How Disasters Made the Modern City of Tokyo

Natural Conditions and Disaster Prevention in Edo

Let us first establish the relationship between disasters and the formation of the city of Edo.

Founded as the castle town of the Tokugawa family at the end of the sixteenth century, Edo was located at the point where the rivers of the Kantō plain flow into the bay, well suited to formation of a large city in an age when sea and river transport of goods predominated. The Musashino plateau, at an elevation of twenty to thirty meters above sea level, spread toward the west, incised with multiple valleys. The Tokugawa built their castle on a promontory at the eastern tip of the plateau. The Arakawa River flowed into the Sumida River and the Tone flowed into the Edo River, flowing through the lowlands to the east of the castle and into the bay.

In order to use this area to the east of the capital as city land that would be convenient to boat traffic, it was necessary to create regular waterways and to take measures against flooding. After the city suffered flood damage in 1619, the following year two important disaster prevention projects were launched. First, the Kanda River, which flowed through a valley in the Musashino plateau passing immediately to the northeast of the castle and directly to the bay, was redirected to the east by cutting a channel through the Hongō promontory and draining it into the Sumida. Second, two elevated embankments were built on either side of the Sumida to the north of the city, called the Nihon Embankment and the Bokutei Embankment. The rechanneled Kanda River became part of the outer moat of Edo Castle, so this project had the additional function of military defense, but the practical result was that streets in the immediate vicinity of the capital were protected from flood damage. One of the elevated embankments along the Sumida ran at a diagonal, preventing floods upstream from reaching the urban area; at the same time, in combination with the northern bank, it constrained the volume of water flowing into the downstream Sumida and protected the eastern portion of the city from floods. The daimyo lords were obligated to fund these projects in exchange for being granted rights to the tax revenue of the fiefs they had been apportioned by the Tokugawa house (see Figure 1). ³

OPEN IN VIEWER

Figure 1.

Greater Tokyo, showing major waterways and flood mitigation projects.

Initially, the daimyo were responsible for providing firefighters, even in the case of large conflagrations. However, after the Meireki Fire of 1657, which destroyed 60 percent of the city and cost sixty or seventy thousand lives, the shogunate organized a firefighting force under its own direct jurisdiction. ⁴ By the end of the seventeenth century, that force had reached over two thousand, divided into fifteen brigades. In addition, after the Meireki Fire, the shogunate restructured the city plan. To prevent fire from spreading to the castle, the estates of Tokugawa collateral families within the western castle vicinity were moved outside of the outer moat and a garden was built in their place to serve as a firebreak. Many of the daimyo had been given land adjacent to the castle for their Edo estates; these were now moved to locations at a distance from the castle and the daimyo were given additional land in the suburbs to serve as places of refuge in the event of fire. Buddhist temples and some of the commoners who had occupied land inside the perimeter of the outer moat were moved out, and the Yoshiwara licensed prostitution quarter was moved to a location on the south flank of the Nihon Embankment. In the built-up area, the shogunate created firebreaks, fire embankments, and wide avenues to limit fires from spreading. ⁵ The shogunate also built a bridge over the Sumida, opening the east side of the river to be laid out for further development. ⁶ The fire embankments were later lost and some of the firebreaks filled with new construction, ⁷ but a cityscape with evacuation sites and some of the features for fire prevention survived, so that although large conflagrations continued, none resulted in the massive loss of life of the Meireki Fire.

Commoners were also ordered to contribute to firefighting after the Meireki Fire. Over time, the shogunate’s own force diminished in size. After another major conflagration in 1717, instead of expanding its own force, the shogunate pursued reorganization of the brigades funded by commoners. As the city prospered, the shogunate increasingly relied on the economic power of commoners. In the mid-eighteenth century, the commoners’ brigades, which now constituted a force of over ten thousand, came to participate in firefighting at the castle itself, while the daimyo brigades were reduced to protecting their own estates and making purely ceremonial expressions of support. ⁸

Edo residents lived on the assumption that they would lose their houses in large conflagrations. Fireproofing was done by applying mud wattle and plaster to the walls of wooden structures. This was done for defensive structures around the castle, some shops, and small storehouses. The storehouses, built with thick mud and plaster walls, were referred to as dozō, and the technique of fireproofing as dozō construction. Daimyo houses and wealthy commoners kept their valuables in these buildings, while retaining stores of lumber at their suburban evacuation sites in preparation for rebuilding if the rest of their property burned. Many households also built underground storage for valuables.

