THE RED TURBAN REBELLIONS AND THE EMERGENCE OF ETHNIC 
CONSCIOUSNESS OF THE HAKKAS IN 
NINETEENTH-CENTURY CHINA 
by 
JAEYOONKIM 
A DISSERTATION 
Presented to the Department of History 
and the Graduate School of the University of Oregon 
in partial fulfillment of the requirements 
for the degree of 
Doctor of Philosophy 
August 2005 
11 
"The Red Turban Rebellions and the Emergence of Ethnic Consciousness of the Haldas 
in Nineteenth-Century China," a dissertation prepared by Jaeyoon Kirn in partial 
fulfillment of the requirements for the Doctor of Philosophy degree in the Department of 
History. This dissertation has been approved and accepted by: 
kaw, Co-Chair of the Examining Committee 
ir of the Examining Committee 
) 
Committee in Charge: Dr. Cynthia Brokaw, Co-Chair 
Dr. Bryna Goodman, Co-Chair 
Dr. Andrew Goble 
Dr. Richard Kraus 
Accepted by: 
Dean of the Graduate School 
111 
© 2005 Jaeyoon Kim 
lV 
An Abstract of the Dissertation of 
JaeyoonKim for the degree of Doctor of Philosophy 
in the Department of History to be taken August 2005 
Title: THE RED TURBAN REBELLIONS AND THE EMERGENCE OF ETHNIC 
CONSCIOUSNESS OF THE HAKKAS IN NINETEENTH-CENTURY CHINA 
Approved: ------- --
~ ~/~thia Brokaw, Co-Chair 
Approved: 
My dissertation, "The Red Turban Rebellions and the Emergence of Ethnic 
Consciousness of the Hakkas in Nineteenth-Century China," focuses on one of most 
important and controversial minorities in China-and a group that significantly shaped 
the country's nineteenth and twentieth century history: the Hakka or "guest people." Han 
Chinese who migrated from western Fujian to Guangdong province in search of new 
economic opportunities over the course of the eighteenth and nineteenth centuries, these 
"guest people" challenged the economic control of earlier settlers in these provinces and 
thereby sparked some of the most violent struggles of late Qing China. I examine, in 
particular, how the participation of the "guest people" in a series of struggles, the Red 
V 
Turban Rebellions (1854-1856) and the Hakka-Punti War (1856-1867) in the Pearl River 
Delta areas of South China, helped create among these people a distinct sense of identity, 
a sharp sense of their own, different, Hakka, ethnicity. 
My study is designed to provide a detailed historical analysis of the construction 
of Hakka identity. I focus on the whole network of different interests and relationships 
that led to the Red Turban Rebellions and the Hakka-Punti War of the mid-nineteenth 
century: the long-standing economic conflicts over land use; the part played by local 
gentry and lineage organizations in Hakka-Punti feuds; the role that the state, and most 
particularly local governments, played in intensifying existing tensions and thus drawing 
"ethnic" lines. In short, in focusing intensively on one particular place and time, my work 
provides a full and rich picture of all the factors--economic, political, as well as social--
that contributed to the definition of Hakka ethnicity. My dissertation thus helps us 
understand more precisely the complex process by which ethnicity is constructed. 
Vl 
CURRICULUM VITAE 
NAME OF AUTHOR: Jaeyoon Kim 
PLACE OF BIRTH: Janghung, Republic of Korea 
DATE OF BIRTH: December 22, 1968 
GRADUATE AND UNDERGRADUATE SCHOOLS ATTENDED: 
University of Oregon 
San Diego State University 
Chonnam National University 
DEGREES AWA RDE D: 
Doctor of Philosophy in History, 2005, University of Oregon 
Master of Arts in History, 1997, San Diego State University 
Bachelor of Arts in History, 1992, Chonnam National University 
AREAS OF SPECIAL INTEREST: 
Chinese History 
Japanese History 
PROFESSIONAL EXPERIENCE: 
Adjunct Faculty, Department of History, Social & Behavioral Sciences, 
Cuyamaca College, San Diego, 2003-2005 
Visiting Scholar, Department of History, Beijing University, Beijing, China, 
2001-2002 
Teaching Assistant, Department of History, University of Oregon, Eugene, 
1998-2001 
Vll 
GRANTS, A WARDS AND HONORS: 
Department of History, University of Oregon: Robert Lang Dissertation 
Fellowship, 2003 
Center for Asian and Pacific Studies, University of Oregon: Freeman Student 
Fellowships for Personalized Leaming, 2001 
Department of History, University of Oregon: Graduate Teaching Fellowship, 
1998-2001 
Department of History, University of Oregon: Paul Dull Memorial Scholarship, 
2000 
Chonnam National University: Full Tuition Exemption Scholarship for Academic 
Excellence, 1987, 1988, 1991 , 1992 
Chonnam National University: University Presidential Award for the Best Article, 
1991 
PUBLICATIONS: 
Jaeyoon Kim, "Mimana Problem in Early Korea and Japan," Daily Chonnam, 
Spring, 1991 
vm 
ACKNOWLEDGMENTS 
I am truly indebted to a number of people, each of whom has contributed to 
making this dissertation possible. 
First, I would like to express my deep gratitude to my Co-advisor, Professor 
Cynthia Brokaw. Cynthia has not only been a first class advisor but also a mentor to me. 
She provided me with the opportunity to fulfill my dream, believed in my ability since we 
first met, and has always been ready with a constructive word during the course of my 
research. 
I would also like to express my sincere thanks and deep appreciation to my Co-
adviser, Professor Bryna Goodman for her continuous precious support and assistance 
during the course of this research. I feel sure that this dissertation would never have come 
into existence without her help, patience and dedication. 
I owe a great deal to Professor Andrew Goble for his teaching and support during 
my study at the UO. I have enjoyed every moment with him including great dinners at his 
home. I would also like to thank Professor Richard Kraus for kindly serving on my 
dissertation committee on a very short notice. 
I am fortunate to have the opportunity to meet with a group of nice people in the 
history department. I want especially to thank Rebecca Lynn and Professor Jeffrey Ostler 
for their support and encouragement to finish this dissertation. 
My research was partly funded by Center for Asian and Pacific Studies, University 
of Oregon (Freeman Student Fellowships). I am also grateful to the history department and 
IX 
the Graduate School, University of Oregon for helping me continue my research through 
the time of hardships. 
I would like to thank my parent, Eung Ryul Kim and Kyungja Hong, for their 
continued encouragement and love during this work and throughout my life. I am 
especially grateful to my wife Kyehee Kim for her help, encouragement, and support. I 
thank her for doing the most important job of being a mother to Ha-eun Mary Kim who is 
the joy of our lives. 
Most importantly, I acknowledge my Lord and Savior, Jesus Christ through 
Whom I "live and move and exist" (Acts 17:28). I give back all the glory to Him, Jesus 
Christ. In Him all things are possible. 
X 
TABLE OF CONTENTS 
Chapter Page 
I. INTRODUCTION .............................................................. .......................................... 1 
1) Objectives ............................. .................... .. ............ ....... ............ ...... .. ..... ....... ...... 4 
2) Methodology and Sources .... ... .. ............. .. ...... ... ... .. ........ ..... .. ..... ..... ..... .... ...... . 12 
II. ENVIRONMENT ............................................................................................ ........ . 22 
1) The Physical Environment ............ .. ............. ....... ............................................. 23 
Geographical Environment .. ..... .. ................. ................ .. ....... .. ..... ... ... ... ... .... .. .. 23 
Regional Differences ............. .... ........ ... ....... ... ..... .......... ...... ... ... .. .. .. ... ... .......... 26 
2) Historical and Social Environment ................ .. .. ........ .. ....................... .. ... .. .. .... 34 
Early History ................... ...... ....................... .................................................... 34 
Population Growth .......................................................................................... .4 0 
Land Reclamation ...... ...... ..................... ..... ................... ...... ... .... ..... ..... ........ .... 43 
Alluvial Land and Lineages ............................................................................. .47 
Feuds .. ....... ....... ... ...... .. ....... .. .. .. ......... .... ... ......... ................ ..... ..... .. ........... ...... ... 54 
Migrations and Ethnic Tensions .......... .... .......... ...... .... ................. ...... ..... ........ 60 
Government Representation of Feuds ... .. ... ... .... .. .. ... .... ....... ............... ..... .. .. .... 69 
3) Economic and Political Environment .............. .... ........ .. ........... .................... ... 72 
Economy ............. ............... ........... ............ ............ ..... .......... .............. ... ........... . 72 
The Opium War and Social Disorder in the Pearl River Delta ... ... ... .... ........ .7 7 
The Taiping Rebellion ............................................ ......................................... 86 
4) Summary ......................................................................................... ...... ....... .. ... 91 
III. THE RED TURBAN REBELLIONS ..................................................................... 95 
1) The Course of the Red Turban Rebellions ...... .... .... ..... ............... ... ... .............. . 97 
A Series oflncidents Lead To The Red Turban Rebellions ........................... 97 
The Beginning and High Tide of the Red Turban Rebellions .... .. ...... ... ...... . 101 
Qing Forces ..... ......... ..... ....................... .. .... ... .. ....... ........ .... .. ..... ... .. ........ ........ 108 
2) The Nature of the Rebellions ......................................................................... 120 
The Composition of the Red Turban Rebels ................................................. 120 
Taiping Connection ................. ......................... ........ ..... ...... ..... ...... .... ....... ..... 137 
3) The Demise of the Red Turbans ....................... .............................................. 143 
Reasons for the Red Turban Failure .............................................................. 14 3 
Foreign Interventions ....... .. ........ ........................... .. ................................ :. .. ... 148 
Militias ......................... ......... ... ..... ..... ....... .................... ......... ... ............... ... .. .. 155 
4) The Red Turban Rebellions Interpreted ......................................................... 164 
Xl 
Chapter Page 
IV. ETHNICITY, LOCAL INTERESTS, AND THE STATE .................................. 172 
1) The Wuyi Region in Historical Perspective .................................................. 174 
Geography of the Wuyi .................................................................................. 17 4 
Early History .................................................................................... ..... .. ....... 180 
2) The Hakka-Punti Conflicts in Retrospect ................................................ ...... 186 
Hakka Controversy .......................................................... ............... .. ............. 186 
Hakka History ................................ ................................................................ 192 
Hakka Settlements in the Wuyi ......................... ............................................ 195 
The Source of Conflict ................... ....................... ......................................... 204 
Hakka-Punti Disputes ........................ ............................................................ 213 
3) History and the Emergence of Hakka Ethnic Consciousness ....................... 221 
Red Turbans in the Wuyi ........... ........... ........... ...... ................. ....................... 221 
The Hakka-Punti War .............................................. ...................... ................ 230 
Aftermath ................ .... .. ............ ............ ........... ..... ... ... ................................... 240 
Hakka Emigration ........................... ........................ .. ......... ...... ...................... 245 
4) Ethnicity and the Hakkas .......... ...... ........... ............................... ..... ................. 253 
IV. CONCLUSION ........... ........................................................................................ ..2 61 
BIBLIOGRAPHY .................................................................. ............................................ 270 
XU 
LIST OF MAPS 
Map Page 
1. Counties of the Pearl River Delta ........................ .. ................. .. ............................... 104 
3. Wuyi (Five Counties) in the Pearl River Delta ....... ..... .. ......... ............ .................. . 175 
Xlll 
LIST OF TABLES 
Table Page 
1. Population Densities (Guangdong Province, 1820) ............... ................ .............. ..... 30 
2. Occupation of the Major Red Turban Leaders in Guangdong, 1854-1855 ...... .. ... 123 
3. Place of Origin of the Rebels Captured in Qujiang County ... ................................ 126 
4. Age Distribution of Red Turban Rebels Captured in Yingde County ......... .......... 128 
1 
CHAPTER I 
INTRODUCTION 
The mid-nineteenth century was a period of widespread insurrections, and with the 
easy success of the Taiping rebellion, the most famous of these, other oppressed people 
were inspired to increasing rebelliousness. One of the most chronically disturbed regions of 
the empire during the nineteenth century was Guangdong province. Generally regarded by 
Chinese officials as unhealthy and hard to govern, Guangdong, even before the Taiping 
rebellion, was noted for its ethnic tensions, its hill gangs and river pirates, and its secret-
society activities. 1 
After about 1820, the swelling opium trade helped motivate a corresponding 
increase in the level of illegal activities and the number of illegal groups engaged in them. 
Guangzhou and the lower Pearl River Delta regions were the hub of these kinds of illegal 
1 Dian H. Murray, Pirates of the South China Coast, 1790-1810 (Stanford: Stanford University Press, 1987), 
2-25 . 
2 
associations, especially the Tiandihui x!-tP;Wf (Heaven and Earth Society).2 At the same 
time, pressures of overpopulation and increasing rural poverty built up in the countryside. 
In response to these multiple factors, the Tiandihui groups gave birth to the Red Turban 
rebellions (1854-56). 
In 1854, a Red Turban leader from Foshan 1,11 LU, rose in revolt against the Qing 
government, declaring his intention of re-establishing the Ming dynasty. The rebels 
attempted to take the city of Guangzhou, 15 miles from Foshan, that same year.3 Though 
they failed at this effort, several district cities such as Heshan U LU, Kaiping I}?] f, and 
Enping ,~, f were captured by the Red Turbans.4 
2 Past studies on the Red Turban Rebellions have suggested that the movement was an expression of the 
Tiandihui (Heaven and Earth Society), a secret society that had inspired countless rebellions throughout 
south China. It is true that secret societies have played roles in anti-dynastic rebellions, though usually such 
activities were more the exception than the rule. Morever, not all such groups participated in revolts against 
the state. Recent scholarship on the Brotherhoods, such as that by Zhuang Jifa, Dian Murray, and David 
Ownby, have shown that the Tiandihui, which were among the most important of the secret societies in 
South China in the pre-1850 pre-Opium War period, did not originate as overtly anti-Qing organizations, 
but rather as mutual-aid and self-protection associations. See Zhuang Jifa m:a~, "Qingdai mimi huidang 
de tantao ~ftfVWwtli!i:1¥J~!;j," Zhongguo lishi xuehui shixuejikan i:p~ff~~Wl~~~flJ, 16 (July 
1984): 153-83; Dian H. Murray, The Origins oft he Tiandihui: The Chinese Triads in Legend and History 
(Stanford: Stanford University Press, 1994); David Ownby, Brotherhoods and Secret Societies in Early 
Mid-Qing China (Stanford: Stanford University Press, 1996); Dian H. Murray, Pirates oft he South China 
Coast, 1790-1810 (Stanford: Stanford University Press, 1987). 
3 Frederic Wakeman Jr., "The Secret Societies of Kw angtung, 1800-1856," in Popular Movements and 
Secret Societies in China: 1840-1950 ed. Jean Chesneaux (Stanford: Stanford University Press, 1972), 43 . 
4 For a succinct account of the activities of the Tiandihui and militias, see Frederic Wakeman Jr., Strangers 
at the Gate: Social Disorder in South China, 1839-186 (Berkeley & L.A.: University of California Press, 
1966), 117-148. 
3 
Faced with a critical situation during the peak of the rebellions in 1854, the local 
magistrates, particularly in the Wuyi Ji B areas (the five counties southwest of 
Guangzhou), then took a step that was later thought to have done more than anything else 
to provoke yet another disturbance, the Hakka-Punti War (1856-68). They called up braves 
from among the Hakkas to fight against the Punti Red Turbans, intending to use the 
hostility between the two groups to help quell the rebellions. Ironically the conflict that 
resulted from this strategy, the Hakka-Punti War, lasted for almost ten years, engulfing as 
many as eight counties-Xinning ffi$, Gaoming r§J BJ!, Heshan, Y angchun ~:ff:, 
Yangjiang ~l[, Enping, Kaiping, and Xinxing ffi~-until 1867, when the Hakka were 
resettled in a newly created county, Chixi $~, in southern Guangdong.5 As witnessed by 
John Scarth, the protracted war cost a million casualties in the province of Guangdong 
within a year.6 Many displaced peasants left the country to escape Qing wrath. It is 
5 S. T. Leong, Migration and Ethnicity in Chinese History: Hakkas, Pengmin, and Their Neighbors 
(Stanford: Stanford University Press, 1997), 74; David C. E. Liao, The Unresponsive: Resistant or 
Neglected?: The Hakka Chinese in Taiwan Illustrate a Common Mission Problem (Chicago: Moody Press, 
1972), 86. 
6 John Scarth, Twelve Years in China: The People, the Rebels, and the Mandarins (Edinburgh: T. Constable 
and co., 1860). pp. 220-40, esp. 238. 
4 
estimated that during the Hakka-Punti War, more than a hundred thousand peasants left 
their hometowns. 7 
The Red Turban Rebellions and the Hakka-Punti War of Guangdong did not have 
nation-wide repercussions, and may appear to be dwarfed in significance by the 
cataclysmic upheavals that China was undergoing at this time. Nevertheless, they are of 
considerable interest and importance since they provide case studies for the interaction 
between local society, a rebellion motivated by problems internal to that society, and 
competing ethnicities. 
1) Objectives 
My purpose in this study is twofold: First, I am interested in studying the cause 
and significance of the Red Turban Rebellions and the Hakka-Punti War and the 
relationship between these two disturbances, as it can be argued that the one spawned the 
other. I will focus my attention on the impact that endemic disorder has on the long-term 
relationships between subethnic groups and organizations within local societies. In 
particular, by looking at the long-lasting economic competition between these subethnic 
7 Chixi Xianzhi #H~M;t, 8.28a. 
5 
groups, at the formation of exclusive groups like secret societies (especially the 
Tiandihui) and lineages, and at the creation of subethnic categories, I will examine their 
impact upon the configuration of their local societies. 
There are a number of excellent related studies on rural revolts,8 on militias, and 
on the role of secret societies in the Tiandihui rebellions of the mid-nineteenth century 
China.9 Surprisingly, however, no English works have treated the Red Turban Rebellions 
and the Hakka-Punti War together in a detailed or systematic manner. Most important as 
far as my study is concerned, however, past studies have also overlooked the most 
fundamental issue of the rebellions of the Pearl River Delta during the mid-nineteenth 
8 Rural movements were endemic in nineteenth century China. The term 'rural ' seems more appropriate to 
the subject of this study than the expression 'peasant.' Defined as a rural cultivator, the term 'peasant' 
excludes other participating categories sharing broad identities of interest and whose condition may in 
many respects is similar. See Henry A. Landsberger, Rural Protest: Peasant Movements and Social Change 
(N.Y.: Harper & Row Publishers, Inc. 1973), 6-22 and Eric R. Wolf, Peasant (Englewood Cliffs: Prentice-
Hall, 1966), 1. 
9 Studies on the subject ofrebel groups and secret societies in nineteenth century China, such as Hsiao 
Kung-ch'uan, Frederic Wakeman, Philip Kuhn, and Albert Feuerwerker, have similarly viewed it as a form 
of popular dissidence, a level of local militarization, or a stage of rebellion. Wakeman's study ( 1966) is the 
first work to present the broad themes surrounding the Red Turban rebellions. However, his work suffers 
from the narrowness of temporal perspective. According to Wakeman, the Opium War and subsequent 
British pressure stimulate a social disorder, anti-foreign ism, the rise of gentry power, and the polarization 
of society. But, we cannot see these developments in conceptual perspective without a consideration of the 
local society before the Opium War. His work suffers from a failure to discuss the pre-1839 period and the 
long-run effect ofregional or ethnic consciousness upon the integrity of the local society. See Hsiao Kung-
chuan, Rural China: Imperial Control in the Nineteenth Century (Seattle: University of Washington Press, 
1967); Wakeman, Strangers; Philip Kuhn, Rebellion and its Enemies in Late Imperial China; Albert 
Feuerwerker, Rebellion in Nineteenth-Century China (Ann Arbor: Center for Chinese Studies, University 
ofMichigan, 1975). 
6 
century: ethnicity and local socioeconomic competition. 10 Very little work has been done 
on the effects of the Red Turban Rebellions and the Hakka-Punti War on the organized 
groups, people and society of the region. These conflicts produced not only the 
strengthening of gentry control over the local society but also the strengthening of Hakka 
ethnic self-consciousness, particularly as a consequence of the Hakka-Punti War. 
Therefore, while defining the problems of understanding the Red Turban Rebellions and 
the Hakka-Punti War in terms of the local context in which they occurred, this study will 
provide a fuller and more accurate history of the uprisings and the Hakka-Punti conflicts. 
Another major purpose of this study is to show how these forces, especially 
through the course of the Red Turban Rebellions and the resulting Hakka-Punti War, 
encouraged the strengthening of the ethnic consciousness of the Hakkas. The Hakkas 
(kejiaren ~%.A) or "guest people" were often in conflict with the Puntis (bendiren 
;;$::1:t!!.A), the earlier settlers, in south China for a long time. It is my contention that 
Hakka ethnic consciousness, though clearly constructed in nature, is more than mere 
1°F or example, Xiao Yishan, Jean Chesneaux, Fei-ling Davis, and Lu Baoqian have only analyzed 
Tiandihui rebellions in terms ofproto-nationalism, social protest, and anti-dynastic dissent. See Xiao 
Yishan ,I -LI.J, Qingdai tongshi r1'H~:im~, Vol. I (Taibei: Taiwan shangwu yinshuguan, 1962); Lu 
Baoqian llfJf-=f-, Lun wan-qing Liangguang de Tiandihui zhengquan ~BiWWi.ll8'9::R:tt!nii!&li (Taibei: 
Zhongyang yanjiujindaishi yanjiusuo, 1975); Jean Chesneaux, Secret Societies in China in the Nineteenth 
and Twentieth Centuries (Ann Arbor: University of Michigan Press, I 97 I); Fei-ling Davis. Primitive 
Revolutionaries ofC hina (Honolulu: University Press of Hawaii, 1971). 
7 
cultural or social construction; Hakka ethnic consciousness instead originated from a far 
more complex process. In the case of the Hakkas, for example, their sense of ethnic 
identity developled from economic competition with the Puntis, lineage conflicts, secret 
society formations, local rebellion in an area of multi-ethnic settlement, government 
policy, and even foreign missionary intrusion, as well as cultural differences. 
Most of the various theories about the origination of the Hakka ethnic identity 
differ depending on the definition of ethnicity used. 11 From the Hakka historians' point of 
view, ethnic identity is rooted in shared sentiments thought to derive from a common 
11 There is also no consensus on how the concept of ethnicity should be defined, but there is general 
agreement that it is a form of culturally based group identity. ln the same vein, others have argued that 
ethnicity is the contemporary expression of a primordial sentiment. Anthony Smith, in a more nuanced 
version of the primordialist thesis, sees a distinctive shared culture and a shared history as being important 
components of the foundation of ethnic community. Barth, in his critique of perspectives that equate an 
ethnic identity with a given cultural content, nevertheless acknowledges the role of culture by noting that 
across in a given group use some cultural features of the variety available to them as over signals or signs-
the diacritical features that people look for and exhibit to show identity, often such features as dress, 
language, house-form, or general style of life. His point in essence is that the ethnic boundary is reinforced 
by a cultural scaffolding of some sort and endures because of persisting cultural differences. See W. W. 
Isajiw, "Definitions of Ethnicity," Ethnicity I (1974): l I 1-24; Brackette Williams, "A Class Act: 
Anthropology and the Race to Nation across Ethnic Terrain," Annual Review ofA nthropology 18 (1989): 
401-4; George de Vos, "Ethnic Pluralism: Conflict and Accommodation," in Ethnic Identity: Cultural 
Continuities and Change, eds., George de Vos and Lola Romanucci-Ross, (Chicago: University of Chicago 
Press, 1975); Charles F. Keyes, "The Dialectics of Ethnic Change," in Ethnic Change, edited by Charles F. 
Keyes (Seattle: University of Washington Press, 1981); Judith Toland, "Introduction: Dialogue of Self and 
Other: Ethnicity and the Statehood Building Process," in Ethnicity and the State, ed. , Judith D. Toland 
(New Brunswick, N.J.: Transaction Publishers, 1993); Anya Peterson, Ethnic Identity: Strategies of 
Diversity (Bloomington: Indiana University Press, 1982); Clifford Geertz, "The Integrative Revolution: 
Primordial Sentiments and Civil Politics in the New States," in Old Societies and New States, ed. Clifford 
Geertz. (Glencoe, Ill.: The Free Press, 1963); Harold R. Isaacs. "Basic Group Identity: The Idols of the 
Tribe," in Ethnicity: Theory and Experience, ed. Nathan Glazer and Daniel P. Moynihan. (Cambridge, 
Mass. : Harvard University Press, 1975); Edward Shils, "Primordial, Personal, Sacred and Civil Ties," 
British Journal ofS ociology 8 (1957): 130-45; Anthony D. Smith, The Ethnic Origins ofN ations (Oxford: 
Basil Blackwell, 1986); Fredrik Barth, "Introduction," in Ethnic Groups and Boundaries: The Social 
Organization of Cultural Difference, ed. Fredrik Barth (Oslo: Universitetsforlaget, 1969), 14. 
8 
primordial past and passed down from generation to generation. These bonds-based on 
a common language, religion, and history-are believed to create the basis for ethnic 
ties. 12 But as Eric Hobsbawm has illustrated in his study of what he calls "invented 
traditions," what is important is that people believe "traditions" to be old, not that they are 
in fact so. 13 Hakka genealogies reach back as far as possible and are considered evidence 
of "pure Chinese" as opposed to "barbarian" status by Hakka historians, but these early 
so-called "Hakka" were not an ethnic group. 
Migration and Ethnicity in Chinese History: Hakkas, Pengmin, and Their 
Neighbors by S. T. Leong (1997) is really the first work to present a complex and 
persuasive account of Hakka migration and evolution. Regional developmental cycles of 
economic expansion and decline, Leong argues, fundamentally shaped the patterns of the 
Hakka migration and their complex relations with native populations. 14 While economic 
upturn in the regional core typically attracted migrants, the Hakkas, an eventual economic 
downturn led to conflicts between the natives and the migrants. It was through these 
12 Leong, " "The Hakka Chinese ofLingnan: Ethnicity and Social Change in Modem China," in Ideal and 
Reality: Social and Political Change in Modern China, 1860-1949, eds. David Pong and Edmund S. K. 
Fung, 287-326. (Lanham: University Press of America, 1985), 302-7. 
13 Eric Hobsbawm, "Introduction: Inventing Traditions," in The Invention of Tradition eds. Hobasbawm 
and Ranger (New York: Cambridge University Press): 1. 
14 Leong, Migration, 28-29. 
9 
conflicts that ethnic consciousness developed. 
Leong takes a situational approach in his analysis of the formation of ethnic 
consciousness. Following the insights of Orlando Patterson and Fredrik Barth, he 
questions the connection between cultural and ethnic groups and instead links the 
development of local expressions of Hakka ethnic awareness to specific regional 
situations of conflict and competition. 15 On this basis, the conflicts in Fujian province 
throughout the sixteenth century between the She (a non-Han minority) and the Han (who 
later became known as the Hakkas) provided the setting in which ethnic identity emerged. 
It was at this time that Leong believes the term "Hakka" made its first appearance. 
Leong's work is impressive for the scope and breadth of his analysis and his 
ability to avoid simplification and over-generalization. Migration and Ethnicity is a 
persuasive application of William Skinner's groundbreaking core-periphery analysis and 
a valuable contribution to current literature on migration and ethnic history within China. 
15 The new scholarship on the Hakka studies, represented by Sow-theng Leong and Nicole Constable, owes 
a fundamental intellectual debt to the work ofFredrik Barth. Nicole Constable focuses on the notion of 
identity, which merges with 'ethnicity.' She provides a very useful review of what is known about the 
Hakkas and sets the essays and case studies into perspective. She looks for variation along some parameters 
for Hakka identity: entrepreneurial ethos, gender roles, Chineseness, ethnic consciousness and language. In 
my judgment, the conclusions are evasive. She doesn't clarify whether the Hakkas are much different from 
other ' subethnic' Chinese or not. For instance, it is not clear whether the consciousness of being Hakka in 
nature is very different from the consciousness of being from Chaozhou or not. See Ibid. , 14- 15; Constable, 
Nicole, Guest People: Hakka Identity in China and Abroad, (Seattle and London: University of 
Washington Press, 1996); Idem., Christian Souls and Chinese Spirits: A Hakka Community in Hong Kong 
(Berkeley: University of California Press, 1994). 
However, Leong has little evidence to prove the development of a self-conscious sense of 
difference among the Hakkas. He simply asserts that the economic conflicts of the 
eighteenth and nineteenth centuries created a distinct sense of identity, but is not able to 
provide any textual references that might support this assertion. Indeed, the label 
"Hakka," as used to refer to a distinct group of people, did not come into common use 
until the early twentieth century, and there is no evidence for its use at all before the very 
early nineteenth century. 16 Certainly, Leong argues persuasively that economic inequality 
created tensions that resulted in conflicts between the Hakkas and native populations 
before the middle of the nineteenth century. But socioeconomic conflict alone does not 
necessarily create ethnic self-consciousness. It was not just economic conflict, but the 
convergence of a range of long-standing economic and social tensions influenced, by the 
presence of western missionaries and traders, and, most significantly, the intrusion of the 
state, that stimulated the development of a distinct Hakka identity. 
More promisingly, recent studies by Emily Honig, Stevan Harrell, and Pamela 
K. Crossley of the Subei people in Shanghai, the Yi people in Yunnan and the Manchus in 
the Qing empire respectively, have suggested that ethnicity should best be seen as 
16 See Constable, introduction to Guest People: Hakka Identity in China and Abroad, ed. Nicole Constable. 
(Seattle and London: University of Washington Press, 1996): 9. 
11 
socially constructed. 17 It is, to be sure, quite useful to see Hakka ethnic consciousness as 
a constructed tradition. But, in analyzing the process of the construction of Hakka 
ethnicity, it is important to understand the full range of contributing factors-not only 
social, but also economic, religious, linguistic, and political-as well as the complex 
interaction of these factors. 
The word Hakka, meaning "guest people," referred to their continuing 
sojourning traditions; they had migrated extensively not only within, but also outside, 
China to Taiwan, Singapore, India, and even America. Even in the midst of immigrant 
societies in these areas, the Hakkas often held themselves aloof from other Chinese and 
were, in tum, often treated as different by Chinese. In many cases, even today, the 
memory of the historical animosity and competition between the Hakkas and Puntis plays 
an important role in the maintenance of a clear sense of Hakka identity. 18 A study of this 
17 According to Honig and Harrell ' s studies, ethnicity is not an objective thing, such as blood ties, but 
rather a process. This idea appears to have been under the influence of discourse in the past two decades on 
modern nations as imagined communities. Members of an ethnic group imagine to share a cultural system, 
much as citizens of a modern nation/state share an imagined community. Based on this idea, Crossley 
argues that there was no primordial basis for Manchu ethnicity and that both Manchu culture and identity 
were created simultaneously with the Qing empire in the 1630s. See Emily Honig, Creating Chinese 
Ethnicity: Subei People In Shanghai, /850-1980 (New Haven and London: Yale University Press, 1992); 
Stevan Harrell, "Ethnicity, Local Interests, and the State: Yi Communities in Southwest China," 
Comparative Studies in Society and History, 32 ( 1990), pp. 51-548; Pamela Kyle Crossley, Orphan 
Warriors: Three Manchu Generations and the End oft he Qing World, (Princeton: Princeton University 
Press, 1990); idem, The Manchus (Oxford and Cambridge, Blackwell, 1997). 
18 The Hakkas in Calcutta is the good case of this. See Ellen Oxfelld. "Still 'Guest People' : The 
Reproduction of Hakka Identity in Calcutta, India," in Guest People: Hakka Identity in China and Abroad, 
12 
series of conflicts, then, will both illuminate the factors behind current ethnic tensions 
and provide a basis for comparative analyses of the interaction of migrant groups within 
the Chinese diasporas. My project, therefore, will contribute not only to the specific 
history of the rural movements of the Pearl River Delta but also to the larger issue of 
ethnicity in Chinese history. 
2) Methodology and Sources 
There are several different sets of secondary sources relevant to my project. First, 
anthropological (a nd some historical and sociological) studies of ethnicity provide 
comparative information and a variety of different theoretical approaches to the 
understanding of ethnicity as a category; these have guided my interpretation of the 
meaning of Hakka identity. Fortunately, too, the number of specific studies of the Hakkas, 
based on field research in Hakka communities in China and other parts of the world, has 
increased significantly in the past two decades. This literature contains little information 
about the mid-nineteenth century conflicts that I am researching and, for the most part, 
accepts the label "Hakka" as an essentialized, unchanging, and unproblematic marker of 
ed. Nicole Constable, 149-175 (Seattle and London: University of Washington Press, 1996) 
13 
identity. Nonetheless, this body of scholarship contributes significantly to my study by 
providing useful information on the continuing evolution of Hakka identity. 19 
Second, the large collection of works by Hakkas, largely devoted to the 
celebration of the Hakka myth, though of little use as a source of information about 
Hakka history, supplies a very rich store of information about the nineteenth- and 
twentieth-century construction of Hakka "ethnicity." Clearly such works have to be read 
very carefully and selectively, but a study of their rhetorical strategies and arguments will 
yield useful insights into the process of the social and intellectual construction of Hakka 
difference. 20 
19 
See Fred C Blake., Ethnic Groups and Social Change in a Hong Kong Market Town, Asian Studies at 
Hawaii, no. 27 (Honolulu: University Press of Hawaii, 1981); Mylon L. Cohen, "The Hakka or 'Guest 
People': Dialect as a Sociocultural Variable in Southeastern China," Ethnohistory 15.3 (Summer 1968), 
237-92; Harry J. Lamley, "Subethnic Rivalry in the Ch'ing Period," in The Anthropology of Taiwanese 
Society eds. Emily M. Ahem and Hill Gates (Stanford: Stanford University Press, 1981 ): 282-318; Leong, 
Migration and Ethnicity; idem, "The Hakka Chinese ofLingnan: Ethnicity and Social Change in Modem 
China," in Ideal and Reality: Social and Political Change in Modern China, 1860-1949, eds. David Pong 
and Edmund S. K. Fung, 287-326 (Lanham: University Press of America, 1985); Burton Pasternak, Kinship 
and Community in Two Villages (Stanford: Stanford University Press, 1972). 
20 Interest in the Hakkas reached its height in the 1920s, a period that has been referred to as the peak of 
Hakka nationalism. It was during this time that Hakka historians began to develop their theories concerning 
Hakka migrations. By the early 1920s, due in large part to the establishment of the Tsung Tsin 
(Chongzheng: ~LIE) association in Hong Kong, the name "Hakka" became more widely accepted as an 
ethnic label. In 1921, in response to a Shanghai Commercial Press publication of The Geography oft he 
World that described the Hakkas as non-Chinese, the United Hakka Association held a conference in 
Guangzhou attended by over a thousand angry delegates representing Hakka organizations worldwide. The 
result was a forced retraction of the offending phrase. Since then Hakka associations continue to publish 
regularly works celebrating the "Hakka Spirit" and what is distinctive about Hakka culture. Astonishingly, 
mainland Chinese scholars have paid little attention to Hakka social history in nineteenth-century 
Guangdong, preferring to focus on the study of Hakka language. According to Erbaugh's essay, there are 
three reasons for this lack of interest in Hakka studies: ideals of Han unity; official categories that do not 
easily accommodate the Hakkas; and the slow, uneven construction of Hakka identity. See Blake, Ethnic 
14 
Finally, though there is little in the way of scholarly study of the Red Turban 
Rebellions or the Hakka-Punti War in English, Japanese, or Chinese, there does exist a 
considerable literature on peasant rebellions in the nineteenth century. This work is 
important for analysis of the series of conflicts that I have chosen to study, for 
comparison with other conflicts will help to identify the distinctive ethnic elements that 
shaped the Red Turbans Rebellions and the Hakka-Punti War. 
Primary historical material on my study can be divided into three different 
categories: official reports and observations and local histories; local community and 
lineage records and genealogies; and foreign accounts of life in Guangdong. Detailed 
reports by local officials from nineteenth-century China are rare, in part because of the 
administrative practice of discarding lower-level reports and in part because of the series 
of rebellions, wars, and revolutions that have destroyed so many primary documents from 
pre-twentieth century China. By a curious irony of history, however, it was the British 
who preserved the collection of Chinese official documents that provides the richest 
source of information about the Red Turban Rebellions. When the British occupied the 
Groups; Hsieh T'ing-yu, "Origins and Migrations of the Hakkas," Chinese Social and Political Science 
Review 13 (1929): 208-28; Luo Xianglin ,l.il;ffi:;M-, Kejiayanjiu dao/un ~*Wr~~Mi, (Xinning: Xishan 
shucang, 1933); Leong, Migration and Ethnicity; Mary S. Erbaugh, "The Hakka Paradox in the People's 
Republic of China: Exile, Eminence, and Public Silence," in Constable, Guest People, 196. 
15 
office of the provincial government in Guangzhou in 1858, they took possession of all 
papers they found there, including both administrative documents and private 
correspondence between the viceroy ofLiangguang ~~ (Guangdong and Guangxi) and 
the Qing court; this collection was retained even after the conclusion of peace with China 
and deposited in the British Legation in Beijing, where it served as a file of intelligence 
materials. Eventually it was sent to London and finally came to rest in the Public Record 
Office, where it is known as the Canton Archive.21 Preserved as it was originally found, it 
is a uniquely comprehensive record of provincial administration in the mid-nineteenth 
century.22 
21 David Pong, A Critical Guide to the Kwangtung Provincial Archives: Deposited at the Public Record of 
Office ofL ondon (Cambridge and London : Harvard University Press, I 975). 
22 Even though Chang Hsin-pao (1964) and Masaya Sasaki ( 1967) identified and introduced these 
documents, surprisingly, very few historians have used them to study the rebellions and ethnicity of the 
Pearl River Delta during the mid-nineteenth century. This source has been tapped to some extent in a study 
by Wakeman (1966); however, his work only relied on materials written in English among Foreign Office 
records, materials that give an incomplete picture of developments in the Pearl River Delta. The few 
accounts by Lu Baoqian (1975), J. Y. Wong (1976), J.A.G. Roberts (1968), and Liu Ping (2003) also used 
some of Canton Archive materials to describe conditions in Guangdong in the connection with the Red 
Turban Rebellions, but all these works ignore the most important issue of the time, ethnicity. See Chang 
Hsin-pao, Commissioner Lin and the Opium War(Cambridge: Harvard University Press, 1964); Sasaki 
Masaya {ti: 7 :.tIE~, Shinmatsu no himitsu kessha filr*O);,t,i&i~ifd: (Tokyo: Kindai Chugoku Kenkyii 
Iinkai, 1967); Lu Baoqian llf Jf =f, Lun wan-qing Liangguang de Tiandihui zhengquan 
ffifufl§eftlr~.!l t¥J':R:ltwii&fl (Taibei: Zhongyang yanjiujindaishi yanjiusuo, 1975); Frederic Wakeman Jr., 
"The Secret Societies ofKwangtung, 1800-1856," in Popular Movements, ed. Jean Chesneaux; Idem, 
Strangers at the Gate: Social Disorder in South China, 1839-18 (Berkeley: University of California Press, 
1966); J. Y. Wong, Yeh Ming-ch'en: Viceroy ofL iang Kuang, 1852-8 (Cambridge: Cambridge University 
Press, 1976); J. A.G. Roberts, 'The Hakka-Punti War', unpublished D.Phil. thesis (University of Oxford, 
1968); Liu Ping )(lj 5f, Bei yiwang de zhanzheng: Xianfeng Tongzhi nianjian Guangdong TuKe da xiedou 
yanjiu Ml~:&;1¥Ja!i:!G4ji- : Jilt.$ [EJ~:q:i'aJ1*± ~j;,::fJJ£4Wfo/t'. (Beijing : Shangwu yinshuguan, 2003). 
16 
Since the Red Turbans were a direct threat to Guangdong province, the official 
reports in the Canton Archive include abundant manuscripts on their rebellions. Of the six 
chronologically ordered divisions of the archive, 23 the section of materials on the 
rebellions, secret societies, military organization and operation, and rebellion suppression, 
1811-1857, is the most voluminous and useful for my work. In addition to military 
materials dealing with the suppression of the rebels from the field, there were also the 
confessions made by captured rebels and reports on such captives after interrogation or 
trial.24 Imperial edicts in these documents also reveal government attitudes and policies 
toward the Hakkas and give us some sense of how the state "read" the ethnic identity of 
this group. 
Some information on the Hakkas and their involvement in the Red Turban 
Rebellions is also contained in the form of memorials sent by provincial officials to the 
emperor, now in the Grand Council Reference Files (Junjichu zouzhe lufu 
23 The six topics are: Opium Trade and the Opium War, Central and local administration, 1765-1857, 
Foreign relations and foreign trade, 181 0s-1857, Rebellions, secret societies, military organization and 
operation, and rebellion suppression, I 811-1857, The second Anglo-Chinese War (the 'Arrow' War), first 
pha5e, 1856-1857, and Maps and illustrations. 
24 David Pong, "The Kwangtung Provincial Archives at the Public Record Office of London: A Progress 
Report," Journal ofA sian Studies, vol. 28, Issue I (Nov. I 968): 139-143. 
17 
]l[;Jjt~~ffl~iiJ) in the First Historical Archives at Beijing.25 The Emperor's response to 
these memorials can be found in the Daqing lichao shilu :::krnfflt~•~- As reports from 
high-ranking officials, these memorials detail the government reactions to the rebels and 
ethnic feuds of the time; the imperial responses outline government policy toward these 
problems. 
The richest sources, however, for viewing the rebels and the Hakka ethnic issue 
from outside Beijing are provincial, prefectural, and county histories or gazetteers. The 
gazetteer is a special genre of literature in Chinese historical writing, and the Qing 
dynasty marked the genre's most fruitful period. County gazetteers (xianzhi M$) from 
Guangdong are particularly useful for my study, as they provide the most detailed local 
information. For example, the dates, locations, and numbers of people involved in the 
rebellions and ethnic feuds can be found in gazetteer sections on previous events or 
25 One of the most exciting developments for historians of China during the last twenty years has been the 
opening of archives on mainland China rich in materials on sensitive subjects of peasant rebellions and 
secret societies. Although there are many publications of archival documents from the Qing by Chinese 
archivists [for example, the edited collection of documents from the Number One Historical Archive in 
Beijing on the pre-Opium War Tiandihui: Qin Baoqi ~-~ and Liu Meizhen IIJ~ft eds., Tiandihui 
:X:tt!nw. 7 volumes (Beijing: The Number One Historical Archive, 1981-89)], thousands of documents on 
various subjects have not yet been published. Thus, many historians still need to visit the Number One 
Historical Archive to pore through the enormous mass of poorly arranged documents for their research. For 
excellent guides to the use of these archival materials, see Beatrice S Bartlett, Ch'ing documents in the 
National Palace Museum Archives, (Taibei: National Palace Museum, 1975); idem, Archival Materials in 
China on United States History, (White Plains, N.Y.: Kraus International Publications, 1985); idem, 
Selected articles from Taipei National Palace Museum bulletin, 1972-1979, (Taibei: National Palace 
Museum, 1972-79); Joseph Esherick, Chinese Archives: An Introductory Guide, (Berkeley : Institute of 
East Asian Studies, University of California, Berkeley, Center f~r Chinese Studies, 1996). 
18 
rebellions. Accounts of local suppression attempts are available in the military and 
biographical sections. Often events are simply chronicled, however, and in these cases, 
the information must be used with care because statistics are often inflated. A more 
important shortcoming is the very apparent bias in these local histories in the accounts of 
the Hakka-Punti War. Only one of the local histories, the Chixi Xianzhi ziFf§lMi;:t, reports 
the war from the Hakka angle. The others, without exception, give a version that is 
wholly on the Punti side. As the purpose of my work has been to give a critical, and as far 
as possible, objective account of what occurred, it has been necessary to use the local 
histories cautiously. 
Biographies, letters, and collections of anecdotes by both prominent officials 
serving in the Pearl River Delta areas and local literati often include their impressions of 
the rebellions and the Hakka-Punti ethic feuds. Thank to the efforts of the history 
department of Zhongshan University, most of these writings as well as local gazetteers, 
on the subject of the Red Turban Rebellions are now collected and published under the 
title of Guangdong hongbing qiyi shiliao ~ * #t~AQ~ .t.l4 (1992).26 Of these sources, 
26 Guangdong hongbing qiy i shiliao $i* YA~~~-'£J4, Guangdong sheng wenshi yanjiuguan, Zhongshan 
daxue lishi xibian (Guangzhou : Guangdong rennun chubansbe, 1992). 
19 
the works by Chen Kun l!t:i:$ and Chen Dianlan l!t~iil are particularly useful.27 Besides 
these sources, I also consulted several biographies and scholarly works of the time not in 
the Guangdong hongbing qiyi shiliao, such as and the Sohak dongjumki iz.g~Jftitric, the 
biography of a Korean gentry member who was traveling in the Guangzhou area during 
the chaotic mid-nineteenth century.28 Like gazetteers, most of these materials were 
written almost entirely from the local ( especially from the Punti) perspective and biased 
by elite chauvinism. But, used with care, these works provide both exceptionally detailed 
accounts of the rebels' activities and information about the sources of Hakka-Punti 
conflict. 
The other local records collected from the Pearl River Delta areas, such as the 
genealogies of the Hakkas and grave, temple and other stele inscriptions, also supply 
crucial data for the study of the Red Turban Rebellions and the Hakka-Punti War. First, 
they describe the actions of influential community figures, important historical events, 
and the details of local conflicts. Second, they also help us to understand the growth of 
Hakka enclaves, their social organizations, ethical codes, and religious practices, and 
27 Chen Kun ~:f:ljl, "Yuedong jiaofei jilue 14 *:l!!JJ l1Uic~ ," in Guangdong hongbing qiyi shiliao; Chen 
Dianlan ~~iii. Gangcheng zhengeji ~:!Jixtt-::JG~c," in Guangdong hongbing qiyi shiliao. 
28 Yong Kweng ®I};!, Sohak dongjumki iffi**ii~c, trans. Kwon Heechul (Seoul: Ulyumoonhwasa, 1974). 
20 
their relations with other local groups. Certainly, their credibility as historical records is 
not unproblematic. Such sources are fragmentary, scattered, sometimes erroneous and 
often elite-biased, and thus must be read critically. They are, however, particularly 
valuable because they present the Hakka perspective on local conflicts and thus provide a 
much-needed corrective to the official and Punti-oriented sources listed above. 
The last primary sources I have used are the writings of nineteenth-century 
foreigners in the central Guangdong area, including tourists' accounts, Consular Reports, 
and, in particular, reports from missionaries.29 These writings are useful as reflections of 
contemporary views of the Red Turban Rebellions, the Hakkas and speculations about 
their identity. These works often contain translations of narratives written by Hakka and 
Punti leaders that supplement the information contained in other local sources. They can 
also be used, to some extent and with much caution, as ethnographic records of Hakka 
life and customs of the time. 
Finally, a note about conventions. First, I use Pinyin Romanization throughout. 
Second, with regard to place names, I have tried to use their late imperial forms, with a 
29 Most of these sources were scattered in church publications such as The Chinese Review and The Chinese 
Recorder and in the twenty volumes of the Chinese Repository and in the British Foreign Office file in the 
Public Record Office, London. 
21 
preference whenever possible for the ca. 1850 forms. Also, although there were various 
names (xiang, ting, etc.) for the lowest-level political unit in the late imperial state, I 
translate all as "county." 
22 
CHAPTER II 
ENVIRONMENT 
By "environment," we refer very broadly to the wider setting--in its physical, 
economic, political, and historical aspects. Although we can as yet propose no elegant 
hypotheses for the precise relationship between rebellions and environ.'Ilental change, we 
are certain of the importance of this exploration.30 Was there an environmental crisis in 
the form of famine, overpopulation, or alteration in the economic activities of the 
Guangdong inhabitants that accounts for the surge in the powerful strain of rural 
movements during the mid-nineteenth century? The testing of this question involves an 
inquiry into local history of a new sort. To be sure, it is regional history, but the actions of 
30 Some scholars of China are ascribing outbreaks of popular movements and rebellions to ecological 
causes and to view them as "survival strategies" in environments, where resources were in short and 
unpredictable supply. For instance, Perry's accounts ofrebellions in Huaibei (I 989) emphasize the 
ecological circumstances that surrounded peasant life in North China. She contends that the Nian rebellions 
result from rational strategies of survival. Whether a particular strategy was "predatory" or "protective," it 
was conceived of as a sustained form of collective action that could, and often did, end in a rebellion when 
ecological conditions worsened. See Elizabeth J. Perry, Revels and Revolutionaries, 1845-1945 (Stanford: 
Stanford University Press, 1980); Robert B. Marks, Tigers, Rice, Silk, and Silt: Environment and Economy 
in Late Imperial South China (Cambridge: Cambridge University Press, 1998), 1-15. 
23 
the locals had ramifications that drew them ultimately into national and international 
spheres where they could not escape contact with outsiders. Therefore, it may be useful 
for us to discuss important events ofthis decade such as the Opium War and the Taiping 
rebellion along with close analysis of the various local conditions in Guangdong and 
especially in the Pearl River Delta before moving into the subject of the Red Turban 
Rebellions. 
1) The Physical Environment 
Geographical Environment 
With the exception of Hainan Island, which was separated from Guangdong and 
redesignated a new province in 1988, Guangdong is the most southerly of the 22 
provinces and 5 autonomous regions in China. It has an area of 85,000 square miles 
representing merely 1.85% of China's total area. However, its coastline is the longest of 
all provinces, amounting to 1,560 miles, or 10.52% of the country's total. This is 
important for coastal and riverine shipping, as well as providing fishing and other marine 
resources. Guangdong's population in 1995 reached 68.68 million, accounting for 5.67% 
of China's population. This gives a population density of 382 persons per sq km, making 
24 
it the ninth densest of all provinces in China.31 In the early nineteenth century it had 
probably about twenty million people, and was one of the richest provinces in the 
empire.32 
Topographically, Guangdong is separated from the rest of China by the east-west 
N anling l¥i ffi range. For the most part the province consists of fluvially dissected, rolling 
to rugged hill country, with more mountainous country to the north and northeast. Only 
23% of the province can be classified as plains: the major areas including the Pearl River 
(Zhujiang :i?KtL) and Han (~) River Deltas, the Leizhou (ffi' 3-M) peninsula and coastal 
plains surrounding Jianjiang (~1-L), and the karst plains and plateaus to the north and 
northeast. Much of the province lies south of the Tropic of Cancer and thus Guangdong is 
the only province, along with Hainan, with tropical and subtropical climates. 
What distinguishes Guangdong from most parts of east China is not its summer 
temperatures but its much higher winter temperatures. Rainfall displays a pronounced 
31 For the best bird's-eye views ofGuangdong, see Guangdongsheng minzhengting ~*ir ~TI&Je, ed., 
Guangdongsheng zhengqu tuce ~*~'TI&lllilffil-, (Guangzhou: Guangdong sheng ditu chubanshe, 1995); 
Yoo Jangkun ru<?~~. "Yangkwangjiyuckyijayunkwa inmoon hwankyung ~JJ::ttg:!;gt.9.l § ~ xl-
AX~il," Joongkookyunkoo (Fall, 1993): 26-43; Robert B. Marks, Tigers, Rice, Silk, and Silt: 
Environment and Economy in Late Imperial South China (Cambridge: Cambridge University Press, 1998), 
16-52. 
32 Ho Ping-ti, Studies on the Population ofC hina, I 368-1953 (Cambridge: Harvard University Press, 1959), 
52. 
25 
summer maximum, with the rainy season lasting from mid-April to mid-October. 
Typhoons coincide with this rainy period and their occurrence can cause widespread 
human and material destruction. The mild winter means that two or three crops of rice 
can be grown everywhere in the province.33 Yet as early as the eighteenth century 
Guangdong had to import rice from neighboring Guangxi and abroad, chiefly from Siam 
(Thailand). 
This was due mainly to the rising population and the shortage of arable land, as 
well as to the fact that much of the arable land was devoted to growing commercial 
crops--sugar, cane, fruits, tea, tobacco, mulberry trees for silk production, etc.--instead of 
rice.34 Most of the arable land was concentrated in the Pearl River Delta and along the 
West River, as well as in the Han River delta in eastern Guangdong. Most of the plateau 
areas and even many of the lower, but steep hills were never cultivated. 
