WINDOW RESEARCH INSTITUTE

In this series, architectural historian Terunobu Fujimori, who has traveled extensively to study buildings of various times and places, will be introducing a selection of notably intriguing windows found in historic buildings from all across Japan, one at a time. The subject of the ninth installment is the Kikugetsutei, which stands within the Ritsurin Garden, a nationally designated Place of Scenic Beauty. Its main space that can be fully opened to the surrounding pond and garden is equipped with sliding storm doors known as amado along its perimeter. Stowed away from view during the day, the amado reappear when the park closes, swiftly transforming the space into a closed room. What makes this quick transition possible is a feature known as a tomawashi.

 

Until quite recently—that is, until the industrialization of housing took off in the 1960s—nearly all houses and other ordinary wooden buildings in Japan consisted of tatami-floored rooms that were enwrapped, in order from inside to outside, by shōji [sliding translucent screens], an engawa [perimeter corridor/veranda], garasudo [sliding glass doors], columns, and amado [sliding storm doors]. The three components other than the columns and engawa—the amado, garasudo, and shōji—were dedicated to closing off the openings. Sometimes buildings also had an external engawa, aptly named a nure-en [wet corridor/veranda], forming an additional layer outside the amado.

Of the trio of components, garasudo did not become ubiquitous until after 1869 (Meiji 2), when the oldest known examples were installed in a barbershop in Yokohama. Amado and shōji hence were the only two components used to separate the inside and outside along the openings of houses prior to Japan’s modern age, or more specifically, through to the end of the Edo period (ca. 1603–1868). The outermost amado shielded against the wind and rain, the engawa ran behind them, and the shōji fitted along the boundary between the engawa and internal rooms filtered in the sunlight while also providing visual privacy.

Amado and shōji thus represent the two solutions that Japan’s traditional wooden architecture arrived at after developing and refining various ways of separating the inside and outside over the course of its long, heralded history. Made of wooden boards and paper, respectively, the two are immediately distinguishable in terms of their materiality, but where they truly differ is in how they function as devices for closing openings. Amado are positioned outside the columns and hence can slide unobstructed over longer distances, crossing beyond the boundaries of the interior rooms. Shōji, on the other hand, are positioned between columns and hence cannot slide beyond a room’s limits (Note: in Japanese, the expression tateru [“place upright”] is used to describe the act of closing an opening with amado or shōji).

The origin of shōji can be traced back to the partitions used within rooms of classical shinden-zukuri architecture, while amado are thought to have emerged not in the classical or the medieval age, but even later, in the early modern age.

How we know this is through the following anecdote about Tokugawa Ieyasu from the Azuchi-Momoyama period (ca. 1568 –1603).

The story goes that in 1586 (Tensho 14), just as Ieyasu and his men were preparing to feast during their first visit to Kyo (present-day Kyoto), a tension rippled through the party as a sudden clamor arose from the surrounding streets. Fearing a surprise assault, they looked outside, only to find that the locals were shuttering their homes with thin boards that seemed to appear from nowhere. They watched on as more boards kept being pulled out from the walls like some kind of magic trick.

Amado were already commonplace in Kyo at the time but had yet to be introduced to the Kanto region, where Ieyasu was based. He surely would have brought the brilliantly practical invention back with him to Edo (present-day Tokyo).

So, after going on to spread nationwide, the amado-shōji door/window fitting duo would be joined by garasudo in the Meiji era, forming the amado–garasudo–shōji trio that has persisted to this day. However, the shift to the duo was limited primarily to residential architecture, and the original duo has remained intact in large-scale structures that place particular importance on tradition, namely temples and shrines. In such buildings, the eaves, which serve to stabilize the top tracks of the amado, are too high up, and amado cannot be that large. Amado must be lightweight in order to be moved from side to side, and they have traditionally been made to a standard height of 1.8 meters.

The great amado–shōji duo that so managed to survive would make the big stage in 1893 (Meiji 26) Chicago, where they put on a show of spatial theatre that has left a lasting mark in modern architectural history.

That was the year of the World’s Columbian Exposition, in which the participating Meiji government presented the Phoenix Pavilion (Hooden), Japan’s national pavilion. The timber building, whose overall layout resembled the Hoodo (Phoenix Hall) at the Byodoin Temple in Uji, was designed by Josiah Conder-trained architect Masamichi Kuru and constructed by Japanese craftsmen.

The event marked the first instance when Japan’s traditional architecture was formally showcased to the world. Many of America’s admirers of Japanese art wandered through the structure, which was crafted in the three architectural styles of shinden-zukuri, shoin-zukuri, and sukiya-zukuri, marveling at the byōbu [folding screens], fusuma-e [partition paintings], and tokonoma [display alcove] that adorned its interiors in a manner that would have felt rather excessive to Japanese eyes. Among them was an architect who was awestruck, not so much by the artwork and ornamentation but by the composition of the architectural space.