In the Ansei Earthquake of 1855, fire destroyed 1.5 square kilometers of the city, in addition to causing widespread building collapse. Following this earthquake, underground storage was lauded and dozō structures condemned as useless. ⁹ This is because these buildings lost their effectiveness against fire if even a part of the mud walls were damaged by the earthquake. Clay tile roofing, which had also been encouraged as a means to limit conflagrations caused by flying embers, also resulted in casualties when it fell in the earthquake. The experience of the Ansei Earthquake thus revealed the limits of the protection provided by the existing building technology designed for fireproofing.

The Modern Water System and Policing

The first global cholera pandemic reached Japan in 1822 but did not enter Edo. ¹⁰ However, the disease arrived in Nagasaki on the U.S. Mississippi in 1858, and this time spread as far as Edo, causing between thirty and forty thousand deaths in the city alone. ¹¹ The opening of Japanese ports thus brought cholera to Edo. In Europe, cholera outbreaks in 1832 and 1849 led to development of piped water and sewage systems in Paris, while the outbreaks of 1848-1849 and 1854 provided the impetus for construction of London’s sewer system. ¹² No major response was evident in Edo, however. Since the human waste from latrines was collected for use as fertilizer, the sewers were simple open ditches carrying washing water and rain and their reform was not considered a significant issue.

A modern water system that delivers spring water or purified water under pressure using metal pipes was effective both for measures against cholera and for firefighting. The Meiji government, which took control in Edo and changed the city’s name to Tokyo in 1868, hired Dutch engineer Cornelis Johannes van Doorn, who submitted a proposal in 1874 for the installation of a modern water system for both fire prevention and sanitation purposes. Takayose Shōzō points out that if the government had invested the enormous sum it spent to rebuild the Ginza district out of brick after the fire of 1872 in a water system instead, the capital could have eliminated water-borne epidemics and remained within budget. ¹³ However, the need for a modern water system was not self-evident to either government leaders or residents at the time.

After the Meiji Restoration, the next cholera outbreak to cause high casualties came in 1877, when 8,027 died nationwide and 459 died in Tokyo. The Home Ministry’s Bureau of Hygiene offered the assessment that “in Tokyo it is hard to call this an epidemic.” ¹⁴ In the outbreak of 1879, 105,786 died nationwide, while deaths in Tokyo were limited to 1,791. ¹⁵ Viewed within the country as a whole, cholera casualties in Tokyo were thus relatively slight during the 1870s, and responding to the disease was not felt to be a matter of great urgency. Firefighting at the time depended primarily on the work of the old commoner brigades, who limited the spread of fires by tearing off roofing and, where necessary, pulling down buildings. They used water when they could but did not rely on it. ¹⁶ It was only in the 1880s that the difficult battle with epidemic disease would make Tokyoites feel the need for a modern water system and, at the same time, that the use of water would come genuinely to be a part of firefighting.

The primary role in this transformation was played by the Police Agency, which was a branch of the national government and not the product of autonomous action on the part of urban citizens. ¹⁷ Tokyo’s police system began in 1871 with an organization of former samurai under Tokyo Prefecture in charge of maintaining the peace. The police were reorganized into the Tokyo Police Agency under the national Home Ministry in 1874. Following the French model, the department bore responsibility for sanitation and firefighting in addition to police administration. In the early period, it also bore the important task of sending armed police brigades to quell uprisings around the country, but after suppression of the Satsuma Rebellion in 1877, the largest of the rebellions of former samurai, by the army together with the police, the police force focused its activities in the capital.

