons, rather than the eclectic mix of military assistance Japan actually received. Moreover, as Horowitz has argued, manifestations of domestic culture are often indistinguishable from (or intimately embedded in) domestic political discourse. Thus, although the explanatory power of cultural models should not be dismissed, the reciprocal interaction between local politics and international agendas may shape the options for, and patterns of, martial diffusion as much as cultural receptivity to change, per se.Footnote ⁵⁹ Accordingly, this article now turns to examine an alternate principal-agent politics of military assistance in nineteenth-century Japan, first under the Shogun and then during the Meiji Restoration, in order to assess the role of conditionality and principal-principal rivalry in the hybrid patterns of Japanese military diffusion.

Military assistance to the Shogun, 1854–68

In 1853, Commodore Perry arrived off Edo with a show of force and a demand that Japan open for business. Unable to respond militarily, the Bakufu had little choice but to sign a series of one-sided treaties granting trading privileges to various Western nations. While the reopening of Japan was thus a triumph of gunboat diplomacy, Western interests in the country varied significantly. For the United States, Japan primarily represented an economic opportunity, as a base for trade, whaling, and coaling. It had neither the inclination nor the means to colonise Japan, especially after the outbreak of the US Civil War. Britain, in contrast, viewed Japan through the lens of its wider imperial network, and primarily sought to prevent other Western powers from using Japan as a platform to threaten British commercial footholds in Hong Kong and Siam.Footnote ⁶⁰ However, Britain's mercantile interest in Japan rose as its trade expanded; not least because, in the words of one historian, British faith in free trade convinced them that if only ‘unrestricted commerce were allowed between all Japanese and the outside world, political discontents would evaporate’.Footnote ⁶¹ France pursued a similar mixture of trade, power projection, and prestige. The Quay d'Orsay was particularly interested in the export of Japanese silkworm eggs to bolster the French silk industry, then decimated by Pébrine disease. Japan also offered a potential strategic harbour for the French navy, while Léon Roches, the French Minister in Japan in the 1860s, likewise hoped that ‘Japan would be for us what China is for England, a French market’.Footnote ⁶² The Western powers were thus united in their aim of opening Japan to further trade, but rivals in their efforts to secure Japan for their own benefit.

For Japan, meanwhile, the treaties underscored the need to acquire modern military capabilities. As one domain official wrote in 1853, ‘with insufficient equipment of fire-arms we will not be able to fight to the best of our abilities in naval engagements, coastal defence, field operations, or any other kind of warfare’, and so ‘will be quite unable to resist their machines’.Footnote ⁶³ Policies like kinken shōbu (‘be frugal and love arms’) and fukoku kyōhei (‘strong army, rich nation’) subsequently aimed to stave off Western encroachment by importing Western arms and practices.Footnote ⁶⁴ However, the treaties also sparked a radical populist reaction, as anti-Bakufu as it was xenophobic. As one British diplomat observed, the Shogun had ‘fallen into national discredit by assenting to the much-dreaded renewal of intercourse with Western people, though indeed this assent was obtained under irresistible pressure’.Footnote ⁶⁵ Moreover, this sonnō jōi movement (‘revere the emperor, expel the barbarian’) produced a series of attacks against foreigners, including consular staff, for which the Bakufu was held responsible. Ernest Satow, a British diplomat and translator, described it as ‘a busy time for Colt and Adams’ as ‘everyone wore a pistol … and constantly slept with one under his pillow’.Footnote ⁶⁶ Hence, Western demands for greater access placed the Bakufu in a double-bind. If the either Shogun or individual domains resisted, Western powers might use this as a pretext for war. Yet, if domains hostile to the Bakufu (like Satsuma and Choshu) were allowed to trade internationally, this might undermine the domestic balance of power on which the Shogunate depended.Footnote ⁶⁷ Thus, while the British viewed domestic stability and unrestricted trade as mutually reinforcing, the Bakufu viewed them as actively opposed, and was primarily interested in limiting territorial access.Footnote ⁶⁸

Caught between its own military inferiority and the divergent interests of the Western powers, the Bakufu initially sought to manipulate the Ansei Treaties to its own advantage. Extra-territoriality, for example, was used to confine foreign merchants to small enclaves at the treaty ports. Most favoured nation clauses were likewise used to balance one foreign power against another, in order to prevent foreign rivalries from spiralling into conquest. As one domain lord, or daimyō, told the Bakufu, ‘The thing most to be feared is not the influx of other countries, but the rivalry between England and Russia.’Footnote ⁶⁹ In neighbouring China, such clauses had contributed to a policy of ‘using barbarians to control barbarians’ by preventing any single foreign power gaining dominance.Footnote ⁷⁰ Although these clauses automatically extended concessions granted to one power to the others, by extension, any limitation had to be unanimously agreed, enabling the Bakufu to leverage each power's individual interests against the others. In 1861, for example, the Bakufu used a series of sonnō jōi attacks to delay to further port openings, first persuading the Americans (after their legation secretary was murdered) and then the British (after their compound was attacked), who then convinced the remaining treaty powers.Footnote ⁷¹ In a similar fashion, the Bakufu enlisted British help to dislodge Russian warships from the island of Tsushima. By playing on British fears of a Russian threat to China and Siam, the Bakufu got the Royal Navy to do what its own forces could not, in a gambit Auslin has described as ‘clearly in line with the bakufu's policy of playing foreign powers off one against another whenever feasible’.Footnote ⁷²

