Guangxi Massacre

The Guangxi Massacre refers to a series of mass killings in the Guangxi Zhuang Autonomous Region of China, principally during 1967–1968, in the Cultural Revolution. The violence encompassed the killing of defeated factional opponents and civilians designated as political or class enemies. Research based on subsequent official investigations documents the participation of local officials, militia organizations, and armed factions, and distinguishes massacres of noncombatants from deaths in factional warfare. Guangxi experienced an exceptionally high death toll within the Cultural Revolution. (cambridge.org)

The events also included documented cannibalism, notably in Wuxuan County in 1968. Investigations conducted after the Cultural Revolution recorded killings and led to criminal proceedings and disciplinary sanctions against some participants. These investigations, local gazetteers, and survivor interviews form the principal evidence for reconstructing the violence. The history of the massacres involves the interaction of national political campaigns, regional struggles for authority, and local persecution; its documentation requires careful distinctions among recorded deaths, estimates, and the different geographical and chronological limits used by sources. (cambridge.org)

Political and Social Conditions before the Massacres

Rural political divisions in Guangxi rested partly on a class-status system established through land reform and subsequent political campaigns. Households had been classified as landlords, rich peasants, middle peasants, or poor peasants; later campaigns identified additional individuals as “counterrevolutionaries” or “bad elements.” Landlords, rich peasants, counterrevolutionaries, and bad elements were collectively known as the “Four Types.” In Quanzhou County, early-1950s land reform classified 10,110 families as landlords and 3,279 as rich peasants, according to the county gazetteer cited by sociologist Yang Su. These classifications supplied local authorities with an established population of designated political enemies before the Cultural Revolution began. (escholarship.org)

Class designation was not equivalent to evidence of an individual’s conduct. The administrative system distinguished family origin (jiating chushen), personal status (geren chengfen), and political performance (biaoxian), although local practice did not always follow these distinctions consistently. A person’s family background could affect admission to universities, recruitment into the army, and membership in the Chinese Communist Party. Children born after land reform could inherit the stigma of a landlord or rich-peasant household, despite having had no part in its earlier economic activities. Such labels therefore identified an attributed social and political background, not demonstrated participation in opposition or organized resistance. (cambridge.org)

Rural administration combined civilian Party authority with an extensive militia structure. In Quanzhou, where approximately 93 percent of the county’s 485,000 inhabitants were rural in 1966, the administrative hierarchy ran from county to commune and then production brigade, broadly corresponding to township and village levels. Party organizations connected these units to higher authorities. Alongside this civilian hierarchy stood the military districts, military subdistricts, and county People’s Armed Departments. A county department was headed by a serving People’s Liberation Army officer, and its subordinate branches commanded citizen militia. Village militia consequently belonged to an institutional chain of command rather than constituting merely informal groups of armed neighbors. (escholarship.org)

The transition into the Cultural Revolution initially preserved much of the machinery of the Socialist Education Movement, also called the Four Cleans campaign. In Guangxi, existing campaign committees were converted into Cultural Revolution leading groups, while senior Party officials initially directed investigations into their own subordinate organizations. In Lingui County, near Guilin, the Socialist Education Movement was still active in June 1966; its work team became a Cultural Revolution leading group and dispatched investigators into Party offices, government departments, schools, and communes. During July and August, approximately 500 people, including 20 county-level officials, were targeted as “antisocialist elements.” In Wuming County, loyalty examinations between September and November placed 58 cadres under formal investigation, of whom 29 were designated for public struggle sessions. Political scrutiny thus reached officials as well as households carrying older class labels. (sociology.stanford.edu)

Village relationships added another layer of affiliation. Yang Su identifies surname lineages and inherited intergroup antagonisms as possible influences on local mobilization, while emphasizing that such divisions alone were insufficient to produce collective killing. These kinship relationships cannot, however, be treated as a uniform Guangxi social structure. In the mid-1960s, approximately 60 percent of the region’s population was classified as Han and slightly more than one-third as Zhuang. Andrew G. Walder estimates that Hakka communities constituted about 9 percent of Guangxi’s population, with only two counties having a narrow Hakka majority. Guangxi therefore differed substantially from neighboring Guangdong’s more extensive Hakka settlement areas; local lineage, linguistic, and settlement relationships require locality-specific evidence rather than a single model of southern Chinese village society. (cambridge.org)