Having taken the former commoner fire brigades under its control, the Police Agency reduced the number of firefighters and had some of them use new hand pumps of French design. Unlike the pumps used in Edo, these had hoses for drawing and directing the stream of water and were thus seen as signs of modern civilization. However, in 1879 the city experienced its first fire since the Restoration to destroy over ten thousand houses, leading the department to leave the commoner brigades to fend for themselves and create a new firefighting force under its direct control, whose members were trained in the use of pumps. In the winter of 1880-1881, with both of these forces operating at once, the city saw the greatest fire damage of any year since its renaming. Four major fires in that season destroyed 22,000 dwellings, burning 0.86 square kilometers of densely built blocks. This was the worst fire damage Tokyo would experience until the 1923 Kantō Earthquake. The city had managed to train the new force based on translated manuals for the hand pumps, but that force had had little experience against Tokyo fires, and its contribution was not sufficient to overcome the opposition of the old brigades to new competitors, hindering the firefighting ability of the force as a whole. ¹⁸

The Police Agency had launched the new fire brigades without consulting the newly established Prefectural Assembly. Seeing what had happened in the fires of 1880-1881, the Assembly refused funding to the new brigades in 1881. In response, the police eliminated the new brigades for over a year, created a city fire department devoted to directing the old brigades, and embarked on a reform of their firefighting practice. From 1884, they introduced small pumps operated by six men, built catch basins into the old gravity-based Edo water system in order to improve water supply for the pumps, and organized the old brigades into a structure that used the new pumps. The Prefectural Assembly welcomed these reforms and elected to purchase a steam pump. A specialized staff was appointed to handle steam pumps in 1887, establishing a system in which members of the traditional brigades ran first to the fire with hand pumps, while a horse-drawn steam pump raced to the scene as back-up in the event they were unable to extinguish the fire. With this, for the first time it was evident to anyone that dousing of the flames was now the central method of firefighting. ¹⁹ The Police Agency had introduced the new technology and the Prefectural Assembly, while criticizing its initial failure, agreed on the point of modernizing firefighting through improved technology.

Following the series of fires in 1880-1881, the Tokyo governor and the chief of police ordered that along major roads and waterways in the central part of the city, which at the time consisted of the area within the outer moat to the east of the castle, all houses were to be built of brick, stone, or dozō construction, and were to be roofed with tiles or other fire-resistant material. ²⁰ This rule was applied not only to new construction after fires but to existing buildings as well. The police executed the plan in collaboration with prefectural authorities until April 1886, when the rebuilding was largely accomplished and sole jurisdiction passed to the prefecture. ²¹ At the time, Tokyo Prefecture had a total of only 1,450 employees, while the Police Agency had a staff of 3,449, which gave the police a large role in directing residents to rebuild. Owners were threatened with demolition if they did not meet the new construction requirements within five years. ²²

It was epidemic disease more than fire that ultimately led to construction of a modern piped water system, although the water system would contribute to firefighting as well. Based on the belief that decomposition of a cholera patient’s excreta produced the contagion, instructions accompanying the Home Ministry’s Cholera Prevention Law of 1877 ordered disinfection with carbolic acid of all places where infected excreta or the contagion were thought to have been found and transport to an isolation ward of all victims who could not be properly quarantined for convalescence at home. The instructions further made it the role of regional offices of the state to enforce this policy on a poorly educated population, explaining “Since this method of prevention has been proven effective in all countries applying it, it is not sufficient to leave it to the care of individual citizens to carry it out. Protection requires compelling people to carry it out through either encouragement or threats.” ²³ In Tokyo, the prefectural government and the police took responsibility. At the time of the 1882 outbreak, the Tokyo Quarantine Bureau, which was under the Home Minister and directed by the Police Agency and the prefecture, gave orders to local government offices and police bureaus. ²⁴ At the time of the 1886 outbreak, however, reflecting the fact that hygiene management for the entire country had been unified under the Police Agency, ²⁵ a quarantine division was set up in the agency. ²⁶ One month after the agency had completed its role in leading the construction of fire-resistant buildings, it began directing the repair and reconstruction of toilets and wells in preparation for the outbreak, which had already appeared in western Japan. By mid-June, these preparations and the dredging of drainage ditches were complete. ²⁷ Despite these efforts of the police, however, the cholera outbreak of 1886 resulted in 9,962 deaths in Tokyo, exceeding the 5,077 deaths in 1882. ²⁸