Simultaneously, the Bakufu attempted to opportunistically exploit multiple sources of foreign military expertise, in order to imitate Western military methods with as little direct Western involvement as possible. During Japan's international seclusion, limited trade was permitted via a single Dutch concession in the bay of Edo. As Japan's only existing Western partner, the Shogun ordered a warship from the Dutch soon after Perry's arrival, and subsequently requested Dutch help to train Japanese naval officers. In 1855, the Dutch naval mission established Japan's first naval training school, largely in order to maintain their trading relationships. However, the new military school that opened that year in Edo saw Japanese instructors teach Western gunnery from translated Dutch manuals.Footnote ⁷³ Simultaneously, the Bakufu assisted Russian sailors to construct a replacement for Admiral Putyatin's flagship, sunk in a tsunami as he negotiated Russia's trade treaty with Japan, and used the knowledge gained to construct ten indigenously built Japanese warships.Footnote ⁷⁴ By 1862, the Shogunate had embarked on its own in-house reorganisation of its samurai troops along Dutch lines, intended to create a force of 13,000 infantry, cavalry, and artillery.Footnote ⁷⁵

However, Shogunal attempts at military emulation were undermined by parallel diffusion among hostile domains, resulting in a growing threat to the Bakufu's domestic power – and its efforts at international balancing. Chōshū, for example, had begun to raise mixed commoner-samurai Kiheitai, or ‘special units’, equipped with imported muskets, while Saga had reverse-engineered a reverberatory furnace to manufacture small arms and artillery.Footnote ⁷⁶ Then, in 1862, Satsuma samurai murdered the British merchant Charles Richardson, prompting British demands for reparations from the Shogun and Satsuma. Despite successful attempts at delay and obfuscation, the Bakufu's inability to make Satsuma pay ultimately resulted in the Royal Navy's bombardment of Satsuma's capital, Kagoshima.Footnote ⁷⁷ Moreover, the Emperor Kōmei issued a proclamation to expel all foreigners in 1863, which the Shogun declined to enforce. Indeed, the British made it clear that any attempt to shut down trade would be taken as ‘a declaration of war by Japan itself against the whole of the Treaty Powers’, and Britain and France landed troops to defend Yokohama.Footnote ⁷⁸ Nonetheless, Chōshū unilaterally began to bombard international shipping in the Shimonoseki Straits, resulting in a joint British, French, Dutch, and American naval campaign to destroy Chōshū's shore batteries during the summer of 1864.Footnote ⁷⁹ Chōshū radicals also launched a failed coup against the Shogun at the imperial palace, but were deposed in a domain power-struggle before any serious fighting with approaching Bakufu troops.Footnote ⁸⁰ The incidents highlighted the Bakufu's enduring material weaknesses vis-à-vis both the West and its own hostile domains, forcing the Shogun to seek greater formal Western military assistance in what D. Colin Jaundrill has described as a ‘major shift in the shogunate's approach to military science’.Footnote ⁸¹

From 1864, Shogunal troops adopted British infantry drills in place of Dutch ones, and requested formal British military training. In so doing, however, the Bakufu exposed itself yet further to Britain's political demands. At first, the outgoing British Minister refused to sanction formal military assistance, instead granting limited access to the British garrison at Yokohama.Footnote ⁸² His successor, Sir Harry Parkes, demanded a further indemnity to offset the costs of the Shimonoseki campaign. Knowing the Shogun couldn't pay, he then sought to make made deferral of the indemnity conditional on further treaty port openings, telling Whitehall in 1866 that the Shogun's ‘application for a delay in the payment of an Indemnity is a lever in my hands’.Footnote ⁸³ With the Bakufu dragging its feet on both port openings and payment, British military assistance remained concomitantly tokenistic. One Japanese official described joint training with the Yokohama garrison as little more than being allowed to ‘peep at the English soldiers’ drill through the fence’.Footnote ⁸⁴ Satow likewise recalled ‘a review and sham fight’ in 1866, jointly held as Shogunal troops prepared for a further expedition against Chōshū, in which the Japanese troops ‘received all the greater praise because they had received no practical instruction. Their officers had got it up from books, the difficult passages being explained to them by ours.’Footnote ⁸⁵ Moreover, the Bakufu's rivals in Satsuma and Chōshū continued to receive significant quantities of arms and ammunition via the British merchant Thomas Glover. During two months in 1865 alone, Chōshū acquired 1,800 rifles, 3,000 muskets, and 6,300 pistols, and its Kiheitai consequently fought the Shogunate to a standstill in the ensuing campaign. In like fashion, Satsuma too adopted British drill over Dutch, in no small part because British manuals were made increasingly accessible.Footnote ⁸⁶

