How Did Japan Find the Men for Its Armies?

In 1873, officials in four villages in Ishizu District needed more than a list of potential soldiers. A surviving register in the Hibi family papers includes lists connected with the national militia, conscription, and exemption from conscription. Finding the men who would serve also required identifying men who would not.^[1]

This was the year the Meiji government introduced its national conscription ordinance. The language of a common military obligation could make the administrative problem sound simple: instead of depending on a hereditary warrior class, the army would draw on the population. The local records preserve the sorting that such a statement leaves out.

Earlier armies had required sorting too. The ancient court organized recruitment through registered households and provincial institutions. Sixteenth-century lords demanded contingents whose size and equipment were specified in relation to a retainer's assessed resources. The national government later applied categories to individual men, while still requiring information about their households.

These arrangements were more than alternative ways to count the same kind of soldier. They identified different people as answerable for delivering military service. Following that responsibility helps explain how an army could be assembled—and why the men available to it were never simply everyone a ruler could name.

Japan's early statutory state had a system for drawing men into military service centuries before the samurai governments. The ritsuryō codes, the laws organizing that state, made military recruitment part of a wider administrative undertaking.

Historian Shimomukai Tatsuhiko traces the system's formation through reorganizations between the mid-seventh century and the Taihō Code of 701. His account begins with forces mobilized through regional chiefs. Successive reforms established local arms depots, grouped households for recruitment, and eventually based conscription on household registers. The aim was to create forces with more regular organization and common training.^[2]

The household was useful because it joined a person to an identifiable local unit. An authority trying to raise men needed somewhere to direct its demand and a means of establishing who belonged there. Registration supplied a framework through which that claim could be made repeatedly.

It also placed military service inside the household's economy. A person assigned to the army had previously occupied time and work elsewhere. Even when the rule was addressed to the man, his absence changed what remained possible for those around him. Recruitment began before anyone reached a battlefield.

The statutory system divided the work of making soldiers available. In Shimomukai's analysis, the military ministry and provincial governors handled peacetime administration; the provincial corps organized soldiers and training; the central council directed mobilization; and commanders led the force assembled for war.^[3]

This division made a recruit's relationship with the state a sequence rather than a single encounter. Being recorded locally, undergoing preparation, and joining an expedition involved different decisions. The official who maintained an obligation did not have to be the commander who later exercised it.

Shimomukai also identifies government control of drums and battle flags as part of the earlier reforms.^[2] Such objects belonged to the organization of command. Men gathered from different households needed a way to act together once assembled. Common training and recognizable signals addressed a requirement that household registration alone could not meet.

The arrangements subsequently changed. The conventional turning point of 792, when much peasant conscription was abandoned, has itself prompted debate. As Sean O’Reilly explains in his survey of the scholarship, Karl Friday distinguishes the kondei, guards assigned to provincial facilities, from a replacement field army. Calling them the successors to the old corps can confuse a change in local security duties with the formation of new campaigning forces.^[13]

The distinction keeps the later rise of warrior forces from becoming the automatic consequence of one decree. Guarding a storehouse, maintaining a trained following, and providing troops for a campaign were tasks that authorities could assign through different relationships. By the sixteenth century, a military lord’s demands could center on a retainer expected to bring a force of his own.

In 1581, Hōjō Ujinao, head of the military house ruling from Odawara in eastern Japan, specified the obligation of his retainer Ikeda Magozaemon-no-jō. Ikeda was to lead fifty-six men, himself included. The document also prescribed their equipment.^[4]

The number was connected to an assessed holding of 389 kan and 100 mon, expressed in monetary units. This included resources associated with yoriko, lower-ranking retainers assigned to Ikeda by the Hōjō. The assessment was a basis for military obligation, rather than a statement that this exact sum sat in Ikeda's cashbox. Spears, firearms, protective headgear, and armor appeared among the specified requirements.^[4]

A lord could consequently demand a body of men through a relationship with one accountable leader. Ikeda's duty was not discharged merely by presenting himself with a weapon. It included bringing an organized contribution of a stated size and composition.

The order attached manpower and equipment to resources controlled through a retainer. Its account included the lower-ranking men assigned to Ikeda, so his contribution was larger than the service of his own immediate household. The lord was assembling forces through an intermediate level of organization, whose leader could be held answerable for what the whole group supplied.

Numbers and equipment also gave the Hōjō a standard against which to judge that contribution. A retainer could not meet a requirement for a specified group simply by offering whichever men and weapons happened to be convenient. Delegated assembly preserved a place for detailed supervision from above.

