Fushimi Castle: The Ground Beneath the Government

In 2007, an excavation below Fushimi Castle found something less spectacular than a palace: a roadside drain and a line of fence posts. They belonged to a property thought to have housed the lord Satake Yoshinobu. The excavators placed these boundary works before the preparation of the residence itself. A street and its edges had been organized before the household's buildings went up.^[1]

That sequence supplies a useful starting point for a castle whose most conspicuous towers are modern. Before a government could occupy an imposing residence, someone had to make a place where its associated households could live. At Fushimi, this involved roads, drainage, the allocation of plots, and large quantities of moved earth. The political center depended on work extending well beyond its principal enclosure.

Its disappearance was similarly uneven. Buildings could be dismantled and transferred; a terrace might remain under later occupation. Following those different fates reveals how a residence became a center of government, and why removing that government did not simply return the ground to its previous condition.

Toyotomi Hideyoshi began a residence at Shigetsu, in southern Fushimi, in 1592. Archaeologists distinguish this initial retirement residence from the larger castle project developed from 1594, and both from the castle subsequently built on nearby Kohatayama. “Fushimi Castle” names a succession of establishments, not a single building slowly acquiring additions.^[2]

Hideyoshi was the military ruler who had brought much of Japan under his authority. A retirement residence for such a household could become a place where political business followed its occupant. The expansion at Fushimi made that possibility concrete: the castle developed together with a surrounding settlement of lordly residences, merchants, and artisans. The Kyoto prefectural account includes these neighboring properties within the extensive outer enclosure.^[2]

Fushimi offered connections as well as ground. South of Kyoto, it stood between routes toward Osaka, Nara, and the country around Lake Biwa. The city's history dates the development of its port for the castle's building materials to 1594. River works separated the Uji River from Ogura Pond, while embankments also carried roads. Bringing construction supplies to the residence meant altering the means of reaching it.^[3]

This is a more practical relationship between residence and government than the familiar image of a ruler surveying the country from a tower. A household established on a hill still needed access from below. The arrival of other households enlarged the task: a single residence could be approached by a path, but a district of compounds required organized boundaries and circulation. Roads and plots gave durable form to the act of gathering people around a political center.

Nor did those arrangements have to be invented anew by every arriving lord. The roadside sequence found in 2007 suggests a distinction between work done for the developing district and work done after an individual plot had been assigned. The castle's neighborhood was being made at more than one scale. That distinction would become particularly important when disaster forced a change of site.

The earthquake of 1596 destroyed the Shigetsu castle. Hideyoshi rebuilt nearby on Kohatayama, occupying the new establishment in 1597. He died there the following year. The two castle zones therefore belong to different stages of his last years; the later hill cannot supply the exact setting for everything associated with the earlier residence.^[4]

Yet the move did not take the political center away from Fushimi's transport connections and developing town. It shifted the principal residence within that landscape. This was continuity at the scale of a district, accomplished through a major interruption at the scale of the castle itself. One can recognize both without supplying an undocumented private calculation about why Hideyoshi chose the new ground.

Archaeology makes the consequences of that shift more interesting than a simple change of address. A 2017 investigation in the southern Shigetsu area found alternating deposits of gravel or broken tiles and clay associated with castle-period ground preparation. Later maps place the investigated area within a residence attributed to Terazawa Hirotaka, another lord. Ground connected with the earlier castle could participate in the residential landscape of the later one.^[4]

The excavation was small, and the identified features did not disclose an entire palace. What they did expose was the labor of making land usable. Laid deposits are evidence of construction even when no standing room survives above them. The ground itself had become part of the establishment's built fabric.

Elsewhere, archaeologist Okui Tomoko has approached the same problem by bringing together records from numerous restricted investigations. Her study identifies substantial preparation west of Date-kaidō, a road through the castle district, while distinguishing documented fill from natural ground and less certain interpretations. Across the reconstructed terrace, differences of several meters separate the inferred earlier terrain from the later surface.^[5]

Seen through this work, “building a castle” includes creating the land on which residences could stand. The hill was neither wholly artificial nor an untouched natural platform. Particular slopes were cut or filled, and their resulting surfaces joined roads and property boundaries. A new ruler might replace a roof or gateway while continuing to use that expensive preparation underneath.

