Hara Castle Ruins: Living Inside a Rebellion

Beside a wall in Hara Castle's main enclosure, archaeologists found nine small dwelling pits in a row. Each was only a few metres across. A narrow passage separated them from the masonry. These were traces of shelters used during the Shimabara-Amakusa uprising of 1637–38, when people from the surrounding peninsula and islands occupied a castle that had ceased to serve as a lord's residence.^[1]

The huts change the scale of the siege. A battle account follows attacks and commanders; here, the ground also records the problem of staying somewhere. Holding a fortress meant finding room for a population, sustaining shared practices and trying to obtain food or a way out. Hara offered defensible ground. Making a life within it required arrangements of the occupants' own.

Hara occupies a promontory in the southern Shimabara Peninsula. Sea and steep ground protected several sides, while marshland obstructed the western approach. The site concentrated access in ways that made it useful for defense. The same geography would acquire another meaning once an opposing force controlled movement around it: protection depended on who could cross the edges, and in which direction.^[2]

The Arima house developed the castle around the turn of the seventeenth century. Its lord, Arima Harunobu, was a Christian, and the castle belonged to a region with established Christian institutions. After the Arima departure and the arrival of the Matsukura as rulers, Hara lost its official role. The later siege therefore took place in an inherited defensive landscape, not in a fortress newly built by its insurgent occupants.^[3]

Abandonment had not removed everything useful. A later local tradition held that much of the stonework had been carried away to Shimabara Castle. Excavations instead recovered substantial defenses corresponding to the siege landscape. An administrative decision to stop using a castle and the physical elimination of its military possibilities were different achievements.^[4]

The uprising began beyond these walls. Heavy taxation under the Matsukura, punishment of Christians who refused to abandon their faith, and famine in 1637 formed its immediate setting. After the killing of a local official, resistance spread. Insurgents failed to take the Matsukura stronghold at Shimabara; across the water, fighting also reached Tomioka Castle in Amakusa. Hara became the place where large numbers eventually concentrated.^[5]

That sequence helps explain why the occupied castle contained more than a force assembled for one planned battle. The conflict had developed across settlements and across the sea. People entered Hara with connections formed elsewhere. The walls brought them into a shared enclosure without giving them identical histories.

An existing fortress could thus change the course of an uprising without having caused it. Hara supplied a place in which resistance could continue after attacks on functioning castles had failed. It also made a dispersed population a target that a surrounding army could attempt to isolate.

The dwelling pits were semi-subterranean shelters: their floors had been cut into the ground. In the investigated row, the excavators found no individual hearths or cooking installations. They proposed that food had been prepared and distributed collectively, partly to reduce the danger of fire among closely packed dwellings. No communal kitchen was identified in that report; the proposal explains the absence of cooking places in the excavated sample.^[1]

That possibility carries a practical consequence. People living together behind defenses could not necessarily continue every household task independently. Sleeping space, passage and the preparation of meals competed for limited ground. A shared cooking arrangement would make individual households dependent on distribution as well as on whatever provisions they had brought. The form of a settlement could change the way its inhabitants obtained a meal.

The regular placement of shelters also raises a question about organization. Nagasaki's heritage account interprets the arrangement in terms of families and village groups. It identifies former retainers of the Christian Arima and Konishi houses among the leaders, including men who had headed Christian associations known as kumi after the prohibition of the faith.^[2]

Those connections offer a plausible bridge between life outside Hara and conduct inside it. A village group or religious association already linked people through repeated dealings. Such a group did not need to begin by introducing all its members to one another. The interpretation is that existing relationships helped organize residence under exceptional conditions—not that an excavated hut can be assigned to a named family.

Work was also occurring elsewhere in the occupied complex. In 2024, the city's archaeologists reported siege trenches near an entrance in the Ninomaru, one of the castle's outer enclosures. In another part of that enclosure, they distinguished earlier leveling and building traces belonging to the castle's development before the uprising.^[6]

The difference keeps two building projects in view. The Arima had prepared a lordly stronghold. Its later occupants made shelters and defensive works within what remained. Their dependence on the old castle did not make them passive users of it: holding the ground required them to alter it.