Factional Formation and the Struggle for Regional Authority

Challenges to Guangxi’s Chinese Communist Party leadership arose within the bureaucracy as well as among students and workers. During the Cultural Revolution’s first months in 1966, officials investigated subordinates and organized “Scarlet Guards” to protect administrative offices from outside rebels. The political balance changed after Chen Boda’s October 16 speech condemned efforts to obstruct rebellion as a “bourgeois reactionary line.” Previously accused cadres sought rehabilitation, while loyal subordinates risked being implicated alongside their superiors. Andrew G. Walder’s reconstruction identifies both processes as sources of cadre rebellion, rather than treating officials as a uniformly defensive group. (sociology.stanford.edu)

By late November 1966, Lingui County’s administration contained more than 50 rebel fighting groups, enrolling 520 of its 883 cadres and staff. In Qinzhou, former rural work-team members turned against superiors who had ordered investigations and subsequently required them to submit to criticism. Cadre rebels participated in the January 23, 1967 seizure of Guangxi’s regional administration alongside student and worker allies; elsewhere, bureaucratic rebels acted independently. In Nanning Prefecture, distinct from the municipal and regional administrations, cadres expelled their leaders on January 25 without student or worker participation. These seizures displaced existing leadership without establishing stable successor authorities. (sociology.stanford.edu)

The resulting coalitions split over military intervention and Wei Guoqing, Guangxi’s first Party secretary since 1961, whom Beijing appointed to head military control in March 1967. Zhou Enlai explicitly invoked Guangxi’s strategic role in supporting Vietnam. On April 19, Wu Jinnan and other former regional Party committee members opposed Wei’s appointment; an April 22 demonstration supplied the opposition alliance’s name. Its rival, the United Headquarters (lianzhi, also translated “Allied Command”), supported Wei. These were umbrella alliances: local disputes and membership differed, but factions favored by county People’s Armed Departments generally joined United Headquarters, while excluded rivals aligned with April 22. (cambridge.org)

Beijing negotiations from May to November 1967 produced a compromise retaining Wei atop a preparatory committee for a new revolutionary committee. It quickly unraveled. County military officers and United Headquarters forces used militia networks against rivals; reports of rural suppression reached Nanning in late November. April 22 organized protests and seized weapons, and armed confrontation intensified into 1968. Factional combat and the killing of captives or civilians were distinct forms of violence, although they overlapped chronologically. (cambridge.org)

Escalation and Suppression in 1967–1968

Collective killing preceded the major suppression campaigns of summer 1968. After negotiations in Beijing, a November 1967 settlement required Guangxi’s rival alliances—the United Headquarters, also translated as Allied Command, and April 22—to cooperate in preparing a revolutionary committee under Wei Guoqing. Reports reaching Nanning late that month described armed suppression of April 22 supporters, including more than 200 deaths in three Yulin Prefecture counties. Further suppressions in January–February 1968 killed 392 people in Fangcheng and 278 in Hepu. These episodes were not simply casualties of battles between comparably armed forces: they accompanied efforts to weaken opponents before new local administrations were established. (cambridge.org)

From March 1968, the establishment of local revolutionary committees became an occasion for further repression. Military authorities relied on county People’s Armed Departments and rural militia to overcome April 22 resistance. On July 3, a central directive approved by Mao Zedong ordered resistance crushed and characterized continued defiance as counterrevolutionary sabotage, including sabotage of assistance to Vietnam. The directive provided political authorization for a substantially broader offensive; the July 25 meeting in Beijing therefore reinforced a campaign already under way rather than initiating the killings. (cambridge.org)

The alleged conspiracy supplied a mechanism for widening the campaign beyond armed opponents. On June 17, Guangxi’s military authorities and the preparatory committee announced the discovery of a purported anti-Communist network, variously translated as the “Anti-Communist Party Patriotic Army” or “Patriots against the Communist.” They alleged that its headquarters were in Guangzhou and its subordinate organization in Guangxi. United Headquarters propaganda then identified April 22 leaders with this supposed network and called for action against them. These were contemporary accusations, not established evidence of an actual organization coordinating the opposition. Yang Su’s research describes their propagation into communes and villages, where allegations of imminent attacks by class enemies helped justify preemptive violence. (escholarship.org)