33 Yoo, "Yangkwang," 26-29; Robert B. Marks, Tigers, 17-31 ;Guangdongsheng zhengqu, passim. 
34 Chen Han sheng, Landlord and Peasant in China: A Study oft he Agrarian Crisis in South China (New 
York: International Publishers, 1936), viii, 1-2; Tang Shen Jj~ and Li Longqian *fffiM , "Mingqing 
Guangdong Jingji Zuowu de Zhongzhiji Yiyi ljJ:j:i-~~~j([~ff.4ml¥ifU1.&g~," in Mingqing 
Guangdong ShehuiJingji Xingtai Yanjiu ~jg }Jf .~::t±lt~i1f:iiJf o/i (Guangzhou: Guangdong Rerun in 
Chubanshe, 1985), 1-21. 
26 
Regional Differences 
The geomorphology of Guangdong is diverse, being made up for the most part of 
rounded hills, cut by streams and rivers, and scattered and ribbon-like alluvial valleys. 
Thus, as a physiographic unit, Guangdong could be divided into four subregions that 
more or less fit together for convenience of our analysis. These regions include (1) 
eastern Guangdong consisting of Chaozhou ~ 1-M prefecture and Jiaying 3 If!. 
independent department; (2) central Guangdong composed of Guangzhou ,- )+I, 
Zhaoqing -~. Huizhou ;!)+I, and Shaozhou ffij)+I prefectures, as well as the 
independent departments (Zhilizhou ~~V-1'1) ofLuoding ~IE, Lianzhou :iJlU-M, and 
Nanxiong P?J1i.t, and the independent sub-prefectures (Zhiliting ~~ft) ofLianshan ill! 
LIJ, and Fogang f~[tqj; (3) western Guangdong composed ofGaozhou ~ )+I, Leizhou, 
Lianzhou ffltfi'I, and Qiongzhou lifM (Hainan) prefectures; and (4) the "water region" 
comprising the "inner ocean" on the South China and the Gulf of Tonkin along 
Guangdong's coast. The regions were further divided into over eighty local level 
administrative units: counties (xian M), prefectures (zhou HI), and subprefectures (ting 
35 For discussion on the administrative subdivisions in the Qing period see H. S. Burnett and V. V. 
Hagelstrom, Present Day Political Organization of China (Shanghai: Kelly and Walsh, limited, 1912); G. W. 
27 
According to G. William Skinner's model of economic macroregions, most of 
Guangdong fell into a single macroregion. Except for the eastern prefectures of 
Chaozhou ~ fM and Jiaying ~ I!, most of the province of Guangdong falls within the 
Lingnan itl¥J macroregion, which also includes most of the province of Guangxi directly 
to the west of Guangdong. Chaozhou and Jiaying prefectures fall into the more 
fissiparous Southeast Coast macroregion, which extended northward and included most 
ofFujian province and parts of southern Zhejiang province. Delineating the eastern edge 
of the Nanling range, the southeastern uplands comprise a separate geologic formation.36 
Skinner, "Cities and the Hierarchy of Local Systems," in City in Late Imperial China, ed. G. W. Skinner, 
275-352. (Stanford: Stanford University Press, 1976); and on Guangdong in particular see "Kwangtung 
Tung-chi, or a general Historical and Statistical Account of the Province of Canton," in The Chinese 
Repository XII (1843), 309-327. Skinner divides Guangdong into nine subregions: Hainan, North Hakka, 
Northeast Hakka, West Peripheral, Northwest, West Central, East Central, Teochiu, and Central. However, 
my discussion on the four regions of Guangdong has mainly benefited from Edward Rhoads, Chinas 
Republican Revolution: The Case of Kwangtung, 1895-1913 (Cambridge: Harvard University Press, 1975), 
9-13 , and Robert James Antony, "Pirates, Bandits, and Brotherhoods: A Study of Crime and Law In 
Kwangtung Province, 1796-1839" (Ph.D. diss. , University of Hawaii, 1998), 23-36 and Murray, Pirates, 
chap. 1. 
36 G. William Skinner's spatial scheme ofmacroregions, cores, and peripheries for explaining economic 
change has been challenged by many scholars. Analyzing China as a tenuous concentration of numerous 
regional systems whose individual political, social, and economic characters are more closely determined 
by internal than by external conditions is certainly appealing. There is an objection on the Skinnerian model. 
Economic activity is not that easily circumscribed. For instance, the Hakka heartland of Guangdong, 
Jiayingzhou, had ties to the Lingnan area, the central Yangzi valley area, and the southeastern coastal area 
(through Zhangzhou), but also formed a socioeconomic sub-unit of its own. As Skinner's model implies, 
there were no convenient lines of communication either overland or via waterways between the central 
areas and the eastern regions, but the two areas were hardly isolated from each other. Coastal shipping was 
highly developed and readily linked eastern Guangdong with other parts of the province. According to Fan 
1-chun ( 1992), interport trade in Guangdong was at least as prosperous as the commerce of many other 
regions more actively involved in interregional trade. It should also be noted that despite the natural 
barriers that separated this region from the rest of Guangdong province, central government and provincial-
level officials observed the political and administrative boundaries of the province when formulating 
economic policies. This was readily apparent in official efforts to encourage land reclamation and to 
28 
Topographically distinct from the rest of Guangdong, the rivers of the eastern 
upland have a distinct pattern, flowing southeasterly in trellis-tributary patterns to the sea. 
East Guangdong is defined by the watershed between the Han River, which flows directly 
into the sea like the other river systems in the southeast coastal region and the East River, 
which has its outlet in the Pearl River Delta. Next to the Pearl River Delta, this area was 
the most developed in Guangdong at the time. Chaozhou prefecture had the second 
highest population density in the province in 1820, with 151.46 people per square 
kilometer. Both rice and sugar were important crops there. The independent department 
of Jiaying in this area was a much less developed one, with a population density in 1820 
estimated at 146.01 people per square kilometer. Most of Jiaying was settled by Hakkas, 
while Chaozhou was mainly populated by a mixture of Hakkas and Hoklos. 37 
alleviate overcrowding in eastern Guangdong during the eighteenth centwy. Official policy envisioned 
overpopulation as a provincial problem and promoted migration from eastern prefectures, including Jiaying 
and Chaozhou, to western prefectures. On the other hand, research on the structure ofGuangdong's rice 
market (Robert B. Marks, 1991) tends to confirm the Skinner model. Statistical tests indicate that Chaozhou 
and Jiaying prefectures were not as closely integrated into the Guangdong grain market. See G. William 
Skinner, "Regional Urbanization in Nineteenth Centwy China," in City in Late Imperial China, ed. G. W. 
Skinner, 212-49 (Stanford: Stanford University Press, I 976); Fan 1-chun, "Long Distance Trade and Market 
Integration in the Ming-Ch'ing Period, 1400-1850" (Ph.D. diss., Stanford University, 1992), 249; Robert B. 
Marks, "Rice Prices, Food Supply, and Market Structure in Eighteenth-Centwy China," Late Imperial 
China 12.2 (December 1991), 64-116. For the debate surrounding the Skinnerian model, see Barbara Sands 
and Ramon H. Myers, "The Spatial Approach to Chinese History: A Test," Journal ofA sian Studies 45, no. 
4 (August, 1986), 721-43 ; Daniel Little and Joseph W. Esherick, "Testing the Testers: A Reply to Barbara 
Sands and Ramon Myers's Critique ofG. William Skinner's Regional Systems Approach to China," 
Journal of Asian Studies 48, no. I (Feb., 1989), 90-99. 
37 Yoo Jangkun, "Yangkwang," 36-37. 
29 
Much of eastern and northern Guangdong consists of dissected uplands, but of 
the greatest economic value are alluvial deltaic plains formed by the three major rivers of 
the Pearl River system. Guangdong province is largely drained by three tributaries of the 
Pearl River, comprising the East River (Dongjiang), the North River (Beijiang) and the 
West River (Xijiang). Only the southwestern and northeastern parts of the province are 
not drained by tributaries of the Pearl River. The Pearl River Delta, referring to the area 
of the West River beyond the confluence, forms the core of not only central Guangdong 
but also of the entire province. It embraces the provincial capital, Guangzhou in the 
fertile and economically advanced Pearl River Delta. Estimates place the population of 
Guangzhou and its surrounding area about 580,000 in 1820, more than a third of the 
province's population (See Table 1) . 
The Pearl River Delta, which is the geographical focus ofthis study, is a filled-in 
bay, rather than a true delta; it is a flat country with isolated hills--formerly islands in the 
bay--rising sharply from the alluvium. The delta covers close to 11 ,300 square kilometers, 
of which approximately one-quarter is either water or hills and the remainder alluvial 
plains.38 The delta has been one of the most intensively cultivated regions in the world. 
38 Chen Zhengxiang ~.IE.ff, Guangdong Dizhi Jjf *ftl!.~ (Hong Kong: Tian di tu shu you xian gong si, 
1978), 162 
30 
Table 1. Population Densities (Guangdong, 1820)39 
County Population Size (km2) Population Density 
Total 21 ,197,741 205,500 103.15 
Guangzhou 5,799,261 18,900 306.84 
Jiayingzhou 1,314,050 9,000 146.01 
Chaozhou 2,180,905 14,400 151.46 
Gaozhou 2,335,516 15,600 149.71 
Zhaoqing 2,516,149 18,600 135.28 
Nanxiongzhou 332,161 3,300 100.65 
Luodingzhou 674,816 6,900 97.79 
Dianzhou 654,256 7,800 83.88 
Huizhou 2,194,896 30,000 73 .16 
Shaozhou 1,021 ,482 15,900 64.24 
Lianzhou :i!fH 298,500 6,000 49.76 
Qiongzhou 1,324,068 34,500 38.38 
Lianzhou 444,870 16,200 27.40 
Fogang 52,299 3,300 15 .84 
Lianshan 54,512 5,100 10.68 
Traditionally, the two main crops were rice and mulberries. Rice is the 
predominant product in the lower areas where conditions for irrigation are ideal and 
where the climate permits two and three harvests. On higher ground, especially to the 
west ofDaliang -f:. ~' mulberries used to take first rank, forming a basis for a silk 
industry. Fisheries in the canals and ponds are common and used to be skillfully 
associated with the silkworm industry as the fish add to the fertilizing qualities of the 
39 This table is based on Liang Fangzhong's study. See Liang Fangzhong ~1Jf9l, ed. , Zhongguo lidai 
renkou, tiandi, tianfu tongji r:p lllnf~A lJ , 133 :l:tl!., 133lWt-i.-t, (Shanghai: Shanghai renmin chubanshe, 
1980), 277-78. 
31 
bottom mud used as manure for the mulberries and paddy fields and, in turn, the 
discarded cocoons are used to feed the fish. 40 
In 1820, the population density over the delta was calculated to reach the then 
already very high figure of 306.84 per square kilometer. This figure reflects primarily the 
intensity of agricultural production. The region, however, has great commercial and 
industrial significance as well and, apart from Guangzhou, there are many large cities 
such as Foshan, Shunde, and Daliang. Beside the core delta, the central Guangdong 
region also includes the areas along the West, North and East Rivers. The West River, an 
important artery for grain supply, originated in Guangxi province and flowed across 
Gaoyao ~~ and Deqingzhou tlJffM, Zhaoqing, Shanshui wl.K, and Guangzhou, on its 
way to the Pearl River Delta. Before the Opium War the North River was the main artery 
linking Guangdong with the north; the imperial highway followed this river to the 
Meiling pass into Jiangxi province and thence finally to Beijing. Nanxiong, Yingde ~ 1l, 
Shaozhou, Qingyuan $i!r:ii, Panyu :ff /Ji, and Nanhai are the major districts along the 
North River. The East River entered Huizhou prefecture from the northeast and passed 
through the commercial centers ofGuishan ~ff, Heyuan iiiJ~, Longquan ffi.)1 1, and 
40 Yoo Jangkun, "Yangkwang," 29-31 ; Miao Hongji ~~~, Zhujiang Sanjiaozhou Shitu Ziyuan 
~rr= fiHJ.lhK ± ~tJ: (Guangzhou: Zhongshan Daxue Chubanshe, 1988), 1-7. 
32 
Huizhou counties, before reaching the Pearl River Delta.41 
Western Guangdong was the least developed region in the province in the 
nineteenth century. It includes the coastal belt, largely the Leizhou Peninsula, Gaozhou 
prefecture, and Lianzhou prefecture, contains considerable lowlands drained by short 
rivers flowing into the South China Sea or the Gulf of Tonkin. Some fifteen-mile offshore 
lies Hainan Island, an essentially mountainous continuation of the same geological 
structure as the southeastern coastal region, with peaks in the central range reaching 
5,500 feet. Narrow alluvial lowlands, mostly in the north facing the Leizhou Peninsula, 
ring the island. Lianzhou was an important sugar producing area. The ethnic mixture of 
the entire region was one of the most diverse in all of Guangdong. There were not only 
Cantonese speakers but also the Hakkas and the Hoklos. On Hainan there was also a large 
population of native Li aborigines.4 2 
Finally, the fourth and largest region encompassed the "inner ocean" off the coast. 
Guangdong is a maritime province with a coastline of over 1,560 miles, and is dotted 
with hundreds of small islands. If the people living inland were dependent on agriculture 
4 1 See G. W. Skinner, "Cities," 275-351. 
42 Edward Rhoads, China's Republican, 11-12; Robert B. Marks, Tigers , 25; Robert J. Antony, "Pirates," 35. 
33 
for their livelihoods, then those living along the coast depended upon the sea for their 
livelihoods as fishermen, sailors, and merchants. In the early nineteenth century, it is 
estimated that seventy percent of the coastal inhabitants were fishermen and that only 
thirty percent were farmers. Dominating the seafaring communities were the Tank.a 
(danjia ~~) "boat people." Treated as outcasts, they made their homes aboard their 
boats and engaged mainly in fishing for a living. When fishing or other legitimate 
endeavors failed they would often turn to piracy and smuggling.4 3 
Guangzhou and the lower Pearl River Delta were also the hub of this vast water 
world. The delta is so dissected by countless river channels that the geographical 
arrangement is like an inland sea dotted with numerous islands. In this geographical 
setting, bandits and pirates could not limit their activities to either the land or the river for 
the simple reason that they could not financially maintain their gangs by plundering a 
limited area. Nor would it be to their advantage to confine themselves to a limited area in 
case they were besieged by the imperial navy or even by other gangs ofbandits.44 The 
43 Robert, "Pirates," 35-36; Murray, Pirates, 7-10. 
44 Many of foreign and domestic merchant ships, as well as native fishing vessels, and coastal villages, fell 
prey to larger, well-armed bands of pirates who virtually ruled this water world during the first decade of 
the nineteenth century. See Robert, "Pirates," 35-36; Laai Yi-faai, "The Part Played by the Pirates of 
Kwangtung and Kwangsi Provinces in the Taiping Insurrection" (Ph.D. diss ., University of California, 
1950), 25-32. 
34 
bandits consequently had to have a fleet or some other means of river or sea 
transportation so as, on the one hand, to extend their activities to a wider area in order to 
obtain more booty and, on the other hand, to enable them to escape to other localities in 
case of necessity. He Liu fiiJ /\, the chiefleader of the Red Turbans (1854-56), and all his 
subordinates had long practiced banditry and piracy in the delta. It is then to these bandits, 
who established rebel kingdoms, that we shall turn our attention in Chapter III since they 
posed the most formidable threat to the stability and well-being of the province. 
2) Historical and Social Environment 
Early History 
The Guangdong region, known as Nanyue 1¥i~ , Baiyue Ef J1f and Lingnan ~ffi 
in the history of China, was originally inhabited by Tai people and until the Ming it was 
culturally divorced from its hinterland. Guangdong had already had contact with the 
northern Chinese during the Spring and Autumn Period (770-476 B.C.). During the latter 
part of the Warring State Period (475-221 B.C.), the Han people started migrating to the 
region from the north, resulting in the gradual integration of the migrants and the native 
Yue people through interaction. Subsequently, the three major dialects spoken in today's 
35 
Guangdong, namely Cantonese, Chaozhou and Hakka, were developed at different 
periods in history. It was during the time of the First Emperor of Qin (221-210 B.C.) that 
closer ties be~ween the region and North China were forged. The Qin state established in 
south China three commanderies, of which Nanhai, with its seat at Panyu (later 
Guangzhou), was one. The Qin emperor brought one hundred thousand or so troops there 
and built the canal linking the region with the north. Subsequently, relatively advanced 
technology was introduced, thus bringing about improvements to its agriculture and 
handicrafts. The Qin-Han city of Panyu was tiny in comparison with later Guangzhou, 
but it established early its position as a port of call for ships coming in from Southeast 
Asia. Later, a Qin general named Zhao Tuo .MifB (d. 137 B.C.) founded the independent 
kingdom ofNanyue, and built Guangzhou's first walls and palaces. By the Tang (618-907 
A.D.), Guangzhou was a major international port, with its Indian and Persian 
communities. The building of the Dayu -:;k~ road in the Tang period greatly facilitated 
communication with, and the migration into, Guangdong from the north.45 Gradually, the 
Guangdong regio.n and China proper merged into one in terms of culture and ethnicity.46 
45 The road connects Guangzhou with Dayu, Jiangxi. 
46 Useful summary accounts of Guangdong history are Jiang Zuyuan li#Jfill~ and Fang Zhiqin 15;::t~ et al. , 
Guangdong Tong Shi. Gudai /lJIUm~. if~ (Guangzhou: Guangdong gao dengjiao yu chubanshe, 1996; 
idem, Jianming Guangdong Shi fffi13A./J:l!~ (Guangzhou: Guangdong renmin chubanshe, 1993); Lu 
36 
However, the large-scale development of the Guangdong did not take place until 
the Southern Song dynasty (1127-1278), when it experienced rapid growth in riverine and 
oceanic trade. Market towns and metropolises shot up along the navigational routes. The 
region's water transport network had had an intrinsic relationship with its social and 
economic growth. Guangdong's links with Central and North China through water 
transport networks and maritime trade facilitated the movement of both goods and people 
between Guangdong and the north, thus furthering the region's integration with China 
proper. Beginning with the Ming dynasty, however, restrictions or a total ban on maritime 
travel and trade drastically weakened the region's link with the outside world. They also 
seriously disrupted Guangdong's once flourishing foreign trade. By the Ming-Qing 
transition years, economically the province was divided into five sub-regions, namely, the 
Pearl River Delta, East, North, and West Guangdong and Hainan Island. Guangzhou, 
endowed with a water transport network, became the economic center of South China in 
the sixteenth century and enjoyed prosperity with a booming foreign trade.47 
Mingzhong § i5 $, "Hanzu nanqian yu Lingnan Baiyue diqu de zaoqi kaifa ~~lfHlrm.~ffis • 
j:jg[Hf9Ej!m.J00~," Zhongguoshiyanjiu tj:i~~lilHE, No. 4 (1984), 17-24; an account on Zhao Tuo may be 
found in Harold J. Wiens, Chinas March Toward the Tropics (Hamden, Conn. : The Shoe String Press, 
1954), 33-141. 
47 For an excellent discussion of Song-era Lingnan, see Han Maoli lU.itfiJ, "Songdai Lingnan diqu nongye 
dili chutan * ft:~ffij:jg 1£./!Jij:jgfijf}J~," Lishi dili ~ ~ j:jg3:11.[ II (1993), 30-34. 
37 
The Manchu invasion of China in the seventeenth century rearranged the 
regional structure of Guangdong. In its early years the Qing dynasty encountered a 
serious challenge from anti-Qing elements. Several princes in the south attempted 
restoration of the Ming dynasty. The first three attempts were, all quickly suppressed by 
Manchu Bannermen and Chinese troops loyal to the Manchus headed by Li Chengdong 
$.,6,x;~. However, Li Chengdong reversed his allegiance and invited another Ming prince 
to return to Guangdong as an emperor in 1648. He was defeated in 1649, and in the 
following year Guangzhou was once again surrounded by Manchu troops. The Manchus 
responded to this resistance with violence and cruelty, killing over 100,000 inhabitants.48 
The Ming loyalists of Guangdong, then emigrated in large numbers to Taiwan to continue 
their struggles under Zheng Chenggong J~px;J}J (Koxinga), whose supporters were 
continuously slipping into Fujian and Guangdong coastal villages for supplies and 
intelligence. In order to forestall any possible collaboration of the coastal population with 
Ming loyalists, the Manchus adopted the proposal of Huang Wu :Jiffi in 1662 and 
ordered all coastal Chinese to move inland about fifty li ( about fifteen miles) and 
48 J. A.G. Roberts, "The Hakka-Punti War," 23. 
38 
prohibited any vessel, mercantile or fishing from going out to sea.49 The implementation 
of these orders was uneven, but Guangdong was very seriously affected. The local 
gazetteer describes the drastic policy of the Qing: 
The government suspected that the people in this area still maintained contact 
with the rebels overseas. The officials surveyed the coastline and built barriers 
along its entire length at a distance of fifty Ii from the beach. All people living 
between the beach and barrier were removed inland behind the 
barrier . .. Violators were immediately put to death. The populace in these areas 
was mercilessly mobilized to build fortresses, bunkers, and watchtowers, and 
were taxed to maintain the military garrison. These things caused tremendous 
suffering among the people, and the whole area was devastated. Many people 
became homeless, and some of them starved to death on the roads.50 
Still, the Sanfan - $ Rebellion, staged by Han Chinese military leaders of 
Guangdong, Yunnan, and Fujian in 1673, was not suppressed until 1681 and gave the 
Manchus a lasting suspicion of the loyalty of the south. Resistance from the Guangzhou 
people took other forms, too, as countless people risked their lives to harass bannermen, 
to circulate revolutionary poetry, or to participate in the secret societies which continued 
to be the champions of anti-Manchu elements in times of distress. However, gradually the 
49 See his biography in Arthur W. Hummel, ed., Eminent Chinese oft he Ch 'ing Period (I 644-19 I 2) 
(Washington: United States Government Printing Office, 1943-44), 355. 
50 X inning XZ, 14:5a. 
39 
Manchus subdued Han Chinese resistance and gained the support of Han elite groups by 
adopting almost wholesale the Chinese system of government. 51 
Records indicate that some eighty five percent of the coastal population of 
Guangdong perished during the period of coastal evacuation. 52 The order to return was 
given in 1684, but the groups who repopulated the coastal regions were in many cases not 
the same as those who had been expelled. Among those moving into coastal parts of 
Guangdong were the Hakkas and Minnan, always alert to the possibility of better lands. 
Considerable competition between the Hakka and Yue communities marked the 
reoccupation of coastal lands during this period, reflecting increasing competition for 
land elsewhere in the Guangdong and Guangxi areas. 53 
51 Wakeman, Strangers, 117-25; Wong, Yeh Ming-Ch 'en, 26. 
52 Ibid. 
53 Aspects ofMing-Qing dynastic transition have attracted a good deal of attention about the brief reign of 
Prince Tang of the Ming in Fuzhou in the 1640's as well as the career of Zheng Chenggong and family. The 
effects of the coastal evacuation in the 1660's were also broadly outlined. On the late Ming court, see Lynn 
A. Struve, The Southern Ming, 1644-1662 (New Haven, Conn. : Yale University Press, 1984); for an Zheng 
Chenggong, see Ralph C. Croizier, Koxinga and Chinese Nationalism: History, Myth, and the Hero 
(Cambridge: Harvard University Press, 1977) and John Robert Shepherd, Statecraft and Political Economy 
on the Taiwan Frontier (Stanford: Stanford University Press, 1993), 91-104; on coastal evacuation, see 
Hsieh Kuo-ching, "Removal of the Coastal Population in Early Tsing Period," Chinese Social and Political 
Science Review 15 (1932), 559-96 and Luo Xianglin, Kejia yuanliukao. 
40 
Population Growth 
With the 1683 Qing capture of Taiwan and the reopening of the coast for 
shipping and foreign trade, Guangdong as well as the rest of China entered a period of 
relative peace for over 150 years. 54 Consequently, the population grew steadily after the 
consolidation of Qing rule in the late seventeenth century. The population of China as a 
whole nearly doubled between the mid-seventeenth and mid-eighteenth centuries, 
jumping from 150 million in 1650 to 270 million by 1776. Guangdong also shared in this 
demographic explosion; the population reached 14.8 million after having_ been recorded 
as 6.4 million in 1749, and it doubled again to 28 million in 1851.55 Although a recent 
reexamination of Qing population history suggests that limitations on the government's 
capability to compile accurate population data during the late Ming may have made the 
increase more dramatic than it was during the transition from Ming to Qing, most 
historians would agree that nationwide, population had roughly doubled by the end of 
54 Historians have eloquently described the grand achievements of this unprecedented era of peace and 
sustained economic and territorial expansion known as the "High Qing." See Frederic Wakeman, Jr., 'High 
Ch'ing: I 683- I 839," in Modern East Asia: Essays in Interpretation, ed. J. B. Crowley, 1-28. (New York: 
Harcourt, Brace and World, I 970) and Susan Naquin and Evelyn Rawski, Chinese Society in Eighteenth 
Century (New Haven: Yale University Press, I 987). 
55 Dwight H. Perkins, Agricultural Development in China, 1368-1968 (Chicago: Aldine Publishing 
Company, 1969), 207, 216; Ho Ping-ti, Studies on the Population of China, 1368-1953 (Cambridge: 
Harvard University Press, 1959), 283. See also Table I. 
41 
eighteenth century. And it is also clear that Guangdong continued to share in its unique 
movement of population growth that did not end until the outbreak of the Tai ping 
rebellion in 1851. 56 
In itself, a rising population does not necessarily imply the existence of an 
economic problem, and under other circumstances it might indicate an increase in 
prosperity. 57 In South China, however, difficulty arose because the increase in population 
was not met by a comparable increase in production. As early as 1723 there were 
indications of what might lie ahead, for at about that time Guangdong became rice-
importing area. 58 According to Ho Ping-ti, the whole country arrived at the "optimum 
condition," that is the point when a population reaches its maximum economic welfare 
for the technological level of the day, between the years 1750 to 1775.59 Thereafter 
56 William Lavely and R. Bin Wong, "Revising the Malthusian Narrative: The Comparative Study of 
Population Dynamics in Late Imperial China," Journal ofA sian Studies 57.3 (August 1998), 714-48. 
57 Although population continued to grow while new sources of high-quality arable land were increasingly 
hard to find throughout the eighteenth century, it would be a serious mistake to suggest that China was on 
the verge of a Malthusian crisis in the eighteenth century. Recent research by William Lavely and R. Bin 
Wong has found "scant evidence that Chinese mortality rates in the late eighteenth and early nineteenth 
century exceeded those of agrarian Europe before the Industrial Revolution," and nutrition and living 
standards were not significantly lower nor were they declining in the early nineteenth century. See Ibid., 
738 
58 Hsiao Kung-chuan, Rural China: Imperial Control in the Nineteenth Century (Seattle and London: 
University of Washington Press, 1960), 381. 
59 Ho, Population, 270. 
42 
further increases in population only led to a lowering of the standard of living. The 
population of China was still rising rapidly and the decline in the standard of living soon 
became noticeable. 
In Guangdong the indications are that in the years following the attainment of the 
"optimum condition," the rise in population continued unabated into the nineteenth 
century. The population pressure was not accompanied by a comparable increase in 
arable land. The amount of cultivated land in Guangdong increased by just 75 percent 
from 1673 to 1853 while the amount ofland (mu) per capita fell by about 50 percent in 
Guangdong during the same period, reflecting the greater population density there. 
Guangdong was among the most land-hungry provinces of the empire with less cultivable 
land per person than all other provinces but Guizhou, Guangxi, Gansu, Fujian and 
Anh · 60 Ul. 
The areas that gained relative weight in Guangdong were in Huizhou and 
Gaozhou prefectures. The Pearl River Delta area, though quite fertile, was much more 
densely populated than the rest of Guangdong. It experienced the largest growth in 
population and, with about 300 people per square kilometer, was twice as densely 
60 See Perkins, Agricultural, 207-240; Marks, Tigers, 280-281 ; and Wakeman Jr. , Strangers, 179. 
43 
populated as Gaozhou, the next most densely populated prefecture (Chaozhou and 
Jiaying prefectures, which also had population densities about the same as Gaozhou).6 1 
Land Reclamation 
Early Qing land reclamation policies during the Shunzhi /1~ mr eign ( 1644-1661) 
were formulated in a context quite different from that of the eighteenth century 
population crisis. According to one estimate, only 30% of the land under cultivation 
during the reign of the Ming emperor Wanli 7-itf! (1573-1619) was being tilled in 1651 
during the Shunzhi emperor 's reign. The land was abandoned not just because people 
died during chaotic war periods, but also because they fled from war or taxes and because 
of the relocation of the coastal population. More land was abandoned than can be 
accounted for by the fall in population. Many either took to the hills to eke out a living 
until better times returned or joined up with any of the numerous bandit and pirate gangs 
operating at the time.62 
The Qing government acted promptly and initiated a series of measures designed 
61 See Ibid. 
62 Jiang Taixin IT::t:ffi and Duan Xueyu f&~:E:, "Lun Qingdai qianqi tudi kentuo dui shehui jingji fazhan 
di yingxiang Mu~rtrwM±:i1!ff tfiltffd:wt~lfHIJ.iJIE~~," Zhongguo Jingjishi y an)iu lf1 ~~~ ~ :iff ~ 
1 (1996), 47. 
44 
to alleviate economic distress and restore the government's fiscal base. Abandoned land, 
which fell into two categories--"abandoned lands without owners" (wuzhu huangdi ~± 
Jit:l:tP.) and "abandoned lands with owners" (youzhu huangdi ~ ±Jit±-tl!)--was a critical 
problem. For example, a survey of nine prefectures in Shanxi province revealed that 80% 
of the land was abandoned.63 In 1644, the Shunzhi emperor issued an edict that allowed 
dislocated peasants to till "ownerless" land, and provided material assistance and tax 
breaks to landlords to encourage them to find new tenants. In 1649, another edict allowed 
dislocated peasants to take permanent possession of ownerless land and to be enrolled in 
the local baojia * l:j3 register. However, peasants and landlords remained cautious in 
their efforts to reclaim land and did their best to hide land from the tax rolls. Peasants 
feared not just taxes but the more onerous corvee labor obligations, while landlords 
sought to avoid taxes. 64 
The Kangxi nti¥~ emperor (1662-1722), with the firm establishment of Qing rule 
63 Peng Yuxin ~ffiffi, Qingdai tudi kaiken shi :m f~± ±tl!.~~..'t (Beijing: Nongye Chubanshe, 1990), 6. 
64 Ibid., 7-12, 34-39. Historically, tax remissions were a form of disaster relief, but during the Qing dynasty 
tax remissions were also used to stimulate economic recovery. Some historians have argued that in addition 
to economic considerations, Qing emperors also saw tax remissions as a means to alleviate class and ethnic 
tensions. Whatever interpretation one prefers, tax remission policies clearly had political as well as 
economic motivations. See Hu Chunf~ ti;lj~ljlfL, Hua Yu ¥;:f:3, Huang Shiqing ~tltffl, and Wen Ji X~, 
"Shilun Qing qianqi dijuanmian zhengze ~Mti~1ltr:IW:t!B~5Ei&m," Qingshiyaryiuji ~~~~~ 3 
(Chengdu: Sichuan Renmin Chubanshe, 1984), 150-65. 
45 
over China, extended tax concessions on newly reopened land to three, six or ten years 
depending on the region and quality of the land in 1693. He also supplied more funds for 
seeds, tools, and the repair of irrigation systems. These efforts apparently succeeded and 
provincial officials no longer emphasized dislocated peasants and abandoned lands as 
serious problems. 65 After the land tax problem was solved with the generous grants of tax 
remissions in the early eighteenth century, however, a new problem arose: there is 
evidence that the growing population was beginning to place strains upon the supply of 
easily available land in Guangdong. In 1710 the emperor, in an edict, observed that the 
population was increasing rapidly there beyond any comparable increase in the amount of 
land for cultivation. 66 
It was this conclusion that had led to the encouragement of schemes for the 
resettlement of farmers on uncultivated land that were to make some of poor counties of 
Guangdong the most important emigration area in the province during the early Qing 
65 Ibid. , 154. Tax concessions, however, created opportunities for tax evasion and brought a problem of 
"hidden land," which referred to reclaimed lands that were not reported to the central government for 
taxation. Apparently, this problem was widespread. Some high-level officials complained that despite 
seventy years ofQing rule, tax collections were only 10% of those collected in the Ming. For the taxation 
and land policy of early Qing government, see Madeline Zelin, The Magistrate '.5 Tale (Berkeley: University 
of California Press, 1984), Chapter 3; Wang Yeh-chien, Land Taxation in Imperial China, 1750-1911 
(Cambridge: Harvard University Press, 1973), 20-25; Peter Perdue, Exhausting the Earth: State and 
Peasant in Hunan 1500-1850 (Cambridge: Harvard University Press, 1987), 72-78. 
66 Hsiao, Rural China, 379. 
46 
period. For instance, Jiaying prefecture, with its ranges of low hills, is the poorest 
agricultural area in the whole province. Its five counties (xian) comprise five out of the 
seven xian in Guangdong that have the smallest proportion of cultivable land. In Xinning 
county, only 2.40% of the land can be considered suitable for cultivation, even by the 
twentieth century estimate.67 The combination of a large population pressing on limited 
food supplies, unsettled political and social conditions, and government resettlement 
policies all contributed to a massive population shift in southeast China. 
The agrarian economy of Guangdong had reached an important turning point 
early in the Qianlong $z~ reign (1736-1795). Easily accessible, arable land was clearly 
in short supply in Guangdong by this time, so the policy of tax exemption and of 
financial support for the reclaiming lands was further extended during this period in an 
effort to open new lands and encourage migration. The land that was brought into 
cultivation during this period was not old, abandoned farmland, but land that had not 
been farmed before, most of which was in the hills in the northern and western sections 
of Guangdong. More residents of Huizhou, Chaozhou, and Jiaying were invited to 
resettle the sparsely populated and unused land of poorer quality in Gaozhou, Leizhou, 
67 T'ien Ju-kang, The Chinese Sarawak: A Study ofS ocial Structure (London: L.S.E. Monographs on Social 
Anthropology No. 12, 1956), 3. 
47 
and Lianzhoufu prefectures, land suitable only for potatoes, paulownia, and tea.68 
According to a survey conducted in 1749, 67,988 mu had been reclaimed from the 75,782 
mu available in the western prefectures.69 
Alluvial Land and Lineages 
Alluvial land was the most important source of new land in Guangdong 
throughout the Qianlong period and was another important category of land that had been 
outside the original tax quota. The Pearl River Delta was mostly a sea of islands in the 
Ming dynasty. But, "river current and sea tide work daily and incessantly to bring 
sediments and to form sandy but exceedingly fertile lands; and by protection and the 
planting of a certain kind of grass thereon, they may be converted into cultivatable fields 
within three years."70 Land formed from silt deposit that collected in coastal areas at the 
mouths ofrivers or along riverbanks, was locally called Shatan tJ;:!:EI (sand flats). 
Through the river deposits, the sea of islands in the Ming dynasty was turned into a fertile 
68 Da Qing Gaozong Chun (Qianlong) Huangdi shi/u ::k~i%*tiE(~~).iF/W'.1Ufc (Taibei: Hualian 
Chubanshe, 1964 reprint), 37:2a-b. 
69 Gongzhongdang Qian/ong chao Zouzhe 'g i:pffl~~im~ffl. Nongye lei J1UtJ~ 33 (Taibei: Guoli 
Gugong Bowuyuan, 1982 reprint), Qianlong 24.3 .28. 
7°C hen Han-seng, landlord and Peasant in China (Westport: Hyperion, 1973), 28. 
48 
cultivatable river delta in the Qing dynasty. These alluvial fields (Shatian ¥.Y 83: sandy 
fields) occupied about one-sixteenth of all the cultivated land in the Guangdong province 
and were most prevalent in the six counties--Na nhai, Panyu, Xiangshan, Dongguan, 
Shunde, and Xinhui--located in the Pearl River Delta. 71 
Some of these sand flats certainly emerged by a natural process, but the majority 
were constructed. The general practice was to build an embankment across the river, and 
then to throw a large amount of big stones or iron pieces into the river to block the river 
current. Once the currents slowed down, silt deposition would be quickened, and alluvial 
sandy fields could be formed in a shorter time. 72 Because of the large amount of labor 
required to create shatan, most projects were undertaken by wealthy families. Before the 
Qing dynasty, the state reserved the right to build embankments in order to prevent the 
concentration of land holdings into the hands of a few gentry. Those few who received a 
state permit were not allowed to enclose more than 1,000 mu. However, in the Qing 
dynasty, when the gentry class had formed lineages and expanded its power at the 
71 Ibid., 22; Nishikawa Kikuko WJIIM'.'.J\ -r, "Qingdai Zhujiang Sanjiaozhou Shatiankao 
~~~iI= :fi3VH 1.'P EH~," in Lingnan Wenshi ~i¥i:X~, tran. Cao Leishi ff~:fi (Gungdongsheng Wenshi 
Yanjiuguan, 1983), 11-14. 
72 Peng, kaikenshi, 164; Marks, Tigers , 81. According to the seventeenth century writer Qu Dajun Jffi )( -Jt;;, 
after three to five years, the shatan would be ready for rice. See Qu Dajun Jffi -Jt;;, Guangdong xinyu 
./Ji:*ffi~ (Beijing: Zhonghua Shuju, 1985), 2:57. 
49 
expense of county magistrate, the gentry de facto had a free hand to build embankments 
and to enclose sandy fields. 73 
As early as 1739, officials complained that wealthy families and great clans were 
monopolizing control of shatan. 74 Recent scholarship on the Pearl River Delta also 
confirms the fact that arable land in that area was usually concentrated in a few hands. 75 
The lineages were the largest landowners of the Pearl River Delta, possessing as much as 
50% of cultivated acreage in the Pearl River Delta and as much as 30% for the whole 
Guangdong province.76 Many lands were owned not by individual landlords but by 
73 See Fu Tongqin {f!A]~, 69-71 , "Ming Qing shiqi di Guangdong shatian aJH~f1;¥WJ:l.1!!/J:JR t.'P 83" in Ming 
Qing Guangdong shehui jingji xingtai yanjiu /J:}Rffd::wI~MMfilHlf~. ed. Guangdong History Institute, 
65-74. (Guangdong: Renmin Chubanshe, 1985). 
74 Xiangshan.XZ, 1789, juan5, 16b. 
75 The He lineage of the Panyu and Dongguan ~ounties studied by Liu Zhiwei is the one of good example 
of this case. Even though Mark Elvin has suggested that since A.O. 1000, manorial landlordism had been 
slowly replaced by small holdings in many regions China, he has also indicated that his study does not 
apply to Guangdong and his reference on the Pearl River Delta shows that the few wealthy lineages 
controlled most of the lands. See Liu Zhiwei, "Lineage on the Sands: the Case of Shaw an," in Down to 
Earth: The Territorial Bond in South China, eds. David Faure and Helen F. Siu (Stanford: Stanford 
University Press, 1995), 21-43 ; Alvin Y. So, The South China Silk District: Local Historical 
Transformation and World-System Theory (New York: State University of New York Press, 1986), 37-45; 
Hsiao, Rural China, 383; Peng, kaikenshi, 34-41; Mark J. Elvin, "The Last Thousand Years of Chinese 
History," Modern Asian Studies, 4.2 ( 1970), 97-114; and Matsuda Yoshiro ~ 83 'a$~, "Minmatsu Shinsho 
Kanton Juko-deruta no Shaden Kaihatsu Shihai no Keisei Katei ~ * rfifJJ}1;;JR~IT7'A,,?7 O) f.'J;83 00 ~ .!:'. 
~JH'-$0)~~ %nt~ f¥," Shakai Keizai Shigaku ifttw/~~51:.~ Vol. 46 (1981) no . 6, 59-66. 
76 Robert Y. Eng, " Institutional and Secondary Landlordism in the Pearl River Delta," Modern China, Vol. 
12 No. 1 (Jan. 1986), 4; Wong, Yeh Ming-Ch 'en, 28. Sinologists have pointed out that powerful lineages 
and strong gentry class tended to locate in a river delta. See Maurice Freedman, Chinese Lineage and 
Society: Fukien and Kwangtung (London: Athlone, 1966); James Hayes, "Rural Society and Economy in 
Late Ch' ing: A Case Study of the New Territories of Hong Kong (Kwangtung)," Ch 'ing-shih Wen-t 'i 3 
(1976), 33-71 ; Jack Potter, "Land and Lineage in Traditional China," in Family and Kinship in Chinese 
Society, ed. Maurice Freedman (Stanford: Stanford University Press, 1970). 
50 
associations (tang ¥'.) formed by groups of land investors and represented by gentry 
members. 77 Powerful lineages and their individual members, as well as schools and 
temples, came to control large tracts of alluvial land. 78 Given the expenses ofreclamation 
and the possibly long period of gestation before profit realization, it was difficult for the 
poor and the weak to develop and hold onto the alluvial fields. The rural elite not only 
usurped the reclamation rights of the poor but also resorted to litigation, illegal methods, 
and even brute force to control alluvial land.79 
Guangdong was well known for strong lineages that served economic as well as 
social and religious functions. There is little argument that before the mid-nineteenth 
century, patrilineal kinship, particularly in the form of lineages, was the major principle 
of social organization in South China, as it was in the rest of the country. 80 Several 
77 During the nineteenth century, some of influential gentry or merchants of the lineage created "tang" to 
control more of lands in the alluvial field. For the details on "tang," see Ye Xianen pt~,~, "Ming Qing 
Zhujiang Sanjiaozhou tudi zhidu, Zongzu yu shangyehua 'lf:l:fi!r~rr = jij¥Jii±JtM-Mlt, *Jj~J=ijiffi~ft," 
Zhongguo Wenhua YanjiusuoXuebao i:p li:Xftliff~Jifi?¢m 6 (Guangzhou, 1997). 
78 One of best-known institutional landlords included the dominant lineages ofShawan 1j;-r!f in Panyu 
county and Mingluntang ll] {$¥:, the institution that managed school land in Dongguan county, controlling 
hundreds of thousands of mu in alluvial lands. See Eng, "Landlordism," 9; Hugh Baker, A Chinese Lineage 
Village: Sheung Shui (London: Frank Case, 1968), 99-13 l. 
79 ShundeXZ, 1853, 3:35-36; 26:29, 32:26;XiangshanXZ, 1920, 16:5-6. 
8°C h'en, Landlord and Peasant in China, 35. Large corporate lineages (zongzu *~) began to form and 
multiply in south China during sixteenth century. Lineage growth was related to the spread of extensive 
local trading networks and the development ofa more highly commercialized economy. See William S. 
Atwell, "International Bullion Flowers and the Chinese Economy Circa 1530-1650." Past and Present 95 
(1982), 79-82; Evelyn S. Rawski, Agricultural Change and the Peasant Economy ofS outh China 
51 
reasons can be given to explain why lineages concentrated in traditional south China. 
Maurice Freedman stressed how the southeast coast provided the crucial material 
environment for kinship principles to blossom into aggressive lineage communities. 81 
Freedman, who relied on lineage data from peasant villages in Guangdong and Fujian 
provinces and adopted a socioeconomic functionalist stance, decided that corporately 
owned land was the basic condition for the existence of Chinese lineages. He adduced the 
predominance of rice cultivation, 82 the need for agnatic cooperation in production, the 
existence of a "frontier" society, and concentrated settlement in lineage for self-defense83 
as reasons in his explanation of why agnates are predisposed to form lineages into "single 
surname villages" in coastal southeast China.84 Also, for Freedman, ancestor worship is a 
required ritual for lineages. The extension from ancestor worship by individuals and 
(Cambridge: Harvard University Press, 1972),64-94. The subsequent spread of corporate lineages, boasting 
ancestral halls and estates with households numbering in the hundreds and thousands, helped to foster an 
uneven division of power and wealth within local society. Strong lineage communities tended to become 
rich at the expense of their weaker neighbors, as did their constituent families and overbearing kinsmen. 
Dominant lineages in the same vicinity also became bitter rivals. 
81 See Maurice Freedman, Lineage Organization in Southeast Kwangtung (London: Athlone, 1958) and 
Chinese Lineage and Society: Fukien and Kwangtung (London: Athlone, 1966) on his various contributions 
to the study of Chinese lineage. 
82 Freedman, Lineage Organization, 129-30. 
83 Idem, Chinese Lineage, 159-65. 
84 Idem, Lineage Organization, 129-30. 
52 
families to corporate ancestor worship by the lineage was for him an essential part of the 
lineage-formation process. 85 
Freedman's arguments are supported by later scholars such as Hugh Baker, Jack 
Potter, and to an extent, James Watson and Rubie Watson, whose premise is that 
corporate estates, ancestral halls, and elaborate biannual rituals of kinship are constitutive 
elements of Chinese lineages.86 According to the land development project described in 
the Qing state policy in 1740, the interpretation of "frontier provinces" ~as exceptionally 
broad, including all provinces with any hilly or mountainous land. Under this policy, the 
frontier situation existed even in the hinterlands of major cities in the Pearl River Delta.87 
The land reclamation, especially in the delta areas, created tensions among 
85 Idem, ed. , Family and Kinship in Chinese Society (Stanford: Stanford University Press, 1970). David 
Faure, however, downplays the structuralist assumptions ofpatrilineal descent in Freedman's work and 
argues that lineage has meant more than kinship to its members. He contends that kinship and territory are 
related but must be treated as analytically separate, and that lineage formation involved their intertwining at 
particular historical moments of state-making. See David Faure, "The Lineage as a Cultural Invention: The 
Case of the Pearl River Delta," Modern China 15.1 (1989), 4-36; David Faure and Helen F. Siu, 
"Introduction," in Down to Earth: The Territorial Bond in South China eds. David Faure and Helen F. Siu, 
1-20 (Stanford: Stanford University Press, 1995). 
86 In Jack Potter's argument, along with that of other later scholars, south China is distinguishable by the 
presence of powerful local groups, organized in the form of lineages and linked together by common 
ancestors, written genealogies, and communal worship at ancestral shrines. Potter, in short, agrees with 
Freedman's interpretation. See Jack Potter, "Land and Lineage in Traditional China," in Family and 
Kinship, ed. Maurice Freedman, 121-38; Hugh Baker, A Chinese Lineage: Sheung Shui (London: Frank 
Class, 1968); and James Watson, "Chinese Kinship Reconsidered: Anthropological Approaches on 
Historical Research," China Quarterly 92 (1982), 589-627. 
87 See Daqing huidian shili, 164; edict quoted in Peng, kaikenshi, 125. 
53 
landlords of the delta, who fought over the construction of embankments and the drawing 
of boundary lines. 88 Sandy fields enclosed within an embankment were turned into 
private property, but nascent sandy farms that began to pile up against the walls of an 
embankment would develop into more extensive plains a few decades later. Since these 
nascent farms lay outside the embankment, they were not the property of the original 
embankment builders. Thus there was no end to this embankment-building venture. In the 
Pearl River Delta, layers and layers of embankments had been built against one another, 
making it very difficult to distinguish one embankment from another. Added to these 
boundary disputes was the perennial problem of flooding caused by imprudent 
construction of shatan and the resulting blockage of waterways. Between 1796 and 1949, 
there were 137 serious floods recorded in Shunde County alone. Many embankments 
collapsed during the floods, further worsening the boundary disputes, since it was 
impossible to demarcate property boundaries without an embankment. 89 As a result, 
intensive lineage conflicts were bound to arise on the issues of embankment-building and 
88 He fiiJ, Li *• Luo Ai, and Long n of the shatian in the Pearl River Delta are some of the lineages that 
vied with one another for the right ofreclamation. See Eng, "Landlordism"; Faure, "Cultural Invention"; 
Sasaki Masaya -fti:-fti::i1(if ijJt, "Juntoku ken kyoshin to Tokaijurokusha JIW{ffi~~Mi c: *te+1'l"t ," Kindai 
Chugoku kenkyu ili:ft 9Cl ~ljlHE 3 (1959), 161-232. 
89 Shunde XZ, 5.33; Peter Purdue, Exhausting the Earth: State and Peasant in Hunan, 1500-1850 
(Cambridge: Harvard University, Council on East Asian Studies, 1987), 219-32. 
54 
sandy farm enclosure. Such conflicts led to considerable local disputes, litigation, and 
intervillage or interlineage warfare.90 
Feuds 
As Harry J. Lamley vividly describes, Guangdong was well-situated for lineage 
feuding. 91 The people ofGuangdong, compared to people in other parts of China, were 
traditionally regarded as adventurous, progressive, and combative. Because government 
authorities were remote and inefficient, the people in the area tended to take law into their 
own hands whenever a dispute occurred, and to solve it by resorting to violence.92 
Chinese officialdom considered the region to be violent and a similar view appears in 
90 Chen, Landlord, 29; David Faure, "The Lineage as a Cultural Invention: The Case of the Pearl River 
Delta," Modern China, Vol. 15 No. 1 (Jan. 1989), 25 ; Eng, "Landlordism." 
91 The Western scholar most closely identified with the study of Chinese lineage feuding (x iedou) is Harry J. 
Lamley, who essentially equates xiedou with feud and highlights the communal aspects of the violence. See 
Harry J. Lamley, "Hsieh-tou: The Pathology of Violence in Southeastern China," Chi 'ing-shih wen-t 'i 3.7 
(I 977), 1-39; idem, "Subethnic Rivalry in the Ch ' ing Period," in The Anthropology of Taiwanese Society, 
eds. Emily M. Ahem and Hill Gates, 282-318 (Stanford: Stanford University Press, 1981); idem, "Lineage 
and Surname Feuds in Southern Fukien and Eastern Kwangtung Under the Ch ' ing," in Orthodoxy in Late 
Imperial China, ed. Kwang-Ching Liu, 255-78 (Albany: State University ofNew York Press, 1990); idem, 
"Lineage Feuding in Southern Fujian and Eastern Guangdong Under Qing Rule," in Violence in China: 
Essays in Culture and Counter culture, Jonathan N. Lipman and Stevan Harrell, 27-64 (New York: State 
University of New York Press, 1990). 
92 Even though Guangdong was not the most violent area, it was more violent than the national average. 
See Lamley, "Lineage Feuding"; Thomas Buoye, "Economic Change and Rural Violence: Homicides 
Related to Disputes over Property Rights in Guangdong During the Eighteenth Century," Peasant Studies 
17.4 (1990), 252-53; E. B. Vermeer, "Introduction: Historical Background and Major Issues," in E. B. 
Vermeer, ed., Development and Decline ofF ukien Province in the J i ha nd I ghC enturies (Leiden: E. J. 
Brill, 1990), 5-34. 
55 
local gazetteers. The local gazetteer in 1734 describes the matter as follows: 
The people have a tendency to embroil themselves in private warfare even for a 
slight grudge. They seldom bring their disputes to government attention. Instead, 
they recruit private forces that often add up to several hundred men on both sides. 
Once a private war erupts, it is prone to involve a dozen large villages; both sides 
use whatever kinds of deadly weapons are available, and these wars, in some 
cases, drag on for years .... Meanwhile, both sides visit wanton upon one 
another .... The result of such warfare is many casualties and utter desolation of 
the communities involved.93 
Simple hostility between members of two different villages could eventually 
involve the entire membership of the respective villages or lineages, and then precipitate 
a series of fierce battles. This form of communal feud became endemic in the coastal 
prefectures of the region during the eighteenth century, and gave rise to serious armed 
affrays that adversely affected local government and society. In the Yongzheng reign 
(1723-35) such lineage and surname feud strife began to be labeled as xiedou ;tJ]x;IM! 