However, Britain was not the only source of military assistance available to the Shogun. Unwilling to concede to the political conditions required for concerted British aid, the Bakufu instead turned to Britain's main rivals, the French. To a certain extent, France represented something of a ‘Goldilocks’ patron; France had sufficient political and military power in Asia to be useful, but appeared less territorially threatening or overtly mercantilist than either the Russians or the British.Footnote ⁸⁷ Importantly, as one contemporary observed, Roches and Parkes ‘hated one another and were as jealous as a couple of women’, mirroring wider tensions in Anglo-French policy.Footnote ⁸⁸ Another member of the British legation remarked that ‘it was amusing to see the French and English Ministers trying to outmanoeuvre each other. Wherever one went, the other also immediately went. When one had an interview with a Japanese official, so had the other’.Footnote ⁸⁹ In 1865, the Shogun requested French assistance to establish shipyards and arsenals, and sought a military training mission the following year. Work began on a dockyard and arsenal at Yokosuka, under the direction of French naval engineer François Verny, while Roches established the Société Française d'Exportation et d'Importation as a vehicle for French arms exports.Footnote ⁹⁰ The military mission arrived in 1867, under the command of Captain Chanoine. It proceeded to train a cadre of Bakufu officers and NCOs, before instructing entire units of Shogunal infantry, artillery, and cavalry.Footnote ⁹¹ British regulations were replaced with French, precipitating a wider reorganisation of Bakufu forces on French lines.Footnote ⁹² By January 1868, the French had trained a force of 1,500 infantry, two cavalry squadrons, five batteries of mountain artillery and a company of engineers, equipped with 1,866 of the latest breach-loading Chassepot rifles.Footnote ⁹³

As with the British, however, French support was neither altruistic nor without caveat. France had initially sought to maintain concerted pressure on the Shogun alongside the other treaty powers, and Roches's predecessor had advised Paris against military assistance in 1862, arguing that ‘the Japanese seek arms and when they see themselves sufficiently provided, they will doubtless seek enemies’.Footnote ⁹⁴ France likewise imposed its own indemnities on the Bakufu for the murder of a member of its garrison at Yokohama, while Roches had himself supported the Shimonoseki campaign, telling the Shogun that if the Emperor ‘continues to insist upon the cancellation of the treaties, he is asking for war’.Footnote ⁹⁵ By extension, France viewed military assistance as a prime opportunity to further its interests in Japan. French-built dockyards were seen as a potential regional base for the French navy, while Roches convinced Paris that assistance would provide an outlet for French arms that would otherwise go to a rival power.Footnote ⁹⁶ Indeed, Shogunal entreaties for ‘arms, munitions, ships and the men which are indispensable in order to carry out the submission of the rebels’ were seen by Roches as an opportunity ‘to persuade the Taikun that we shall be his friends and natural allies’ and thereby ensure ‘that his own interest is linked to ours’.Footnote ⁹⁷ Even so, Paris still refused to entertain any waiver of the Shimonoseki indemnity, on the basis that military assistance alone was sufficient for suppression of domain threats to the treaties, and even scaled down the proposed size of the French military mission so as ‘not to give at first too great an importance to the mission which will be formed’.Footnote ⁹⁸

Importantly, the provision of French military assistance aggravated Anglo-French rivalries, allowing the Bakufu to manipulate this pre-existing principal-principal conflict to its own ends. At face value, conditional French assistance appeared to facilitate French influence. The Bakufu told Roches that ‘France will be for us the Older Brother who guides and assists his Younger Brother’, and dispatched the Shogun's own younger brother to France for education in 1867.Footnote ⁹⁹ Roches became directly involved in discussions on governmental reform, while the Quay d'Orsay declared that he had obtained ‘so legitimate an influence’ that ‘when the Japanese ministers have wished to have a plan adopted by their sovereign they have asked him to support it’.Footnote ¹⁰⁰ Yet, to secure his position, Roches recommended that the British provide naval instructors alongside the French military mission.Footnote ¹⁰¹ French military assistance had unnerved Parkes, who complained to London that that Roches ‘does not care to support very warmly the commercial policy of England’ but instead ‘prefers to minister to the military aspirations or vanities of the Japanese’ in the hope of becoming the Shogun's ‘Military Mentor’.Footnote ¹⁰² Consequently, Parkes made it his mission ‘to divert their attention from military glitter to industrial enterprise’, and was instrumental in the collapse of a French scheme to finance the Bakufu's military reforms.Footnote ¹⁰³ He even described the Yokosuka dockyard as ‘our Suez Canal question’, and was particularly concerned at any prospect of Shogunal support for French colonial expansion in Asia.Footnote ¹⁰⁴ Despite earlier British parsimony, Parkes now performed a complete volte face, writing that ‘England should have her fair share in such matters’ and arguing that a naval mission offered ‘as fair an opportunity of exercising a legitimate influence on their action as the French with their military schemes’. He even informed London that ‘I have had a little trouble to manage this point, for I thought I saw a disposition on the part of our French friends to monopolize arrangements that might minister to their influence.’Footnote ¹⁰⁵

