Navigation – Plan du site
II. Codes et espace politique

Encoding “The Capital” in Edo

Encoder « la Capitale» à Edo
Timon Screech
p. 71-96


Cet article se propose d’examiner la façon par laquelle Edo (l’actuelle Tôkyô) a été construite au sein de la plaine déserte de Musashi, et de quelle manière la capitale shôgunale est devenue le centre symbolique du pays au pont de Nihon-bashi. Nous montrerons également comment le plan d’Edo évoquait certains aspects cruciaux de l’ancienne capitale (Kyô ou Kyôto). Edo constituait une ville entièrement nouvelle et ses maîtres, les Tokugawa, pouvaient la façonner à leur guise. Dans la direction la plus importante, le secteur nord-est (la « porte des démons » ou kimon), ils choisirent de rassembler un certain nombre de bâtiments de façon à faire intentionnellement allusion à la capitale impériale. Il s’agissait là d’« encoder » de façon délibérée Kyôto au sein de la ville qui, après 1602, devint la capitale shôgunale. L’organisation d’Edo a déjà été considérablement étudiée, mais les points examinés ici n’ont jamais fait l’objet d’une analyse systématique.

Haut de page

Texte intégral

  • 1  This paper forms part of my book, Edo toshi keikaku no shigaku, Kôdansha, 2007. Some of the import (...)

1This paper will engage with the theme of “codes” and “encoding” by considering how Edo (modern Tôkyô), the shogunal capital during the Edo or Tokugawa period (1603-1867), a new city created out of a military encampment, built itself into a fine and dignified metropolis, worthy of the shogunal regime. I shall maintain that it did this by imbricating into its plan and structure aspects of the prior city, then called Miyako or Kyô, which both mean “capital” (modern Kyôto), home of the dairi (emperor). Although the planning of Edo has been intensively studied, this matter of urban borrowings from the older cultural heartland of Kamigata (modern Kansai) has not previously been comprehensively addressed.1

Defining Edo

  • 2  Since the creation of the Japanese state in the late 6th century, the home of the dairi was always (...)

2As is well known, when the Tokugawa, under their head Ieyasu (1542-1616) arrived in the Kantô Plain in 1590, there was little there. The venerable city of Kamakura was close, but had suffered during mediaeval times, and its most imposing structure, the Great Buddha Hall (Daibutsuden), had been washed away, leaving the statue, as it stands today, in the open air. Many of Kamakura’s finest temples were in poor repair.2

  • 3 Man’yô-shû, verses 3375 & 3376.
  • 4 Kokin waka-shû, verse 865.

3Although there was a small settlement at Edo, the Kantô Plain, and Musashi province (kuni) around it were largely wildernesses. Not only was little there, but the nothingness of Musashi was its most famous characteristic. The area’s central expanse was Musashino, where no indicates open, vacant country. The tradition of classical Japanese poetry (waka), where verses are often set in real places, underpinned much of Japan’s cultural topography. To invoke Musashino, in poetic terms, was to refer to all that lacked and failed to be civilized. Musashino stood in opposition to the refinement of Kyô. The locality already appears in Japan’s earliest anthology, the Man’yô-shû, probably compiled in the early 8th century, where the poets speak of its crying quail (kigisu) and the wild grasses (ukera-ga-hana).3 A verse in the first official waka anthology, the Kokinshû, of 905, written by Ariwara no Narihira (who will be met below) added references to its “purple grasses” (murasaki-gusa), or heather.4 Perhaps the locus classicus was by the poet Kujô Yoshitsune (1169-1206), anthologised in the subsequent New Kokinshû (Shin-kokin-shû) of 1205:

  • 5 Shin-kokin waka-shû, verse 422, in Shin Nihon koten bungaku taikei, Iwanami Shoten, 1992, vol. 11, (...)

At our destination,
The clouds come together as one
At Musashino.
From the field of grasses
Come forth the lunar shadows.5

  • 6  For a general discussion, see, Edward Kamens, Utamakura, Allusion, and Intertextuality in Traditio (...)

4Sites of poetic note (or notoriety), known as “poetic pillows” utamakura, dotted the landscape, colouring the sensations of real or imaginary travelers.6 Musashino could not escape the meaning accorded it by the waka tradition.

  • 7  See, Katsuhiko Kôzai, “Les protections religieuses de la capitale”, in Nicolas Fiévé (ed), Atlas h (...)

5It is true that some other sites, likewise designated as no (moors), especially those around the capital, were regarded not purely as negative spaces, but as liminal ones, at which it was possible to mediate with another world, religiously.7But Musashino was not accorded that capacity. It was too remote and too far to have gained any status among poetic travellers as a site of mediation.

6Poetic pillows were often rendered as scroll or screen paintings, and those of Musashino showed the birds and tall grasses evoked in verses, specifically, a large moon rising up behind. Most waka carry a clear season, and paintings of poetic sites retained this season. Musashino stood for autumn - a time of falling leaves and colder days, thus associated with sadness and loss. Such paintings would be displayed at the designated time (here, autumn), so as to evoke the appropriate seasonal sentiment (here, desolation).

7Such an aura was not ideal for the Tokugawa power-base, much less, after nomination of Ieyasu to the defunct title of shôgun in 1603, for the new shogunal capital.

  • 8  For a discussion of the etymology of Fuji, see, inter alia, Hirazawa Taka, “Edo meisho izen ni oke (...)

8Musashino had one other site to offer, not actually within its borders, but visible from it, in neighbouring Suruga. This was Mt Fuji. No classic poets had thought to link Musashino with Fuji, but once the Tokugawa had settled in the region, paintings of Musashino became much more common, and ateliers unilaterally added the mountain to the established poetic iconography, as a kind of compensation, or valorisation (pl. 12-13, p. 103). Fuji is not visible from Kyô, and thus could serve as a symbol of Edo as independent cultural space, just as the old depiction of Musashino symbolized it as culturally defective. Fuji was unrivalled, literally so, since authors punned on its name : fu can mean “not” (不) and ji “two” (二), hence “peerless”. Since ji could also mean “death” (死), the cone of Fuji offered Edo immortality, suggesting a longevity for the emergent Tokugawa regime.8

9Depictions of Fuji had been known since antiquity, but in Edo became widespread. By convention, an image of Fuji came to be displayed as part of the New Year decorations. It was popularly said (though why is unclear) one should dream of Fuji on New Year’s Day (on the second and third nights, one should dream of aubergines and hawks) to ensure a successful year (fig. 1). Since the lunar New Year marked the arrival of spring (unlike the solar one which falls mid-winter), pictures of Fuji denoted the return of warm weather, and once assigned to the Edo region (rather than to Suruga itself) the image of Fuji removed this part of Japan from the melancholy autumnal orbit into which waka had placed it, and associated it with the regenerative assertiveness of spring.

