Kanbara: Making a Village Again

The stair below Kanbara Kannon-dō continues farther than the ground allows it to show. Fifteen steps remain above the material deposited by Mount Asama’s eruption of 1783. Excavations begun around 1980 traced another thirty-five below them. At the bottom, archaeologists found the remains of two women. The hall had stood high enough to survive; much of the village around it had not.^[1]

The stair gives the catastrophe a vertical scale. It also poses a question about everything that happened afterward. Once the deposit had become the surface on which people would live, what would make that surface a village?

Kanbara, in present-day Tsumagoi in Gunma Prefecture, was rebuilt above the buried settlement. Local history identifies ninety-three survivors, a remnant of a community that had lost most of its population.^[2] Their recovery involved more than replacing roofs. Houses had lost their inhabitants as well as their walls. Old claims to land confronted a changed landscape. Water had to be made available, and the dead remained beneath the place where the living were beginning again.

Mount Asama’s 1783 eruption unfolded over months. Activity began in the spring and intensified into the destructive events of early August. On 5 August, an avalanche of rock and earth associated with the eruption swept down the northern side and overwhelmed Kanbara. Material entering the Agatsuma River continued downstream as a destructive mudflow, carrying the disaster into the Tone River system. The broad lava field called Onioshidashi belongs to the same eruption, but it is not interchangeable with the deposit that buried the village.^[3]

These different movements produced different aftermaths. Ash could cover cultivated ground; a deep avalanche deposit could place a former street well below the new surface; sediment in a river could carry trouble far beyond the mountain. There was no single task called clearing away the eruption. The work depended on where the material had gone and what it had left usable.

The people sent to investigate could not even assume that survivors had watched the whole event. Negishi Kurōzaemon, an official of the Tokugawa shogunate, the military government based in Edo, investigated the affected region later in 1783. In a report quoted by disaster historian Inoue Kimio, he recorded conflicting accounts of where the destructive material had emerged. People had been occupied with escaping, not with observing its source. Negishi subsequently directed government works in severely affected areas.^[4]

His report separates two requirements that might otherwise be confused. Explaining the eruption required comparing imperfect observations. Helping a damaged settlement required decisions before every physical question had been settled. A household could need shelter even when investigators disagreed about the source of the debris.

Kannon-dō supplied an enduring point in that changed geography. Dedicated to Kannon, the Buddhist figure of compassion, it had been a religious destination before becoming a place associated with refuge. The surviving height around the hall and the buried ground below it established a difference that reconstruction could not simply remove. Returning to the locality would mean living with a new relationship between the old settlement and the surface above it.

Negishi’s collection of anecdotes, Mimibukuro, preserves an account of the social rebuilding. In the modern rendering published by the Disaster Prevention Research Institute, three local organizers urged the survivors to regard one another as kin despite differences of family standing. Among the organizers was Kuroiwa Chōzaemon, a headman from neighboring Ōzasa. The account describes remarriages between bereaved spouses and the placing of orphaned children with older people who had lost children. Food and drink marked the new arrangements.^[5]

The problem addressed was different from supplying a building. A roof could shelter people who happened to have survived together. A household established continuing responsibilities among them: who would live together, who would care for a child, and who would belong to the next generation of that home. Rebuilding required both an address and people able to make a life at it.

Negishi foregrounded family standing because it affected whom people would ordinarily treat as social equals. In his telling, the organizers addressed that obstacle before arranging new relationships. The appeal to kinship was meant to make cooperation possible across distinctions that had mattered before the disaster. The language of family then acquired concrete forms through marriage and adoption.

This is an approving official’s account, not a collection of the survivors’ own explanations. It does not establish that every match was freely chosen or every new household harmonious. Its evidence is narrower: household composition itself became an object of organized reconstruction. The people directing relief did not expect buildings alone to restore a population whose family relationships had been broken.

That distinction helps explain why Kannon-dō cannot be understood only as the end of an escape. Survival left another set of decisions. A person who had reached safety might still lack the relatives with whom the former home had been shared. The work of rebuilding therefore concerned the distribution of care and membership as well as materials. A new household could continue life in Kanbara without recreating the family that had occupied the lost house.

The settlement’s reconstruction also left a spatial record. A 2023 architectural-history graduation study at Waseda University compared published village documents with a reconstruction map. It identifies an initial relief phase followed by more organized rebuilding. The new settlement used elongated plots along a road, while fields outside that regular arrangement complicated any simple picture of equal rectangular allotments. The study also finds that former household names were carried into the reorganized community and that some earlier land relationships persisted.^[6]

These findings qualify the language of beginning again. A declaration that survivors should act as relatives could address social conduct without abolishing every distinction in property or household identity. Likewise, a more regular street layout did not make all land beyond the street identical. A village was being reassembled from what remained, not created on a surface without prior claims.

The persistence of household names is particularly revealing. A name could maintain an identifiable unit while the people belonging to it changed. That offered a way to connect a reorganized family with the village’s remembered and recorded past. Continuity did not have to mean that the same relatives returned to the same rooms.

The two forms of reconstruction consequently worked at different scales. Marriage and adoption addressed the people who would share a home. Plots, roads, and household identities placed those homes within a settlement. One arrangement could not do all the work of the other. The road needed neighboring properties; the household needed members; both needed some relationship to land from which a living could be made.

The rebuilding described in these records combined planned physical change with retained elements of an older social landscape. A more orderly line of houses could coexist with a less regular pattern of holdings beyond it. Kanbara’s new ground covered the earlier settlement, but the surviving people and records still connected the two.

