The 8 p.m. Battle Cry: The 1923 Earthquake and the Korean Sawagi in Central Tokyo


June 15, 2022

The 8 p.m. Battle Cry: The 1923 Earthquake and the Korean Sawagi in Central Tokyo
The 8 p.m. Battle Cry: The 1923 Earthquake and the Korean Sawagi in Central Tokyo

Volume 20 | Issue 12 | Number 4

Article ID 5717


Abstract: “Korean sawagi” was the contemporaneous labeling of the rumor-driven commotion and massacres following the Great Kanto Earthquake of 1923. This article examines the Korean sawagi as a staged event through a battle cry exercise in the Imperial Palace Plaza in Tokyo on the night of September 2.

Keywords: Korean sawagi, Kanto Massacre, Great Kanto Earthquake, Imperial Palace, aural domination




In the early stages of scholarship on the Kanto Massacre, the question of whether the anti-Korean rumors triggering the massacres were first manufactured from above by state authorities or arose naturally from the Japanese masses was a central topic of debate. Kang Tŏk-sang and Saitō Hideo argued the former position (kanken-setsu) while Matsuo Takayoshi argued the latter (shizen hassei-setsu).1 While the two camps took opposing positions regarding the origins of the rumors, more significantly, they agreed that state authorities played a central role in diffusing the rumors. The first instance of anti-Korean rumor to arise after the September 1 earthquake will never be known. However, this unknowability is insignificant in assessing the nature of the devastatingly systematic rumor diffusion behind the massacres—a process that Kang, Saitō, Matsuo, and subsequent scholars have recognized as fundamentally state-led in nature.

A related but separate question that has not received sufficient attention concerns the nature of the rumor diffusion process by state authorities. Kang Tŏk-sang is the scholar who has contributed most to furthering our understanding of the state-led rumor-diffusion campaign. In his landmark book, there is an unresolved tension over the portrayal of state officials as, on the one hand, inadvertently setting off a mass panic through their fearful measures against potential colonial unrest, and on the other hand, deliberately manipulating the Japanese masses against Koreans to forestall social unrest.2

As conflicting tendencies existed within the state leadership and the spectrum between panic and opportunism can be fluid even within individuals, this tension cannot be completely resolved. Nevertheless, in our effort to better understand the events following the September 1923 earthquake, it is important to maintain the distinction between a “panic” over Korean attackers that swept through all levels of Japanese society, and a “commotion” (sawagi) deliberately manufactured by state authorities.

Chōsenjin sawagi” (Korean commotion), the contemporaneous naming of the rumor-driven mass agitation and massacres following the Great Kanto Earthquake of 1923, has been shunned by scholars of the massacre due to its dismissive and obfuscatory nature. As Andre Haag has pointed out, the label of Korean sawagi served to shroud the massacre in ambiguity. The term referred to rumored activities as diverse as rioting and carnivalesque merry-making, and Korean sawagi could refer to the rumored acts of violence carried out by Korean people, the violence by Japanese that resulted, or both.3 Due to this problematic ambiguity, “Korean Massacre” (Chōsenjin gyakusatsu) has become the accepted naming and object of inquiry in Japanese language scholarship, while “Korean Panic” has become the most common translation for the broader social responses of the Korean sawagi in English language scholarship.4 Yet, “Korean sawagi” was how the realities of the rumor-driven commotion and massacres were staged and enacted in the terror-filled days after the 1923 earthquake. As such, it requires critical re-examination.

The Korean sawagi was staged by state authorities from the initial aftermath of the 1923 earthquake targeting two complementary policing signs that have generally been misconstrued as oppositional: “unruly Koreans” (futei senjin) and “rumors” (ryūgen higo). The sawagi was preemptively inversive, aimed at forestalling a mass sawagi against the state by inciting a wave of collective violence that was not completely controllable but that was, more importantly, contained within the framework of joint security operations against enemies of the state expressed in these policing signs. It achieved its aim through agitational and silencing effects, and through the integrated process of spreading and quelling rumors of Korean attackers. These mutually constitutive elements of the sawagi manifested themselves in differential configurations and chronologies depending on local circumstances. An exceptionally early display of the agitational-and-silencing effect of the sawagi was the 8 p.m. battle cry (toki no koe) exercise of September 2 in the heart of the imperial capital.