10The image of Fuji offered not only a means to promote the Edo region, but also one to invert the cultural hierarchy that placed Kyô above Edo, and specifically, within the frame of waka. One of the greatest works in the classical tradition was the poem-prose collection, the Tales of Ise (Ise monogatari) (c. 900). The protagonist, called only “the man” (otoko) but since ancient times associated with the courtier Ariwara no Narihira (825-880), travelled to the Musashi region. As he passed Fuji, he noted that it was much taller and grander than the finest mountain in Kyô – Mt Hiei, at the city’s north-east:

  • 9 Ise monogatari 9, in Shin Nihon koten bungaku taikei, Iwanami Shoten, 1997, vol. 17, p. 88. The tra (...)

If we compare that mountain [Fuji] with the one here [in Kyô], it is like as many as twenty Mt Hieis piled up together, and its shape is [as perfectly conical] as a salt cake.9

  • 10  For a discussion of the Saga editions, and their diffusion, see, Katagiri Yôichi, “Edo jidai no Is (...)

11This observation gave Edo pride of place. The Tales of Ise was often painted on narrative picture scrolls, and the work was published in the 1610s, as one of the widely-acclaimed Saga-bon editions (the first printed versions of several classic texts).10 Its illustrations standardised the iconography throughout, and here showed “the man”, with companions, passing under Fuji, heading from Suruga into Musashino (fig. 2). This image was extracted from the Tales and soon became an independent theme in itself, under the title of the Descent to the East (Azuma kudari).

Absorbing Fuji

  • 11  For this discussion, I am indebted to Yukio Lippit, The Birth of Japanese Painting History: Kano A (...)

12Paintings of Fuji in Japanese art – including those showing the Descent to the East – always showed the mountain from the Western side, that is, the Kyô side, as it would be seen by travellers from the capital. Among the countless depictions, by the Edo Period the one by Sesshû Tôyô (1420-1506) in the late 15th century had attained the status of most revered (fig. 3).11 Sesshû showed Fuji rising to the left, the temple of Seiken-ji below it, and the pine-covered split of Miho-no-Matsubara to the right.

13After Tokugawa Ieyasu died in retirement in 1616, one of the first acts of his son and successor, Hidetada (1579-1632), who now took full control of government, was to institute the office of Painter in Attendance (goyô-eshi). He nominated to this position the young but admired Kanô Tan’yû (1602-1674). Tan’yû haled from a formidable family of artists, and his grandfather, Eitoku (1543-1590), had decorated some of the greatest castles of the pre-Tokugawa Warring States period. Tan’yû had arrived in Edo, with some other members of his family, after the establishment of the shogunate. Tan’yû is much associated with depictions of Fuji, and particularly with variations on Sesshû’s formulation. Sesshû’s work was a hanging scroll, but as paired folding screens became the typical elite format in the mansions of Edo potentates, Tan’yû had to adapt the format, and he invented a construction with Fuji on the left-hand screen, and Miho-no-Matsubara on the right.

  • 12  What follows contradicts the view recently put forward by Beatrice Bodart-Bailey, “Urbanisation an (...)
  • 13  This elision is known to have occurred at Nihon-tsutsumi / zutsumi (“embankment”) in the vicinity (...)

14Tan’yû still followed the established norm, and showed Fuji from the Kyô side. There is no case where he represents it unambiguously from Musashino (though there may be lost instances). It was left to artists like the maker of Plates 12 and 13, probably rather after Tan’yû, to do that. But often he left the angle of vista deliberately unclear. Fuji was symmetrical until an eruption in 1706 caused a lump to appear on the northern, that is, right-hand side when viewed from Edo. If Miho-no-Matsubara and the Seiken-ji are omitted, or, conversely, if grasses, birds and the moon are left out, it was moot, before 1706, which side of the mountain was being shown. This permitted Edo, subtly, to take possession, and to absorb Fuji as truly its own, while at the same time retaining Sesshû’s canonical positioning (pl. 14, p. 104). The shogunate went further in the cooptation of Fuji. In 1603, the Tokugawa built an architectural scheme in celebration of their new shogunate, centring on a vista leading to the main gate of their castle.12 The focus of this installation, intriguingly, was a bridge, from which the vista led towards the castle. Although in time it would be surpassed, at the time of construction this was the largest bridge in the region, and perhaps the finest in all Japan. It seems the bridge was first known as the Two-Track Bridge, from its generous width. This is pronounced nihon-bashi (“two” ni, “track” hon and “bridge” hashi / bashi, 二本橋).13 But since nihon also means Japan, in another pun the name was re-written “Bridge of Japan” (日本橋). Standing on it, a person looked down a waterway, initially nameless, but later called the Nihon-bashi river, towards the castle. The morning sun would flood the shogunate’s gateways, and the evening sun would illuminate its towers from behind. Nihon-bashi was declared the centre of Edo, and by extension, of the whole realm, as its name seemed to declare. The main Kyô-Edo highway, the Tôkaidô, began here, winding up to Kyô, fifty-three stations, or about ten days’ walk ahead. From the other end of the bridge extended the Nikkô Dôchû highway, which, after burial of Ieyasu at Nikkô, in 1617, led to the most sacred of Tokugawa sites. Thus, standing on Nihon-bashi and looking towards the castle, one knew that southwards, or leftwards, ran the route of linkage with inherited civilization and to the imperial court, while to the right ran the route of linkage with shogunal commemoration.

  • 14  The Dutch East India Company began making trips to Edo in 1609, and these trips became annual from (...)
  • 15  The building was rebuild several times in the Edo period, but always on the same site. For a discu (...)
  • 16  Tsunoyama Sakae, Tokei no shakai-shi, Chûô Kôron-sha, 1984, p. 78.