At 1 a.m. on July 25, Zhou Enlai, Kang Sheng, and six other senior central leaders met Guangxi factional representatives in Beijing’s Great Hall of the People. Their speeches favored United Headquarters over April 22. Zhou and Kang affirmed the existence of the alleged Anti-Communist Party Patriotic Army and portrayed it as extending beyond Guangxi, without presenting concrete evidence. As Su emphasizes, this treatment endorsed the regional leadership’s depiction of factional opponents as participants in a counterrevolutionary conspiracy, rather than treating both alliances as equally legitimate participants in the Cultural Revolution. (cambridge.org)

Killings intensified in July and August as the remaining resistance was suppressed. Guangxi’s regional Revolutionary Committee was established on August 26, 1968, but administrative consolidation did not mark an immediate end to lethal persecution. The subsequent investigations specifically examined a wave of mass killings in August–September. During the mid-1980s, the Guangxi Communist Party committee commissioned investigations into severe factionalism and these killings; the resulting local chronologies, incident reports, and statistics document violence continuing across the transition to the new regional authority. (cambridge.org)

Organization of the Killings and Selection of Victims

Village militia connected local killing operations to Guangxi’s military and administrative hierarchy. County People’s Armed Departments, headed by serving officers of the People’s Liberation Army, commanded militia through subordinate branches. Their alliance with the United Headquarters faction—also translated as Allied Command—gave that faction organized forces extending into rural communities. Andrew G. Walder’s reconstruction identifies revolutionary committees, public-security departments, commune and brigade leaders, militia, and factional organizations among the perpetrators. These roles were not interchangeable: military authorities supplied command structures and political backing, while local cadres and armed groups participated in operations. Evidence of administrative authority does not, by itself, establish personal participation in a particular execution. (cambridge.org)

Victim selection combined factional identification with imposed household status. Defeated April 22 adherents and people associated with them were principal targets, but accusations of counterrevolution also encompassed families already stigmatized as enemies of the regime. In Walder’s event dataset, more than half of those killed were identified with April 22; the categories nevertheless overlapped, and local patterns varied substantially. Political association must therefore be distinguished from inherited family classification: membership in a stigmatized household did not demonstrate participation in an armed organization, or even support for one. (cambridge.org)

The extension of violence beyond active opponents included women, children, and older people. Yang Su’s Collective Killings in Rural China during the Cultural Revolution documents public executions of such noncombatants from local archives, government records, and survivor interviews. A concrete example discussed in his earlier research concerns Quanzhou: former landlord Liu Xianyan, his wife of poor-peasant background, and their children aged one and three were caught in a collective killing directed by militia. The case illustrates how an imposed family identity could override differences within a household, rather than recording a shared political activity by its members. (cambridge.org)

Organization could involve preparatory meetings, encirclement, detention, and subsequent killing. Su’s account of Fengshan records a joint meeting of public-security officers on August 18, 1968, followed by mobilization through county armed departments. Approximately 4,400 United Headquarters personnel surrounded members of the dissenting “7.29” organization who had fled into the Nanshan and Beishan areas. More than 10,000 people were detained; the cited account attributes 1,016 deaths to the siege and subsequent detention together, not exclusively to executions. Fengshan’s revolutionary committee was established on August 25. This sequence distinguishes the mobilization and capture of opponents from their treatment after detention. (escholarship.org)

Public denunciation also provided a setting for lethal violence, rather than merely a preliminary to imprisonment. Andrew G. Walder and Yang Su’s 2003 study, citing the Yangshuo county annals, reports that 2,513 people underwent violent struggle sessions during one month in 1968 and that 639 were beaten to death. The same study distinguishes accounts of factional fighting from imprisonment, torture, and collective executions in Guangxi’s published regional chronology. These documented forms of coercion cannot all be treated as battlefield casualties: victims included people already detained or publicly assembled under the control of their persecutors. (cpb-us-e2.wpmucdn.com)

Geographical Distribution and Documented Local Episodes

The documented localities extended across central and northeastern Guangxi rather than forming a single cluster around the regional capital. In the prefectural geography used by Andrew G. Walder and Yang Su, Binyang and neighboring Shanglin belonged to Nanning Prefecture; Mengshan and Zhongshan belonged to Wuzhou Prefecture, with Zhongshan adjoining Hunan; and Wuxuan lay along the southern edge of Liuzhou Prefecture. These county jurisdictions must be distinguished from their county seats and surrounding villages. Nanning itself housed three separate administrative headquarters—those of the autonomous region, Nanning Municipality, and Nanning Prefecture—so references to the prefecture do not necessarily describe events within the city. (cpb-us-e2.wpmucdn.com)