At the time of the 1882 outbreak, the governor of Tokyo Prefecture consulted the Prefectural Assembly about allocating funds to expand isolation hospitals, but the argument was made that the sick would disperse disease microbes in the process of being taken into custody, thereby exacerbating the spread of the disease, and the funds were denied. In response, the governor made a detailed report to the Home Minister, following government protocol, and was able to realize his original proposal under direction of the ministry. ²⁹ In 1886, funds for hospital construction were approved but a portion of the funds for disinfecting was denied and the same process was repeated, allowing the governor to pursue his original plan. ³⁰ In contrast to the handling of funds for firefighting, measures against epidemic disease were thus taken on the basis of state authority, ignoring the intentions of the Prefectural Assembly.

Execution of the policy also relied on force. If someone vomited and was believed to be suffering from cholera, a policeman would bring a team of workers and disinfect the house from the roof down to under the floorboards with carbolic acid using the hose of one of the old fire pumps, then burn all of the person’s clothing and belongings. ³¹ Even the editors of the newspaper Chōya shinbun, which criticized the unforgiving disinfection policy of the police, praised the practice at the same time as responsible for preventing the return of a tragedy like that experienced in 1858, when the first arrival of cholera in the city had taken the lives of tens of thousands. “If we were to abandon the present prevention law and leave things to nature,” the editors asked, “how could we guarantee this would not result in the disease strengthening its virulence to a far greater degree, ultimately calling before us a tragic scene like that of the fifth year of Ansei?” ³² Thus, although there had been no epidemic of comparable scale since, the memory of that late Edo-period catastrophe could still be invoked to support stringent measures.

In imposing these measures, the national government functioned in the mode of a developmental dictatorship, seeking to claim legitimacy by demonstrating the effectiveness of Western public hygiene, regardless of local concerns. The harsh response of the police and the governor’s imposition of state authority overriding the wishes of the Prefectural Assembly heightened the awareness that epidemic disease suppressed people’s freedom as well as causing heavy losses. Expectation increased for a modern water system that could prevent it. As Tokyo newspapers called for the implementation of improvements to the water system, entrepreneur Shibusawa Eiichi and his associates established the Tokyo Water Company in 1887 and commissioned a British military engineer to design the system. In response, the national government enacted the Urban Improvement Ordinance the following year, establishing a commission for the planning of Tokyo that included state bureaucrats and members of the Prefectural Assembly. This body revised the Urban Improvement Plan of 1884, adding the construction of the new water system to the plan, which gave it the prospect of winning funding approval from the assembly. In 1889, the Tokyo City government was created, and the prefectural governor, serving concurrently as city mayor, took charge of the project, leaving the budget to Tokyo City Assembly deliberation. The City Assembly approved a bond issue of ten million yen, of which 7.5 million would be spent on rebuilding the water system and the remaining 2.5 million on other aspects of the Urban Improvement Plan. There was opposition to the Improvement Plan as a whole, but on the one point of the water system, the assembly offered its approval. ³³ A state subsidy of 2.25 million yen was granted the same year, against the backdrop of another cholera outbreak. Construction of the modern water system began in 1891 and was completed in 1898. ³⁴ After 1895, Tokyo ceased to see epidemic outbreaks of cholera.

