Isawa and Takkoku: What Followed the Fighting

An excavated document from Isawa Castle reports soldiers absent through illness. The paper survived because it acquired another use: old writing was reused for purposes such as covering lacquer containers, and absorbed lacquer helped preserve it underground. An administrative record outlasted the office that needed it.^[1]

A very different account of military life survives at Takkoku no Iwaya, the cliff sanctuary near Hiraizumi. Its founding tradition credits the commander Sakanoue no Tamuramaro with defeating enemies and establishing worship of Bishamonten, the Buddhist protector associated with his victory.^[9] Here the remembered general belongs to a continuing religious relationship. His work did not end, in the temple's account, with the departure of an army.

Isawa, in present-day Ōshū, and Takkoku, in Hiraizumi, are separate places. They were never two sections of one castle. Read together, they disclose different things that followed frontier fighting: arrangements for administering people, and accounts through which later generations gave the fighting a sacred meaning. The first required a record of who could report for duty. The second could make one commander's name useful long after his lifetime.

The imperial court's northern campaigns entered a region that already had inhabitants and political organization. Ōshū's historical account records an attack on Isawa in 776 and a court defeat by forces led by Aterui in 789. Aterui and another leader, More, surrendered in 802. That was also the year Tamuramaro built Isawa Castle; the establishment of Isawa district followed in 804.^[1]

These events describe a contested extension of government, rather than a capital discovering land without an owner or population. The people resisting the court had names before they acquired a place in the history of its castle. Nor should the creation of an administrative district be confused with an account of how every resident experienced incorporation.

The word Emishi needs that political setting. Aomori Prefecture's historical synthesis explains that the term's geographical reach and meaning changed over time. A major element was a population's position outside direct state rule. As the reach of government changed, so could the people to whom the designation applied.^[2]

It is therefore a poor substitute for an unchanged ethnic label. Archaeologist Kudō Masaki, reviewing the debate in 1994, challenged the choice between simply identifying the ancient Emishi as Ainu or treating them as already Japanese in the later sense. He emphasized overlapping cultural histories and different subsequent developments.^[3] The people in an early ninth-century court account cannot be assigned wholesale to a modern identity by translating one word.

At Isawa, the new government's presence took a substantial physical form. Its enclosure was approximately 670 meters square, bounded by a wall of compacted earth. Within it stood facilities for administration, ceremony, and provisioning. By 808, the Chinjufu, the headquarters supervising northern military affairs, had moved here from Taga Castle farther south.^[4]

This was a government compound, not a later samurai castle with a commanding keep. Its problem was how to maintain an establishment in a region into which court power had recently advanced. Winning a campaign could change who claimed authority. Keeping a headquarters supplied and staffed required decisions that would recur after the campaign's exceptional effort was over.

Another Isawa document records the ages and home localities of soldiers assigned to the castle. The Iwate Prefectural Museum's catalogue identifies two of those localities with places in what is now Miyagi's Shibata district. The museum's catalogued object is a reconstructed replica of document number 43, not the original sheet.^[5]

The geographical information expands the castle beyond its earth wall. Men entered the institution from elsewhere; their place of service and their place of origin were recorded separately. The headquarters depended on relationships between settlements and an assigned military establishment. It could not be staffed simply by declaring that a new square on the landscape was now a castle.

Age and origin are also different kinds of knowledge from the statement that an army had won. They make people identifiable within an administrative system. The illness report adds the possibility of a discrepancy between assignment and attendance. A list could identify the men expected to serve; another document could explain why some were unavailable. Those surviving categories reveal work that a commander's biography rarely includes.

The documents do not supply a complete account of military service. They nevertheless prevent the garrison from becoming an anonymous, permanently available mass. The institution had to register differences among people and changes in their capacity to work. Its continuing operation depended on information smaller than a proclamation and more frequent than a battle.

Materials arrived through other connections. Tiles excavated at Isawa are associated with the Seyago kilns in the Esashi area. The supply of roofing thus linked specialist production outside the enclosure to construction inside it.^[1] A finished roof concentrated the results of making, transporting, and fitting material; the castle's boundary did not contain all the work embodied in its buildings.