Background to the Korean sawagi

From late 1917, state authorities began reorganizing the police system in Japan, sending police officials on inspection tours to the West to expand their knowledge on policing techniques. One of the officials sent to revolutionary Russia and Europe on these tours was struck by the foreboding sense that Japan would be engulfed by “an awful sawagi” if it failed to make innovations to policing methods.5 Such anxiety was widely shared within the government and was exacerbated by ensuing events: the Rice Riots of 1918, the March First Movement of 1919, and the mass unemployment resulting from the 1920 economic crash. These crises erupted while Korean migrants continued to flow into urban Japan as what Ken Kawashima has called the “uncontrollable colonial surplus” of peasants uprooted by the colonial policy to transform the Korean countryside into a source of cheap rice exports to the metropole.6

One of the ways that the police responded to these developments was preventive policing measures against the futei senjin, or unruly Korean, a figure that became a policing sign that criminalized all Koreans and legitimated preemptive violence under the rationale that their potential secret plots against the Japanese nation were not always identifiable. Preventive policing also entailed extending police work to welfare organizations such as the Sōaikai, a state-funded but officially “private” Korean welfare organization that operated as the “obscene, violent supplement of state power” in policing the boundaries of acceptable Korean behavior. As the inversion of the futei senjin, the Sōaikai cooperated with state authorities after the massacres in placing Koreans under protective custody in its dormitories and having its workers engage in highly publicized volunteer reconstruction work. Such zealously loyal contributions paid off for the Sōaikai, which grew massively in size and established itself as the colonial regime’s prime procurer and manager of Korean day laborers.7

Newspapers and magazines also repeatedly transmitted sensationalist stories about Korean conspiracies to the Japanese public, thereby establishing the futei senjin as folk devil whose manifestation could provoke powerful popular reactions of fear and loathing.8 Through its figuration as a liminally subversive villain freely violating geographical, ethnic, and ideological boundaries beyond the imperial state’s field of vision, the futei senjin aided police officials in their effort to “policify the Japanese masses”9 through their vicarious participation in tantalizingly non-panoptic surveillance over the unruly colonials.10 Media narratives of the futei senjin also established a discursive style that collapsed the distinction between operational and aspirational plots against the state, resulting in the routine production of stories free from the constraints of reality.11 The officialized reports of Korean attackers after the 1923 earthquake arose from this discursive style. With law enforcement temporarily incapacitated by the natural disaster, citizens habituated to the vicarious surveillance over the colonized in print media were now called upon to rise up in the real world as vigilantes to defend their communities against Korean attacks.

After the earthquake struck on September 1, 1923, Japanese state authorities conducted a rumor-based anti-Korean mobilization campaign and triggered the collective violence of the Kanto Massacre. The orders for vigilance in this campaign were disavowed as “rumors” (ryūgen higo) after the massacre; a less common but more accurate label was “propaganda” (senden). Around 8 p.m. of September 2, as battle cries arose from evacuees in the heart of the imperial capital, a naval officer approximately twenty kilometers to the east at the Funabashi Naval Transmission Station went into combat mode after being flooded with official and unofficial reports of Korean riots. Responding to dire reports of Korean troops advancing toward his station, he later transmitted desperate SOS radio messages pleading for reinforcements, only to be later subjected to an internal investigation for his embarrassingly public panic. He plaintively explained to investigators that he was “absolutely not informed of the propaganda etc. being conducted in the Tokyo region until after things had settled down.”12

Military and police leaders, for their part, tended toward reticence regarding their own propaganda operation, portraying themselves as equally uninformed and bewildered by the uncontrollable spread of “wild rumors” of the Korean Panic. A notable exception was the Commissioner of the Tokyo Police Akaike Atsushi, who lauded the “highly effective” activities of his “propaganda forces” that energetically diffused information useful in managing the people’s fears amidst the mass confusion after the earthquake. While the operation in mass psychology was not completely controllable, the outbreak of some unfortunate incidents did not interfere with the ultimately favorable outcome: “The riots and the like that we feared most were almost completely nipped in the bud.”13

During the “era of popular violence” that began with the Hibiya Riots of 1905, the diffusion of public grievances by a sympathetic press played a key role in the mass mobilization process leading to the eruption of city-wide riots. Parodic inversion of state rituals was a common repertoire of protest in these riots and police facilities were frequently targeted for destruction.14 The proliferation of the policing sign of the futei senjin in the mass media at the tail end of this era enabled the re-inversion of riotous repertoires of collective action against the state into a sawagi targeting the accursed other threatening state and society alike.