Influence of Fire and Earthquake on the City’s Architecture

With the opening of the ports, the technology of building in brick, long practiced in China and the West, was brought to Japan. After the large Ginza fire of 1872, which destroyed 4,879 buildings, including a number of former daimyo estates that had been converted to government offices, the Council of State announced that all buildings in Tokyo should in the future be constructed of brick and went about constructing and selling two-story brick rowhouses in the fire-destroyed area of Ginza based on designs by a foreign engineer. ³⁵ Those who rebuilt at their own expense were only permitted to do so in brick or stone, with the result that the former commercial center on Ginza Avenue was soon lined with brick buildings. Dozō construction was permitted off the front street. By 1879, 924 brick structures (and 91 in stone) had appeared in central Tokyo, including 451 built by the government, containing 1,436 households. Outside of the Ginza area, there were only 44 brick structures in all of Tokyo at the time. ³⁶

The Ginza district escaped further conflagrations. Brick was not broadly adopted, however. In 1880, following the large fire of that year, the governor proposed to the Prefectural Assembly that 32 brick storehouses be built to form a firebreak, but was asked by the assembly to change them to dozō construction, on the grounds that “the present state of brick manufacture is unrefined and manufacturers lack experience, so it would be better to build in the traditional Japanese dozō style.” ³⁷ It was apparently on the basis of this response that after the further conflagrations of 1881, new restrictions on construction permitted use of dozō in addition to brick and stone. What emerged as a result was a streetscape of dozō storehouse construction that inherited the architectural technology of the Edo period and has thus been referred to as the “final completion of Edo.” ³⁸

The issue of earthquakes becomes important in considering the introduction of brick and stone construction. An architect among the first graduates of the new Imperial College of Engineering wrote in the Journal of the Society of Japanese Architects (Kenchiku Zasshi) in 1887 that construction of brick buildings as firebreaks was “of great benefit to the state,” then recalled the earthquake of 1855:

Since buildings sturdy enough in ordinary circumstances will not survive without damage when struck by an earthquake of the kind that occurred in the Ansei period, it hardly requires an expert to establish that buildings inferior in all their parts will suffer more severe damage. Much of the architecture we have at present was built with the sole priority on cost and is of flimsy construction.

He pointed out that the inexpensive brick buildings that were beginning to appear at the time would be particularly likely to fail in an earthquake. ³⁹ The prefectural assembly’s rejection of the governor’s suggestion of brick thus reflected more than an amateur opinion. In fact, in the Nōbi earthquake of 1891, centered near Nagoya, major brick buildings in that city collapsed, including the Owari Spinning Mill, the Electrical Light Company, and the Post and Telegraph Office. ⁴⁰ Some structures of brick failed even in distant Osaka and Tokyo. Criticism of brick strengthened as a result, with some advocating its complete elimination. ⁴¹

Englishman Josiah Conder, professor of architecture at the College of Engineering and a powerful influence in the Japanese architectural world, examined the damage from the Nōbi earthquake. He concluded that brick buildings had collapsed because of “unreliable and cheap construction,” and insisted that under the “careful direction of experienced experts,” observing “strict laws of brick construction . . . with care and preventative measures,” Japan should continue to build in brick. ⁴² In fact, the Bank of Japan building, which was under construction at the time, was built under strict supervision, and it was reported that “no mason or bricklayer was permitted to lay a single stone or brick except in the presence of a foreman, allowing not the slightest doubt that the mortaring had been done with care.” ⁴³ The building survived the Kantō Earthquake and stands today.

Nevertheless, following the Nōbi earthquake, the idea of requiring construction in stone or brick as a general solution of the fire problem was abandoned. In 1890, the year before the earthquake, the city had issued a regulation that all public bathhouses be built in brick or stone for fireproofing. ⁴⁴ However, after the earthquake, allowance was made for bathhouses to remodel without rebuilding. In the end, the law was altered to require only that the furnace rooms for heating the baths be built of fireproof material. ⁴⁵ Construction of government buildings used brick in the 1880s, and the Mitsubishi development of the Marunouchi district in the 1890s was built of brick, making brick characteristic of the central architectural examples of Tokyo’s modernization in commerce, business, and government, but brick construction remained restricted to a small collection of special cases in which ample funds could be invested.