15To the right of the bridge, that is, on what we might term the “Tokugawa” side, the shogunate began to position sites of political importance to itself, and specifically three. The gold monopoly (kinza) was established here, to issue coin. Near to it was built the hostel, or embassy, of the Dutch East India Company, called the Nagasaki House (Nagasaki-ya).14 By the 1630s the United Provinces (“Holland”) was the only foreign country with a permanent presence in Japan, and the Company used this building on its annual trips to Edo, when it would pay respects to the regime.15 Thirdly, and added in the period between the monopoly and the hostel, was Edo’s principal time-bell, donated by Tokugawa Hidetada to regulate the city.16 This accretion of emblematic structures occurred gradually, and there was probably no initial blueprint. But by the time Edo suffered its first disastrous fire in 1657, Nihon-bashi had become a centre of what we may call a Tokugawa civic iconography. After the fire, the scheme was rebuilt (it would be several times more, as conflagrations demanded), reinforcing the vista from Nihon-bashi as the shogunate’s prime iconic core.

16A person standing on the bridge looking westwards knew (but could not see) the three above-named institutions of Tokugawa power – for minting and monetary policy, the summoning of foreigners, and regulation of time. They would feel the world divide under their feet, as one end of the bridge led to Kyô and the other to Nikkô. Ahead, the vista was laid out so that the castle was not perfectly in the centre, but somewhat to the right. This device was in order to balance it with another soaring form to the left – the summit of Mt Fuji. The name for the area on the “Tokugawa” side of the bridge was Suruga-chô, in reference to the province containing the mountain, and binding Edo to it.

  • 17  There are, of course, exceptions, notably the screens of Edo (Edo-zu byôbu) of the Idemitsu Museum (...)

17The vista from Nihon-bashi was not depicted in the 17th century, for there was barely any practice of representing the formal spaces of the city.17 It was not until the latter part of the 18th century that such pictures emerge. One of the best examples dates from later, c. 1835, and is from Katsushika Hokusai’s (1760-1849) series, Thirty-six Views of Mt Fuji (Fugaku sanjûrokkei), entitled“Fuji from Nihon-bashi” (pl. 15, p. 104). Hokusai had mastered vanishing-point perspective, enabling him to extend the vista, making the Nihon-bashi river look much longer that it really is. Note that Hokusai also interrupted this perspectival scheme to elevate the castle and Fuji. Correctly drafted, he should have placed both as dots in the distance. Since this would have, literally, diminished them, he broke them out of the spatial grid, allowing them to float ethereally above.

  • 18  Hamada, Edo bungaku chimei jiten, q.v.

18Hokusai also depicts another bridge, built between the castle and Nihon-bashi, by his time known as Ichikoku-bashi, from its placename, but its formal name was Yatsu-hashi, or “eight-bridges”.18 That appellation was immediately recognisable as a reference to the Tales of Ise, where “the man” rests beside an eight-plank bridge (yatsu-hashi) immediately before beginning his “descent to the East”. A sense of turning point, between the old culture of Kyô, and the environment of Musashi, as recounted in the Tales, was therefore embedded into the vista too.

  • 19  For a convenient reproduction, see, Henry Smith (comm. and intro.), Hokusai, One Hundred Views of (...)

19This detachment of Fuji from Suruga and its absorption into Edo was practised on the ground at Nihon-bashi and Suruga-chô. Pictorial representations were able to do more, and fictionalised, or elaborate on the actual space. It is impossible to reconstruct exactly how the vista from Nihon-bashi would really have looked. Often Fuji would have been concealed in haze. The castle possibly did not rise so perfectly contrapuntally against the mountain. But artists could make it seem to. In his One Hundred Views of Mt Fuji Hokusai offered a “Fuji of Edo”, showing castle and peak in perfect unity, and he added a bird perched in the apex of the turret.19 Hokusai is being deferential: since birds are easily startled and fly away at any sound or movement, a bird at rest is a sign of enduring peace.

The Perfect City

  • 20  This subject has been much discussed, but see, Fiévé, “Une ville et son modèle: De Chang’an à Heia (...)

20The building of Nihon-bashi, and its careful incorporation of Mt Fuji, was not sufficient for Tokugawa needs. Across the Chinese cultural world, that is to say, in most of North and South-East Asia, there was a clear concept of how a city should look. The model had been set in the 5th century, when the Northern Wei dynasty built their capital at Luoyang.20 Ever after, it formed the proper model, and, with minor modifications, became the fixed and static norm. This perfect city should be built as a grid, with streets running north-south and east-west, with a main thoroughfare going through the centre, north-south, and dividing the city into two vertical blocks. This broad road would connect the palace, set in the north, to the main gate, set in the south. In the 8th century, Heijô-kyô (modern Nara), Japan’s first truly permanent capital, was codified like this (fig. 4).

21Heijô-kyô was followed, briefly, by Nagaoka-kyô, and then by the enduring Heian-kyô (modern Kyôto), from 794 to 1867, both similarly patterned on the same model.

22Cities were, and should be, visibly different from mere aggregations of residences and workspaces. As homes to a king, or emperor, capitals were intended to represent order out of chaos, and the perfection of human life under the gyration of the heavens. In contrast to the multitude of higgledy-piggledy towns across the countryside, there was only one royal residence, and there was only one capital. In Japanese this city was designated, as we have seen, as miyako or kyô. For most of history, including in Edo times, “Kyô” was used generically to denote the city more formally known as Heian-kyô, or Kyôto.

23This perfect capital could be compared to a human body. In the north, that is, at the top, was the controlling head (the palace), with its principal opening opposite, in the south, at the bottom (the gate). Like a body, the city was laterally symmetrical across the main thoroughfare, which could be thought of as the spine. Mansions, administrative offices and other institutions were ranged equally on either side; thus if there was an east market, there would be a west market too; if there was a temple in the east, there would also be one in the west.

24But a human body is not entirely symmetrical, as its most vital organ, the heart, is on the left. The city was ordered similarly. Though buildings and institutions were paired, what was on the left was deemed more vital than what was on the right. Throughout, sakyô (the left part of the capital) came before ukyô (the right part). The Ministry of the Left was superior to the Ministry of the Right, and the chief of government was the Great Minister of the Left (sadaijin), the Great Minister of the Right (udaijin) following second. This schema was conceived as viewed from the palace. For someone entering the city from the south, what is called “left” is therefore actually on the right (so too, ones heart is on the left to oneself, but on the right to others).

  • 21  There are numerous studies of this system, but for the best assessment of its role in Edo, see, Mu (...)
  • 22 See, Kuroda Hideo, Ô no shintai ô no shôzô, Tôkyô: Heibonsha, 1993, p. 18-54.

