What Were Japan's Prisons For?

In 1755, the Kumamoto domain established a place where offenders could be confined and made to work for terms of one to three years. Part of their earnings was retained to provide resources on release. The legal historian Takashio Hiroshi describes the arrangement as a punishment intended to prepare people for a livelihood afterward.^[1]

This was more than a century before the end of Tokugawa rule in 1868. Yet it would be equally misleading to imagine every Tokugawa jail as an earlier version of that institution. Many held people awaiting the decision about what should happen to them. Confinement could precede the punishment rather than constitute the punishment itself.^[2]

The distinction changes the question we ask of a prison. Its walls tell us that people could be kept inside. They do not tell us why they were there, what the authorities expected to accomplish, or whose decision could let them out.

Following those questions from Tokugawa institutions into the Meiji period reveals a history larger than the replacement of wooden jails by brick prisons. Confinement became a more extensively specified punishment, while retaining older functions and acquiring other demands. A law could count a prisoner's years. Someone still had to decide how those years would be lived.

In his account of Edo's penal order, the legal historian Ishii Ryōsuke distinguishes ordinary jail custody from imprisonment imposed for a fixed period as the punishment. The first was central to the jail's work; the second was not the general organizing principle. Ishii also identifies exceptions, including forms of prolonged or punitive confinement. Even this basic distinction therefore needs to be made without claiming that Tokugawa Japan never punished anyone by keeping them locked up.^[2]

Consider what waiting for judgment required of the institution. Its immediate purpose was to keep a person available while officials investigated and decided the case. The question governing the stay was what those officials would decide next. A sentence of three years, by contrast, made a specified duration part of the decision already reached.

Those arrangements could produce similar outward scenes while establishing different relationships with time. In one, custody belonged to an unfinished process. In the other, continuing custody carried out its result. That difference mattered even when both prisoners experienced confinement as suffering.

Nor did every accusation require a place in the city jail. Ishii notes that people involved in lighter cases could be kept under the responsibility of households or local communities.^[2] The authorities had more than one way to secure a person's continued availability. A locked institution and an obligation imposed on people outside it were alternative means of custody.

Inside Edo's jail, responsibility was layered too. Ishii describes officials and staff answerable through the city magistrates, alongside positions held by inmates within the institution.^[3] Formal control from outside and the organization of life among prisoners were connected but distinct. Calling the jail a government institution does not mean that a magistrate personally directed every encounter within it.

Kumamoto's establishment belongs against this background. It was a particular domain's use of confinement, work and a term of punishment. Its existence prevents a later national system from being mistaken for the invention of every practice it brought together. It does not make one domain's arrangement a description of the whole country.

The Ishikawajima ninsoku yoseba, established in Edo in 1790, presents a different problem. The name designated an institution for compulsory work and the supervision of people classified as mushuku. That category concerned removal from household registration; it should not simply be translated as either “homeless” or “criminal.”^[1]

Takashio explains that the original arrangement chiefly concerned people whose punishment had already ended. Their further confinement was not formally the same thing as a new criminal sentence. Earnings could accumulate toward a livelihood after release, while accumulated resources or an accepted guarantor could help make departure possible.^[1]

The gate was therefore answering a question that the completed punishment had left unresolved. Where would this person go, and under what arrangement could the authorities allow him to leave? A declaration that a penalty was over did not automatically supply work, recognized connections or someone prepared to take responsibility.

The accumulated earnings had a double function. They could make an eventual departure more feasible. While officials controlled their accumulation and release, they also helped organize the person's continued dependence on the institution. Resources offered for a future outside could be produced through labor that the person was not free to refuse inside.

The guarantor made the relationship wider than a prisoner and a guard. An acceptable person beyond the gate could become part of the conditions for leaving it. Release depended partly on recognition by others, not simply on a change that an official could observe within the confined individual.

In 1820 the institution also began receiving people sentenced to banishment. Their confinement acquired a prescribed term, but release still had to respect the prohibition on returning to the excluded area.^[1] Leaving Ishikawajima could therefore end one restriction while another remained in force.

This change is a warning against assigning a single timeless purpose to a named establishment. Its original work of compulsory reintegration and its later relationship with a banishment sentence were not identical. The population admitted, the reason for custody and the conditions of departure could change while the institution's name remained familiar.

The useful comparison with Kumamoto is consequently not which place was more “modern.” It is how confinement related to the judgment. In one arrangement, a term of compulsory work was the punishment. In another, officials could continue to control a person after the punishment had ended because they considered the conditions for life outside unresolved.