Here, the Shogun's approach to military assistance mirrored long-standing efforts to manipulate principal-principal rivalries in order to bypass unpalatable treaty conditions. Moreover, as Roches's perceived influence with the Bakufu grew, so he became increasingly concerned with undermining British leverage – even at the expense of French policy. Despite Paris’ desire for increased trade with the Shogun, Roches wrote of the need ‘to avoid everything on our part which could cause unpopularity for his government’.Footnote ¹⁰⁶ He then deliberately undermined British attempts to use the Shimonoseki indemnity to force open new ports, pretending to be ill during the talks.Footnote ¹⁰⁷ Roches later declined overtures from Chōshū and Satsuma, demanding that they first subordinate themselves to the Shogun, in spite of Paris’ concerns that Roches was placing all France's eggs in one basket.Footnote ¹⁰⁸ Indeed, Roches’ dispatches home have been subsequently described as ‘not only uninformative … but also misleading’.Footnote ¹⁰⁹ For their part, the British described their naval mission ‘as a counterpoise to the French’, and evidently felt that the Shogun shared these sentiments.Footnote ¹¹⁰ Parkes relayed repeated Bakufu entreaties for the swift arrival of the British mission, and likewise observed Shogunal unease at the prospect of a French invasion of Korea, ‘chiefly occasioned by the apprehensions that appear to have crossed their minds as to the uses to which their new docks may be put’.Footnote ¹¹¹ Yet, such concerns did not prevent the Bakufu from appealing for further French naval assistance just two months later, ostensibly on the grounds that Britain might create a Hong Kong in Japan, to gain agreement for Bakufu naval cadets to gain experience on French warships.Footnote ¹¹²

While the Bakufu's adroit exploitation of Anglo-French rivalries undoubtedly reduced the political strings attached to foreign military assistance, the quality of military diffusion this balancing produced did not provide the desired domestic security. Chanoine, for example, remained sceptical of Bakufu military readiness, dismissing Roches's inflated perceptions of Shogunal strength. In late 1867, the Shogun resigned his office in a desperate bid to stave off civil war, though he refused to surrender the Tokugawa chieftaincy. This came as a complete surprise to Roches, highlighting his limited real sway.Footnote ¹¹³ In the ensuing Boshin War, a Satsuma-Chōshū led coalition systematically outfought the Bakufu's largely unmodernised levies, proclaiming the restoration of power to the Emperor Meiji. Pro-Tokugawa clans in the north continued to resist until the end of 1868, while the Bakufu navy established the Ezo Republic on the northern island of Hokkaidō, until finally overcome in the spring of 1869.Footnote ¹¹⁴ Roches joined the other Western powers in a pact of neutrality, largely in the belief that this would curtail British aid to the imperial faction. Unfortunately, it also curtailed Western aid to the Shogun, and Roches was recalled to France.Footnote ¹¹⁵ As the French Admiral Ohier remarked, ‘our situation in Japan has become difficult as a result of a revolution which had been forecast for a long time by everyone, but which our Minister refused to believe in’.Footnote ¹¹⁶

Admittedly, the extent to which parallel British support to Satsuma and Chōshū deliberately facilitated the Shogun's defeat has been the subject of some controversy. Gordon Daniels, for example, has argued that ‘the first objective of Foreign Office policy was to prevent civil war’, blaming Satow's later memoirs for misrepresenting British policy.Footnote ¹¹⁷ Satow afterwards claimed that Parkes had contributed to the ‘downfall’ of the Shogun ‘as far as lay in his power’, and his own sympathies certainly lay with the imperial faction.Footnote ¹¹⁸ Nonetheless, contemporary letters from other British legation staff echo Satow's observations. Willis, the legation doctor, described Parkes as ‘eminently aggressive’ and ‘a well placed man … to dissolve the established order of things’.Footnote ¹¹⁹ Moreover, in Daniels's analysis, Glover's arms shipments to Satsuma cannot be taken as evidence of British duplicity, given the secretive nature of the anti-Bakufu pact between Satsuma and Chōshū.Footnote ¹²⁰ Yet, Satow appears to have been aware of Satsuma's involvement with the imperial faction, and Satsuma agents even met the foreign secretary in London to push their cause.Footnote ¹²¹ Willis himself noted that Satsuma was ‘deeply in debt to the House of Glover & Co.’, providing ‘a great lever’ for Parkes to exploit in building relations with Satsuma.Footnote ¹²² Parkes certainly appears to have reconciled himself to the prospect of civil war, describing it as a potentially ‘powerful purifying agent’, and he later provided Glover with an introduction to the Foreign Office Permanent Secretary, describing him as one who had ‘done a great deal to reconcile the Daimios of the West to foreigners’.Footnote ¹²³ Although Satow's claim that Parkes ‘had felt the pulse of the Japanese people more carefully … than our rivals’ probably benefited from more than a little hindsight, the British undoubtedly hedged their bets by backing both sides – and condemned the Shogun in so doing.Footnote ¹²⁴