25This rational, geometric, but also metaphorical system, was enhanced by geomancy (onmyô-dô).21 The main north-south thoroughfare was called Suzaku-ôji, in reference to the geomantic contention that the southerly direction was governed by a “red bird” (suzaku). Coloured creatures defined all cardinal directions – north was the direction of the Dark Warrior, east of the Blue Dragon, and west of the White Tiger. Each of them was treated as the guardian of its respective cardinal direction. The colours had their significances in relation to the sun: red from the brightness of the south, and black because the north is dark; sunrise is blue, and a cloudy sunset is white. The colours also referred to the king or emperor for the ruler always sat in the north facing south, so that the north was dark because it was behind him; the south was red because the sun shone on his face.22

26Much more could be said about the regulation of the perfect city, or capital, though this must be saved for another occasion. One additional aspect, however, is crucial and must be addressed here. Geomancy held that spiritual forces (ki) pervaded the universe, and tended to move in a north-east to south-west direction. This meant ki entered the city from the north-east, which required that direction to be protected. Purifying institutions were erected to wipe incoming ki free of pollution. This gave rise to a second exception to the perfect symmetry of the city, created in the north-east. The north-east was known as the kimon, perhaps originally “gate of ki” (気門), but written as “gate of demons” (鬼門). The name emphasised the potentially lethal quality of that direction. Perfect cities had to be positioned with a hill to the north-east, and a temple complex was built on top of this, to debar bad ki and admit only the purified sort. Many perfect cities ended up with an large excretion to the north-east, compromising their grid layout, but ensuring all entering ki was clean.

27The north east was not dangerous in itself, but only because it offered ki ingress to the city. Again, this was like a human body, not imagined as standing frontally, but as looking east. The person rises at morning and looks towards the sunrise, his mouth turned to the north-east. Ki entered the city as matter enters the mouth, and as the mouth has ways to protect itself (lips, teeth, a tongue to taste, and, in extremis, an ability to vomit), the hill and temple afforded the city analogous protection.

  • 23  Otherwise, aku-basho, see, Hirosue Tamotsu, Aku-basho no hassô, published as Hirosue Tamotsu chosa (...)
  • 24  See, Takayanagi Kaneyoshi, Edo burakumin no seikatsu (Yûzankaku, 1981).

28Heijô-kyô had a lump to the north-east to accommodate the vast complex of its protecting temple, the Konmyô Shitennô Gokoku-ji, or popularly, the Tôdai-ji (“Great Eastern Temple”, actually, north-east, not east). Heian-kyô had a more significant eminence in that same direction, Mt Hiei, with the plethora of holy sanctuaries that made up the Enryaku-ji monastery. By Edo times, beyond the buffer formed by the hill and temple, city authorities placed the “vile places” (akusho) that every city needed, but which polluted civic space.23 These were the execution grounds, slaughterhouses, leatherworking and tile making workshops, and the homes of those forced to perform such jobs – the eta or hinin, today referred to as burakumin or “untouchables”.24 The early metropoles were less formally arranged, but when Edo was laid out, also with a north-easterly hill and temple, the “vile places” were concentrated in Senju, and after the fire of 1657, the official pleasure district, the Yoshiwara, was moved there too, as, from 1841, were the kabuki theatres.

29One associated geomantic requirement was that a river should run from north-east down to south-west, along the city’s eastern fringe. This was to further assist the flow of ki. Even purified ki, if impeded in its movement, becomes malignant. Heijô-kyô had two such streams, the major Saho and the smaller Higashibori; Heian-kyô had the Kamo; Edo had the Sumida. These rivers might be thought of as the city’s alimentary canal, running down from the point of intake at the mouth to an exit in the south-west, which could be likened to the anus. Sustenance enters, passes through and is digested, and waste is then excreted.

Perfection and Reality

30The capital was conceptualised in a way that was theoretical and so did not always work in practice. Japan’s longest-lasting imperial city soon diverged from its intended grid. Cities do not evolve in ordered ways. A river along the eastern side was not practical, for people naturally wished to live with access to water, thus there was a tendency to abandon the western sections and slide the city eastwards. Already in early times Heian-kyô had moved distinctly eastwards, and by Edo times, about one third was east of the river. Within a generation, the city of Edo also jumped the Sumida too.

31But a more serious shortfall became obvious in Japan during the 16th century. The perfect city was hopeless in wartime. Since everyone would look the same, an invading army could find its way, without plans or maps, to all places of administration and power. And to capture the palace, an enemy only needed to enter from the north, where the ruler sat exposed.

32During the mediaeval period, the palace of the Japanese ruler was moved to the centre of Kyô, so as to be more defendable. As warfare raged, the straight streets were broken and few, or none, ran the full length of the city. The body analogy, structured around a dichotomy of left and right, was abandoned, with the city split into “upper” and “lower” sections. In the late 16th century, the potentate Toyotomi Hideyoshi (1536-1598) erected a wall (odoi) around Kyô, in an attempt to hold it together. To all appearances, the capital degenerated into a jôka-machi, that is, a “castle town”. Jôka-machi were built primarily for defence, though many did pay lip service to the notion of perfection, with streets in a grid pattern, only that the grids were partial and came together in patchworks, lacking overall coherence, thus precluding assault on the castle, which would be lost somewhere in their middle.

  • 25  For further discussions of the cultural role of bridges, see, John Kieschnick, The Impact of Buddh (...)

33Edo was a castle-town, but from 1600, with the Tokugawa victory at the battle of Sekigahara, it was clear Edo would become especially important. In 1603, with declaration of the shogunate, Edo had to define itself differently. Centring the city on a bridge was a new idea, not part of any codified earlier civic layout, or what I have called an iconography. A bridge had a range of meanings. It stood for freedom of movement and an end to the bloodshed that had divided the country in the past. Yet, as history demonstrated, a bridge could also be removed in time of need.25

34Having the castle to the centre raised a problem. Those who lived north would look south on the castle and its occupants, and this was considered insolent. The rationale for placing the palace in the north, in the perfect city, was to allow the ruler to survey all, with the sun on his face. No inferior should look south on him.

35This conundrum was resolved in the Edo case, by placing direct Tokugawa retainers (hatamoto and gokenin) to the north. They constituted an extended shogunal family, or in-group, and, as servants, it was not inappropriate for them to protect their master from the rear. The northern part of Edo became a kind of honorary extension of the castle.