Binyang provides a particularly concentrated example of rural killing. Walder and Su’s 2003 study cites pages 111–116 of the Guangxi wenge dashi nianbiao (Chronology of the Cultural Revolution in Guangxi), published by Guangxi People’s Publishing House in 1990, for 3,681 executions during one ten-day period in summer 1968. Almost all those killed were village residents. The figure therefore concerns a specified episode within the county, not deaths in Binyang’s county seat alone, the county’s entire Cultural Revolution death toll, or a regional total. Its geographical significance is that extensive killing reached the village population, rather than remaining confined to urban factional strongholds. (cpb-us-e2.wpmucdn.com)

The accounts of Mengshan offer a different combination of town-centered and village-level evidence. Donald S. Sutton, drawing on Zheng Yi’s investigation, describes 850 people killed, mostly by militia in mid-June 1968 and mostly in one area of the county town. He also recounts a rural household episode, based partly on interviews with surviving witnesses, in which a man classified as a landlord and his sons were killed while his wife survived. For Zhongshan and Binyang, Sutton identifies killings conducted through mass denunciation meetings; for parts of Mengshan and Rong’an, he notes cases in which male family members were killed but wives were spared. These accounts document different local settings and victim-selection patterns, not a uniform sequence reproduced across every county. (gwern.net)

Shanglin and Zhongshan also illustrate the uneven detail preserved in different sources. Walder and Su’s comparison of Zheng Yi’s accounts with published county annals lists 1,906 deaths against 171 for Shanglin, 850 against nine for Mengshan, and 625 against 63 for Zhongshan. These are comparisons between reported county-level figures, not counts for equivalent ten-day episodes such as Binyang’s. Wuxuan, by contrast, shows close agreement—524 against 526—although agreement in totals does not establish that every locality or incident was described equally fully. The same study notes that county annals generally organize narratives around county institutions and towns, while their coverage of individual communes, production brigades, and villages varies considerably. (cpb-us-e2.wpmucdn.com)

Cannibalism in Wuxuan

Documented cannibalism in Wuxuan County occurred between May and July 1968, during the Cultural Revolution. Donald S. Sutton identifies a May 4 incident in Tongwan commune, preceding the defeat of the local April 22 faction on May 11–13. Subsequent incidents accompanied denunciation meetings and executions across the county town and rural communes; they were not confined to battlefield deaths. These county-level cases must be distinguished from Guangxi’s much larger regional death toll. (gwern.net)

The evidence includes a July 4, 1983 list naming 64 victims and an internal report dated May 1987, reproduced or cited by Zheng Yi and examined by Sutton. These records identify victims and locations rather than merely repeating anonymous allegations. Zheng also interviewed victims’ relatives and people acknowledging participation. Sutton rejected Zheng’s extrapolation that 10,000 Wuxuan residents had eaten human flesh as unconvincing; documented cases do not establish that estimate. (gwern.net)

A named victim was Wu Shufang, a geography teacher at Wuxuan Middle School. An Agence France-Presse investigation published in May 2016 described students killing Wu after denunciations, followed by the consumption of organs at the school. Its account also reported that another teacher had been compelled at gunpoint to participate in the handling of the body. The episode illustrates a collective institutional setting and the presence of coercion, rather than an isolated act involving only a killer and a victim. AFP interviewed a former senior investigator who reported at least 38 victims in Wuxuan. That attributed minimum and Sutton’s documentary list are different reported counts, not figures to be added together. (scmp.com)

The same former investigator attributed the cannibalism to intensified class struggle and expressions of political hatred. AFP also interviewed scholar Ding Xueliang, who distinguished the episodes from famine cannibalism: his explanation emphasized political events, ideology, hatred, and ritual rather than food shortages. These retrospective explanations concern the political setting; neither establishes an identical personal motive for every participant. (scmp.com)

Sutton’s January 1995 study, “Consuming Counterrevolution,” interprets the communal meals as extensions of the political ritual of denunciation, execution, and collective reaffirmation. He emphasizes participation by cadres and Communist Party members, not simply rebellious youths. This is an explanatory interpretation of patterned conduct, not proof of a universally shared motive. He also cautions that an inherited Zhuang cannibalistic tradition was unestablished and that participants were not identified specifically as Zhuang. (gwern.net)

