Why Does Japan So Often Change by Restoring the Past?

The Charter Oath of 1868 promised to seek knowledge throughout the world. It also called for abandoning harmful customs inherited from the past. These were striking commitments for a government established in the name of restoring imperial rule. A restored emperor was being made the authority for departing from established practice.^[1]

There was no necessary contradiction. The leaders of the new government could distinguish the emperor's rightful authority from the arrangements through which Japan had recently been governed. Recovering the first did not oblige them to preserve the second. Indeed, the claim to restoration gave them a reason to change it.

This was not the only occasion on which an appeal backward supported a movement forward. In the fourteenth century, renewed imperial government produced new machinery for deciding land disputes. In eighteenth-century Kyoto, reconstructing an older palace required negotiation about present needs and costs. Under Meiji, an ancient imperial past could justify institutions that had never existed in that past.

The resemblance does not establish a timeless Japanese preference, or a frequency that can be ranked against other countries. It identifies a recurrent political possibility. An older authority could help people challenge a more recent arrangement. What made the strategy useful—and what happened when different people wanted different things restored?

In 1333, the Kamakura warrior government fell, and Emperor Go-Daigo attempted to establish a new governing order. The short-lived enterprise is conventionally called the Kenmu Restoration. For an institution holding land, the change raised an immediate question: would its existing claims still be recognized?^[2]

A document preserved at Matsunoo Taisha, a shrine in Kyoto, brings that question down to the scale of one estate. An imperial rescript issued in 1333 confirmed possession concerning the Yamamoto estate in Settsu Province. It was addressed to a former court councillor of the Ichijō family, not directly to the shrine. Matsunoo's research institute interprets the document as part of the shrine's effort to secure its interests under the new regime.^[3] The address and the interpretation belong together: an estate could connect several institutions and families, rather than have one straightforward owner.

The need for that document reveals the practical difficulty of restoration. A victorious ruler could announce a change in government. That announcement did not identify every valid claim to an estate, determine which earlier grants should stand, or persuade an occupant to leave. Land supported livelihoods, religious observance, and military service. Changing the authority that confirmed possession exposed those interests to reconsideration.

Nor was an estate necessarily one property belonging wholly to one owner in the modern sense. The question could concern particular rights and revenues within a landscape already burdened with obligations. A general promise to recover rightful order could therefore generate several incompatible requests. One claimant's restoration might require another claimant to surrender an income.

Historian Mori Shigeaki's study of the Kenmu government's legal arrangements shows how the pressure entered its institutions. Unresolved disputes inherited from the Kamakura period burdened the new regime. The emperor and the court's existing adjudicative machinery could not manage the business adequately. A new body, the Zassō Ketsudansho, was established in 1333 to handle disputes over land and associated claims.^[2]

Creating this tribunal changed how restored imperial authority was exercised. Recovering imperial government did not mean that all the necessary governing arrangements were waiting intact in an old court office. New procedures and divisions of work had to be developed. An imperial rescript could provide an authoritative decision. The government still needed procedures for adjudicating other claims and for giving its decisions effect outside the capital.

This was the point at which restoration encountered competing authority. A document mattered because people recognized its issuer; its effectiveness also depended on those able to enforce it. Mori traces adjustments to the relationship between imperial commands and adjudication. His account makes the regime's legal activity more substantial than a series of royal pronouncements, without turning its innovations into proof of lasting success.

The wider political settlement failed. That outcome should not erase the institutional work undertaken within it. Kenmu exposes a recurring difficulty in restoration projects: a claim to ultimate authority can open many questions that it cannot, by itself, answer. The return of the ruler creates business for a government; it does not complete that business.

More than four centuries later, restoration posed a different kind of problem in Kyoto. The great fire of 1788 destroyed the imperial palace. Rebuilding required money, materials, and decisions about design. It also required an answer to a historical question: should the court replace the palace it had just lost, or recover arrangements associated with an earlier imperial world?^[4]

The court was not independently in command of the resources needed for the project. The Tokugawa shogunate, the warrior government based in Edo, had to participate in making reconstruction possible. An appeal to ancient imperial practice thus entered a negotiation between institutions that occupied different positions in the governing order.