Importantly, French assistance gained far more traction in the absence of British competition, highlighting the negative effects of principal-principal rivalry on both French conditionality and recipient military change. During the civil war, for example, members of the French military mission absconded to Ezo with the tacit blessing of the French Army.Footnote ¹²⁵ These officers successfully used conditionality to exert significant control over Ezo's military and political preparations. The leading French officer, Charles Brunet, demanded that the Tokugawa officers:

Without any other foreign power to turn to, the Ezo leadership appears to have agreed. Visiting French officers remarked in wonder at how ‘everything passes through his [Brunet's] hands. The simple Japanese are puppets whom he manipulates with great skill.’Footnote ¹²⁷ Ultimately, this was not enough to prevent defeat at the hands of a larger and better equipped Meiji force, but Brunet's leverage with Ezo was enough for Paris to consider further French assistance to the Tokugawa. A French ministerial note of 1869 contemplated taking ‘indirect possession of the island of Hokkaido, perhaps even of a larger part of Japanese territory’ via support to Brunet, potentially enabling ‘our officers, transformed into daimyos themselves’ to ‘become the arbitrators and even the masters of the empire’. Ultimately, though, the scheme was rejected, not because the prospect of such control of the Bakufu was far-fetched, but because ‘one could, of course, expect some strong reactions from the Powers, especially England’.Footnote ¹²⁸ This same principal-principal conflict had enabled the Shogun to bypass unpalatable conditionality attached to French and British military assistance, but in so doing, subordinated the quality of military diffusion to political independence. Consequently, when tested against the parallel diffusion of its domestic rivals, aided and abetted by a British principal ambivalent about its selection of local agent, the Shogun's efforts at modernisation proved too little too late.

Meiji military diffusion, 1868–94

In many respects, the new Meiji regime was confronted with a very similar politico-military dilemma as its Bakufu predecessor. It inherited the Bakufu's international treaty obligations, while facing similar challenges to its domestic authority. Moreover, the national desire to check foreign influence and regain Japanese sovereignty remained much the same, fuelled by significant popular sonnō jōi expectations. Importantly, though, the new government placed far greater emphasis on treaty revision over limiting physical access. As one official remarked, ‘We must defend our imperial country's independence by revising the trade treaties … foreign countries’ troops have landed in our ports, and when resident foreigners break our law, they are punished by their countries’ officials. It can be said that this is our country's greatest shame.’Footnote ¹²⁹ Yet, the Restoration proved no less vulnerable to Western power than the Bakufu, at least at first. In early 1868, imperial troops from Bizen fired on Western representatives at Kobe, while troops from Tosa killed a party of French sailors at Sakai soon after. In response, the treaty powers collectively seized four Japanese steamers as a ‘material guarantee’, highlighting the ease with which Western neutrality might have shifted against the imperial cause during the still-ongoing civil war.Footnote ¹³⁰ To prevent a rift with the treaty powers, the Meiji regime elected to contain both incidents quickly, making immediate arrests and executions in marked contrast to the Shogun's previous foot-dragging.Footnote ¹³¹ Moreover, the Meiji leaders declared their intention to ‘open the country and cooperate with other powers’, and obtained imperial ratification of the Ansei Treaties as a demonstration of good faith.Footnote ¹³² Even so, Japan was still obliged to sign further treaties with Spain, Sweden-Norway, the North German Federation (Prussia's successor), and Austria-Hungary in 1868–9; the latter winning fresh concessions that accrued to all the treaty powers. Consequently, the necessity of further military diffusion formed a central pillar in efforts to build a strong, centralised Japanese nation-state on Western lines, and thereby buttress the Meiji regime's domestic and international position.Footnote ¹³³