The Satake-attributed property shows how the larger undertaking met the requirements of one compound. After the roadside works, builders leveled a slope falling toward the northwest. Stone foundations and paired pits containing foundation stones belonged to the subsequent construction. The excavation team tentatively interpreted these as a gateway and associated structures facing Yamato-kaidō. Roof tiles carrying the Satake family's fan-and-moon emblem strengthened the proposed identification of the household.^[1]

The sequence is more informative than the emblem alone. A crest can attach a powerful name to an object; the arrangement of deposits shows how a residence took its place in a neighborhood. First came a usable frontage. Then the assigned ground was prepared for buildings. The possible gateway addressed the road already running along it. Even without reconstructing the rooms, the evidence connects the household's entrance to the district's organization.

Other excavations across Fushimi have recovered residence boundaries, roads, drains, gates, ceramics, and decorative tiles. The archaeological institute's district survey describes a broad arrangement of lordly compounds on the western slopes and commercial occupation farther west. Its individual examples include boundaries between neighboring properties, rather than simply fragments of the ruler's own palace.^[6]

These discoveries change the scale at which the castle can be understood. A map dominated by a central keep encourages attention to one household. The excavated district introduces the households assembled around it, each needing a defined place in relation to streets and neighbors. Political proximity acquired a material cost: enough level ground, a boundary, an entrance, and a connection to the routes serving the center.

The traces also set limits on how confidently that world can be peopled. A tile does not record everyone who occupied a compound, and a proposed gate cannot reveal who passed through on a particular morning. There is nevertheless a substantial history between an anonymous hole in the ground and an invented scene. The builders' sequence tells us that the residence joined an organized district rather than simply appearing beside a finished castle.

For the castle's next occupants, that distinction mattered. Reusing Fushimi meant taking over a place already equipped to accommodate political households. Destroyed buildings did not necessarily mean that all the work of assembling the center had to begin again.

Hideyoshi's death did not immediately end the Toyotomi household. His son Hideyori, born in 1593, was still a child. Ieyasu initially entered Fushimi as a senior councillor serving the young heir.^[10] A powerful man's presence in the castle therefore meant different things at different moments. The establishment first accommodated him within the arrangements surrounding Hideyori; it later served his own government. The sequence helps explain why identifying every phase simply by the most famous resident can be misleading: a residence, its occupant, and that occupant's authority did not always change together.

Fushimi burned in the fighting of 1600, during the conflict that culminated at Sekigahara. Tokugawa Ieyasu subsequently rebuilt the castle. The prefectural archaeological account emphasizes that the Tokugawa establishment retained the basic arrangement of the Toyotomi earthworks and much of the earlier stonework, despite the destruction of buildings.^[2]

This was a consequential form of inheritance. A predecessor's prepared ground could remain useful even when its superstructures and occupants changed. The evidence does not make every surviving wall “Hideyoshi's,” but it does rule out treating the Tokugawa castle as an entirely separate construction imposed on untouched land. The rebuilding drew upon an existing physical framework.

Fushimi also continued to receive people whose business required a political meeting place. In 1603, an imperial envoy came here to convey Ieyasu's appointment as shogun, the head of the military government. In 1615, his successor Hidetada assembled the regional lords at Fushimi and promulgated the Buke shohatto, regulations for the warrior houses.^[7] These were distinct occasions, separated by twelve years, but both depended on Fushimi being an active setting for government rather than a monument to its previous ruler.

The second occasion is especially useful for understanding the relationship between castle and neighborhood. Regulations addressed to houses governing different territories were announced to lords gathered at a particular place. The national scope of the document did not abolish the local work of assembling its recipients. Fushimi provided a setting in which scattered territorial authority could be addressed together.