(armed affrays) by concerned Qing officials.94 
93 Kaiping XZ, 21 : 16a. 
94 The tenn "xiedou" appears to have come into common usage only in the 1720's (as a contraction of such 
four-character phrases as chixie xiangdou fif~ffl IMJ), suggesting that xiedou were a Qing phenomenon 
even if feud like violence had a long history. Gradually, the xiedou designation was also applied to 
interlineage conflicts elsewhere as well as to communal feud strife between other types of rival groups 
throughout China and among Chinese settlement overseas. Nevertheless, this tenn was used most 
consistently in reference to the endemic lineage feuding that prevailed in Fujian and Guangdong where 
xiedou outbreaks were first reported and continued to persist throughout the nineteenth and over much of 
the twentieth century. See David Ownby, "Communal Violence in Eighteenth Century Southeast China: 
The Background to the Lin Shuangwen Uprising of 1787" (Ph.D. diss., Harvard University, 1989), 92-98 
56 
Disorder of this type was not new in the nineteenth century, nor does it appear to 
have been a direct result of the increasing strain on resources. Freedman considers that 
organized violence of this sort "cannot be dismissed as the sign of a particular crisis in 
the southeastern China." The wealth and unity of powerful lineage was tolerated in South 
China because in some ways they promoted stability in rural areas.95 But the danger 
inherent in the system, that powerful lineages would use their position to further their 
interests at the expense of their neighbors, was recognized. And so the memorials of 1766, 
referred to above, traced the cause of feuds to the unequal division of property among 
lineages: 
If the [ritual] land is used lawfully to consolidate and harmonise [kinship 
relations] ... it is not a bad practice at all. But if [it induces people] to rely on the 
numerical strength or financial power of their clans, to oppress other villages, or 
even worse, to assemble mobs and fight weapons ... [s uch a practice] surely 
should not be allowed to spread. This wicked custom is especially prevalent in 
Fukien and Kwangtung province .... 96 
Issues of local contention included land and boundary disputes and controversies 
over water rights. Attempts to monopolize local markets, transportation systems, and 
95 Maurice Freedman claims that lineages in Fujian and Guangdong graduaIJy increased their wealth and 
power until they were able not only to oppose rivals in their areas but also to fend off and make use of the 
state. See Freedman, Chinese Lineage, 115-16. 
96 Quoted by Hsiao, Rural China, 355. 
57 
mineral resources also created rivalries between lineage communities, as did competition 
over grave sites and even marriage in localities where shortages of marriageable females 
existed. Furthermore, intangible matters that were felt to be vital to communities at large 
became issues involving feud rivalry. The redemption of debts, misunderstandings 
concerning reputation and pride, and disputes hinging on ancestral and lineage honor are 
common examples of such sensitive issues.97 
During the late Qing period, as feuds expanded among lineage, increasingly large 
segments of the population were affected by communal feud violence as well. These 
villagers and townspeople also became subject to reprisals on the part of vengeful 
adversaries and targets of abuse by venal local authorities. Closed off and under great 
tension, they too grew more susceptible to the hostile, group inclinations of the feud. It 
was not always the whole lineage that became involved in fighting; sometimes it was 
merely individuals or small groups. But even when the origins of fighting were trivial, the 
97 Given these factors relating to historical causation, some scholars identified lineage feuding as an 
extreme fonn of competition for scarce resources, benefits, and things of an intrinsic value. See Lang 
Jingxiao N~~W, "Zhongguo nanfang xiedou zhi yuanyinji jizuzhi 9'1 ~1¥JJJ~IMl:z.~ lzsl.:&;J,U,fi~," 
Dongfang zazhi }l:[JJ~~ 30 no. 19 (1933), 81-96; Niida Noboru t#fEr@:, Chugoku no noson kazoku 
q:i~O):l:M*~ (Tokyo, 1952); Hsiao, Rural China. Depending on the setting, the general character of 
local feud competition varied to some degree. Thus Elizabeth J. Peny has depicted feuding in the 
impoverished Huaibei ?tfdt region of North China as a fonn of predatory violence. Viewed in the context 
of the stark ecology of this region, the communal feud seems to have been reduced to more of a struggle for 
the basic needs for human survival. See Peny, Rebels, 74-80. 
58 
resulting conflict was potentially serious.98 
For members who took part in the fighting in a full-scale lineage battle, a lineage 
provided incentives and rewards. The men would often be feasted before taking the field 
and, if successful, on return. If they fell on the battle-field, their widows and families 
would be cared for by the lineages, and their sacrifice would be immortalized by the 
inclusion of their ancestor tablets on a special 'heroes' altar' in the main ancestral hall. 
The lineage would, of course, provide the arms and ammunition and whatever other 
logistic support was required. A household levy might be made if lineage funds were not 
sufficient for the purpose.99 Wealthy lineages or villages at feud often hired outside 
mercenaries to fight their battles for them: 
If lineages engaged in xiedou are lacking in numbers, then they pay large sums of 
money to hire people to help. If a helper dies in the xiedou, his family is given a 
sum of money, and ifhe is injured, he is given money with which to convalesce. 
These fees come out of the ancestral coffers, or are assessed per amount of land 
98 During the nineteenth century, lineages at feud in south China resorted usually to more temporary 
alliances. These alliances also were generally formed through surname attachments. Because the same 
surname in China implies a common ancestor, at least in the distant past, common-surname (tongxing 
!AJtt) identities enabled lineages to associate more readily on the basis of presumed descent. However, in a 
region where most of the population bore only a few surnames in common, feuds sometimes occurred 
between lineages (or sublineages) of the same surname. Then inter-lineage feud alliances were apt to be 
based on different-surname (ybcing ~tt) identities or more solely on territorial considerations. See Lamley, 
"Surname Feuds," 262. 
99 See "Clanship among the Chinese," The Chinese Repository, Vol. IV no. 9 (1836), 413 ; Public Record 
Office (London) F.O. 228/470, no. 46, Robertson to Alcock (21, May 1869); Lamley, "Lineage Feuding," 
45. 
59 
owned. The idle and the unemployed are happy to receive such employment. 
Even if they die they do not regret it. 100 
Later in the nineteenth century, with the accelerated import of more Western 
arms, so called gunmen (chongshou i!t-¥ or qiangshou ~-¥), trained in the use of the 
new guns, were contracted as mercenaries by feuding lineages. 101 Brotherhood 
associations and secret societies also participated in xiedou: 
The people of Guangdong are willful and perverse. In every trivial incident, 
every affair involving fields and mountains, they do not await official 
adjudication, but rather hire outside bandits, choose a time and engage in xiedou. 
Those who lead the fighting are, in the main, people despised by the lineage 
heads, lineage gentry, and ritual leaders. They say it is a matter of two lineages, 
but in fact it is a bunch of temporarily hired bandits and salt smugglers, plus a 
kind of professional mercenary dare-to-die. Those who are called together 
number in the hundreds if not thousands, involved in several if not several tens of 
villages. 102 
These hired mercenaries or members of secret societies added to the terror 
100 Cited in Niida, Chugoku, 377-78, n.29. 
101 Lamley, "Lineage Feuding," 49. 
102 Cited in Lang J ingxiao, "Qingdai Yuedong xiedou shishi l1f ft~ }l! ~IMJ 5e. ff," Lingnan xuebao 
!lifi~~ 4.2 (1935), 144. Qing officials frequently linked xiedou and brotherhood associations and depict 
leadership of violence as having fallen into the hands of young toughs who used the process to enrich or 
empower themselves. Zhuang Jifa has related the emergence of surname alignments and secret society 
bands in Guangdong to local feud traditions. See Zhuang Jifa #f-=g~, Qingdai Tiandihui yuan/iugao 
l1fft.:R:tt!nff L\1-vft~ (Taibei: Shilin, 1982), 11-18. 
60 
spawned by feud strife. When not employed in xiedou violence, they were contracted for 
other vengeful acts of violence, such as assassinations, or they robbed on their own. More 
harmful to localities were the pillaging, acts of rape, and wanton destruction of fields, 
corps, and villages. 103 As a result, villages became more heavily fortified, with stout walls, 
cannon emplacements, and watchtowers. 104 
Migrations and Ethnic Tensions 
In some cases feuds were fought between villages belonging to different ethnic 
groups. Unlike the lineages, ethnic differences were less significant in the formation of 
systems of social organization. In South China, ethnic differences were perceived as a 
matter of cultural rather than racial attributes, for, at least by the eighteenth century, 
inhabitants of south China were predominantly Han. Native tribes accounted for only two 
percent of the whole population in south China. After the massive migration of the Han 
into South China, most of the aboriginal groups (such as the Li and the Yao) either went 
103 Zhang Zhidong *:z.WJ, Zhang Wenxiang Gong Quanji ~>Cl!!-~~(Taibei: Wenhai, 1963; reprint 
1937)juan 14, zouyi 14 (1886), 9a-b. 
104 In Guangdong there were plenty of indications to show how common and widespread the conflicts were, 
either in the fortifications that surrounded villages and market places, or in the damage and neglect 
resulting from the fighting. See Niida, Chugoku, 386. 
61 
across the border into mainland Southeast Asia, lived in remote areas in southwestern 
China, or were assimilated by the Han. Nonetheless, South China was culturally more 
heterogeneous than North China. While the North China plain served as a melting pot, 
South China was the recipient of successive waves of migration from the north. It is also 
hillier and conducive to the isolation of one social group from another. Hence, as late as 
the nineteenth century, the Han people in South China still divided themselves into 
various subgroups based on such differences as date of settlement and provincial origin of 
their ancestors or cultural differences such as customs, habits, and dialects. In Guangdong 
province, the Chinese inhabitants categorized themselves into four large groups: Punti, 
including the Cantonese and the people of Chaozhou; Hakka; Hok.lo; and Tanka. 105 
Subethnicity was important in South China before the twentieth century, since it 
was responsible for frequent subethnic feuds and one major local war during the mid-
nineteenth century between the Hakka and the Punti people. This war affected many 
105 The Puntis were the earliest Han people to settle in South China. They migrated to the area during the 
tenth century along the North River route. A large majority lived in the Pearl River Delta, along the West 
River, and in western Guangdong. The Hoklo came to South China during the thirteenth century from 
Fujian province. They settled in the Han River drainage, particularly in the area ofChaozhou. At a later 
point in history, some Hoklo migrated to parts ofHainan Island. The Hakka came from North China during 
the seventeenth century. They settled in the watersheds of the Han, the East, and the North rivers. The 
Tanka were exiles during the Yuan Dynasty (1289-1368), when they were so discriminated against by local 
people that they had to live on boats. They were also excluded from taking the civil service examination. 
See Chen, Guangdongdizhi, 68-69, 141,205,229. 
62 
counties in central Guangdong, claiming more than one hundred thousand lives. The 
origin of the war dates back to the seventeenth century. Under the Qing, vast inter-
regional migrations of population took place, in particular Sichuan, Jiangxi, Guangxi, 
Manchuria, and to Taiwan. In some of these areas, for example Sichuan, the population 
had been decimated by wars at the end of the Ming period, and there was consequently 
land available for cultivation. In other areas, there were extensive upland regions, which 
hitherto had been little used agriculturally. With the introduction of new crops, in 
particular maize and sweet potatoes, these areas could now be developed.106 Though 
much migration took place voluntarily, other migrations took place with the active 
encouragement of the Qing government, which saw the opening up of new land a way of 
increasing revenue and of extending its control over under-populated areas. 
What is significant here is that the Hakkas played a major role in several of these 
migrations. When moving to Sichuan, Jiangxi and Taiwan, the Hakkas were joining main 
streams of migration that involved other groups in South China. In their movement 
westwards across Guangdong, to Hainan and to Guangxi however, they played the 
leading part as colonists of hilly areas. In Hainan and Guangxi there were still, at the 
106 Ho, Population, 136-68; Marks, Tigers, 291. 
63 
beginning of the Qing period, large areas available for settlement, and these naturally 
attracted Hakka migrants. But, in contrast, those who settled in Guangdong were moving 
into areas where there was already a substantial and settled population. 107 
Some of the earliest Hakka settlers in central Guangdong went to those districts 
that had been affected by the policy of coastal evacuation. After the restrictions had been 
lifted, these areas were first re-occupied by the former inhabitants and their descendants. 
After twenty years of disruption, however, their numbers had been much reduced. By the 
time the ban was lifted, nearly 80% of the population had succumbed in eastern 
Guangdong areas alone and about 9 million mu had been abandoned in the coastal 
provinces of Guangdong, Fujian, and Zhejiang. 108 The land left vacant attracted early 
Hakka settlers. The direction of the movement was principally to the center of the 
province, which is, to the areas already inhabited by Cantonese. 
In a number of cases the early Hakka settlements were officially encouraged 
under the system of military colonies (tuntian ~ 83 ). The system was one of great 
antiquity, dating as far back as the Han dynasty. Its purpose was to secure regions that 
107 For the comprehensive descriptions of the Hakka migrations, see Leong, Migration, Chapter 2 and Luo, 
Kejia, Chapter 2 and 3. 
108 Chixi XZ, 8:2a; Peng, tudi kaiken shi, 61 . 
64 
had recently been conquered or resettled. Military colonists were farmers who obtained 
their land rent-free in return for certain military and guard services. In addition, their land 
could not be alienated. Under the Manchus the system was also used in the South, and 
military colonies were established in Xin'an ffi~, one of the districts affected by the 
removal of the coastal population. 109 The farmers in this case were the Hakk:as, who were 
brought into the area from Jiayingzhou, Huizhou and also from Fujian and Jiangxi. The 
date of the first establishment of these colonies is not known, but by the end of the 
seventeenth century they could be found in the core counties of Dongguan, Huaxian, and 
Heshan. By 1716 the number of colonists was sufficient to warrant their being given a 
quota in the military examination. 110 
Another example of official encouragement of resettlement, and one which is of 
particular interest in view of its later bearing on the Hakka-Punti War, was that to the 
southwest of Guangzhou during the Yongzheng ~IE period (1723-35). The policy of 
109 During the Southern Han dynasty, military colonies of this sort had been established in the present Hong 
Kong New Territories, and the Puntis there had been associated with them. This method of colonization 
thereafter fell into disuse, but was revived by the Ming, who used it to secure strategic regions along the 
Great Wall and by Qing for land reclamation. It survived until their final abolition in the twentieth century. 
See, Ho, Population, 137; T. R. Tregear, Hong Kong Gazetteer to the land Utilization Map of Hong Kong 
and the New Territories, with China and English Names (Hong Kong: Hong Kong University Press, 1958). 
110 Leong, Migration, 60; Luo Xianglin, Hong Kong and its External Communications before 1842 (Hong 
Kong: Institute of Chinese Culture, 1963), 134, 147. 
65 
encouraging the settlement of wastelands had been reiterated at the beginning of the reign 
in an edict calling upon officials to report the existence of wasteland within their 
jurisdiction. 111 In the following years a number of edicts offered loans and remission of 
taxes for persons opening up new land.112 In mid-1732, the Governor of Guangdong, a 
Manchu named E-mi-da ~ 51:ii, suggesting resettling the poor--especially those in the 
"overpopulated" eastern Guangdong's Huizhou, Jiaying and Chaozhou prefectures--in the 
districts southwest of Guangzhou on the basis of information he had derived from local 
officials. 11 3 
E-mi-da had called upon influential businessmen to collect poor Hakka people 
from eastern Guangdong, and to provide them with assistance in the way of lodgings, 
rations and working capital. 11 4 From this it appears that the migrants were to be treated as 
laborers or tenant farmers, and that the cost of their settlement would be recovered from 
11 1 In his edict, the Yongzheng emperor reasoned the land reclamation policy: "Population has increased of 
late, so how can [the people] obtain their livelihood? Land reclamation [kaiken Im!!] is the only solution." 
See Da Qing Shizong Xian (Yongzheng) Huangdi Shilu ::kl'l!ft!t*~(~itIE).§!7!fjf~ (Taibei: Hualian 
Chubanshe, 1965 reprint), edict ofYZ 1.4, 6:25. 
11 2 Yongzheng zhupi yuzhi ~ iE f(tf!t!! ~ (Taibei: Wenhai chubanshe, 1965 reprint), Memorial dated 
YZl0.6.9, 56a-58b. 
11 3 A report from the Grain Intendant Tao Zhengzhong ~ .iE q:i about the wasteland in the southwestern side 
of Guangzhou initiate in resettlement of Hakka people into Xinning (Taishan) and other counties. See 
Xinning XZ, 2:38b; Leong, Migration, 60. 
114 Another example of official encouragement given to the Hakkas to migrate to the west was to the district 
of Guixian ~ ~ in Guangxi during the Daoguang period. 
66 
their services. In this way a sufficient supply of settler families had been obtained. Each 
group of five families had been allotted an area of 100 mu to cultivate, and for security 
reasons each of these groups had been organized into ajia, with a person responsible for 
their behavior. The initial plan had been to settle 1600 families, and by the time of E-mi-
da's report some 300 families had already been established in Heshan and their number 
was said to be daily increasing. By early 1735 over 200,000 mu had been resettled by 
7,760 tenant farm families in Heshan, Enping, and Kaiping counties. 115 
Officially sponsored migrations, while interesting as illustrative of the official 
attitude, were only a small part of the general pattern of migration. By far the greater 
proportion of Hakka migration westward was without government assistance or official 
recognition. By the end of eighteenth century, the Hakka settlements in the districts to the 
southwest of Guangzhou had attained sizeable dimensions. 116 But most of the central 
Guangdong areas were already densely inhabited by the Puntis, and in particular the areas 
suitable for wet rice had already been brought under cultivation by the time of arrival of 
115 Guangzhou FZ, 20:2 la. All of these areas became the major battle grounds of the Hakka-Punti War in 
the mid-nineteenth century; I will be talking about this more in detail in the ensuing chapter. 
116 Up to the end of the Qianlong period, the Hakkas had established settlements in all the districts of the 
Guangzhou and Chaoqing and had extended to Guangxi and Hainan. See Chixi XZ (1920), 8:3b; Herold J. 
Wiens, China's March Toward the Tropics: a discussion oft he southward penetration of China's culture, 
peoples, and political control in relation to the non-Han-Chinese peoples ofs outh China and in the 
perspective ofh istorical and cultural geography (Hamden: Shoe String Press, 1954), 272. 
67 
Hakkas. So, in central Guangdong, the districts in which the Hakkas settled in the 
greatest numbers were once again those that contained hilly land. To the southwest of 
Guangzhou, in the districts involved in the Hakka-Punti War, the topography was on the 
whole more favorable to agriculture than in Jiayingzhou. Over twenty percent of the land 
there was cultivable, except in the district of Enping, where the proportion fell to 
approximately thirteen percent. 117 But, as elsewhere, the best areas were already under 
cultivation. In Xinning the Hakka found that the only vacant land was in the hills, in 
regions that were desolate and barren. 118 Occasionally they settled in areas still occupied 
by non-Chinese groups. 119 In these hilly regions they cultivated cash crops like tea. But 
hilly areas of Guangdong had suffered badly from erosion by the nineteenth century and 
the Hakkas and other settlers eventually had to migrate down toward the fertile fields of 
the lowlands. 120 
When the growth of population and shortage of land had become apparent, 
migrant farmers such as the Hakkas appeared to the local inhabitants as dangerous 
117 J.A.G. Roberts, "The Hakka-Punti War," 45. 
118 Chixi XZ, 8:4a. 
119 Enping.XZ, 13:23a. 
120 By the nineteenth century, erosion has caused the devastation of large areas of the province. See Articles 
in China Mail (11 Nov., 1858 and 14 Aug., 1856). 
68 
intruders. During the second quarter of the nineteenth century, the continuing expansion 
of the Hakka population in the region stretching from the northeast to the southeast of 
Guangzhou, especially in the vicinity of Sihui, Kaiping, Enping, and Taishan, led to 
increasingly frequent conflicts with the Cantonese. One example of this was the feud 
fought between Hakka and Punti villagers in the district of Guishan ~ LU to the east of 
Guangzhou. In 1843 a quarrel occurred over the payment of rent for a market that the 
Hakka had leased, and there was a clash between the two groups. Seven years later the 
feud was revived, and the Hakkas, finding support from ninety villages in the 
neighborhood, succeeded in capturing the principal fortress of the Punti village. The 
fighting did not end at this point. After a further six years there was another outbreak, 
which left the area devastated by the effects of the feud.121 In Xin'an in the same period 
feuding between Hakka and Punti villages was common. One German missionary 
reported that the Puntis and the Hakkas in these districts were in constant warfare: 
On some occasions, when there is a grand fight, the forces on either side have 
been known to number nearly a thousand persons, armed with old gingalls, 
spears, or stones. The magistrates are totally unable to suppress these feuds. 
Fortunately, the fighting is of that description which does not cause much loss of 
life. These engagements frequently last three or four days. 122 
121 The Chinese and Japanese Repository 3 (1865), 283-4, quoted by Hsiao, Rural China, 366-7. 
122 Lindesay Brine, The Taiping Rebellion in China: a narrative of its rise and progress, based upon 
69 
The district magistrate was neither willing nor able to intervene, and so feuds 
dragged on in a sporadic fashion. In the southwest of Guangzhou, it appears that, until 
1854, the animosity between the Hakka and the Punti was only one aspect of inter-village 
feuding; after that time, however, the animosity between the two groups became the 
principal motive for fighting between villages, and that this rapidly led to an expansion of 
the feuds into more serious clashes. The most violent battles occurred after 1856 and 
lasted until 1867, when the governor of the province decreed the establishment of the 
sub-prefecture of Chixi as a sort of reservation where Hakka displaced by the fighting 
could reestablish themselves. 
Government Representation of Feuds 
Such feuds were so commonplace in Guangdong that in 1823 the Ministry of 
Justice had to make an exception in the law for this province, ordering that in the case of 
conflicts involving the loss of lives, a distinction should be made between crimes 
committed by individuals and large-scale fighting between lineages. 123 And yet, in 
original documents and information obtained in China I by Commander Lindesay Brine ; with map and 
plans (London: John Murray, 1862), I 07-8. 
123 F.O. 931.1193 , Xu and Yeh to Emperor (31, Jan. 1850). 
70 
contrast to the frequent mention of the headaches caused by xiedou, records of official 
prosecution of actual xiedou incidents are rare. The local authorities, in this situation, 
could have sent regular troops to stop the feud. They generally refrained from doing so, 
however, for the following reasons: frrst, they had to bear the burden of financing the 
troops; second, they would have exposed themselves to the danger of being cashiered by 
their superiors for being incompetent to deal with the controversy before it was too late; 
third, the feud between the large lineages involved thousands of members, too many for 
the local authorities to control. 124 TheXunzhoufuzhi mfl·IJt-f]t says: 
None of the local officials wanted to be exposed to the charge of mal-
administration by memorializing the emperor that the region under their 
jurisdiction was infested with bandits or disturbers of the peace. Naturally the 
best thing for them to do was to pretend to be ignorant of the feuds; they hoped, 
by so doing, that they could escape the punishment for maladministration by 
transferring to another post in some other district or even province after the end 
of their term of office. 125 
This partly explains why lineage feuds were common and why the lineages could 
have their own way in precipitating private war. Nor was this all. Powerful lineages made 
124 Acquisitive-minded civil and military officials, along with their avaricious underlings and soldiers, were 
often attracted to lineage feuds. For feud affrays offered the underpaid personnel of the Qing government a 
chance to reap lucrative bribes and fees. See Bian Baodi -tJUi, Bian Zhijun (Songchen) Zhengshu i-1tU1fI 
(~i::2)i&if (Taibei: Wenhai, 1968 reprint), 240,277; Lamley, Surname Feuds, 268. 
125 Xunzhou Fuzhi ~fM!f.f ;t;, 56: l la- l 8b, quoted in Long, Jingdetang, 4: 12b. 
71 
use of their influence to encourage their members to commit various crimes, ranging 
from burglary to plundering. Lin Zexu 1-tJ'l iJ ~, the Imperial Commissioner for the 
suppression of opium (1839-40), was not a man to shy away from difficult administrative 
problems, but even he found the lineages daunting: 
In Huizhou and Chaozhou there are clans who made their fortunes through 
banditry; in such cases, the fact that their accomplices are numerous prevents the 
government from laying hands on them, since no one dares inform against them 
and no official dares arrest them. Sometimes entire clans or entire villages are 
bandits. Any attempt to arrest them may cause disturbances . . .. This is the reason 
that it is difficult to check the rampancy of banditry. 126 
The state was clearly not making a serious attempt to control the area in question. 
The strength of lineage organizations was no doubt related to the weakness of central 
political and military control, especially during the late imperial period. Under these 
conditions, officials expected to find groups of armed men in the backwaters of South 
China where state presence was at a minimum, and they recognized that a certain amount 
of violence was inevitable. No special effort was made to enact laws to contain xiedou 
simply because xiedou conformed to the social language of the south China--territory and 
lineage. When Chinese officials looked at xiedou, they saw lineage and village leaders, 
126 Lin Zexu ~!'!U~ , "Liangguang zougao firi.ll~~." in Lin Zexu ~!'!U~, Lin wenzhonggong zhengshu 
~X,'iE',~iSc~ (Taibei: Taiwan Shangwu Yinshuguan, 1968 reprint), 3: 17b-19a. 
72 
not marginal young toughs or criminals. Lineage and village leaders were part of the 
Confucian hierarchy, which was governed by "the rule of men" rather than by "the rule of 
law." These considerations may explain the reticence of local officials to intervene in the 
ongoing Hakka-Punti War of mid-nineteenth century. Subethnic xiedou remained more 
tolerable than the uprisings led by secret societies. 
3) Economic and Political Environment 
Economy 
By the early Qing, Guangdong merchants had traded far and wide, both in and 
out of the province and abroad. In internal trade the West, East and North Rivers served 
as the links between Guangdong and the neighboring provinces. Foreign trade boomed 
during the Ming and Qing dynasties, as Guangzhou became the leading port of China. In 
the early Qing, Guangdong's booming handicrafts and commercialization of agricultural 
products led to the rise of its commerce. 127 
Guangdong's sugar refining industry can be traced back to the Eastern Han 
dynasty and fine sugar had been manufactured around the sixth century. In the Ming and 
127 Wong, Yeh Ming-Ch 'en, 29. 
73 
Qing, the Guangdong sugar industry experienced remarkable growth, with the acreage of 
sugar cane plantation almost equal to that of rice paddies in the Pearl River Delta. 128 
Cotton was brought to the region from India no later than the Song and cotton textiles 
experienced rapid development in the Qing dynasty, a development that paralleled the 
already flourishing silk industry in the province. Foshan, Shantou nlJi'rn (Swatow) and 
Qiongzhou 311-Ms prang up as centers of the textile production. 129 Foshan was also 
important for the production of ironware, fireworks and parasols. Other industries like 
porcelain flourished at Chaozhou and Dongguan, and embroidery at Guangzhou, Huizhou 
and Shaozhou tfl 1-M. 130 
With merchants from Guangzhou and Chaozhou among the most prominent, 
Guangdong merchants expanded their activities rapidly in the province and subsequently 
into the rest of the country. The Chaozhou merchants even traded as far as in Southeast 
Asia. Merchants from other provinces also converged in Guangzhou in the early Qing, 
128 Such eta Mazumdar, "A History of the Sugar Industry in China: The Political Economy of a Cash Crops 
in Guangdong, 1644-1834" (Ph.D. diss. , UCLA, 1984), 290-292. 
129 Ibid. 
130 David Faure, "What Made Foshan a Town? The Evolution of Rural-Urban Identities in Ming-Qing 
China," Late Imperial China, Vol. 11, No.2 (Dec., 1990), 1-31 . 
74 
giving rise to its bustling trade activities. 13 1 Guangdong provided the largest share of 
customs revenues in the coastal trade from 1735 to 1812, in large part due to the 
transshipment of foreign goods.132 From the first quarter of the eighteenth century to 
1800, foreign trade in the provincial capital of Guangzhou, as measured in thousands of 
tons, grew eightfold. 133 
The specialization and commercialization of agriculture in Guangdong was 
further fueled in 1759 when the provincial capital, Guangzhou, became the only legal 
port of entry for foreign trade. 134 Great amounts of silver imported to Guangzhou from 
overseas were mostly used to purchase goods produced in the province. Basically, the 
city's development was in line with its trade activities. Movement of goods in and out of 
13 1 Li Hua *~, "Qingchao qianqi Guangdong de shangye yu shangren trUJfiw:lmJJJIHl"JFJ3~~FJ3 A," 
Xueshu Yanjiu ~~~~' No, 2 (1982), 39-44. 
132 Fan 1-chun, "Long Distance Trade," 241. 
133 Susan Naquin and Evelyn Rawski, Chinese Society in the Eighteenth Century (New Haven: Yale 
University Press, 1987), 103. 
134 Qing court made the trade on the Chinese side a monopoly under the control of a superintendent of 
maritime customs (the "hoppo" 1!};1iUUii'li) and a group of thirteen hong merchants (the Cohong ~fi), 
who served as his agents. Guangzhou prospered even more under this monopoly, exporting the tea and silk 
of central and eastern China, and importing silver, copper, arms and other foreign goods. On the Canton 
system See John K. Fairbank, Trade and Diplomacy on the China Coast (Cambridge: Harvard University 
Press, 1953), chaps. 1-4; Ch ' iang T'ing-fu, "The Government and Co-hong of Canton, 1839," Chinese 
Social and Political Science Review, 15.4 (1932), 602-7; Chang Hsin-pao, Commissioner Lin and the 
Opium War (Cambridge: Harvard University Press, 1964), 14. 
75 
the province and abroad led to the rise of Guangzhou as metropolitan port city. 135 It is 
important to note that throughout most of the eighteenth century China enjoyed a 
favorable balance of trade with the Europeans. Chinese merchants and possibly silk 
producers benefited from foreign trade during this period. The deleterious effects of 
foreign competition on some handicraft industries and the rise in opium imports were not 
acutely felt until the nineteenth century.136 
The impact of foreign demand for Chinese products, especially for Chinese silk, 
greatly contributed to the transformation of agriculture in the Pearl River Delta, where 
farmers switched from rice cultivation and the fish pond and fruit tree system of farming 
to the mulberry embankment and fish pond system in order to increase the supply of 
mulberry leaves, silkworms, cocoons, and silk. The dike-pond system of the Pearl River 
Delta is composed of three essential components: fish ponds, mulberry dikes and sugar 
cane dikes. When the Qing government closed all ports except Guangzhou to foreign 
135 The price of silk, an important export commodity, rose steadily throughout the eighteenth century, 
thereby increasing the amount of silver circulating in the Guangdong economy. See Yeh-chien Wang, 364-5, 
"The Secular Trend of Prices during the Ch ' ing Period (1644-191 I)," Journal oft he Institute of Chinese 
Studies oft he Chinese University ofH ong Kong 5.2 (Dec., 1972), 347-68. 
136 While the impact of imported textiles clearly hurt some native producers, in some areas the availability 
of cheap yarn boosted the development oflocal weaving. See Albert Feuerwerker, Studies in the Economic 
History ofl ate Imperial China: Handicraft, Modern Industry, and the State (Ann Arbor: University of 
Michigan, Center for Chinese Studies, 1995), 123-63. 
76 
trade, and at the same time limited the export ofTaihu (::t1itl1) silk, the demand for Yue 
(Guangdong) silk increased abruptly and prices soared. Thus the conversion of rice fields 
to the dike-pond system accelerated. This system was best developed in the central part of 
the delta, where it centered principally on Shunde county, together with parts of 
neighboring Nanhai, Zhongshan, Xinhui, and Heshan counties. 137 
With an extensive water transport system, peasants in Guangdong could devote 
more time to the cultivation of cash crops. The importation of rice from abroad--mainly 
from Thailand--and from neighboring provinces began as early as 1723. By the mid-
nineteenth century the Pearl River Delta had become a rice-deficit region. 138 A 
nineteenth-century official reported, "Guangdong is the richest province in China and is 
universally acclaimed for its wealth .... Sugar and silk are transported to the north and 
west. It cannot be said that the profit attained is not vast. But the people are poor and 
dependent because the production of rice is insufficient." 139 Whereas such cash crops 
137 See Kenneth Ruddle and Gongfu Zhong, Integrated Agriculture-Aquaculture in South China: The Dike-
Pond System oft he Zhujiang Delta (Cambridge: Cambridge University Press, 1988), 2-17; Marks, Tigers , 
129. 
138 An ever-increasing population and a shortage of arable land led to a rice shortage in Guangdong. See 
Hsiao, Rural China, 381; Chen, Landlord, viii, 1-2; Tang Li, "Guangdong Jingji, 1-21. 
139 Li Wenzhi $)CW;, et al. , comps., Zhongguojindai nongyeshi zi/iao !fl~ili:~.f::~3':_jf~ (Beijing: 
Sanlian Shudian, 1957), 1:472 cited in Marks, Rural Revolution, 47. 
77 
undoubtedly added to the wealth of the rich landowners and merchants, the plight of the 
common folk was aggravated by the rapidly rising price of rice. By the middle of 
eighteenth century, the demand for rice in this region led to the development of a well-
integrated market for rice centered on the provincial capital, Guangzhou. 140 As a result, 
peasant households became dependent on the market to purchase rice and the impact of 
commercialization was felt early and profoundly in this region. 141 
The Opium War and Social Disorder in the Pearl River Delta 
Foreign pressures on China started just about the time that the dynasty's internal 
difficulties began. The background to the onslaught was the growing disrespect for China 
as the eighteenth century closed and the nineteenth century dawned and, more generally, 
the increasing economic exploitation by the Europeans. The first results of this situation 
were British attempts to resolve certain problems of trade and relations with China in 
ways that violated Chinese laws. These attempts led to confrontation and, eventually, to 
140 Robert Marks, "Rice Prices, Food Supply, and Market Structure in Eighteenth-Century China," Late 
Imperial China 12.2 (Dec., 1991), 64-116. 
141 Violent disputes over property rights were a recurrent problem in the Pearl River Delta during the Qing 
dynasty. See Thomas M. Buoye, Manslaughter, Markets, and Moral Economy: Violent Disputes over 
Property Rights in Eighteenth-century China (Cambridge: Cambridge University Press, 2000), Chapter 5. 
78 
combat: the Opium War of 1839 to 1842 and another conflict, sometimes known as the 
Second Opium War, from 1856 to 1860. The overwhelming foreign victories in both wars 
and the treaties that came into existence as a result established the basis for Western 
power in China for the next one hundred years. 142 
Military defeats in the Opium War not only contributed directly to the weakening 
' 
of the Chinese state, but also increased the financial hardship of the rural people. First, 
the war made it possible for opium smuggling to be carried on more brazenly, without 
further fears of punitive action. Consequently, the import of opium increased to 70,000 
chests a year in the 1850s, and after legalization in 1858, imports of drug touched the 
high point of 81,000 chests in 1884. 143 The resulting efflux of silver steadily raised its 
value in terms of copper cash and exacerbated the problems of tax-paying peasants. 
During the first years of the reign of the Daoguang emperor, which started in 1821 , silver 
was 1,200 cash a tael; it was 1,600 cash in 1838 and 2,300 cash in 1848. 144 As Kung-
142 The following works have been especially useful in discussing relations with the European powers in the 
early nineteenth century: Chang Hsin-pao, Commissioner Lin; John K. Fairbank, Trade and Diplomacy on 
the China Coast (Cambridge: Harvard University Press, 1953); Frederic Wakeman, Jr. , "The Canton Trade 
and the Opium War," in The Cambridge History ofC hina, Vol.. 10, eds. Denis Twitchett and John K. 
Fairbank, 1978-87 (Cambridge: Cambridge Unjversity Press, 1980); idem, Strangers at the Gate; James M. 
Polachek, The Inner Opium War (Cambridge: Harvard University Press, 1992); and Peter Ward Fay, The 
Opium War, 1840-1842 (Chapel Hill: University of North Carolina Press, 1975). 
143 Spence, "Opium Smoking," 151. 
144 Luo Ergang itffl~, Taiping Tianguo Shigang :k-Zf'-:R~~ffl (Shanghai, 1937), 16. 
79 
chuan Hsiao has remarked: "By the middle of the nineteenth century such deterioration 
[of the rural environment] had reached a critical point; large numbers of rural inhabitants 
in various parts of the empire were driven by destitution and hardships to the point of 
despair." 145 
Second, the war increased the fiscal problems of the state. This and subsequent 
conflicts ultimately led to bankruptcy. China suffered a loss of over $8,223 ,700 in its 
trade with England alone. 146 In addition, at the end of the Opium War Beijing had to pay 
23 million silver dollars (Mexican) as indemnity; at the end of the 1860 war, Beijing was 
forced to pay 16 million silver taels. 147 After that China was continually in debt until the 
establishment of the People's Republic in 1949. Beijing imposed several additional levies, 
hoping to relieve the exchequer, but these further impoverished the already suffering 
peasantry. 148 
Last, the Opium War had not only exacerbated poverty in the rural society, but 
145 Hsiao, Rural China, 41 l. 
146 Chinese Repository, XII (1843), 514. 
147 1 tael was 70 U.S. cents in local value. 
148 The import of foreign cloth and yam also caused a certain amount of decline in the native handicraft 
textile industry driving out Chinese native products and cutting thus the subsidiary income of the rural 
people. 
80 
had also revealed the military weakness of the Manchu dynasty, contributing to social 
disorder. One major condition has generally been given as both cause and result of the 
military weakness: government corruption and inefficiency. Local officials did not 
hesitate to keep the ruler in the dark by tolerating and falsely reporting the troubles. In the 
Guangzhou fuzhi we find this statement: 
Where does the source of the present-day turbulence lie? It lies in those 
magistrates who are greedy and cruel . . .. Lower officials depend upon higher 
ones for protection, while higher officials rely on lower to serve as their outside 
agents in their common malpractices. The people have hated them all in the 
depth of their hearts for a long time. As soon as the crisis comes and the alarm is 
sounded, wicked elements take advantage of the situation . .. . 149 
The Manchu banners had already lost their military effectiveness during the 
eighteenth century. Their pay remained the same and could not have easily been 
increased under the fixed government budget system. Under steady, slow inflation, prices 
had risen, so that the real pay of the troops had declined. The number of paid soldiers in 
the banners also remained the same, while the banner population increased. In addition, 
the life of the idle banner elite had corrupted morale and reduced their value as a fighting 
force. The Chinese professional troops, the Green Standard, had also been affected by the 
149 Guangzhou FZ, 127:24b. 
81 
general corruption and demoralization. Their salaries, which were low to begin with, 
were often pocketed by officers, and the underpaid soldiers lived off the land. Training 
was negligible and equipment was lacking; the army had become a typical part of the 
corrupt bureaucratic organization, totally unprepared to deal with any major 
emergency.1so To supplement inadequate government troops, local officials hired 
mercenaries; and to protect their towns and villages, the gentry as well as some 
commoners formed militia, composed of either mercenaries or peasant minutemen. Thus, 
the Opium War in the Pearl River Delta, as Frederic Wakeman and Philip Kuhn have 
demonstrated, led to a militarization of that region's hinterland. 1s1 
The Opium War had had other effects on the internal Chinese development, 
especially in the Pearl River Delta. Guangzhou was the principal city of Guangdong 
province and the political and military headquarters of the area. Guangzhou had seen 
most of the fighting during the Opium War and the changes brought about by the Treaty 
of Nanjing had destroyed the trade monopoly formerly held by the merchants of the city. 
The opening of Shanghai as a new center for trade brought the foreign merchants closer 
150 See Franz Michael, "Regionalism in Nineteenth-Century China," introduction to Stanley Spector, Li 
Hung-chang and the Huai Army (Seattle: University of Washington Press, 1964) . 
151 
. Wakeman, Str,angers, 12-70; Kuhn, Rebellion, 69-76. 
82 
to the tea and silk production areas, 152 which meant that the thousands of coolies and 
boatmen, who earlier had been employed in transporting these commodities to 
Guangzhou, were thrown out of work. 153 Coastal pirates, cleared from the high seas by 
the British navy, invaded the river systems of Guangdong and Guangxi to harass and 
pillage. The ranks of the lawless were further swelled by the soldiers discharged from 
local militias after the conclusion of the war. 154 
In addition to the above problems, Guangdong province, and still more, Guangxi, 
was settled by diverse ethnic populations as we have seen earlier. The Puntis and Hakkas, 
major Chinese groups in the region, had maintained their special traditions and dialects 
and were often involved in communal rivalries and clashes. A large number of tribal 
people also lived in the two provinces, especially in Guangxi. People like the Miao ES 
and Yao :Et, who had made up the pre-Han population of the area, had been pressed back 
by Han settlement to less desirable land and mountainous regions. These tribal people 
152 Since the foreigners had already won the Opium War, they did not have to stay in Guangzhou and they 
started their business enterprise anew in Shanghai in which the foreigners found scope and opportunity for 
the full exercise of their new facilities and privileges and trading methods with more freedom than under 
the domination of the old Cohong system in Guangzhou. See Banister T. Roger, A History oft he External 
Trade of China, 1834-81 (Shanghai: Inspector General of Chinese Customs, 1931), 25; Consular Report 
Vol. 23 (Washington D.C.: U.S. Government Printing Office, 1889), 382. 
153 Wakeman, Strangers, 100. 
154 Chinese Repository, V. 3 (1834-5), 62; Laai Yi-faai, "Pirates ofKwangtung," 19-25. 
83 
had a latent hostility against the Chinese officials and farmers who had taken land and 
imposed taxes. One re~ult of this was the Yao rising in the 1830s, the last determined 
attempt at resistance by a non-Chinese group on the mainland of Guangdong: 155 
The dislocation of the regional economy in the mid-nineteenth century, along 
with the diversity of ethnic composition and of occupational groupings, consequently 
made the province the favorite ground for the establishment of secret societies and bandit 
gangs, whose membership was drawn from the disaffected populations. Jobless porters, 
impoverished boatmen, and discharged soldiers and militiamen from the Opium War 
were ready recruits to underground secret societies. Discontented peasant tenants also 
joined the secret societies. The concentration of landownership in a few hands (the 
lineages and powerful gentry), meant that the majority of the farmers were tenants. By 
the early nineteenth century, the pressure of population and the shortage of arable land 
had considerably raised the price of land as well as rents, and the tenants were badly 
affected. 156 It is not surprising then, that tenants should "harbor deep resentment against 
155 For this rebellion, see Wei Yuan~~' Shengwuji ~ftUc (Shanghai: Shanghai Guji Chubanshe, 1995 
reprint), 7:35a-39a, 4 la-52a. 
156 As the population doubled during the eighteenth century and continued its growth in the following 
century, the limits of cultivable land in Guangdong reached by 1850. In spite of the fact that the amount 
increased from about 7 percent around 1713 to 13 percent in 1853 in terms of the percentage of the total 
Lingnan land area under cultivation, the amount of cultivated land in Guangdong did not increase after 
1853, indicating that the limits of cultivable land had been reached by then. See Hsiao, Rural China, 3 83-
84 
the clan leaders who enforced high rents or usury."157 More and more began to join secret 
societies, the champions of the oppressed, with their underground ideology transcending 
lineage authority. 158 The formation of the secret societies was rampant around the Pearl 
River Delta after the Opium War: 
In August 1843, a thousand men or more, Triads and members of the Ngo Lung 
Hwai (Sleeping Dragon Society) fought together with arms in the village of 
Yung-ki, in the district of Shun-te (Shunde) .... In 1844, the secret societies began 
to appear in public to entice people into the society. At first but a few scores 
would assemble for the purpose, and by night; but, in the course of time, bodies 
of several hundreds held their meeting publicly and in broad day .... Later, the 
secret societies began to rob merchants on a massive scale in daylight. Even on 
the White Cloud Mountains, close to the provincial city, meetings for enlistment 
were held at all times and seasons .... 159 
All the elements discussed above combined to imprison the majority of the 
Chinese in a state of poverty that breeds discontent. The Chinese rural population still 
might have had the miraculous resiliency to rebound from their hardship. But once a 
famine struck, this discontent developed into riots and eventually into revolts. Augustus F. 
385; Marks, Tigers, 307. 
157 Wakeman, Strangers, 110, citing Hu Hsien-chin, The Common Descent Group in China and its 
Functions (New York, Johnson Reprint Corp., 1948), 90. 
158 Wakeman contends that the society was polarized into wealthy and poor by 1845 and that class interests 
were no longer softened by the lineages. See Wakeman, Strangers, chap. l 0. 
159 George W. Cooke, China and Lower Bengal: Being the Times Special Correspondence from China, 
(London: Warne & Routledge, 1858), 435-45. 
85 
Lindley quoted Sir John F. Davis' observations about Chinese famine from 1838 to 1841: 
"During the years 1838-1841, many parts of the empire became plunged in misery and 
want;--so severe was the famine, that many thousands perished, while multitudes were 
driven to insurrection."160 By 1843, the Tiandihui groups were actively recruiting 
members in the settled villages around Guangzhou. In some cases, entire villages joined 
voluntarily because their inhabitants were faced with an economic crisis: the cottage 
weaving industry had declined, taxes had increased dramatically, and 1848-50 and 1852 
were years of poor harvest. 161 
Other parts of China also fell again and again into the clutch of famine in the 
mid-nineteenth century. The calamities included drought, food, storm, sandstorm, hail, 
and failure of crops. 162 It was under these circumstances that not only the Red Turban 
Rebellions but also large-scale rural rebellions arose to threaten the Qing dynasty. 163 The 
greatest and best organized of these anti-dynastic insurrections was, no doubt, the Tai ping 
160 Augustus F. Lindley [Lin-le], Ti-ping Tien-Kwoh: The History oft he Ti-ping Revolution, Vol. I (London: 
Day and Son, Ltd., 1866), 101. 
161 Wong, Yeh Ming-ch 'en, 96; Wakeman, Strangers, 127, 139-44; "The Secret Societies ofKwangtung," 
37-41 
162 Luo, Shigang, 17-18. 
163 This was the larger context in which the Red Turban Rebellions of the 1850s took place. The numbers of 
district towns were seized before the Red Turbans (Tiandihui lodges) were defeated but not eliminated, as 
subsequent events were to be studied in detail in the ensuing chapters. 
86 
Rebellion (1850-1864). This uprising, which covered the whole of South China, was 
accompanied or followed by several others: the Nian Rebellion in Anhui and Shandong 
(1853-1868), the Southwest Muslim Rebellion in Yunnan (1855-1873), and the 
Northwest Muslim Rebellion in Gansu, Shanxi, and Chinese Turkestan (1862-1877). 
The Taping Rebellion 
The growing disorder in China culminated in the Taiping Rebellion and the near 
overthrow of the dynasty. In the environment of Liangguang, the rebellion had much in 
common with the Red Turban Rebellions and the Hakka-Punti War, as seen earlier in this 
chapter, and in its course it was to dominate much of the political and military scene for 
the next fifteen years. 164 The immediate impetus to the Taiping Rebellion was the 
dissatisfaction of a minority group, the Hakkas, in the provinces of Guangdong and 
Guangxi, where the revolt began. As seen earlier, the Hakkas remained comparatively 
unassimilated into the local society, spoke their own dialect, and were differentiated by 
special customs. They also suffered discrimination at the hands of the local majority, and 
164 The basic reasons behind the Taiping rebellion and the other rebellions were the general problems of the 
nation: increasing poverty, natural disaster, alienation, frustrated ambition, and a government too weak and 
corrupt to tackle the problem effectively. In addition, each uprising had its own special set of causes, 
depending on local circumstances and the sorts of people involved. 
87 
conflict between the Hakkas and others was often a spark for political and social conflict. 
The Hakkas provided the basic cadre for the rebellion in its early stages, and 
discrimination against them was an important source of discontent. 
In addition to the special circumstances of the Hakkas, the foreign presence at 
Guangzhou added fuel to local problems. The Qing government had suffered a loss of 
prestige among the people of the South through its easy defeat at the hands of the 
foreigners. In addition, the shift of trade to other ports in the aftermath of the Opium War 
resulted in commercial dislocations in the area and added to its economic woes. The 
foreign presence had also brought zealous, dedicated Christian missionaries to China's 
shores. 
The Taiping Rebellion began in 1851 in Guangxi province among a 
predominantly Hakka group called the Society of God Worshippers, which was under the 
leadership of Hong Xiuquan ~~~ (1814-1864). The Taipings were significantly 
different from the other uprisings of the time, because of the influence of Christianity. 
The society's god was the Christian God, and its leader Hong Xiuquan claimed to be the 
second son of God, who had received from the Heavenly Father the mandate to restore 
88 
China to a state of Christian grace. 165 In practical terms, this mission translated into 
eradicating Confucianism, Taoism, and Buddhism and seizing political power from the 
Manchu rulers who the Taipings demonized. Hong and his group envisioned an 
egalitarian, classless society based on communal ownership of property and the 
communal worship of God. 166 
By 1853 the Taiping army, now over a million strong, had advanced from its base 
in Guangxi, overrun Hunan and Hubei, defeated Manchu armies in Anhui and the lower 
reaches of the Yangzi River, and established the capital ofTaiping Tianguo .::t5¥-:::R~ 
(The Heavenly Kingdom of the Great Peace) at Nanjing. Taiping influence extended over 
six of the richest provinces of China. Hong installed himself as the Heavenly King. China 
now had two dynasties, two kingdoms, vying for control of the country. The Tai ping 
government, however, lacked the stable underpinnings of the traditional Confucian state 
165 Hong's Christian utopia is a hybrid version. Hong's real intention was probably to use the new faith as a 
powerful means to launch his rural revolution. Most of Hong's preaching came from the Old Testament. 
Many Chinese traditional values and beliefs, including the concept of Taiping were combined with the alien 
Christian gospels. See, Jonathan Spence, The Taiping Vision ofC hristian China (Waco, TX: Baylor 
University, 1996), p. 14. 
166 Most useful on the Taiping Rebellion have been Eugene Boardman, Christian Influence upon the 
Taiping Rebellion (Madison, 1952); Philip A. Kuhn, "Taiping Rebellion"; Franz Michael with Chang 
Chung-Ii, The Taiping Rebellion: History and Documents, Vol. I, History (Seattle: University of 
Washington Press, 1966); Vincent Y. Shih, The Taiping Ideology: Its Sources, Interpretations, and 
Influences (Seattle and London: University of Washington Press, 1967); Jen Yu-wen, The Taiping 
Revolutionary Movement (New Haven: Yale University Press, 1972); The Taiping Revolution (Beijing: 
Foreign Language Press, 1976); Jonathan D. Spence, God's Chinese Son: The Taiping Heavenly Kingdom 
ofH ong Xiuquan (New York: W. W. Norton & Company, Inc., 1996). 
89 
that served the Qing. Its leadership lacked unity and the organizational capacity that 
might have tilted the final outcome in its favor. The Taiping Rebellion attained the high 
point of its success and power by 1856; thereafter the movement lost its vigor and the 
upstart dynasty was liquidated in 1864. 
After the forces of the Tai ping moved north in 1851, the main body never 
returned to the Guangdong-Guangxi border region, where the rebellion had had its 
origins. However, other bodies in sympathy with the Taipings succeeded in crossing the 
Guangdong border. The first of these was a body of outlaws bearing the Ming banner, 
which reached the district of Qingyuan to the north of Guangzhou in the summer of 1850. 
In the following year secret society adherents in the neighboring district of Conghua 1ft {t 
joined them in revolt. In the summer of 1851, a Taiping supporter named Ling Shiba ~ 
+J \. also crossed the border in the west and captured the city ofLuoding ~5:E, where he 
held out for a year until finally being bought off. 167 
In part the failure of the Taipings to achieve success in Guangdong can be 
explained by the military events of the first month of the movement. The principal access 
route to Guangdong was, of course, down the West River, but after the original agreement 
167 Wakeman, Strangers, 132. 
90 
between the Taiping and pirates who controlled the river had broken down, this route was 
blocked to the main force. 168 Even after the capture ofYong'an 1)(~, the Taiping still 
intended to carry their rebellion into Guangdong, but found their way barred by Imperial 
forces. 169 It seems probable, however, that by this date Hong Xiuquan had already 
conceived the idea of capturing Nanjing. 170 
The Taipings bypassed Guangzhou. This vast upheaval, however, left its mark on 
the every side of China including the local societies around the Pearl River Delta. The 
Chinese state that emerged from this blood-letting was in many significant ways different 
from what it had been in 1850, fourteen years earlier. 171 Because of the importance of its 
effect on affairs in the whole of China at the time, the subject of the Tai ping Rebellion 
will be referred to again in the following chapter. 