A second Hōjō order shows those requirements being activated for an occasion. The city catalogue dates it tentatively to 1561, because the document lacks a year. Addressed to Daitō Shikibu-no-jō and groups of foot soldiers, it called for the prescribed numbers and presentable equipment when meeting forces from the allied Takeda side.^[5]

Here the anticipated encounter made a standing obligation urgent. Someone had to turn the expected contingent into men who could appear together. Counting them and inspecting their preparation were ways of testing whether an established service relationship could produce what it promised.

The orders record demands, not attendance returns. Their importance lies in the mechanism they make visible: a military lord could mobilize resources through subordinate leaders, then hold those leaders to a specified contribution. Calling everyone involved “samurai” would conceal much of that internal organization.

Another order from the same Hōjō government went to a carpenter.

In 1579, Sōzaemon of Tawara was summoned to Odawara the day before castle repairs were to begin. He had to report regardless of the weather; even illness did not excuse him from receiving instructions. A day’s delay carried a threat of punishment.^[6]

Fortification work followed Takeda Katsuyori’s alliance with Uesugi Kagekatsu in 1578, which placed the Hōjō under threat of attack.^[6] A change in relations among military houses reached a craftsman as a deadline for reporting.

This did not transform the carpenter into an infantryman. It shows why the question of military manpower cannot be settled by counting combatants alone. A castle requiring repairs imposed demands on skilled work outside the fighting contingent. The people directing preparations needed that work at a specified time, rather than whenever its provider happened to be free.

The distinction becomes useful when thinking about what a territorial lord controlled. A register of warriors might identify men obliged to fight. The carpenter's summons reached through another relationship, linking occupation, locality, and command. Different channels could feed the same military undertaking.

The illness clause left the assignment with the authorities: Sōzaemon was required to report before the question of his work could be settled. His response is not preserved in the catalogue account.

Placed beside the contingent orders, the document reveals the variety concealed by a general statement that a lord “raised an army.” Some people were expected to lead others. Some had to appear within the required body of foot soldiers. A craftsman received a separate summons because his work served the fortifications.

A shortage in one kind of work could constrain the rest. More combatants could increase the numbers assembled without supplying the craft skill required by a damaged fortification. Military preparation depended on matching demands to capacities outside the army as well as filling positions within it.

That need for differentiated preparation did not disappear when warfare became less frequent under the Tokugawa order.

Under the Tokugawa shogunate, territorial domains retained military establishments of their own. The Uesugi records from Yonezawa preserve returns organized by commanders and units in 1631. Other records specified equipment for lower-ranking retainers and foot soldiers. The archive also contains an 1862 copy of a military-service register from 1654, stamped by the domain's record office.^[7]

The two dates on that last item matter. It is evidence that an earlier arrangement was copied in the nineteenth century, not proof that every seventeenth-century obligation remained unchanged or was fulfilled in 1862. Yet the act of copying shows that military organization remained something a governing house preserved and consulted.

A campaign imposes an immediate deadline. Between campaigns, authorities face a different problem: how to keep an obligation recognizable when no current battle is testing it. Specifying units and contributions gives later officers a starting point. The paper can preserve what someone is supposed to supply even while the people, resources, and practical demands change.

This continuity also helps explain why nineteenth-century military reform was more complicated than replacing an absent army with a newly invented one. There were existing service relationships to alter, expectations to override, and organizations through which experiments could be made.

D. Colin Jaundrill's study of military service places the origins of the later conscript army within reforms extending back before 1868. Its account of Western-style musketry in the 1840s and 1850s concerns both the shogunate and domains, rather than a national army already organized under one command.^[8]

Changing the desired military performance could disturb an inherited allocation of obligations. A ruler might know which household owed service without having the right arrangement for teaching and coordinating a different kind of unit. The ability to demand men and the ability to organize their work remained connected tasks, neither reducible to the other.

The transition to national service therefore involved a change in who could make the demand and in the units through which men were supplied. The new government would increasingly claim the individual male subject directly. But it still had to turn that principle into an institution with an answer to a very practical question: which men should report this year?

The conscription ordinance of January 1873 specified a process of selection. Men reaching twenty were to undergo examination; those who passed were subject to selection by lot for three years in the standing army. Four further years brought liability for wartime recall. A wider register encompassed men aged seventeen to forty in the national militia.^[9]

These were different relationships with military service. A name in the wider register did not mean that its bearer was currently living in a barracks. Eligibility for examination did not guarantee selection. Completing an initial term did not end every military claim upon the person.