That architect was a 26-year-old Frank Lloyd Wright. An employee at Louis Sullivan’s firm at the time, he grew curious about the peculiar-looking Phoenix Hall during his worksite visits to the Transportation Building that Sullivan designed for the fair. Upon visiting the completed pavilion, he found the answer to a design problem he had been exploring.

He had been trying to figure out how to free architecture from compositions consisting of aggregations of wall-enclosed spaces, instead creating spaces that continuously flow from one room to the next and from the inside to the outside.

The timber-framed Phoenix Hall effortlessly solved what seemingly was an impossible problem to crack using walls of heavy masonry or brick. Its solution: create an orderly arrangement of columns defining the corners of the rooms, partition the rooms by inserting shōji and fusuma between the columns, and wrap the building with amado along the outside of the perimeter columns.

Wright studied ways to translate the spatial continuity (fluidity) created within and around the wooden building into an architecture of brick or masonry appropriate to his time and place. His solution: in plan, minimize the number of partition walls and utilize horizontally elongated openings to enhance the sense of continuity, and in section, extend the eaves far out to articulate the sense of horizontality.

In 1910, he released a book of drawings and photographs of the new houses he designed through this approach, such as the Robie House (1908). He compiled the book in Italy and published it through the German publisher Wasmuth in German. He chose German because Germany, at the time, was at the forefront of the budding movement of modernism.

It hit its intended mark: architects including Walter Gropius and Mies van der Rohe of Germany and Gerrit Rietveld of the Netherlands followed his example and produced designs that placed an emphasis on continuity, such as the Fagus Factory (1911) by Gropius and the Schröder House (1924) by Rietveld. What’s more, this trend was cemented as a principle of modern architecture by Gropius’ Bauhaus Building (1926) in Dessau and eventually spread worldwide, including to Japan.

The amado–shōji duo’s show of spatial theatre thus circled the globe, first performing in Azuchi-Momoyama Kyoto, then in 1893 Chicago, and then in 1910 Germany, before returning home to post-1926 Japan, where numerous famed works of early modernism were born.

There are countless surviving examples from throughout history where one can see this duo that unexpectedly played a pivotal role in reshaping the world. Masterpieces of sukiya-zukuri architecture, such as the Katsura Imperial Villa and Sankeien Rinshunkaku, immediately come to mind, but here I would like to discuss the Kikugetsutei at the Ritsurin Garden, a designated Place of Scenic Beauty in Takamatsu, Kagawa.

While the Katsura Imperial Villa and Rinshunkaku are works that are appealing in a wide variety of ways, one could say that the Kikugetsutei bests both in purity because spatial continuity is its singular point of focus.

The site of the Kikugetsutei, the Ritsurin Garden, was initially designed as a private garden for a local powerful family in the late sixteenth century. After the Ikoma daimyō clan implemented extensive flood control works around the site in the early seventeenth century, opening up the possibility for its significant expansion, the garden became the property of Matsudaira Yorishige, the first daimyō of the Takamatsu Domain, in 1642 (Kanei 19). It was then completed during the rule of the fifth daimyō, Matsudaira Yoritaka, in 1745 (Enkyo 2).

Rooted in the Heian Pure Land garden tradition, this landscape of water, greenery, and rocks was designed with the assumption that important guests would arrive at the Kikugetsutei from across the pond by boat. The building’s affinity with the pond is maximally enhanced by its low placement near the water’s surface. It is staged splendidly from the viewpoint of the interior as well in that it gives you the sensation of being on a boat when looking out at the garden.

I would like you to first see the Kikugetsutei’s most crucial room, the Ninoma [Second Room], with all its amado and shōji hidden. The inside and outside are completely blended. The shōji have been removed from between the columns and stowed away, while the amado that stood along the engawa have been drawn aside one after another and stacked upright inside tobukuro [door storage boxes] situated at the back corners of the room.

This whole procedure should feel straightforward to those familiar with old Japanese wooden houses. But think again about the position of the amado along the railing at the front end of the room. Shouldn’t it be impossible for the flat amado to turn 90 degrees around the corner tracks?

Is there a way to rotate the heavy amado without lifting them? The carpenters back in the day surely would have scratched their heads over this question, too. Their solution: the tomawashi [lit. “door-turning”] technique. I was shown a demonstration in the north wing of the Shoenkan, which adjoins the Kikugetsu.

I do not know whether tomawashi was performed at the Phoenix Pavilion, but if it was, Wright undoubtedly would have been utterly astonished.