While the futei senjin figure criminalized all Koreans, the policing sign of “rumors” criminalized all transmissions of information that could potentially lead to subversive collective actions against the state. “As a rule, the proliferation of various rumors precedes the outbreak of riots,” a Justice Ministry study on the Rice Riots of 1918 reported. The state of rumors was the “barometer for monitoring the instability of the people’s hearts (jinshin fuan).”15 From a public security standpoint, it was of crucial importance that preventive policing measures be taken in the “incubation period” before the outbreak of undesirable rumors.16 While it was in 1938 that the above report was submitted to the Justice Ministry, state officials were already implementing counter-riot rumor diffusion during the 1918 Rice Riots, with Home Ministry officials propagandizing through the mass media that the riots were being organized by burakumin agitators who were deceiving the masses through seditious rumors.17 The mobilization of local firefighting brigades, youth groups, and military reservist groups to guard against burakumin-led unrest was also attempted, though with limited success, during these riots.18 While Koreans had yet to be targeted as the primary villain in this mobilization campaign, the governor of Osaka made an early attempt by remarking amidst the riots: “Various rumors are being spread and continue to threaten the people to this day. As for me, I cannot help but suspect that Koreans are behind these rumors.”19

While state authorities were unable to manufacture a full-fledged sawagi against a stigmatized enemy of the people to prevent the outbreak of the Rice Riots, the earthquake of five years later provided them with the opportunity to fully implement this preemptive tactic. The Metropolitan Police reports that it reached a decision at 3 p.m of September 2 to “conduct a grand propaganda operation (daisenden) to prevent the outbreak of rumors and stabilize the people’s hearts.”20 The Home Ministry official Kawamura Teishirō elaborates that the systematic transmission of orders for vigilance against planned anti-Japanese terrorist attacks through military and police channels on September 2 “successfully prevented the outbreak of various rumors and alleviated the instability of the people’s hearts.”21 This simplified narrative of the post-earthquake propaganda campaign elides the collateral damage of the resulting massacres. Nevertheless, it conveys the essence of the larger framework of the Korean sawagi subsuming the Kanto Massacre: an operation to claim control over the people’s hearts through the massive proliferation of fear and hatred against the futei senjin. With this preventive policing measure for rumor management, the state of rumored Korean attackers became the barometer registering the amplified fears of state officials over the people’s hearts. The orders for vigilance against these rumored enemies were “rumors” only insofar as the Sōaikai was “Korean.” Both were instances of state power insinuating itself into criminalized categories to invert them from within, with the resulting preemptive violence subsequently disavowed as unfortunate outbursts of the undisciplined masses.


Agitate and Silence

The official police report of the Great Kanto Earthquake details the emergency security measures taken in the heart of the imperial capital in the immediate aftermath of the earthquake. After the quake struck at 11:58 a.m. of September 1, 1923, the emergency police headquarters was initially set up on the inner moat of the Imperial Palace, outside the front gate of the Metropolitan Police Headquarters building. At 1 p.m., it was moved inside Hibiya Park and emergency operations were started at this time. With the spread of fires throughout the downtown area, Hibiya Park and the plazas fronting Tokyo Station and the Imperial Palace began to be flooded with refugees carrying their possessions. Responding to the perceived threat of disorderly crowds in such sites, the police began preparations for security operations in this area. 120 police cadets led by the head of the training academy joined forces with officers from neighboring police stations who congregated at the emergency headquarters to join these operations.22


The Metropolitan Police Headquarters (September 1, 1923)


Police officials were concerned about the “operation of crowd psychology” among the masses on the streets and took preventive crowd-control measures which included the “transmission of orders” and “propaganda.”23 Around 2:30 p.m., with the Metropolitan Police Headquarters building in flames, a newspaper car in the area was announcing through its loudspeaker: “A riot has broken out in the Sugamo Ōtsuka area. Everybody be vigilant.”24 Shortly thereafter, canon shots rang out from the Imperial Palace and rumors began spreading that Korean riots had broken out at the instigation of socialist agitators as police officers, soldiers, and community volunteers scrambled to contain the fires in the area.25 In Kōjimachi, a neighborhood bordering the western side of the Imperial Palace and located just beyond the reach of the fires, rumors of Korean rioters led to vigilante lynchings of suspected Koreans in the evening.26 Early outbreaks of killings on the night of September 1 are also recorded in accounts from eastern Tokyo and Yokohama.27