The Meiji government's prison rules of 1872 attempted to specify how confinement should be organized. Their associated plans and equipment designs made the prison a subject of national prescription. Ohara Shigeya, who helped prepare the rules after studying prisons overseas, was concerned with arrangements that could be reproduced rather than merely with the name given to an institution.^[4]

The rules' preamble described the prison as a place of restraint and correction. It expressly associated its purpose with humane care rather than cruelty, and correction rather than the infliction of suffering. Yet the same documentary collection prescribed restraints and disciplinary measures. The stated intention to avoid cruelty did not remove the institution's power to compel.^[4]

That combination deserves more attention than either half alone. A rule about treating prisoners humanely could supply a standard against which conduct inside was assessed. It did not by itself settle which forms of coercion the rule-makers considered compatible with that aim.

Plans also shifted the level at which decisions could be made. A centrally prescribed arrangement offered an answer to questions previously confronted within particular establishments: where custody would take place, how people would be separated and what equipment would be provided. It made the interior of a prison something that officials elsewhere could attempt to define.

Specification was not the same as completion. A drawing required a building; an instruction required staff and supplies. The existence of the 1872 documents establishes a program and its declared purposes. It cannot certify that every prison immediately acquired the prescribed rooms or that every prisoner experienced the treatment promised.

The earlier cases help place the change. Compulsory work and plans for life after release already had Japanese institutional precedents. What the new rules added was an effort to organize prison administration across the new national state. This was a change in the reach and form of prescription, not a sudden discovery that confinement could have purposes beyond keeping a suspect available.

The Penal Code promulgated in 1880, effective from January 1, 1882, shows more precisely what the state was asking its institutions to do. Its provisions distinguish forms of custody by work, location and duration. The following comparisons concern that code, not the meanings those legal terms acquired under later laws.^[5]

Articles 22 and 23 provide a particularly clear pair. Both divided the punishments they described into heavier terms of nine to eleven years and lighter terms of six to eight. Under Article 22, the prisoner entered a mainland penal establishment and performed prescribed labor. Under Article 23, the prisoner entered a mainland prison without that labor obligation.^[5]

The matching durations expose the difference. A term was not a complete description of the punishment. The law could take the same number of years and assign a different obligation within them. To know what a sentence demanded, an administrator needed its legal category as well as its length.

Location mattered separately. Other articles provided for dispatch to islands, again distinguishing labor from confinement without prescribed labor. The code also retained fines and the death penalty.^[5] The growing elaboration of imprisonment did not turn every punishment into a uniform quantity of time behind the same kind of wall.

Article 49 addressed the apparently small business of calculation. A day contained twenty-four hours, a month thirty days, and a year followed the calendar. The first day counted even if it began partway through; the day of discharge did not. Subsequent provisions dealt with the starting point of a sentence, the effects of an appeal and time spent at large after escape.^[6]

Such clauses connected a court's decision with the work of keeping someone confined. “Three years” had to become a date from which officials could calculate and an endpoint they could recognize. A sentence expressed in time created a practical obligation to account for that time.

It also constrained the question an official was supposed to answer. A person serving a specified term was not simply to remain until somebody found him socially convenient to release. The judgment and the rules for calculating it supplied an endpoint. The code's additional provisions could qualify that endpoint, but qualification itself had to be described.

Work produced another account. Article 25 directed that earnings from prescribed labor be divided, under prison regulations, between institutional expenses and the prisoner. It excluded the first hundred days of actual labor from the prisoner's allocation.^[7] The statute was not promising that every hour would immediately produce personal savings available to spend.

The division joined two purposes that might otherwise be narrated separately. Labor could help sustain the establishment while also producing an allocation for the person required to perform it. The provision defined a relationship among work, institutional cost and the prisoner; it did not supply evidence that the resulting resources were sufficient for a livelihood afterward.

There was also a division between the code and the rules that would give it effect. Article 11 left detailed methods of executing punishments and restraining offenders to separate regulations.^[5] A national statute made the basic categories explicit while leaving substantial work to administration. Reading only the list of sentences misses that second layer of decisions about daily life.

The history of penal labor in Hokkaido makes that daily life difficult to imagine solely through a permanent cell. At Abashiri, prisoners were dispatched to road construction in 1891 and accommodated in temporary quarters along the work. As sections were completed, the camp moved. The Abashiri Prison Museum describes more than a thousand prisoners involved in that year's undertaking.^[8]

Its reconstructed temporary quarters represent a prison capable of changing address. The exhibit was built in 1983 to illustrate the earlier arrangement; it is not a camp building preserved unchanged from 1891.^[8] The historical point lies in the organization it describes: guarded custody accompanied the labor force.