168 Laai Yi-faai, "River Strategy: a place of the Taiping's Military Development," Oriens 5 (1952), 303. 
169 Lindesay Brine, The Taiping Rebellion in China (London, 1862), 151. 
170 Wakeman gives us two reasons for the lack ofTaipings in the Guangdong areas: 1) The real seedbed of 
discontent and zoned ofrecruitment for the Taiping forces was not Liangguang but Hunan and Jiangxi, full 
of unemployed boatmen and coolies; and the Yangzi valley, with its impoverished peasants and 
propertyless vagabonds. 2) The Heavenly Kingdom centered on Nanjing. See Wakeman, Strangers, 131. 
171 See Jen, Revolutionary; Robert P. Weller, Resistance, Chaos and Control in China: Taiping Rebels, 
Taiwanese Ghosts and Tiananmen (Seattle: University of Washington Press, 1994); Michael, Taiping 
Rebellion; Ssu-yu Teng, New Light on the History oft he Taiping Rebellion (New York: Russell & Russell, 
1966). 
91 
4) Summary 
The Pearl River Delta is geomorphologically formed by sediment carried down 
by the West, East and North Rivers on their way to the South China Sea. The Delta is 
distinctive in that it is not a single extensive flat plain of low relief. Instead, it is a 
composite of several basins drained by a number of rivers and their tributaries. The 
confluence of the rivers and the conjunction of their basins take place near Guangzhou, 
the biggest urban center of the region. This composite delta is interlaced with many rivers 
and their branches, most of which are navigable year-round. 
The delta is also distinguished by its natural endowment, which is most favorable 
for agricultural production. With a subtropical location, the delta enjoys warm 
temperatures (21-22 degrees Celsius yearly average), abundant precipitation (1 ,600-1 , 700 
millimeters annually), and ample sunshine, the result being a year-long growing season 
suitable for double-cropping rice. Such favorable conditions, combined with superior 
fertile soil and a well-established waterway system for transporting and marketing farm 
products, have led to the development of an intensive farming system based on the 
production of rice, sugar cane, mulberry, and fruit, as well as silk cocoons and pond 
92 
fish. 172 Thus, despite the fact that the delta region was not richly endowed with energy 
and mineral resources, agricultural and aquatic production provided a raw material base 
diversified enough for flourishing small-scale manufacturing. 
The increased output in farming and manufacturing had in tum led to increased 
trade and prosperity as marketing and transport networks developed and grew. The 
concentration of commercial activities and the specialization of production facilitated the 
agglomeration of population in towns and cities. By the late nineteenth century, the delta 
region had become one of the most urbanized economic regions in China, next only to the 
lower Yangzi region. 173 In addition, the frontier location of the delta, separated from 
China proper, has made the region serve as China's traditional southern gateway for 
foreign trade and sea transportation. Guangzhou, the biggest port city in the delta, was 
one of China's earliest trade outlets and enjoyed its role as the only port for international 
trade before the Opium War.174 
172 Zheng Tianxiang i!l::R~, Zhujiang Sanjiaozhou Jingji Dili Wangluo ~il:-=-Jl'I rfli~i1f±ll!Jlll.~~ 
(Guangzhou: Zhongshan University Press, 1991), 42. 
173 G. W. Skinner, "Regional Urbanization in Nineteenth Century China," in The City in Late Imperial China, 
ed. G. W. Skinner, 211-49 (Stanford: Stanford University Press, I 977). 
174 Since the Tang dynasty, Guangzhou had been the center of commerce and shipping not only for inland 
merchants but also for foreign traders. By 1757 the Qianlong emperor had restricted all foreign trade to 
Guangzhou, and Sino-Western economic relations were governed for almost the next hundred years by the 
so-called Canton system. Each year hundreds of foreign vessels from Southeast Asia, Europe, and America 
sailed into the Pearl River estuary to do business with the Chinese. 
93 
The unique geographical location of the delta also merits special attention. By 
virtue of its location, the delta is relatively remote, situated at the southern end of the 
mainland far away from the political center of Beijing. This remoteness is further under 
scored by the existence of a wide range of high mountains (nanlingshan i¥i-ffl LU), which 
physically separates the delta from the vast territory of the dynasty. 175 It is such 
remoteness that has given the local people considerable flexibility in seeking 
development, and in some circumstances, the possibilities of rebellions. Then, around the 
nineteenth century, with the population continuing to grow and the limits of cultivable 
land filled up by 1850, a critical point had been reached. For the rural population, with 
the extent of cultivable land in Guangdong reached by mid-century, with farming 
techniques as developed as possible, and with yields stagnating, the question became one 
of obtaining more land at the expense of one's neighbors. Thus, fights between neighbors 
over land, water, and hills with trees became endemic in South China176 by the middle of 
the nineteenth century, contributing to the lineage conflicts and feuds that so 
175 Diana Lary, Region and Nation: the Kwangsi Clique in Chinese Politics, 1925-1937 (Cambridge: 
Cambridge University Press, 1975). 
176 For examples in eastern Guangdong, see Marks, Rural Revolution, 60-75 . 
94 
distinguished the region, 177 to the conflict between the Hakkas and the Puntis, including 
the rise of the Tai pings in Guangxi province, to the rise of secret societies and sworn 
brotherhoods, and to the Red Turban Rebellions in the 1850s. 
On the eve of the Tai ping Rebellion yet another event of profound importance 
took place, namely the Opium War of 1839-1844. The Opium War demonstrated the 
weakness of the military forces. In Guangdong the braves who had been levied to 
supplement them had, in many cases, become bandits, so adding to the disorder around 
the Pearl River Delta region. Until the Taiping Rebellion this disorder remained 
unorganized and undirected, and its main effect was a devastating and unsettling 
influence on the countryside. But the rapid success of the Taipings, and in Guangdong of 
the Red Turban Rebellions, exposed the degree of discontent that had developed by· the 
middle of the nineteenth century. The connections between the emerging environmental 
crisis and the various social movements that originated in Guangdong, particularly in the 
Pearl River Delta region are more than coincidental. In short, the economic and social 
crisis around Guangzhou after the Opium War directly caused the massive rebellions of 
the 1850s. 
177 Harry Lamley, "Hsieh-tou." 
95 
CHAPTER III 
THE RED TURBAN REBELLIONS 
In the previous chapter, I examined various factors-the Opium War, lineages, 
the Taipings, land reclamations, feuds, and the Hakka migrations-that facilitated chaos 
and lawlessness around the Pearl River Delta areas. The major focus of this chapter is on 
the Red Turban Rebellions, a series of little studied but significant uprisings by Tiandihui 
members. 178 These rebellions, though themselves not explicitly _about ethnicity, were 
fueled by demographic and ethnic competition for land. Particularly in the Wuyi areas, 
178 Scholars agree that there were two large secret societies in China: the religiously inclined White Lotus 
Society, which prevailed in North China for several hundred or possibly even more than 1,000 years, and 
the Tiandihui. It is generally supposed that most southern secret societies were derived fonn the Tiandihui. 
The group of societies on which this chapter is focused is the Tiandihui, popularly known as ' The Triads' or 
the ' Hong league.' According to Xiao Yishan, the name ' Hong' was used only among the members 
themselves. In general, however, all these groups were usually lumped together under the name Tiandihui 
(Triads; Heaven and Earth Society), and this convention shall be followed here for convenience whenever 
discussing these societies in general. The term Triad was coined by the missionary William Milne in the 
1820's, and has stuck with us ever since. By that time one of the most common names used by the secret 
societies in South China was sanhehui = ir~, which Milne translated as "The society of the Three United, 
or the Triad Society." This association, he claimed, had earlier been known as the Tiandihui, or "The 
Society that unites Heaven and Earth." See William Milne, "Some Account of a Secret Association in 
China, entitled the Triad Society," Transactions oft he Royal Asiatic Society ofG reat Britain and Ireland I 
(1827), 240; Xiao Yishan ;ffi-Ll!, Jindai mimi shehui shiliao :llr{~*..01tf±~9 ::f--l- vol. I , (Changsha: Yuelu 
shu she :Hunansheng xinhua shudian faxing, 1986), 4; Chesneaux, "Secret Societies," 2. 
96 
the Red Turban Rebellions emerged out of a local Tiandihui network and an alliance 
between this network and the Puntis, to this network, an alliance that created ideological 
differences between the Hakkas and the Puntis, thereby intensifying pre-existing cultural 
differences. 
Eventually, the Red Turban Rebellions in the Wuyi region justified the Puntis to 
mobilizing against their Hakka competitors. By inviting the formation of loyal Hakka 
militia to suppress the largely Punti rebels, Qing administrators militarily pitted the 
Hakkas against the Puntis and permitted the Hakkas (by virtue of their enhanced identity 
as loyal Qing supporters) to take violent punitive action against the Puntis. These 
disruptions of local society, particularly in the Wuyi region, set the stage for the Hakka-
Punti War. Thus an understanding of the dynamics of the Red Turban Rebellions is 
essential to the analysis of the process whereby Hakka consciousness was formed. 
Although national events had major impact on the timing and evolution of the 
rebellions, the contention here is that the origin and activities of the Red Turbans are 
inextricably linked to an ongoing process of adaptive competition within the Pearl River 
Delta area. I will now proceed to a discussion of those activities and the rebellions 
undertaken by the Tiandihui at the Pearl River Delta. 
97 
1) The Course of the Red Turban Rebellions 
A Series of Incidents Lead To The Red Turban Rebellions 
In Guangdong the year following 1850 was a period of growing disorder. The 
trouble had begun with the Tai ping incursions across the border from Guangxi. Of these, 
the one headed by Ling Shiba had been the most difficult to subdue. Ling, a native of the 
prefecture of Gaozhou, went to Guangxi in 1847 and joined the Taipings. Then in May 
1851 , he returned to his native land and set up Taiping headquarters at a place called 
Luojing m~. Severe fighting between the rebels and the Qing forces continued 
throughout the year of 1852. The war effort by Xu Guangjin, a Viceroy ofLiangguang, 
against Ling Shiba in Gaozhou made little headway. 179 In same year, the border city of 
Wuzhou ffi 1-1'1 was under siege by a Tiandihui rebel group led by Tian Fang EB 7i . Other 
Tiandihui rebels closer to the delta were inspired to attack Conghua, only tens of miles 
from Guangzhou. 180 
179 Xu Guangjin was a Viceroy ofLiangguang. 
180 To save the city from imminent attack, a thick barrier of garrison troops and hired mercenaries was set 
up between the delta and the borderland. Xu Guangjin and Ye Mingchen, successive viceroys of 
Liangguang, personally led armies into the mountains. By August 1852 they had blocked the rebel descent, 
decisively crushing Ling Shiba at Luoding m~. See F.O. 931.1190. Correspondence regarding a rebel, 
Ling Shiba (1850); F.O. 931.1281 . Draft memorial by Xu Guangjin and Ye Mingchen regarding Ling Shiba 
(April 10, 1850); Daqing Lichao shilu, Xianfeng period :k~ lf.UJUUJ, rnz_:.~ (Taibei, 1964 reprint), 
32.11 ; Wakeman, Strangers, pp. 132-133. 
98 
Rebel incursions and the failure of the provincial authorities to deal with them 
effectively were not the only cause of discontent. The gentry had become increasingly 
alienated by the extent to which they were called upon to finance operations against 
rebels, not only in Guangdong and Guangxi, but also in other provinces. This burden fell 
particularly heavily on the gentry of the Pearl River Delta. They were further incensed by 
the discovery that a local magistrate had concealed a remission of taxes within the 
province in 1851. In a number of districts, they reacted violently refusing to pay any land 
taxes. The situation further deteriorated in 1852, after heavy flooding in the delta had 
destroyed much of the harvest. The provincial authorities, however, were in such 
desperate need of funds that they continued making demands on the gentry. In 1853 they 
were asked to supply one month's rent on all their property to the provincial treasury. 
Eventually, these burdens were passed on to the tenant farmers who had already suffered 
much over the past several years. 181 
By far the most significant event took place in 1853, when the Taipings captured 
Nanjing and established this city as their Heavenly Kingdom. Local disaffected elements, 
181 Nishikawa Kikuko gs:};:§~ r, "Juntoku danren sokyokuno seitatsu Jl!Ji:~l!l~~fiVO)Jit:tl," in 
Toyobunka kenkyusho kiyo Jl:t?$Jtfl::1iJf~mff.c~105 (Tokyo University, 1988), 14-16; Liu Weizhang ~f-i; 
1lr, "Taiping tianguo gemingshi Qingchao Guangdong caizheng .:t:-5¥::ROOiJ-®fl't~!j'JJJ*Y!tI!ili," in 
Taiping tianguo yujindai zhongguo .:t:-5¥:;ROOl;j:ili:f~i:p 00, (Guangzhou: Guangdong renmin chubanshe, 
1993), 366; Wakeman, Strangers, pp.132- I 36; idem, "Secret Societies of Kwangtung," pp.3 8-39. 
99 
especially the secret societies, were very much excited by these disturbances, and they kept 
the local officials and officers very busy indeed. One of the secret society proclamations, 
which appeared among British Foreign Office papers, reads, "The ancient books tell us that 
once in five centuries some man of talent beyond his fellows will appear, on whom the hope 
of the nation will depend. That period has elapsed since the rise of the Ming dynasty, and it is 
full time that a hero should come forward and save the nation."182 
In 1853, soon after the Taiping occupation of Nanjing, disturbances fomented by 
the Tiandihui once again occurred in the vicinity of Guangzhou. Large numbers were said 
to be joining the Tiandihui. 183 The society was anti-dynastic, but in its early stages this 
fresh burst of activity consisted merely of undirected attacks and banditry, prompted by 
economic grievances and the shortage of rice. In that same year Amoy and Shanghai were 
captured by secret society bands. In Guangdong they were not to achieve the same 
success, but they were convinced that it was high time for them to take action. 
Fearful ofTaipings penetration into Guangdong, Qing authorities made strenuous 
efforts to raise money for defense. 184 Their fund-raising generally took the form of 
182 F.O. 17.126, Bowring-Clarendon, Desp. 18 (Jan. 9, 1855). 
183 China Mail (Hong Kong, 1845-1911), 14 (April 1853). 
184 The Qing government was mortally afraid that the Taipings would link up with the Tiandihui rebels and 
100 
surtaxes or of "contributions" that were anything but voluntary. These exactions roused 
opposition in Shilong n ~. a village of Dongguan county. Many local Tiandihui 
members actively participated in the tax-resistance movement. 185 In October 1853, the 
situation further deteriorated when a band allied with the Small Sword rebels of Amoy 
were driven out of their center at the Bogue, to the south of Guangzhou, and established 
themselves in Huizhou. In the following months severe disorders were reported in 
Huizhou and Dongguan.186 The provincial commander-in-chief was sent to the area, and 
reported a victory that was punished with great severity. Banner troops were unleashed to 
wipe out not only alleged local Tiandihui members but also entire villages. 187 The 
disturbed state of the province provided an ideal setting for the further widespread 
disturbances known as the Red Turban Rebellions. 
magically transfonn them into a military juggernaut. 1n the sources, an occasional statement can even be 
found to the effect that the Taiping leaders instructed the Tiandihui leaders and others on tactics and molded 
them into effective troops. A memorial ofYe Mingchen said, 'The rebellious bandits of Jiangnan had 
secretly sent their cohorts back to Guangdong to get all evil elements in league with them and start riots at 
the same time. Gaoming, Shunde, Huilai and other districts and the city of Chaoqingfu had been lost to 
them.' See Yong Kweng, Sohak, 63 ; Luo, Taipingtianguo shiqi vol. 2, 54-55; "Taipingtianguo chunguan 
zheng chengxiang zhuodeng zhi Folingshi tongbing dayuanshuai Lideng diewen ;;t'(f- ~!Jil~'gjf!E;;ffi* 
~ 3&1,ll lhtrtfg-JfA::kJG YrJi*~~ )( " in Hongbing Qiyi, 62-63 ; F.O. 931.1599 Note concerning 
interrogation of Gan Xian if5t. 
185 Nishikawa Kikµko , "Juntoku," 13- I 6. 
186 China Mail 27 (October 1853); Guangzhou FZ, 82.2b. 
187 Wakeman, Strangers, 138. 
101 
The Beginning and High Tide of the Red Turban Rebellions188 
By the beginning of 1854, Ye Mingchen, Governor-general of Liangguang, was 
already receiving urgent dispatches from his agents informing him that every day in the 
suburbs of Guangzhou, hundreds upon hundreds of people were secretly assembling and 
then as secretly dispersing. These assemblies were in fact secret society gatherings held 
to admit new members. 189 
In early 1854, the Tiandihui groups in the Pearl River Delta areas apparently 
embarked on a greater enterprise under the leadership of a Tiandihui leader, Chen Song 
~ti. 190 In March 1854, Chen sent a Tiandihui agitator, Zhong De ~ffi, into Xinqiao ffi 
188 For a comprehensive outline of the history of the Red Turban rebellions, See Wakeman, Strangers, 
Chapters. XIV and XV; idem, "Secret Societies." 
189 F.O. 931.1512. A report on the Sanhehui (1854/1855); F.O. 931.1513. A summons to war against the 
Sanhehui in Foshan (1854/1855) .. 
189 ShundeXZ, 23:5.b, translated in Wakeman, Strangers, p. 139 
190 Nothing concrete is known of Chen Song's personal background other than his Tiandihui activities in 
Guangdong. Through his activities, several Tiandihui lodges were created in Guangdong in the mid-
nineteenth century. The Tiandihui in mid-nineteenth century Guangdong was organized in a loose 
confederation of local lodges (shantang L1.J '.!it) that retained complete control over their individual courses 
of action and finances . Every branch of the Tiandihui adopted the name of a certain mountain as its fancy 
name. The name adopted did not necessarily correspond to an actual mountain. There were also various 
fancy names of "halls," derived from the Zhongyitang :!',~'.¥'. (the Hall of Loyalty and Righteousness) in 
the novel, Shuihuzhuan ;,J(f.4{$. The most notable Tiandihui local lodges around Guangzhou areas at the 
time were He Liu's Hongyitang #!;~'.!it , the Hongshuntang #!;JI~ '.¥: led by Chen Kai and Li Wenmao, the 
Hongxingtang #l;Ji '.!it led by Chen Jinkang ~~ijU, and the Hongdetang #!;~'.!it led by Liang Peiyou 
~~R - See Hirayama Shu, "Zhongguo," 81-82, Kuhn, Rebellion and its enemies, 168; Davis, Primitive, 
103; Hu Zhusheng tiJHl~, Qingdai Hongmenshi iw1-\:#Ul.9:. (Shenyang: Liaoning renrnin chubanshe, 
1996), 309. 
102 
fit and Dongguan to contact the leaders of local Tiandihui lodges such as Chen Xianliang 
~~~, He Liu {ii};\ and Yuan Yushan ~=E: L.lJ in the hope of stirring up a general 
uprising. 191 After Zhong De's arrest by the Qing authorities, Chen Song appointed Li 
Wenmao as a leader of the Tiandihui lodges around Guangzhou (Pingjingwang Zp-~.::f. ) 
and planned an uprising for early July. 192 But the first major outbreak came a month 
earlier than their original date, with the revolt of He Liu. 193 
On 17 June 1854 He Liu, a leader of a Tiandihui local lodge (Hongyitang tA~ 
¥'.) in Shilong, animated by Chen Song's direction, seized the opportunity to rise up by 
proclaiming revenge for his blood-oath brothers who had been killed during the 
suppression of the Qing purge at Shilong. He Liu and his followers attacked the district 
191 Wakeman argues that there is no evidence to show that the rebellion was planned in advance. "Rather, 
rebellion engendered rebellion, in a distinct crescendo of disorder after He Liu 's revolt." But Wakeman 
failed to provide an explanation of how such a large number of people could have gathered so suddenly 
and make an organized attack on an administrative seat. In contrast to Wakeman's contention, many 
documents show us that there was plotted conspiracy under the Tiandihui leader, Chen Song. The shadowy 
secret society networks with which Chen Song !lifl~ were involved provided a useful organizational 
framework that allowed much swifter mobilization of followers than would otherwise have been possible. 
See "Huifei zongtoumu Chen Song shiyou wr~~M § llifl~ $ El3,"in Sasaki Masaya ftcftc:;tIE~, edit. 
Shinmatsu no himitsu kessha-Shiryo hen ~*1¥Jit,W~ifrt -ji~Ji,(Tokyo: Kindai Chugoku Kenkyii 
Iinkai, 1967), 21; F.O. 931.1090. Information regarding bandits in Guangzhou (1850s); idem, 1439. 
Reports on Chen Song (1854); idem, 1484. List of36 Red Turban rebels in the Nanhai area (1854). For 
Wakeman 's argument, see Wakeman, Strangers, p, 139. 
192 "Huifei zongtoumu," in Sasaki, Shimatsu, . 21; F.O. 931 .1090. Information regarding bandits in Henan 
(1850s). 
193 He Liu was from a merchant family in Shunde county and said to have moved into Shilong village, 
Dongguan county, where he come under the influence of the Tiandihui leader Chen Song. See F.O. 
931 .1439. A report on Chen Song (1854). 
103 
city, Dongguan and sacked it. 194 This was soon followed by outbreaks of rebellions in all 
the districts surrounding Guangzhou. Rebellions soon engulfed the Pearl River Delta 
areas. Town after town fell to insurgents, who held several administrative seats 
continuously for several years (see Map 1). The government was unable to suppress these 
rebellions until early 1855, when massive reinforcements were transferred from 
neighboring provinces. 195 
All the insurrectionists and their collaborators wore red turbans as an identifying 
sign, which earned them the name ofHongjinze }I rh ~ (Red Turban Bandit) or 
Hongtouze }IM~ (Red Head Bandit). They were also called Hongbing ~A (Vast 
Soldiers or Triad Soldiers) since the Red Turbans used the insignia and watchwords of the 
Tiandihui. 196 The principal prize was, of course, Guangzhou itself, and in the summer of 
1854 the Red Turbans came close to taking the city. 197 
194 A number of factors had to come together to precipitate the revolt, including maladministration by the 
local magistrate, Jiang Zhaoen ITff,l~L See Junjichutang zouzhe lufu • ~~~~ii~~ffi (Taibei: Guoli 
Gugong Bowuyuan, 1982 reprint), Xianfeng 4.7.25; Dongguan XZ, 35.7. 
195Jian Youwen ffiDZ.X, Taiping Tianguo quanshi j;; 112:xlll-i:j: (Hong Kong: Jianshi mengjin shudian, 
1962), 888-93 . 
196 Soda Yo ffi l:E 7$, "Kokinko f,I rh ~," in Toyoshi kenkyu *?'$~!ff~ , 38.4 (March 1980), 38-4, 57; 
Wakeman, Strangers, Chapter 14. 
197 The sources regularly mention almost 34 major rebel leaders and many minor ones. See F.O. 931 .1497. 
A list of34 Red Turban leaders (1854); Lu, Tiandihui Zhengquan, 133-142. 
104 
At Dongguan, the Red Turbans halted their attacks on Guangzhou and took time 
out to organize their rebel groups more fully. The rebels needed time to recoup and 
reorganize their forces, incorporating the many new recruits they had gained. More 
important still may have been the need to clarify the chain of command among the 
Tiandihui leaders. The critical military situation that would face them at Guangzhou 
made it all the more necessary to strengthen the fighting spirit by a promise of great 
rewards. 
Map 1. Counties of the Pearl River Delta. 
105 
After the sudden arrest of Chen Song by the Qing authorities on October, 1854, 
all Tiandihui lodges concentrated their bands in the Dongguan area, where they formed 
an alliance and elected Li Wenmao and Chen Kai, co-leaders of the strongest group 
(Hongxuntang ~/l!Wi'.§t), as mengzhu M±: (Lords of the Alliance). 198 
In late 1854, the Tiandihui leaders assembled in a temple at Foshan and started 
their rebellion with a sacrifice to the five-colored banner indicating the legendary 
founders of Tiandihui. 199 This matter of formality symbolized something concrete in the 
Tiandihui organization. For example, during the ceremony of sacrifice to the Tiandihui 
banner, the rebels used to recite verses such as this: 
Five men began fighting against the Qing troops; They founded the invincible 
Family of the Hong Brothers; After we have taken province, district, and region, 
We shall overthrow the Qing and restore the Ming.200 
The concentration of bands under the leadership of Li Wenmao and Chen Kai 
indicates that the Tiandihui groups were trying to build a unified force out of scattered 
198 Tiandihui symbolism marked many aspects of the rebellions. Li and Chen's chosen title-"alliance 
leader" (mengzhu)--resonates with the terminology of brotherhoods and associations. Rebels added the 
hong character to the ranks of some of the generals. Perhaps more importantly, rebels retained the language 
and some of egalitarian practices associated with brotherhoods. See F.O. 931.1439. Report on Chen Song 
(1854); "Dongguanxian shilong difang qishi yuanyou *~~E l'fi:1-mJJ~ Jelai:~ 83," in Tiandihui, 21. 
199 See Enping XZ, 14.2 
200 As Wakeman depicts the anti-Manchu slogan by the Red Turbans, "Overthrow the Qing and restore the 
Ming. (fan-Qingfu-Ming) was "more than a heroic echo of the past." It gave the rebels political relevance, 
even providing a form of social respectability. See Liu, Tiandihui, I 0.32; Wakeman, Strangers, 35. 
106 
and incoherent branches. The sacrifice to a flag implies not only that the disunited and 
disorganized Tiandihui groups had rallied under one standard, but that they were pledged 
to a greater enterprise-rebellion. The unification of the numerous branches formerly led 
by local tangzhu marked a higher degree of centralized organization over the previously 
scattered forces. 201 Yet this coalition, though representing greater unity, did not last, 
because it was not yet founded on an institutional reality. Immediately after its formation, 
the combined force met a test that resulted in disaster. 
Li Wenmao divided his lodge into four groups for the campaign: Chen Xianliang 
l!t~ .Bl, with his army of about 30,000 from the east; Gan Xian if -!Jr:, from the north; Li 
himself from the west; and Lin Guanglong #b'tl3! from the south-all were to attack the 
city. Chen Kai l!tOO also agreed to lead his army, upwards of 100,000, to join in the 
attack on Guangzhou. Li also asked He Liu ofHongyitang to bring his troops, 10,000 
strong, to reinforce Chen Xianliang in attacking Guangzhou from the east. They started 
with conspicuous success, capturing the important city ofFoshan, only 15 miles from 
20 1 Though the supreme leader of the complex Tiandihui groups always made the major decisions 
concerning external affairs, the sub-leaders provided a secondary level of command. In the event of the 
leader's death or absence, they could take over to prevent the alliance's disintegration, taking primary 
responsibility for their own followers while seeking to restructure the rebel groups according to the 
previously established hierarchy. Li Wenmao and other top leaders could thus keep the Red Turban forces 
without jeopardizing their control after Chen Song's death. See lbid. ; F.O. 931 .1440. A Red Turban 
Proclamation ( 1854 ). 
107 
Guangzhou, on July 14. The Red Turban leader there, Chen Kai, proclaimed his intention 
to restore the Ming dynasty, and made some attempt to institute a popular program.202 In 
the time-honored tradition of Chinese rebels, Chen Kai used imperial trappings to bolster 
his legitimacy. Chen declared that a new reign period was in e:ffe ct: it was now the first 
year of the Daning *'¥. (Great Peace) reign.203 
In other parts of Guangdong Red Turban bands first attacked the chief towns of 
the districts in which they had originated. To the southwest of Guangzhou, in the districts 
that were later to be involved in the Hakka-Punti War, the pattern of events was similar. 
Risings are recorded as having taken place in Xinning by on July 21, in Kaiping on 
August 12, and in Gaoming and Heshan at about the same time. The district cities of 
Heshan and Kaiping were captured by the Red Turbans (Hongdetang #f.J!ti:) led by 
Liang Peiyou ~:!:-g:1;z, and the district city ofEnping was also besieged.204 The Red 
202 F.O. 931.1558 Deposition regarding rebel activities in Zengcheng, Dongguan and Foshan areas in 1854 
(April, 1855). 
203 Imperial imagery was mixed. Chen himself assumed the title of Zhennanwang ~ ffi :E and was 
particularly emphatic in his claim to be restoring the Ming dynasty, a claim that explains-and adds 
credence to-the tales linking his revolt to the machinations of the Tiandihui. Edicts and proclamations 
were invariably issued in the name of the Ming as well. SeeXunzhou FZ, 27.5; F.O. 931.1532. A report on 
the Red Turbans in Panyu (1855); F.O. 931.1569. Deposition of the rebel (July 1855). 
204 Enping XZ, 14:7b; Guangzhou FZ, 82: 10b; Kaiping XZ, 21 :3a; Su Fengwen 91' )XI. X . "Pinggui jilue ff£ 
tc.~," in Hongbing Qiyi shiliao, 772-813 ; Idem., "Gufei zhonglu /N ~ ,g,;lK," in Hongbing Qiyi shiliao, 
814-840. 
108 
Turbans were particularly active in the prefectures of Guangzhou and Zhaozhou; in these 
areas more district cities fell to the insurrectionists, such as Huaxian, Qingyuan, 
Zengcheng, Conghua, Longmen, Kaijian Im~, Yingde, and Changluo. The Red Turbans 
also terrorized the prefecture of Jiayingzhou.205 It is estimated that the total number of 
rebels around Guangzhou was estimated more than 200,000.206 
Qing Forces 
The initial failure of the Qing forces against the Red Turban rebels revealed the 
failure of the methods of rural control employed by the Qing government. To prevent the 
development of local disorder, the Qing had continued and improved the sub-
administrative apparatus of control that had existed under the Ming. The most important 
of these was the baojia. For the Qing the baojia had two purposes. All households and 
individuals were compelled to register and to be formed into groups of households, the 
bao andjia. In this respect it resembled a form of census, at the same time singling out 
205 Su, "Pinggui," 800-812; Wakeman, Strangers, Chapter 14; Jian Youwen ffiDZ.X, Taiping tianguo 
quanshi :::t:2P-:.Rm~~. Vol. II, Chaps. I and 2. 
206 Qing force at Guangzhou had to rely on the braves from Chaozhou and delta braves from Dongguan. 
There were about two thousands Chaozhou braves and four thousands Dongguan braves stationed at 
Guangzhou. See Dongguan XZ, 72.4a; Scarth, Twelve Years, 229; Wong, Yeh Ming-Ch 'en, p. 96. 
109 
persons responsible for carrying out the system. Its other function, which was probably 
considered the more important, was the duty imposed on baojia heads to report the 
occurrence of crimes and the existence of criminals and rebels in their neighborhood.207 
Although the baojia system of collective security was often shown to be 
defective in the details of its operation, in the relatively peaceful years of the eighteenth 
century there seemed little to indicate that it might be inadequate as a method of rural 
control. By the beginning of the nineteenth century, however, there was plenty of 
evidence to show that the system could not control the growing disorder. Even when 
working well, the baojia as a system of maintaining local security had limitations. 
Particularly in Guangdong, where lineages were strong, the system was modified to take 
into consideration local organization. For the purposes of the baojia, lineage villages 
were not divided, but treated as a single unit under a local head. This conformed with the 
Qing policy of encouraging clan solidarity, but it did little to advance the purposes of the 
baojia, for the interests of the lineages did not necessarily coincide with those of the 
authorities, nor could the clan be expected to act against members of the lineage. The 
gentry were incorporated into the baojia, although in recognition of their status they were 
207 Hsiao, Rural China, pp. 43-83 . 
110 
relieved of the duties prescribed under the system. But in order to maintain a balance of 
power in rural areas, the gentry were not allowed to take control of the baojia. These 
organizational arrangements were only partially successful in practice. In a number of 
ways the gentry opposed their inclusion under the baojia, and in the words of Hsiao 
Kung-chuan "probably prevented police control from being completely or consistently 
effective in every part of the empire."208 Until the end of the eighteenth century they 
made little attempt to try to gain control of the baojia, and rather stood aside from it. But 
in the nineteenth century the gentry, impelled by fears for their own security, became 
much more interested in, and often gave active support to the baojia.209 
The baojia was intended to prevent the development of disorder in rural areas. 
But to deal with more serious threats to security, the Qing government had at its 
command a large standing army. This was organized in two great divisions, the Banner 
troops and troops of the Green Standard. In Guangdong the Banner troops, nominally 
numbering about 5,000, were under the command of the Manchu General and all 
stationed at Guangzhou. The Green Standard troops, numbering approximately 70,000, 
208 Ibid., p. 69. 
209 F.O. 931.1751 Regulations for the prevention of bandits. (n.d.); F.O. 931.1173 Regulations regarding 
organization of baojia for the prevention ofrebels (1850s). 
111 
were distributed throughout the province. The system of military control had remained 
largely unchanged since the Manchu conquest, and the disinclination of the Manchus to 
place complete trust in the Green Standard forces was displayed by the practice of 
dividing the army into separate commands. The greater part of the forces was under the 
Provincial Commander-in-Chief(~~), and was stationed in Huizhou. Under him also 
were the naval forces. But both the Governor-general and the Governor had separate 
forces under their command, the former largely stationed at Zhaoqing, and the latter at 
Guangzhou. The situation was further complicated by the fact that, although the military 
headquarters of the various commands were territorially separate, there was no territorial 
division of military responsibility, and the forces of each command were distributed 
throughout the province.210 
Certainly military garrisons were distributed in most of the rural areas of 
Guangdong; however, there was insufficient protection of the local officials against the 
threat of rebellion. While the magistrate had overall responsibility for the security of his 
district, the forces at his disposal were extremely limited. Military garrisons were not 
210 The commander-in-chief of the land forces was stationed at Huizhou. His troops consisted of four 
brigades (zhen ijt), namely, Shaozhou, Chaozhou, Gaozhou and Qiongzhou :l:t1'1'1. Humen, on the other 
hand, was the headquarters of the marine forces, which consisted of four squadrons, two on each side of 
Humen along the coast. For details on the distribution of military forces in Guangdong, See China 
Repository, Vol. 4 (1835-6), 284, and Vol. 20 (1851), 54, 254. 
112 
under his command. In order to make arrests, he could use the yamen-runners, who in 
some ways served as a police force. But if there was any disorder in the district, the 
magistrate had either to find local aid to deal with it, or to call for assistance from outside, 
laying himself open to a charge of incompetence. This problem is rooted in the absence 
of central government structure below the county level, coupled with laws prohibiting 
magistrates from serving in their home districts. During the Qing at the village level there 
were local constables, headmen, and gentry families, and even government units for 
purposes of taxation, but "no formal government of any sort existed below the zhou 
[department] and xian [county] levels. "211 
In this period, most counties had a population ranging from 100,000 to 250,000, 
numbers very difficult to keep under close surveillance.212 To be sure, the district 
magistrate had subordinates charged with collecting taxes and information, but they could 
be bribed or intimidated by powerful local interests, including the secret societies 
themselves. Other difficulties of county officials included insufficient staff, lack of funds, 
and short terms in office, all of which in some situations could lead to unfamiliarity with 
21 1 Ch'u T'ung-tsu, Local Government in China under the Ch 'ing (Cambridge: Harvard University Press, 
1962), 1. 
212 Hsiao, Rural China, 5. 
113 
local conditions.213 With no more than a small constabulary at their immediate command, 
most magistrates were not equipped to deal with large, well-armed rebels. In any case, 
from the viewpoint of the average local official, for whom the slightest hint of unrest 
could mark the end of his career, any attempt to suppress local trouble was extremely 
perilous. Unless sure of success, most magistrates sought to ignore the suppression 
directives that periodically descended upon them from their less vulnerable superiors.214 
The incompetence of local magistrates against rebel threats was only part of the 
Qing local security problem; another problem lay in the soldiers themselves. There had 
been several cases of Manchu brigadier generals who reported that they had won the 
great victories, though in fact they had fled from their posts at the approach of the rebels. 
The lies were later discovered after an investigation by Qing censors.215 Often brigadier 
213 Chang Chung-Ii, The Chinese Gentry (Seattle: University of Washington Press, 1955), pp.52-53 
2 14 One can find herein another reason-government negligence and connivance- for the rebels' 
continuation and development. In 1803 the Jiaqing emperor said that since avaricious officials took 
advantage of people's legal disputes and feuds for the purpose of blackmail, the people first resented, then 
hated, and finally fought against the officials. "Are the people to be blamed for their conduct?" the emperor 
queried. "They were actually incited by local officials." The emperor noticed that, because the disorders in 
a district would bring disgrace and penalty to the magistrate, many local officials in recent years had 
ignored important local disorder cases. This was especially true in later period in Humen where the 
oppressiveness of the garrison soldiers had led many local people to join secret societies, eventually 
resulting in armed attacks on the soldiers. See Da Qing Renzong, 117 .18; 347, 17-18; Wakeman, "Secret 
Societies," 36. F.O. 931.1163 A report on the longstanding hostility between the soldiers garrisoned at 
Humen and the local population. (1850s). 
215 For instance, a Provincial judge, Xu Xiangguang ~ifit::3/t was impeached by Governor-general, Ye 
Mingchen for bad military organization, false victory reports regarding the fall and recovery ofXingan, and 
corruption. See F.O. 931.1412 Draft Memorial ofYe Mingchen (Nov. 20, 1853). F.O. 931.1395 A 
114 
generals were too old to totter along, but they were still commanders at the front.216 
Under such weak generals and poor political and social conditions, the morale of the 
soldiers was very low.217 For instance, two Mongolian officers each commanded more 
than one thousand cavalry, all of whom had been in the service for more than ten years; 
they were tired, ill, and homesick, and they had no desire to fight.218 The Green Standards, 
too, had deteriorated. They were in arrears on pay for years on end. And without good 
officers, the soldiers became inefficient. The guard posts also disintegrated. By the mid-
nineteenth century, a foreign eyewitness, the Reverend Krone, wrote that on his visit to 
some of them, ' he found neither guns nor soldiers', and that 'the places themselves 
showed no signs of fortification save a dilapidated wall'. 219 
The average county garrison numbered between 150 and 300 soldiers raised from 
among the young men of the county, whose main responsibility was to ensure security 
confidential letter from the prefect ofHuizhou regarding the falsifying reports of the magistrate for 
exaggerating a military emergency and shunning responsibilities. 
216 F.O. 931 .1662 An Edict regarding a case of alleged negligence by Qing forces. (Oct. 6, 1857). 
217 The phrase "the officers and soldiers had no desire to fight (shi wu douzhi ± $ilMJi$)" appears 
frequently in government documents. See Ibid., F.O. 931. 1395. 
218 F.O. 931.1760 List of soldiers and an officer who interfered with official matters. (n.d.) 
219Krone, "A notice of the Sanon District,' China Branch, Royal Asiatic Society, Transactions, ( 1859), 71 -
105. 
115 
within the county boundaries. When troubles were reported in outlying areas these troops 
were supposed to take independent action to suppress the rebels. However, the local 
garrison, apart from a few well-meaning youths who volunteered to protect their home 
county, generally attracted the same rough sorts who joined the rebel forces or the 
government army. Indeed, the same man could even be bandit one day and soldier the 
next. At worst, someone in the magistrate's office itself might even be in collusion with 
the rebels, selling them information about suppression plans.220 What happened at Foshan 
may serve as a good example. On 25 May 1855, three suspects were arrested during a 
house-to-house check. The suspects confessed that they were soldiers from a company 
stationed in the neighborhood, but at the same time they were members of the Tiandihui 
and had joined the forces of Chen Kai, who had occupied the town the year before.221 
Statements made in the same case provided evidence of the soldiers ' drug-addiction, 
gambling, prostitution, illicit dues, and affiliations with the Tiandihui.222 Often greedy 
22°F or instance, a Brigade General, Wang Pengnian :E)Jll~- was impeached by Xu Guangjin for his attempt 
to conceal the facts in connection with a case of piracy. Some minor Xiangshan officials also collaborated 
with the Red Turban rebels. See F.O. 931.1285 Memorial of Xu Guangjin (April 28, 1851); F.O. 931.1504 
A letter reporting on the Red Turbans under Chen Kai (1854). 
22 1 F.O. 931.1522. Report on three rebels. (1855); F.O. 931.1493 and F.O. 931.1133. Reports on three 
executed rebels at Foshan (June 1855). 
222 Ibid. Soldiers and militiamen of the Opium War were discharged when the foreign menace was 
apparently disposed of after 1849 and many of them became bandits or joined secret societies as soon as 
they were discharged. According to Wang Tianjiang, there were approximately sixty to seventy percent of 
116 
soldiers relished the opportunity to collect the loot of the rebels for themselves. For 
example, during the battle against Chen Ji ~ a of Shunde, they immediately fell on the 
booty left by the rebels who were routed, despite the order to pursue the rebels and wipe 
them out. 223 
Banner troops stationed at the provincial city had similar problems. The standard 
of living of the average Banner ~older had declined considerably by the mid-nineteenth 
century. Groups of idle young Bannermen wandered about the city of Guangzhou, 
provoking trouble wherever they could, and became a serious problem. As a result, it was 
suggested to the governor-general that he take advantage of the urgent demand for 
soldiers during the siege of Guangzhou in 1854 to recruit the jobless Bannermen into a 
mercenary force attached to their own Banner to strengthen the defense of the city.224 
In late 1854 Red Turban rebels appeared everywhere in the Pearl River Delta and 
government forces were in disarray. In mid-nineteenth century, Guangzhou was supposed 
to have an authorized 15,000 government troops. In reality the total number of soldiers 
the Green Standard and Hunan Annies in the secret societies during the 1860s. See Wang, "mimi shehui," 
85-6. 
223 F.O. 931.1722. Note on rebellious societies along the coast of Guangdong ( c. 1855) 
224 F.O. 931 .1744. Five proposals for the defense of Guangzhou (1854). 
117 
available in Guangzhou was less than 5,000, and, since the generals cared only for the 
protection of their own provinces, the isolated units were frequently surrounded and 
harassed by their foes.225 For instance, Ye Mingchen in 1855 made a sharp complaint to 
the throne against the Qing forces from other provinces. He said that he could not 
command any troops in the neighboring provinces, and the soldiers were either still at 
their original stations or had purposely slowed down in their movements so that others 
might go first and bear the brunt of fighting. Thus no other governor made much of an 
effort to aid in the defeat of strong forces of the Red Turbans in Guangdong.226 In other 
words, the incompetence and low morale of the Qing officers and men was one of the 
major reasons for the prolonged war against the rebel forces. 
But most importantly, the poor relationship between soldiers and civilians long 
delayed of military success. For many years the government could not count on civilian 
support for the imperial soldiers, who did innumerable horrible things to the people. The 
soldiers commonly occupied people's private houses, burning the tables, chairs, and other 
225 More than two thirds of the troops were temporarily disabled because of on-going troubles in rural areas 
and in neighboring province. For instance, upwards of 3,000 soldiers were sent to fight the Taipings in 
Hunan and many were also transferred to north Guangdong to fight against the rebels. See F.O. 17 .217. 
Desp. 59, Incl. , Morrison's report (Nov. 9, 1854); F.O. 931 . 1098 A list of officers and numbers of troops 
(1850s). 
226 F.O. 931 .1609 Ye to Commander-in-chief of Guangxi (I 855); Jian, Quanshi, 852. 
118 
furniture to keep warm. They swarmed in city streets, and blackmailed shopkeepers. 
Taking vegetables, fruit, and miscellaneous articles without paying, they often beat 
peddlers to death when payment was demanded. The devastation the soldiers caused 
made the people flee from their town and villages when they heard that government 
troops were to pass through. 227 Such behavior created bad relations between the 
government and people. In the words of the classic aphorism, "the officials compelled the 
people to rebel (guanbi minfan 'g~.§:&)."228 Such events were a constant theme in real 
life as well as in fictionalized, romantic accounts of banditry like the Shuihu zhuan, 
implying that behind every rebel was the action (o r inaction) of some uncaring official. In 
Guangdong, the situation was certainly bad. The conduct of officials serving in the 
province was regarded as the basic cause of disorder in the early years of the Red Turban 
Rebellions: 
Where does the source of the present day turbulence lie? It lies in those 
magistrates that are greedy and cruel. Let us speak of western Guangdong. When 
disorder had not yet arisen in western Guangdong, officials who held posts in 
this region ... perceiving that the custom there is primitive and the people are 
ignorant, treated the latter as if they were birds, beasts, or savages that deserve no 
affection or care. 
227 Qinding Daqing huidian shili ~'.:\E:ktrf wr :ll!!-¥H7'1J, (Taibei: Xinwenfeng, 1899 edition), 779. 13a-19b. 
228 Shi Naian, The marshes of Mount Liang : a new translation oft he Shuihu zhuan, trans. John and Alex 
Dent-Young (Hong Kong: Chinese University Press, 1994), Pt. 1, 25. 
119 
The writer goes on to say that superior officials are themselves corrupt, and abet 
their subordinates, whom they treat as their agents. As a result: 
The local people have hated them all in the depth of their hearts for a long time. 
As soon as the crisis comes and the alarm is sounded, wicked elements take 
advantage of the situation [and start a revolt]. Vagabonds join them; and ignorant, 
destitute people, after having been subjected to some pressure [applied by the 
rebels] and having found it impossible to get food, follow them gladly. The 
conditions for a great upheaval are thus brought to completion.229 
Of course, famine and economic depression were also factors facilitating the Red 
Turban development. The worst floods along the Pearl River during the years of 1852 and 
1853, for instance, compelled many helpless people to tum to Tiandihui groups.230 The 
inflation of silver, the devaluation of the coinage, and the rapid increase of prices for 
daily commodities affected the Pearl River Delta areas. During the widespread famine of 
1853 and 1854, rural insurgents led by local Tiandihui lodges simply overran the countrys 
ide.231 
229 Guangzhou FZ, 129.24b, quoted by Hsiao, Rural China, pp. 469-70. 
230 ln the summer of 1853, Guangdong had severe floods, ruining many crops and eventually leading much 
of rural population to starvation. See Daqing lichao shilu, Xianfengchao, 67 .19 Imperial Edict (Sept. 11 , 
1852); Idem., 104.25. Imperial Edict (Sep. 26, 1853); F.O. 931 .1319 Report on forts damaged by severe 
weather (1852). 
231 Records in local gazetteers during the 1850s give enough evidence to show that climatic calamities fell 
120 
2) The Nature of the Rebellions 
The Composition of the Red Turban Rebels 
What social groups or classes then dominated the Red Turban forces? It is 
difficult to be precise about the social composition of Red Turban followers. One source 
dismisses them as a group of vagrants, monks, sorcerers, and artisans. 232 Others state that 
rebel groups assembled several tens of thousands of poor people, presumably peasants, to 
oppose tax abuse.233 The leadership of the Red Turban rebel groups is certainly of critical 
importance in any assessment of their character. Among the rebel leaders, there were 
fortune-tellers, bean-cake sellers, carpenters, sailors, actors, barbers, blacksmiths, rice 
pounders, miners, medicine men, merchants who dealt in illegal trades such as smuggled 
salt and opium, hired laborers, robbers, pirates, pawnshop owners, and members of the 
lower intellectual class such as clerks and graduates of public examinations; to these 
many more might be added, such as petty officers of the courts, military deserters, 
runners, peddlers, and smugglers. All these groups were not distinctly separate, but 
upon almost the entire area of the Pearl River Delta. See Guangzhou FZ, I 63.35b-36b; Huizhou FZ, 
18.21b; Deqing Zhouzhi, 15.29a; Xinning XZ, 14.18a; Wakeman, Strangers, 136. 
232 Sasaki Masaya, "Kampo yonen Kanton Tenchikai no haran roX:~~if-1':l!:::R:i'l!r~O) ,&:~L," in Kindai 
Chugoku Kenkyu Senta iho j[r~i:p~lff~t /'~ ~ffi, 2.3 (1963). 
233 Maeda Katsutaro AA EB Mf~~~. "Shindai no Kanton ni okeru nomin toso no kiban n!f fl.:;Jl:l!: 1:: Hit Q 
Jl~4* C7J ~ffit," in Toyo gekuho l!:7'$~ffi, 51 (March 1.969), 1-38. 
121 
overlapped each other. 234 
Chesneaux identified these people as semi-proletarian intellectuals or 
impoverished semi-proletarian vagrants who had fled from their home villages.235 The 
principal Red Turban leader, Li Wenmao, was a wandering actor; and many other Red 
Turban leaders were semi-proletarian vagrants.236 Some of these vagrants were drawn to 
the search for a new political authority because of many setbacks that they had 
encountered after leaving their native places; others embraced a new political vision 
because their social or occupational positions allowed them to see the degenerate reality 
of the ruling state. These people inhabited a world very different from that of the peasants 
involved in local rent and tax resistance, for even the poorest tenant was tied to his land 
and even the propertyless cultivator clung to the plot he rented in order to go on living. 
The objective conditions that brought them to the leadership of political rebellions were 
supplied by the great mass of desperate marginalized vagrants who were appearing in 
greater numbers in nineteenth century. 
There is little doubt that the majority of Red Turban groups were led by members 
234 Ibid; Lu, Liangguang de Tiandihui, 133-142. 
235 Chesneaux, Peasant Revolts, 16. 
236 See Table 2, Occupation of the major Red Turban leaders in Guangdong, 1854-1855. 
122 
of the local secret societies. Equally indisputable is the fact that poor people constituted 
the majority of the rank and file Red Turban members. Wang Tianjiang's study reveals 
that the bulk of secret societies in the nineteenth century were composed of six main 
types of social persons: dispossessed peasants, artisans, small traders, small owner-
mangers of various land and water transport vehicles, laborers or porters, and disbanded 
soldiers of the Taiping and imperial forces. 237 
According to Chesneaux, smugglers, sorcerers, geomancers, victims of floods, 
droughts and other disasters were also numerous in secret societies.238 The features 
common to all these categories of people were of course their relative poverty and 
mobility. They came to be known as "the drifting population (youmin ¥}ff-~)." In brief, 
members of secret societies tended to come directly from those classes of society 
conspicuous for their poverty, their despised occupations and their inferior status, which 
also implied an absence of kinship ties or residence. 
237 Wang Tianjiang :;R~, "l 9tt!:tc.T-'¥-1¥J i:p 00:1:ik~~:t±~ I 9shiji xiaban de zhongguo di mimi shehui," 
lishi yanjiu JJj ~Wf~ (Beijing, 1963) no. 2, p. 83 . 
238 Chesneaux, Secret Societies, 6-20; Idem., Peasant Revolts, 70-72. 
123 
Table 2. Occupation of the Major Red Turban Leaders in Guangdong, 1854-1855239 
Occupation Names Number 
Artisan Zhu Hongying * #t~, Chen Jingang ~~~ij 2 
Bandit Chen Songnian ~tl::¥, Gao Mingyuan r@j ~;l!, Che 6 
n Liangkang ~~~~. Fan Yayin ffi:§ if, Chen Zhigu 
ang ~;t )ji, Liang Peiyou ~ :1-: tt~ 
Boxer Chen Kai, Di Huagu W!k~, Huang Zhenshan ~fl 5 
W, Zhang Jiaxiang ~:g:ff, 
Yao Xinchang ~~~ ~ 
Fortune Teller Gu Shengyang tr!!im 1 
Low Degree Holder Wu Lingyun !R:~~ ' Zhou Yingnian ~ 1J<1f. 2 
Monk Heshang Neng to ri!6 at Seng Liang f'ltf ~ 2 
Not Known Yin Changying ;}3"~~' Zou Xinmao i ~ffia 4 
Deng Xingque ~J!!f, Liang Xianfu ~If~! 
Servant Su He ~)l, Lu Cuijin 8~-ff 2 
Soldier Guan Ju IUJj§, He Mingke fpJ ~ t4, Huang Jinliang 7 
~~~. He Wan Wat, 
Zhang Gaoyou ~r@j~, Huang Dingfeng ~W~JI., 
Feng Liu <,~h 
Strolling Player Li Wenmao * ""!I:.a 1 
Tough Tanya Erman ~§7G¥ffi, Li Shibao*:E* 2 
This picture of the social composition of secret societies largely holds true for the 
composition of the Red Turban rebels. For these drifting populations, most of whom were 
likely to be men of an economically low status, secret societies were sometimes explicitly 
political tools, to be wielded against the local authorities or even the dynasty, either for 
reform or -more likely-rebellion. Table 2 presents information on the occupational 
239 Source: Lu, Liangguang de Tiandihui, 134-142. 
124 
background of thirty-four major rebel leaders. Over eighty percent of these leaders were 
from the rootless sector--disbanded soldiers, strolling players, local toughs, vagrants, 
etc.240 
Many rebel followers were famine refugees, another fruitful source of rebel 
recruits when they failed to find work at the end of their wanderings. What most 
characterized these rebels was their landlessness. The social and psychological 
implications of this fact were profound, for possession or non-possession of land was a 
basic criterion of peasant society. The landless were accordingly the "out" class in much 
of rural China, forced to live on the outskirts of the village, and subject to various kinds 
of discrimination.241 Unable to accumulate wealth, the accepted avenues of social 
advancement closed, people like these had nothing to lose and everything to gain from 
confrontation. The decision to become a secret society member, soldier, bandit, or a rebel 
was less agonizing than usual since there was little alternative. 