Initially, however, the Meiji government sought to disentangle itself from foreign military assistance, preferring instead to pursue emulation as the Bakufu once had. This was in part a cost-cutting measure, though Meiji leaders equally distrusted foreign missions who had until recently been advising their Bakufu enemies.Footnote ¹³⁴ In an effort to maintain France's leverage with the new government, Roches replacement Maxime Outrey adopted a dual tack. He sought to build bridges with the Meiji regime, disavowing Brunet's activities in Ezo, and even waived some of the Sakai executions.Footnote ¹³⁵ Simultaneously, he applied coercive conditionality to military aid to assert French interests. Outrey used the Shogun's debts as a pretext to maintain control of the Yokosuka arsenal, while similarly threatening ‘pains and penalties in the shape of compensation for non-fulfillment of a contract’ should the French military mission be dismissed.Footnote ¹³⁶ Unfortunately for Outrey, the Meiji oligarchs also continued with the Bakufu's policy of exploiting divergent French and British interests to play one military partner off against the other. A British loan promptly settled the Shogun's debts, securing Japanese ownership of Yokosuka, though Verny remained superintendent. Parkes informed London that a ‘French design I think has been effectually checked both in regard to these Docks (or Arsenal) which might easily have become French territory, and military influence’. The British even agreed to withdraw their own naval mission in order to pressure French withdrawal. In justifying this to Whitehall, Parkes explained that the new Meiji government ‘cordially desired to be relieved of the French’ but ‘felt their difficulty in dismissing one [mission] and retaining the other’. In truth, the Bakufu wanted to be rid of both missions, but for Parkes, the important point was that ‘the removal of our men carried with it that of the French’ which meant that ‘a cherished project on the part of the latter has been destroyed’.Footnote ¹³⁷

Instead, the Meiji government attempted to emulate Western national armies, as part of wider efforts to dismantle Japan's feudal system. In 1869, a new Ministry of Military Affairs was established, and all domain forces were directed to adopt a standardised British naval and French military model the following year. The previous proliferation of Dutch, French and British systems presented real problems for centralised control. As Satow explained, ‘Japan could not be strong as long as it was open to every daimio to withdraw his assistance at his own pleasure, and each prince to drill his troops after a different fashion.’Footnote ¹³⁸ However, many domains ignored the new injunction, while Kishū even began to reform along Prussian lines.Footnote ¹³⁹ Moreover, the Meiji oligarchy was itself divided on how best to raise a national force, with Chōshū favouring a conscript army while Satsuma preferred a professionalised samurai force. Indeed, the prospect of conscription led to the Army Minister's assassination, while an initial quota-based attempt in 1870–1 produced mediocre results.Footnote ¹⁴⁰ In Jaundrill's words, ‘the lack of standardization revealed the relative weakness of the new government vis-à-vis the domains’, compelling the new government to create a standing Imperial Guard from their own loyal domain troops in 1871. Shortly after, the hereditary domains were dissolved and converted into national prefectures, with their forces either disbanded or subsumed into imperial garrisons. The near bankruptcy of many clans precluded serious resistance, but enduring indiscipline in the new regional garrisons raised serious questions about the reliability of former domain troops, reinforcing the necessity of renewed military assistance.Footnote ¹⁴¹

The inherent difficulty of constructing a modern national army from diverse Japanese fiefs obliged the Meiji leadership to seek renewed Western military assistance. In 1870, the Meiji government approached Outrey for a new French military mission, and fresh instructors were finally appointed in 1872. Inevitably, the intervening onset of the Franco-Prussian War delayed the French response, but the hiatus equally reflected renewed French efforts to use military assistance as a tool of leverage in wider Franco-Japanese relations. At the time, Outrey was threatening naval action in response to Meiji persecution of Japan's indigenous Catholic population, and sought to make military assistance conditional on religious toleration. Outrey eventually relented, suggesting that the potential strategic benefits of military cooperation were ultimately more important to France than Catholic emancipation. Nonetheless, his decision to view religious toleration as an expected outcome (rather than a necessary precondition) for military assistance was also the result of enduring principal-principal rivalries, with protestant British and Prussian propaganda casting aspersions on Catholic proselytisation gradually undermining Outrey's stance.Footnote ¹⁴² Indeed, the importation of an alternate Prussian military model was certainly favoured by some Meiji officers at the time, though the Shogun's previous investments in French instruction ultimately won out – along with a desire to be seen as reliable in international engagements – resulting in the confirmation of French assistance despite their recent defeat to Prussia. Importantly, Outrey himself believed that the French had been reappointed largely to counter the British, who commanded over half of Japan's international trade, and were likewise requested to provide new naval instructors in 1872.Footnote ¹⁴³ With hindsight, it seems more likely that the British were intended to balance the French. Of the 66 foreign advisors to the Navy Ministry in 1874, 29 were British and 36 French; the latter concentrated at Yokosuka. Certainly, British instruction was not especially popular; cadets at the naval academy even claimed to be ‘more than happy to see their parents fall ill’ just to get leave.Footnote ¹⁴⁴