The scholar Uramatsu Kōzen had studied the earlier palace through written sources. In Fukuda Toshiaki's account, that research helped the court explain why the contracted spaces it had been using were inadequate for certain ceremonies. The older arrangement supplied an argument about what a functioning palace needed. Matsudaira Sadanobu, directing the shogunate's response, accepted a selective revival adapted to contemporary circumstances and expense.^[5]

The resulting reconstruction, completed in 1790, was neither a replacement drawn solely from recent memory nor an attempt to rebuild the entire early capital. The revived Seiryōden, a hall associated with imperial domestic and ceremonial life, referred to arrangements of the late twelfth century. Furnishings and designated spaces formed part of the recovery.^[6]

That specificity matters. “The ancient palace” sounds like one recoverable object. In practice, there were different buildings, dates, descriptions, and uses from which to choose. Historical scholarship helped make a particular past available. Ceremonial requirements helped determine which parts were worth recovering. The shogunate's participation set practical limits on what could be built.

A claim to authenticity could consequently contain several kinds of judgment. Which written description should carry authority? Which arrangement would allow a rite to proceed? How much space could the project obtain? The answer emerged through decisions made by people who belonged to the eighteenth century, however seriously they took their medieval references.

Kōzen's great compilation on the imperial complex was formally presented in 1797, after the rebuilding. Fukuda distinguishes that presentation from the work's earlier completion and use. The chronology prevents a misleading picture of builders working before the relevant research existed. It also shows that research could enter political and architectural decisions before its later formal recognition.^[5]

The practical success of this revival made it an inheritance in its own right. When fire again destroyed the palace in 1854, the rebuilding completed the following year largely followed the 1790 precedent.^[4] A selective eighteenth-century recovery of older forms had become the design that nineteenth-century builders received as their model.

Here, restoration reinforced an existing relationship between court and shogunate. It did not necessarily prepare a revolution against that relationship. The same broad practice—appealing to an older order—could serve accommodation as well as confrontation. Its political meaning depended on who was making the claim and what they wanted changed.

The palace case also helps distinguish restoration from simple preservation. The purpose was to keep certain imperial activities possible. That could require new buildings, recreated furnishings, or a temporary arrangement rather than the unchanged survival of one room.

In February 1867, the Seiryōden was occupied by the funeral arrangements for Emperor Kōmei. His successor's accession rite therefore used the Kogosho, another palace building, as a substitute. Spaces and furnishings were assigned the ceremonial relationships needed for the occasion.^[7] The court did not have to suspend succession until one preferred hall became available.

This was a different undertaking from abolishing a government. It involved people who knew the requirements of a rite well enough to accommodate unusual circumstances. The inherited practice supplied constraints, but those constraints did not prescribe that every event must happen in the same unchanged physical setting.

Seen from a distance, a ceremony can appear to certify that nothing has altered. Behind that appearance may lie substantial work: interpreting instructions, moving objects, assigning rooms, and deciding which relationships must remain recognizable. The activity's continuity depends on successful adjustment.

That does not make tradition an empty disguise. The requirements were consequential precisely because the participants treated them as requirements. Nor does it mean every change was acceptable. A substitute setting had to sustain the associations that made the act intelligible as the proper ceremony.

The distinction becomes useful when returning to political restoration. Both a reconstructed hall and a restored government invoke an inheritance, but their claims operate at different levels. A palace adaptation need not redistribute governing power. A political restoration may do exactly that while continuing to use the palace's ceremonial forms. Similar language can conceal very different stakes.

The Meiji government's founding promises combined a powerful claim to imperial authority with considerable uncertainty about institutional design. The Charter Oath did not provide a finished constitution or a detailed administrative system. It announced purposes broad enough to support further decisions, including learning from outside Japan.^[1]

Those decisions could produce unfamiliar combinations. The governing arrangements announced in 1868 used the old name Daijōkan, or Council of State, while incorporating new deliberative and administrative structures. Repeated reorganizations followed. In 1885, a cabinet system replaced the Daijōkan arrangement.^[8] The historic name had not committed the government to reproducing the ancient office in all its functions.

A public notice issued by Chōshū forces acting for the imperial side in 1868 shows the change at a more immediate level. Following the fighting around Kyoto, the notice declared former Tokugawa territory subject to confiscation. Yet it also instructed people to continue existing tax and other arrangements temporarily, pending further decisions. It addressed the removal of intolerable practices and assistance to poor people while awaiting the court's directions.^[9]

A change in the authority claiming the land did not mean that officials could immediately replace every procedure through which the land was administered. Revenue still had to be collected. Local matters still required decisions. The notice combined a sharp statement about sovereignty with provisional continuity in everyday government.

This makes it difficult to divide the transition into a purely conservative restoration followed by an unrelated period of modernization. The new authority was already making selective choices about what to retain and what to reconsider. It could preserve an arrangement because it remained useful, while reserving the right to alter it in the emperor's name.

The choices also operated on different timetables. A proclamation could change the political identity of a territory before new procedures reached every household within it. The old rule surviving for another season was not necessarily evidence that the change in sovereignty meant nothing. Conversely, the new proclamation did not prove that every local official had acquired new instructions or the means to carry them out.

