How Did a City of Wood Live with Fire?

Inside a nineteenth-century firefighter's jacket, a hero confronts a monstrous centipede.

The garment, preserved in the Musée Guimet, was made from thick, closely stitched cotton intended to hold water. Its wearer could soak it for some protection while working near a fire. The decorated lining presents a different kind of defense: the legendary Tawara Tōda facing an extraordinary enemy.^[1]

The contrast gives us two ways of remembering Edo's firefighters. One makes them figures of daring. The other returns us to the limited protection available to a person approaching heat, smoke, and collapsing buildings.

Both belong to the history. Neither is enough by itself.

Courage did not decide where a firebreak would go, who would pay for a safer roof, or how people would eat after their homes burned. Nor could regulations implement themselves while frightened residents carried possessions into the street. The city needed people willing to act, but it also needed arrangements that made their actions useful.

Edo's experience reveals a problem larger than firefighting: how does a community protect itself against a danger continually reproduced by the ordinary things that make its life possible?

Edo, the city now called Tokyo, expanded around the Tokugawa shogunate and the households serving it. Merchants and craftspeople supplied the growing population. In busy districts such as Nihonbashi and Kanda, shops and crowded housing stood close together, much of it built from combustible materials. Dry winter weather and strong winds made a spreading fire especially dangerous.^[2]

This was not a city that used fire only when someone behaved recklessly. Cooking, warmth, and light required controlled burning. A household could be careful and still depend on a flame several times a day. Avoiding every source of ignition would also mean abandoning essential parts of ordinary life.

The difficulty was multiplied by proximity. One building's roof could become the next building's problem. An owner deciding what repairs were affordable was making a private calculation with consequences beyond the property. A neighbor might suffer without having chosen the material or permitted the original flame.

Recognizing that connection does not make the solution effortless. Separation uses land. Stronger protection requires materials and work. A family needing somewhere to live cannot always wait until the safest possible building is available. The choice facing people after a fire might be between an imperfect shelter and no shelter, not between two equally accessible architectural systems.

There is therefore little explanatory value in imagining Edo's residents as people who simply failed to notice that wood burned. The problem lay in coordinating necessary activities under constraints that no single household could remove.

Nor did official attention begin only after catastrophe. Before the Meireki fire, towns were already subject to requirements concerning watches, buckets, water supplies, and responses to an alarm.^[3] Those rules did not provide an effective answer to every large blaze. They do show that repeated loss was not the same thing as complete indifference.

A city can understand a danger and remain exposed to it. Knowledge of the hazard is only the beginning of deciding what to change, who must change it, and how much change people can sustain.

The Great Fire of Meireki broke out in 1657, during the rule of the fourth shogun, Tokugawa Ietsuna. The name brings together three major outbreaks beginning on the eighteenth and nineteenth days of the first month of the contemporary Japanese calendar—March 2 and 3 in the Gregorian calendar. Strong winds followed a prolonged dry spell. Tens of thousands died, although estimates of the total vary.^[3]

Calling it one fire is convenient. It can also make the event seem simpler than the experience of those trying to escape it.

Separate outbreaks meant that a place beyond one advancing fire was not necessarily beyond danger. The failure of a building was only one part of the disaster. People also needed to move, locate others, and choose a destination while conditions changed faster than they could be understood.

The castle's keep burned too. Its destruction connected the most imposing center of government with the much less imposing homes lost elsewhere. But a shared exposure did not create identical losses. The government could retain officials, claims to revenue, and buildings outside the devastated area. A household might lose the few things through which it earned a living.

The tower was not rebuilt, though a new foundation was prepared.^[4] Its absence has become one of the best-known signs of the disaster. Yet an empty platform is a poor measure of suffering. It records a conspicuous construction decision, not the full number of interrupted lives.

A later story linked the catastrophe to a burning kimono. The historian Iwamoto Kaoru notes that comparable accounts are absent from contemporary records.^[5] The story gives the disaster a memorable beginning; it cannot safely substitute for an explanation of how fire spread or why so many people were vulnerable.

The historical questions begin where the satisfying story stops. Which defenses failed? What could be changed before the next outbreak? And how could a government keep people alive long enough for rebuilding to mean anything?

Edo did not possess one uniform firefighting organization throughout the Tokugawa period.

Early official arrangements placed particular responsibility on great lords and focused on the castle and warrior residences. After Meireki, permanent firefighting establishments were created under direct shogunal retainers. The town brigades associated with magistrate Ōoka Tadasuke were organized from 1718, under the eighth shogun, Yoshimune.^[6]

The sequence matters. The familiar firefighter of popular imagery belongs to an evolving system, not to an institution fully formed when Ieyasu established his government.