In the series of urban riots following the 1905 Hibiya Riots, the police adopted the tactic of dividing its men into uniformed “regular troops” (seihei) and undercover “commandos” (kihei) to conquer riotous crowds.28 Undercover police units were mobilized in the immediate aftermath of the 1923 earthquake as well and their operations, which aimed for the agitational and silencing effects of “arousing the fighting spirit” and preempting subversive actions by “delinquent gang members,” were reported to have produced “exceedingly favorable results.”29 In central Tokyo, the Metropolitan Police mobilized its off-duty officers into undercover surveillance units in addition to patrol troops that circulated around the city in groups of four.30 By the night of September 1, a massive crowd of evacuees was packed into the Imperial Palace Plaza–an alarming situation that the police official Shōriki Matsutarō described as “a sawagi so great we could not even find space to stand” in the plaza.31 Amidst the confusion, “four to five men walking in single file” forced their way through the crowd, repeatedly shouting out warnings about Korean rioters preparing to invade the palace. “Extinguish your lights! Koreans are coming to attack!” The refugees put out their lanterns and candles and awaited motionless in nervous silence.32

The warnings of Korean attackers at times terrorized the evacuees around the Imperial Palace into silence; other times, they agitated them into shouting battle cries.33 These battle cries intermittently piercing through terror-filled silence to fend off the rumored Korean attackers were heard in scattered sites throughout the Tokyo-Yokohama area on this night. In the eastern Tokyo neighborhood of Mukōjima, vigilantes ordered residents to evacuate to their second floors to protect themselves from the effects of a tsunami and Korean bomb-throwing arsonists. As locals moved upstairs and waited in nervous silence, they heard fire bells and battle cries in the distance. A resident recounts that when the commotion reached their neighborhood, “We did as instructed, crying out ‘Wa–!’ with all our might to keep the rioters from coming.”34 Such battle cries were the product of the agitational effect of the Korean sawagi. In Yotsugi, approximately one kilometer east from Mukōjima, the shouting was accompanied by killings punctuated by cries of “Banzai!” from the night of September 1.35

While some residents along the Sumida River moved upstairs to protect themselves from rumored tsunamis and Korean attackers on the night of September 1, others evacuated toward the Imperial Palace. Unlike the following night, the crowd congregated at this site without being ordered to do so by the police. “If we are going to die, we might as well die at our Honored Emperor’s place,” one father told his family as they ignored a police officer’s frantic calls to evacuate in the other direction.36 Another father led his family to the palace plaza and told them that if the fires jumped the moat, they could flee into the palace grounds.37 The kabuki actor Nakamura Kan’emon also found himself in front of the Imperial Palace after evacuating to Hibiya Park from his Tsukiji home in response to tsunami warnings that night. There, soldiers began chasing down suspected Koreans and terrorizing captured suspects. Feeling endangered, Nakamura and his group fled the park and settled down at the outer edge of the Imperial Palace Plaza. Relief was not to be found there either, as the wind-driven conflagration encroached upon the area and threatened to incinerate the large crowd of refugees and their possessions.38 It was in this context, with the overwhelming forces of the natural disaster giving rise to screams among terrified congregated people, that man-made rumors about Korean attackers were prompting battle cries directed against the alien security threat. On this night, the primary security operations in the area centered on the fires not their rumored Korean origins.

Rumors of Korean attackers agitated and silenced Japanese residents and refugees across the Kanto region from the night of September 1. However, the area surrounding the Imperial Palace may have been the only site where these rumors were not only spread, but also quelled, from this night. “Koreans are not coming this way,” a man announced to the crowd after about 30 minutes after a warning of Korean attackers had been shouted out.39 The heart of the imperial capital may also have been the only area where martial law was announced on the night of September 1. According to the detailed recollections of Miki Torirō, a military officer arrived on horseback in the palace plaza along with several cavalry soldiers and shouted out a message to the evacuees on this night. Miki could not make out the officer’s words, but his father explained to him, “Martial law has been declared.”40 In the Kanto region, it was not until around September 5 that the government’s rumor suppression campaign eventually put an end to the Korean sawagi and massacres. In the heart of the imperial capital, this rumor suppression campaign started in the first days after the earthquake together with the exceptionally early implementation of de facto martial law.


The 8 p.m. Battle Cry

On the morning of September 2, in a scene reminiscent of the 1918 Rice Riots, looting broke out at the fancy food store Meiji-ya, located one block away from the Imperial Palace Plaza.41 In the afternoon, Yokohama police officials arrived on foot at the army headquarters in central Tokyo and reported on the complete destruction of their city and its descent into anarchic confusion.42 At night, as the monstrous mass of flames that had dominated the senses of the crowd around the Imperial Palace the previous night moved northward along the Sumida River, the anti-Korean rumors took center stage in central Tokyo and throughout the Kanto region.43

The rumors could terrorize residents and refugees into silence. The writer Takami Jun describes one such scene in Azabu, where people on the streets began shouting out warnings of Korean attackers armed with guns and bombs advancing toward their neighborhood. “Koreans are massacring all Japanese, including women and children!” Residents of the neighborhood retreated into their houses and the streets turned deadly silent. Military reservists, youth group members, and others joined vigilante patrol operations to capture any suspected Koreans amongst those that remained.44 The rumors could also mobilize vigilantes into producing deafening sounds of community alarm and warfare—fire bells, gun shots, and battle cries, accompanied in some areas by massacres.