The Ninomaru excavation produced another connection between fighting and life within the walls. The city's 2024 report identified lead musket balls remade into crosses and other devotional objects in the siege trenches. The find puts religious manufacture within a defensive setting, rather than leaving belief as an explanation supplied only by later writers.^[6]

Earlier investigations had already recovered irregular lead crosses. In a commentary on those finds, Ōura Yumiko describes their weight as corresponding to musket shot and discusses their interpretation as objects made during the siege. Medals and rosary beads were recovered too. A rosary is a sequence of beads used to keep track of repeated prayers; these finds concern practices that could be carried into a temporary shelter.^[7]

Changing a piece of ammunition into a devotional object gave the same material a different use. It does not follow that its maker had stopped resisting, or that everyone inside had made the same choice. In the trenches, the two activities occupied the same wartime environment. Religious practice did not have to wait until the work of defense was finished.

A larger object connects the uprising with public religious imagery. The banner traditionally associated with Amakusa Shirō, a leader of the uprising, depicts a chalice beneath the Eucharistic host—the consecrated bread of Catholic worship—with angels praying on either side. It is preserved in the Amakusa Christian Museum, across the sea from Hara. Its date and original setting remain disputed: the municipal record gives both a possible production inside Hara by the painter Yamada Emosaku and an earlier use at a seminary.^[8]

The image consequently belongs to the uprising's material history without supplying a secure scene of someone painting it during the siege. The portable banner and the objects excavated from the ground also reached the present by different routes. One was preserved away from the battlefield; the others emerged from it.

Together they establish a Christian presence expressed through things people could use, make and display. They do not require an answer to the more difficult question of what every person in the crowded fortress was prepared to endure for that faith. A message attributed to someone inside offers a notably different kind of evidence.

A letter preserved by the Tsuboi family is identified as a yabumi, a message attached to an arrow and sent from the insurgent side. A reproduction belongs to the interpretation at the Arima Christian Heritage Museum. The writer is unnamed, and the route by which the Tsuboi obtained the document is uncertain; it may not have reached the senior officials for whom it was intended.^[9]

In the modern Japanese rendering published by Nagasaki Prefecture, the writer connects conversion with harsh taxes and the Christian practice of helping people without property. The appeal then offers to abandon Christianity permanently if that is the shogun's wish. Economic distress, mutual assistance and a possible religious concession appear in the same argument.

The message does something more specific than announcing that the insurgents had mixed motives. It tries to make their conduct intelligible to an adversary. Rather than treating prohibited belief as an inexplicable refusal of authority, it describes circumstances under which people might have entered a community that offered help. The proposed renunciation then presents a possible way to end the dispute.

That is an argument seeking a result, not an impartial survey of conversion. We cannot assume its unnamed author spoke for Shirō, for the other leaders or for all the people in the huts. The letter is valuable precisely as a particular intervention: someone associated with the occupied castle was trying to describe a future other than continuing resistance under the existing terms.

Read beside the crosses, it prevents an easy choice between “religious revolt” and “tax revolt.” Objects of devotion demonstrate religious activity. The letter makes Christian charity part of a complaint about material conditions, then offers a concession concerning the faith itself. These are different acts within the same conflict, not answers that cancel one another.

The uncertain delivery matters too. Writing a proposal is not the same as getting it considered. The castle's occupants faced a problem of communication as well as a problem of defense. An appeal could cross a wall without entering the decision-making process that might alter the siege. Hara's boundaries separated people from supplies, but also separated an intended argument from an assured audience.

For the attacking forces, prolonged resistance created problems of their own. Itakura Shigemasa, an early shogunal commander, died in an unsuccessful assault. The siege continued under Matsudaira Nobutsuna, with forces supplied by several domains. Hara's defenses had enabled resistance to direct attack; blockade offered another means of bringing the occupied ground under control.^[5]

A letter by Nobutsuna reveals a small but consequential movement across the lines. Writing to Hosokawa Tadatoshi, the lord of Kumamoto, on the twenty-sixth day of the second lunar month of 1638, he referred to information from a woman who had come out the previous day. Having considered what Tadatoshi had conveyed, Nobutsuna wanted instructions sent to capture Shirō alive. He would discuss the matter further in person.^[10]

The surviving wording gives neither the woman's name nor the substance of her account. It does show her information passing through a domain lord into the shogunal commander's plans. A person leaving the enclosure could become a source of intelligence about it. The besiegers' understanding was still being revised near the end of the campaign.