36The third shôgun, Tokugawa Iemitsu (1604-1651), promulgated the law of “alternate attendance” (sankin kôtai), requiring the 250-odd hereditary regional rulers (daimyô) to reside intermittently in Edo, and leave their wives and inheritors there permanently. Daimyô mansions in Edo were zoned with care. None was ever placed north of the castle. The loyalist (fudai) group who had fought with the Tokugawa at Sekigahara were allotted plots to the north-east of the castle, that is in the kimon. Their loyal virtue would further purify ki after it had flowed in from the hill and temple complex. By contrast, the suspect daimyô group (tozama), who had opposed the Tokugawa at Sekigahara, but were too powerful to be eliminated, were placed in the south-west, looked south on by the shôgun, and if ki became sullied as it passed over their mansions, it would soon flow harmlessly out of the city.

Encoding Kyô

37At Nihon-bashi, Edo was given an entirely new, sui generis symbolic core. Throughout the rest of the city, negotiations had to be made between perfection and necessity. Edo never pretended it was other than a castle-town, but the Tokugawa also sought to give it aspects of a kyô.

38There was also a third dynamic at work, for several of the sites and monuments of emerging Edo were specifically constructed to refer to, or match, those of Kyô, and thus to enhance Edo’s dignity. This was prominently done at the point of most fundamental geomantic importance, namely, the kimon. It so happened there was an ancient temple, established in the 7th century, more or less to the north-east of what became Edo, the Sensô-ji. Initially, it was used as the guardian of the north-east, but it did not satisfy geomantic requirements very specifically, as it was not located on a hill, and was more to the east-north-east than the north-east of the city. Furthermore, the Sensô-ji was dedicated to the Pure Land Buddhist school (Jôdo-shû), which failed to match the kimon expectations : the Enryaku-ji in Kyô was of the Tendai Buddhist school. In the 1620s, therefore, the shogunate built a new complex, denominationally Tendai and positioned more exactly north-east on a hill. The hill was Ueno, and the temple was named the Kan’ei-ji. To make the parallel with Kyô even clearer, the temple was given the sanmyô (“mountain name”, which all temples have) “Mt Hiei of the East” (Tôei-zan). This necessitated stripping that honorific title from the Kita-in temple in Kawagoe, which previously had been regarded as the Enryaku-ji equivalent, overall guardian of the Kantô, and prime Tendai establishment in the region – but was not in Edo.

39The name Kan’ei-ji is significant. “Kan’ei” is the name of the era 1624-1643, when the temple was founded. But it is highly unusual to name a temple after its foundational era. The precedent, of course, was the Enryaku-ji, named after the era of its foundation too, 782-805.

40Beyond Mt Hiei is Lake Biwa, Japan’s largest lake and a place of religious importance thanks to a shrine to Benzaiten on the islet of Chikubushima in its middle. There was already a small pool at the foot of Mt Ueno, and this was now dug out and turned into an “Eastern Lake Biwa”, although that name was not used, and it was known as Shinobazu Pond. A small island was created in the middle, and equipped with a shrine to Benzaiten (pl. 16, p. 105).

41The precincts of the Kan’ei-ji were used to replicate other landmarks from Kyô. Several strategies were attempted. Perhaps Kyô’s most splendid temple was the Kiyomizu-dera, known for its wonder-working image of the divinity Kannon, and for the sensational construction of its main hall, with a gantry that jutted out over the hillside. The Kiyomizu-dera traced its history to Heian times (794-1185). It had burned down in 1469, but was rebuilt fifteen years later, only to burn again in 1629. In the early Kan’ei period, when Ueno was laid out, the Kiyomizu-dera of Kyô was no longer there. On this occasion, rather than fabricating Edo as an Eastern Kyô, it proposed itself as Kyô’s successor. In 1631, the shogunate paid for a building of a Kiyomizu-dera in Edo, in the grounds of the Kan’ei-ji, although in deference to its lost original, it was built on a smaller scale, and modestly called the Kiyomizu Hall (Kiyomizu-dô). But it looked like the original, with a gantry jutting over the hill (pl. 17, p. 105). Only when the Edo version was nearing completion did the shogunate offer funds to rebuild the original Kiyomizu-dera, which was completed in 1633, and survives today, slightly later than the Edo one (also extant).

42Probably the second most astounding temple in Kyô was the Sanjûsangen-dô (“Hall of thirty-three bays”) at the Rengeô-in, in the south-east of the city. It had been built originally in 1164 and still houses 1001 life-sized statues of Senju (“Thousand hands”) Kannon bodhisattva made by the sculptor Tankei (1173-1256) and his atelier. Of course, it had to be very large, or more particularly, very long.

  • 26  See, Smith (ed), Hiroshige, One Hundred Famous Views of Edo, New York: Braziller, 1986, n 69.

43A version of the Sanjûsangen-dô was built in Edo in 1642, not in the Kan’ei-ji precincts, but close to the Sensô-ji monastery, thus still part of the larger kimon district; it was also built with private funds, not by the shogunate. But the Edo building was the same length as the original, over a hundred metres, though only half as wide, and contained but one Senju Kannon. Not only the hall, but the rituals associated with its Kyô precedent were matched in Edo too. The original building was known for an archery contest held annually, the objective being to shoot an arrow the entire length, difficult to achieve because of the protruding rafters. It was performed also at the Edo site, and indeed, it is said to have been an archery master who petitioned the shogunate to build the Edo Sanjûsangen-dô in the first place.26 In Kyô, the competition was held at the back of the building, since this was not the structure’s real purpose, But in Edo, it was held at the front, since shooting was the motivation for erection.

44In 1644, the Edo hall was moved, supposedly because the archery master could not pay the merchant who supplied the wood. This merchant took over ownership and moved the building to Fukagawa, on the other side of the Sumida. In 1698, it burned down, but was rebuilt at another location, near the Eitai-ji. Hiroshige depicted this third manifestation (fig. 5). The hall was destroyed by a storm in 1872, and not rebuilt.

Other Sites

  • 27  A man approaches holding a huge rice spoon, and, as the caption above the artist’s signature state (...)