Its role also outlasted Ieyasu's regular residence. Historian Wakabayashi Masahiro describes Hidetada returning with lords after the destruction of the Toyotomi house and Ieyasu's death, using Fushimi for territorial reassignments and negotiations with the imperial court. The castle remained an operating Tokugawa establishment until the third shogun's accession in 1623, after which it was dismantled.^[8]

The sequence resists an abrupt ending in 1615. Fushimi did not become useless the instant the Tokugawa defeated their principal rival at Osaka. It continued to serve particular gatherings and negotiations. Its eventual dismantling ended that residential and political function, rather than merely commemorating a victory already won elsewhere.

Once the castle ceased to operate, its buildings could have futures separate from their Fushimi setting. The best way to follow that dispersal is through individually documented structures, not a catalogue of everything later said to have come from the castle.

Fukuyama Castle offers a particularly strong example. The city's account records that Hidetada granted buildings from Fushimi for the new castle completed in 1622. During dismantling repairs in 1954, an inscription cut into a beam of Fukuyama's Fushimi Turret identified it with the eastern turret of Fushimi's Matsu-no-maru enclosure. The material discovery supports a specific transfer, rather than a resemblance between two imposing buildings.^[9]

Its timing is instructive. The new castle's completion preceded Fushimi's formal dismantling in 1623. Transfer and final abandonment need not have been one simultaneous event. Nor did the move preserve the turret's old relationship to neighboring buildings, roads, or households. A structure could continue to function while the establishment that had given it its first position was being dispersed.

The local stonework tells a smaller, different story. In 1977, development exposed a substantial stretch of foundations belonging to a narrow castle enclosure. A section was moved to Momoyama Higashi Elementary School for preservation. It is genuine castle fabric, but its present setting is the result of a modern conservation decision.^[10]

Unlike the turret, this wall section was not transferred to serve another working castle. It was saved as evidence. Relocation preserved stones and a portion of their assembly while separating them from the full enclosure whose ground they had helped define. Reading it as a surviving boundary in its original position would lose the very history that made its survival possible.

The two cases illuminate what dismantling could and could not erase. Timber and masonry were separable from a place and valuable elsewhere. The roads, levels, and relationships among compounds were much harder to pack up. Once the households left, a transferred turret could still be used, but it could not carry Fushimi's political neighborhood with it.

The former castle landscape did not remain an empty diagram. At the Shigetsu investigation, the archaeological report traces later woodland, military use, cultivation, and residential development. Substantial modern fill lay above the earlier deposits. The surface encountered by an excavator consequently recorded later projects as well as the castle's making.^[4]

Across the district, other transformations became much more visible. Emperor Meiji's Fushimi-Momoyama Mausoleum occupies the area of the later castle's principal enclosure. The archaeological institute's map locates the imperial burial ground within the old castle terrain; the Imperial Household Agency identifies and administers it as Meiji's mausoleum.^[11] Its approach leads to an imperial place of remembrance, not into an exposed castle palace.

Farther north, the paired towers built for an amusement development in 1964 supplied a different attraction. Kyoto City's account identifies them as imitation castle architecture in an area of the former northern enclosures, not the historical main keep. Their silhouette belongs to the twentieth-century presentation of Fushimi's past.^[12]

Neither later use recreated the neighborhood that the excavated roads and residential plots reveal. The mausoleum gave part of the hill a new institutional purpose; the towers made a recognizable castle image. Below and around them, fragments of prepared ground survived without necessarily announcing what they were. Visibility and historical continuity no longer ran together.

The town below had its own continuation. Wakabayashi's account of Fushimi's water transport describes the former castle town developing as a lodging and distribution center along the routes linking Osaka and Kyoto. The port's later work did not require a ruler to keep his residence on the hill.^[13]

This is the particular legacy of Fushimi's short career as a political center. Its construction had joined a ruling household, neighboring residences, prepared land, and transport connections. Its dissolution pulled those elements apart. Some buildings traveled; some boundaries disappeared; some ground remained useful under new occupation. To recover the castle's history is to put those relationships back into view—not to put every surviving stone beneath the nearest castle-shaped roof.