The deposit also changed the practical problem of obtaining water. Geographers Suzuki Hidekazu, Sekigawa Yūki, and Miyazaki Mitsuo describe difficulties reaching stream water and groundwater from the rebuilt settlement. The supply eventually known as Kanbara’s water channel originated in a spring near the edge of the Onioshidashi lava. Yet it did not immediately serve the village in the form later remembered.^[7]

The spring initially discharged hot water. Kuroiwa Chōzaemon arranged to take it to Ōzasa, where a bathhouse offered free bathing to residents and travelers. Construction employed people who had lost homes or work in the disaster. After the water cooled and the bathhouse ceased operating, the supply was redirected to Kanbara for domestic use. It continued serving the settlement until a piped water system was installed in 1958.^[7]

This history does not describe one recovery project reaching an inevitable final form. The first undertaking used heat; the later one needed a dependable supply for different purposes. Cooling made the original use less viable while leaving water available for another. A change in the spring’s condition altered what its users could ask it to do.

It also separates employment from the finished facility. Construction could provide work to displaced people before the bathhouse admitted its first user. The completed baths then served another constituency, including travelers passing through Ōzasa. A project associated with recovery could therefore have benefits at different moments and for people who did not all live in Kanbara.

For the rebuilt village, the significant achievement was delivery. Water at a spring and water available to households are different resources. A route had to connect the source to the place where it would be used. That requirement makes the surrounding volcanic landscape part of the village’s history long after the destructive flow had stopped moving. The mountain had changed the ground; rebuilding demanded new connections across it.

The channel’s succession of uses is also a reminder that recovery has no single completion date. A house might be occupied while its water supply remained difficult. A useful facility elsewhere could later become part of the village’s solution. The relationships among source, route, and users continued to change after the immediate emergency had passed.

Later archaeology recovered evidence of what reconstruction had left beneath the surface. The Tone River System Sabo Office’s account follows investigations at Kannon-dō, the buried house site at Tōkanokubo, and Enmeiji temple. The finds included building remains and objects of domestic and religious use. The research opened separate windows into the settlement rather than uncovering a complete town for permanent inspection.^[8]

The difference between those windows matters. A stair establishes a former ground level and route of approach. A house site can locate everyday life within the buried settlement. Temple remains identify another place of religious activity, separate from the hall that survived. Together they prevent Kannon-dō’s survival from standing in for the experience of every building around it.

Inoue, who visited the Enmeiji excavation, reports that much of the recovered timber and household material there was not charred. His observation is incompatible with imagining every part of Kanbara consumed in a uniform sheet of burning lava.^[1] It does not by itself settle every question about the avalanche’s origin or temperature. It does restore a material difference concealed by the general word “eruption.”

The human remains near the stair’s lower end give the burial a different gravity. They locate deaths beside a route associated with safety. They do not establish the women’s relationship to one another or provide a transcript of their final actions. The strength of the evidence lies in its position, without requiring an invented story to make the loss real.

The adjacent Tsumagoi Folk Museum and the hall’s setting now offer complementary ways of approaching this past. The museum interprets excavated material; the upper precinct preserves a place around which people continued to gather.^[9] Most of the old village remains below ground. Its presence is not cancelled by the fact that later residents built above it.

That vertical relationship gives Kanbara a distinctive form of continuity. The new settlement did not have to expose the old one in order to remain connected with it. Archaeological investigation could make particular connections visible much later, while mourning had already supplied another means of acknowledging who remained beneath the inhabited landscape.

Around the turn of the twenty-first century, researchers Mieda Yasuyo and Hayakawa Yukio documented Kanbara’s mawari nenbutsu, a rotating Buddhist memorial observance. A scroll naming the victims and ritual equipment passed among participating households. Remembrance acquired a circuit through the living settlement rather than remaining confined to one monument. Their account also records a practical adjustment in 1999: the meal expected of a host was simplified to reduce the burden of providing it.^[10]

The rotation gave the names a destination. A household did not merely know that the catastrophe was commemorated somewhere; it could become the place responsible for the next observance. Passing the scroll onward transferred a task as well as an object.

Food made that responsibility tangible. A ceremony requiring a household to receive others depends on time, preparation, and resources. Reducing the expected meal could change the work of hosting without abandoning the reason for gathering. The adjustment belonged to the continuation of the practice, not simply to its decline.

This was another kind of reconstruction. The participants were not arranging marriages after the eruption or digging out a buried home. They were adjusting the obligations through which a much later community remembered those losses. Their decisions show that continuity could require negotiation over what living households could reasonably supply.

The circuit also joined two geographies. The victims belonged to the village buried in 1783; the hosts occupied the settlement that had developed afterward. Carrying their names between homes brought the earlier population into a recurring relationship with the later one. It did not require every present household to be identical in descent, belief, or circumstances.

A separate stone monument to the Asamayama Funka Daiwasan, a Buddhist hymn about the eruption, was erected in 1985.^[11] Its date places it among later acts of transmission. A stone can hold a title or text in one location. A rotating observance asks people to take something up and pass it on. Kanbara’s remembrance has used both forms, leaving room for a fixed place and a recurring undertaking.

The hall and its stair can make survival seem like a line: above it, people lived; below it, they died. The subsequent history requires several lines running in different directions. Household reconstruction joined people after the loss of relatives. Land and names connected new homes with older claims. Water traveled from a source beyond the village. Memorial objects moved between houses rather than staying beside the steps.

These arrangements did not restore the old settlement intact. They made it possible for a population to continue here through forms that were partly new. Some of the work was immediate; some depended on resources repurposed later; some remained a responsibility for generations who had not experienced the eruption.

The physical place preserves the difference. The old stair descends beneath the ground on which the later community stands. The missing steps are neither a route through an open archaeological town nor merely an illustration of a disaster statistic. They mark the depth separating two inhabited landscapes. Kanbara’s history is the work of reconnecting them: making homes above the buried village while continuing to give its people a place among the living.