45Other Kyô sites were replicated too. Edo had a hill that was constructed to match Mt Atago, which overlooked Kyô. If Mt Atago was of dramatic height, the Edo hill was only some thirty metres, though it did offer excellent views of the city and bay beyond. The Edo hill was named Mt Atago. Hiroshige depicted it too, showing a ritual being performed, although one specific to Edo, and not copied from Kyô (fig. 6).27

  • 28  Hamada, Edo bungaku chimei jiten, q.v.

46Kyô was the centre of cultural and sacred authority, and Edo naturally drew its references from there. But there is at least one instance of an Ôsaka site being recreated in Edo. This is of a more mercantile nature, and was affected at two islets in the mouth of the Sumida. They were built up and made into homes for Edo’s fishermen and their families. The facts are unclear, but the first Tokugawa shôgun, Ieyasu, is said to have brought fishermen up to Edo, to practice in his new city the skills they were known for at Tsukuda-mura, an island at the mouth of the River Kanzaki, at Ôsaka.28The Edo site was named Tsukuda-jima. Because the character tsukuda is rare, the connection would have been immediately apparent. In 1644, the island was expanded to a square of almost two hundred metres each side, and by the early 18th century had become home to almost two hundred people.

47Tsukuda-mura was known for a shrine to Sumiyoshi Daimyôjin. An equivalent was therefore built on Tsukuda-jima in Edo. The Edo shrine did not replicate the architectural structure of the original, but only bore its name. A Sumiyoshi festival was held there every three years, on 29th day of 6th month, which was said to be the anniversary of the Ôsaka fishermen’s move to Edo.

  • 29 Hakurakuten, in Nihon koten bungaku taikei, Iwanami, 1994, vol. 41. For a translation, see, Paul Wa (...)

48Sumiyoshi Daimyôjin was the god of fishermen, but also of literature. This rather curious coupling is accounted for by an old story, best known from the drama Hakurakuten.29 The anonymous play tells how the Chinese poet Hakurakuten, that is, Bai Juyi (772-846), of the Tang dynasty (618-907), came to what was then called Sumiyoshi Bay, to test the Japanese skill at verse. If the Japanese were not good, the Chinese emperor would invade and conquer them. While still in the boat, Hakurakuten held a poetry competition with the first Japanese he saw, who happened to be a fisherman. In fact, it was the God of Sumiyoshi, who had taken the form of a fisherman to protect Japan’s honour. The verse was excellent, and Hakurakuten was overcome with admiration. In the play, a kamikaze (“divine wind”) rises and expels the poet back to the Tang. A chorus ends the performance,

How true are the customs of Yamato!
How true are the customs of Yamato!
The fisherman has a heart that knows this.
Truly this is beneficial learning!

49Set at the sea door of Edo, this little island recalled the encounter of Japan with Continental cultural norms, in an event where Japanese cultural autonomy had been clearly stated. It was useful to have fishermen live by the mouth of the bay, but the location of them and their shrine also indicated Japan’s scorn for foreign cultural arrogance, and capacity to deal anyone who might contemplate invasion.

50The single largest building in Japan, and the most visited temple in Kyô has not yet been mentioned. The Kiyomizu-dera and the Sanjûsangen-dô were sensational to behold, but pre-eminent in Kyô was the Great Buddha Hall, built by Toyotomi Hideyoshi as part of the Hôkô-ji temple, founded in 1586. The progenitor for such a structure was, of course, the Great Buddha Hall at the Tôdai-ji in Heijô-kyô, although again, the ultimate precedent was Luoyang. A Great Buddha had been built at Kamakura, in 1252, when it was the capital of the first shogunate, which offered a challenge to the new shogunate at Edo.

  • 30  Oddly, this image has been the object of no scholarly attention at all, but it can be seen in this (...)

51A Great Buddha was therefore constructed at the Kan’ei-ji. Little is known for sure about the negotiations that went into its building, or even about its form. Curiously, the Great Buddha of Edo was seldom illustrated or written of, though we know it was erected in the 1630s. Such pictorial evidence as exists, suggests it was close to the Edo branch of the shrine to the deified Tokugawa Ieyasu at Nikkô, the Tôshô-sha (later Tôshô-gû).30

  • 31  See, Screech, The Lens within the Heart, p. 242-247.

52It was not until 1793 that some enterprising Edoites decided to create a Great Buddha to be a popular focal point. It was built at the Kaian-ji, a Zen temple founded in 1251 at Shinagawa, in the south-west, and unrelated to the shogunate’s grand scheme for the city. Unlike other colossi, which were built at government expense, funds were private, and the image was made of cheap basketwork coated in resin (kappa). Apparently, pictures were ever made of this Buddha, though there are stories and verses.31 Proximity to the populace was made clear in the nomenclature, for the characters for “Great Buddha” were not read in the Sino-Japanese way as daibutsu, but in the vernacular, ôbotoke. The image is, amazingly, said to have been forty metres tall, which makes it twice as high as any other (Nara is 16 m, Kamakura 11.5, and Kyô was 20 m). People regularly crossed Edo Bay to enjoy views of the city, sometimes using telescopes, and now they could also see an icon rising up, as shown by Hokusai, though in a fictional work, where the image is not said to be the Kaian-ji ôbotoke, though that is clearly the real referent (fig. 7).

53The popularity of this Great Buddha was too much for Edo’s Magistrate of Temples and Shines (jisha bugyô) who demanded it be pulled down. The ôbotoke had lasted only sixty days. Five years later, in 1798, Kyô’s Great Buddha was burned to nothing too.


54The planning of Edo has been studied before, and forms a central feature of the discipline of Edo-Tôkyô Studies. There have also been previous scholarly accounts of the creation of Japan’s older capitals, which became known as Nara and Kyôto. But a concerted consideration of how these earlier capitals were themselves used in the building of Edo has never been fully addressed. This paper attempts to open some lines of enquiry, and so ask how Kyôto, specifically, was “encoded” within the political fabric, and also within the literary memory, of a new and developing shogunal metropolis, that became today’s Tôkyô.

