Donate Menu


September 1, 2023
Volume 21 | Issue 8 | Number 2
Article ID 5784

Keywords: Disaster, Great Kantō Earthquake, Korean massacres, economy, historiography


“Fire Ruins of Kanda Main Street" (Imaging Kantō, 2017).


Four major segments of the earth's lithosphere converge close enough to the Kantō plain that their interactions have had, and will continue to have, powerful, harmful effects on the region’s inhabitants and built environment. These tectonic plates are all in motion, but not in the same direction, and not at the same rate. Where the Philippine and North American plates collide along the Sagami Trough, the former is slipping—subducting—beneath the latter.1 The historical record and paleoseismic studies suggest that one of the characteristics of that process in that location has been the sudden slippage of parts of the Philippine plate along a large surface of the subduction zone.2

Just before noon on September 1, 1923, along a fault extending roughly northwest by southeast deep beneath the Sagami Bay, that pattern repeated. Some sections of the Philippine plate are estimated to have moved by as much as eight meters laterally; parts of the Bōsō and Miura peninsulas were thrust vertically upwards by almost two meters, all in a matter of seconds. The contours of the ocean floor under the bay were also radically altered by the shift. The tsunami that resulted swept swimmers away at Kamakura, and towered twelve meters above the shoreline when it crashed ashore in Atami.

On land, the energy released by the slippage had a catastrophic effect on communities all across the Kantō plain, and especially in Tokyo and Yokohama. The abrupt, violent movements of the ground beneath the densely populated lower city wards of Honjo, Fukagawa, and Asakusa in Tokyo and in parts of Yokohama caused the immediate and complete collapse of many thousands of homes and other structures there. Almost 10,000 of the two cities’ residents are thought to have been killed immediately or nearly so as buildings fell down around them. Landslides and tsunami killed hundreds more.3

Some of the first fires after the tremors ended reportedly began within collapsed homes and factories and spread quickly from those locations, house to house and then neighborhood to neighborhood. Firefighters and civilians tried desperately to extinguish or at least contain the flames, but could not. Damaged equipment, inadequate supplies of water, blocked roads, and poor communication hampered even the professionals’ efforts. There were few wide boulevards or other potential fire breaks that might have slowed the fires’ spread in either Tokyo or Yokohama. In the capital, as the conflagration quickly engulfed more and more of the city, surviving residents sought shelter where they could. Many fled to the wide-open and plaza-like grounds of the former Honjo Clothing Depot, not far from the east bank of the Sumida river, hoping that there they would be safe as the city burned around them. They were not; the firestorm that swept over and through the crowd that afternoon killed almost everyone that had gathered there; 30,000 are thought to have died at that site alone (Schencking, 2013, 25–26, 29–35). Of the more than 105,000 people thought to have died as a result of the Great Kantō Earthquake, most—close to 90 percent—were victims of the fires. 

Those who survived the earthquake faced a nightmarish landscape. The tremors and the fires had left roughly half the capital’s population homeless. Forty percent of all the buildings in Tokyo had been completely destroyed; in several of the lower city wards, almost all had burned to the ground, the devastation in those areas almost total. Yokohama fared no better; the earthquake damaged or destroyed 90 percent of the structures in the port city. Those unhoused by the disaster in the two cities made do with makeshift shelters in parks, other open spaces (including right outside the grounds of the Imperial Palace), and wherever else they could find some sense of security, however temporary.

Some escaped the earthquake’s immediate dangers only to fall victim to acts of violence at the hands of other survivors and agents of the state. Rumors that seditious Koreans were poisoning wells, or starting fires, or about to launch an assault on the capital—these packets of misinformation took many forms and none of them were true—spread quickly across Tokyo, Yokohama, and eventually the rest of the country as well within the first 48 hours after the earthquake. Assemblages of vigilantes (jikeidan), enabled by the rumors, the police, and the state’s initial failure to intervene, began assaulting and killing Koreans (and not a few individuals mistakenly identified as such) at sites throughout the region. The violence went on for days, and ended only after the government made a belated effort to disarm and disband the vigilante groups.4

The police would later report that as many as 233 Koreans had been killed by members of the jikeidan, a figure that eyewitness accounts at the time and researchers afterwards insisted was far too low. Recent studies suggest that more than 1,700 Koreans were massacred in Tokyo after the earthquake, and that perhaps as many as 4,000 were killed in Yokohama.5 Very few of the massacre’s perpetrators were arrested or tried, and fewer still convicted (Schencking, 2013, 28). Deliberate acts of violence after the earthquake also took the lives of a number of labor and political activists. Their killers too were seldom held to account for what they had done (Gordon, 1991, 61–64).