Much as before, renewed instruction was seen by France as a means to secure its interests in Japan, while also conferring a decent helping of much-needed military prestige.Footnote ¹⁴⁵ Indeed, assisting with the creation of Japan's first truly national army afforded the French mission a central role in the Meiji government's wider nation-building agenda. Conscription, in particular, was an important step in the dismantling of Japan's feudal system, accompanied by the successive revocation of samurai class distinctions and privileges. Here, the 1873 conscription law was heavily influenced by French practice, down to the inclusion of numerous French exemptions (adjusted to Japanese society) and even the ill-advised Meiji usage of the French slang term l'impôt du sang, or ‘blood tax’, to describe Japan's new citizenship obligations.Footnote ¹⁴⁶ French officers likewise assisted with the new army's organisation and doctrine, establishing specialist schools for training NCOs, military skills, and technical branches, alongside a new officer academy modelled on St Cyr. Indeed, the mission's duration was repeatedly extended, and its scope similarly expanded to include the Imperial Guard, rising from and initial strength of 16 instructors to a high of 45.Footnote ¹⁴⁷ Simultaneously, though, Parkes also viewed British naval instruction as a means to block rival powers’ inroads. In 1870, he pushed the Royal Navy to accept Japanese naval cadets onto British warships, after Japan accepted places at the US Naval Academy. He likewise supported the renewal of a British naval mission, informing Whitehall that ‘if we were to show any hesitation in furnishing them with Naval officers they would doubtless be able to obtain these without difficulty from other Powers’, having ‘already obtained from the French Government a numerous corps of Military officers’.Footnote ¹⁴⁸

Importantly, this balancing allowed the Meiji government to insist on contracting new instructors individually, rather than receive official delegations as the Bakufu had done, enabling Japan to place greater formal limits on the influence of its foreign advisers. The French accepted contracting provided the Japanese paid handsomely and contracts remained subordinate to wider French interests in Japan. Military instructors, for example, still had to be approved by the French War Ministry, and stayed under French command and jurisdiction.Footnote ¹⁴⁹ Indeed, both Britain and France expected further Japanese political concessions as their contracted missions became established, and looked to renegotiate the Ansei Treaties early as a result. France even demanded the complete opening of the Japanese interior in return for tariff reform, which Japan could not accept.Footnote ¹⁵⁰ Together, however, contracting and balancing increased the precarity of military advisors, providing a greater degree of Meiji control over the hiring and firing of foreign officers, and concomitantly limiting the political leverage assistance generated. In 1874, for instance, the Meiji government dispatched an ill-advised expedition to Taiwan to placate domestic hard-liners. The French minister recommended the military mission be withdrawn in the event that the conflict spillover into war with China, in order to bring Japan to heel. Yet Paris baulked at such a risk, and the mission was actually extended as a result of the debacle. Similarly, the lack of French command authority allowed Japanese officers to obstruct unpalatable French training and gerrymander new promotional processes, despite their ostensibly privileged position.Footnote ¹⁵¹ In 1875, the head of the British naval mission returned home before his contract expired, after a comparable spat over the extent of his authority. Moreover, 23 British naval instructors were summarily dismissed in 1879 after Britain spearheaded efforts to block tariff reform, while French contracts were allowed to expire shortly after.Footnote ¹⁵²

Instead, Japan looked to a new principal, Germany, to complete its military modernisation. From 1884, Germany provided instructors for the Japan's new staff college, precipitating a wider restructure of the army overseen by the Prussian Major Meckel. Japan adopted Prussia's field service regulations and divisional structure, and conscription was also revised to create a Prussian reserve system in 1889, removing French-style exemptions.Footnote ¹⁵³ This shift has traditionally been explained through a mixture of functional and normative imperatives. In 1877, Government victory over reactionary samurai in the Satsuma Rebellion required eight months and the entire Meiji army of 32,000 troops, plus a further 10,000 reserves, revealing inadequacies in the army's higher command and control later attributed to French instruction.Footnote ¹⁵⁴ By then, the French army had itself adopted a Prussian-style staff system, while Imperial Germany may also have seemed a more appropriate model than republican France to Meiji oligarchs concerned by domestic interest in democratisation.Footnote ¹⁵⁵ Yet, equally, this shift from French to Prussian assistance also reflected Japan's established policy of playing rival Western principals off against each other to get what it wanted. While the Satsuma Rebellion undoubtedly affected Japanese perceptions of the French military mission, French contracts were still renewed afterwards, while Prussian assistance was not sought until a full seven years later. Moreover, French officers were already beginning to reform Japan's staff processes when they left in 1879, and France was again requested to provide fresh instructors in 1880, but refused to accept the pay Japan offered.Footnote ¹⁵⁶ Moreover, if German assistance was primarily sought for functional reasons, it is unfortunate that Meckel's preferred views on tactics and doctrine differed significantly from mainstream Prussian military thinking, and were in many respects quite reactionary.Footnote ¹⁵⁷ Importantly, while further Meiji attempts at treaty renegotiation stalled in 1882 due to Anglo-French obstinacy, Germany was seen as increasingly sympathetic.Footnote ¹⁵⁸