Haut de page



aku-basho 悪場所

akusho 悪所

Ariwara no Narihira (825-880) 在原業平

Atago (Mount) 愛宕山

Azuma kudari 東下り

Bai Juyi (772-846) 白居易

Benzaiten 弁財天

Bishamon-ten 毘沙門天

Biwa (Lake) 琵琶湖

burakumin 部落民

Chikubushima (islet) 竹生島

daibutsu 大仏

Daibutsuden 大仏殿

daimyô 大名

dairi 内裏


Edo meisho (series) 江戸名所

Edo-zu byôbu 江戸図屏風

Eitai-ji (temple) 永代寺

Enryaku (era) 延暦

Enryaku-ji (temple) 延暦寺

eta えた

Fugaku sanjûrokkei 富嶽三十六景

Fuji (Mount) 富士山

fudai daimyô 譜代大名

Fukagawa 深川

gokenin 御家人

goyô-eshi 御用絵師

Hakurakuten 白楽天

hatamoto 旗本

Heian-kyô 平安京

Heijô-kyô 平城京

Heike monogatari 平家物語

Hiei (Mount) 比叡山

Higashibori (river) 東堀川

hinin 非人

Hôkô-ji (temple)方向寺

Ichikoku-bashi 一石橋

Ise monogatari 伊勢物語

Isoda Koryûsai (fl. 1764-1781) 磯田湖竜斎

jisha bugyô 寺社奉行

Jôdo (school) 浄土宗


kabuki 歌舞伎

Kaian-ji (temple) 海安寺

Kamakura 鎌倉

kamikaze 神風

Kamo (river) 鴨川

Kan’ei (era) 寛永

Kan’ei-ji (temple) 寛永寺

Kannon 観音

Kanô Eitoku (1543-1590) 狩野永徳

Kanô Tan’yû (1602-1674) 狩野探幽

Kansai (region) 関西

Kantô (région) 関東

Kanzaki (river) 神崎川

kappa 合羽

Katsushika Hokusai (1760-1849) 葛飾北斎

Kawagoe 川越

ki 気

kigisu 雉子


Kita-in (temple) 喜多院

Kiyomizu-dera (temple) 清水寺

Kiyomizu-dô (temple) 清水堂

Kokin (waka-)shû 古今(和歌)集

Konmyô Shitennô Gokoku-ji (temple) 金妙四天王護国寺

Kujô Yoshitsune (1169-1206) 九条良経

kuni 国

kyô 京

Kyôto 京都

Luoyang 洛陽

Man’yô-shû 万葉集

Miho-no-Matsubara 三保の松原

miyako 京

murasaki-gusa 紫草

Muromachi 室町

Musashi (province) 武蔵


Nagaoka-kyô 長岡京

Nagasaki-ya 長崎屋

Nanasato fûki 七々里富貴

Nara 奈良

Nikkô 日光

Nikkô Dôchû (highway) 日光道中

no 野

nô 能

ôbotoke 大仏

odoi 御土居

onmyô-dô 陰陽道

Ôsaka 大阪

otoko 男

raku 洛

Rengeô-in (temple)蓮華王院

sadaijin 左大臣

Saga-bon 嵯峨本

Saho (river)佐保川

sakyô 左京

Sanjûsangen-dô (temple)三十三間堂

sankin kôtai 参勤交代

sanmyô 山名

Seiken-ji (temple) 清見寺

Sekigahara (battle) 関ヶ原

Sensô-ji (temple) 浅草寺

Senju 千住

Senju Kannon (divinity) 千手観音

Sesshû Tôyô (1420-1506) 雪舟等楊

Shinagawa 品川

Shin-kokin (waka)-shû 新古今(和歌)集

Shinobazu (Pond) 不忍池

shôgun 将軍

Sumida (river) 隅田川

Sumiyoshi (bay) 住吉

Sumiyoshi Daimyôjin住吉大明神

Suruga (province) 駿河

Suruga-chô 駿河町

suzaku 朱雀

Suzaku-ôji 朱雀大路

Tang (dynasty) 唐

Tankei (1173-1256) 湛慶

Tendai (school) 天台宗

Tôdai-ji (temple) 東大寺

Tôei-zan 東叡山

Tôkaidô (highway) 東海道

Tokugawa 徳川

Tokugawa Ieyasu (1542-1616) 徳川家康

Tokugawa Hidetada (1579-1632) 徳川秀忠

Tokugawa Iemitsu (1604-1651) 徳川家光

Tôshô-gû (shrine) 東照宮

Tôshô-sha (shrine) 東照社

Toyotomi Hideyoshi (1536-1598) 豊臣秀吉

tozama daimyô 外様大名

Tsukuda-jima (islets) 佃島

Tsukuda-mura (island) 佃村

Ueno (hill) 上野

Uji (bridge) 宇治

ukera-ga-hana 朮が花

ukyô 右京

udaijin 右大臣

Utagawa Hiroshige (1797-1858) 歌川広重


waka 和歌

Wei (dynasty)魏

Yatsu-hashi 八つ橋

Yoshiwara 吉原

Haut de page


1  This paper forms part of my book, Edo toshi keikaku no shigaku, Kôdansha, 2007. Some of the important and readily-accessible works on Edo city planning are William Coaldrake, “Edo Architecture and Tokugawa Law”, in Monumenta Nipponica 36, 1981, p. 235-84; Nicolas Fiévé & Paul Waley (eds), Japan Capitals in Historical Perspective: Place, Power and Memory in Kyoto, Edo and Tokyo, London: Curzon, 2003; Hidenobu Jinnai, “The Spatial Structure of Edo”, in Chie Nakane & Shinzaburô Oishi (eds), Tokugawa Japan: The Social and Economic Antecedents of Modern Japan, Tôkyô, 1992; Noël Nouët, The Shogun’s City: A History of Tokyo, Paul Norbury Publications, 1990; Gilbert Rozman, Urban Networks in Ch’ing China and Tokugawa Japan, Princeton, 1973.

2  Since the creation of the Japanese state in the late 6th century, the home of the dairi was always in the Kansai region: in the 8th century, it was in Heijô-kyô (Nara), and from 794 to 1867 in Heian-kyô (later, Miyako or Kyô, now Kyôto). From the 13th century and until the end of Edo period (mid-19th century), Japan was ruled by a shôgun, that is a military dictator, whose government was established successively in Kamakura (Kantô), Muromachi (Kansai) and Edo (Kantô), and who technically deferred to the dairi and was appointed by him. Thus, there were always two centres of authority and power, one at the court of the dairi and one at that of the shôgun.

3 Man’yô-shû, verses 3375 & 3376.

4 Kokin waka-shû, verse 865.

5 Shin-kokin waka-shû, verse 422, in Shin Nihon koten bungaku taikei, Iwanami Shoten, 1992, vol. 11, p. 132. The translation is my own.

6  For a general discussion, see, Edward Kamens, Utamakura, Allusion, and Intertextuality in Traditional Japanese Poetry, New Haven: Yale University Press, 1997.