Takashi Fujitani has pointed out how Foucault’s “monarchical power” and “disciplinary power” emerged simultaneously in modern Japan, producing the effect of “visual domination” with the Meiji emperor not only newly visible as the center of power but also invisible as the anonymous overseer behind the panoptic gaze.45 The Korean sawagi took place during a transitional crisis in monarchical rule, where the visual domination established under the Meiji emperor was incapacitated together with the sickly Taisho emperor, and state authorities were preparing a newly participatory style of imperial rule under Crown Prince Hirohito.46 In addition to the visual domination analyzed by Fujitani and Hara, aural domination, or the mobilization and repression of Japanese subjects through emotive sounds and aura, emerged as a key dynamic during and after the Kanto Massacre. The silencing and agitational effects of the Korean sawagi were mutually constitutive elements that together, in differential configurations depending on local circumstances, produced this dominating effect. The peripheral border zones of Tokyo and Yokohama were the sites of the largest massacres. The heart of the imperial capital was the site of an exceptionally early display of the agitational-and-silencing effect of the sawagi, where the police spread, and subsequently quelled, rumors about Korean attackers in an integrated crowd control exercise.47

The Nihonbashi shop owner Konishi Kihei provides a detailed description of a police action in the Imperial Palace Plaza on the night of September 2.


As it started getting dark, a policeman came and requested that we relocate to the Imperial Palace Plaza. He told us that it was not desirable from a security standpoint that people be scattered in different places. He also told us that there were rumors circulating and we needed to be vigilant as it was a time of instability of the people’s hearts.48


Unlike evacuation orders in other areas, the order Konishi received to relocate to the Imperial Palace Plaza was not accompanied by rumors of Korean attackers. Rather, it was an atypically candid warning about the proliferation of rumors and the potential for mass unrest. Abandoning hopes of sleeping on tatami mats in a nearby building, Konishi and his family joined the flow of refugees into the crowded plaza and camped out under the statue of Kusunoki Masashige. A student in the area recounts that around 5 p.m., a group of police officers assembled and held a meeting. One of the officers later approached his group and prohibited them from lighting fires. He also instructed them to guard vigilantly against Koreans who it was said would be attacking the area at 8 p.m.49 Konishi’s group was told that police and military were battling with rioters around Shinagawa. A group of men circulated through the crowd announcing:


This Marunouchi area is bound to be attacked sooner or later. When the attackers come, we want all the men to take up arms and fight. When we see the malcontents, we will alert you. At that time, everybody in this area needs to shout out in unison a battle cry (toki no koe). Please crush the spirit of the malcontents by showing them how many of us are here.


Around 8 p.m., a thunderous battle cry arose from the crowd in front of the Imperial Palace. “Here they come!” On cue, the evacuees shouted out with all their might as instructed.


“Wa–!” Like a tsunami, the battle cries repeatedly roared in, subsided, and roared in again. The desperately heroic human cries reverberated across the sky, red and black mirroring the conflagration on the ground.


Moments after the final battle cry, a uniformed police officer arrived with his deputies and issued a threatening-and-reassuring warning to the crowd:


Who just shouted? Who was deceived by the ridiculous rumors and shouted a battle cry? The rumors about malcontents coming to attack are completely baseless. Please rest assured and trust us. Do you know where you are? This is the Imperial Palace. If you forget your respect for this place and shout out again, we will arrest you immediately!50


With this, silence returned to the area. This time, it was the silence of a crowd cowering not from the specter of Korean attackers, but rather the self-referential threat of their own hearts being deceived by false rumors.



Earthquakes, tsunamis, and fires could not be conquered by state violence, nor could they be disciplined through state propaganda. Aftershocks gradually subsided after the initial days but only after frequently and randomly striking with complete disregard for the police order for vigilance on the night of September 1 that warned of major aftershocks at 11 p.m. and 3 a.m.51 The dreaded massive tsunamis did not strike the Tokyo-Yokohama region. However, this was not due to the Home Ministry’s authoritative pronouncement timed with the implementation of martial law: “There will be no more major aftershocks and there will absolutely be no tsunamis.”52 The wind-driven conflagrations in the eastern half of the imperial capital grew to massive scales impervious to human intervention and generated uncontrollable movements of evacuees fleeing and congregating across the city, consuming all material that lay in their whimsical paths before extinguishing itself.