Each arrangement also answered the question of responsibility differently. Calling on a lord's retainers used an existing chain of command. Keeping a permanent establishment reduced the need to assemble everyone only after an alarm. Organizing townspeople drew more directly on the people who lived and worked among the buildings at risk.

These were practical choices with political implications. A service centered on the ruler's compound was not the same as an equal promise to protect every household. Extending organized protection into town districts could increase capacity without abolishing the distinctions between different parts of the city.

It is important not to mistake this development for a neat replacement of privilege by public service. Older and newer arrangements could coexist. The people involved had different obligations, resources, and relationships to authority. The challenge was to make that uneven collection capable of acting against a danger that did not respect administrative boundaries.

Thinking about responsibility also changes the meaning of preparedness. A tool stored somewhere is not yet a response. Someone has to know where it is, have reason to use it, and arrive while its use can still make a difference. A permanent station and a neighborhood organization address those problems in different ways.

This is one of the lessons that recurring fires forced into government. A city could not depend indefinitely on bravery appearing at the right place by chance. It had to organize the expectation that particular people would act.

Those people often went toward a fire intending to destroy buildings.

Edo's firefighters relied heavily on demolition to interrupt the spread of flames. Hand-operated pumps existed, but their reach and capacity were limited. Hooks and saws helped crews take structures apart. Construction workers formed an important part of the town brigades because they possessed useful skills for dangerous work on buildings.^[6]

That is a very different picture from a modern engine directing a powerful stream of water at a burning room. The aim might be to deprive an advancing fire of the next thing it could burn.

Knowing how buildings were put together became valuable in taking them down quickly. Work normally used to make the city habitable could be turned toward removing part of it. The same person's competence could serve construction, firefighting, and the rebuilding that followed.

Yet demolition also exposed a difficult question. From the perspective of a threatened district, sacrificing one structure might preserve others. From the perspective of the people losing that structure, the sacrifice was immediate and concentrated. A decision made for a larger area could still leave a particular household homeless.

That is an ethical consequence of the method, not evidence that every such action was disputed or that every owner was treated in the same way. It explains why effective firefighting required authority and cooperation as well as physical skill. Someone had to decide where to make a break, and people acting together had to accept enough of the decision to carry it out.

The material limits should remain visible too. Wet clothing could help; it could not make its wearer immune to heat. A well-chosen gap could interrupt a line of fuel; it could not guarantee that windborne embers would stop at the edge. Admiration for the work is more meaningful when we remember what the workers could not control.

The hero on the jacket faces a creature with a recognizable body. A city fire had no single opponent to defeat. It required a succession of local judgments, sometimes including the deliberate loss of something worth saving.

After a great fire, a map can make rebuilding look seductively simple. Burned blocks become spaces in which a planner appears free to draw wider roads, move institutions, and provide room for the next emergency.

But destruction of buildings does not automatically erase the relationships attached to land. Owners, tenants, religious communities, and powerful households still have interests. A government proposing a new arrangement has to confront those interests, even when the old roofs are gone.

Iwamoto's study of the Meireki disaster challenges the familiar story of a single, comprehensive urban redesign. Comparing maps and records, he found that temple relocations had peaked before the fire. He describes the shogunate as negotiating among interests rather than imposing one finished master plan. Some lords also used the occasion to seek larger residences.^[5]

This does not mean the fire had no spatial consequences. It means that change after a disaster should not automatically be attributed to a plan created by the disaster. Some projects continue earlier developments; others emerge through bargaining over what is now possible.

Firebreaks demonstrate both the achievement and the difficulty. Clearing land could interrupt combustible buildings and make room for movement. But land kept empty also had other possible uses. The benefit of an open space might become most obvious on a day nobody wanted to experience, while the wish to build or trade there was present every ordinary day.

Around Ryōgoku Bridge, built across the Sumida after Meireki, broad open areas at the approaches became famous places of entertainment. The National Diet Library identifies their fire-prevention purpose as well as their later popularity; the Edo-Tokyo Museum traces the bridge and the attractions gathered around it.^[7][8]

That history cannot be reduced either to successful planning or to people carelessly undoing it. Space could support several uses, provided those uses did not destroy the function for which the space was needed. Keeping such a balance required continuing management.

The important achievement was not simply drawing an empty area on a map. It was maintaining an agreement about what people could do there after the memory of the latest fire began to fade.

Building rules offered another way to slow the movement of fire. They also show that official policy could change its mind.

A 2024 study of the shogunate's chronicle, the Tokugawa Jikki, traces a reversal from earlier restrictions on tiled roofing to its encouragement during the Kyōhō period, 1716–1736. Officials also promoted protective roof treatments using oyster shells. In 1729, an inspection in Kōjimachi found shells laid on roofs without the required plastering. Orders demanded correction; penalties reached property holders and a local headman.^[9]

The episode is valuable because it takes us beyond the announcement of a sensible measure. A rule had reached the roof, but not in the form the authorities intended.