7  See, Katsuhiko Kôzai, “Les protections religieuses de la capitale”, in Nicolas Fiévé (ed), Atlas historique de Kyôto, forthcoming.

8  For a discussion of the etymology of Fuji, see, inter alia, Hirazawa Taka, “Edo meisho izen ni okeru meisho kannen no hatten to bunka-teki keikan”, in Nara kokuritsu bunkazai kenkyû-jo gakuhô kenkyû kiyô 58 (1999), p. 76.

9 Ise monogatari 9, in Shin Nihon koten bungaku taikei, Iwanami Shoten, 1997, vol. 17, p. 88. The translation is my own.

10  For a discussion of the Saga editions, and their diffusion, see, Katagiri Yôichi, “Edo jidai no Ise monogatari”, in Gotô Museum (ed), Ise monogatari no sekai, Gotô Bijutsukan, 1994, p. 171-176.

11  For this discussion, I am indebted to Yukio Lippit, The Birth of Japanese Painting History: Kano Artists, Authors and Authenticators of the Seventeenth Century (unpub. Ph.D thesis, Princeton University, 2003), p. 101-83.

12  What follows contradicts the view recently put forward by Beatrice Bodart-Bailey, “Urbanisation and the Nature of Tokugawa Hegemony”, where she writes that at Nihon-bashi, “no attempt was made to express political authority in spatial design”, in Fiévé & Waley (eds), op. cit., p. 115.

13  This elision is known to have occurred at Nihon-tsutsumi / zutsumi (“embankment”) in the vicinity of the Yoshiwara district, see, Hamada Giichirô et al. (eds), Edo bungaku chimei jiten, 2nd ed., Tôkyô-dô, 1973, q.v.; and Nishiyama Matsunosuke et al. (eds), Edo-gaku jiten, 2nd ed., Kôbunsha, 1994, q.v.

14  The Dutch East India Company began making trips to Edo in 1609, and these trips became annual from 1633. It is not known when the Nagasaki-ya was first built. See, Hamada, Edo bungaku chimei jiten, q.v.

15  The building was rebuild several times in the Edo period, but always on the same site. For a discussion, see, Timon Screech, The Lens within the Heart: The Western Scientific Gaze and Popular Imagery in Later Edo Japan, London: Curzon, 2000, 2nd ed., p. 15-21.

16  Tsunoyama Sakae, Tokei no shakai-shi, Chûô Kôron-sha, 1984, p. 78.

17  There are, of course, exceptions, notably the screens of Edo (Edo-zu byôbu) of the Idemitsu Museum, and the National Museum of Japanese History.

18  Hamada, Edo bungaku chimei jiten, q.v.

19  For a convenient reproduction, see, Henry Smith (comm. and intro.), Hokusai, One Hundred Views of Mt Fuji, New York: Braziller, 1988, p. 152.

20  This subject has been much discussed, but see, Fiévé, “Une ville et son modèle: De Chang’an à Heiankyô”, in his Atlas historique, forthcoming, & Takahashi Yasuo et al. (eds), Zushû Nihon toshishi, Tôkyô University Press, 1993, p. 1-56.

21  There are numerous studies of this system, but for the best assessment of its role in Edo, see, Murayama Shûichi et al. (eds), Onmyô-dô sôsho, vol. 3: Kinsei, Meichô shuppan, 1992.

22 See, Kuroda Hideo, Ô no shintai ô no shôzô, Tôkyô: Heibonsha, 1993, p. 18-54.

23  Otherwise, aku-basho, see, Hirosue Tamotsu, Aku-basho no hassô, published as Hirosue Tamotsu chosaku-shû, vol. 6, Eishobô, 1997.

24  See, Takayanagi Kaneyoshi, Edo burakumin no seikatsu (Yûzankaku, 1981).

25  For further discussions of the cultural role of bridges, see, John Kieschnick, The Impact of Buddhism on Chinese Material Culture, Princeton: Princeton University Press, 2003, and Edo no bigaku 28, 1988. The famous battle affected by such destruction was the Battle of Uji Bridge in 1180, when the bridge-planks had been ripped up to prevent crossing, as told in the Tales of the Heike (Heike monogatari).

26  See, Smith (ed), Hiroshige, One Hundred Famous Views of Edo, New York: Braziller, 1986, n 69.

27  A man approaches holding a huge rice spoon, and, as the caption above the artist’s signature states, he is the messenger of the divinity Bishamon-ten. This ritual was performed on the 3rd day of the New Year (the same day when one should dream of hawks). Hiroshige included the year Ansei 4 (1857) in the top left, as if written on a votive plaque.

28  Hamada, Edo bungaku chimei jiten, q.v.

29 Hakurakuten, in Nihon koten bungaku taikei, Iwanami, 1994, vol. 41. For a translation, see, Paul Waley, The No Plays of Japan, London: George Allen & Unwin, 1921. The translation that follows is my own.

30  Oddly, this image has been the object of no scholarly attention at all, but it can be seen in this location in the anonymous Edo-zu byôbu folding screen (mid-17th c.) in the National Museum of History, Sakura, right element, leaf 4. For a convenient reproduction, see, Suzuki Jun (ed), Edo-zu byôbu, Heibonsha, 1971, pl. 19. Today, the face only of the image is to be seen, on a mound in Ueno Park.

31  See, Screech, The Lens within the Heart, p. 242-247.

Haut de page

Pour citer cet article

Référence électronique

Timon Screech, « Encoding “The Capital” in Edo », Extrême-Orient Extrême-Occident [En ligne], 30 | 2008, mis en ligne le 01 octobre 2011, consulté le 24 avril 2017. URL : ; DOI : 10.4000/extremeorient.100

Haut de page


Timon Screech

Il est professeur à la SOAS (School of Oriental and African Studies) de l’Université de Londres. Spécialiste de l’histoire de l’art japonais. Professor at the University of London (SOAS), in Japanese history of art. A notamment publié: Sex and the Floating World: Erotic Images in Japan, 1700- 1820, Reaktion, 1999; Japan Extolled and Decried: Carl Peter Thunberg and the Shogun’s Realm, 1775-1796, Routledge, 2005; Secret Memoirs of the Shoguns: Isaac Titsingh and Japan, 1779-1822, Routledge, 2006.

Haut de page

Droits d’auteur


Haut de page