The chain is unusually revealing: a woman came out; Tadatoshi communicated what she said; Nobutsuna considered it and proposed an instruction. None of these steps was performed by the celebrated figure at the center of the rebellion. Shirō was the intended object of the order, while a person largely lost from the record helped change what the attackers wanted to do with him.

This was not a general offer of safety to the population. The instruction singled out a leader; it did not propose terms for the households living beside the walls. Nor does the document tell us that the unnamed woman could secure the release of anyone still inside. Crossing the lines changed her relation to the siege, but the letter preserves that change chiefly through its usefulness to the men conducting it.

The two surviving messages therefore illuminate different directions of movement. The arrow letter argued for an altered outcome from within; the commander's letter used information from someone who had left. Both expose decisions being attempted while the castle still held, rather than a population already committed in every particular to the ending that followed.

The siege lasted for months. Food restriction wore down the occupants, and the final assault in 1638 brought mass killing. The dead included men, women and children. Some people had surrendered or left during the conflict, but most of those caught in its final destruction did not survive. Published totals vary; none is needed to recognize the scale of the catastrophe.^[11]

The failure was not that Hara had provided no protection. Its defenses had made sustained resistance possible. The problem was that continued defense could not restore access to the resources needed by the people behind it. A wall changes the conditions of an assault; it cannot grow food for the population it encloses.

The possible communal provisioning suggested by the huts belongs on this side of the story as well. Distribution can organize a supply. It cannot by itself replenish one. Any arrangements that helped people remain together were operating within an increasingly restricted situation. Organization and vulnerability were present at the same time.

The defeat was followed by work on the castle itself. Nagasaki's archaeological center identifies systematic disabling of the defenses, with masonry broken and buried. Its interpretation places extensive destruction after the uprising, separate from the earlier cessation of official use.^[12] Hara had demonstrated what a disused stronghold could still enable; the later damage removed much of that capacity.

The destroyed site was not subsequently rebuilt as a working castle. This is why its most important surviving form is so unlike the towers associated with other famous fortresses. Open ground and broken masonry record the undoing of an occupied place. Restoring a complete imagined castle would risk concealing the very action that explains the ruins.

Human remains give that destruction a further dimension. Excavations around the entrance recovered bones beneath large facing stones, smaller packing stones and soil. At a turret platform, corner stones had been removed. Ōura's account connects these patterns with the thorough dismantling of the complex.^[13] The evidence brings the killing and the disabling of the fortress into the same physical sequence, without requiring us to supply the private thoughts of the men moving the stones.

Later commemoration returned people to the emptied landscape in a different way. The Honekami Jizō, a Buddhist memorial figure at the main enclosure's entrance, is associated with an eighteenth-century effort by a local abbot and village headmen to gather and honor the dead. Local tradition places the undertaking in 1766, more than a century after the siege.^[14]

The memorial is not a surviving object from the insurgents' worship. It records a later response to their remains. In a place where excavated crosses attest Christian practice, a Buddhist figure became part of the work of commemoration. The difference in religious form does not make either layer disappear.

Archaeology has since added another kind of attention. A hut floor restores the question of habitation; a trench locates defensive work; a remade piece of lead identifies an act of devotion. The messages add attempts to explain conduct and change instructions. None speaks for every occupant, but each recovers an action that the final destruction could otherwise overshadow.

Hara's people shared a predicament more securely than we can establish a shared purpose. They made arrangements to stay, practiced a prohibited faith, communicated across the lines and sometimes left. The stronghold gave these actions a common setting while restricting what could follow from them.

The ruins preserve the end of that setting. Reading them through the lives organized inside—rather than only through the armies around them—makes the loss more specific. What was destroyed was a defensible castle, but also a place where people had continued trying to live.