Death Tolls and Statistical Limits

Published totals for Guangxi measure different categories of mortality, not a single, consistently defined massacre toll. Andrew G. Walder’s Rebellion and Repression in China, 1966–1971 compares deaths recorded in published local histories with a larger total from internal investigations. Its category of politically induced “unnatural” deaths is broader than executions or killings of noncombatants. The principal figures therefore require their source, geographical coverage, and counting period to remain attached to them. (sociology.stanford.edu)

The first figure appears in Walder and Su’s 2003 discussion; the middle pair comes from Walder’s comparative table; the final figure comes from his later reconstruction using the internal investigations. They are overlapping accounts, not quantities to add together. (cpb-us-e2.wpmucdn.com)

The statistical distinction between deaths and missing persons is particularly important. Walder’s 2022 article reports nearly 20,000 people whose disappearance remained unresolved, separately from the 89,810 confirmed deaths. His suggestion that the total could approach 100,000 assumes that approximately half the missing had died; it is an estimate, not an additional enumeration. Likewise, combat fatalities, executions, suicides associated with persecution, and other deaths grouped under broad political-mortality headings cannot all be relabelled massacre victims. A massacre-specific total requires evidence about the circumstances of death, rather than subtraction from an undifferentiated regional aggregate. (cambridge.org)

Differences between local gazetteers and internal reports reflect selective publication as well as uneven detail. Walder’s coding rules recorded zero when an account supplied no numerical toll, even if it described extensive violence. His published-annals total represented 62.9 percent of the corresponding internal investigation count. This measures the completeness of published reporting, not the proportion of deaths that investigators could identify by name. Administrative boundary changes also required reconciliation to prevent omissions and double counting. (sociology.stanford.edu)

County comparisons show why a uniform correction factor is unreliable. Walder and Su’s 2003 table contrasted 3,951 deaths in other accounts of Binyang with 40 recoverable from its gazetteer, and 850 in Mengshan with nine. Wuxuan, however, showed close agreement: 524 against 526. Some comparison figures covered only a short period, further limiting equivalence. Aggregate numbers in such tables are not necessarily published nominal lists of victims. Nor do a few accessible counties constitute a representative sample from which missing regional deaths can simply be extrapolated: the completeness of reporting differed sharply even within Guangxi. (cpb-us-e2.wpmucdn.com)

Post–Cultural Revolution Investigations and Sanctions

The early-1980s investigations proceeded through both county inquiries and interventions directed from Beijing. In Wuxuan, an investigation team examined the killings and cannibalism in 1981–1983, more than a decade after the events. Donald S. Sutton’s 1995 study records that investigators compiled a list of people who had been eaten and identified a number of leading participants. This county inquiry overlapped chronologically with, but should not be treated as identical to, the wider investigation of violence during the Cultural Revolution throughout Guangxi. Its documented sanctions concerned Wuxuan rather than the autonomous region as a whole. (cambridge.org)

According to Andrew G. Walder’s reconstruction, the Chinese Communist Party Central Committee dispatched two successive waves of investigators after evidence emerged that Guangxi officials were concealing deaths and responsibility for them. In 1981, twenty teams drawn from agencies in Beijing bypassed regional authorities, examined city and county Party and government archives, and interviewed more than 700 people. Those interviewed included officials, ordinary residents, and prisoners and labor-camp inmates. A second intervention in 1983 brought higher-ranking investigators and a more extensive examination of archives and survivor testimony. Across the Beijing-led investigations, more than 100 officials employed close to 100,000 local personnel to prepare reports covering every county and city. These were region-wide investigative operations, not simply an enlargement of the Wuxuan cannibalism inquiry. (cambridge.org)

The investigations fed into the Guangxi Party Committee’s mid-1980s rectification campaign, which sought to establish the causes of factional conflict and the mass killings of August–September 1968 and to punish those responsible. The resulting documentation combined detailed reports on major incidents, extended local chronologies, and accompanying statistics. Materials were assembled for the regional administration and its cities, prefectures, and counties. The eighteen-volume Guangxi “wen’ge” dang’an ziliao—Archival Materials on the Guangxi Cultural Revolution—was compiled under the regional Party committee’s rectification leading-group office and dated 1987 in Walder’s 2016 bibliography; his later account specifies that volumes were issued in 1987–1988. Most exceeded 500 pages. Classified as secret, the collection recorded local events and punishments rather than constituting a publicly released judicial accounting. (sociology.stanford.edu)