Besides their generally impoverished conditions, the other characteristics of the 
240 See Table 2. See also Davis, Primitive Revolutionaries, p. 93; she includes the following list of39 
Guangzhou Tiandihui leaders which was confiscated by officials in 1855: Number of Salaried Workers -
10, of Small Traders [Peddlers?] - 8, of Farmers - 6, of Fishermen - 5, of Artisans - 5, of Mendicants -2, 
of Smugglers - 1, of Minor government employees -1, of Gentry - I, and ofTotal-39. 
241 Martin C. Yang, A Chinese Village, Taitou, Shantung Province (New York: Columbia University Press, 
1945). 
125 
rebels were their chronic underemployment and high geographic mobility.242 As 
examined in Chapter 2, Sino-British tensions as well as the shifting of the trade routes 
that caused the decline of exports in Guangzhou had a very significant effect upon the 
economic situation of the Pearl River Delta regions. This meant the unemployment of 
thousands of people involved in foreign trade, especially boatmen and porters. The 
following passage reveals the desperation that drove these people: 
At present the bandits along the West and North Rivers have not been pacified. 
Yet trade between us and the barbarians has long been stagnant. The boatmen 
and the porters, being poverty-stricken people, when they are once unemployed 
do not have other means of making a living, and are forced to join the bandits.243 
Had they been financially able in normal times to save enough to take care of 
themselves in bad years, their financial condition would not have been so serious. As a 
matter of fact, however, they earned just enough to maintain subsistence. The question of 
hunger probes the important issues of poverty, misery, and marginalization. 
242 Zhuang Jifa found that most of the men who joined the secret societies in the nineteenth century were 
poor people who had difficulties in earning an adequate living. Most secret society members were men who 
did not hold steady employment, whose work was very precarious and unpredictable, and whose families 
were landless. See, Zhuang Jifa, Qingdai mimi huitang, 1793-1795. 
243 Xianfengchao chou ban yi wu shi mo J,jUf.lm tM~r!Ml* 40 juan (Beijing, 1928 reprint), 19.8b. 
126 
Table 3. Place of Origin of the Rebels Captured in Qujiang County.244 
Place of Origin Number 
Qujiangxian BBtL~ 10 
Huaxian fEM 8 
Nanhaixian JW#ij:M 6 
Yingdexian ~t!M 1 
Qingyuanxian ~~M 2 
Dabuxian -Jc!mM 1 
Renhuaxian 1=:.tEM 1 
Nanxiongzhou 1:W1lt3'M 1 
Jiayingzhou ~ l!3'M 1 
Shixingxian ~t Jl!M 2 
Hepingxian l□ 1¥ M 1 
Nankangxian 1:W~M (Jiangxi IT1ffi) 1 
Changningxian ~$M OiangxiIT1ffi) 1 
Ganxian ftM (JiangxiIT1ffi) 1 
Guiyangxian :It ~M ( Hunan ¥iiJJ 1W) 1 
Yizhangxian '.§:$M ( Hunan1MW'J ) 1 
The lack of adequate employment opportunities accounts in a large measure for 
the high geographic mobility of these men. Table 3 shows us that in a group of thirty-nine 
convicted rebels, twenty-nine men had joined a rebel group outside their native counties, 
and the rest had joined within their native county. These were exactly the sorts of 
people-hired workers, porters, peddlers, boatmen, itinerant actors, fortune tellers, etc.-
who were constantly traveling up and down the rivers and roads, and who were the most 
244 Source: F.O. 931.1753 List of 39 Rebels (n.d.). 
127 
apt to seek the help and protection of brotherhoods while away from their homes.245 
The Red Turbans had no apparent restrictions on recruiting their forces. The rebel 
participants ranged widely in age and even several women were allowed to join the rebel 
forces. 246 As presented in Table 4, the youngest were in their early twenties and the oldest 
in their late forties. However, most members were adult males aged between twenty and 
thirty-two years old. The government report of the captive of twenty rebel leaders in 
Hongbing Qiyi Shiliao confirms that most of rebel followers were poor and marginalized 
immigrants who did not have local kinship ties: nine had no living parents and siblings, 
twelve had never married, and only three had living parents and siblings.247 Lacking 
family or village ties, linked only to regional markets or even the road itself, their 
rootlessness made rebellion a relatively easy decision when an economic downturn 
destroyed the system's cohesion. 
245 Simply because these rebels did not belong to any settled community or family, they were most likely 
viewed with suspicion and fear by members of settled communities. So, in addition to lacking the 
protection and economic ties provided families and attachment to a native place, these men were often 
ostracized or treated with some hostility by resident of settled communities. See Table 3. 
246 One of powerful rebel leader, Di Hua'gu ~X¾li is a good example offemale rebel participant. See Wu 
Bingheng :!R*~, "Xinfeng shouchengji shilue nf ft."1':9iX!c.JJ~," in Hongbing Qiyi, 1126-27. 
247 Hongbing Qiyi, 95-129. 
128 
Table 4. Age Distribution of Red Turban Rebels Captured in Yingde County.248 
Age No. Age No. 
23 2 24 3 
25 1 27 5 
28 3 32 11 
33 1 34 1 
35 2 36 1 
37 2 38 1 
40 1 42 1 
43 1 45 1 
47 1 Total Number 39 
The Red Turban method of recruiting seemed to follow the tradition of the 
Tiandihui, both attracting volunteers and compelling others to join.249 Numerous accounts 
in government sources tell us that the Red Turbans used generous distributions of money 
and food to the poor iri order to enlist new members. After sacking government treasuries 
and private pawnshops, the Red Turbans distributed part of the booty to the people to win 
their hearts.25° For example, the Red Turban leader Chen Kai wasted very little time once 
he had taken Foshan in proclaiming his promise of good treatment to population, officials, 
248 Source: F.O. 931.1729 List of 35 Rebels (n.d.) 
249 Members were recruited to the Tiandihui in a number of ways: by voluntary enlistment or persuasion, or, 
failing this, sometimes by blackmail, physical threat or kidnapping. See Schlegel, Thian Ti Hwui, pp. 57-8. 
250 Xiangshan XZ, 22.58; Shunde XZ, 23.6. 
129 
and officers if they offered no resistance. 251 Despite his connection with the Tiandihui, 
Chen relied largely upon secular slogans to rally his followers. His call for an end to 
extraordinary tax levies was a far more potent source of support than his Ming 
restorationism.2 52 Many families were in arrears, and dissatisfaction was widespread. In 
his proclamations, Chen railed against the harsh tax burden and other abuses perpetrated 
by the government. He abolished surcharges and remitted the land tax for three years, 
after which it was to be collected at the basic rate. One source states that he acquired 
251 F.O. 931.1476 A Proclamation by Chen Kai (Nov. 29 1854). 
252 Latent traditions ofrebellion contrary to Confucianism but instead bound to an egalitarian set of mores 
re-emerged at such times. Red Turbans marched under colorful banners inscribed Taiping :;{l;;.Ijl (Great 
Peace) and Pingjun .Ijl:f$;J (Equality). The Tiandihui lodges most often perceived a blissful past that had 
been overtaken by the current strife and oppression. Each era of agrarian crisis, these dreams of a golden 
ancient past revived and the lodges and even the sects took them up. Rituals of fellowship were 
appropriated from the popular epics The Romance oft he Three Kingdoms and The Water Margin; concepts 
of redistribution and equalization of land-holdings were derived from the probably apocryphal ancient 
system called 'well-field' ; and even the mode ofrebel dress and the character of military formations were 
borrowed from ancient uprisings. At least in their ideals, the Red Turbans, as Tiandihui brotherhoods, often 
shared this egalitarianism. And in some cases, the rebels distributed goods to the poor and carried banners 
reading 'kill the officials, kill the rich, spare the poor.' To this egalitarian emphasis were added what look to 
be Ming restorationist slogans as well as symbols of legitimacy derived from a variety of tradition. At first 
glance, the restorationism seems obvious from the explicit mention of the Ming on the banners used by the 
rebels and from the selec
(Darning douyuanshuai *tio n of title ofrebel leaders as "Generalissimo of the Great Ming Dynasty a)Hf~5Ggrb)." Given the historical connection to the pro-Ming stance, one can 
assume that the Red Turban Rebellions expressed a simple desire to oust the Qing and return to Ming rule. 
However, when we examine more closely the rebel forces, it is more difficult to determine the nature of the 
ideology of the rebels. With the exception of the banners, there is no explicit expression of pro-Ming 
sentiment in rebel confessions or other accounts of the mass action. For egalitarian slogans and policy of 
the rebels, see F.O. 931.1440 A Red Turban Proclamation; F.O. 931 .1466 Proclamation of Chen Kai (Aug. 
27, 1854); F.O. 931.1477; "Tao Daqing xiwen g-;f7'.inffl: X," in Hongbing Qiyi, 38-39; "Oath Taken by 
Members of the Triad Society, and Notices of Its Origin," The Chinese Repository, Vol. Xviii, no. 6 (June, 
1849), Article 7 .. For the tenns related to Ming restorationism, see F.O. 931.1531 Proclamation by 
. "*C ommander in Chief for the Restoration of Ming," (1855); "Darning Douyuanshuai xiwen ajj 1f~5Ggrbffl:X," in Hongbing Qiyi, 50-51; "Fuming zongbing dayuanshuai hong gaoshi, ~ l)JHt~ :X7G 
y1jJ#ti§- ~ ," in Hongbing Qiyi, 61. 
130 
several thousand followers in this way.253 
More often, however, the Red Turban rebels forced young men to fight for them. 
According to a local gazetteer, Foshan Zhongyi Xiangzhi, "Wherever the Red Turban 
bandits went, they set houses on fire and forced men into their service."254 The following 
document dates from a somewhat later period but recounts well how outlaw bands 
developed into secret societies in the late nineteenth century: 
Hui-fei (society bandits) have been most rampant in Szechwan, Hunan, 
Kweichow, K wangsi and K wangtung. At first they were lawless wandering 
people who burned incense and organized societies. When their organizations 
have waxed strong and their members become numerous, they rely on their 
strength to tyrannize their neighborhoods and victimize the good people. The 
humble people, being helpless, may join their societies for self-protection. Such 
societies assume diverse names, for example, Ke-lao (Elder Brother), An-ch 'in 
(Contentment-affection), T'ien-ti (Heaven-earth), and San-tien (Three-dots). 
They operate in every chou and hsien. Each local unit numbers several dozen or 
several hundred men. All the units keep in touch one with another.255 
During periods of widespread uprisings and social unrest, as was the case in 
1854, whole communities joined in the disturbances. At such times rebel forces were 
253 X unzhou FZ 2 7.4-5. 
254 Foshan Zhongyi Xiangzhi {~Ill ,'it',~~*• 11.14. 
255 Quoted in Hsiao, Rural China, 460--a memorial submitted to the Emperor by the Governor of 
Guangdong in 1899. 
131 
flooded with new recruits made up of the normally settled peasant farmers, and not just 
the rootless poor. Much of the initial Red Turban success was due to sudden switches of 
allegiance by the local people. Whole villages and blocs of villages might hoist the red 
flag and temporarily join the insurrectionist camp.256 Some might have done so out of 
patriotic Chinese anti-Manchu sentiments, but the majority were primarily concerned 
with their own security or motivated by the thought of gain. Depending on the specific 
context, joining the insurrectionists might afford advantages in the conduct of inter-clan 
or inter-village relations.257 Similarly, the non-joiners were not all motivated by reasons 
of loyalty to the Manchus or to the established order; they might simply be acting in 
accordance with their own interests.258 
While the Red Turbans seem certainly to have come largely from the humbler 
segments of society, rebels also managed to recruit several members of the local elite. 
The principal Red Turban leaders, men like He Liu and Li Wenmao, were illiterate, and 
many other Red Turban leaders were also uneducated.259 But, though bravery was 
256 Xiangshan.XZ, 15.30-40; Huizhou FZ, 18.25; Guangzhou FZ, 82.8.4. 
257 Wakeman, "Secret Societies," 34-35. 
258 Ibid. ; "Shundexian huifeilan duanyuanyou ~"*'~~~Li4'M~EB." in Sasaki, Shinmatsu no himitsu 
kessha .Shiry8 hen, 51; Xu.xiu nanhai XZ ffl{~ffi¥tlJ"*'i't, 21.4. 
259 Dongguan.XZ, 35.4;Xu.xiu nanhai.XZ, 26.13, 18.14; Panyu.XZ, 22.27. 
132 
essential for fighting, brains were needed for planning. This need was filled by a number 
of"half-baked" scholars, some of whom had received the first degree through the civil 
service exarnination.260 
These political advisers transformed Tiandihui leaders into rebel kings. A notable 
example of the work of these scholars was Chen Kai's proclamation as Zhennanwang ~ 
1¥J.:E (Guardian King of South), which was planned by a licentiate (xiucai ':f§::t"), Lii 
Zigui 8 -=fti. He helped Chen perform the imperial ceremony of sacrificing to heaven 
and earth; together they drew up regulations and formed a military system. Chen Kai also 
rewarded his chief lieutenants with titles and offices, issued proclamations, stepped up his 
recruitment efforts and finally created his own kingdom, Dachengguo, in 1855, all acts 
presumably recommended by Lii or other educated advisers.261 The military adviser for 
Chen Kai 's rebel kingdom was also a stipendiary shengyuan :i:ln, Chen Dingxun ~~§~ 
l/] _262 
Theoretically these intellectuals should have been loyal to the imperial dynasty, 
260 There are some instances of lower degree holders working with the rebels . See Jian, Quanshi, vol. 2, 
824; Nanhai XZ, 15.12a; F.O. 931 .1500. A note on gentry rebel who had led rebels in an attack on Xinhui 
(1850s). 
26 1 "Lil Zigui gongci 8r:ti#Uii!," in Hongbing Qiyi, 108-110; Xu, Lingnan, 236. 
262 Yu/in zhouzhi W#fl'l;i\;,, 18.60. 
133 
but in fact, a small number of them broke the moral bonds that had been weakening since 
the middle of the nineteenth century. At that time the weakness and corruption of the 
Manchu empire was being uncovered by foreigners and by the Taipings. Some alert and 
educated men began to wonder at the subtle distinctions between obedience and rebellion. 
Certainly some local elites participated in the reb_ellions, but the image of rebels 
mainly composed of a ' drifting population (youmin) ' still emerges as the dominant one. 
However, merely depicting the social background of the rebel participants is not 
sufficient. A discussion of the groups according to their ethnicity would give us a much 
clearer understanding of the significance of the Red Turban Rebellions. The Qing 
archival records used in this study shed much light on the identities of rebel participants. 
The ethnic composition of the Red Turban rebels was no simple, static matter, but rather 
tempered by regional variation.263 
As we have previously seen in the case of the Increase Brothers Society in Boluo 
county, feuding in Guangdong also created rebels. Often the local Tiandihui lodges took 
263 The types of Red Turban groups also varied along with differences in cropping and landholding patterns. 
For instance, Northeastern Guangdong, with a less productive agriculture based upon wheat , maize, and 
beans, was the home of more egalitarian Red Turban units in which the leaders-often peasants 
themselves- were treated as brothers. The central region saw the development of Red Turban units 
comprised of demobilized soldiers, brigands, and local bullies. On the other hand, the Wuyi regions of the 
Central Guangdong, known for high rates of tenancy without sufficient lands, saw Punti tenants dominate 
Red Turban units known as a tang. See Yoo, Kundae, 258-275. 
134 
advantage of ongoing feuds to enhance their position. One of the most enduring of such 
feuds was the ethnic struggle between the Puntis and the Hakkas in the Wuyi regions.264 
Although the Puntis alleged that their Hakka neighbors were Red Turban rebels, 265 
archival records from Central Guangdong show, on the contrary, that the Hakkas were 
notably absent in the lists of rebel-participants.266 There were certainly some Hakkas 
among the Red Turbans but most of Hakka rebel participants were from East 
Guangdong.267 In view of what happened later, there seems no doubt that, in the districts 
to the southwest of Guangzhou, the Wuyi regions, most of the rebels were Puntis.268 
The Tiandihui members were more numerous in the Wuyi region during the mid-
nineteenth century, in part because of the mass emigrations of the early nineteenth 
264 The Wuyi region indicates five counties adjacent to Guangzhou and lying to the southwest of 
the Pearl River delta, a region that has sent thousands of its residents to North America. The 
inhabitants of these five counties (Xinning, Kai ping, Enping, Xinhui , and Heshan), speak the 
same dialect and believe that they share the same subculture. 
265 This accusation was brought by the Punti tenants during their struggles with Hakka tenants in Xinning. 
SeeXinningXZ 14.34b. 
266 According to Lu Baoshan's research, there was only one Hakka rebel'leader among twenty-eight major 
Red Turban rebel leaders. See Luo, Lun wanqing, 133-142. 
267 A band ofTiandihui rebels who attacked Kowloon on 19 August 1854 were reported to have been nearly 
all Hakka stonecutters from Xin ' an county. See C.O. 129.47. Governor of Hong Kong to Sir George Grey 
(Aug. 21 , 1854). 
268 "Enp ing Tuke hudou yanyo~ ,l~,\t f ±~ 1[ J;J,~ El3 ," in Qin, Tiandihui, 243-244; "Huitang Xiao xi 
-frlirlt@-," Jin Yudi ~frtj!& ed., Taiping Tianguo shiliao *ifl-Rli9:f4 (Beijing: Zhonghua shuju, 1953), 
497-499. 
135 
century, when millions fled there from starvation. The new settlers provided eager 
recruits for the ·protection that Tiandihui membership offered, and some even utilized 
their Tiandihui affiliations to launch attacks against their long-standing neighbors, the 
Hakkas. The Hakkas often responded by forming militias aiding Qing forces, to secure 
greater organizational strength against the newcomers. 269 
Red Turban efforts to recruit Wuyi people increased in 1854 and continued in the 
period immediately following the first Tiandihui attack on Guangzhou. However, 
interethnic hostility proved to be more powerful than either the Tiandihui "cause" or 
general popular animosity toward the Qing state. In late 1854, ethnic strife pushed the 
Punti tenants and vagabonds in Heshan county into open rebellion against the Manchus 
under the guidance of Liang Pei you. Although local details of rebel organization in 
Heshan at the time are scarce, sources clearly indicated that Punti resentment of the 
Hakkas was a subsidiary cause of the Red Turban Rebellions.270 
But soon, this ethnic division brought the failure of the rebellions in the Wuyi. 
Before the arrival of Qing forces, the rebels' main opponents were ethnically organized 
269 Enping XZ, 14. 7b; Guangzhou FZ, 82.1 Ob, Kaiping XZ, 21.3a. 
270 See China Mail 22 May 1856; IO May 1855. 
136 
bodies of Hakka militias. Local Tiandihui leaders, for instance, those leading the Red 
Turbans in Xinning, could not have raised forces district-wide without the mobilizing 
power of ethnic anger. On the other hand, the Qing fight against the Red Turban rebels 
would have been even more protracted without the aid of the Hakka militia. The Red 
Turbans' efforts to forge a pan-ethnic alliance failed, and much of the violence of the 
rebellions pitted the Punti rebels against the Hakka militias. In this sense, ethnicity was 
central to the failure of the rebellions.271 
As we have stated before, a great variety of groups became rebels. Among these 
groups were many whose interests conflicted. It is clear from the above discussion of the 
Punti relations with the Hakka in the Wuyi regions that, at least in these regions, the great 
driving force that welded the rebels together, despite conflicts among themselves, was the 
dialect, Cantonese, shared by all the groups joining the rebels. In other words, through the 
Red Turban Rebellions, there developed a kind of tradition of violence that, in areas 
particularly vulnerable to socio-economic competition like the Wuyi region, could easily 
explode into conflict drawn on subethnic lines, as exacerbated by ideological difference 
( or the different positions of rebel/loyalist that developed with the Red Turbans) and also 
27 1 Ibid. 
137 
by the actions of the state. By focusing on the Red Turban Rebellions in the Wuyi region, 
I will trace the development of subethnic conflict, explaining the outbreak of the Hakka-
Punti War. 
Taiping Connection 
Another opportunity for the development of the Red Turbans was the Tai ping 
rebellion. There is no doubt that the advent of the Taipings had a considerable effect on 
most of China. It encouraged the spread of rebellion even among those who knew little or 
nothing about its ideology. It certainly did so in Guangdong. There were a number of 
observable similarities between the Taipings and the Red Turbans and some Taiping 
leaders gave the Red Turbans a ready welcome as recruits. Both shared a hatred for the 
Manchus, went unshaven, wore clothing after the style of the late Ming dynasty, and 
dressed their heads with red turbans. Some Red Turban rebel groups tried to imitate the 
Taipings in their imperial ambitions. For instance, Chen Kai adopted the presumptuous 
title of king and flags that similar to the Tai ping flags. The uniforms of the Red Turbans 
were largely imitations in color and style of the Taipings. After occupying Huaxian, Chen 
Kai even put the characters on his flag, "Heavenly King of the Tai ping Kingdom -Jc 2P-* 
138 
~lE-€ir.::R3:." Why did these Red Turbans do such things? It was probably because the 
Taipings had greater power and prestige than did the Red Turban organization; the latter 
tried to copy the Tai pings' appearance in order to overcome their timid governmental 
opponents. Another motive for imitation may have stemmed from the good reputation 
and propaganda of the Tai ping expedition, which inspired numerous rebels and affected 
the reshaping of the Red Turban organization from small to larger units. 272 
Various facts also demonstrate substantial Taiping influence over the Red Turban 
Rebellions. Both groups caused the Manchu dynasty a great deal of trouble and both dealt 
it near-fatal blows. Both were able fighters in the eyes of the imperial government, and 
they were the products of the same general political and social conditions. In Guangdong 
and Guangxi, beset by a wide variety of natural disasters and political conflicts, subethnic 
alignments were a natural organizational device for disparate speech groups. This was in 
fact the case that the alignments on the basis of subethnicity could result in ever 
expanding aggregates of people being pitted against each other. And this is what 
272 See Luo Baoshan ~'.i:~, "Guangdong Hongbing qiyi luelun 1*~~~.:X.~ii?:-," in Lingnan Wenshi 
ilitWJX~ (Guangzhou: Guangdongsheng wenshi yanjiuguan, 1983), 87 ; Zhong Zhenwei fil[ f:ttl ," Chen 
Kai, Li Wenmao lingdaode hongjinjun qiyi Wf-:ff, 2}:JC"it-@i~ a{_J tr rjJ '.¥~.:X.," in Zhongxue lishijiaoxue 
r:r?¥:in~~?¥:9 (Guangzhou: Guangdong renmin chubanshe, 1958), 17; Cai Xiaoqing, "Lun 
Taipingtianguo yu Tiandihui guanxi ~j;;:f~~~~±t!nwlfflfj," Zhongguojindai huidangshi yanjiu 
r:r~ili1~ i'"itj:m3t, 158-111. 
139 
happened in Guangdong and Guangxi. The organizational means for spreading the 
conflict had been provided by Hakka gentry and their militia units and by Tiandihui 
differentiated on the basis of dialect.273 
Certainly the Hakka-Punti conflict was a relevant factor for both rebellions. Yet 
the differences between the Taipings and the Red Turban rebels were manifold. 
Interestingly, the key difference between the Red Turbans and the Taipings lies also in the 
way subethnicity was linked to other patterns of social differentiation. For instance, the 
Hakka settlements in the Wuyi, Guangdong, were older and more stable than those in 
Guangxi, and this had contributed to the development of a more influential Hakka elite 
group. Unlike the Hakkas in Guangxi, therefore, the Wuyi Hakkas were called upon to 
assist local Qing force against the Red Turban rebels. In addition, the Red Turbans' 
political and military institutions were less highly organized than those of the Tai pings. 
273 Philip Kuhn argues that the religious notions of the God-worshippers provided Hakka ethnicity in the 
conflicted frontier environment of Guangxi with a new set of concepts for its expression (just as it 
demanded a new organizational framework-the God-worshipping Society). However, Kuhn's position is 
ultimately unsatisfactory. It proceeds from an equation of prior events in Guangdong province in which the 
Hakkas and the Puntis engaged in long-term quarrels forming a backdrop to the onset of the Taiping 
Rebellion, to argue that these same factors were operative in Guangxi. But Guangxi was a much more 
complex social context in that there was present a third major group, the Zhuang. The Zhuang, a ethnic 
minority, were more numerous than the Hakkas. Any explanation which analyzes ethnic tensions and yet 
fails to embrace the diverse ways in which the Zhuang participated in the rebellion cannot be fully 
satisfactory. See Philip A. Kuhn, "The Tai ping Rebellion." in The Cambridge History ofC hina, Volume l 0, 
late Ch 'ing, 1800-19 JJ , Part I. ed. John K. Fairbank (Cambridge: Cambridge University Press, 1978); 
idem, "Origins of the Taiping Vision: Cross-Cultural Dimensions ofa Chinese Rebellion," Comparative 
Studies in Society and History, Volume 19, (l 977), 365; Huang Xianfan ~fJ\IJI, Zhuangzu tongshi fr~ 
:im..t.. (Nanning: Guangxi minzu chubanshe, 1988). 
140 
The Taipings also had a clear political, economic, social, and religious ideology bridging 
ethnic and localistic loyalties.274 The Taiping message went beyond simple economic or 
subethnic antagonism. The Taipings distinguished not between the rich and the poor or 
the Ha.kkas and the Puntis, although they paid some attentions to this distinction, but 
fundamentally between the group of believers and the non-believers.275 This religious 
ideology eventually permitted the Taipings to appeal to more groups than could the Red 
Turbans, whose main members were the Puntis. The Taipings were something more than 
the Red Turbans and their more complex message permitted them to appeal to various 
groups including intellectuals, peasants, and various ethnic groups. 
The Taipings also exercised political and military control over a definite territory, 
but after 1856, the Red Turbans simply sought food and safety. Although later some Red 
Turban forces did become Tai ping allies in the more desperate stages of the fighting 
against the imperialists, the alliance was born of military necessity and never constituted 
an organic union. As Hong Xiuquan himself recognized, an alliance or partnership 
274 For good summaries on this, see Vincent Y. C. Shih, The Taiping ideology; its sources, interpretations, 
and influences, (Seattle: University of Washington Press, 1967); Teng Ssu-yu, New Light on the History of 
Taiping Rebellion, (New York, Russell & Russell, 1966); Jen Yu-wen, The Taiping Revolutionary 
Movement, (New Haven, Yale University Press, 1973). 
275Shih, The Taiping Ideology, xiii. 
141 
between the two organizations was rendered impossible by fundamental differences in 
political objectives, religious beliefs, and moral ideals and practices: 
Though I never entered the Triad Society, I have often heard it said that their 
object is to subvert the Tsing and restore the Ming dynasty. Such an expression 
was very proper in the time of Khang-hi (K'ang-hsi), when this society was at 
first formed, but now after the lapse of two hundred years, we may still speak of 
subverting the Tsing, but we cannot properly speak of restoring the Ming. At all 
events, when our native mountains and rivers [the Empire] are recovered, a new 
dynasty must be established. How could we at present arouse the energies of men 
by speaking of restoring the Ming dynasty? There are several evil practices 
connected with the Triad Society, which I detest: if any new member enter the 
society, he must worship the devil, and utter thirty-six oaths; as sword is placed 
upon his neck, and he is forced to contribute money for the use of the society. 
Their real object has now turned very mean and unworthy. If we preach the true 
doctrine [Christianity], and rely upon the powerful help of God, a few ofus will 
equal a multitude of others. I do not even think that Sun-pin, Woo-khi, kung-
ming, and others famous in history for their military skills and tactics, are 
deserving much estimation, how much less these bands of the Triad Society.276 
Theodore Hamberg described the story of the eight chiefs of the Tiandihui who 
wanted to join the Taiping forces with their bands. Hong Xiuquan demanded from them 
and their followers the acceptance of the Tai ping religious beliefs and code, which was to 
be taught to them by sixteen Tai ping tutors. When one of these tutors retained the money 
given by the bandit chieftain instead of handing it over to the common treasury, he was 
276 Theodore Hamberg, The Visions of Hung-siu-tshuen, and Origin oft he Kwang-si Insurrection, (Hong 
Kong, 1854), 55-56. 
142 
executed.277 This example gave the eight chiefs an uncomfortable feeling and they left the 
Taipings, complaining that these laws were too strict.278 
There is no doubt that splinter groups of Tai ping did come in contact with the 
Red Turban rebels.279 But the relationship was not always harmonious and was rarely 
close. The earliest contact between the Taipings and the Guangdong Tiandihui groups, we 
may remember, occurred in 1851, through Ling Shiba's revolt. Loyal to the Taiping cause, 
Ling Shiba fought from Guangxi to Xinyi, Guangdong, where his unit was mercilessly 
wiped out by Governor-general Xu Guangjin in July 1852.280After Ling's revolt, the 
Taipings apparently sent agencies to stir up a rebellion in Guangdong; however, no 
formal alliance with the Tiandihui groups was concluded. This was a time when the 
Taipings were at the height of their strength; cooperation with a motley group of local 
277 fbid. , pp. 54-55 . 
278 Immediately after the Taiping rose up, the local Tiandihui leaders Zhang Zhao * ,t1J and Tian Fang, who 
had migrated inland from the Pearl River Delta, were similarly at war with local society but saw for 
themselves a political future : they consider linking their forces with the Taipings and actually joined their 
encampment at Jintian for a short time. Unable to stomach the Taipings' strict discipline and stem religion, 
they departed. ln 1852 they sold their services to the Qing side for a time but soon returned to outlawry and 
were killed in 1853. See Zhou Yum in ftij~ ~ . "Zhongguo huidang wenti yanjiu shuping i=p OO~Jt fA1 JlliJf 
~iiW," in Zhongguo huidangshi yanjiu hui 't1 OO~Jt.'1:11Jf~~ ed., Huidangshi yanjiu ~;'t~1i}f~ 
(Shanghai: Xuelin Chubanshe, 1984), 340. 
279 Ibid. Some members of the Tiandihui, including Ling Shiba, Luo Dagang, and Zhang Zhao joined the 
Taipings at different times. 
280 X iny i XZ ffi1[~~. 8.4; Guo Tingyi W~t.J., Taiping tianguo shishi rizhi -;k 2P-:Rm9:. ~ B tt. (Taibei: 
Shangwu Printing Co. , 1976), p. 73 .. 
143 
secret societies in Guangdong probably seemed a liability to be avoided.281 Several years 
later the situation had changed dramatically. The Tai pings suffered from serious internal 
dissension and severe military losses. Inasmuch as the Red Turbans had also undergone a 
series of defeats, it was to the advantage of both groups to effect an alliance. 
3) The Demise of the Red Turbans 
Reasons for the Red Turban Failure 
In the districts around Guangzhou, the Red Turban successes continued 
unchecked until early 1855. After nearly seeing the city fall into rebel hands, the 
provincial authorities were saved less by their own efforts than by four factors that 
eventually brought about the defeat of the rebellion. The first was the failure of the rebels 
to coordinate their activities. One of the greatest weakness of the rebel forces in the 
campaigns against the Qing force was lack of harmony and cooperation among the rebel 
groups. The alliances of the various Red Turban groups were a normal part ofTiandihui 
strategy, but the unity thereby achieved was often circumscribed by the need of individual 
28 1 At this time, the Taipings rigidly rebuffed alliances with the ' impure' pro-Ming secret society lodges. A 
point well put in Kuhn's volume, Rebellion and its enemies in late imperial China, militarization and social 
structure, 1796-1864, (Cambridge: Harvard University Press, 1970). 
144 
leaders to retain their followers' loyalty. Although lesser leaders might recognize 
another's superior ability, their ultimate motivation was either survival or self-
advancement. 
Not surprisingly, a rebel groups' internal harmony varied according to its 
fluctuating fortunes. Although Chen Kai, the rebel supreme leader surrounded himself 
with other Tiandihui confederates right up to his death in 1861 at Xunzhou, the problem 
of maintaining authority over his sub-leaders had always been a thorny one for him. Chen 
Xianliang's experience provides useful examples. After failing to be a king of Dacheng 
kingdom, Chen bore a grudge against Chen Kai and returned to Guangdong without 
following Chen Kai's force in Guangxi.282 Li Wenmao was always on bad terms with 
Chen Kai. For instance, Li played safe to protect his own headquarters at Liuzhou fYDJ'i'I a 
nd refused to cooperate with Chen Kai to protect Xunzhou from the counter attack of 
Qing forces in 1861.283 These are only a few examples of the bad relations that prevailed 
among the rebel leaders fighting a common foe. The disagreements among rebel leaders 
282 Hu, Qingdai Hongmenshi, 320. 
283 Chen Kai, after establishing his rebel government, minted new money, "Hongfu tongbao ~~:ilii ~ ." 
The fact that Li Wenmao circulated his own money, "Pingjing shengbao f~/m~" while he was still in 
collaboration with Chen to fight against Qing force indicated the uneasy relations between Chen and Li as 
well. See "Chen Song shiyou ~tl•El3," in Hongbing Qiyi, 212-213; "Gufei zonglu J1U[U~,3K," in 
Hongbing Qiyi, 824-830. 
145 
naturally caused a lack of coordination in fighting. 284 During the protracted attack against 
Guangzhou, there were frequent disagreements and disorganization among the rebel 
groups near the front. These gave the Qing force the time to bring in more troops from 
other provinces.285 In 1861 , the governor-general ofLiangguang, taking advantage of the 
internal trouble of the rebels, summoned troops from Hunan to destroy the rebel front. 286 
On the basis of presently available material, one may conclude that although the Red 
Turbans did have a united force, it was not very solid or systematic. 
Another reason for the Red Turban military failure was that they did not have 
enough soldiers and weapons to enlarge their rebellion. Nor was the Red Turban army a 
professional one, being composed mainly of marginalized wandering people.287 Even 
though they claimed to have a huge following and the imperial government officials 
reported large rebel armies in order to gloss over their own defeats, the actual Red Turban 
combatants were not large in number. From beginning to end most records about rebel 
284 The authorities were equally aware of this fact, and frequently exploited it in attempts to put down the 
rebels. See Wang Xin .3::~ , Wangzhuang wugong weiji .3:)!±ftt:i;/J:l:!Hl, 13 .319 (1892). 
285 Foreigners observed that ' there seems no unity of purpose nor combined action among the different 
bands ofrebels.' See F.O. 17.218, Bowring-Clarendon, Desp. 235, (Dec. 23, 1854). 
286 Hu, Qingdai hongmenshi, 318-320. 
287 Zhou Yumin ,ffll~~. Zhongguo banghuishi i:p OOffl~!t'. (Shanghai: Shanghai renmin chubanshe, 1993), 
154. 
146 
numerical strength are exaggerated. The total number of the Red Turban rebels who 
participated in the first attack on Guangzhou reported at approximately 200,000 in 
1854.288 But one scholar estimated that the actual rebels numbered only around 40,000 in 
1856.289 This is a surprisingly small fighting force and reveals the inadequacy of the rebel 
army. 
Popular support for the Red Turbans began to disappear when they failed to enter 
Guangzhou. The failure to take the provincial capital must have had great psychological 
effect on the Red Turban leaders as well as on their potential followers; they had no more 
prestige and no bright future. They made several abortive attempts to attack Guangzhou. 
Instead of distributing food and valuables to attract followers, the Red Turban leaders had 
to force the peasants to join them by holding swords to their necks.290 Soon the rebels 
would have to risk alienating the local populace. Large-scale recruitment to join the 
Tiandihui seems to have stopped with the failure of the Red Turban attacks on 
Guangzhou in 1855.291 
288 Jian, Quanshi, 830-8. 
289 Zheng, "Dachengguo," 19. 
290 Guangzhou FZ 82.17 and 134.25. 
291 Rebels ' desperate resort to plunder antagonized the general populace particularly after their failures on 
attacking of Guangzhou in 1855. See NanhaiXZ, 13.40a, 17.13a-15b; F.O. 17/215, Incl. I in Desp. 112 
147 
In addition, muddle-headed leadership generated political corruption among the 
rebels. The leaders of Red Turban forces were often short-sighted and were limited in 
education and ability. Incompetent leadership created internal trouble. Most rebel leaders 
were stubborn and narrow-minded and could not take advice from others. For instance, 
Su Qiumei iUX~, a military adviser for the Red Turban force at Poling {tit, advised 
Chen Kai to hire gentry member for the purpose of efficient rural control, but his 
proposal was never accepted.292 Li Wenmao was even not interested in holding an official 
examination in his rebel government to select more literate advisers.293 Often the rebel 
government alienated the literati by their unnecessary cruelty.294 Even though a few 
lower-level local elite member participated in the Red Turban forces, on the whole, the 
Red Turbans lacked men of ability to serve as staff members or military officers. Because 
of this problem, the Red Turbans failed to keep a unified military force with 
interchangeable military commands for a long time, nor did they achieve effective central 
control over appointments. 
(July 20, 1854); Wakeman, "Secret Societies," p. 42. 
292 See "Foling Tiandihui junji wenfang sishi tiaochen 1t~5'-::J.m w, '.ijL~::il::m i:i ~JH~~." in Sasaki, 
Shimatsu, 218. 
293 Jian, Quanshi vol. 2, 935 . 
294 Panyu.XZ, 212.218; Xuxiu nanhai XZ, 212.215 ; Foshan zhongyi xiangzhi, 11.14. 
148 
A lack of modem weapons and ammunition was another cause of the Red Turban 
failure. Qing forces were able to use powerful weapons supplied by foreigners. On the 
other hand, the Red Turbans used swords, lances, spears, tridents, matchlocks, and jingals. 
The Red Turbans were aware of their shortcomings in this respect and tried their best to 
acquire foreign guns, pistols, and even steam-ships. Unfortunately for them, the weapons 
they bought were ' foreign frauds. ' 295 
Foreign Interventions 
The second factor was the part played by foreigners in the suppression of the 
rebels. Officially the foreign powers, such as British and France, proclaimed a policy of 
' strict neutrality' from the beginning of the revolt.296 However, this policy wavered when 
the British residents became anxious about the fate of foreign property in Guangzhou. By 
early 1855, having taken Dongguan, the Red Turban forces under the leadership of Chen 
295 F.O. 17/352, 28 Mar. 1861; F.O. 931.1749 lntelligence Report (1854). 
296 The British policy of neutrality on China as formulated by Bonham was approved by Clarendon, 
followed by Bowring and others, and served as an official line in China for almost a decade. The British 
home government preferred neutrality; the aggressive Palmerston had left the Foreign Office in December 
1851, and his successor, Clarendon, stood for mediation. The rising power of the Second Empire under 
Napoleon III, and the deepening Crimean crisis, also necessitated a policy of non-commitment in the Far 
East. British diplomats in China liked such a policy because it gave them ample room for action. Bonham's 
strategy was to sit tight until hjgher Chinese officials asked for aid, so that greater demands could be made 
with greater assurance of acceptance. See F.O. 17 .200, no. 17, Hong Kong, IO Mar. 1853 ; F.O. 17 .218, 
Desp. 226, Incls., Robertson-Bowring, (Dec. 2 and 5, 1854); Wakeman, Strangers, 146. 
149 
Xianliang, who set up the rebel headquarters at Xinzao ffi~ (Blenheim Reach), had 
surrounded Guangzhou on four sides, and were ready for the assault on the city. From the 
east Lin Guanglong was to sail up the Pearl River with 145 junks; He Liu and his fleet 
were to attack from the south. The combined forces of Kan Xian and Zhou Chun 
including about 20 junks, were to attack from the west. Finally Li Wenmao was to attack 
from the north.297 At this point, problems arose between the Red Turbans and the 
foreigners, whose presence in Hong Kong was of major importance. The river 
communications were immediately interrupted by the insurgent fleets that extended ten 
miles along the Pearl River.298 To prevent a severance of communication between Hong 
Kong and Guangzhou, the British and American authorities agreed to grant permission 
for the British and the United States flags to be flown aboard.Chinese boats.299 This 
measure had two major effects: it secured the supply line of armaments to the Qing forces 
from the foreign merchants,300 and it inaugurated the Arrow incident in 1856.30 1 
297 Jian, Quanshi, pp. 859-60. 
298 F.O. 931.1086 A report giving a brief account of military action in north of Guangzhou (Jan. 1855); Jian, 
Quanshi, 844. 
299 W. C. Costin, Great Britain and China, 1833-1860, (Oxford, The Clarendon press, 1937), 178; Zheng 
Piexin 1~-{ij\,~, "Dachengguo de fanqing qiyi -J::.ft.lE B<J &mm:l'J..," in Shixueyuekan .'t~ J=J f iJ (Henan 
renmin chubanshe), Dec. 1958, 19. 
300 The Peninsula and Oriental Steam Navigation Company (P & 0), among others, had done a roaring 
trade by using steamers to tow "up the River to Canton, Chinese vessels and goods in the nature of supplies 
150 
The motive for abandoning the policy of neutrality was primarily the protection 
of British commercial interests including opium, the import of which was then 
increasing. 302 The other motive was the British desire to show a favorable response to 
Qing court's conciliatory policy. After signing the peace treaty with the Qing government, 
there was a clear realization that more commercial advantages would be forthcoming 
from a recognized government. The British were not willing to support the rebels. 303 The 
other motive was reaction against the rebels' attitude of superiority. The Red Turbans' 
first contact with foreign diplomatic and commercial representatives can1e in the year 
1854 when the British received Chen Xianliang's invitation to his camp. However, 
suspecting correctly that foreign ships were supplying armaments and grain to the Qing 
forces, Chen changed his attitude to the foreigners. He treated the British officers as 
and munitions of war." The rebels gave notice of a blockade of Guangzhou. See F.O. 17.226, Stirling to 
Bowring 12 Jan. 1855, enclosed in Bowring to Clarendon, Desp. 31, 15 Jan. 1855. Stirling was the British 
rear-admiral commanding the China station of the Royal Navy. 
30 1 Thus the present study confirms the fact that an important origin of the Arrow War may be found in the 
Red Turban Rebellions and the Taiping Rebellion, which simultaneously ravaged central and southern 
China. The Qing government was exhausted by these upheavals, thus presenting a golden opportunity for 
the British again to wage a war on China, a war they could expect to win easily. Such a consideration 
makes strategic sense, and may have swayed Parkes and Bowring when they reported the unrest to London. 
See F.O. 228.213 Desp. 157, Parkes to Bowring, (Oct. 13 1856). 
302 F.O. 17.358, Lt-Col. Neale's report, Peking, (Dec. 20 1861) 
303 Stanley Wright, The origin and development oft he Chinese Customs Service, 1843-1911 An historical 
outline. (Shanghai, 1939), 14 7. 
151 
tribute bearers, and wanted them to recognize him as the lord of all kingdoms, demanding 
that they stop selling armaments and grain to the Qing forces. Meanwhile, the rebels 
irritated the foreigners more and more by continuous bombarding and looting of foreign 
merchant ships. The foreigners were furious and began to lose their patience with the 
rebels. This ineptness in foreign relations arose naturally from the very narrow experience 
of the leadership that eventually became a fatal weakness for the Red Turban rebels.304 
When the revolt spread along the whole Pearl River system and Guangzhou, the 
capital of Guangdong province, was besieged for months, the Qing authorities in 
Guangzhou were in such a desperate situation that they were eager to seek the aid of the 
British in the suppression of the insurgents.305 Though not actively responding to the 
Chinese request, the British authorities did send their naval forces to Guangzhou under 
the pretext of protecting their national interests.306 In Guangzhou, Sir John Bowring allied 
304 Most of rebel leaders had never seen or made the acquaintance of a foreigner. Only Chen Xian Jiang had 
business dealing with foreigners in Guangzhou and he presumably spoke some pidgin English. Ineptness in 
foreign relations was one of major factor ofrebel demise. See Wakeman, Strangers, 146-147; F.O. 17.228, 
Desp. 108, Incl. , Robertson-Ye, (Feb. 23, 1855); F.O. 17.228. Bowring-Clarendon, Desp. 108 (Feb. 28, 
1854). 
305 Ye had been asking the British help since the beginning of December, 1854. He made the request that 
the British ships of war which were 'also in the river for purposes of protection' should join in the 
suppression of' the thieves in this river.' See F.O. 17.218, Desp. 230, Bowring-Claredon, 11 (Dec. 1854); 
Mai Bingjun ~JIH1J ed., "Heshan Maishi zupuji yutujishi lunlue ~w~ t:£;nj®&~OO-ic.JJ ffi~," in 
Hongbing Qiyi, 1016-1017. 
306 Various aspects of British assistance to Qing officials during the Tiandihui siege of Guangzhou are 
documented in both British and Qing sources. See F.O. 17.218, Des. 230, Incl. , Bowring-Ye, (11 Dec. 
152 
the British fleet with Governor-general Ye Mingchen's forces, and thus contributed to the 
'extermination which desolated the province of Guangdong.'307 Ye said that 'the persons 
executed by his order were over 100,000.'308 Ye dared to do this because he was being 
supported by the British forces. Bowring denied any responsibility, claiming that it was 
no part of his duty to interfere with the domestic affairs of China, but that he had gone to 
Guangzhou with a large naval force to protect the factories. 309 The presence of the British 
naval force along the Pearl River impeded seriously the river fleet operations of the 
insurgents in the regions. This was one of the reasons, as claimed by some authorities, for 
the sudden collapse of the Red Turbans.310 
After some disorganized fumbling, the authorities mustered several thousand 
1854); Scarth, Twelve, p. 278; "Heshan Maishi," in Hongbing Qiyi, 1016-1017; Luo Ergang, Taiping 
tianguo shiliao kaoshiji j;:5¥-j(wJllj:l.;J.~~~. (Beijing: Shanlian, 1956), p. 143; North China Herald, No. 
301 (may 3, 1856), and No. 359 (June 13, 1857); Thomas T. Meadows, The Chinese and Their Rebellions, 
(London: Smith, Elder, 1856), 453 ; Jian, Quanshi, 2, appendix to ch. 11, illustration 7; Parliamentary 
Papers, 1857, vol. XII, [2173] Correspondence relative to entrance into Canton, 1850-55, pp. 27-28. 
Commissioner Ye to Sir J. Bowring, Dec 7, 1854; Sir J. Bowring to Commissioner Ye, Hong Kong, Dec. 11 , 
1854. 
307 G. W. Cooke, China, 406; Luo Baoshan, "Guangdong Tiandihui qiyi qijian zhongwai fandongpai de 
goujie 1*j;:±l!!,~)1:g5CJ'tijfaji:p,'j-,&i-;tJ¥]KB'-]~~." in Taipingtianguoxuekan j;:5¥-j;:OO~f!J (Beijing: 
Zhonghua shuju, 1983), 416-429. 
308 Ibid. 
309 Stanley Lane-Poole, The life ofS ir Harry Parkes, K.C.B., GC.MG, sometime Her Majesty's minister to 
China and Japan. , Vol. i. (London, New York: Macmillian and co., 1894), 221. 
310 Eitel, p. 304; Hirayama Shu Zp-Li.J~, Zhongguo mimi shehuishi i:p~f,0ffijfd:fr~ (Shanghia, 1912), 
1935 edition, p.p. 26-27. 
153 
troops to deal with the rebels. Included among them were reinforcements from a 
formidable Qing general Shen Dihui tt~1$, hundreds of specifically recruited 
Chaozhou ~ fl·! mercenaries, militias from the Ninety-six villages of Nanhai county and 
an augmented personal militia of a military Juren ~A, Zhou Guoxing JWJ ~ ~. In March 
1855, the siege ofXinzao fr~ was lifted and the city ofFoshan was recaptured. 
Guangzhou was at last saved. Chen Kai and Li Wenmao chose to avoid direct 
confrontations with government troops as much as they could. They moved toward 
Guangxi with their men and established the Dacheng Kingdom (Dachengguo *1Pt~) in 
Xunzhou mfH.311 
The other Red Turban armies such as the bands of He Liu, Chen Jinkang,3I2 and 
Zhou Chun )WJ;g: (alias Dou Pichun .RJsllf:)31 3 were also repulsed. Their flight now 
described a rough arc: first they moved north through Yizhang 1r~, Chenzhou #!UH and 
Guiyangzhou tt~fl'I of Hunan province for a final stand. Government troops, mostly 
those under the command of Wang Xin .3:.~, were in constant pursuit of the rebels.314 
3 11 Jian, Quanshi, pp. 859-60; F.O. 931.1692. List of captured rebels in area south of Guangzhou (n.d.); F.O. 
931 .1606. Details about militia in Nanhai district (1855); F.O. 931.157. A list of Red Turban leaders (1855). 
312 "Gufei zonglu," in Hongbing Qiyi, 836; "Sihui XZ [!j ~-f!.$:," in Hongbing Qiyi, 1593-1594. 
313 "Zhen'geshi biji tt:J<;l:£;~-i,2,," in Hongbing Qiyi, 853. 
314 "Qing Wenzong Xianfeng Shilu tili:iOJ~~.$~::!R," in Hongbing Qiyi, 1152-53. 
154 
The Red Turbans periodically paused to attack towns, but without success. For the most 
part, they just tried to stay ahead of their pursuers. He Liu was killed in action on January 
3, 1856 near Chenzhou and the rebel ranks thinned steadily in combat. Kan Xian led 
remnants of the band back to Guangdong by a roundabout route and was himself killed 
near Huaxian in September.3 15 
In early 1856, Qing forces took their revenge on the helpless civilian population 
in the villages surrounding Guangzhou. Some claimed that a million people were killed in 
the "White Terror" on suspicion of involvement in the rebellions. Mass executions were 
held at Guangzhou and Zhaoqing .!f. Thousands were rounded up and sent in chains to 
the capital, where they were executed in a barbaric fashion. 316 The gentry took advantage 
of the government policy of indiscriminant repression to present their private enemies as 
rebels to be executed, using the dynasty's severity as a means of settling their own local 
grievanc~s. Around Guangzhou the gentry used this opportunity so effectively that, with 
the eclipse of the local administration and the destruction of their opponents, they gained 
315 "Yahu Caoshi zupu :fflJv.Je- B;;Mcif,t" in Hongbing Qiyi, ed. Cao Fengzhen e-ffel.!J!, 851; 'Kan Xian 
Shilue if .$t;~~'' in Sanyuanli renmin kangying douzheng shi/iao = 5Glt!.A ~J,;t~4~ ~f4, 20 I. 
316 According to Yong Kweng's account, Sohak Dongjumki, 75,000 people were executed by Ye 
Mingchen's order in Guangzhou alone and more than half of these people were not directly related to the 
Red Turban rebellions. See, Yong Kweng, Sohak, pp. 63-65. See also G. W. Cooke, China: being 'The 
Times ' special correspondence from China in the years 1857-8, with corrections and additions, (London, 
1858), 407; Wakeman, "Secret Societies," 246. 
155 
complete and undisputed control of the countryside. 317 
Militias 
Lastly, resistance from the gentry organized militias helped defeat the Red 
Turban armies. In face of the failure of the regular troops in time of crisis, the first 
response of the Qing military leadership was to hire paid fighters, or yang ~, to 
supplement the regular forces. Broadly speaking, the term yang was used to denote any 
irregular forces of a loyalist character. Thus the militia of a tuanlian ~ ~l association was 
sometimes described as tuanyang. But generally yong referred to forces on a higher level 
of militarization: men who were entirely detached from their communities and who 
depended for their sustenance upon pay or loot. Some such units were recruited directly 
by government officials, such as notorious yang from Chaozhou.318 There were 
numerous mercenaries employed by the Qing at Guangzhou at the beginning of the Red 
317 Ibid. ; Wakeman, Strangers, 149-151. However, regaining control of urban areas did not mean the end of 
the rebellions; countless marauding bands were still at large, plaguing the countryside. Qing anny was still 
engaged in a vigorous campaign of rural recovery as late as the end of 1857. 