The surviving official account does not tell us what each resident was thinking. Ignorance, expense, expedience, and resistance are possibilities, not motives we can assign without evidence. What the account does establish is a difference between prescribing an improvement and obtaining it.

That difference is built into the problem. A safer building benefits its occupants, but it may also protect people nearby. The person expected to pay is not necessarily the only person who gains. Where owners and occupants differ, deciding who bears the cost adds another relationship to the calculation.

Enforcement can make compliance more likely while placing new pressure on those least able to rebuild. Relaxing a requirement can ease an immediate burden while leaving a hazard in place. There is no reason to assume that every official confronting these choices had either a benevolent solution or a purely cynical purpose.

A study by Yuji Morishita and his colleagues emphasizes the interaction between Edo's firebreaks, firefighting arrangements, and building regulations. These measures developed across decades rather than operating as three self-sufficient inventions.^[10] A stronger roof would be more useful within a district whose organization and open spaces also helped contain a blaze.

The larger lesson is not that regulation failed whenever a city burned again. Nor is issuing a rule proof that protection had improved. We need to examine what people actually changed, how that change was sustained, and what demands it placed on them.

Fire safety was becoming a question of ordinary administration precisely because it could not be completed by one extraordinary response.

Surviving the Meireki fire did not end the danger.

The Cabinet Office's historical report describes emergency food distribution at six locations, including outside Zōjō-ji. A feeding program initially intended to last nine days had to be extended. Cold and snow added to the distress of people without homes. Rebuilding brought shortages and rising demand for materials as many people tried to replace what they had lost at once.^[11]

These are different problems from extinguishing a blaze. They require food, shelter, transport, and purchasing power rather than a better position from which to pull down a roof.

The distinction matters when we speak of a city recovering. From a distance, fresh buildings may make recovery visible. For a survivor, the more urgent test may be whether there is something to eat tonight or a place for a child to sleep. A reconstruction program operating on the scale of months can coexist with an emergency operating on the scale of hours.

Relief also had to pass through existing social arrangements. The report records assistance to townspeople as well as warriors; some distributions were calculated through property frontage rather than by counting individual survivors.^[11] That method made aid administratively manageable, but it does not tell us how fully the needs of every tenant or dependent were met.

We should neither assume that nothing reached them nor mistake a recorded allocation for proof that everyone had been provided for. The archive's unit of administration and a person's unit of need are not always the same.

Rebuilding had a related inequality. A household that retained savings, a place in a trading network, or help from relatives faced different possibilities from one that had lost its tools and support together. This is a distinction in what recovery requires, not a claim that the surviving records allow us to balance every family's losses.

It also warns against treating the demand for new buildings as compensation for the disaster. Some workers and suppliers could find business in reconstruction. That did not restore the people who had died, or make the losses of those without resources economically harmless.

A city can be rebuilt while many of its inhabitants remain far from recovered. The new skyline is evidence of activity, not a final account of what happened to everyone below it.

Edo's repeated rebuilding is sometimes easier to admire than to explain. The city burned and returned; its inhabitants appear to have possessed an extraordinary ability to begin again.

That description recognizes effort, but it can conceal the arrangements that made effort effective—and the lives for which a new beginning was much harder.

The cases suggest no single Japanese law of resilience. They reveal several distinct tasks. People needed to notice and report danger, organize a response, interrupt the path of a fire, support survivors, and rebuild without reproducing every vulnerability. Each task involved different skills and a different distribution of costs.

A firefighter could act courageously without being able to choose the city's housing materials. A magistrate could demand a safer roof without supplying the means to build it. A merchant could benefit from an open riverside while adding to the pressure on space reserved for emergencies. The problems were connected, but nobody experienced all of them from the same position.

The institutions that helped Edo endure were therefore not simply barriers against nature. They were ways of arranging obligations among people. Their effectiveness depended on whether orders became practices, whether useful spaces remained usable, and whether assistance reached people in time to matter.

Even successful arrangements required renewal. New buildings appeared. Personnel changed. A protective measure became expensive to maintain or inconvenient to follow. The next generation inherited both the benefit of earlier work and the temptation to reduce its cost.

This helps explain why living with a recurring danger is different from accepting it. People may continue an exposed way of life because their work, relationships, and resources are there, while still trying to make it less dangerous. Remaining in the city does not prove that its losses were willingly chosen.

Return to the firefighter's jacket. The spectacular image belongs to its history, but so do the stitches that enabled the fabric to hold water. One makes courage imaginable. The other records a modest attempt to improve the wearer's chances.

Edo needed both the person who would approach the flames and the less celebrated work that began before an alarm or continued after it. Its survival was not an effortless cultural gift. It was a contested, uneven achievement, repeatedly made by people who could never be certain that the preparations of this year would be enough for the next fire.