For Wuxuan, Sutton reports fifteen people jailed and 130 Party members and cadres disciplined. The two figures describe different forms of sanction: imprisonment belongs to criminal punishment, while Party discipline and administrative sanctions concern organizational membership or official employment and cannot be counted as additional criminal convictions. Sutton’s aggregate disciplinary figure does not itself supply a breakdown between those categories or establish whether the groups overlapped. Nor does it specify individual offenses and sentence lengths. The fifteen and 130 therefore remain county-specific outcomes associated with the Wuxuan investigation, not totals for prosecutions or disciplinary action across Guangxi. The regional archival collection separately recorded sanctions against officials, Party members, and other participants in the localities it covered. (cambridge.org)

Documentary Sources and Public Knowledge

The most detailed administrative evidence comes from retrospective investigations commissioned by the Guangxi regional committee of the Chinese Communist Party during the mid-1980s. Their findings were assembled in the eighteen-volume Guangxi “wen’ge” dang’an ziliao (Archival Materials on the Guangxi Cultural Revolution), completed in 1987 and classified as secret. The collection contains reports on major incidents, extended chronologies, and statistics for the regional administration and its subordinate prefectures, cities, and counties; most volumes exceed 500 pages. Andrew G. Walder’s 2016 study demonstrates how these files identify officials, organizational affiliations, and sequences of action that published accounts often compress into a few sentences. They are post-event investigative compilations, however, rather than an unmediated record produced during the killings themselves. (sociology.stanford.edu)

Published local gazetteers, or difangzhi, provide broader geographical coverage but generally less detail. Their treatment of the Cultural Revolution was shaped by political review: national guidelines issued in 1985 prescribed brief treatment of sensitive political mistakes, with material dispersed among different sections. Researchers therefore have to combine entries in major-event chronologies with chapters on Party organization, public security, government, and political campaigns. Regional chronologies offer another scale of reconstruction. The Chronology of the Cultural Revolution in Guangxi, published in 1990, supplies dated entries connecting local incidents to the regional sequence; Walder and Yang Su used it to document the concentrated executions in Binyang in summer 1968. Such a chronology establishes timing and reported scale without necessarily supplying the underlying witness statements or individual case files. (cpb-us-e2.wpmucdn.com)

*Scarlet Memorial* brought local narratives to a substantially wider readership. Zheng Yi’s Chinese-language Hongse jinianbei appeared in Taipei in 1993; Westview published the English edition, Scarlet Memorial: Tales of Cannibalism in Modern China, in 1996. Zheng had travelled to Guangxi in May 1986 to investigate reports of cannibalism. The English edition, translated and edited under the collective pseudonym T. P. Sym, was abridged selectively: according to its editorial note, the sections documenting Guangxi incidents were translated in full apart from minor modifications, while much of the political and cultural discussion was compressed. Walder and Su subsequently identified the book’s documented cannibalism accounts as an important source of public awareness of rural violence. (api.pageplace.de)

Later scholarship combines these narrative and administrative sources rather than treating either as sufficient alone. Yang Su’s 2011 Collective Killings in Rural China during the Cultural Revolution draws on local archives, government documents, and survivor interviews in Guangxi and Guangdong. Retrospective testimony provides a different perspective from an administrative chronology: it records remembered experiences and local relationships, while dates, official affiliations, and reported totals can be compared with documentary entries. Su’s chapter “On the Record” explicitly distinguishes the relatively systematic but censored gazetteers from more revealing internal documents that reached researchers through occasional leaks. Uneven access consequently affects which counties can be reconstructed in depth. A figure reproduced in a public gazetteer may originate in an earlier investigation, rather than constitute an independent count; its evidentiary value depends on its provenance and reporting scope, not simply on the number of later publications repeating it. (cambridge.org)

Established Findings and Historical Explanations

The documented relationship between regional repression and local killing is more specific than a general association between political disorder and violence. Investigations conducted in the 1980s show that village militia participated in the suppression of rebel organizations across Guangxi in 1967–1968 and also killed large numbers of noncombatants. The victims therefore cannot be understood simply as casualties of fighting between armed factions. Andrew G. Walder’s reconstruction distinguishes factional warfare from the killing associated with the defeat and suppression of rebels: the latter extended coercion to people who were not combatants. This evidence establishes an organizational connection between repression and massacres without demonstrating identical intentions among military authorities, local cadres, militia members, and other participants. (cambridge.org)