318 The river valleys of eastern Guangdong, particularly in the area of Chaozhou and Jiaying, which 
notorious for their tradition of fierce lineage feuding, furnished recruits for both orthodox and heterodox 
forces in the civil wars of the mid-nineteenth century. The Hot-blooded Chaozhou mercenaries (Chaoyong) 
that stiffened Qing annies in the early years of the anti-Qing struggle were as unruly and contumacious as 
they were aggressive in battle; and proved thereby a headache to their government masters and a plague to 
the populace. See Jian, Quanshi, Vol. l , 370. 
156 
Turban Rebellions, but as the war wore on, there was an acute shortage of government 
funds,319 and the gentry were induced by the governor-general, with the promise of 
honors, titles or even offices, to use their own resources to hire their clansmen as 
mercenaries and to organize their clans into self-financing militia units.320 
The local gazetteers point to a trend of local militarization that began in the early 
nineteenth century. This upsurge of local militarization was historically new. Although 
Guangdong traditionally had militias organized to defend its villages from pirate attacks, 
the formation of militia was strongly discouraged by the Qing government on the grounds 
that it gave the gentry military and political power at the expense of the local 
government.321 Moreover, lineage conflicts and kinship divisions in the Guangdong 
countryside also prevented the militia from growing into a regional organization. 
Therefore, lineage militia before the 1840s remained illegal and localized. 
However, the threat of foreign conquest had served to realign the power structure 
319 Shortage of military funds was another reason for the prolonged war. The financial structure of the Qing 
dynasty was rigid and incompatible with the needs of special circumstances. The Ministry of Finance could 
barely make ends meet even in time of peace, and in wartime the central government could only issue 
mobilization orders and dispatch troops to the front, without being able to give them adequate weapons and 
provisions. These had to be furnished by the locality where the war occurred. See F.O. 931.1092. A report 
regarding military preparation for attacking rebels (1850s). 
320 F.O. 253A. 3. 56. Regulations against banditry, etc. (1855) 
32 1 Wakeman, Strangers, :3 l 
157 
of Guangdong in new ways. First, militias became organized on a higher level than the 
village. The Sanyuanli militia, for instance, involved 103 villages and tens of thousands 
of people.322 Second, militias began to gain formal legitimacy under the guise of being 
local academies or shexue )f!±~, which were mainly educational institutions before the 
l 840s. At the peak of the antiforeign movement, however, the local academies were 
transformed into organizations to recruit men and to raise money for regional defense.323 
These organizations were called local academies rather than tuanlian (militia) because the 
former term sounded eminently Confucian and could, thereby, receive official blessings 
from the Qing court.324 Third, while the local academy was a gentry organization, it 
received more mass support than before. The drain of silver out of China, the importation 
of opium, and the threat of foreign conquest had given the gentry class an°ideological 
basis to unite both the rural population and the city craftsmen to fight against foreign 
322 Hsiao, Rural China, pp. 493-95 . 
323 Started as gentry-led institutions to promote local education, the academies became centers of political 
networking. In times ofrelative peace, they created channels by which kinsmen could go through the civil 
service examinations. In times of dynastic decline, such as from the mid-nineteenth century on, they were 
commissioned by county officials to organize regional self-defense. They were also entrusted with rent 
collection for urban-based corporations that owned vast estates in the sands. By 1850s, the shexue gave 
practically no public lectures but devoted all their time to keeping the peace. As a result, newly created 
institutions of this type could rather be calledju (gongju 2:lm), which meant a local council. See F.O. 
931.1493. and 1522, 1142 and 1143. Reports on rebels, etc. (1855) 
324 Hsiao, Rural China, p. 495; Wakeman, Strangers, p. 64. 
158 
intrusion. 325 In sum, the militia now organized on a wider regional scale with semiofficial 
blessing and mass support. As a result, the gentry class was able to strengthen its power at 
the time of foreign conquest.326 Although the militia units formed during the Opium War 
were disbanded, the same organizations and the same gentry leaders were to take a 
prominent part in the suppression of the Red Turban Rebellions of 1854. 
The Red Turbans' drive through the Pearl River Delta areas in the 1850s 
galvanized residents to re-mobilize for self-defense. Now, in the cities and the larger 
market towns, officials and elites resurrected the procedures used in Sanyuanli in 1841-
the formation of militia or self-defense bureaus, the recruitment of mercenaries, the 
organization of baojia as a patrol system and conscription pool. The mobilization appears 
to have been much fuller than in 1841 , giving the defense efforts of the 1850s a depth that 
the earlier ones lacked. Whereas the pertinent sources for 1841 are virtually mute on the 
subject of militarization in places other than cities and the larger market towns, the 
325 Hsiao, Rural China, p. 493 . 
326 The militia was one of those key institutions in China that to some extent represented a convergence of 
the interests of state and society. For rural inhabitants, a viable defense structure means protection of their 
livelihood against the threat of banditry. To the government, the militia was a means of maintaining its 
mandate in the face of rebellion. Official encouragement oflocal defense was forthcoming when the regime 
found itself unable to cope with serious rural unrest. For a cogent analysis of the role of militia in the 
history of China, especially in the late Qing periods, See Kuhn, Rebellion and Its Enemies. For an excellent 
overview of militia activities during the Red Turban Rebellions in Guangdong, see Wakeman, Strangers, pp. 
144-156. 
159 
materials for 1850s speak repeatedly of the formation of militias in countryside. These 
rural defense groups, most commonly called "white soldiers" (baibing B ~ ) for the 
white cloth they wrapped around their heads to distinguish themselves from the ' red-
turbaned' rebels, are usually identified by the name of the small market town or village 
that served as their headquarters, suggesting that natural, as opposed to administrative, 
units formed the basis of organization.327 
The most formidable militia complex around the Pearl River Delta was organized 
in Shunde county. Like other counties, Shunde county was badly hit by the Red Turbans. 
In addition to the usual village-by-village devastation, there were also ruinous raids of 
Red Turbans from other counties. When the Red Turbans captured the county capital in 
1854, the country magistrate of Shunde fled and the local government dissolved.3 28 After 
the downfall of the local government, the gentry class in Shunde country quickly 
reasserted itself. It reorganized militia in villages and in a few months, drove the Red 
Turbans out of Shunde county. It was reported that 13,000 rebels were executed in the 
course of the suppression. Subsequently, when the order from the provincial governor to 
327 Xuxu ~~' Lingnan gujinlu ilitw4°i!i-4'3R (Guangzhou: Guangdong renmin chubanshe, 1992), 236-237. 
328 Shunde XZ, 1929, 23 .5-7. 
160 
form a militia arrived, the gentry greatly expanded the scope of the local militia to form 
the Shunde County Central Militia Bureau on May 7, 1855.329 
Another powerful militia unit was formed, called the Ninety-Six Villages in 
Nanhai county and situated somewhere between Guangzhou and Foshan. They were so 
well organized and efficient that they not only successfully defended their villages, but 
also brought a halt to the advance of Chen Kai's troops on Guangzhou.330 The 
contributions by these militias towards the suppression of rebellions and policing of the 
local areas were considerable. 
In mid-west Guangdong, especially the Wuyi regions, however, the revolt could 
not be defeated, nor could the areas be pacified in the same way. In these districts there 
was not the tradition of gentry organization and gentry-led militia that had developed in 
the districts around Guangzhou. As the area was cut off from provincial capital by rebel 
bands, there was no hope of obtaining assistance to deal with the revolt. District 
magistrates had no forces at their command, and on their own they were powerless. 
Moreover the magistrates of Heshan and Kaiping had been killed, or had committed 
329 Ibid., 3.1, 16.3, 23.7. 
330 This was a vital for the survival of Guangzhou because it prevented Chen Kai joining forces with the 
rebels besieging the city. See F.O. 931 .1455. Report from three leaders of the gentry on campaign against 
rebels near Foshan (Aug. 1854). 
161 
suicide when their cities had been captured. Faced with a critical situation, in a number of 
districts the magistrates took a step that was later to be considered to have done more 
than anything else to provoke the Hakka-Punti War- they summoned Hakka braves to 
fight against the Red Turbans, most of whom are the Puntis.331 The Hakkas now formed a 
gentry-led militia similar to the militia around Guangzhou and defeated the rebel forces at 
the gate of Xinning district town, a victory that eventually stemmed the rebel tide in this 
area.332 
While the need to use local forces at a time of emergency was recognized, 
officials regarded the raising of militia as potentially dangerous. As a consequence, there 
was a difference between the concept of the tuanlian held by officials, and that held by 
the gentry. For officials the tuanlian was an immobile and non-professional force, the 
direction of which should come under the district magistrate, who should also approve 
the immediate leaders. For the gentry, the tuanlian, unlike the baojia, system offered 
them direct influence in local affairs.333 At the time of the Red Turban Rebellions, the 
33 1 This incident eventually created the sharp ethnic consciousness of the Hakka and the process of this will 
be dealt in detail in the following chapter. See Enping XZ, 14.7b; Kaiping XZ, 21 .3b. 
332 Xinning XZ, 14. I Sb-21 b. 
333 Philip Kuhn, "The T'uan-lien Local Defense System at the time of the Taiping Rebellion," Harvard 
Journal ofA siatic Studies 27 (1967) 
162 
weakness of the government allowed the gentry to exploit their potential strength as 
leaders of the rural community. For instance, having such a strong county militia bureau, 
the gentry class in Shunde was able to established gentry hegemony in the aftermath of 
the Red Turban Rebellions. Since nearby Xiangshan county was still under the threat of 
the Red Turbans, the gentry of Shunde county had an excuse to take over the control of 
the police forces of the polder farm in Xiangshan county, which was known collectively 
as 'Donghai Shiliusha * $-t 7\ ¥~ (Sixteen delta lands of the Eastern Sea). ' 334 
A few years later, the Arrow War also provided the gentry from the districts 
around Guangzhou with the opportunity of taking the lead in the resistance movement. 
The uncoordinated units of tuanlian south of Guangzhou were provided with leadership 
by a body known as the Guangdong Militia Bureau, which, until 1859, pursued a policy 
of harassing the European forces. The effect of permitting the gentry to take control of 
334 The Shunde gentry then asked the polder farm owners to donate 0.6 tales per mu annually as a protection 
fee. Since there were 460,000 mu ofpolder far in Xiangshan county, this meant an annual revenue of 
276,000 tales to the gentry in Shunde county. With such a strong financial base, the gentry in Shunde 
county were able to donate huge sums of money to finance the provincial military as well. In the year after 
the Red Turban Rebellions, the Shunde gentry donated 900,000 taels-about one-seventh of the total 
provincial budget. Another source ofrevenue for the gentry class was lijin JI~-namely, transit tolls levied 
on articles of commerce. Each subdivision (called gongyue Z-~) of the Shunde Central Militia Bureau was 
functionally charged with the management of funds in conjunction with the hiring of militia, the 
distribution ofrelief, and the reconstruction of local schools. The subdivision obtained those funds by 
erecting customer barriers around major market towns. For instance, the Longjiang gongyue ~IT Z-~, 
which had I 00 braves and an annual expenditure of I 0,000 derived its revenue by collecting donations 
from the cocoon, butcher, and mulberry activities. See Shunde XZ 3.2-11 , 23.7; Xiangshan.XZ, 1920 16.5-7. 
163 
local defense was described in an article in the Overland China Mail: 
In availing themselves oflocally directed efforts of mercenary 'braves,' whether 
among the Hakkas or in other parts of the country (among the Puntis), the 
authorities in Canton were unwittingly calling into existence a power which was 
designed to usurp, in a very few years, a large share of the functions of 
government. The 'gentry' ... being the persons most capable of influencing their 
fellow villagers, were now everywhere formed into committees, and invested for 
the first time with official powers. They levied and disposed of funds, led armed 
bodies into the field, executed prisoners, were universally employed as 
pacificators, and may be said to have saved the Province. The official standing 
thus acquired (their private leadership in their own villages had always, naturally, 
existed) has grown more extensive year by year; and the Provincial Government 
of today finds that no fiat can be executed in the country without the concurrence 
of the Kung-ku, Committee or Municipal Council-by whom the affairs of the 
District are administered as ad-latus to the actual Magistrate. 335 
In sum, the Red Turban Rebellions had again realigned the power structure of the 
local Chinese communities, with the gentry and the Hakkas emerging from the rural 
upheavals considerably strengthened at least in some areas.336 In the 1850s, the gentry 
obtained the legitimacy to reorganize regional militia, to have its own tax base to finance 
the militia, and to use the militia for regional hegemony.337 
335 Overland China Mail 11 (July 1863) 
336 This kind of case was still not happened in the Wuyi, for the position of the Punti gentry could not be 
considered undisputed so long as the Hakkas could still hold out against them. 
337 For example, in just a few years after the rebellions, 8,641 mu of communal land were acquired to build 
a new educational institution called Qingyun Wenshi ~~Jtjjii (Blue Cloud Academy). See Shunde XZ 
4.34-36. 
164 
4) The Red Turban Rebellions Interpreted 
In the great wave of rural insurgencies that enveloped China in the mid-
nineteenth century the Red Turban Rebellions stand as only a minor incident, 
overshadowed by more massive upheavals such as the Nian and Taiping rebellions. The 
unstable situation, particularly created by the Taipings, undoubtedly influenced the local 
Tiandihui leaders and their followers in the Pearl River Delta. Yet as the preceding 
analysis has shown, the Red Turban Rebellions were not simply spontaneous products of 
the influence from the Taipings. Their history-the social relations and conditions out of 
which they arose, the specificities of their development and their place in the social 
process-reveals a multi-layered complexity. Involving a multiplicity of identities, 
informal networks and invisible connections amidst social and economic change, the 
movement draws attention especially to the interpenetration of class and ethnicity and 
Tiandihui organization in the emergence of insurgent politics. 
In this chapter I have attempted to focus on the set of events that led up to the 
Red Turban Rebellions and to describe the rebellions themselves. The factors giving rise 
to the Red Turbans and the details of their development are of interest for a number of 
reasons. 
165 
First, the local conditions of the mid-nineteenth century suggest that the level of 
conflict in the Pearl River Delta was high, that outbursts of collective violence were 
common, and rationales for rebellions were readily available within Chinese popular 
culture. The evidence suggests that the origin of the Red Turban Rebellions owed much 
to the special historical circumstances of mid-nineteenth-century Pearl River Delta-
crippling natural disasters exacerbated by inefficient government, increased taxes, and the 
success of the Tai ping rebels. Most of the rebel groups were spurred to rebellions by a 
combination of economic conditions and government demands. Not only natural disasters, 
but also government abuses and excessive tax demands could easily push people below 
the margin of subsistence. Therefore, government commutation of taxes was often a 
major issue for the Red Turban rebels. Violence was already endemic to the Pearl River 
Delta. 
Second, the local violence was further aggravated by the heavy-handed practices 
of the Qing state against local Tiandihui lodges which became centers for the organizers 
and perpetrators of the Red Turban Rebellions. While many of the individual outbreaks 
that collectively made up the Red Turban Rebellions were directly linked to events that 
threatened the livelihood of the people, subsistence crises alone do not provide an 
166 
adequate explanation. In the Pearl River Delta, the decision to rebel seems to have been 
made deliberately in almost every case. 
Most of the rebels did not rise against the government on the spur of the moment, 
as someone acting out of blind fury would do. Both Chen Kai and Li Wenmao spent 
several months preparing for revolt under the leadership of Chen Song, and their risings 
were both related to subsistence-threatening tax demands and to the Tiandihui 
direction.338 He Liu's Tiandihui lodge spent more than a year organizing to resist 
excessive tax demands related to commutation.339 To borrow the phraseology used by the 
Tillys in the European context, "There is outrage, all right, but nothing like blind fury. 
The whole sequence bespeaks premeditation and prior organization. "340 The same can be 
said about the Red Turban Rebellions; the rebels-that is, the Tiandihui leaders and their 
committed members-weighed the alternatives and made their decision. In other word, 
the ambitions of certain Tiandihui leaders that intensified and transformed local 
discontents into massive rebellions against the state. 
338 Foshan Zhongyi Xiangzhi, 11.14-15; "Chen Song Shiyou," in Hongbing Qiyi, 212-213; Yoo, Kundae, 
284. 
339 He Liu organized the Tiandihui in his area for at least a year before he was pushed into large-scale, 
active rebellion. See "He Liu deng shilong qishi yuanyou fiiJ7\~1:ift~$~El3," in Hongbing Qiyi, 214. 
34°C harles Tilly, Louise Tilly, and Richard Tilly, The Rebellious Century, 1830-1930 (Cambridge: Harvard 
University Press, 1975), 179. 
167 
Third, although people of different ethnic groups and different social strata 
banded together against the Qing government, the revolt particularly in the Wuyi regions 
was caused largely by a pervasive subethnic hostility. At the low level of society, the 
Punti rebels in the Wuyi regions tended to direct their hostility primarily at Hakka people. 
The solidarity demonstrated by Punti rebels and by Hakka militias is one of interesting 
findings of this study. After the suppression of the Red Turban rebels, the Hakka-Punti 
armed conflicts continued for more than ten years (1856-1867) and covered most of the 
Wuyi regions along the Pearl River valley. These had an unquestionable influence on the 
development of the ethnic consciousness of the Hakkas, as we shall see in the next 
chapter. 
Fourth, the Red Turban Rebellions had again realigned the power structure of the 
local Chinese communities, with the gentry emerging from the local disorder 
considerably strengthened. What G. William Skinner has termed "community closure" 
began to take effect.341 Lineages, sublineages, and villages turned inward for protection 
while competition and feuding among communities intensified. Economic activity at the 
regional level and economic mobility became ~onstricted; protective confederations and 
341 G. W. Skinner, "Chinese Peasants and the Closed Community: An Open and Shut Case," in Comparative 
Studies in Society and History, 13 (1971 ): 270-281. 
168 
boundaries hardened. As magistrates hesitated to interfere, as lineage leaders were able to 
defy civil servants, and as disaffection with the government spread, Qing governance 
became ever more superficial. Villagers took the law into their own hands. In the 1850s 
during the Red Turban Rebellions, the gentry obtained the legitimacy to reorganize 
regional militia, to have its own tax base to finance the militia, and to use the militia for 
regional hegemony. 
Last, the suppression of the Red Turban Rebellions with the aid of foreign troops 
was not only brutal and devastating but also shortsighted in terms of political action. The 
human and material destruction was immense and incalculable, sowing profound 
bitterness and xenophobia in the minds of survivors. Thousands of these had been driven 
from their homes and livelihood, while others emigrated abroad, some as indentured 
laborers in the Americas. In addition to hundreds of villages, some thirty-two different 
administrative seats in Guangdong ranging from the district to the prefectural level were 
lost to the insurgents during the period of 1854 to 1855. Most were quickly recaptured, 
but several were held by the rebels contirtuously for more than several years. For example, 
Deqingzhou in Zhaoqing county was held by the rebels until April 1858.342 The 
342 Yoo, Kundae Zoongkuk 'yi Bimil kyulsa, p. 279. 
169 
provincial capital never fell, but it came perilously close on two occasions. Towns were 
reduced to rubble, fields were overgrown with weeds, and large tracts of the province 
were depopulated. 343 
A foreign traveler in the late l 850s observed that after defeating the Red Turban 
rebels, "The mandarins burnt all the villages they had occupied. It was a heart-rending 
sight to see these deserted desolate ruins ... Land, meanwhile, remains uncultivated, 
famine follows, and trade is disarranged."344 In this period, the Shunde gazetteer also 
indicates that the local economy had greatly deteriorated during the rebellions and many 
roads and bridges were damaged.345 The other problem involved the flooding of the area. 
343 The devastation in the Pearl River Delta, including the burning of numerous houses, was not caused by 
the rebels alone, but by the imperial troops and local militias as well. The living of many of the displaced 
refugees were lost not only through the fighting, but through hunger, pestilence, and other tragedies as well. 
The Red Turban advance meant war, and war meant the destruction of crops, the disruption of trade routes, 
pillage, rape, and death. As reiterated in account after account, Qing troops were responsible for much of 
the wrack and ruin in Guangdong, to the point that, as one gentryman lamented "the people fear the Qing 
soldiers, but not the Red Turban rebels." See F.O. 931.1461 A letter advising against the demolition of 
civilian house as a method of defense. (July 9, 1854). 
344
· Scarth, Twelve Years, pp. 230-232. 
345 Shunde XZ, 3 .26. At Guangzhou, the Red Turban rebels besieged that city in the latter part of 1854 and 
into 1855. After their defeat, rebel discipline broke down quickly, and bands of marauders ravaged the 
countryside. The instability cast doubt on the ability of the Chinese to maintain a steady supply of silk. Not 
surprisingly, British merchants began looking to alternative sources of supply. There were already such 
alternative sources outside China. By the time of the Arrow War, India, Turkey, Egypt, and Holland had 
become substantial rivals to China as suppliers ofraw silk to the United Kingdom. Once a trading 
relationship had been consolidated with these alternative sources, businessmen would be reluctant to go 
back to the old supplier even after conditions in China returned to normal. Thus the rebellions had one 
serious damage to the local silk industry as well. For statistical information about the quantities ofraw silk 
supplied by the various parts of the world to the United Kingdom in 1850s, see Shunde XZ, 4.1; Parliament 
Papers 1859, Session 2, v. 23, 319 and v. 65, 292. 
170 
For instance, Shunde County is a part of the Pearl River Delta and for the most part lies 
only slightly above high tide line in the summer rainy seasons. In July and August, these 
low lying lands were subjected to severe flooding, thereby reducing rice cultivation from 
the usual two crops per year to one crop annually. In the 1850s, this flooding problem 
was getting worse as a result of the lack of repair of embankments after the Red Turban 
Rebellions. 346 
The cost of the rebellions was appalling in terms of lives, property damage, and 
funds required for suppression. Reliable figures are difficult to find. But, more than 
26,000 people in Panyu and Shunde counties and 10,000 people in Zhaoqing county were 
executed by Qing government.347 According to one scholar's estimate, about 400,000 
people died in Guangdong province because of this "White terror" by Qing court.348 
In sum, the Red Turban Rebellions was compounded of many ingredients, 
including adverse ecological conditions, which led to increased competition for 
346 Due to the weakening of state controls after the rebellions, the state could not prevent the gentry from 
depositing stones in the course of a river, which led to the blocking of river flow, the quickening of silt 
deposits, and the extending of the river delta toward the sea. This artificially created river delta quickened 
the pace with which the gentry accumulated land ownership, and caused the water level to rise above the 
embankments, enhancing the danger of flooding. See Shunde XZ, 4.1 ; Nanhai XZ, 1873, 7.7-8. 
347 Luo Baoshan ~-~. 'Taiping tianguo shiji Guangdong Tiandihui qiyi shulue :;t.zi:;:7C~~;ttJ!/JJI! 
7C:l:t!r~~~3Zlt~,• Zhongshan Daxue xuebao q:iL1Jj::~~¥R, 4-1 (1981), 58-59. 
348 Xu Xu, Lingnan, 244. 
171 
resources; government abuses and rnal-adrninistration; communal frictions; and the 
activities of the Tiandihui. The Tai ping rebellion also played an important role in the 
timing of the Red Turban Rebellions. Without it, unrest in the Pearl River Delta would 
probably not have exploded with such intensity. It was the corning together of a number 
of different socioeconomic and historical rhythms at a particular time that brought about 
series of rebellions, of such massive dimensions. 
Seemingly, the explosive situation underlying the Hakka-Punti War was 
concentrated in certain local communities, the Wuyi region rather than evenly permeating 
the peasantry of the Pearl River Delta. Therefore, in the following chapter, I examine in 
detail the emergence of the ethnic consciousness of the Hakkas, focusing on the Wuyi. 
172 
CHAPTER IV 
ETHNICITY, LOCAL INTERESTS, AND THE STATE 
Subethnic rivalry arose in the Pearl River Delta during the Qing period as a result 
of heterogeneous Chinese immigration and settlement. In particular the intercommunity 
competition among the two major subethnic groups,349 the Hakkas and the Puntis, which 
emerged within the South China, influenced internal migration and local patterns of 
settlement. In the case of the Wuyi region in Guangdong, however, this rivalry spread and 
intensified during the Red Turban Rebellions and grew even after the end of the 
rebellions. 
In the face of the Red Turbans most Hakka clans remained loyal to the Manchu 
349 Tribal, national, and racial ethnicity is the most prevalent forms of ethnicity. However, in China- as in 
other parts of the world where pervasive cultures or civilizations with "great traditions" have existed-
major subcultural distinctions have often led to separate ethnic identities apart from tribal, national, or 
racial ascriptions. Hence "subcultural ethnicity or subethnicity" may also be recognized as a widespread 
type. The term "subethnic" refers to the shared village, native place, and the shared branch or subgroup of a 
lineage on a small scale. On a large scale it is the shared lineage, speech group, the shared surname, as well 
as the sense of different species of people. See Cynthia H. Enloe, Ethnic conflict and political development 
(Boston: Little, Brown, 1973), 23-25; Bryna Goodman, "The Locality as Microcosm of the Nation?-Native 
Place Networks and Early Urban Nationalism in China," Modern China, Vol. 21 No. 4 (October 1995), 
387-419. 
173 
officials. Most of the Punti tenants, in contrast, joined the Red Turbans until brutally 
suppressed by Ye Mingchen, Governor-general ofLiangguang. Feuds among the two 
subethnic groups intensified when the Puntis attacked Hakka villages in retaliation for 
their support of the Manchu government during the Red Turban Rebellions. This local 
rivalry among the Hakkas and the Puntis also spread and sparked subethnic disturbances 
in other areas. Extensive feud patterns formed between the Punti and the Hakka 
groupings in many parts of the Pearl River Delta estuary. The struggle turned into open 
warfare in which the officials were powerless to interfere and which continued long after 
the end of the Red Turban Rebellions. It is estimated that subsequent intercommual wars 
involving the Hakkas and the Puntis led to some half-million deaths in Guangdong. 350 
Subethnic conflict of this magnitude eventually perpetuated local tensions and 
consolidated the boundaries between contending communities and enclaves, thereby 
enabling subethnic rivalry to become an enduring and deeply rooted phenomenon in the 
local society. Most importantly, these conflicts were in tum important in consolidating the 
ethnic identity of the Hakka people. 
My primary aim in this chapter is to analyze the historical causes of the 
emergence of ethnic consciousness of the Hakkas and to show how conflicts between the 
Hakkas and the Puntis during the Red Turban Rebellions contributed the strengthening of 
350 Luo, Kejia yanjiu daolun, 3. 
174 
Hakka identity. In order to do this, it is necessary to provide a description and analysis of 
the relationship between the Hakkas and the Puntis in the Wuyi region. 
1) The Wuyi Region in Historical Perspective 
Geography of the Wuyi 
The main channel of the West River of Guangdong forms the west side of 
the Pearl River delta. West across the river the pronunciation and vocabulary of 
spoken Cantonese changes to that of the Siyi sub-group (see Map 2). Siyi 123 B 
(Sze-yup; four counties) is the collective term referring to the four counties of 
Xinhui, Xinning (Taishan), Kaiping, and Enping. To this groups is added Heshan 
so that is now becomes the Wuyi E. B (Ng-yup; five counties). 351 
Descriptions of the Wuyi terrain follow two main themes. The first 
focuses upon the beauties of land. Most of the Wuyi regions were described as "a 
beautiful abundant place at the edge of the Southern Sea, inland there are flat 
351 See HIM Mark Lai, In Search of Roots, A Course (Chinese Cultural Center: University of San 
Francisco); Chen Chengxi ~ ;ifi;.fff , Guangdong dizhi ./it:*:ftl!~ (Hong Kong: Tiandi chubanshe, 1978), 3-4, 
31 ; Chunfat Lau, The Decline oft he General Hakka Accent in Hong Kong: A Comparison of "Old-Style" 
and "New-style " as spoken by the indigenous inhabitants (Milnchen: Lincom Europa, 2000), 2; 
Guangdongsheng fangyan diaocha zhidaozu 1 *1lf1i"§ir.l lH~-@~J:I., ed., Siyiren xuexi putonghua shouce 
[/] e!A ~ }:J ~ iii !5 .:f-M (Guangzhou: Guangdong renmin chubanshe, 1958). 
175 
plains, hills and high mountains. "352 However, the more common and grimmer 
second theme emphasizes the limited arable land for agriculture. 
Map 2. Wuyi (Five Counties) in the Pearl River Delta . 
.1  H UN A N ) 
/ Jiangxi ,l Fujian 
,,.... .. .J.... 
..r ··" · . ·. / •.  .•  ....:, \. . 
\ 
Guangxi 
ONG 
o f Pearl River 
MA CAO 
5 £ A 
I N A • 
C H 
1. Heshan 
2. Xinhui 
3. Xinning 
4. Kaiping 
5. Enping 
-- Inte r na t ional boundary 
0 50 100 _ ., _ Provincial boundary 
MIL E S ----- Dist r ic t boundary 
Certainly the Wuyi region is a pretty place with blue skies, white clouds, 
and fields laid out in a patchwork of varying shades and textures of green tucked 
352 Tan Xin iif,C.,, ed. , Siyi f engguang [g el,Ji}t (Hong Kong: Wanye Chubanshe, n.d.), 49. 
176 
between tree-covered hills. These hills, which are so picturesque, pose the greatest 
obstacle to economic development because their limited productiveness make it 
difficult to increase the amount of cultivated land and they make the cost of 
building roads, railroads or canals, which might improve commerce, prohibitively 
expensive. Xinning, for instance·, by 1893 could produce only enough grain to feed 
its inhabitants for half a year. Such limits to economic growth posed serious 
problems. 353 
Much of the Wuyi is rugged: hills, tablelands, and high mountains 
occupy an estimated 7 6 percent of the total area; plains and basins merely 24 
percent. The narrow alluvial plains along the eastern coastline were quite 
extensively settled during the Qing period. Foothills rise sharply from the coastal 
plains and then spread inland. Many of the habitable foothill areas were also 
settled during the Qing period, as were the larger interior valleys. 354 The region' s 
physical features tended to divide it into highland and lowland sections. The vast 
353 Although the land is broad, rural areas are narrow, probably sixty or seventy percent mountains 
and seas. An examination of production reveals that it is certainly not enough to satisfy necessity. 
The people of this place consider this a matter of life and death. The insufficiencies suffered by 
the people are why many of the Wuyi people hire themselves out overseas. See Xinning XZ, 8.6b; 
Xinning xiangtu dili f}r$jffl±:f.m:f:lll. (Taishan: n.p., 1900). 
354 Chen, Guangdong, 3-4, 31 , 65 . 
177 
interior featured mountainous terrain indented by river valleys and upland basins. 
To the east, only the lowland valleys and basins, together with the scattered 
alluvial plains along the seacoast, provided relatively level landscapes.355 
In addition, the Wuyi region faces the unenviable situation of being 
surrounded by neighboring areas better endowed by nature to become wealthy. It is 
close to three international ports: Guangzhou, Hong Kong, and Macao, and sits 
near the market-oriented Sanyi r!s region (three counties, Nanhai, Panyu, Shunde, 
located south of Guangzhou). 356 Unfortunately, the Wuyi region has neither the 
natural waterways nor the acres of rich alluvial soil needed to duplicate the 
success of these nearby regions.357 Graham Johnson describes central Xinning; 
" ... Its prospects for growth and development were, until recently, hampered by 
355 Most of the Wuyi regions are landlocked except some parts of Xinhui and Kai ping counties and, 
like the rest of western Guangdong, has a hilly topography. Some 70 percent of its land is barren 
and unproductive. Kaiping county is, however, somewhat more fortunate than other counties, for 
the Tan River ?llIT and some of its tributaries flow through Kai ping on the way to the important 
port of Jiangmen. Most of Kaiping's arable land is found in the valleys along the Tan and its 
tributaries, where the bulk of Kaiping cash crops are cultivated. See Kaiping XZ, 56-62. 
356 A number of factors account for the vitality and prosperity of the Sanyi region, which had been blessed 
by nature in various ways. Overall, it had the most fertile and productive land within the province. What is 
more, it had an ideal natural environment for internal and international trade. Its many inland waterways 
served to link different counties within the region, and the long, curved coastline offered fine harbors, 
connecting the region to economies overseas. As early as 1730, the Emperor Yongzheng noted: " .. region is 
surrounded by the ocean on three sides, where merchants arrived from various provinces and foreign 
barbarians come with money to purchase goods. Trade is very heavy." See Zengcheng XZ, Introduction, 10. 
357 Liang Rencai ~t~, Guangdong Jingji dili ./JUIHrf~ttg}I (Beijing: New Science Publishers, 1956), 
21-24, 78-79. 
178 
inadequate communications. "358 
Agriculture was the cornerstone of the Wuyi region 's local economy, as it 
still is today. Late nineteenth century statistics reveal that four hundred out of six 
hundred thousand people in Xinning county depended upon agriculture in full or in 
part to make a living.359 Double-cropping of rice is common, with cereals and 
sweet potatoes being the staple crops. But, only one-quarter of the food supply 
was grown locally. 360 The raising of cattle, goats, and pigs as well as fishing 
were major activities, and the chief exports included peanuts, fish, meat, animal 
hides, hemp, clothing, grass sacks, grass mats, tobacco, and tea. 361 
Dependence upon the land was particularly risky in a poorly endowed 
place like the Wuyi region. Frederic Wakeman estimates that in 1812, there was 
only 1.67 mu of land for each person in Guangdong province, or one-quarter of an 
358 Graham Johnson, "Family Strategies and Economic Transportation in Rural China: Some Evidence from 
the Pearl River Delta." In Deborah Davis and Stevan Harrell, eds. , Chinese Families in the Post-Mao Era 
(Berkeley: California University Press, 1993), 112. 
359 Zheng Dehua llffi~, "Shijiou shiji mo Taishan qiaoxiang de xingchengji pouxi 
-J-.fLt!ttic.*ii'L.1.Jfit~1¥JM$;.bt.g:Utfr," Guangzhou huaqiaoyanjiu /Ji1'1·1~{\WHff~ 4 (1986), 68. 
360 Unlike the central Sanyi region near the provincial capital, where land was relatively fertile and 
conditions sometimes prosperous, the Wuyi region was capable of maintaining themselves when conditions 
were stable but slow to recover from disruptions. The hilly land yielded little, and productive capacity was 
low. The importation of rice from abroad-mainly Thailand- and from neighboring region began as early 
as early eighteenth century. See Hsiao, Rural China, 3 81. 
361 Liang, Guangdong, 21-24, 78-79. 
179 
acre. 362 In the Wuyi region, the situation was more dire because of the poor 
quality of the soil, which was red or yellow in color and often mixed with stones, 
sand or salt. 363 The local gazetteer states that the Wuyi region was not a good place to 
live and work. "This section of the country is generally remote, desolate, and 
sickly .... Until a generation ago, politically undesirable and criminal elements were 
banished into this region."364 Virtually all local gazetteers depict the area, time and again, 
as a distant and unwholesome place. They also suggest in unequivocal terms that the 
Wuyi was considered an underdeveloped and poor region compared to neighboring 
counties. Therefore, during the fall and winter seasons and during times of poor harvest, 
people from these areas, especially Xinning and Enping counties, went to other counties, 
such as the Sanyi region, to seek employment as artisans and other types of workers.3 65 
People from this less developed Wuyi region often entered the delta's commercial 
362 Wakeman, Strangers, I 79. This figure contrasts sadly with the 4.0 mu needed for bare subsistence. See 
Ng Chin-keong, Trade and Society: The Amoy Network on the China Coast, 1683-1735 (Singapore: 
Singapore University Press, I 983), 15-16. 
363 For example, fifty-six percent of the surface area of the Xinning county is taken up by hills and 
mountains and only 1,659, I 00 mu, or 34.9 percent of total land area of 4,748,000 mu, is cultivable. See 
Taishanxian dang'anguan f,;-LiJ!!,. ;j;~~'cit ed., Taishan di/izhi f,;-w±-1!!. lllL'it (Taishan: Zhongguo Taishanxian 
weiyichuanbu, I 984), 30; Zheng Dehua and Cheng Luoxi Jil<;;,ijffi, Taishan qiaoxiang y u X inning tie/u 
irLIJfit~~ffi~ ffljjlg (Guangzhou: Sanlian shudian, 1991), 2. 
364 Enping XZ, 1.2a. 
365 Enping XZ, 4.12. 
180 
network as suppliers of labor. 
Early History 
The Wuyi region had originally been the homeland for aborigines called "Yao" 
tribes (Yaozu ~ ~ ). The Yao people remained culturally stagnant for many centuries, 
and had subsisted primarily by hunting and gathering.366 The Han Chinese, who them-
selves were a racial intermixture from north China, began to colonize Guangdong 
province as a whole in 234 B.C.367 In this early stage, the Han Chinese were reluctant to 
venture out from their civilization. Consequently, the original colonizers of this remote 
region were said to have been banished criminals and political exiles.368 However, 
southward migration of the Han Chinese increased whenever Han areas in the north were 
devastated by the onslaughts of nomadic people from the steppes and deserts further 
north. Other causes of migration were natural calamities in the forms of flood and 
drought, which brought famine. For these various reasons the Han Chinese moved south 
366 Herold J. Wiens, Han Chinese Expansion in South China (Hamden, Conn .: Shoe String Press, 
1967), p, 51 , 106 
367 Ibid., 132. 
368 Enping XZ, 1.2a, 1. 7. 
181 
and east in a constant search for peace and arable land during the fifth and sixth centuries. 
Meanwhile, the cultural frontiers of the Han Chinese expanded steadily toward the south 
along with the migrating people. With their superior farming technology, the newly 
settled Han Chinese were able to cultivate the choice lands and build teeming villages as 
they advanced to the south. 369 In the settling process of the Han Chinese, the 
aborigines either ran away into the mountains or were absorbed into the dominant 
group. 
An intensive effort for the economic development of Guangdong began during 
the Ming dynasty. Increasingly Han Chinese penetrated and settled in areas inhabited by 
other ethnic groups such as the Yao, She, Zhuang and Li. During the mid-fifteenth 
century the Ming government abandoned its policy of indirectly ruling non-Han peoples 
through appointed native headmen. Instead these areas were put under direct Han 
Chinese rule and the people encouraged the adoption of Chinese language and customs. 
This policy of sinicization led to friction and resistance. This was reflected in historical 
369 Kaiping XZ, 2. l 9a-20a. 
182 
chronicles by the occurrence of numerous Yao, She, Zhuang and Li uprisings in 
Guangdong during this period. 370 
During the sixteenth century the Portuguese arrived off the China coast. They 
established one of their earliest bases on Shangchuan _lJII (St. John) Island off the 
Xinning (Taishan) coast. In 1551 Father Francis Xavier landed on the island seeking 
permission to enter China to launch a missionary effort. But there he sickened and died 
and was buried in 1552. The Portuguese abandoned this base after they occupied Macau 
in 1557.371 
The population of the Wuyi regions was then depleted by Qing attempts 
to eradicate the pirate Koxinga (also known as Zheng Chenggong), by forcing all 
Chinese living on the coasts of Guangdong and Fujian to move approximately 15 
kilometers inland. Enacted in 1661, this drastic policy forced people to leave the 
Wuyi areas altogether. Many never returned. After this order was rescinded in 
1683, the population of the Wuyi regions began to grow again, by natural increase 
370 Ibid. 
37 1 Despite this proximity to the ocean and early foreign contacts, the Wuyi region never became a center 
even for local, much less coastal or international trade. See Georg Schurhammer, Francis Xavier, his life, 
his times, Vol. 4 (Rome : Jesuit Historical Institute, 1973-1982), 640-644. 
183 
and through official edicts. In 1683, government attempts to resettle vacated 
coastal areas attracted groups of Hakka settlers.372 
When competition for land became fierce in the mid-nineteenth century, these 
differences ignited a major conflict between the Cantonese speaking people and their 
"guests." But during the eighteenth century, tensions had yet to arise, and as late as 1732 
a government official, Tao Zhengzhong, inspected barren lands in Xinning and assigned 
Hakka to come and till them. During the course of the Qianlong (1736-95) and Jiaqing 
(1796-1820) reigns, 30,000 Hakkas came from Huiyang, Chaozhou, and Jiaying 
prefectures and settled in southeastern Xinning. Now two subethnic groups, Cantonese 
and Hakka, prevailed in the Wuyi region. Most of the Cantonese speakers inhabited the 
eastern lowlands and coastal areas. The Hakkas were concentrated in the rugged 
hinterland but they were also spread over the highlands of Enping, Xinning and 
mountainous Heshan counties. 373 
Lineages were common throughout the region among both the Cantonese 
372 For a detailed description of this see Chapter 2 above. 
373 The local history records that wherever a Hakka moved onto a piece of land there soon developed 
a teeming village. For a period of 110 years, from 1740 to 1850, the Hakka population within 
Xinning county grew from a handful to 300,000 . This was a little less than half of the total 
population of the district, and they lived off one-third of the tilled lands in the district . See Chixi XZ, 
8.5b-6a. 
184 
and Hakka population. Variations in lineage size and structure in different areas 
seem to have reflected mainly ecological influences rather than subethnic 
distinctions. Lineages located in lowland areas where the population was dense, 
arable land relatively extensive, and riverine and sea trade active tended to be 
larger and wealthier on the whole than those situated in impoverished highland 
areas containing poorer resources and fewer communications. Powerful lineages 
were concentrated more in the prosperous lowlands and coastal centers of Xinhui 
and Jiangmen. These lineages varied widely in size, ranging from several hundred 
to ten thousand or more. Single lineages of this size sometimes occupied entire 
villages. In some areas stretches of villages were occupied by people with the 
same surname and same lineage or related lineages. For example the Huang Ji 
(Wong) clan was numerous in the north half of the Xinning county. One source 
listed 163 Huang hamlets, or approximately 40 percent of Huang hamlets in the 
enti· re county. 374 
Highly organized armed feuds reportedly developed first among the 
flourishing lineage concentrations especially in Cantonese-settled areas like 
374 Most of the prominent lineages had their headquarters along the lower and middle courses of the Tan 
River. See Kaiping XZ, 64, 90. 
185 
Kaiping and Jiangmen. From these Cantonese areas the feuding fashion also 
radiated to the less productive coastal areas of Xinhui and Xinning and inland to 
the Hakka-dominated hinterland of Enping and Heshan as well as to other 
mountainous sections of the region over the later half of the eighteenth century. In 
the more thinly populated highland areas there were fewer concentrations of 
powerful lineages. Hence local feud patterns appear to have developed on a less 
extensive scale. Severe conflicts between the Punti and the Hakka people were 
still not reported in areas of mixed habitation in the Wuyi.375 
Sometime during the eighteenth century, however, the balance tipped, 
and Wuyi people began to feel insecure on the amount of land they had to farm. As 
the Hakka population grew, the acquisition of lands in the Wuyi by the Hakkas 
also increased. Hence, simultaneously with the advancing tide of Hakka 
settlements, the tide of Punti settlements receded. This situation in the Wuyi united 
the Punti tenants and the Punti landlords against the Hakkas: the Punti tenants 
were in competition with the Hakkas for land, and the Punti landlords were 
alarmed at Hakka assertiveness on acquiring of more lands. Hence, the Puntis and 
375 Zhang Fenqian ~lllfiJ, "Kejia minxi zhi yanhua ~*~*z~{-t," Taiwan Wenxian lf~X~ 13.4 
(Dec. , 1962), 60. 
186 
the Hakkas were increasingly divided into two hostile camps. These hostilities 
finally exploded in 1854 in the Wuyi, and this communal war, called the Hakka-
Punti War, continued to devastate the community for the next decade. 
2) The Hakka-Ponti Conflicts in Retrospect 
Hakka Controversy 
Let us now proceed to the sources of conflict among locals. Local societies 
were derived from two subethnic lines-the Hakkas and the Puntis who were the 
principal cause of conflict within Wuyi local societies in the mid-nineteenth century. The 
myths and realities of migration and cultural diffusion in south China have 
always held a fascination for scholars with an interest in ethnic relations. In the 
nineteenth century, few who lived in the prosperous towns and villages of the 
Pearl River delta would have questioned their identity as legitimate members of 
the Chinese empire. They might have acknowledged differences in dialect, living 
habits, and marriage customs among themselves, but all traced their origin from 
the central plains (zhongyuan g=i !Jr-: ). 376 
376 See Wolfram Eberhard, Social Mobility in Traditional China (Leiden: E. J. Brill, 1962); Luo, Kejia 
yanjiu daolun . 
187 
Both the Hakkas and the Puntis were all obviously Han. Yet a sense of 
subethnic identity clearly existed among them.377 Particularly after the subethnic 
group of the Hakkas, another Han Chinese group who arrived during the early 
modern period from farther north in search of better lands, many Cantonese felt 
rather besieged in their own homelands. Since ethnic consciousness is often 
strengthened when a group must compete with others for land, it is not remarkable 
that the Cantonese were more likely to think of themselves as a people apart than 
are many subethnic groups among the Han. 378 The sense of the Cantonese as 
somehow different from other Chinese cuts both ways. In a chapter titled "We and 
They" in a book describing the early contests of south China with the Europeans, 
Frederic Wakeman states, "Chinese from other provinces often stressed the 
uniqueness of the Cantonese. They were considered uncommonly bellicose, and 
they were often looked down upon as serpentine Yeman !l!Hf (savage southern 
barbarians) whose habits were bizarre and uncouth." The Cantonese responded 
377 The Cantonese-Hakka conflict raged mainly within the territory of the Wuyi . It is convenient to 
regard the Puntis as an indigenous Cantonese group for this writing, although many Punti surnames 
claim by their legends that they too had entered Guangdong from the north. Similar legends of 
common origin are recorded among the Hakkas, and are given of the Dan by the Puntis. See Chapter 
Il.1-2 above. 
378 Ross Terrill goes so far as to describe "Cantonese chauvinism." See Ross Terrill, 800,000,000: 
The Real China (Boston: Little, Brown and Co., 1971), 68. 
188 
with an intense, though defensive, pride in their origins.379 
For a long time, the Hakkas were described by foreigners, even by 
anthropologists and linguists, as if they constituted a non-Han minority nationality 
within China. In 1946, the Encyclopaedia Britannica was still calling them a people 
of "doubtful" origin who "may be related to the Burmese and Siamese" and contrasting 
them with the "true Chinese. "380 Hakka people were also consistently stereotyped by 
other Han Chinese groups. They were often described as minority people who liked to 
live in the mountains, sing special 'mountain songs' and lamentations and tended to be 
poor by comparison with other people living in the plains. Certainly it is the case that 
there are differences between the Cantonese and the Hakkas. If Cantonese and Hakka 
people were principally distinguished by their mutually unintelligible dialects, 
they could also be differentiated by other features. 381 Hakka women, for example, 
did not bind their feet and favored distinctive hair styles. There were also religious 
differences in regard to festivals, temple names, and deities. Distinctions in dress, 
379 Wakeman, Strangers, 57. 
380 Encyclopaedia Britannica, a new survey of universal knowledge (Chicago and London: Encyclopaedia 
Britannica, Inc., 1946). 
381 The Hakkas are somewhat taller than other South Chinese groups. See S.M. Shirokogoroff, 
Anthropology of Eastern China and Kwangtung Province (Royal Asiatic Society, North China Branch, 
Extra Vol. IV: Shanghai, 1925). 
189 
architecture, and eating habits likewise led Cantonese and Hakka to associate with 
their own kind and to discriminate against others. The Hakkas preserved detailed 
traditions which claimed for them an origin in the north of China. These traditions were 
collected and recorded in the nineteenth century by European missionaries who were 
working among them.382 Since then the subject has been discussed at greater length by 
modern scholars.383 
A great deal of the speculation about the Hakka people is colored by the 
term "Hakka" of which usage led some Chinese to look upon the Hakkas as if 
they were really outsiders-that is, as if they were a non-Han people. 384 In reality, 
however, the Hakkas are but another of the subethnic groups among the Han. 385 The 
382 Hamberg, "The Visions ofHung-Siu-Tshuen"; E. J. Eitel, "An Outline History of the Hakkas"; Charles 
Piton, "On the Origin and History of the Hakkas," China Review 2 (1873); R. Lechler "The Hakka 
Chinese," Chinese Recorder 9 (1878); George Campbell, "Origin and Migrations of the Hakkas," Chinese 
Recorder 43 (1912). 
383 Luo Xianglin, Kejiayanjiu daolun; Hsieh T'ing-yu, "Origin and Migrations oftbe Hakkas," Chinese 
Social and Political Science Review I 3 (1929); J .A.G. Roberts, The Hakka-Punti War; S. T. Leong, 
Migration and Ethnicity in Chinese History; and Nicole Constable, Guest People, Hakka Identity in China 
and Abroad (Seattle and London: University of Washington Press, 1996) 
384 Leong, Migration, 63-67. 
385 Luo Xianglin iUl=ft, "Kejia yuanliukao" '.:g'.*~vlt:~ in Kejia yanjiu '.:g'.*lilf,'E, eds. Zhang Weidong 
iHHtrli: and Wang Hongyou .:E~it, 3-5 (Shanghai : Tongji daxue chubanshe, I 989); Luo, Kejia yanjiu 
daolun; S. T. Leong, "The Hakka Chinese ofLingnan: Ethnicity and Social Change in Modem China," in 
Ideal and Reality: Social and Political Change in Modern China, 1860-1949, eds. David Pong and Edmund 
S. K. Fung (Lanham: University Press of America, 1985): 287-326 As the Hakkas were al so granted Guest 
Families (kehu '.:g'.p: legally registered status) status, they used this to identify and call themselves Ke ( '.:g'.) 
or Kejia('.:g'.*), or in their language "Hakka." The Punti (Mandarin: bendi :;$::i'IH) are the earliest Han 
Chinese inhabitants of a region. This term is often translated as "natives" or "indigenous inhabitants." 1n 
190 
usage appears to represent nothing more complex than an archaic way of 
distinguishing the newcomer from the original inhabitant. There is historical 
precedent in various parts of China for using the twin terms "host" (zhu ± ) and 
"guest" (ke ~) to distinguish an original population from more recent arrivals. 
One Song dynasty census, for example, divided the population of Guangdong 
Province into two basic categories: the hosts, or native non-Han peoples, and the 
guests, who were at the time the Han settlers. 386 
Usage of the term "Kejia" to mean an ethnic group or language was 
never seen in books or papers before the seventeenth century.387 County gazetteers 
of many "pure" Hakka-dwelling counties did not refer to their communities as 
being Hakka before the nineteenth century. In Northern and Western Guangdong, 
the Hakka speech group was named "Ngai" (derived from the Hakka word ngai 
meaning "I" or "me"). In Guangxi, they were called the "New People (xinren 
Guangdong the Punti are Yue(~ Cantonese speakers); in Fujian they are Min(~ Hokkien). In both cases 
this distinguishes the Chinese who were already settled in the area from the Hakkas and other later arrivals. 
Often the Hakkas regarded themselves as the Punti in their heartland, Meixian (m~)-
386 Herold J. Wiens, China's March toward the Tropics (Hamden: Shoestring Press, 1954), I 05-06, 269-72. 
387 Leong, Migration, 64; Hu Xizhang i5JH!H~, et al., Kejiafenghua ~*}X\,{.t (Guangzhou: Guangdong 
renmin chubanshe, 1997). 
191 
fr A)" or the "Newcomers (lairen *A)."388 In western Fujian, some Hakka 
speakers think that "Hakka" are those who live in Guangdong sharing their 
language.389 The label "Hakka" did not begin to appear in county gazetteers of 
mixed Hakka dwellings counties (particularly in the Wuyi) until around 1700. 
Also, "Kejia" was never mentioned in historical documents of Hakka speaking 
areas in Taiwan or Sichuan, instead, the people are recorded as "Yueren" or 
"Guangdongren."390 The term "Kejia" has never been found in genealogical 
records of Hakka speakers.391 On the contrary, identification as "Chinese" or 
Tangren ~ A was always emphasized.392 In other words, Hakka-dialect 
communities are not found to have called themselves Hakka until quite recently. 