Amidst the overwhelming powerlessness of national defense measures in the face of the natural disaster, the 8 p.m. battle cry operated like clockwork. The phantom Korean attackers arrived on time and triggered a rousing battle cry from the Japanese crowd.53 The commotion was subsequently quelled by the police announcement that the rumors of Korean attackers were “ridiculous” and “completely baseless.” However, the resulting silence was exceptionally localized and fleeting. Military, police, and vigilante forces around the Imperial Palace and throughout the Kanto region continued to be mobilized in security operations against rumored Korean attackers into the following days. The ensuing massacres under martial law resulted not so much from a “panic” over Korean attackers as a staged sawagi that preemptively inverted the specter of “Korean riots” to wrest control from the natural disaster the physical and psychological movements of Japanese subjects.




The author thanks Mark Selden, Jae-jung Suh, Sayaka Chatani, Max Ward, and the two anonymous reviewers for their insightful and helpful comments.




Kang Tŏk-sang, “Tsukuridasareta ryūgen: Kanto daishinsai ni okeru Chōsenjin gyakusatsu ni tsuite,” Rekishi hyōron v.157 (1963), 9-21. Saitō Hideo, “Kanto daishinsai to Chōsenjin sawagi,” Rekishi hyōron v.99 (1958), 25-39. Matsuo Takayoshi, “Kanto daishinsaika no Chōsenjin gyakusatsu jiken (jō),” Shisō v.471 (1963), 44-61. Matsuo Shōichi, “Kanto daishinsai-shi kenkyū no seika to kadai,” in Hirakata Chieko and Ōtake Yoneko eds., Seifu kaigenrei kankei shiryō (Tokyo: Nihon Keizai Hyōronsha, 1997), 7-43.


Kang Tŏk-sang, Kanto daishinsai (Tokyo: Chūkō Shinsho, 1975), 43-57. Kang Tŏk-sang, Kanto daishinsai gyakusatsu no kioku (Tokyo: Seikyū Bunkasha, 2003), 57-78.


Andre Robert Haag, “Fear and Loathing in Imperial Japan: The Cultures of Korean Peril, 1919-1923” (PhD diss., Stanford University, 2013), 275.


Haag, “Fear and Loathing in Imperial Japan” Chapter 5; Alex Bates, The Culture of the Quake: The Great Kanto Earthquake and Taisho Japan (University of Michigan Center for Japanese Studies, 2015), Chapter 8.


Obinata Sumio, Keisatsu no shakaishi (Tokyo: Iwanami Shoten, 1993), 122.


Ken Kawashima, The Proletarian Gamble: Korean Workers in Interwar Japan (Durham: Duke University Press, 2009), 44.


Ibid., 134-154.


Haag, “Fear and Loathing in Imperial Japan: The Cultures of Korean Peril, 1919-1923,” 74.


“The massification of the police, and the policification of the masses” was the slogan under which the Japanese police reorganized itself after World War I. See Kawashima, The Proletarian Gamble, 133-38; Max Ward, “Toward a Genealogy of the Police Idea in Imperial Japan: A Synthesis,” International Journal of Asian Studies (2021), 11-13.


For the Foucauldian panoptic gaze in Japan’s modern emperor system, see T. Fujitani, Splendid Monarchy: Power and Pageantry in Modern Japan (Berkeley: University of California Press, 1996), 24-25.


Haag, “Fear and Loathing in Imperial Japan: The Cultures of Korean Peril, 1919-1923,” 78-79. This style had been established in the 1911 High Treason case with anarchists as target. The court proclaimed: “Because of their beliefs we may surmise that they planned to carry out their crimes. Their chief motivation was belief.” Robert Thomas Tierney, Monster of the Twentieth Century: Kōtoku Shūsui and Japan’s First Anti-Imperialist Movement (Oakland: University of California Press, 2015), 127.


Kang Tŏk-sang and Kŭm Pyŏng-dong eds., Gendai shiryō vol. 6, Kantō daishinsai to Chōsenjin (Tokyo: Misuzu Shobō, 1963), 29. The “propaganda” mentioned here referred to what the officer had been falsely informed in the public relations disaster mitigation phase after the massacres: that the rumors of Korean attackers were part of a propaganda campaign by traitorous leftist ideologues seeking to undermine colonial policy.


