Matsumoto Castle: Where the Wall Opens

At Matsumoto Castle, a room beside the main keep can be opened on three sides. Its thin wooden panels can be removed, leaving the roof supported above a space bordered by a vermilion-railed veranda. This is the Tsukimi-yagura, the moon-viewing turret. It belongs to the same connected group as thick-walled buildings equipped with firing openings.^[1]

The contrast invites a ready explanation: a fortress made for war acquired a room for peace. There is truth in the difference of purpose. But the building has more to say about how those purposes could coexist, and recent research has changed the chronology behind the familiar account.

The useful question is architectural as well as historical. How could part of a defended stronghold become so open without the whole ceasing to be a stronghold? The answer lies in the relationship among its surrounding enclosures, the timber frames that carried its roofs, the passages joining its buildings, and the particular places where walls were made to open.

Matsumoto is a flatland castle, built in the basin of central Nagano rather than on an isolated mountain summit. Its historical defenses extended through three lines of water moats and successive enclosures. The main keep occupied part of the innermost enclosure, alongside a palace; other residences and storehouses stood farther out. The tower group was one component of a much larger arrangement.^[2]

That larger setting changes the meaning of an open room. The moon-viewing turret did not provide an unguarded entrance from an ordinary street. Its veranda belonged to buildings within the castle's defended landscape. An occupant could have a broad view without making the route to that room equally open.

The water also posed a construction problem. The main keep's foundation occupied soft ground near the edge of an alluvial fan. Investigation identified sixteen substantial timber supports inside the stone platform, connected into a framework beneath the building. The castle's technical account interprets them as members placed during the platform's construction, not piles later driven through completed stonework.^[3]

The visible masonry and the building above it therefore had different parts to play. A stone face retained the platform; concealed timber helped transmit the weight of the superstructure toward the ground. Looking only at the stones makes it easy to imagine a solid pedestal carrying everything in the same way. The internal arrangement was more discriminating.

This is the first reason to read Matsumoto from inside as well as outside. The surface enclosing a space is not necessarily the structure bearing all the weight above it. Timber posts and beams could carry roofs while the intervals between posts received walls, windows, or movable panels. That distinction would become particularly conspicuous in the moon-viewing room. Before considering its openness, however, it is worth looking at how carefully the earlier defensive spaces used their smaller openings.

The main keep and its northern companions contain openings for arrows and firearms. These connect a protected interior with the space outside while exposing much less of the occupant than a broad doorway would. The castle's conservation inventory treats the firing apertures, windows, doors, and downward-facing openings as distinct architectural elements, rather than as one category of hole in a wall.^[4]

Their directions matter. An opening through a vertical wall addresses the ground beyond it. One beneath a projecting part of the building allows attention to the sloping masonry immediately below. A person close to the base of a wall and someone farther across the moat do not occupy the same line of sight. The combination extends what could be watched or threatened from within the structure.

The interior needed to make those positions usable. On the main keep's first floor, a perimeter passage runs below the level of the central floor. The castle's building account explains the difference through a doubled arrangement of foundation timbers beneath the inner portion. A step encountered inside the room has a structural explanation as well as consequences for movement.^[5]

It is tempting to make every awkward feature a trick intended to defeat an attacker. The timber arrangement offers a more concrete account. The building had to carry its weight, fit its platform, and allow people to move around its edges. Those requirements interacted; they were not all disguised weapons.

The surviving apertures establish provision for defense, not a record of a successful engagement. Their arrangement is informative enough without supplying an imaginary siege. A useful reading follows the actual sequence: shelter behind a wall, access along its inner face, an opening directed toward a particular area outside. The moat, stone slope, and occupied floor are parts of the same defensive relationship.

That relationship also explains why openness should be measured locally. A fortress could contain many apertures while remaining difficult to enter. The essential question was what could pass through each one, in which direction, and under whose control.