They did not use the name "Hakka" by themselves, but they were named or told to 
388 Mantaro J. Hashimoto, The Hakka Dialect: a Linguistic Study of its Phonology, Syntax and Lexicon 
(Cambridge: Cambridge University Press, 1973), l. 
389 David Prager Branner, Problem In Comparative Chinese Dialectology: the Classification ofM iin and 
Hakka (Berlin and New York: Mouton de Gruyter, 2000), 42-44. 
390 Cui Yingchang ~~ §\, Sichuan Fangyanyu Bashu wenhua ll!l)il1f"8 Ej E.~--Jt:.{t (Chengdu: Sichuan 
daxue chubanshe, 1996). 
39 1 Chen Ganhua ~T$, "Yuexi Lianjiang de Kejiaren .l!!fjffiJjjIB"J~~CI-..," Kejiayanjiujikan ~%:ii)f~ 
~ftJ , Vol. 9 (1996), 55-60. 
392 According to recent studies on the Hakkas, there were no people calling themselves "Hakka" and no 
ethnic group existed as Hakka before the end of the Ming dynasty in Southern Jiangxi Hakka speaking 
areas. The Pengmin Wl~ of northern Jiangxi, whose ancestors were also composed largely of Hakka 
speech group, did not get named Hakka, nor do they have any sense of being "Hakka." See Leong, 
Migration, 109. 
192 
name themselves in this fashion. The development of a positive Hakka identity is a 
process that took several decades; the first sign of it is appeared in early 
nineteenth century by a few Hakka gentry, but then the conflicts later in the 
century, particularly through the Red Turban Rebellions and the Hakka-Punti War, 
confirmed and expanded that sense of identity. 
Hakka History 
Concerning the origin and history of the Hakka people, there have been many 
theories based on different studies. In many parts of South China, these "guest people" 
are still treated as outsiders and intruders even though everyone now concedes that they 
are Han Chinese. More interestingly, the Hakka people today adopted the term for a 
guest as their ethnic name call themselves "Hakkanyin" (Kejiaren ~*A: Guest 
Person) to differentiate themselves from the neighboring people. Most Hakka people 
today follow the conclusions of Luo Xianglin, a Hakka historian, who claimed that the 
Hakkas are the "noble pure blood Han from the Central Plain," and have been 
migrating to the South since the third century in five waves.393 Luo categorized the 
393 Mostly based on the Hakka genealogies, Luo reconstructed five stages of migration of Hakka people: (1) 
A.O. 311-873, starting in Eastern Jin dynasty when the northern China was occupied by northern 
193 
ethnic groups in Guangdong into two major types, Han and non-Han. He argued 
that the extant sub-groups within the Han included the Cantonese, the Chaozhou, 
and the Hakkas. In terms of blood purity, the Cantonese were least pure, for they 
inherited the blood of many non-Han groups including not only the Miao , Yao, Bai, 
Dan, but also Negroes, Persians, and Arabs . The blood of the Chaozhou was purer 
than that of the Cantonese, but even that still mixed substantially with that of the 
She and the Dan minorities. Among the three Han ethnic groups, Luo asserted, the 
blood of the Hakka was comparatively the purest, although it still might possibly 
have assimilated with the blood of the She. 394 The conclusion was, therefore, that 
in Guangdong it was the Hakka that were closest to the Han. 
Luo 's claim about the origin of the Hakka serves as the spiritual support of 
the Hakka-speaking people today. However, this view is now challenged by many 
modem scholars. Here I do not note this fact to nullify Luo 's theory, which has been 
barbarians; (2) 874-1276, investigating the impact of Huang Chao Rebellion on the migration of Hakka 
people to western Fujian and southern Jiangxi ; (3) 1276-1682, the migration to eastern Guangdong during 
the Song-Yuan transition; ( 4) 1682-1867, the Hakka's movement into the Sichuan, Jiangxi, Fuj ian, 
Guangdong and even to the Taiwan partly because of the population increase in this period; and (5) 1867 to 
present, discussing about the migration of the Hakkas from eastern Guangdong to other parts ofGuangdong, 
Jiangxi, Taiwan, other parts of Asia, Pacific Islands, America and Africa. See Luo, Kejia yuanliu kao, Chap. 
2. 
394 LuoXianglin, "Guangdong minzu gailun '*~~~it," Minsu zhoukan ~{ftmflJ 63 (special 
issue), 1929. 
194 
considered elsewhere. 395 The most important weakness of this theory, however, lies in 
the derogatory view of the neighboring groups, a Hakka version of racism. Luo also 
failed to keep his promise of a detailed account Cantonese and Hoklos, which makes a 
contrast of the three groups impossible. 396 In addition, the notion of "Ke" or "Kehu 
~J=i" in the Jin dynasty at the fifth century as the origin of the Hakka suggested by 
Luo could not have been correct, because "Kehu," as a term simply indicating 
immigrants, was ubiquitous in China. There is also evidence to suggest that at least the 
first three of the five waves of migration were common to all Southern Chinese. By 
comparing the genealogical records of the Hakkas, Cantonese, and the Hoklos, Chen 
Zhiping (1997) found out that these speech groups shared common ancestors as late as 
the Ming dynasty. He further argued that Hakkas were not only similar to other non-
395 Recently, there are objections to Luo's theory, such as Fang Xuejia m~* who suggests that the Hakka 
people are sinicized and the so-called Hakka-migrations are pure inventions. According to the Fang's study, 
there was indeed some northern ethnic people migrated to south, however, the number of people migrated 
from north was really small and there was no evidence of massive migration from northern to southern 
China from the historical records. The mixture of both Yue and northern culture brought by these small 
numbers of migrants eventually produced the special Hakka culture. The primary ancestors of the Hakkas 
were Yue people in the south not very few northern immigrants. Zhang Kuangyu 1lO't-$, on the other hand, 
contends that the Hakka people is a mixture of Han and the She 11;- minority. Therefore, the definition of 
"Hakka people" is itself controversial. See Fang Xuejia m~*• Kejia yuanliu tan 'ao ~%~mt~~ 
(Guangzhou: Guangdong gaodengjiaoyu chubanshe, 1994); Zhang Kuangyu 1IO't-$, "Lun Kejiahua de 
xingcheng it~%i! Ef{]*f5x:," QinghuaXuebao m~?¥:ffl, Vol. 25.2 (1995), 161-180. 
396 Luo himselfalso admitted in his book, Kejia y uanliu kao (p. 43), the descriptions in the Hakka 
genealogies are fragmented and give very vague information. He also mentioned the error of listing few 
counties in Fujian as pure Hakka areas but actually are pure Hoklo areas. See Chen Zhiping ~3z: 5¥-, Kejia 
yuanliuxinlun ~%~tmffiit (Nanning: Guangxijiaoyu chubanshe, 1997), section 2.23 . 
195 
Hakka in their origins, but also not much different in culture, customs and civilization. 
Therefore, the notion of a pure Hakka line is more of an imagined identity based on racial 
prejudice than a fact. 397 When, then, does the "Hakka" become a named identity for the 
Hakka speakers indicating all of the Hakka speakers? Let us examine the general stages 
of Hakka settlement in the Wuyi during the Qing period to answer the above question. 
Hakka Settlements in the Wuyi 
Like other Han Chinese residents of southeastern China, Hakka speakers traced 
their ancestral roots to areas of central China. However, as relative latecomers to the 
south, they were often forced to settle in less desirable upland regions populated by 
groups of non-Han Chinese such as the She, with whom they intermixed and 
intermarried.398 Alternatively, they pushed on to the flatlands, where they became tenant 
farmers for earlier settlers like the Cantonese. 
According to S.T. Leong (1985), the dispersal of Hakka sojourners over a wide 
territory of southern China and across different economic regions mitigated against the 
391 Ibid. 
398 Mary S. Erbaugh, "The Hakka Paradox in the People's Republic of China: Exile, Eminence, and Public 
Silence," in Guest People: Hakka Identity in China and Abroad, ed. Nicole Constable (Seattle: University 
of Washington Press, 1996), 210. 
196 
development of a uniformly shared Hakka identity. The expansion of the Hakka speech 
group began after the middle of the Ming dynasty as a result of economic cycles. They 
began to move to Northern Guangdong, Huizhou, Boluo, and also south to Chaozhou 
speaking regions. This resulted in local disputes, but as the government officials regarded 
them as Chinese, they ended up having the right to stay there. Because of the xiangyue 
~B*J system,399 they were always grouped together and they built up their own villages, 
which resulted in Hakka dialect islands. A xiangyue was a collection of neighboring 
villages (but not necessarily all the villages in one area) operating as a unit and 
sometimes holding a common estate. Often the complexes of xiangyue were formed in 
the Hakka villages to protect their interests against the natives. But these Hakka speech 
399 Xiangyue, a term with a long history in Chinese local government and administration, appears first in 
Northern Song, when a Confucian scholar set out a scheme for a kind of self-governed village in which 
country people were to promote among themselves morality, education, social solidarity, and mutual aid. 
The plan seems to have enjoyed some vogue in Ming times, but the early Qing rulers took over the term to 
give it a new meaning: Xiangyue became a public lecture system by means of which the masses were to be 
indoctrinated with the political ethics of Confucianism in a kind of adult education. Yet by the nineteenth 
century xiangyue had once again undergone a transformation, a lecture system having developed into a 
framework of state control to the point where xiangyue was sometimes taken to be synonymous with baojia 
and lijia. On the other hand, a contrary process was also at work, moving xiangyue back some distance 
towards the kind of self-government which had originally been conceived under its name. It is on record 
that in places in Guangdong province the heads of Xiangyue assumed roles of leadership in such a way as 
to take command of local affairs. Moreover, xiangyue were used as a framework for organizing militia for 
local defense. Through these xiangyue, the Hakkas in Guangdong developed extra-local organization for 
interethnic conflict later. For the studies on the xiangyue, see David Faure, The structure of Chinese rural 
society : lineage and village in the eastern New Territories, Hong Kong, (Hong Kong ; New York : Oxford 
University Press, 1986); James Hayes, The Rural Communities ofH ong Kong: Studies and Themes (Hong 
Kong: Oxford University Press, 1983); and P.H. Hase, "The Mutual Defense Alliance (Yeuk) of the New 
Territories," Journal oft he Hong Kong Branch oft he Royal Asiatic Society 29 (1989): 384-88; Hsiao, Rural 
China, 66. 
197 
immigrants had not named themselves "Hakka," nor were they recognized as being so, 
right after their settlement in the area. 400 
As discussed in Chapter II above, the Hakkas who settled in the northeastern 
part of Guangdong were always ready to leave the area because of immense population 
pressure. In 1670 the Hakkas took the golden opportunity for expansion by accepting the 
Manchu government's invitation to reclaim and settle empty lands on the coast. The Qing 
even granted the new immigrants money and seeds if they moved to the coast. To 
distinguish themselves from the aborigines, they could register themselves as "Ke" or 
"Keji ~ii" so that they could receive subsidies. Thousands of people from Huizhou, 
Chaozhou, and Jiaying prefecture, which comprised the central core of the Hakka 
speaking people, responded because their homelands were too crowded.401 They were 
eager to be "Keji" because they would receive benefits.402 This was the beginning of 
Hakka settlements in the Wuyi. A memorial by the governor of Guangdong in 1730 set 
400 There is no mentioning of "Hakka" as minority in the county gazetteer. Instead, many Hakka 
expressions were listed as part of the dialect of the county. When one searches into the historical documents, 
one cannot find the term "Hakka" as a name for a group before the eighteenth century. Even in the 
recognized center for the Hakkas, Meixian identified themselves as Chengxiangren ~~A as late as the 
middle of the eighteenth century. Hakka as a dialect was first mentioned in a book describing the dialects of 
Xinfeng ffi .:_ county in 1756. See Jiaying ZZ, 7.86b; Zhou Rijian fiiJ B 19!, Xinfengfangyanzhi iFJT$1f1§~ 
(Guangzhou: Guangdong gaodengjiaoyu chubanshe, 1990);Leong, Migration, .64-64. 
401 Chixi XZ, 8.3-4. 
402 Kaiping XZ, 20 .21 a. 
198 
these forth in some detail: 
The waste land of Kai ping and Enping is also abundant, probably in excess of ten 
or twenty thousand mu. At the moment, five thousand odd mu have been 
surveyed out from the whole. On account of the sparseness of the local 
population, although money-subsidies for the opening up of the waste land are 
available in the treasury, very few are willing to take them. Your minister has 
instructed the wealthy merchants and gentry families to recruit indigent people 
from Huizhou and Chaozhou, and to provide them with cottages, food and 
provisions, and working capital. For a unit of five households, they should be 
organized into the baojia system, and their names entered into the local 
government registry and be provided with a hundred mu of land. Further, in view 
that they had come from distant places and although at the moment they have 
secured a livelihood by farming, they might be apprehensive of future disputes 
over the title of the land with the landlords, it is necessary to reassure them by 
long-term planning . ... At the present moment, the poor people from the two 
prefectures ofHuizhou and Chaozhou who have come to cultivate and registered 
at Heshan already number more than three hundred households. Their number is 
rapidly increasing, for as news of their decent treatment travels back, more 
people are to be persuaded to join them ....4 03 
It is obvious that the Hakka speaking migrants came as tenant farmers under 
the joint patronage of the state and of the local men of substance.404 It was not the 
migration of a people en masse. The first immigrants were the marginal men who could 
403 Huangqing zouyi ~;ij!f~~, comp. Qinchuanjushi ~J ll lls- ± , (Taibei: Wenhai chubanshe, 1967), 30.16-
18 
404 Liu, Bei y iwang de, 112-114. 
199 
not scrape together a living in their homeland went first.405 In the initial stage of their 
arrival, they settled interspersed among the Puntis who provided them with land, 
provisions, implements, and even housing. This arrangement was mutually profitable. As 
the Punti landlords were short of manpower, this increased the value of their land, and it 
afforded sustenance to the indigent Hakkas. There was no irresolvable conflict of 
interests.406 The Punti tenants had pre-empted the best land available, they were the 
overwhelming majority, and were moreover backed by strong kinship and territorial 
orgam•z at1• 0ns. 407 
In an early stage of the Hakka settlement, their dispersed dwellings were 
consolidated into more compact settlements. This development was necessitated by the 
pooling of manpower and resources for irrigation as well as for defense, and was 
reinforced by the fact that people invariably seek their own kind for congenial 
companionship. Once they had established a foothold in any locality, their more remote 
kinsmen in their homelands who felt the urge to emigrate in search of a better life were 
405 Myron L. Cohen, "The Hakka or "Guest People," Dialect as a Sociocultural Variable in Southeast 
China," in Guest People , Constable, 51. 
406 The Puntis were at first quite pleased to see the arrival of the new immigrants because they could not get 
enough manpower to work in their fields. The Hakkas and Puntis formed a kind of"symbiotic" relationship 
for a few decades. See Ibid., 53-54. 
407 Ibid., 50. 
200 
likely to follow the line of least resistance and come to join them. Thus throughout the 
eighteenth century and the first half of the nineteenth, a trickle of Hakka immigration into 
these districts had continued uninterrupted. The Puntis were still the dominant majority, 
but the initial extreme of numerical and economic disparity had diminished to a 
considerable extent.408 
The additional acreage of land to absorb the ever-swelling agrarian Hakka 
population could be conditioned by several factors. The most important of these involved 
the system of land tenure in the Wuyi. As examined in the Chapter II above, much of the 
land reclamation efforts of the Pearl River Delta areas were done by the elite, the locally 
powerful and wealthy households. Many plots were owned not by individual landlords 
but by associations (tang) formed by groups of land investors and represented by gentry 
members, as well as schools and temples, which came to control large tracts of reclaimed 
lands.409 The concentration of land in the hands of lineages reached its peak in the 
districts around Guangzhou, but in the Wuyi the tendency was still very marked. In 
408 Barth, Bitter Strength, 25 . 
409 According to the Robert Eng's study, an institutional land such as lineage land accounts for as much as 
50% of cultivated acreage in the Pearl River Delta. See Robert Y. Eng, "Institutional and Secondary 
Landlordism in the Pearl River Delta, 1600-1949," Modern China, Vol 12 No. 1 (Jan. 1986), 4. 
201 
Xinning it reached 50% and in Enping, Kaiping and Heshan, 40%.410 Not only were these 
lineages large landlords, but they also tended to be absentee landlords, because 
undeveloped, unoccupied, or mountainous regions were remote and sparsely populated. 
Tax evasion provided another cause for prevalent absentee landlordism. With increasing 
urbanization and the flight of the gentry and the wealthy to towns and cities since the 
early Qing, it appears that fewer and fewer landlords took a direct hand in reclamation 
and management after establishing their claim with the government. Zhang Zhidong in a 
later period reported that most landlords rented out to tenants, who would pay no rent for 
several years while they were reclaiming the land.411 In the Wuyi, according to Eitel 's 
report, the Hakkas were nearly everywhere to be found "either dispersed in small villages 
and paying top-soil rent to the Punt is, or congregated in larger villages ...." 412 
This kind of land tenure system along with the local situation developed 
in the Wuyi region was very favorable to Hakka tenants and was directly 
responsible for the flourishing social organization and self-direction evident there 
410 Chen Han-seng, Landlord and Peasant in China: A Study oft he Agrarian Crisis in South China (New 
York: International Publishers, 1936), 34. 
411 Zhang Zhidong *z.~, Zhang Zhidong quanji *z.~1:~ (Shijiachuang: Hebei renmin chubanshe, 
1998), 28.16a. 
412 In many places the Hakkas depended on land which they rented from Punti absentee landlords. See E. J. 
Eitel, Notes and Queries 1, 49. 
202 
at the mid-nineteenth century. The vastness of the reclaimed lands made it difficult 
for town-based lineage estates to extend their reach. They relied on tenant 
contractors who managed the outposts and secured harvests . These functionaries 
were mainly local Hakka farmers who rapidly acquired enough power to carve out 
their own base.41 3 Some of these Hakka tenants retained a portion of the rented 
land after reclamation and cultivated it with hired short-and long-term labor. Often 
these tenants had subleased most or all of the land they rented. In effect, these 
primary tenants functioned as landlords and were recognized as such in popular 
parlance as "secondary landlords. "414 
During the Qing effort to pacify Taiwan, the coastal people were ordered to 
move inland in 1653. Consequently, many of the alluvial fields in the Wuyi were also 
abandoned and the owners were forced to rent to the secondary landlords at rock bottom 
4 13 Junjichu zouzhe lufu il!ttlt€~ffl~i1J : Nongmin yundonglei :zy. ~iEz;/J~, 3.166.8947.14 (July 28, 
1864), Microfilm 2296-2298 . 
41 4 Secondary landlordism closely resembled the "two lords to one field" (yitian liangzhu - EE ivi±) 
system that had become increasingly prevalent throughout China since the Ming-Qing transition. Under 
such a system, the landlord owned the subsoil whereas the tenant possessed the topsoil, which amounted to 
the right of permanent tenancy and which could be sold without reference to the landlord. The main 
difference between secondary landlordism and the "two lords to one field" system lies in the terms of 
tenancy, which was for a fixed period for the former and permanent for the latter. Typically, the length of 
the tenure for the Guangdong secondary landlords was ten years. See Evelyn S. Rawski, Agricultural 
change and the peasant economy ofS outh China (Cambridge: Harvard University Press, 1972), 19; Eng, 
"Institutional," 3-37. 
203 
rates. Tenants could also co-opt income from the offspring alluvia (zisha T fj;- ) that were 
unknown to the owners of the mother alluvia (musha -ffJ:lP). Thus, some of the Hakkas 
settled in the Wuyi were lucky enough to find tracts of suitable land on which they could 
establish permanent settlements. The western Xinning was such an area, and the Hakka 
settled in large numbers on the low hills there.415 In settlements of this sort they were 
largely independent economically from the Puntis.416 Around the mid-eighteenth century, 
Hakka tenants came to control land in newly-established villages near the sea as well as 
in distant mountainous areas. The Hakkas purchased land which the Puntis had already 
brought under cultivation. With their lower standard of living, they had an advantage over 
the Punti tenants in bidding for land leases from Punti landlords.4 17 
The Hakkas have a reputation for industry, and as their women played a role in 
providing agricultural labor, they had a certain advantage over the earlier settlers. They 
first served as "hired laborers, but through industry and thrift acquired most of the land of 
their masters."418 For instance, in Kaiping, after the Yongzheng period, the Hakkas were 
415 Chixi XZ, 8.37a. 
416 Cohen, "The Hakka," 53. 
417 F.O. 228/213 , Robertson to Bowring (July 16, 1856). 
418 Barth, Bitter Strength, 25. 
204 
said to have flourished to such an extent that they were able to buy up cultivated fields 
and houses.419 After arriving in the Wuyi as poor farmers, the Halek.as had in the space of 
a few generations greatly improved their position. One indication of their progress can be 
seen in the fact that by the time of the Hakka-Punti War, there were Hakka gentry in all 
the districts which were involved. In Xinning in particular the Hakkas prospered. No less 
than thirty Hakka gentry played a role in the war, and of these, eight had obtained their 
degree by purchase. From this one may conclude that some Hakkas at least had become 
landowners on a considerable scale.420 
The Source of Conflict 
Subethnic considerations had of course played a role in migration and 
settlement patterns in the Wuyi, and violence did on occasion follow subethnic lines. 
Before 1800, however, the relative sparseness of settlement and the common challenges 
of wresting and reclaiming polder farmland in many instances led to cooperation between 
the natives and migrants. Prior to the nineteenth century, Guangdong had been 
characterized by both frequent violence and subethnic social organization, but not 
4 19 Kaiping XZ, 17.4a. 
420 Xinning XZ, 12.10-11 b; Chixi XZ, 8.7 . 
205 
consistently by subethnic violence within the Wuyi areas. Except for the eastern part of 
Guangdong (especially Jiayingzhou), the Hakkas were concentrated in widely separated 
areas. Even though some Hakka communities joined together to fight against their 
neighboring Punti communities, there was no centralized plan for anti-Punti resistance 
before the Red Turban Rebellions. No unified command or Hakka leadership ever 
emerged. That is to say, the Hakkas were never a united ethnic group. 
In the eyes of the Punti people, the Hakka-speaking people were no more than a 
group of migrants from the eastern mountains. They did enjoy a "honeymoon period" of 
peace and symbiosis at the beginning when the Punti landlords needed labor.421 However, 
the situation worsened when the Puntis felt that these migrants threatened their survival. 
More and more Hakka people bought land from their Punti landlords and built their own 
towns. The acquisition of new holdings brought in its train new settlements, while 
conversely, loss of the land holdings brought decline and eventual termination of the old 
settlements. Hence, simultaneously with the advancing tide of Hakka settlements, the tide 
of Punti settlements receded. In 1832, Long Qirui depicts this change in quite absolute 
terms. "The Hakka people have been hired to cultivate the lands. For several generations, 
421 Cohen, "The Hakka," 54. 
206 
their descendants multiplied so greatly that they [outnumbering the Punti] reverse the 
dominance of the Punti."422 Some Punti people began to flee from their own homes and 
left the whole area for the intruding Hakka people.423 A Protestant missionary described 
this kind of phenomenon in the nineteenth century Guangdong: 
It is an impressive picture, that of the old Punti villages, enclosed by ivy-covered 
brick walls and towers half in ruins and surrounded by ditches, and all about the 
rows of white houses of the Hakkas, into whose hands the great part of the lands 
have fallen, soon after the appearance of a Hakka house inside the walls of a 
Punti village, the Punti disappeared completely. It is, however, in the main, a 
peaceful acquisition by diligence and thrift, though in some instances trickery 
and force are said not to have been entirely absent from the methods of 
occupati· on ....4 24 
Naturally the displaced Punti people did not entertain charitable opinions of the 
Hakka migrants who displaced them. In the face of the tangible perils of a common 
enemy their sense of community of interests was aroused. They could not help viewing 
with suspicion and fear the compact Hakka settlements with their teeming thousands.425 
One way in which the Puntis expressed their resentment was by blocking the attempts of 
422 Long Qirui, Jingdetang wenji (Taibei: Yiwen, 1970), 2.24a. 
423 Mai ~ family in Shenzhen is a good example of this. See Xiao Guojian ffiOO -o!, Shenzhen diqujiazu 
fazhan ~ :ijlitt!!. IR*.nj'.&,M (Hong Kong: Xianchao Shushi, 1992). 
424 D. Oehler, article in Milton T. Stauffer, ed., Christian Occupation of China (Shanghai, 1908), 351. 
425 Chixi XZ, 8.32b-33. 
207 
recent Hakka immigrants to register with the local government. Without local registration, 
a man was not eligible to participate in the local civil service examination, for which each 
district had a fixed quota. As registration matters were in the charge of the taxation 
secretary of the district magistrate, to be managed at his discretion, the Puntis were 
generally successful in thwarting the Hakka immigrants' attempts to register. They were 
apparently not willing to admit the latter into the avenue for academic success and social 
prestige, which would have afforded their rival a further lever for dislodging or 
undermining them. With the consent of the central government, the impasse was resolved 
by compiling a separate registry of residence for the Hakkas of these districts and 
assigning them a special quota of shengyuan 1=.~ for each examination, over and above 
that in force. 
Under normal circumstances, Hakka-speaking shengyuan were barred from the 
prefectural and district schools which dispensed stipends. The separate quota for the 
Hakka speaking immigrants in the Wuyi was established late, the earliest record was in 
1789.426 Whereas the intention of the government in providing special quotas for 
particular categories of persons appears to have been intended to satisfy their desire to 
426 See Xinning XZ, 12.10-11 b, Chang Chung-Ii, The Chinese Gentry (Seattle: University of Washington, 
1955), 81. 
208 
participate in the system and to avoid a possible source of grievance between them and 
the dominant group in which they were living, in practice the policy proved ineffectual 
and short-sighted. Referring to the special quotas for the Pengmin in Jiangxi, Chang 
Chung-Ii comments that "this policy may have sometimes increased the psychological 
barrier between these groups and others under special quotas.',427 This would also seem 
true of the special quotas provided for the Hakkas, which left neither side satisfied, and 
created friction between the two groups.428 
On the Hakka side, they could, with some justice, say that the quotas were at 
the best only a grudging recognition of the extent of their settlement in the Wuyi. 
Whereas in some cases, for example the Pengmin mentioned above, the special quota 
gave the candidates who were eligible a better chance of success than those who 
competed in the regular quota, this was not true in the case of the Hakkas. At the most the 
special quotas only provided them with eight places, out of a total of nearly four hundred 
places of the prefectures of Guangzhou and Zhaoqing. Also, there were restrictions 
427 Chang, The Chinese Gentry, 8 I. 
428 But institutionalizing the linguistic groupings actually hardened the communal lines, for it erected a wall 
between the gentry classes who were the natural leaders of their respective communities and who should 
have been the most promising element in reconciling and uniting the divided populace. Because of the 
nature of this dispute, it played an important part in placing the gentries of both Hakka and Punti in the lead 
when hostility between the two groups developed into the war. See F.O. 228/213 , Robertson to Bowring 
(July I 6, I 856). 
209 
placed on the special quotas which had been granted to them.429 On the Punti side the 
idea that the Hakka should be allowed to take part in the examinations was itself 
distasteful, and every attempt by the Hakkas to increase their quotas was greeted with 
opposition. This opposition of course came from the Punti gentry, for it was they who 
found their position in the examination system threatened when it was amended to make 
provision for the Hakkas. Then they were faced with a challenge to their social position 
with the emergence of a local Hakka gentry.430 
By the middle of the nineteenth century, the Hakkas constituted between a 
quarter and a third of the total population in some districts of the Wuyi; in Xinning, for 
instance, their settlements were alleged to cover a third of the territory. Some of their 
consolidated settlements sprawled across county boundaries and actually enabled them to 
enjoy a numerical superiority over the Puntis in some restricted localities. One of these 
localities lay in the junction between Xinning, Kai ping, and Enping, and another was 
astride the borders of Kaiping and Heshan. Nearly every large Hakka community could 
boast of a few shengyuan in its midst, who, like their counterparts elsewhere in China, 
429 Chang, The Chinese Gentry, 80-81. 
430 Enping XZ, 14 .11 b. 
210 
held the reins of power in village-level politics.431 
The numerical increase in the Hakka population was accompanied by an effort 
to expand and secure the land which they were farming. Such activities on the part of a 
population definable in linguistic terms led to the formation by both Hakka-speakers and 
Cantonese-speakers of common dialect groups in order to pursue interests relating to the 
securing of agricultural holdings. 
In the early nineteenth century, partly in response to Hakka examination 
successes in Guangdong province, and partly in reaction to struggles taking place 
between the Hakkas and other groups, a few Hakka intellectuals also began articulating 
positive statements of a unified and historically constituted Hakka culture. S. T. Leong 
(1985) identifies the first spokesman for the Hakka ethnic group as the scholar Xu 
Xuzeng ~ Afl, ~.4 32 Xu, as a Hakka leader, began to shape the term "Hakka" into a totem 
43 1 Leong, Migration, 75-77. 
432 Leong identifies the first consciously "ethnic" statement articulating the origins and "ethos" of the 
Hakkas as appearing around 1808, in a lecture presented by a scholar named Xu Xuzeng (~Afl. iit). Many 
ofXu's views, and those of other isolated Hakka spokesmen of the nineteenth century identified by Leong, 
were reiterated by Hakka intellectuals during a period of strong Chinese nationalism after the turn of this 
century. Certainly by the 1920s, with the emergence of international Hakka associations and the visions and 
rhetoric they produced, Hakka ethnicity became the basis for the action. However, the puzzle arises why 
"Hakka" is never used before the Qing dynasty, if Hakka was a glorifying term to represent "true Chinese 
traits." See S. T. Leong, "The Hakka Chinese of Lingnan: Ethnicity and Social Change in Modem China," 
in Ideal and Reality: Social and Political Change in Modern China, 1860-1949, eds. David Pong and 
Edmund S. K, Fung, 287-326 (Lanham: University Press of America, 1985), 303-308. 
211 
suggesting that Hakka people were the true Chinese from the North who had migrated 
southward in the twelfth century to avoid Mongol rule. He "introduced all the themes that 
would be repeated tirelessly in Hakka ethnocentric writings ... of the claim to 
Chineseness that was equal, if not superior, to that of other Chinese. "433 
The Punti responded with an intense, defensive, pride in their origins. 
Their dialect (Cantonese) was more like ancient Chinese than was Mandarin, the 
Puntis pointed out; they were purer representatives of Chinese culture, they 
asserted, since the north had long been defiled by barbarian influences culminating 
in Manchu conquest; they were more observant of the family duties prescribed by 
the Confucian tradition, since the lineage system was far stronger in south China 
than in north China.434 In this framework the Puntis could present themselves as 
the most faithful bearers of the Great Tradition and looked down upon the Hakka 
migrants who had fallen away from the ancient norms. 
The Puntis often tended to dismiss the Hakkas as an uncultured and poverty-
433 Then this idea got supports from many Hakka scholars who reached a maximum momentum by Luo 
Xianglin. See Ibid., 303.; Xu Xuzeng, "Fenghu Zaji ~ml~~c," in Kejia Shi/iao huipian , Luo Xianglin, 
297. 
434 Punti literati argued that compared with the dialects of many other provinces, Cantonese came 
closer to Sui and Tang dynasty pronunciation, because most Guangdong people migrated to the 
south from central China in as early as the Tang dynasty. See Wakeman, Strangers, 57; Panyu XZ, 
6.16b. 
212 
stricken people. Punti scholars also commented on the Hakka dialects in negative 
terms. In the 1820 edition of Zengcheng county gazetteer it is stated: "More and 
more Hakkas have moved into the county. They have not changed their village 
dialect, which is so noisy that you can tell without asking them their hometowns 
that these people have come from other places. "435 
Yet despite this long period of conflict between the Hakka speakers and the 
Puntis, even by the end of the Qing it is unlikely that all of the Hakka speakers of 
Guangdong had any notion of their ethnic consciousness. Even Hakka ethnic 
assertiveness in the Pearl River Delta "in the nineteenth and early twentieth centuries 
should not be constructed as an all-Hakka phenomenon. The 'Hakkas' were segmented by 
regional contexts of time and place, and some had more reason than others to assert their 
ethnicity."436 Certainly there had been land disputes between the Hakkas and the Puntis in 
the Wuyi, but most struggles seem to be locally focused and were isolated events. 
Nonetheless, the symbiotic relationship between Hakka tenants and farm 
laborers and the Punti landlords was upset during the mid-nineteenth century. As 
we have seen in Chapter II above, armed conflict was not confined to the Wuyi, 
435 Zengcheng XZ, 2.29a. 
436 L eong, M 1. grat.w n, 63 . 
213 
and fights between subethnic groups were a regular feature of the Pearl River 
Delta as well. There was, however, an eccentricity to the society of Wuyi that, in 
an unquantifiable but very real way, created a predisposition toward the formation 
of Hakka ethnic solidarity through intensive violence and lawlessness. This 
peculiar feature was, as we shall see, the byproduct of government action. 437 
Hakka-Punti Disputes 
By the late eighteenth century the Hakkas in some part of the Wuyi, 
especially in Xinning, not only had the necessary population but had also 
developed to the point where Hakka inhabitants had started to look to an end to 
dependence on wealthy Punti families. By approximately the early nineteenth 
century the Wuyi region started to have enough Hakka gentry leaders for it to have 
had some access to the local government, which would have facilitated any moves 
to throw off dependence on the Punti landlords. Finally, by the early nineteenth 
437 Disputes between new comers and local inhabitants are no big deal in the past history of China. This 
happened almost everywhere and every century. Sometimes it resulted in bloodshed. But normally, there 
are three outcomes: ( 1) The incoming group has either higher culture, large number of people or higher 
political status. The local people become assimilated into the incoming immigrants. (2) Opposite to ( 1) , the 
incoming people are of small number, poorer or less educated. They learn the way of life of the local 
people. (3) Both group are of similar number, education background and political status. The incoming 
group will maintain their culture in their new home. The third model is less common and is always the 
source for ethnic dispute. "Hakka" belonged to the third type of immigrants when they moved to the Wuyi . 
214 
century, at least some Hakkas had enough wealth to act as local leaders.438 Since 
the late Qing some of these rich Hakka tenants retained a portion of the rented land after 
reclamation and cultivated it with hired short- and long-term labor. The majority of these 
tenants had always subleased most or all of the land they rented. They found collection 
from tenants easier than supervising workers. Often the subletting of land did not stop at 
one level; tenants of secondary landlords in tum might sublet to other tenants, and so on 
down the line.439 
However, this kind of land tenure system could prevail only when the 
fertility of the soil was high or land was abundant; the rent charged the ultimate 
tenant must have been astronomical to support several levels of landlords and 
intermediaries.440 Unlike fertile alluvial fields in the Sanyi areas, however, the 
Wuyi regions did not have sufficient land. As previously indicated, the Wuyi region 
was traditionally a land-hungry area. The local gazetteer pointed out that the land-
438 The wealthy Hakka areas were those from which the men who had invested in the reclamation 
had come Kai ping, Heshan, and Xinning. These areas seem to have produced many of the early 
Hakka leaders. The Hakkas in these areas formed the Confederation to fight against the Red 
Turbans later. 
439 Eng, "Institutional and Secondary Landlordism." 
440 Majority of theses tenants had always subleased most or all of the land they rented. For instance, as 
many as a five layering of landlordism could be found in the delta. See Ibid. 
215 
population ratio reached a saturation point by the third quarter of the eighteenth century. 
The only remaining wasteland in the northwestern comer of the region was filled by 
Hakka migrants before the turn of the century.441 The swelling population during the 
eighteenth century made the situation worse and the local economy was thus heavily 
strained. Population had increased dramatically over the course of seventeenth and 
eighteenth centuries. For instance, in 1838 the population of Xinning county reached 
196,972, an increase of 1,440 percent since 1657.442 The tremendous population 
increase of the Puntis and the Hakkas brought the two peoples into closer physical 
proximity, and intensified conflicts between the two hostile groups over the issues 
of precious land titles and all important tomb sites became more and more frequent. 
These signs of overpopulation that had become apparent by the end of the eighteenth 
century caused elites and local officials to worry about possible social disorder.443 
At this point, some of the Hakkas, who were beginning to dominate the 
economic life of the Wuyi, came to be blamed for having contributed to the pauperization 
of the Punti tenants, many of whom became so impoverished that they turned to banditry. 
441 Kaiping XZ, 20.21 a-21 b. 
442 Taishan dilizhi, 21-22. 
443 Xinning XZ, 14.28a. 
216 
The poor Punti were jealous of the affluence of the wealthy Halek.as. A more concrete 
source of resentment was the alienation of Punti lands. The Punti demanded the return of 
land alienated to outsiders, usually the Halek.as. They looked upon the land as a sort of 
ancestral heritage, which had been illegitimately taken away from them. In practical 
terms, they found it increasingly difficult to produce enough for subsistence as their land 
shrank to ever smaller dimensions.444 The Punti refuges who had migrated into the Wuyi 
recently, escaping from the disorder of other parts of Guangdong, were in particularly 
dire straits, barely eking out a living from their miserable land. They would open 
marginal tracts of polder farms, but often they had to endure suffering from the floods 
and droughts. Many apparently were squatters or tenants, and they were usually much too 
poor to be in a position to fleece the Hakkas.445 Resentment toward the Hakkas 
occasionally boiled over into riots in which a Punti mob would attack the Hakka 
community, beating up the villagers and plundering their houses. On the one hand, there 
was widespread Hakka resentment against the presence of the newly arrived Puntis in 
their midst, when they regarded as having caused their own impoverishment. On the 
444 Liu, Bei y iwang de, 114-115. 
445 Junjichu zouzhe lufu, 3.166.8947.29 (Mar. 13, 1866), Microfilm 2383 . 
217 
other hand, both the Hakka and Punt i tenants harbored a deep hatred of the Punt i absentee 
landlords whom they accused of frequently subjecting them to oppression and 
harassment. 446 
Conditions were ripe for a violent upheaval in the Wuyi region. A severe 
drought in the year 1851 was followed by one of the worst recorded floods the following 
year. Consequently, an inevitable famine that began in 1852 added to discontent in the 
community.447 Although an outward quiet prevailed in the area until the year 1853, it was 
the calm on the eve of a storm. The local residents must have sensed the uneasy peace in 
their villages, where people felt a sullen lethargy caused by hunger, but filled with a 
potential for violence which could explode at any moment. Easily cultivable land had 
been exhausted. Therefore the young peasants, especially Punti peasants, many of them 
under-employed, sought membership in the Tiandihui.448 They were forced to cultivate 
miniscule plots too small to provide a living, sometimes serve as tenants on the land of 
the Hak.kas, or leave the land to engage in banditry on lakes and rivers. Many 
446 Ibid.; Chen Kun, "Yuedong," 400; Liu, Bei yiwang de, 79-82, 98-99. 
447 Xinning XZ, 14.18a; Kaiping XZ, 21.3a. 
448 At this time the regions were already infested with the Tiandihui groups under the leaderships of Tian 
Fang, Zhang Zhao ~taiJ, and Liang Peiyou. See Su Fengwen ff AX, Gufei zonglu !Nl1U~jit 
(Taibei: Xiangsheng chubanshe, 1975), 3.3a, 9.a; Liu, Bei yiwang de, 115. 
218 
impoverished Punti tenants left the area to join the secret societies of Guangzhou area.449 
In this shift migrant workers and rural poor, especially Punti tenants, formed 
new solidarities aimed precisely at upper-strata Hakka peasants and others in local 
society who, in their eyes, had become exploitative violators of communal morality. 
Many of these Punti peasants felt the need to band together either for self-protection or to 
take revenge. The Tiandihui thus provided a convenient means for such purposes. It was 
only natural that the Tiandihui would eventually become involved in a direct 
confrontation with the Hakkas. Indeed, several months before the outbreak of the Red 
Turban Rebellions, several members of the Tiandihui in the Wuyi area were tortured and 
killed for having illegally entered a Hakka farm to collect firewood.450 
449 The key feature of the Punti tenants was that in addition to agriculture and cotton craft production, 
various fonns of(below subsistence) wage labor fonned a major or primary source of family income. By 
the early nineteenth century, the interconnected effects of tenancy, the small size of holdings, and the 
overcrowding created a situation where agricultural and craft income could no longer cover family 
subsistence. Hence they need for additional income. To obtain this, males from the communities along the 
river often worked as dockers, loaders and laborers on the commercial boats the piled the Pearl River and, 
seasonally, as coolie-transporters and agricultural laborers in the Sanyi districts to the north. Although these 
people signified declining economic status of Punti peasants, by routinely placing peasant-workers beyond 
the boundaries of the village and local marketing networks, and in freeing them from the direct constraints 
of the Wuyi's powerful resident landlords, it created a tactical latitude which family fanners did not enjoy. 
Scholars have identified the importance of this type of tactical mobility in the emergence of rebellion. The 
greater geographical mobility which it engendered not only exposed Punti tenants to a world beyond the 
village, but especially, in the years prior to the planned rising, it enabled them, either directly or indirectly, 
to gain and transmit news of the Red Turban advance. See Ibid.; Eric. R. Wolf, Peasant Wars oft he 
Twentieth Century (New York, Harper & Row, 1969); Elizabeth J. Perry, Rebels and Revolutionaries in 
North China, 1845-1945 (Stanford: Stanford University Press, 1980). 
450 Zhao, "Hongbingjishi," 103 ; EnpingXZ, 14.7b. 
219 
The Tiandihui activity began to increase after the late eighteenth century in the 
Wuyi, fed by the growing number of men forced by financial hardship to leave their 
native places and seek employment elsewhere.451 As economic conditions worsened over 
the course of the nineteenth century, the Tiandihui became bolder. In the mid-nineteenth 
century, bandits from Xinhui and Xinning often held meetings in broad daylight to solicit 
new members. Several hundred would gather at a crossroad, post guards with guns, and 
encourage local peasants to join. Farmers left responsible for their own protection joined 
the Tiandihui for 300 copper coins to avoid onerous extractions. During these times, even 
subclerks, yamen runners, and district magistrates in local government became Tiandihui 
members.452 
The fact that these Punti tenants joined the Tiandihui immediately after the 
initial rising of the Red Turbans indicates that there had already joined members of the 
Tiandihui in the Wuyi. In this situation the Tiandihui served as a reservoir to receive any 
Puntis who were dislodged by the Hakkas. The Tiandihui was in turn becoming not 
merely a secret society but also the champion of the Punti tenants in the Wuyi. As an 
451 China Mail (May 10, 1855 and May 22, 1856); Guangzhou FZ, 82.1O b; Kaiping XZ, 21.3a. 
452 F.O. 931 .1512. A Report on the Sanhehui (l 854/ l 855);Liu, Bei y iwang de, 68-78. 
220 
organization championing the interest of the Punti tenants, the Tiandihui extended its 
influence quickly into Punti districts. This partly explains why its members were mostly 
found in the Punti villages of the Wuyi.453 Looking for the cause of the Red Turban 
Rebellions in the Wuyi in 1854, Qing officials invariably concluded that the social-
economic-ethnic conflicts between the Hakkas and the Punti were one significant 
factor. 454 
Animosity between the Hakkas and the Puntis continued to grow. E. J. Eitel 
wrote that "the Hakkas ... congregated in large villages, and then continually fighting 
with the Puntis for the ownership of the hills and fields occupied by them." 455 Had the 
local authorities intervened in time, the feuds could have been nipped in the bud. But the 
bureaucrats at the time had already degenerated into a very corrupt group. The Wuyi was 
a poor region. Intervention in the local feuds on the part of the authorities meant the 
necessity of sending troops , which, in turn, meant the expenditure of money not provided 
for in the regular budget of the local government. This extra money would have to be 
453 Zhao, "Hongbing jishi," I 00-06; Junjichu zouzhe, 3 .166.894 7 .18 (April 25, 1866), Microfilm 2288. 
454 [bid.; Guangzhou FZ, 129.18a-20a. 
455 E. J. Eitel, "Ethnographical Sketches of the Hakka Chinese," Notes and Queries on China and Japan I 
(6), 65-67. 
221 
obtained from the superior officials, but none of them wanted to be exposed to the charge 
of maladministration by reporting that their jurisdictions were infested with bandits or 
disturbers of the peace. Naturally the best thing for them to do was to pretend to be 
ignorant of the feuds ; they hoped, by so doing, that they could escape the punishment for 
maladministration by transferring to another post after the end of their term of office.456 
All these factors-the quarrel between the Hakkas and the Puntis, the disturbances 
created by the Tiandihui, and the negligence on the part of the provincial authorities 
toward the area- produced the volatile situation in the Wuyi before the coming of the 
Red Turbans. 
3) History and the Emergence of Hakka Ethnic Consciousness 
Red Turbans in the Wuyi 
In the year 1854, natural calamities of flood and famine recurred.457 The 
price of rice jumped so high in the Wuyi that "innumerable people starved to 
death. "458 Worse still , it was during this year that the Red Turban rebels spread into 
456 Long, Jingdetang werifi, 4.12b. 
457 Liu, Bei y iwang de, 73. 
458 Kaiping XZ, 2 l .3a-3b. 
222 
this area.459 Unlike the frequent clan feuds , these were great province-wide armed 
rebellions which were tinged with secret society activities; all were obliged to take 
sides either for the rebels or for the government. The principal efforts of the Red 
Turbans in 1854, however, were still directed at establishing control over the Sanyi region 
where the Wuyi Punti peasants regularly sought employment. All during the 1854 then, 
through their movement on the river and in and out of the accessible areas of the Sanyi, 
the Punti tenants from the Wuyi areas were in a position to acquire and transmit 
knowledge of the Tiandihui.460 By late 1854 news of the rebel victories had in fact 
spread throughout the Wuyi.461 The poor and oppressed, particularly Punti peasants, 
joined the rebel ranks, while most of the Hakka clans under conservative leadership 
prudently remained loyal to the government.462 Consequently, the Hakkas were more 
immune from suspicion of collaboration with the rebels. More and more the Punti 
459 Tan Zuen u'liffl.,~, "Xinhui jingbian shilue ffi~~5HRH~," in Hongbing qiy i, 953-958. 
460 Many of the Tiandihui members in the Sanyi areas especially in Sbunde were the Puntis drawn from the 
areas of the Wuyi, particularly from the Xinhui and Xinning. See Donghua quanlu (daoguangchao) 
Jl!UiHtt.k (:li:3/ttl1) , (Taibei, 1963 reprint), 51 .5b. 
461 Xinning XZ, 14.18b-2 lb. 
462 Hakka people, by the middle of the nineteenth century, were accustomed to defending their local rights 
against outside demands, and the ambiguities of state policy toward migration reinforced their expectations. 
The early Qing government encouraged Hakka migrants clear wasteland by providing them seeds, tools, 
cattle, and tax exemptions. In each case, local Hakka community members had reason to expect official 
support for their mobilization against the hated outsiders and naturally more royal to the officials. See 
Junjichu zouzhe, 3.166.8947.18 (April 25, 1866), Microfilm 2288; Zhao, "Hongbingjishi," 104-113. 
223 
absentee landlords were worrying about the situation in the Wuyi where the Punti 
peasants joined the rebel force. 463 
As more of Punti peasants joined the rebel forces, the Red Turban 
movement tried to gather momentum in this area to form a genuine peasant uprising 
against the Manchus. The frantic government officials, thereupon organized a 
mercenary force out of the Hakka populace and positioned it against the advancing 
rebels. This was a clever scheme designed to exploit the old animosity between the 
Puntis and the Hakkas. Hakka and government forces succeeded in checking the 
rebel advances in this area. Thus, in the villages of the Wuyi area new resentments 
against Hakkas were added to older, more deeply rooted ones at the precise moment 
when the Red Turban Rebellions were gaining ground in a nearby region. Conflict 
between the Hakkas and the Puntis may have made the state's job easier-it could use 
one faction of people against another. Ironically, in reacting to these threats by forming 
Hakka militia, the local government created a new organizational field for further 
disturbance in the region. 464 
463 Punti landlords were expressing their frustrations when they were attacked by the Punti peasants in 1855. 
See Zhao, "Hongbingjishi," 108-111 ; Yoo, Kundae, 270-274. 
464 Junjichu zouzhe, 3 .166.894 7 .18 (April 25, 1866), Microfilm 2288; Chixi XZ, 8.36. 
224 
In Enping, when the revolt broke out, the district magistrate had summoned 
both the Punti and the Hakka militias to defend the district city, and to destroy rebel 
forces within their area of his jurisdiction. The Hakka braves distinguished themselves in 
the north of the district and captured and executed many rebels.465 In Kaiping rebel forces 
had converged on the district city from all directions, and the district magistrate, faced 
with defeat, had committed suicide. A Hakka named Cai Yunsheng ~ti*, who had 
played a role in the defense of the city, also killed himself rather than face capture.466 
For local Hakka leaders, the escalating local disorder that accompanied the 
rebel invasion was a major preoccupation. Feuds between residents of the Wuyi had been 
a problem for years. At this time, the local Tiandihui lodges took advantage of other 
ongoing feuds to enhance their position. In the midst of this chaos, the Tiandihui began to 
gain strength throughout the area, recruiting new members to their ranks. Crowds of 
Punti residents, variously identified as local bandits, peasants, and tenants joined the Red 
Turbans and looted rich households. Again the targets were often, though by no means 
always, selected on more specific grounds than wealth.467 In some outbreaks, punishment 
465 Chen, "Yuedong," 71 I; Enping XZ, 14.7b. 
466 Liu, Bei y iwang de, 98101 ; Kaiping XZ, 21 .3b. 
467 Liu, Bei y iwang de, 68-79. 
225 
was reserved for all those rich households who, relying upon their influence to collect 
their rents, had incurred the long-standing resentment of the Punti peasants and those who 
were harsh in rent and debt collection. Most often these targets were the Hakka secondary 
landlords.468 
Some of the fiercest and most extensive conflicts took place in Heshan county, 
where the numbers of Hakka villages were roughly the same as the Punti. The Red 
Turbans captured the city and overran much of the countryside. The Hakkas at Yunxiang 
~~~ offered them stout resistance. A wealthy Hakka, whose youngest son had been 
kidnapped and murdered by the Punti Red Turbans, vowed to avenge his murder and put 
up all he was worth to combat them. With the assistance of other like-minded people, he 
raised a private army and put it under the charge of two military juren • A, Ma 
Conglong and Zhang Baoming ~-~.469 .~{ft~ 
On October, 1854, the Red Turbans made a foray into Heshan but were 
intercepted by the Hakka forces and defeated. The Hakkas were then entrusted by the 
468 Mai Bingjun ~)!H1J, "Heshan Maishi zupuji yutujishi lunlue ~w~~~tf &~OOi-c.¥~~," in 
Hongbing qiyi, 1014-1018. 
469 In actual fact, as often happens, the quarrel over the unfortunate boy was only a comparatively minor 
incident in a long-standing rivalry between the two communities, but it was sufficient to trigger the 
outbreak of hostilities. The real reason was undoubtedly the struggle for " living room" in the Wuyi between 
the Punti and the Hakka. SeeJunjichu ZL, 3.166.8947.18 (April 25, 1866), Microfilm 222288, 2292; 
Kaiping XZ, 2 l .4a-b. 
226 
Qing force with the principal role in the recapture of the city. There is no record of a 
militia bureau being formed at the time, but Ma Conglong received some form of warrant 
to do so from the Governor-general, Ye Mingchen, who had deliberately used the hostility 
between the Hakkas and the Puntis as a means of containing the revolt. Consequently the 
Hakkas participated in the extermination campaigns and their mission took them farther 
outside the confines of Heshan.470 
A similar situation developed in Xinning, where a Red Turban force under Tan 
Yashou IDf52~ and Yu Zhaobiao #;?)~~ crossed the Xinhui river at Dihai J.:kre and 
attempted to attack the district city. The magistrate, when organizing the defense, called 
on a Hakka named Yang Jinlan m~?I to collect Hakka braves, and with the assistance 
of braves collected by the Punti gentry, the Red Turbans were driven back to 
Changsha.471 
As in other areas where the Hakkas and the Puntis lived side by side, the 
relationship between the two groups had often been antagonistic. It had further 
470 The makeshift solution to which local officials most frequently resorted was to pit rival groups against 
each other, "using the people to control the people" ( t-J.§;; 'itU .§;;). the government's fonnal recognition of 
anned Hakka bands as a legitimate militia in the Wuyi and the awards bestowed on these forces for their 
stand against rebels during the Red Turban Rebellions provide a notable instance. See Ibid.; Chixi XZ, 8. 6b 
and Kaiping XZ, 21. 4a-b. 