The importance of principal-agent politics in explaining this shift is reinforced by the fact that some French reforms were retained, and new French instructors were engaged alongside the Prussians. This ensured neither party gained a free hand, melding both military traditions in the process. These latest French officers instructed at the military academy and specialist schools, and so did not benefit from the same level of access or prestige as Meckel's staff mission. However, the simultaneous employment of both German and French advisors stimulated significant rivalry, actively exploited by Japanese officers. The senior French officer was promoted to prevent Meckel out-ranking him, and promptly used his platform at the officer academy to compete with Meckel, delivering rival staff lectures with the active encouragement of senior Meiji commanders. For their part, the Germans vociferously objected to the French presence, and structural reforms in 1887 conveniently reduced the status of Japanese initial officer training – where the French were based.Footnote ¹⁵⁹ Nonetheless, Meckel himself was sent home on suspicion of spying in 1888, to be replaced by another Prussian officer. France withdrew their officers soon after, in a desperate bid to pressure Japan into renewing their own chief instructor's contract, but to little avail.Footnote ¹⁶⁰ Even the British were not immune to such rivalries. Although unconcerned by Prussian military involvement, Britain repeatedly offered additional naval aid in an effort to maintain its standing with the Japanese navy. In 1884, the British agreed to educate Prince Yamashina Sadamaro at Dartmouth Royal Naval College to prevent him joining the German navy, and subsequently attempted to prevent Japan from buying French warships over British, claiming that naval effectiveness required the adopted system to ‘be uniform and continuous’.Footnote ¹⁶¹

While this approach mirrored previous Bakufu policies, the Meiji government's centralisation of domestic power significantly improved its efficacy, precluding Western sponsorship of alternate domain agents. Admittedly, each power instead cultivated coteries of sympathetic Meiji officers as a vehicle for their own national interests, exacerbating inter- and intra-service politics.Footnote ¹⁶² Nonetheless, as one foreign advisor observed, Westerners remained ‘simply helpers and servants, not commissioned officers, and have no actual authority’, while another concluded that ‘the Japanese had made up their minds to make what use they could of their foreign servants, but in no case to have them become masters, or to invest them with any power’. Indeed, foreign advisors were colloquially known as yatoi in Japanese, or ‘hired menial’.Footnote ¹⁶³ Occasionally, competition between the powers did impede Meiji goals. Foreign garrisons remained at Yokohama long after the direct threat to Western citizens had passed, in part because the British refused to withdraw unilaterally while French felt their presence counteracted British ‘ideas of absolute predominance’.Footnote ¹⁶⁴ Yet, the exploitation of these rivalries remained central to Meiji strategies for preserving its agency. One French diplomat decried the ‘lack of entente’ among his Western colleagues:

By 1880, the French Minister complained that ‘Japan has been able to throw enough dust in our eyes to make us believe … it was going to be our champion’, while his successor eventually concluded that the Japanese ‘seek to borrow from us our methods of civilization only in the hope of being able one day to make use of them against that civilization and against us’.Footnote ¹⁶⁶

It might reasonably be argued that the shape of Meiji military modernisation responded to domestic instability more than external threat, reducing the significance of international principal-agent tensions in the patterns Japanese diffusion. Notably, conscription has been described as a poor mechanism for external defence but vital to break samurai power, while the ensuing focus on army-building over naval development is seen as indicative of domestic rather than international priorities.Footnote ¹⁶⁷ However, Meiji domestic consolidation should not be seen as distinct from its international relations, and military diffusion featured prominently in Meiji responses to both issues. To the extent that conscription mirrored European norms, this must be seen in the light of British statements that Japanese sovereignty would be returned only ‘in precise proportion to their advancement in enlightenment and civilization’ in Western eyes.Footnote ¹⁶⁸ Equally, while conscription undoubtedly provided a vehicle for nation-building, this was itself a response to Western exploitation of Japan's feudal divisions. Conscription was also expected to free up financial headroom for expensive naval defences, while the extended time it took to train naval officers made army reform a more rapid route to deterrence.Footnote ¹⁶⁹ Although Kublin dismissed talk of invasion as a ploy to justify high military expenditures required for internal security, Meiji commanders spent significant time preparing territorial defence plans during the 1880s, with senior officers ‘consumed’ by fears for national survival that, in Jaundrill's view, bordered on ‘paranoia’.Footnote ¹⁷⁰ By then, Meiji interest in the trappings of Western ‘civilization’ came a firm second to military diffusion, and the revision of the unequal treaties in 1894 ultimately reflected a hard-won symmetry in military power between Japan and the West.Footnote ¹⁷¹