Archaeology reveals that Isawa's governing arrangements changed within the enclosure. One excavated building stood on the main north–south axis, about thirty meters south of the government compound's southern gate. It was added after the initial construction. Identified by investigators as the denmon, it occupied a position where an approach might otherwise be expected to continue unobstructed.^[6]

In his excavation study, Itō Hiroyuki traces changes from a gate-like arrangement toward more enclosed building forms. The structure used posts set into the ground and was not roofed with tiles. He interprets it as a place to which certain functions were transferred, particularly ceremonies involving Emishi brought within the court's administrative relationships. Meanwhile, the principal government compound expanded.^[6]

That combination is the revealing part. A new building across an approach might look like an obstruction if the only question were how quickly people could get inside. It becomes intelligible when the approach itself acquires a task. Activities can be received and organized in front of the principal enclosure rather than carried straight into it. Itō's interpretation connects architectural change with a differentiation of institutional work, not simply a headquarters shrinking into decline.

Ceremonial reception had material requirements. Aomori's historical account describes northern officials inviting Emishi leaders to feasts, conferring ranks or titles, and distributing goods such as iron products and textiles. Such gatherings commonly took place at fortified administrative sites.^[2] Hospitality could be a means of establishing a relationship with the court, as well as an occasion on which people received something useful.

The two aspects were inseparable in the arrangement. Food and goods made a gathering materially consequential. Rank and formal reception placed the participants within an order defined by the institution hosting them. Calling the event a feast does not make it politically neutral; calling it administration should not make the provisions disappear.

The proposed use of Isawa's denmon belongs within that wider practice. It remains an archaeological interpretation, rather than the recovered transcript of a ceremony in one room. Its value is that it explains a particular alteration to the site through a particular kind of recurring work. The castle was being adapted to manage relationships, not merely retained as the footprint of a victory.

Isawa's later development also requires a scale larger than its own plan. In a study of northeastern fortifications, archaeologist Yagi Mitsunori identifies a consolidation into six principal establishments by the middle of the ninth century. He relates this to changes in the state's military and settlement policies and to the wider responsibilities assigned to the remaining centers.^[7]

Under that interpretation, fewer headquarters did not necessarily mean less work for Isawa. A surviving institution could acquire a larger area of responsibility when others ceased to perform comparable functions. The growth of one compound and the slowing of territorial expansion are therefore compatible. Its buildings need not be read as a simple graph of an army moving ever farther north.

The modern historical park makes one later phase especially legible. Opened in 2019, its presentation of the outer South Gate area is based on the late ninth to early tenth centuries. Recreated earth walls and indications of posts establish the positions of the gate and associated structures. They do not reproduce the moment of construction in 802.^[8]

That choice gives the visitor an important advantage: a coherent arrangement can be understood on the ground. It also selects one period from a longer sequence. The square enclosure, the headquarters' transfer, the added denmon, and the park's displayed phase belong to different moments. Isawa remained in use into the later tenth century.^[4]

There was time for an institution founded during conquest to acquire a history of its own. By the period represented in the park, the founding commander was long gone. The practical relationships exposed by soldier lists, regional supply, and ceremonial space had required successors to keep doing the work. At Takkoku, his name followed a different path: instead of receding behind an institution's routine, it became part of the institution's continuing reason for worship.

Takkoku's founding story makes the rock shelter a stronghold occupied by Tamuramaro's enemies. The temple's published account names Akuroō and his companions, describes their violence, and places their defeat in 801. It then explains the sanctuary as the commander's offering of thanks for Bishamonten's protection. The stage-like form of the hall is linked, in the same tradition, with Kyoto's Kiyomizu-dera.^[9]

The story gives each element a role. The cliff shelters the opponents; victory demonstrates protection; worship acknowledges the source of that protection. A place associated with danger becomes a place from which protection can be sought. This is a religious explanation of the sanctuary, not a contemporary campaign report.

The named antagonist should remain within that distinction. Aterui belongs to the historical account of resistance and surrender at Isawa. Akuroō belongs here to the temple's transmitted founding narrative. The materials used for this essay do not establish that they were the same person. Substituting one name for the other would turn the moral roles of the legend into a biography that the evidence does not provide.