Yang Su’s Collective Killings in Rural China during the Cultural Revolution (2011) proposes a “community model” to explain why collective killing occurred in some localities but not in otherwise comparable communities. Su contrasts this with a “state-policy model,” in which a central policy of elimination is implemented through a coordinated administrative or military apparatus. He argues that the rural killings of 1967–1968 did not require such a nationwide extermination policy. Instead, state mobilization and institutional breakdown interacted: political campaigns supplied categories of enemies and mobilized potential perpetrators, while weakened restraints allowed local actors to turn political conflict into collective killing. The model’s principal explanatory task is geographical variation across provinces, counties, townships, and villages, rather than the assumption that one centrally specified mechanism operated everywhere. (cambridge.org)

In Su’s account, local social relationships helped determine how political categories were applied. His discussion of lineage rivalries treats historically formed group boundaries as conditions that could facilitate violence, not as sufficient causes in themselves. Other mechanisms included the construction of “class enemies,” the weakening of legal protections, and the presentation of political conflict as an existential war. The Guangxi meeting of July 25, 1968, provides a concrete example of the last mechanism: Zhou Enlai and Kang Sheng favored the United Headquarters alliance and announced an alleged counterrevolutionary conspiracy called the Anti-Communist Party Patriotic Army. Su treats such political framing as a context in which local killing could emerge, rather than evidence that every subsequent incident followed a single operational directive. (cambridge.org)

Walder’s Anatomy of a Regional Civil War: Guangxi, China, 1967–1968, published in the Spring 2022 issue of Social Science History, shifts the analytical emphasis toward province-wide suppression. Drawing on the post–Cultural Revolution investigations, he argues that Guangxi’s extreme mortality resulted from a counterinsurgency campaign in which village militia were important agents. Whereas Su foregrounds communities, inherited class status, and variation in local participation, Walder emphasizes the regional political and military process connecting dispersed killings. The disagreement concerns the relative explanatory weight of community dynamics and organized suppression; militia participation and the killing of noncombatants are evidence to be explained, rather than mutually exclusive alternatives. (cambridge.org)

For the study of rural political violence, Guangxi also changes the choice of analytical unit. Su’s research combines local archives, government documents, and survivor interviews to examine collective executions involving women, children, and older people—events less visible in histories centered on urban student activism. Examining counties and villages makes it possible to distinguish participation in armed conflict from victimization through imposed social status or alleged political association. It also permits investigation of how broad political campaigns acquired different local outcomes, without treating the countryside as a uniform social setting. (cambridge.org)

References

1. State Sponsorship or State Failure? Mass Killings in Rural China, 1967–68escholarship.org
2. Chinese Society under Mao: Classifications, Social Hierarchies and Distributioncambridge.org
3. Anatomy of a Regional Civil War: Guangxi, China, 1967–1968cambridge.org
4. Rebellion of the Cadres: The 1967 Implosion of the Chinese Party-Statesociology.stanford.edu
5. Community and Culturecambridge.org
6. Framing War (Chapter 7) — Collective Killings in Rural China during the Cultural Revolutioncambridge.org
7. State Sponsorship or State Failure? Mass Killings in Rural China, 1967–68escholarship.org
8. Collective Killings in Rural China during the Cultural Revolutioncambridge.org
9. The Cultural Revolution in the Countryside: Scope, Timing and Human Impactcpb-us-e2.wpmucdn.com
10. Consuming Counterrevolution: The Ritual and Culture of Cannibalism in Wuxuan, Guangxi, China, May to July 1968gwern.net
11. PRC ‘flesh banquets’ remain unspoken, 50 years ontaipeitimes.com
12. Rebellion and Repression in China, 1966–1971sociology.stanford.edu
13. Consuming Counterrevolution: The Ritual and Culture of Cannibalism in Wuxuan, Guangxi, China, May to July 1968cambridge.org
14. Scarlet Memorial: Tales of Cannibalism in Modern Chinaapi.pageplace.de
15. On the Recordcambridge.org
16. Anatomy of a Regional Civil War: Guangxi, China, 1967–1968cambridge.org
17. Patterns of Killingcambridge.org
18. Collective Killings in Rural China during the Cultural Revolution: Chapter Summariescambridge.org