Akaike Atsushi, “Daishinsai tōji ni okeru shokan,” in Kŭm Pyŏng-dong ed., Chōsenjin gyakusatsu ni kansuru chishikijin no hannō 1 (Tokyo: Ryokuin Shobō, 1996), 213.

[1] Andrew Gordon, Labor and Imperial Democracy in Prewar Japan (Berkeley: University of California Press, 1991), 33,42-50.


Andrew Gordon, Labor and Imperial Democracy in Prewar Japan (Berkeley: University of California Press, 1991), 33,42-50.


Jinshin fuan is a ubiquitous term in state documents on the Korean sawagi. It refers ostensibly to “fear in the people’s hearts” and is synonymous with mass panic. Here and elsewhere, however, the more appropriate translation is the perceived “instability of the people’s hearts,” or elite fears over the people’s hearts.


Shihō Keiji Kyoku ed., Iwayuru kome sōdō jiken (Kyoto: Tōyō Bunkasha, 1974), 226.


Fujino Yutaka, Tokunaga Takashi, Kurokawa Midori eds., Kome sōdō to hisabetsu buraku (Tokyo: Yūzankaku, 1988), 46. Jeffrey Paul Bayliss, On the Margins of Empire: Buraku and Korean Identity in Prewar and Wartime Japan (Cambridge: Harvard University Asia Center, 2013), 134-40.


Ōe Shinobu, “Teikokushugi seiritsuki nōson no mujun,” in Ōe Shinobu ed., Nihon fashizumu no keisei to nōson (Tokyo: Azekura Shobō, 1978), 69-70.


Fujino, Tokunaga, Kurokawa eds., Kome sōdō to hisabetsu buraku, 51. The circular inversion of spreading rumors about Koreans spreading rumors reappeared during the 1923 Korean sawagi in the form of residents of eastern Tokyo massacring Koreans after being forced to flee for their lives by alarming shouts of an impending tsunami and subsequently being told by military reservists and policemen that Koreans were deliberately spreading false rumors of tsunamis to commit burglary. Kanto daishinsai ni gyakusatsu sareta Chōsenjin no ikotsu wo hakkutsu shi tsuitō suru kai ed., Kazeyo hōsenka no uta wo hakobe (Tokyo: Kyōiku shiryō shuppankai, 1992), 46-47; Nicchō kyōkai Toyoshima shibu ed., Minzoku no toge (Tokyo: Nicchō kyōkai Toyoshima Shibu, 1973), 39.


Keishichō ed., Taisho daishin kasaishi (Tokyo: Keishichō, 1925), 459.


Kawamura Teishirō, Kankai no hyōri (Tokyo: Kawamura Teishirō 1933), 170.


Keishichō ed., Taisho daishin kasaishi, 18-20.


Ibid., 24-25.


Tokyo-to Shinagawa-ku ed., Daijishin ni ikiru (Tokyo: Shinagawa-ku, 1978), 149.


Teishin-shō, Taisho daikasai tsūshin (Tokyo: Teishin-shō, 1923), 4; Matsune Tōyōjō, “Taishin risaiki,” Shibugaki (October 1923), 13.


Tokyo-shi gakumu-ka ed., Tokyo shiritsu shōgaku jidō shinsai kinen bunshū (Tokyo: Baifūkan, 1924), 7.


Kantō daishinsaiji ni gyakusatsu sareta Chōsenjin no ikotsu wo hakkutsushi tsuitō suru kai ed., Kazeyo hōsenka no uta wo hakobe (Tokyo: Kyōiku shiryō shuppankai, 1992), 46-49; Kenji Hasegawa, “The Massacre of Koreans in Yokohama in the Aftermath of the Great Kanto Earthquake of 1923,” Monumenta Nipponica v.75.1 (2020), 98.


Obinata, Keisatsu no shakaishi, 111-112. See also Miyachi Tadahiko, Shinsai to chian chitsujo kōsō (Tokyo: Kurein, 2012), 33,173,228. Miyachi categorizes this anti-riot tactic in the hardline “harsh punishment” law enforcement approach (genbatsu shugi) while highlighting the significance of the alternative preventive policing approach of “benevolent guidance” (zendō shugi). While his book contains important insights into the conflicts among state officials before, during, and after the Kanto Massacre, I do not agree with its characterization of the Korean sawagi as the product of mass panic, with only aberrational instances of state involvement.


Yasukōchi Asakichi, “Shinsai jōkyō hōkoku,” v.2, 326, Kanagawa Prefectural Archives.