The earliest textual reference identified by the national heritage record is much later than the supposed foundation. The medieval chronicle Azuma Kagami places a visit by Minamoto no Yoritomo at Takkoku in 1189, after his destruction of the northern Fujiwara's rule. Its account of the visit also relates the sanctuary to Tamuramaro's earlier campaign.^[10]

Two military pasts therefore meet in the account: the expedition associated with the sanctuary's origin, and the later campaign during which Yoritomo came to Hiraizumi. The chronicle makes Takkoku a place where one conquest could be understood in relation to another. It does not give direct access to an encounter at the cliff in 801.

This is already a substantial history for the site. By the medieval period, the sanctuary could place a visitor within a remembered northern landscape whose significance extended beyond the buildings of his own time. The general's name was carrying an account of why the place existed. Archaeology helps establish what stood there when that account was being transmitted.

Excavation in front of the Bishamonten hall in 1988 found an east–west alignment of river stones, interpreted as a pond bank. Numerous unglazed earthenware dishes, known as kawarake, resembled late twelfth-century examples from central Hiraizumi. The national designation account uses these finds to identify a medieval arrangement in which water lay before the hall.^[10]

This supplies something different from the founding story: evidence for the sanctuary's developed setting in the period of Fujiwara Hiraizumi. It connects Takkoku with the material world of nearby religious and administrative centers without requiring every part of its tradition to describe a verified ninth-century event. Pottery and stonework establish a place being used; the transmitted narrative explains how that place's origin was understood.

Water also changes the cliff's architectural relationship. The sanctuary was not simply a building inserted wherever rock allowed. The arrangement brought a hall, a forecourt, and a pond into relation with the natural face behind them. The designation account connects this layout with the broader religious landscapes of medieval Hiraizumi.^[10]

The hall now standing there belongs to another rebuilding. The temple dates it to 1961, following the destruction of its early modern predecessor by fire in 1946.^[9] The present structure restores a worship space against the cliff; it is not the timber erected by the commander of the founding tradition.

That rebuilt relationship remains consequential. At Isawa, interpretation marks the positions of vanished institutional structures. At Takkoku, a new hall continues to organize worship around the rock shelter. Both places contain modern work, but the work has different purposes. One makes an archaeological arrangement intelligible. The other provides an active setting for practices whose authority includes a remembered past.

The temple's explanation of Tamura faith, published in 2005, describes a further transformation: Tamuramaro himself is understood as a manifestation of Bishamonten. The commander is not only someone who received a deity's help. He becomes a figure through whom that protective relationship can be expressed.^[11]

The same statement associates the space beneath the hall with the returning spirits of ancestors and with accounts of refuge and renewal. These are the temple's descriptions of sacred space, not verified biographies of everyone said to have sheltered there. They nevertheless show why the cliff cannot be reduced to a remembered military position. Its religious meaning extends from danger and victory to the dead and the hopes of people still living.

A paper talisman makes the extension particularly concrete. The temple describes its goō hōin as produced through the New Year observances, including twenty-one offerings to Bishamonten and other rites. Recipients are instructed to place it high in the home or at an entrance, where it is believed to ward off harmful forces and invite good fortune.^[12]

The protective claim has moved from a commander's campaign into a household. The paper carries a connection to rites performed at the sanctuary; its destination is a threshold elsewhere. Remembering Tamuramaro does not require the recipient to be a soldier, recreate his expedition, or live beside the cliff. The temple makes the relationship available through another object and another setting.

Its published 2026 calendar includes recurring offerings to Bishamonten and the sequence of New Year observances culminating in the distribution of the talismans.^[13] The old account is thus accompanied by scheduled work in the present. A foundation story explains why protection belongs here; ritual provides occasions on which that protection is sought again.

This is the connection that makes Isawa and Takkoku worth reading together. Frontier fighting produced continuing tasks, but those tasks did not all remain military. Isawa's records organized obligations and service within an expanding—and subsequently reorganized—government. Takkoku's traditions and rites made an account of victory available for religious purposes beyond the campaign itself.

One place preserves a soldier's absence because someone had to account for it. The other keeps a general present because people continue to invoke the protection associated with him. Between them lies more than the distance from a castle field to a cliff hall: there are two different ways in which an army's history could enter ordinary life after the fighting.