Reinforced concrete was used in building construction in Tokyo for the first time in 1909. The Urban Building Law enacted together with the Urban Planning Law of 1919 delineated two standards of fireproofing: structures in class-one fireproof districts were required to have exterior walls of reinforced concrete at a minimum thickness of four sun (121 mm) or one shaku (303 mm) of brick or stone; while class-two semi-fireproof districts were permitted a lower standard. Thus, given the same building size, reinforced concrete structures could attain larger floor areas. Central Tokyo was designated for the higher class-one standard of fireproofing in August 1922, while buildings along major roads were to meet the semi-fireproof standard, but this applied only to new construction and modifications. Before the new regulations could yield results, the city was hit by the Kantō Earthquake on September 1, 1923.

At the time of the Kantō Earthquake, the water system failed, hampering firefighting and resulting in casualties exceeding the Meireki Fire of 1657. Thirty-five square kilometers burned, comprising 43 percent of the urban area. Afterward, subsidies were offered for fireproof construction in class-one fireproof districts that had burned. ⁴⁶ In April 1925, the class-one fireproof area was expanded and main streets previously designated class two were upgraded to class one. ⁴⁷ Large structures had to be built in fireproof materials, resulting in increased use of reinforced concrete in commercial buildings, schools, and other structures. However, since temporary structures were exempted from demolition after the earthquake even in class-one districts, ⁴⁸ as late as the end of 1939, a total of only 10,161 concrete buildings stood in Tokyo, surpassing the 5,618 dozō and 3,464 brick buildings, but a bare 0.94 percent of the total of 1,078,909 buildings altogether, making up 9.2 percent of total floor area. ⁴⁹

Those who had the capital to do so built in reinforced concrete, which was in practical terms the only fire- and earthquake-resistant building technology, as well as the only technology suitable for building above three stories. But others held out the hope that they could seek temporary exemption from the fire-resistant construction requirement in class-one fireproof districts. ⁵⁰ It is well known that there was strong opposition to the reconstruction plan as a whole from some members of the reconstruction commission within the government, although members were not opposed to fireproofing. However, because of the opposition, the city was forced to proceed with a plan for the burned areas that focused primarily on street widening and regularization, and without sufficient funding from the national government, this was limited to a procedure whereby landowners were made to give up a fixed portion of their land without compensation. ⁵¹ It was likely because of this that the city had no choice but to give in to the demands of property owners to be released from the obligation to rebuild in fireproof materials.

Although reinforced concrete buildings had increased in number by the 1930s, they still made up only a very limited part of the cityscape. In contrast, what now stood out were the numerous examples of what architecture historian Fujimori Terunobu has called “signboard architecture”: wood-built structures clad with cement or other fire-resistant material. ⁵² The use of wire mesh and cement, first applied in 1911, made it possible to cover traditional wood-built structures with a façade of concrete. ⁵³ With some improvements, this method continues to be used. The 1920 Urban Building Law had permitted a building with a wood frame to be considered semi fire-resistant if the exterior walls were covered with 1.2 sun (37 mm) of cement mortar or concrete. Since dozō buildings were classified as semi fire-resistant only if their total wall thickness including clay walls and plaster were over 3 sun (91 mm), cement-clad wood became the more expedient choice. ⁵⁴ For the shops and businesses that faced main streets and therefore had to be built to a certain standard of fire-resistance, reinforced concrete was 1.5 times more expensive than wood, while wood construction with cement cladding was 1.2 times the cost. ⁵⁵ Since the cement cladding allowed building owners to display a Western-style façade that looked similar to reinforced concrete, it was widely adopted not only on main streets outside the designated class-two fireproof districts, but also in temporary structures built in the reconstructed districts where the streets were otherwise lined with reinforced concrete buildings. ⁵⁶

Flooding: Waterway Engineering and the Boundaries of Greater Tokyo

As noted above, the northern and eastern areas of Tokyo, along with southern districts extending to the Tama River, all lie close to sea level and frequently experienced floods as a result. Much of this land had originally been rice paddies. As the urban area expanded into this farmland, flood control emerged as an issue. The application of modern riparian technology began in the 1910s. Without this technology, much of the suburban expansion of the modern city would not have been possible.