Naimushō, “Shinsaigo ni okeru keikai keibi ippan,” (1923), 19.


Shōriki Matsutarō, “Sono hi no keishichō,” in Kŭm Pyŏng-dong ed., Chōsenjin gyakusatsu ni kansuru chishikijin no hannō 2 (Tokyo: Ryokuin Shobō, 1996), 468.


Nishizaki Masao ed., Kanto daishinsai Chosenjin gyakusatsu no kiroku (Tokyo: Gendai Shokan, 2016), 321.


Abe Kojirō, Sakana ichidai: Abe Kojirō jiden (Tokyo: Ginrinkai, 1969), 40; Iwasaki Katsumi, “Kudan saka ue kara jigoku wo mita hito tachi,” Masukomi jânarizumu kenkyū v.3 (1995), 47.


Tokyo-to Shinagawa-ku ed., Daijishin ni ikiru, 31-32.


Kantō daishinsaiji ni gyakusatsu sareta Chōsenjin no ikotsu wo hakkutsushi tsuitō suru kai ed., Kazeyo hōsenka no uta wo hakobe, 46-49.


Murata Kimi, Watashi no jinsei kaidō (Tokyo: Fudan-ki zenkoku gurūpu, 1974), 4.


Miki Torirō, Seishun to sensō to koi to (Tokyo: Heibonsha, 1994), 55.


Nakamura Kan’emon, Jinsei no hanbun (Tokyo: 1959), 71-72.


Kitazono Kōkichi, Taishō Nihonbashi Hon-chō (Tokyo: Seiabō, 1978), 202.


Miki, Seishun to sensō to koi to, 56.


Meiji-ya honsha ed., Meiji-ya 73 nen-shi (Tokyo: Meiji-ya Honsha, 1958), 55.


Nishizaka Katsuto, “Shinsai issūnen no omoide,” Keisatsu kyōkai zasshi 290 (October 1924), 47.


Matsune Tōyōjō, “Taishin risaiki,” Shibugaki (October 1923), 18.


Nishizaki ed., Kantō daishinsai chōsenjin gyakusatsu no kiroku, 411-12.


Fujitani, Splendid Monarchy, 24-25, 142-45.


Hara Takeshi, Kashika sareta teikoku (Tokyo: Misuzu Shobō, 2011), 276.


For the spatial configuration of repression and reform in the post-massacre Peace Preservation Law apparatus in colonial Korea and the Japanese metropole, see Max Ward, Thought Crime (Durham: Duke University Press, 2019), 57-74,126-129,155-160.


Konishi Kihei, Kantō daishinsai no omoide (Tokyo: Konishi Kihei, 1925), 33.


Chūō Shōgyō Gakkō ed., Kugatsu tsuitachi (Tokyo: Sankōsha, 1924), 126.


Konishi, Kantō daishinsai no omoide, 34-36.


Tokyo-shi gakumu-ka ed., Tokyo shiritsu shōgaku jidō shinsai kinen bunshū, 8.


Kawasaki-shi ed., Kawasaki-shi shi shiryō-hen 3 (Kawasaki: 1990), 686.


For how historical denialism of the imperial period in contemporary Japan has operated as a battle cry exercise with the notion that “the side that shouts louder wins” irrespective of facts, see Yamaguchi Tomomi, Nōgawa Motokazu, Tessa Morris-Suzuki, Koyama Emi eds., Umi wo wataru “ianfu” mondai (Tokyo: Iwanami Shoten, 2016), 9.

Share with a colleague:

Volume 20 | Issue 12 | Number 4

Article ID 5717

About the author:

Kenji Hasegawa

Kenji Hasegawa is Associate Professor of Modern Japanese History and Coordinator of International Student Exchange at Yokohama National University. He is the author of Student Radicalism and the Formation of Postwar Japan (Palgrave Macmillan, 2019).

The Asia-Pacific Journal: Japan Focus is a peer-reviewed publication, providing critical analysis of the forces shaping the Asia-Pacific and the world.

    About the author:

    Kenji Hasegawa

    Kenji Hasegawa is Associate Professor of Modern Japanese History and Coordinator of International Student Exchange at Yokohama National University. He is the author of Student Radicalism and the Formation of Postwar Japan (Palgrave Macmillan, 2019).


    Our monthly newsletter provides readers with an in-depth analysis of forces shaping the Asia-Pacific and the world.

      Since 2002

      Asia Pacific Journal: Japan Focus has produced critical reporting on geopolitics, economics, history, environment, and international relations.