This special issue of Japan Focus helps mark the 100th anniversary of the Great Kantō Earthquake by bringing together eight reflections on the event and its legacies from the perspectives of history, history of art, literary studies, and seismology, by scholars with long-standing interests in the disaster. One of our goals in assembling these short articles is to share with readers some of the approaches and questions that have shaped the recent research on the catastrophe’s multiple, overlapping tragedies. That body of scholarship has grown considerably over the past decade or so, much of it produced by the contributors to this collection. The essays collected here build on that work, in some instances by re-visiting arguments and sources that were showcased in that earlier wave of scholarship on the Great Kantō Earthquake, and in others by highlighting bodies of evidence and analytical concerns that are shaping research on the disaster and its legacies in the present. Most do a little of both, and all draw attention to the many different mediums and methods through which the Great Kantō Earthquake was constructed as an event and as an idea.6

Alex Bates’ contribution to the special issue, “Fiction from Unstable Ground: The Imagination of Disaster in the Aftermath of the Kantō Earthquake,” for example, explores literature’s continuing role in the creation of “a shared imagination of disaster” in Japan, one shaped by encounters with trauma and with the differential vulnerabilities that are the inevitable products of modern catastrophes. Bates’ earlier scholarship on the disaster, meanwhile, including his 2015 monograph The Culture of the Quake: The Great Kanto Earthquake and Taishō Japan, analyzed its proximate impacts on a wide range of literary and cinematic genres, from I-novels to popular melodramas, and on the writers and filmmakers responsible for the work that constituted each. 

Janet Borland’s Earthquake Children: Building Resilience from the Ruins of Tokyo (2020) looked at the many forms of cultural and didactic work that targeted the capital’s youngest residents, and brought to light the abundance of effort that went into documenting, trying to understand, and ultimately attempting to re-shape what the disaster meant to them. Her contribution to this special issue, “Giving Earthquake Children the Voice they Deserve One Hundred Years Later,” explores how primary school students throughout the city came to be instructed by their teachers to write essays in which they were to describe what they experienced or witnessed on September 1 and in the days that followed. Borland goes on to show how collections of those essays were originally shared with the public not long after the earthquake, and explains how (and why) the same materials are once again being deployed, this time as part of the city’s ongoing efforts to encourage Tokyo’s residents to prepare themselves and communities of which they are part for the city’s next disaster. 

Seismologist Robert. J. Geller, who became the first tenured foreign faculty member in the University of Tokyo’s history when that institution hired him as an Associate Professor in 1984, has devoted much of his scientific career to the development of new methods for mapping the interior structures of the earth and to the detailed analysis of the results of those efforts. Since the early 1990s, Geller has also been one of Japan’s most vocal and persistent critics of that country’s commitment to short-term earthquake prediction (and of other prediction-adjacent projects) as a scientific project and as a disaster mitigation strategy. Geller’s critiques have appeared regularly in professional journals as well as in media intended for general audiences; the Japanese press has covered his anti-prediction efforts at some length over the years.7 In his article for us, Geller examines the “Imamura-Omori Forecasting Debate,” a well-known episode in Japanese seismology’s early twentieth century history in which the Great Kantō Earthquake played a definitive role. Geller’s assessment of seismologist Imamura Akitsune’s claim to having successfully forecast that event appraises its specific shortcomings, while also modeling an approach to thinking about the many similar assertions that have surfaced since.8

Andre Haag’s research has long focused on the production of what he has elsewhere described as “the tensions and terrors of cultural engagements linking Japan and Korea from the era of imperial literature to postcolonial Zainichi cultural production.”9 For this special issue, Haag has written a nuanced genealogy of sorts of the phrase “Chōsenjin sawagi”—a derogatory description that was ubiquitous in media and official accounts of the “confusion” and “panic” that supposedly set in soon after the shaking stopped. Its use in that context was in part a means of deflecting responsibility for the subsequent mass killings of Koreans onto the victims themselves; Haag also shows that “Chōsenjin sawagi” did that work in part by linking fears of violence in the metropole to deeply felt anxieties about Japan’s ongoing imperial projects.