The expansion of Tokyo in the 1920s and 1930s is generally explained in terms of the growing commuting area with the development of suburban rail. ⁵⁷ This projects backward from later conditions. In 1995, about two-thirds of residents either working or attending middle school or above in the 23-ward area of Tokyo held monthly passes for public transport. In contrast, however, in the fifteen wards of Tokyo City in 1930, only one in three working or attending school commuted from outside of the ward where their workplace or school was located. ⁵⁸ If we include the four surrounding counties that would be incorporated into the expanded 35-ward Greater Tokyo two years later (establishing the boundaries of present-day Tokyo’s 23 wards), the proportion of commuters drops to one in four. ⁵⁹ Commuting by public transport was not yet the norm, as it would later become. Viewing Tokyo’s growth only through the expansion of the rail network thus projects the conditions of late twentieth-century Tokyo back onto the past, and for the prewar years, depicts only the city as it was experienced by new middle-class white-collar employees who commuted to downtown offices. The same 1930 survey reveals that in contrast with the central five wards, which had large numbers of out-of-ward commuters, in the four eastern wards, more than eighty percent of residents working or attending school did so within their own ward. In these so-called Shitamachi districts, it was usual to work at home (in small shops and light industry) or to commute to workplaces or schools on foot. The same was true in the northern and eastern portions of the newly incorporated districts of Greater Tokyo. ⁶⁰ For these areas in the north and east, zoned under the 1929 city plan as industrial, it does not work to explain urbanization mainly in terms of new residential construction following the expansion of rail.

In the Meiji period, these northern, eastern, and southern areas flooded frequently from swelling of the Arakawa, the Sumida, and the Tama Rivers, as well as high tides from Tokyo Bay, but since they were predominantly rice paddies at the time, serious flood management measures were not taken. Waterway normalization also played little part in the Tokyo Urban Improvement project begun in 1889. ⁶¹ Tokyo City launched an improvement project at the mouth of the Sumida River in 1906, anticipating revenues from landfill construction, but did not get around to initiating fundamental waterway reforms for the purpose of flood prevention. After flood damage in 1896, the Prefectural Assembly sought national government funding for work on the Tamagawa, Arakawa, and Edogawa rivers, but the funding was denied. ⁶²

Between 1885 and 1904, flood damage in Tokyo Prefecture as a proportion of the total nationwide was no more than 0.54 percent. With massive flooding in 1907, that rose to 3.1 percent. ⁶³ Following the 1907 floods, the prefectural and city assemblies announced plans for a survey in preparation for a water management project that would include the opening of the Arakawa Drainage Canal. Unable to realize the project themselves, they proposed the survey to seek a response from the national government. The prefecture and the city both pursued surveys and submitted their results to the national government. ⁶⁴

The year 1910 saw unprecedented flooding nationwide. Flood damage in Tokyo was particularly severe, constituting 4.7 percent of the total national losses. ⁶⁵ Prime Minister Katsura Tarō, who also held the Finance Ministry portfolio at the time, established a special commission of inquiry on flood control to explore a long-term plan beyond the existing bureaucratic framework. The plan targeted sixty-five rivers, focusing on twenty in the first phase. Prior to this time, the national government had directly managed flood control projects on ten rivers nationwide, none of which were in Tokyo. This time, Tokyo’s Arakawa, Nakagawa, and Tamagawa Rivers were included, and construction began in 1911 on the Arakawa, where the surveying work had advanced furthest. ⁶⁶ The work focused primarily on cutting the channel for the drainage canal in Tokyo that is today known as the Arakawa River, while also regularizing the channel of the Nakagawa, a tributary. Mechanical excavation and dredging equipment was used. Water began flowing in 1924 and the project was completed together with construction of embankments upstream in 1930. ⁶⁷ With its completion, eastern and northern Tokyo were largely saved from severe floods. Between 1918 and 1933, the downstream portion of the Tama River was also re-engineered under national management. As a result of these projects, stable development of new industrial zones followed, creating districts dominated by closely situated homes and workplaces and, since the land was flat, suited also to bicycle commuting. Thanks to improved flood control, the topography, and the wide use of bicycles, workers who could not afford commuter rail were able for the first time to leave the inner-city slums and live in better surroundings. ⁶⁸