The ordinance also exempted categories of men from standing-army service. They included household heads, government employees, students at public institutions, and people undertaking certain professional studies. Men who paid a prescribed fee could also be excused from that service.^[9]

The exemptions belonged to the system's stated rules. They should not all be grouped under evasion, as though the only authentic form of conscription were one in which every eligible man actually served. The authorities were defining a broad claim while distinguishing obligations and activities that could qualify it.

That arrangement distributed the burden unevenly. Two men of the same age could face different demands because they occupied different household positions or had access to different institutions and resources. A common language of national service did not eliminate those differences; it made them part of the question of who would be selected.

Selection by lot addressed another distinction: the pool of men who might serve could exceed the number taken into the standing army. An obligation to make oneself available was wider than the act of entering that army. Keeping those categories separate is essential to reading the local registers.

The national change was nevertheless substantial. The Hōjō order had demanded an equipped contingent from a retainer. The conscription process assessed individual men through rules that the national authorities could apply across localities. A household's place in a particular lord's service no longer supplied the only route into military obligation.

The principle did not make recruiting uncontested. Jaundrill's account of the early conscription years includes resistance among communities approached to replace soldiers who had deserted posts in Kyushu. He presents these episodes as conflicts within the establishment of national service, rather than evidence that the population entered its new military role without objection.^[10]

A law could establish the category. Recruitment still had to happen among people whose lives were organized around other expectations.

The village register from Ishizu belongs at this point in the story. Its different lists were tools for making distinctions required by the new arrangements: men within the wider military register, men connected with conscription, and men recorded as exempt.^[1]

Their coexistence is more informative than an isolated total would be. Even before examining individual names, the archive's description shows that local record-making was concerned with several possible outcomes. National service produced paperwork for absence as well as presence.

The authorities needed the work done to a schedule. A Gifu prefectural notice of 7 December 1873 called for urgent preparation and submission of the second conscription and national-militia registers. The document was addressed through local household-registration administration.^[11]

This was one of the points at which a national claim depended on local knowledge. The central rule could define a category, but someone nearer the household had to establish which people belonged within it. A list also had to reach the authority that required it. A rule sitting in one office and information remaining in another could not yet select a recruit.

Household circumstances made the process more than copying names. An Army Ministry file from 1883 concerns a query from Tochigi about exemption for a replacement heir. The internal discussion distinguished succession following a predecessor's death from replacing a designated heir without an accepted reason. The proposed response denied exemption unless the replacement followed the earlier heir's death.^[12]

The significance here is administrative rather than a general account of adoption. Officials needed a sequence of events. A person's current description as heir did not by itself establish the exemption: how he had acquired that position could change the answer.

Rules of this kind made military selection depend on the history of relationships outside the army. They also generated questions that had to travel back upward for interpretation. National law could standardize the terms of a decision while leaving officials substantial work in establishing which term applied.

The surviving file is a proposed ruling on a query, not a measure of how frequently such cases occurred. Its narrowness is useful. It reveals a particular place where a general military obligation encountered a household's specific history—and where the decision about one man's service required more than his age or physical presence.

Across these cases, the decisive question is who was answerable for producing military service.

The statutory court worked through households, registers, and provincial military organizations. The Hōjō demanded specified contingents through retainers while issuing separate orders for necessary skilled work. The national conscription system applied rules to individual men, using local records to decide which of them were available, exempt, or subject to a continuing claim.

These were not successive steps from no organization to organization. Each system had its own means of counting, specifying, and demanding. Nor did a change in the formal unit remove every earlier relationship. Household circumstances continued to affect the selection of national conscripts, just as a lord's order to one retainer reached the subordinate men for whom he was responsible.

What changed was the arrangement of those relationships. A commander receiving an already organized contingent depended on another leader to deliver it. A conscription authority selecting individual recruits depended on a different set of officials to identify and classify them. The responsibility for assembling the force moved, even though both systems continued to need information and cooperation beyond the camp.

The same comparison explains why recruitment cannot be read directly from the number of men a society contained. Population supplied possible people; an institution had to establish a claim upon particular ones. That claim could be limited by an exemption, mediated through a superior, or directed toward work other than fighting.

The men who stayed outside the army also remained part of its history. Some were formally exempt. Some supplied labor that military preparations required. Others belonged to households from which somebody had been taken. Their different positions shaped what the armed force could demand without appearing as equivalent entries on a battlefield muster.

Japan's armies were assembled through these changing divisions of responsibility. A list was useful because it connected a name or a household to a demand that officials could pursue. But every such connection still had to reach a person. The history of recruitment lies in how that passage was organized, where it met resistance, and which other obligations were allowed to interrupt it.