The group now designated a National Treasure consists of the main keep, the smaller Inui keep, the Watari passage turret, the Tatsumi annex, and the Tsukimi turret. The names describe connected buildings, not five interchangeable towers. Their roofs and internal floors do not follow a single scale.^[6]

On the north, the first floor of the main keep connects with the second floor of the Watari passage and Inui keep. Toward the southeast, it connects with the first-floor level of the annex and moon-viewing room. The conservation plan and its published floor drawings record these offset relationships.^[7]

The practical result is easy to miss in a photograph. Crossing a connection does not necessarily mean climbing another storey, even when the destination's floor number changes. Each building starts from its own relationship with the platform below. The passages make those unequal structures usable together.

Their carpentry is different too. Inui uses round as well as squared posts, and its standard interval between columns is smaller than the main keep's. These differences help identify the small keep as its own building rather than a reduced copy of its neighbor. They do not, by themselves, provide a date for its construction.^[8]

Nor do five apparent roof stages give the main keep only five floors. It has six. The third floor lies largely within the zone enclosed by the roof below, while a south-facing gable provides light. Inui's corresponding roof-enclosed level lacks that source of daylight.^[8] Roof form and useful internal space do not have to advertise the same count to someone looking from outside.

Together, these features produce a more demanding picture than a vertical stack of identical defended rooms. The castle joined buildings with different starting levels, column intervals, and relationships to their roofs. Movement through the group required the connections between them. That is why the attachment of another room could change the range of uses without requiring the earlier buildings to be remade from the ground up.

The broad historical setting is secure. Ishikawa Kazumasa entered Matsumoto in 1590 and began enlarging the castle establishment; his son Yasunaga continued the work. Assigning each surviving building to an exact year has been harder. The city's older account favored 1593–1594 for the main group while acknowledging that its evidence did not settle the matter conclusively.^[9]

A study released in 2025 added timber dating to the architectural investigation. Dendrochronologist Mitsutani Takumi identified original main-keep timbers felled in 1596. Architectural historian Fumoto Kazuyoshi combined that evidence with construction details to propose a main-keep building campaign around 1596–1597. The study also argues that the Tatsumi annex may belong to the original campaign, rather than to a later addition with the moon-viewing turret.^[10]

The implications extend beyond moving a date by a few years. The familiar silhouette had encouraged the eastern attachments to be treated as a pair. Construction evidence can separate them. Two adjacent buildings need not have arrived together, just as two different kinds of interior need not represent different centuries.

At a public presentation of the research in November 2025, the investigators explained the revised sequence; the organizer's account reported that the city would consider changes to its official interpretation alongside further investigation.^[11] The result should be presented as a research conclusion, not silently substituted for a dated construction inscription that has suddenly appeared.

Tree-ring dating answers a precise question about timber. A construction date requires a further judgment about that timber's place in the building and the interval between felling and use. Those are connected questions, but not identical ones. Here, the combination of dated wood and architectural relationships has made the later-1590s case substantially more specific.

The useful consequence for reading the castle is straightforward. Its parts cannot be divided into “military” and “peaceful” simply by grouping whichever roofs happen to touch. The sequence must be established from the buildings themselves. Their different functions can then be understood within that sequence, rather than used as a substitute for it.

The Tsukimi room offers the most striking change in how the outside could be encountered. Its three openable sides use mairado, wooden panels strengthened with narrow horizontal strips. Its veranda and shaped ceiling give it the character of an inhabited viewing room, rather than a firing position. Yet “open-sided” describes an available configuration, not an obligation to leave it permanently exposed.^[1]

The distinction is practical. Closing the panels restores an enclosure; removing them allows a broader exchange of light, air, and view. The roof remains in place. Its support does not disappear with the panels, because the frame and the infill perform different tasks. An opening can therefore become much wider than a gunport without the building losing its shelter overhead.