Conclusion

Facing the problems of flooding and conflagration in the seventeenth century, the Tokugawa shogunate ordered the daimyo to contribute to a project to protect the castle area from flooding, then re-planned the built-up area of the city for fire control. These steps defined the shape of the city of Edo.

The opening of ports in the mid-nineteenth century on one hand brought cholera to the city for the first time, while on the other hand making possible the importation of Western building technology. The new government, after making Tokyo the de facto capital of the country, immediately upon gaining the authority to impose a national tax, applied strict building regulations and poured in funds from the national coffers to unveil the new brick-built Ginza district. This represented a hasty application of technology imported from advanced countries. Natural conditions gave disasters a different character in Tokyo, however, so that the introduction of technology from advanced countries required a certain period of preparation and selection, together with fusion with native technologies.

The Meiji government and large capitalists built structures using Western architectural styles and techniques and took pride in their strength, but since earthquakes are frequent in Japan, there were limits to the general diffusion of brick construction. In order to compel people to adopt fire-resistant architecture, it was necessary to persist with the traditional method of dozō construction. In contrast, only a modern piped water system based on imported technology offered a truly effective response to both epidemic disease and fires. Yet before the need for this technology could be widely recognized, firefighting based on dousing rather than demolition had first to become the norm, and society at large had to understand the significance of public health measures against contagion. For roughly ten years, the Police Agency, as an arm of the state, pursued the reform of firefighting and implementation of epidemic disease prevention measures by trial and error. As a result, the City Assembly eventually agreed to funding for the modern water system and the national police came to occupy a solid position in urban society, having acquired the role of bearers of modern civilization.

Construction of the modern water system was realized through approval from local elected bodies representing city residents, but the plan itself came under the leadership of a foreign expert hired by the state. The national government viewed modern waterworks for the imperial capital as too important a matter to be left to local authorities and proceeded by fiat, with the municipal government responsible only for the cost. Similarly, large-scale reconstruction of the city’s waterways for flood prevention followed demands from locally elected officials and surveys conducted by the prefecture and the city, but was ultimately carried out as a state project. Entering the twentieth century, when the need increased to expand Tokyo’s urban area for factory construction, the cost of flood damage served to legitimize state investment in prevention, and thanks to waterway reconstruction projects carried out by the state and assisted by increased national tax revenues during World War I, the city expanded northward and eastward. The reconstruction plan following the Kantō Earthquake too was put together by the Imperial Reconstruction Commission, a state agency.

Tokyo’s residents came to grasp the necessity of a modern water system as a result of the Police Agency’s deployment of new technology. They responded with native technology to the order to build in fire-resistant materials; they worked in districts made safe by waterway engineering; and they responded to the post-earthquake reconstruction plan by giving up land for street widening and in exchange escaped the requirement to rebuild in reinforced concrete. While the state led the reshaping of the city in response to disasters, residents found realistic ways to respond to the state initiatives, and together these two forces changed the cityscape. After American firebombing destroyed the majority of the city in 1945, the state attempted to pursue its postwar urban reconstruction plan, but with reduced authority and finances because of the defeat. Under postwar democratization, the Tokyo Metropolitan Government was transformed from a state institution to a municipality. Tokyo residents and municipally employed engineers, who had had little experience in determining the shape of the city on their own initiative, found themselves unable to control the chaotic process of reconstruction through comprehensive planning. In light of the process that had shaped Tokyo before the war, it is not surprising that under these circumstances the postwar plans yielded little.

Translated by Jordan Sand