The contrast with Kenmu is instructive. Both projects made imperial authority central to a new settlement. Both then faced claims and administrative obligations that required more than a declaration. The nineteenth-century government eventually established a much more durable order, but that outcome cannot be read backward into the word “restoration.” The word named an authority and a direction; it did not supply the institutional result.

Restoration could become especially disruptive when its advocates identified a long-established practice as a departure from an older, purer condition. In that case, much of what people had inherited might become an obstacle to the past the reformers wanted.

The Meiji separation of kami worship from Buddhism illustrates the difficulty. For centuries, shrines, Buddhist institutions, clergy, images, and divine identities had been connected in numerous ways. The separation orders of 1868 acted against those accumulated relationships. Buddhist personnel serving shrines had to change their position or leave; names and objects understood as Buddhist were challenged or removed from shrine settings.^[10]

Sakamoto Koremaru's account places the measures within arguments that treated such combinations as later distortions of an original imperial religious order. The original condition invoked by these arguments was itself an object of interpretation. It was not a recently documented administrative arrangement that officials could simply reinstate.

The consequences were tangible. Separating institutions meant deciding who could continue to serve, which objects could remain, and what an establishment should be called. People and property could be reassigned by a policy presented as the recovery of rightful distinctions.

Anti-Buddhist destruction developed alongside the separation measures, but the two should not be treated as identical instructions uniformly implemented across Japan. Local officials and other participants pursued actions that went beyond separation, sometimes exploiting opportunities involving property and finance. The central authorities' stated limits did not prevent all such actions.^[11]

This is restoration's harsher possibility. Selecting an earlier past can delegitimize centuries of subsequent history. To someone attached to a combined shrine-temple institution, what the reformer calls recovery may be experienced as loss. The issue cannot be settled by asking which side favors tradition. They are defending different inheritances, and one has acquired the authority to reorganize the other.

The Kyoto palace had shown how selection could make an old practice workable. Religious separation shows how selection could make established practices unacceptable. Looking backward did not prescribe a gentle pace or a conservative outcome.

The emperor's provincial journeys offer another way to see the relationship between revived forms and new government. Six major tours between 1872 and 1885 brought the Meiji emperor through parts of Japan after the replacement of the domains by prefectures.^[12]

Historian Kasahara Hidehiko interprets these journeys partly through the revival of older royal progress traditions and their usefulness to the new system of regional administration. His emphasis differs from accounts that concentrate on the public representation of monarchy and the making of national identification. The tours supported more than one historical question, and their occurrence cannot by itself reveal what every spectator believed.

For the present argument, their institutional setting is decisive. An imperial journey recalling an older form was taking place amid a new distribution of administrative authority. Its significance did not depend on recreating the territory or political relationships of the period being recalled. The inherited ruler was moving through a reorganized country.

Placed beside the palace, this produces a useful contrast. An eighteenth-century shogunate could help rebuild the settings of imperial tradition while continuing to govern. In the nineteenth century, imperial tradition could support arrangements that displaced that shogunate and its regional political order. The identity of the legitimating institution persisted across a profound change in what was being legitimated.

This is why restoration cannot be treated as an explanation by itself. It directs attention toward the past being invoked. The historian must then follow the offices, resources, obligations, and participants that give the invocation a practical effect.

Across these cases, an appeal to the past did several kinds of work. It supplied a standard by which a recent arrangement could be criticized. It connected a proposed change with an authority already entitled to respect. It also gave participants something sufficiently recognizable around which to organize action before every detail had been decided.

None of that removed conflict. Go-Daigo's authority generated demands for recognition that needed adjudication. The Kyoto court's historical claims required a funded building project. Meiji's restored sovereignty left tax procedures temporarily in place while other inherited arrangements were dismantled. The practical question was always which claims would become enforceable, and through whose work.

The differences are as important as the recurrence. Kenmu's legal innovations did not secure the regime's survival. Palace reconstruction helped sustain an existing accommodation. Meiji restoration authorized a continuing sequence of institutional departures. They cannot be reduced to one national habit of changing without admitting it.

Nor must every restoration claim be dismissed as cynical. Historical forms could matter deeply to those recovering them, while also serving current needs. Sincerity does not eliminate selection. A person convinced that an older order is right still has to decide what belongs to it and how it can operate under changed conditions.

Japan's history therefore suggests a more useful question than whether restoration was really conservative or really revolutionary: what did the chosen past authorize people to change? The answer may be a tribunal, the dimensions of a hall, the staffing of a shrine, or the structure of a government. An older authority could open the way. The new arrangement still had to be made.