The section drawing made for the twentieth-century repairs shows the moon-viewing room beside the taller annex, with occupied space above a lower level and the stone platform beneath. It is an addition fitted to the castle's built edge, not a pavilion set down independently on a garden lawn.^[12]

That connection supplies the historical point. The new room could draw on the stronghold's location and controlled approaches while offering an experience unlike that of its defended compartments. A broad view and restricted access were entirely compatible. “Open” described the room's wall, not the social availability of the castle.

The 2025 study dated a moon-turret beam to felling in 1626 and proposed construction around that time. This challenges the traditional association with preparations for a shogunal visit in the 1630s.^[10] A particular high-ranking guest is unnecessary to explain the architecture. The room's surviving configuration gives firmer evidence of the kind of occupation it accommodated than a story assigning it to one anticipated evening.

What changed was the treatment of a particular edge: instead of demanding a narrow, defended opening, it permitted a room to expand toward its surroundings. The neighboring keep did not have to surrender its own walls for that to happen. The two arrangements could remain connected precisely because they did different work.

The older buildings also resist an exclusively martial reading. On Inui's upper level, curved katōmado windows use a form associated with Buddhist architecture and prestige. Such windows also appear in the Tatsumi annex. Their shape participates in the building's presentation as well as admitting light.^[8]

A decorated window is not proof of a peaceful construction date. It is evidence that a fortified building could possess an architectural language beyond the simplest provision of cover. Different openings within the same group could address different expectations: a firing aperture, a framed view, a place where an interior announced its standing.

Weather created still another requirement. Inui's window sills have drainage holes that allow water collected while the shutters are closed to escape outward.^[8] This tiny passage reverses the emphasis of a gunport. Its purpose is not to provide a person with access to the outside, but to keep unwanted water from accumulating inside the joinery.

The main keep's black lower boarding belongs to the same practical world. It shields earthen walling where the roof cannot keep off all the rain. A wall section removed during repair reveals timber lathing, earth, and plaster, together with an additional outer layer applied in a later campaign.^[5]

The wall was therefore an assembly with vulnerable surfaces, not an abstract mass of “defense.” It had to remain sound in ordinary weather if its military provisions were to remain usable at all. Protection from an attacker and protection from rain could meet in one piece of architecture without becoming the same problem.

These details give substance to the coexistence visible at the moon-viewing room. The castle did not wait for a new era before needing light, drainage, shelter, or a distinguished appearance. What varied was their arrangement and emphasis. A large openable room made one of those possibilities unusually conspicuous; it did not invent every use of a castle that involved something other than fighting.

The evidence is available because Matsumoto's buildings were maintained and investigated, not because they stood untouched. Major dismantling repairs took place from 1950 to 1955. Work exposed concealed construction and produced the detailed drawings through which the connected group can now be studied.^[13]

The distinction between preserving a building and preserving every surface is especially important here. The city's conservation plan separates elements whose material itself should be retained from elements requiring periodic repair or replacement while maintaining their form and finish. Structural timbers and many openings belong to the first concern; plaster and roof coverings require different treatment.^[4]

That policy corresponds to a real difference in the evidence. An old timber may preserve a growth sequence, tool marks, and traces of former joints. A renewed plaster surface can protect the frame without offering the same chronological record. The work of keeping rain out can help preserve the pieces through which an earlier construction history becomes recoverable.

The surviving group consequently permits more than admiration of a unified silhouette. It allows the joining of unequal buildings to be examined, the directions of openings to be compared, and proposed dates to be tested against material. Its historical value does not depend on every conventional attribution remaining unchanged.

The moon-viewing room completes that reading without dissolving the fortress around it. Its removable panels offered a different relationship with the world outside, within approaches and boundaries that remained controlled. The change was accomplished by adding and connecting particular spaces, not by converting every defended floor into a place of leisure.

Matsumoto's achievement lies in that capacity for differentiated use. A wall could provide cover, shed rain, frame a view, or withdraw behind a set of removable panels. The castle became more versatile at the places where those choices met. Its open room is most revealing when the rest of the stronghold stays in view.