An open construction site was therefore not the opposite of imprisonment. What made it a place of confinement was the continuing power to direct people's movements and prevent their departure. Walls were one means of exercising that power. Guards and a moving camp supplied another.

The work also gave penal administration a purpose beyond the individual sentence. A road was useful outside the institution that supplied its labor. Officials could consequently judge the undertaking through progress toward a public project while a prisoner experienced it as compulsory work under custody.

Tsukigata's municipal history records the costs of related road-building campaigns. It describes inadequate food, poor sanitary conditions, injuries, illness and numerous deaths. Its account also notes that in 1889 the warden of Kabato Prison held district administrative responsibilities alongside his prison office.^[9]

That combination of offices makes the competing demands concrete. A person charged with confining prisoners could also have responsibilities for the region their labor was being used to develop. The municipal account does not permit us to assign every decision to one motive. It does show why an institution's work cannot be assessed solely by repeating its stated aim of correction.

The evidence of illness and death changes the comparison with the early promise of preparation for life outside. Producing something valuable beyond a prison does not establish that its labor prepared those performing it for release. A system could organize and account for a sentence while exposing its prisoner to harms that the number of years alone did not describe.

The former Nara Prison, completed in 1908, gives a different material form to these questions. The national cultural-property record identifies its radial arrangement: prison wings connected to a central guard area, with other buildings for work and detention. The separate hospital building was part of the complex too.^[10]

The arrangement's most immediately visible effect was on supervision. The institution's architectural archive explains how the central position gave views along the corridors of the radiating wings. That is a specific line of sight, not proof that one guard could see every person inside every cell.^[11]

The wings did not all impose the same social arrangement. In the original scheme, some combined individual confinement with work inside the cell. Another separated prisoners at night while bringing them together for daytime labor. A further wing accommodated communal custody.^[11]

These alternatives divided the prisoner's day in different ways. Separation could govern sleeping without governing all work. A person alone in a cell could still be required to produce something. The pairing of room and activity matters more than a simple contrast between an old communal jail and a new solitary one.

It also reveals a problem for anyone trying to recover experience from the plan. The plan identifies intended arrangements and the possibilities of supervision. It cannot establish how a particular prisoner experienced isolation, how officers used their authority on a given day, or whether work followed the intended schedule. Those questions require other records.

Nara also contained a distinct remand establishment.^[11] Waiting for a case to be decided had not vanished because confinement was now elaborately specified as punishment. The complex accommodated custody serving different stages of the legal process.

This is where the comparison with Edo becomes most useful. The later prison did not merely replace waiting with sentencing. It incorporated both, while developing different spaces for them. A visitor looking at the brick exterior would need the institutional distinctions before knowing why a person behind any particular door was being held.

The building thus made administrative classifications tangible. Different arrangements of people, work and observation could be maintained within one establishment. Its apparent unity as “a prison” concealed several answers to what custody was for.

A sentence measured in years might appear to remove personal judgment from release. The 1880 code did not go that far. Article 53 allowed conditional release for eligible prisoners who had completed three-quarters of a fixed term, observed prison rules and shown signs of reform. Eligibility was not an automatic right to leave on the day that fraction was reached.^[12]

The provision joined arithmetic to an assessment. An official could establish how much time had passed by consulting a record. Deciding whether conduct demonstrated reform required a different kind of judgment. The promise of early departure therefore gave behavior inside the institution a possible consequence beyond the next day's discipline.

Other clauses preserved restraints after conditional release. A person conditionally released from the island-labor punishment still had to reside on the island. Special supervision continued for the unexpired term.^[12] The prison door could open without every boundary imposed on the person disappearing.

Compare this with Ishikawajima's earlier reliance on resources and an acceptable guarantor. Both arrangements required more than an unlocked gate, but the questions were not identical. The earlier institution asked whether an arrangement for departure could be accepted. The code combined a specified portion of the sentence with an official evaluation of conduct, then maintained conditions afterward.

This comparison does not establish a direct line of borrowing. It identifies a recurring difficulty: leaving a place of custody and recovering unrestricted control of one's life are separate events. Institutions can connect them in different ways—or keep them apart.

Nor did the new prison's concern with correction exclude other outcomes. Article 12 of the same code prescribed execution by hanging inside the prison under official supervision.^[5] A prison could be organized for labor, observation and possible release while also containing a procedure that ended a prisoner's life. No single reassuring statement of purpose describes every person held within it.

Japan's prisons were places for keeping people available, imposing specified punishments, directing compulsory work and judging conditions for departure. The balance changed across institutions and periods; the older functions did not simply vanish as new ones appeared. To understand the locked gate, ask three questions: what decision brought the person there, what could the authorities do during the stay, and what decision would be needed to end it?