Janet Hunter is currently at work on a monograph-length analysis of the economic impacts of the Great Kantō Earthquake, and her paper for us emerges out of that research. In it, Hunter draws our attention to the interesting contrast between the close attention that Japanese economists and other researchers devoted to documenting and analyzing the disaster’s effects in its aftermath, and the relative disinterest economic historians have shown to those concerns since. Using examples from Ikeuchi Yukichika’s book Shinsai keizai shigan (Personal View of the Earthquake Disaster Economy), published in December 1923, along with other reports from the same time period, Hunter shows that Ikeuchi’s methods and conclusions can tell us a great deal about how he and his contemporaries conceptualized the structures of and tensions within the Japanese economy, and the effects a disaster might have on them. That he and his cohort’s approaches to understanding how the earthquake affected the economy in many ways prefigure work being done by disaster economists in the present day is perhaps indicative of how transformative contemporary observers believed the Great Kantō Earthquake to be. 

Charles Schencking’s The Great Kantō Earthquake and the Chimera of National Reconstruction in Japan (2013) explored the social and political responses to the disaster in rich detail. One of that book’s several interventions was to use the contentious debates over what a rebuilt, reimagined Tokyo might look like to outline the limits on the Taisho political regime’s capacity to bring about the sort of meaningful, lasting change that Gotō Shimpei and some other visionaries had sought. Schencking’s contribution to this special issue, meanwhile, takes one of the few “feel good” stories to emerge from the disaster—that of the unprecedented outpouring of support for Japan by generous donors from all across the United States—and unravels some of the assumptions surrounding it. Schencking affirms that Americans were indeed enormously generous when it came to offering aid in various forms to the Japanese after the earthquake. The largesse was real. What he goes on to show, however, is that businesses in America’s industrial heartland gave with the explicit expectation that their charity would be rewarded many times over by its Japanese recipients, if not immediately then soon, in the form of business contracts, purchase orders, and any number of other money-making opportunities. That corporate America’s embrace of humanitarian aid for Japan was essentially a marketing strategy, Schencking argues, unsettles much of what we thought we knew about the dynamics between its U.S. donors and their Japanese recipients.

The lingering impacts of the catastrophe in 1923 are evident throughout Kerry Smith’s forthcoming book, Predicting Disasters: Earthquakes, Scientists, and Uncertainty in Modern Japan.10 His focus in that work is in part on Japanese seismology’s efforts after the Great Kantō Earthquake to engage—or not—with research and disaster mitigation efforts in which the pursuit of earthquake prediction played a central role. The Imamura-Ōmori debate, the focus of Geller’s article in this issue, is one of the many legacies from 1923 that later generations of earth scientists were forced to confront. When experts in the 1960s began warning policy makers and the public about the growing seismic risk to Tokyo, for example, they included vivid descriptions of what would happen to the city in its present form if an earthquake just like the one that had struck the capital district in 1923 did so again. In his article in this issue, Smith looks at the Great Kantō Earthquake’s legacies in a different setting, and explores how security forces in Japan in the early 1960s looked to the unrest that followed the 1923 disaster for insights into the challenges they would face in the aftermath of Tokyo’s next disastrous earthquake. 

The 2012 publication of Gennifer Weisenfeld’s masterful Imaging Disaster: Tokyo and the Visual Culture of Japan's Great Earthquake of 1923 marked the beginning of what has since become an ongoing exploration of that event from many different perspectives by scholars working across an array of disciplines and methodologies, as we’ve seen herein. It is perhaps fitting that this issue concludes with an example of Weisenfeld’s work, which showed us, among many other things, how artists, filmmakers, photographers, and the media that disseminated their materials made the disaster legible in some registers and not others, and with what consequences. Weisenfeld drew on her book for an article that originally appeared in this journal in 2015; in it she surveyed some of the “technologies of seeing” used to interrogate the disaster’s effects, and the narratives they made possible. Weisenfeld has updated that piece for inclusion here with a brief overview of “Imaging Kantō,” an interactive digital complement to Imaging Disaster.

In closing, we hope that these essays will be of use to readers, whether as first steps toward a deeper dive into the growing body of scholarship on the Great Kantō Earthquake, or as resources to draw on amidst the proliferation of claims and denials that are sure to surface in proximity to September 1, 2023 and the disaster’s 100th anniversary. 



Chūō Bōsai Kaigi, Saigai Kyōkun no Keishō ni kansuru Senmon Chōsakai (2006) 1923 Kantō daishinsai hōkokusho (Dai ippen).

Clancey, G.K. (2006) Earthquake Nation: The Cultural Politics of Japanese Seismicity, 1868–1930. University of California Press, Berkeley.

Geller, R.J. (2011) Nihonjin wa shiranai ‘Jishin yochi’ no shōtai. Futabasha, Tokyo.

Geller, R.J. (2018) Gerā-san Nippon ni mono mōsu. Tōkyōdō Shuppan, Tokyo.

Gordon, A. (1991) Labor and Imperial Democracy in Prewar Japan. University of California Press, Berkeley.

Haag, A. (2019) The Passing Perils of Korean Hunting: Zainichi Literature Remembers the Kantō Earthquake Korean Massacres. Azalea: Journal of Korean Literature & Culture, 12, 257–299.

Hasegawa, K. (2020) The Massacre of Koreans in Yokohama in the Aftermath of the Great Kanto Earthquake of 1923. Monumenta Nipponica, 75, 91–122.

Hasegawa, K. (2022) The 8 p.m. Battle Cry: The 1923 Earthquake and the Korean Sawagi in Central Tokyo. The Asia-Pacific Journal: Japan Focus 20(12) No. 4,

Horowitz, A. and Remes, J.A.C. (2021) “Introducing Critical Disaster Studies.” In Critical Disaster Studies (Eds, Horowitz, A. and Remes, J.A.C.), University of Pennsylvania Press, 1–8.

Imaging Kantō (2017) "Fire Ruins of Kanda Main Street." (accessed August 31, 2023).

Kitahara Itoko, Matsuura Ritsuko and Kimura Reo (Eds.) (2012) Nihon rekishi saigai jiten Yoshikawa Kōbunkan, Tokyo.

Satake, K. (2023) Recurrence and Long‐Term Evaluation of Kanto Earthquakes. Bulletin of the Seismological Society of America,

Schencking, J.C. (2013) The Great Kantō Earthquake and the Chimera of National Reconstruction in Japan. Columbia University Press, New York.

Smits, G. (2014) When the Earth Roars: Lessons from the History of Earthquakes in Japan. Rowman & Littlefield, Lanham.




Readers may notice that some analyses and maps of the tectonic setting in the northeast Japan subduction zone describe a smaller segment of the lithosphere, the Okhokst plate, as interacting with the others along the Japan trench and the Sagami trough, not the North American plate. Both labels describe what is for our purposes the same structure.


For overviews of seismicity in the vicinity of the Kantō plain and of the 1923 earthquake specifically, see (Schencking, 2013, 15-20) and (Satake, 2023). (Smits, 2014) is another useful resource on these topics. 


The data on casualties and other damages, and analyses of the hazards associated with the earthquake are drawn from (Kitahara Itoko et al., 2012, 423-433) and (Chūō Bōsai Kaigi, Saigai Kyōkun no Keishō ni kansuru Senmon Chōsakai, 2006).


Some of the more recent work on the post-earthquake violence against Koreans includes (Haag, 2019) (Hasegawa, 2020) and (Hasegawa, 2022).


See (Hasegawa, 2020, 93) for a brief discussion of these figures.


The description of disasters as socially constructed events and ideas is from (Horowitz and Remes, 2021, 2).


Geller’s two trade books for the Japanese market, Earthquake Prediction: What the Japanese Don’t Know (Nihonjin wa shiranai ‘jishin yochi’ no shōtai) and Hey Japan, Mr. Geller Has a Few Things to Say to You (Gerā san Nippon ni mono mōsu) were published in 2011 and 2018, respectively.


For more on the earthquake’s effects on Japanese seismology, see (Clancey, 2006, 225-245).


Asian American Writers’ Workshop, n.d. “Andre Haag.”


Forthcoming in 2024 with University of Pennsylvania Press in its series on Critical Studies in Risk and Disaster.

This introduction is part of the Special Issue: The 100th Anniversary of the 1923 Great Kantō Earthquake. See the table of contents here.


Kerry Smith is an Associate Professor in the Departments of History and East Asian Studies at Brown University. His most recent book, Predicting Disasters: Earthquakes, Scientists, and Uncertainty in Modern Japan, is forthcoming from the University of Pennsylvania Press.