471 Chixi XZ, 8.6b. 
227 
deteriorated in the atmosphere of disorder in the years since mid-nineteenth century.472 
But it was the disorganization brought about by the Red Turban Rebellions that really 
brought into the open the antipathy existing between them. The relaxation of rural control 
probably accounted in large part for this, but a particular grievance among the Punti was 
the way in which the Hakka braves summoned to deal with the Red Turban rebels had 
taken the opportunity to make attacks on them. In 1856 the British consul at Guangzhou 
described: 
At the time of the troubles in that quarter nearly two years ago they readily 
responded to the call of the authorities to form themselves into train-bands, and 
whilst the Pentis [Punti] showed little disposition to assist the government, the 
Hakkas retook from the Triads the District cities of Ganping [Enping] and 
Kaeping [Kaiping]. They have of course obtained by this behavior a high 
position in the opinion of the authorities, and they now avail themselves of the 
power to avenge their old wrongs.473 
Therefore, the first major clash between the Hakkas and the Puntis occurred 
while the Red Turban Rebellions were still at their height. After having helped in the 
relief of the district city of Enping, the Hakkas in the north of the district were said to 
472 Guangzhou FZ, 82.30b. 
473 F.O. 228/213 , Consul Canton to Bowring, (July 16, 1856). In the summer of 1854, a similar account of 
the sequence of events happened in Enping where the Hakkas, after their success over the Red Turbans, had 
begun to despise the Punti and to plot against them. See Enping XZ, 14. 8a. 
228 
have armed themselves in preparation for an attack on the Puntis.474 Elated by the 
success of their service to the government, the emboldened Hakkas began to take 
held of lands belonging to the Puntis. The Puntis were naturally very alarmed by 
Hakkas, land grabbing, which exacerbated their deep seated and long-standing 
bitterness toward the Hakkas. 475 On November 1854, the Hakkas were alleged to have 
attacked a Punti village in Kaiping.476 Two months later there was a series of clashes 
between Hakka and Punti villages on the Kaiping and Enping border. On one occasion 
Punti braves from Kaiping crossed the border to assist a village in Enping which had 
been threatened by the Hakkas. On another, the Hakkas were reported to have collected 
together confederates from the neighboring districts of Heshan, Gaoming, Xinxing, 
Enping, Kaiping and Yangchun, in an attempt to destroy the Puntis in the area.477 
474 Enping XZ, 14.8a. 
475 In the atmosphere of growing suspicion, the long-standing grievances relating to land-ownership and 
payment of rent became increasingly bitter. Hakka tenant farmers, taking advantage of the disturbances and 
the authorities' reliance on them to assist in the suppression of the revolt, took the opportunity to withhold 
payment of their rents. While the Punti gentry might have accepted Hakka assistance in fighting the rebels, 
as landlords this action would have incensed them in particular. See Junjichu zouzhe, 3. 166.8947.18 (April 
25, 1866), Microfilm 2288; Ibid. 
476 Kaiping XZ, 21.4a. 
477 Under the guidance of the Hakka gentry, the Hakka villages of these counties gathered under the banner 
of the Hakka Confederation of Six Counties. The Hakka forces of this Confederation were called "Black 
Soldiers (Heibing ~A)." See Mai, "Heshan Maishi zupu," 1018-1019; Enping XZ, 14.8a-10a; Zhao, 
"Hongbingjishi," 127; Liu, Bei yiwang de, 104-107. 
229 
At this time the Qing authorities were still preoccupied with quelling the Red 
Turban rebels throughout the province, and had no time to interfere in the quarrel which 
was developing between the Hakkas and Puntis. In Kaiping the Punti gentry had at last 
organized a force of braves, and offered their service to the local government to quell the 
Red Turbans. The district city was finally recaptured on early January, 1855, but a body 
of Red Turbans held out in the south of the district for a further month.478 By then the 
rebels were in general retreat around Guangzhou and the prefectural city of Zhaoqing had 
been recaptured.479 
The defeat of the Red Turbans in Guangdong was followed by a "White Terror" 
in which thousands of persons were executed on suspicion of having been implicated in 
the rising.480 This was a turbulent period and often the normal procedure of judicial trial 
for crimes carrying the death penalty was suspended. Particularly in the Wuyi region, 
most of the people who perished in these summary trials were the Puntis-a trend which 
could hardly be viewed with complacency by the large number of the Puntis who had 
collaborated with the Red Turbans during their period of ascendancy. Given the reservoir 
478 F.O. 228/213 , Consul Canton to Bowring (July 16, 1856). 
479 Chen, "Gangcheng," 902-909 .. 
480 Ibid. 
230 
of hostility toward the Hakkas which had been building up for over a time, it took no 
great ingenuity for these men to convince their fellow Puntis that the Hakkas were 
capitalizing on the Governor-general's mandate to supplant them, that the Puntis as a 
group were in jeopardy and that something had to be done to avert the danger. Being 
well-armed and with a tradition of pugnacity nurtured in local feuds, they naturally 
resorted to force. Once fighting got under way, the bloodshed of kindred people and the 
spreading tales of enemy atrocities set in motion the self-perpetuating mechanism of 
suspicion, fear, and hatred. It tended to spread by contagion to adjacent areas where the 
kindred populations of the feuding parties lived.481 
The Hakka-Punti War 
The Hakka-Punti clashes, which had begun in the confusion caused by the Red 
Turban Rebellions, continued after the Red Turbans had been dispersed. The incidents 
were most frequent in Enping. The Hakkas again attacked the Puntis in the district and 
raided as far as the Yangjiang border, looting as they went in 1855.482 Punti villagers were 
481 Kaiping XZ, 2 l .4b-5 ; Chixi XZ, 8.6b. 
482 Chixi XZ, 8.8b. 
231 
forced to flee, and the whole area became involved. Both the Hakk:as and the Puntis from 
villages across the border of Enpingjoined in the fighting. There the Punti gentry leaders, 
Guan Weicheng ~~~ organized a "Mutual Security" hall (Lianbaotang ~~*~) and 
sent arms to help their fellows in the north.483 In 1856 the relationship between the 
Hakkas and the Puntis in these districts worsened, and from this year the so-called the 
Hakka-Punti War began.484 Lechler made a succinct narration ofthis event: 
The rebels ... found numerous adherents among the Punti people, whereas the 
Hakkas (there] remained loyal, and assisted the Mandarins against the rebels. 
This exasperated the Puntis very much, and they swore vengeance against the 
Hakkas . .. . The Puntis commenced hostilities with the Hakkas. The district of 
Heshan was the first in which disturbance broke out, and the fortunes of war 
were variously experienced on both sides, until finally the Puntis, being stronger 
in men and means, conquered the Hakkas in this and other districts, and expelled 
those who were not killed .. .. 485 
The most significant feature of the clashes between the Hakkas and the Puntis 
483 Guan Weicheng had played a leading part in the organization of the Punti militia that had been raised in 
the previous year to assist the magistrate in the recovery of the district city. This new organization, 
Lianbaotang, would appear to be a direct descendant of the previous one that was to play a role against the 
Hakka organizations. See Kaiping XZ, 21.5b-6a. 
484 Of course, this was not a war in the sense of an armed conflict between nations, nor was it, in ordinary 
terms, a civil war, for the Hakkas do not seem to have been intending a rebellion. The term "war" is used, 
because this is what to contemporary observers the struggle seemed to be. It had developed beyond the 
stage of feuds and had become a protracted conflict spreading over a large area. For the complete details of 
the Hakka-Punti War, see Liu, Bei y iwang de; Zheng Dehua ~ll~, "Guangdong Zhonglu Tulce Xiedou 
Yanjiu )1l}.I:! i:p ~ ±~~IMHiF-f ~," (Ph.D. diss., University of Hong Kong, 1989); Roberts, "The Hakka-
Punti War." 
485 Rudolf Lechler, "The Hakka Chinese," Chinese Recorder 9 (1878), 354. 
232 
was the way in which the disorder was showing signs of developing from inter-village 
feuding, which in the Pearl River Delta at the time was so common that it might be 
regarded as the rule rather than the exception, into subethnic feuds in which 
reinforcements were brought in across district borders. Now, those involved in the Punti-
Hakka conflicts had little freedom to choose between the warring parties. They 
automatically belonged to one or the other side by virtue of their speech. 486 
Fairly large raiding parties were launched by both sides at infrequent intervals. 
Destruction was horrifying whenever a village was taken by assault. In this initial stage 
of the war, both the Hakka and the Punti gentry began to take an active part in the 
struggle.487 In July 1856 fighting commenced on a larger scale encompassing Kaiping, 
Enping, and the whole districts ofXinning and Xinhui.488 At this time, the Guangdong 
provincial government at Guangzhou was engrossed in a diplomatic crisis with the 
British over the Arrow Affair, which was to usher in armed hostilities and eventually, a 
two-year occupation of the city by the combined British and French forces.489 With the 
486 Punti villages fought under the red flag and the Hakkas under the white flag. See Chixi XZ, 8.9b. 
487 Chi.xi XZ, 8.9a. 
488 China Mail (July I 0, 1856). 
489 J. Y. Wong, Deadly Dreams. 
233 
intervention of the provincial government ruled out, the local government authorities with 
their token military forces were helpless in the face of armed clashes of this nature. 490 
Initially, the Hakkas seemed to have the advantage, not only holding their own, 
but also harassing the Punti villages that surrounded theirs. In June 1856, they stormed 
the Punti stronghold of Songhai mountain (~ fB LlJ) in Enping, annihilating more than 
two thousand of their adversaries.491 Certainly, the principal battle-fields of the Hakka-
Punti War, particularly Xinning and Enping, were indications that the dispute had reached 
the stage where the two groups had come to regard each other indiscriminately as threats 
to their security and livelihood. However, many of the smaller Hakka settlements in the 
eastern region of the Wuyi were ill-prepared for a showdown and after some preliminary 
encounters were abandoned. Their inhabitants sought refuge in the bigger Hakka 
settlements or in hastily constructed quarters in the mountains. Thus, many different 
Hakka villages solidified into a bigger group based on their speech. The Puntis "being 
stronger in men and means, eventually defeated the Hakkas in this area and expelled 
those not killed."492 
490 Xinning XZ, 12. 16-2 I. 
49 1 Kaiping XZ, 21.5-6. 
492 Lechler, "The Hakka," 354; Xinning XZ, 10.11 b-16b. 
234 
It is hard to estimate the actual numbers of the Hakkas and Puntis who 
participated in the Hakka-Punti War, but for the Wuyi-wide concern, there is no doubt 
that the Puntis far outnumbered their adversaries. For example, the number of Hakka 
villages in Xinning county approximated those held by the Puntis, but as Punti villages 
were usually two or three times larger than Hakka villages, this still afforded them a 
numerical superiority of at least two to one.493 The local gazetteer narrates this: 
At this time there was plundering and looting, and seizing and murdering, 
without a day's respite. The Hakka villages were overwhelmed one by one, and 
those who escaped fled elsewhere. In those villages that had not yet fallen, the 
people, thinking the local people numerous and the Hakkas few, and fearing that 
it would be difficult to resist, also for the most part removed to Dongshan *L .lJ, 
Chishui ~7.K, Nafu m~, and Shenjing ~:fi:.494 
By the end of 1857 the dislodged Hakkas fled with their wives and children, 
and in the course of their flight were intercepted, massacred, or taken prisoner by the 
Puntis. The only Hakka settlement in the east that survived was the one Caochong l!f fqi, 
situated on a sparsely-populated peninsula.495 
After the full scale eruption in 1856, the events in the Wuyi were overshadowed 
493 The situation could not have been much different in the other districts. See Chixi XZ, 8.45, 10.11 b-16b. 
494 Chixi XZ, 8.15a. 
495 Junjichu zouzhe, 3.166.8947.25 (Oct. 25, 1867), Microfilm 2351-2355; Chixi XZ, 8.21 b. 
235 
by the Arrow War and the assault on Guangzhou. In the local histories it was the 
hostilities with the British and French that occupied the pages for the years between 1857 
and 1858.496 Information on the Hakka-Punti War is therefore meager. What there is 
suggests that fighting continued between the Hakkas and the Puntis. There was only 
desultory fighting, however with no major changes in the balances of power and no major 
incidents were reported.497 In 1860 and the first half of 1861 there was a lull in the 
fighting between the Hakkas and the Puntis, which offered hope that a settlement might 
be achieved. Some time early in 1860, the Guangzhou provincial government appointed a 
Punti gentry, Liang Yuangui ~7Ctf to mediate. Liang, together with the magistrate of 
Enping, called for cooperation between the Hakkas and the Puntis, and summoned Hakka 
leaders from Enping, Kaiping, Heshan and Gaoming to take part in discussions. The 
gentry of both sides established a committee, and began to discuss the question of 
restitution of property. 498 In May 1861 the provincial authorities issued instruction to all 
the Wuyi to bring cooperation between the Hakkas and the Puntis.499 
496 J. Y. Wong, Deadly Dreams. 
497 Chixi XZ, 8.22b; Nanhai XZ, 14.30. 
498 F.O. 228/344, Canton Intelligence Report (May I , 1863). 
499 Guangzhou FZ, 82.39a. 
236 
However, the truce came to an end just two months after their initial 
cooperation effort. According to one report, fighting began first in Gaoming, provoked by 
"some obscure circumstances."500 Once the news of fighting in Gaoming and Heshan had 
reached Kaiping and Enping, the uneasy truce in these counties quickly came to an 
end.501 In 1862 Hakka-Punti fighting was reported throughout the Wuyi region and even 
as far north as Gaoyao. According to a later commentator, it was at this time that the 
contest was at its height, and local governments were entirely suspended in several 
counties.502 No external factor can account for this sudden change, nor was there, as far 
as is known, any particular catalyst of events. It was rather that the feeling aroused in the 
previous years of disorder, and the revelation of the impotence of the authorities, led the 
gentry on both sides to believe that they could act without danger of interference. In the 
report from the Overland China Mail at the time also stated that after the first incident in 
Gaoming "the Punti villagers rose en masse to avenge the memory of 1854."503 In other 
words, popular hated of the Hakkas, especially of the role they had played in the Red 
500 Overland China Mail, (July 11 , 1863). 
501 Kaiping XZ, 21 .8a; Chen, "Gangcheng," 908-943 . 
502 China Mail (June 29, 1867). 
503 Overland China Mail, (July 11 , 1863). 
237 
Turban Rebellions, provided massive support for a renewed Punti attack on the Hakkas. 
After the wave of attacks, the Punti gentry asserted that they were the victims 
of a conspiracy formed by the Hakkas and branded the Hakkas bandits and rebels, and 
even characterized the Hakkas as not a Chinese race to justify their attacks on the 
Hakkas. 504 Within three years, the task of pacifying the "Hakka rebels" was completed. 505 
The Punti forces had begun to reduce the Hakka settlements in the Wuyi one by one. By 
the end of 1862, the Hakkas were outnumbered and driven towards the coast. The 
dislodged, dispossessed Hakkas became roaming bandits who descended in vengeance 
upon unexpecting Punti villages. There was no means for reconciliation, as the Hakkas 
only wanted liberty to live while the Puntis refused to be sheltered under the same heaven 
with them. 506 The desperate Hakkas then stormed and occupied the fortified town of 
Guanghai JJ:Wfi, west of Macao on their way to Caochong, the only surviving Hakka 
settlement in the whole district. 507 
504 This attack was even fostered in the county gazetteers by the Punti gentry. Thus, in the section recording 
the feuds , the Hakk.as were labeled as "fei " (bandits) in the gazetteers. See Xinhui XZ I 0. 7; Xinning XZ 
8.6b, and Gaoyao XZ 2.29b-30b. 
505 Kaiping XZ, 21.8-10. 
506 Eitel, "Outline History of the Hakkas," Rev. II, 163-164. 
501 Junjichu zouzhe, 3.166.8947.19 (April 25, 1866), Microfilm 2290-2293; Xin 'ning XZ, 14.31. 
238 
At this point the Punti gentry petitioned the provincial authorities to send troops 
so that these "Hakka rebels" could be settled by force. The Punti gentry already had an 
ally in the persons of the local government. Local enforcement, often undertaken by local 
Punti gentry or militias under the command of the magistrate or other imperial officials, 
partook of the local Punti cultural ethos, one which was often actively and vociferously 
anti-Hakka. Virulent anti-Hakka sentiment may be found in the memorials of many local 
Qing officials. 508 
The news of this outbreak of hostilities had also reached Beijing. In a decree 
dated May 6, 1862, the Governor-general was taken to task severely for having failed to 
report the clashes that had occurred between the Hakkas and Puntis in the Wuyi and the 
death of an official near Guanghai where the Hakkas had dared to capture the garrison 
town, an act tantamount to overt rebellion.509 For the first time the Hakkas were officially 
508 The powerful factor working in the Punti's favor was their accessibility to the government authorities. 
Clustered as they were in the environs of the district cities, they had the ear of the district magistrates, and 
also of the provincial and central governments, since the Puntis had a shengyuan 1:Jl quota several times 
greater than that of the Hakkas and consequently were better represented at these levels of power. By 
pressing their version of the story first and then constantly repeating it, they usually succeeded in 
prejudicing the minds of government officials, who generally had neither the opportunity nor the 
inclination to verify what they were told. Government intervention on behalf of the Pun tis was a crucial 
factor in the collapse of the Hakkas' resistance effort all along the line. See Ibid. , 14.26b; Enping XZ, 
14.22a. 
509 Chixi XZ, 8.26-27. 
239 
condemned for their activities; "they really have no regard for the laws."510 The Imperial 
government had just weathered one of its most serious political crises,5 11 and now with 
the Tongzhi Restoration, was in a mood to resume again the civil administration of the 
country. Under orders from the central government, the provincial troops of Guangdong 
were mobilized into two expeditions to chastise the "Hakka bandits." The Hakkas were 
massacred or driven into hiding. 512 
In September, 1866, the Governor of Guangdong, Jiang Yili ~ ~}~ sent 
eight thousands troops to Caochong to compel the Hakkas to give up their arms and 
disperse. To prevent further difficulties with the Puntis, two hundred thousand taels 
were set aside for distribution to the Hakka (eight taels to each adult and four to 
each child), with passes and protection to enable them to migrate to Guangxi , 
Hainan, and other regions where wastelands existed in abundance. The small district 
of Chixi was marked out from Xinning and set apart for the Hakkas to reside in 
1867.513 
510 Daqing Muzong Yi (Tongzhi) Huangdi shilu ::k:ffl~*l.lt (IE]~) £ WJ !f ~, (Taibei: Taiwan huawen 
shuju, 1968) 24.32b. 
511 The occupation of Beijing by the British and French forces. 
512 Chixi XZ, 8.40b. 
513 Ibid., 8.26b-27, 43b-44b; Junjichu zouzhe, 3.166.8947.19 (April 25, 1866), Microfilm 2290-2293. 
240 
Aftermath 
It is difficult to form a satisfactory estimate of the numbers who died, for 
there are no accurate figures for the population of the areas concerned; nor is it 
known how many moved away from the area during the course of the war. But as 
early as 1865, the memorial from the Governor-general estimated that the dead on 
both sides exceeded a million. 514 According to Luo Xianglin' s study, the killed, 
wounded and missing on both sides amounted to 500,000-600,000. 51 5 While there is 
no consensus on the casualties that resulted from the Hakka-Punti War, it is safe to 
say that the war did have severe and damaging consequences to the areas involved. 
Many of the population died, not from wounds, but from disease and starvation, the 
inevitable concomitant of a war in which large numbers of peasants were uprooted. 
Even twenty years after the war ended, a missionary who visited the Wuyi wrote 
that the "blighting effects" of the "Hakka rebellion" were still visible. 516 
The end of the Hakka-Punti War also led numerous officials to put forward 
programs intended to redress the problems that had caused the war and to initiate 
514 Ibid., 8.38a. 
51 5 Luo, Kejia yanjiu daolun, 3. 
5 16 B.C. Herny, Ling-Nam, or Interior Views ofS outhern China (London: S. W. Partridge, 1886), 86; 
241 
effective reconstruction measures. There was a flurry of activity, and a few changes did 
take place; but most of them were superficial and temporary; no fundamental change 
emerged in the relationship between the Hakkas and the Puntis. The war ended in a 
clear victory for the Puntis, for they succeeded in the almost complete expulsion of 
the Hakkas from the districts involved in the war. The principal exception to this 
was the Hakka enclave which was allowed to remain at Chixi. As far as the Hakkas 
now confined in Chixi were concerned, the long ordeal that they had together 
endured naturally increased their bond of Hakka solidarity. In other words , the 
effect of the establishment of the sub-prefecture was to reinforce the pre-existing 
subethnic spirit of the people by giving it a more strongly territorial grounding, as 
well as an official sanction.517 
Yet the settlement was a forced one, and it was received with 
dissatisfaction on both sides. On the Hakka side the feeling of resentment was 
stronger, and lasted much longer. Despite the efforts to appease them, in particular 
517 In 1912, the provincial government ofGuangdong, in a drive for economy and the rationalization of the 
local administrative system, decided to abolish the sub-prefecture of Chixi as a distinct administrative unit 
and merge it into the district ofXinning. Its principal reason for this decision was that Chixi 's fixed quota 
of seven hundred-odd taels of silver for land tax was far below the minimum cost for the operation of the 
new district government. But this contemplated move was strongly opposed by the Hakkas ofChixi who 
voluntarily raised the tax quota six fold to four thousand taels. This incident dramatizes the subethnic spirit 
of the Hakkas in Chixi. See Chixi XZ, 4.12. 
242 
by the provision of examination quotas in Chixi and in the districts to which they 
had been sent with assistance, the Hakka gentry found it difficult to accept the 
settlement. Soon after the end of the War in 1878, a Hakka named Lin Daquan 
;t.t~ 7':R. wrote a note in which he condemned the Guangdong administration' s action 
towards the Hakkas, and the injustice of their expulsion. Lin circulated a treatise, 
"keshuo ~~," on the Hakkas and attacked the local authorities for referring to the 
Hakkas involved in the feuds as "Hakka bandits" and for aiding the Puntis in driving 
them out. 518 His treatise was a glorified and romanticized image of their past history: 
"true sons of the Golden Age, descendants of the upper class of the Central Plain, 
with a speech that showed a clear affinity with the ancient speech of the north."519 
This coincided with the views of the gentry who had been expelled from the Wuyi , 
and on a number of occasions they attempted to obtain permission to return. In 
about 1875, some of the Hakka gentry who had moved to Guangxi contacted others 
who had been sent to Gaozhou and Leizhou and jointly presented petitions to the 
Censorate asking that they should be allowed to return to the Wuyi. Nothing came 
518 Luo, Kejia yanjiu dao/un, 5. 
519 Lin Daquan it~ ~. "Keshuo ~~," in Chayang sanjia wenchao 1k~ - ~Cti.'!J, ed. Wen Tingjing 
lK,l~~, 131 -35 (Taibei: Wenhai chubanshe, 1966). 
243 
of this, but the matter was raised again in 1885 and in 1905. However, the proposals 
were continuously rejected on the advice of local officials, who pointed out the 
possible dangers of such a course, and declared that the Hakkas had no cause for 
grievance. 520 
The attitude of condescension and contempt harbored by both the Hakkas and 
the Puntis toward each other remained essentially unaltered and even worsened. The 
Puntis were also resentful that the Hakkas had been allowed to retain the Chixi 
peninsula, small though that was. Now the period of armed conflict between the two 
groups took the form of a literary battle. The Punti gentry in the late nineteenth-century, 
could not refrain from writing that the Hakkas were ignorant, their practices stupid, and 
their natures fierce. Such attitudes were fostered in the county gazetteers that labeled the 
Hakkas as bandits (fei ~ ).521 Some gazetteers even described the Hakkas as Yao or Lao 
aborigines. 522 
The Hakkas, at times stridently, insisted that they belonged to the mainstream 
of Chinese culture through their own versions of the local gazetteers. Wen Zhonghe 
52°C hixi XZ, 8.54a; Kaiping XZ, 2 l .13a-b. 
521 Xinhui XZ I 0.7; Xinning XZ 8.6b, and Gaoyao XZ 2.29b-30b. 
522 GaoyaoXZ, 25.39b; YangjiangXZ, 1.4, 20. 
244 
ml. {qi f D, a Hakka Hanlin academician, made cases in the 1880 edition of Zhenping 
county Gazetteer (m2P-\i%;G, Zhenping Xianzhi) and in the 1898 edition of Jiaying Sub-
Prefectural Gazetteer (RrfJ. 'l'i'l ;G, Jiaying zhouzhi) that include extensive treatises on 
Hakka customs and rituals to identify the Hakkas with the Han Chinese. 523 Like the 
Cantonese, Wen Zhonghe also associated the Hakka dialect with the Sui-Tang 
speech in order to show the northern origins of the Hakkas. Wen concluded that "the 
Hakka dialect should be regarded as one of the oldest of Chinese dialects" since 
certain words in Tang poems rhyme only in Hakka but not in other dialects, showing 
that "Hakka" was true spoken Chinese language. 524 
Based on these gazetteers and Hakka genealogies, Luo Xianglin, in his book, 
Kejia Yanjiu Daolun (1933), also fought the idea that "the Hakkas are non-Chinese" and 
concluded that the Hakkas were originally from northern China and more Chinese than 
non-Hakkas.525 Therefore, the name 'Hakka' became more and more glorious since Luo 's 
publication of his book among the Hakka-speaking world. An editorial preface from the 
Hakka yearbook clearly demonstrates this kind of glorification: 
523 Zhenping XZ ijpp-,m;=t; Jiaying ZZ Jrrffl1+1 ;:t. 
524 Ibid., 7.84b. 
525 Luo, Kejia yanjiu daolun, Chapter 2. 
245 
Our people's distant ancestors migrated from the fertile (northern) plains into the 
barren hills (of the south) where they forged a character ofresistance and 
nonappeasement. Here our people have steadfastly served the ancestors by 
multiplying and carving out a niche in which to survive. By developing 
independently, our customs have remained distinct from the local people's, and 
we have not mixed as easily as does water with milk. 526 
These writings, compiled or written by respected members of the 
community as a means of mobilization for ethnic interaction fostered "group 
cohesiveness and maintained the ethnic boundary, even though these were not their 
only purposes. "527 
Hakka Emigration 
For the great majority of the Hakkas who survived the war, their lot was 
willing or unwilling departure from the area. Having failed to maintain their lands in 
the Wuyi , some returned to the east, and many resumed the general direction of the 
Hakka migration during the Qing, towards the west. Official assistance was 
provided for those Hakkas who wished to leave the province, and some accepted the 
526 Chongzheng Editorial Staff(1971 :I), quoted by Blake, Ethnic Groups, 50. 
527 Leong, Migration, 81. 
246 
offer, were given rations and sent under escort to Guangxi. The Hakkas were 
guaranteed land on which to settle in the districts to which they were to be moved, 
and promised that a quota would be provided for them in the civil service 
examinations. 528 
The Hakka dispersal took three years, from 1866 to 1869. In a number of 
instances the arrival of the Hakka settlers in their new areas provoked unrest. For 
instance, when Hakka dispersed from Kaiping and Enping reached Hainan early in 
1867, the local authorities had failed to make arrangements for their arrival. The 
Hakkas settled down apart from the existing population, and in Danzhou ft 3-H on 
occasion hired themselves out as braves to hostile villages. This brought a reproof to 
the prefect of Hainan. 529 
Another phenomenon that added to the highly volatile nature of this situation 
was the animosity between the native villagers and Hakka Christian converts among the 
migrants. Some of Hakka people found refuge in the reserves that the Christian 
missionaries were instrumental in creating during the nineteenth century, and in this 
528 Chixi XZ, 8.29b, 3 la. 
529 F.O. 228/448, Intelligence Report, (April 9, 1868). 
247 
context the missionaries played an important part in the construction of Hakka identity. 
Roman Catholics had penetrated the Wuyi area early in the eighteenth century and had 
made several thousand converts especially from the Hakka communities by the early 
nineteenth century.530 The missionaries' goal was to 'civilize' the people on the mission 
reserves, by teaching them to plough, to read, and to assist in the administration of their 
communities. To this end, they saw fit to make ethnic distinctions amongst the refugees 
who flocked to their protection. They sought to distinguish between the Hakkas and the 
natives, regarding the Hakkas as more amenable to civilization than the Punti people. 531 
Rev. Rudolph Lechler in his article, "The Hakka Chinese" described the Hakkas from a 
missionary point of view: 
On the whole the Hakkas are not as bigoted as the Puntis, and the Gospel has 
found easier access to them than to the latter. It is also comparatively easier to 
make friends of them than of the Puntis. It is perhaps owing to their standing 
constantly in fear of their own countrymen, the Puntis, that any sincere sympathy 
which is shown them by foreigners finds more reciprocity, and is thankfully 
availed of. 532 
53°F .O. 228/407, Mayers to Alcok, (October 30, 1866). 
531 F.O. 228/407, Mayers to Alcok, (October 30, 1866). 
532 Rudolph Lechler, "The Hakka Chinese," 352-359, Chinese Recorder and Missionary Journal, Vol. 9 
(Sep-Oct., 1878), 358. 
248 
Another missionary who had lived among the Hakkas once depicted the Hakka 
as "the cream of the Chinese people," and soon this statement became a favorite citation 
used by Hakka scholars from time to time. 533 
Typical of all Christian missionaries elsewhere in China, the foreigners were 
overzealously protective of their converts, some of whom frequently took advantage of 
their foreign connection to breach the law and bully their native neighbors. For instance, 
the enmity between the Hakka Catholic migrants and the local inhabitants was noted by 
the officials of the Leizhou peninsula. About 11 ,000-12,000 Hakka settlers arrived 
Leizhou in 1867, where French missionaries had already been active before Hakka arrival 
in the area. The Hakkas had moved from the Wuyi and were Catholic converts. 534 Soon 
after their arrival, the prefect of Leizhou received a letter from Monseigneur Guillemin, 
the Catholic bishop of Guangzhou, which stated that the land on which the Hakkas were 
settled was very inferior, and he wished to have the island of Weizhoutun Ill f l·I~ , off the 
mainland, for their settlement. Even though local Qing government did not accept this 
533 Historically, Christian missionaries were important in the ethnogenesis of the Hakka. Accounts written 
by Basel missionaries Eitel and Lechler were used by Hakka scholars to bolster the legitimacy of Hakka 
identity. These late-nineteenth-century missionary descriptions continue to be cited by contemporary Hakka 
specialists. See Ellsworth Huntington, The Character of Races (New York: Amo Press, 1924), 168; Eitel, 
"Ethnographical Sketches" ; Lechler, "The Hakka Chinese"; Constable, Christian Souls; Lutz and Lutz, 
Hakka Chinese. 
534 Choubanyiwu shimo iUi~Mi* (Beijing, 1930), 44.32a-b. 
249 
proposal, some Hakka converts moved on to the island.535 The aggressive tactics of 
Roman Catholic missionaries who were supporting the interests of Hakka settlers 
alarmed the native population and several anti-foreign and anti-Hakka incidents occurred. 
Three times between 1867 and 1870 Hakka churches there were burnt down.5 36 
One of the protestant missions, the Basel Mission, also concentrated on Hakka 
areas. A missionary from the Basel Mission, the above-mentioned Rudolf Lechler, visited 
Xinning in 1865 and took charge of the Hakka refugees who reached Hong Kong.537 
Some 3,000 of the Hakka refuges were brought to the colony, and as they were destitute, 
they were taken care of by the government and merchant community. Some of these 
refugees settled permanently in Hong Kong and some migrated to the Hawaii through the 
arrangement of Lechler. 538 
535 F.O. 228.448, Service Report (August 7, 1868). 
536 F.O. 682.341 ; Paul A. Cohen, China and Christianity: The Missionary Movement and the Growth of 
Chinese Antiforeignism, 1860-70 (Cambridge: Harvard University Press, 1963), 219. 
537 Lechler, "The Hakka," 355 . 
538 The Basel Mission was in their missionary politics aware, that to grant independence to the indigenous 
sub-churches as soon as possible was of important help to their goal : to spread Christianity even if it was 
adapted to local or societal circumstances. This led to the early foundation of the Tsung Tsin Associations 
(~IE~). first in Hong Kong and later with new branches in other countries where Hakka had emigrated. 
The process of this has had certain consequences for the Hakka ethnic consciousness: it has helped the 
creation of the "pan-Hakka" identity. See Leong, Migration, 87; Tin-yuke Char and Wai Jane, Chinese 
Historical Sites and Pioneer Families of the Island ofH awaii (Honolulu: University of Hawaii Press, 1983), 
100. 
250 
Although Chinese emigration overseas began before the Opium War, massive 
emigration to sell labor power started in the mid-nineteenth century. Besides the change 
in economic structure, there were other important reasons. During the mid-nineteenth 
century, after the abolition of slavery and the termination of the black slave trade, the 
large plantations in Central and South America faced a critical labor shortage. 539 In 
addition, the discovery of gold in California, the construction of transcontinental railways 
in the U.S. and Canada, and the development of mines and plantations in Southeast Asia 
all required cheap labor.5 40 Thus, the Westerners turned to Asia, and to China specifically, 
attempting to replace black slaves with Chinese coolies. In the meantime, under the 
military and diplomatic intimidation of the foreign powers, the Qing government changed 
its traditional isolationist policy and legalized foreign labor recruitment activities in China. 
The surrender of Hong Kong to the British and the opening of five ports to foreign trade 
facilitated this recruitment. Agents of foreign capitalists established labor recruitment 
539 Marshall K. Powers, "Chinese Coolie Migration to Cuba" (Unpublished Ph.D. Dissertation, University 
ofFlorida, 1953), 2. 
540 One of the most effective inducements was offering simple-minded peasants a conveyance to the gold 
hills in California or Australia. A peasant with an acquisitive bent easily fell into such a trap laid by one of 
the ubiquitous coolie agents. See Kil Young Zo, Chinese Emigration into the United States: 1850- / 880 
(New York: Amo Press, 1978), 39. 
251 
stations in Guangzhou, Swatow (Shantou), Amoy (Xiamen), and other coastal cities. 
Hong Kong and Macau became centers for the coolie trade.541 
An important source of coolie labor was prisoners of war taken in the frequent clan 
wars in southeastern China. During and after the Hakka-Punti War in the Wuyi areas, able 
bodied prisoners were sold to the coolie agents by their captors.542 Still, many war refugees 
gave themselves up out of despair to the coolie recruiters and willingly signed up to be 
contract laborers overseas. As noted more succinctly by the Xinning Gazetteer of 1893: 
"Ever since the destruction of the Red Turban Rebellions and the Hakka-Punti War, the 
numbers of people going abroad have increased dramatically. The strong and healthy 
regularly go to the four comers of the earth."543 These emigrants received a few dollars, 
advanced by the coolie agents, which they sent back to their starving families. The Gazetteer 
estimated that the Hakkas thus shipped to the Western Hemisphere who survived the ordeal 
of being coolies numbered up to 30,000 men by the turn of the century.544 
541 Yen Ching-hwang, Coolies and Mandarins: China 's Protection of Overseas Chinese during the Late 
Ch 'ing Period (1 851-1911 ), (Singapore: Singapore University Press, 1985), 54, 76; Persia C. Campbell, 
Chinese Coolie Emigration to Countries within the British Empire (London: Oxford University Press, 
1923), 94. 
542 Chixi XZ, 8.28a. 
543 X inning XZ, 8.6b. 
544 Chixi XZ, 8.28a. 
252 
After the Hakka-Punti War, both contract laborers and free immigrants from 
Guangdong formed permanent communities throughout the Western Hemisphere. There 
were about 7,000 Hakkas in Hawaii alone. The second largest district association among 
the Chinese community in California was also created by those who came from the Wuyi; 
and there was also a separate association for Hakka emigrants.5 45 Hakkas were also 
officially authorized and encouraged to emigrate to Taiwan after 1876. In the second half 
of the nineteenth century the main migratory movement, however, was to Southeast 
Asia. 546 The Malayan tin-mines attracted many emigrants, and later there was a steady 
demand for labor there and in the Dutch East Indies. As considerable numbers of these 
came from the Wuyi; they included many who had emigrated as a direct consequence of 
the war and others who had taken part in the hostilities or who had bitter memories of the 
conflict. 
Consequently among these overseas Chinese, subethnicity has remained an 
important factor in social organization until the present day. Clashes of interest have often 
coincided with subethnic grouping. It is not therefore surprising to find indications that 
545 Barth, Bitter Strength, 90; Stephen Themstrom, ed., Harvard Encyclopedia ofA merican Ethnic Groups 
(Cambridge: Harvard University Press, 1981 ), 231. 
546 Clyde Y. Kiang, The Hakka Odyssey & Their Taiwan Homeland(Elgin: Allegheny Press, 1992), 68-69. 
253 
the hostility between the Hakkas and the Puntis continued even when they had moved 
away from the scene of the war. In Hong Kong, in 1866, by which time a substantial 
number of Hakka immigrants had reached the colony, it was reported that clashes on a 
small scale occurred between the two groups. 547 The best information on the extension of 
Hakka-Punti hostility overseas comes from Malaya. In Perak, Malaya, Larut Wars broke 
out between the two groups over possession of the Larut tin fields between 1862 and 
1873. These disturbances have been supposed to have been in some way connected with 
the Hakka-Punti War.548 Subethnic conflicts between the Hakkas and the Puntis continued 
long after the end of the Hakka-Punti War in various places and it was clearer than ever in 
Guangdong at least, who was Hakkas, and who were not. 
4) Ethnicity and the Hakkas 
In this chapter, I provided background information on the Hakkas, and 
specifically explored their settlement in the Wuyi area. Focusing on ethnographic cases in 
547See Overland China Mail (February 15, 1866). 
548 Thus Leon Comber has written: "This bitter struggle will help to explain the traditional secret society 
hostility between the Cantonese and Hakka emigrants in the Straits Settlements during the same period." 
See Leon Comber, Chinese secret societies in Malaya; a survey oft he Triad Society from 1800 to 1900 
(Locust Valley, N. Y. , Published for the Association for Asian Studies by J. J. Augustin, 1959), 29. 
254 
the Wuyi, I have also continued the treatment of the Hakka-Punti distinction less in terms 
of static ethnic categories and more in terms of intense relationships enacted in the 
process of major events, the Red Turban Rebellions and the Hakka-Punti War especially 
in the Wuyi area. 549 The fluid and constructionist nature of Hakka identity is not only 
manifested in relation to 'place' and local interactions, but also through time and 
changing state-sociery relationships. 
Many Hakkas today perceived their identity as being inherited from their 
ancestors. The history of family and ancestors plays an important role in Hakka culture. 
The Hakk.as have a strong sense of identity, emphasizing their common language and 
blood relationship even among settlements in various countries across the world. 
Research by Hakka scholars, especially by Luo Xianglin, strengthened the idea of a 
'common imagination' of the Hakkas across regional boundaries: the Hakkas may have 
lived in the Wuyi, Jiayingzhou, Hong Kong, Taiwan, Malaysia and America, but they are 
the Hakkas sharing the same blood, and speaking the same language. Differing from Luo, 
549 I see ethnicity as a fluid, artificial boundary drawn between groups whose economic and political 
interests are locked in competition. In other words, ethnicity is not a "thing" but a summation of a set of 
relations. See Choi Chi-chenug, "Descent Group Unification and Segmentation in the Coastal Area of 
Southern China," (Ph.D. diss., University ofTokyo, 1987); James Hayes, The Hong Kong region, 1850-
1911 : institutions and leadership in town and countryside (Hamden, Conn. : Archon Books, 1977); John L. 
Comaroff, "OfTotemism and Ethnicity: Consciousness, Practice and the Signs oflnequality." Ethnos 52. 3-
4 (1987): 301-23. 
255 
however, Chen Zhiping (1997) contends that the Hakk:as do not have a bloodline that 
differs significantly from other groups in Fujian, Guangdong and Jiangxi. He sees the 
blood relationships among these groups to be mingling and mixing. 550 Nicole Constable 
(1994) also points out that the key to Hakka identity is not language, shared political 
interests, shared cultural practices, religion or native place, but it is the way in which 
these and other elements are invariably tied to Hakka history. 551 
Of course cultural symbols are in fact potent shapers of ethnic consciousness. 
For instance, ethnic language or dialect is considered a powerful symbol of ethnic identity, 
and it is in this category where a sharp contrast between the groups is obvious. Thus the 
"Indian Hakkas" pride themselves on their preservation of the Hakka dialect even in 
India; they regard their dialect as an important attribute of their ethnic identity. 552 
Difference in customs, too, did play a role in the construction of Hakka identity. But 
exactly when and where Hakka cultural differences coalesced into an expression of social 
group differences is difficult to say. Moreover, it is clear that a conscious, collective 
55°C hen, Kejia yuanliu, 147. 
551 Constable, Christian Souls, 76. 
552 Dialect served as an external indicator of Hakka ethnic identity in Calcutta, India, and a symbol through 
which the Hakkas could be differentiated from other Chinese groups. Oxfeld, "Still 'Guest People '." 
256 
group identity did not come form in all regions where Hakka speakers existed. 553 Thus, in 
the case of the Hakka, language and customs-"culture"-were not sufficient markers of 
ethnic difference. They alone cannot explain the formation of Hakka ethnic 
consc10usness. 
Political action (and inaction) also played a role. The state, in fact, played a 
crucial part in the development of subethnic rivalry in the Wuyi. Ineffective restrictions 
and government policy of migration helped to bring about the heterogeneous immigration 
and settlement that gave rise to Hakka subethnicity in the Wuyi. Subsequently, weak 
governmental control allowed and, indeed, encouraged the formation and preservation of 
bellicose rival communities. In other words, official concessions to one or other 
subcultural group or local group (e.g., examination quotas) not only fostered subethnic 
rivalry but also set precedents for competition of an economic, social, and political sort. 
More overtly influential was the active support of the Qing court in the formation of the 
Hakka Confederation by some Hakka gentry in 1856 to combat the Red Turban threat. 
In sum, it was in the course of a complex series of events-socioeconomic 
conflicts between local groups, Red Turban Rebellions, and the resulting Hakka-Punti 
553 Constable, Guest People, 12. 
257 
War-that the early cultural solidarity of the Hakka-dialect speakers was forged into a 
full ethnic consciousness. To summarize the crucial factors here: The Punti absentee 
landlords of the Wuyi areas, in the process of enclosing polder farms and reclaiming land, 
fought among themselves under the disguise of lineage feuding, and they made use of the 
newly immigrated Hakka tenants in these contests to carry out frontier cultivation. This 
enabled some Hakkas eventually to become secondary landlords. But then, as these 
Hakka secondary landlords became more prosperous, they threatened the security and 
dominance of the Puntis (both landlords and tenants), and economic conflict between the 
two groups escalated. This conflict helped create the tensions that sparked the Red 
Turban Rebellions. Then, in the effort to defeat the Red Turban rebel forces, the state was 
placed in the position of legitimizing the Hakka gentry's efforts to organize their own 
confederation. This confederation did assist in the defeat of the Red Turbans, but it also 
attempted to dominate the Punti tenants and to enclose polder farms owned by the 
absentee Punti landlords. As a result, the power struggles in the Wuyi areas during the 
Red Turban Rebellions came to take the forms of subethnic feuds rather than the 
conventional form of interclass struggle or lineage feuds. 
This ethnic struggle analysis also highlights the complex connection between 
258 
human agency and structural constraints. Throughout this chapter, both Hakka and Punti 
elites were portrayed as the actors in the Wuyi areas. Both sides skillfully procured the 
endorsement of the state in legitimizing their attacks on their enemies and they 
manipulated subethnic sentiments in dominating the local regions, thus eventually 
creating a clear demarcation between the Puntis and the Hakkas. 
The Hakkas, first, were identified as different by people who looked down upon 
them and explained their difference in derogatory terms. Then, as economic and social 
conflicts developed and as the Hakkas began to do rather well, this sense of difference 
was sharpened on both sides, and eventually the Hakkas began to define themselves as 
distinctive in more positive terms, as the lieratus Xu Xuzeng did in 1808. As the Hakkas 
and the Puntis were drawn into armed struggle, encouraged by the government, the sense 
of difference was heightened and extended, so that after the Hakka-Punti War, Hakka 
"theorists" like Wen Zhonghe wrote extensive definitions of Hakka customs, language-
in short, of Hakka identity. 
Religion, both Chinese and foreign, added an important dimension here. Hakka 
identity was historically influenced by Western imperialism and nineteenth-century 
Christian evangelism. In the midst of the turmoil of the nineteenth century, the Puntis 
259 
were able to gain some security and mutual aid from "traditional" Chinese religious 
organizations and secret societies (like the Red Turbans) that were based in settled Punti 
communities. In contrast, the fragmented kinship and settlement patterns of the Hakkas in 
Guangdong provided no firm social base, so the Hakkas utilized the ideas and institutions 
of the foreign missionaries for their protection. 554 In some cases, turning to the 
missionaries was probably a last resort in an effort to escape poverty and homelessness. 
Alliance with the missionaries, powerful because they had the backing of their 
governments, also provided the Hakkas with a useful counterweight to the Puntis. 
In any event, European missionaries found that the Hakkas provided more fertile 
ground for sowing the seeds of the gospel than the Puntis. In converting their charges, the 
missionaries, pursuing what Stevan Harrell has named "the Christian Project," hoped to 
elevate the Hakkas to what they perceived as the superior level of Western civilization. 
Confronted then by these missionary attempts to educate them to be "civilized," Hakka 
communities were forced to reflect on who they were and how they were different from 
the foreigners attempting to "civilize" them. Thus contact with Western missionaries 
554 Kuhn, "Origins of the Taiping," 365. 
260 
served to stimulate the ethnic self-consciousness of the Hakkas. 555 
In short, this chapter focused on the whole network of different interests and 
relationships that led to the Red Turban Rebellions and the Hakka-Punti War of the mid-
nineteenth century in the Wuyi: the long-standing economic conflicts over land use; the 
different religious allegiances and associational ties that divided migrants from natives; 
the part played by local gentry and secret societies in Hakka-Punti feuds; the role that the 
state, and most particularly local governments, played in intensifying existing tensions 
and thus drawing "ethnic" lines. 
555 Stevan Harrell, ed. Cultural Encounters on China's Ethnic Frontiers (Seattle and London: University of 
Washington Press, 1994): 20-29. 
261 
CHAPTERV 
CONCLUSION 
It is difficult to overestimate the role that ethnicity and perceived ethnic 
differences have played in late imperial Chinese history. At many turning points in late 
Ming (1368-1644) and Qing (1644-1911) history, underlying subethnic factors, 
perceptions of group interests at the regional level, and local value systems have been 
crucial elements in events that shaped the future of the empire. Even today, the 
perception of one Chinese about another is likely to be colored by knowledge of his 
native place or of the dialect group to which he belongs, as evidenced by his 
pronunciation of Standard Chinese. Thus, the survey of subethnic groups within China 
not only reveals the complexities of dialect or local folklore, but also becomes necessary, 
a basic step, in the effort to understand Chinese history and society better. 
The recent proliferation of works by historians and anthropologists on ethnic 
262 
groups in China brings us closer to this understanding.556 Yet only the most preliminary 
work has been done on one of the most influential and controversial minorities in China: 
the Hakkas. Taking off from (and to some degree contesting) the important work of Sow-
Theng Leong (1997), I have examined the formation of a distinct Hakka identity in the 
nineteenth century. Through a study of Hakka involvement in the Red Turban Rebellions 
(1854-56) and the Hakka-Punti War (1856-68), I have focused on how the Hakkas 
conceived of their identity and how they expressed it through their voices, their 
settlements and migrations, and their conceptions of the way their local history was 
linked to broader Hakka and Chinese history. I contend that Hakka ethnic identity is 
constructed in relation to antagonistic interactions between the Hakkas and the Puntis. 
Hence, even though there was some basis for positive Hakka identity before the conflicts 
later in nineteenth century,557 it was the experience of the Red Turban Rebellions and the 
resulting Hakka-Punti War that reinforced and extended that sense of identity. 
556 See Emily Honig, Creating Chinese Ethnicity: Subei People In Shanghai, 1850-1980 ( New Haven and 
London: Yale University Press, 1992); Stevan Harrell, "Ethnicity, Local Interests," 51-548; Pamela Kyle 
Crossley, "Thinking about Ethnicity in Early Modem China," Late Imperial China, 11 (1990): 1-35; Dru C. 
Gladney, Muslim Chinese: Ethnic Nationalism in the People's Republic ofC hina (Cambridge: Harvard 
University, 1991); David Ownby, "The Ethnic Feud in Qing Taiwan: What is This Violence Business, 
Anyway? An Interpretation of the I 782 Zhang-Quan Xiedou," Late Imperial China 11 (1990): 75-98. 
557 Xu Xuzeng's speech, made in 1808, is a good example of the development of a positive ethnic 
consciousness before the mid-nineteenth century. 
263 
The Hakkas were originally set off from the Puntis by their more recent 
settlement in the Pearl River Delta and by their distinctive dialect (along with some 
distinctive cultural practices). Economic tensions between the two groups, as land grew 
scarce, intensified the differences between the two groups, leading the Puntis to develop 
derogatory labels and descriptions of the Hakkas. The state failed to ease these tensions 
and inadvertently intensified them by establishing inadequate quotas for the examination 
system. 
By early in the nineteenth century, at least some Hakka literati (represented by 
Xu Xuzeng) had developed a positive sense of their separate identity, responding to Punti 
slurs with an ideology of Hakka superiority. Increased economic decline in the early 
nineteenth century sparked lineage feuds and the growth of the Tiandihui among the 
Puntis in particular. Then, the Red Turban Rebellions spread disorder and resulted in the 
greater impoverishment of the countryside; and the governments' reliance on the Hakka 
braves to defeat some of the rebels particularly in the Wuyi region further fueled ethnic 
tensions. Shortly after the defeat of the Red Turbans, these tensions sparked the Hakka-
Punti War, which, by explicitly and violently pitting the Puntis against the Hakkas, served 
to consolidate and strengthen Hakka ethnic consciousness. 
264 
Protestant and Catholic missionaries reinforced the sense of the difference 
between the Hakk.as and the Puntis by embracing and perpetuating the Hakka ethnic 
ideology. And finally, the state, in creating a new and separate county for the defeated 
Hakka of the Wuyi, confirmed and gave political approval to the Hakka-Punti distinction. 
In sum, the construction of a distinct Hakka ethnic identity, one that claims the 
allegiance of Hakka even today was a long-term process, beginning with the arrival of the 
"Guest People" in Guangdong. It was also a complex process, shaped not just by cultural 
difference (language and customs), but also by socioeconomic conflict, state action (and 
inaction), and religious and ideological difference, fuled in part by the presence of foreign 
missionaries in Guangdong. 
The influence of subethnic identities on Chinese society and culture has become 
a topic of increasing interest among scholars of China recently and over the past years 
scholars have also become increasingly sensitive to histories and construction of Hakka 
identities.558 While I was conducting archival research for this writing in Beijing from 
August 2001 to January 2002, a special documentary TV series on the Hakka people was 
regularly on the air on the CCTV (China Central Television). That documentary program 
558 See Leong, Migration; Constable, Guest People. 
265 
described Hakka kinship and village structures (including the famous circular walled 
villages of northwestern Fujian), agricultural practices, folk arts, folk songs, and ritual 
practices. The Hakka world and customs portrayed by the TV special documentary 
program was very colorful, meaningful and intrinsically appealing for those who desired 
to identify with the greater world of Hakka culture. 
In fact, the special documentary program on Hakka customs and culture was a 
product of the recent surge of interest in Hakka culture occurring among Chinese both in 
the PRC and overseas. The Annual International Conference on Hakkaology is only one 
of the most recent examples. Its more than 100 scholarly papers written by Chinese from 
the PRC and overseas communities attests to a vigorous interest in Hakka history, 
language, and culture both within and outside China. This sort of ethnic revitalization, 
according to Mary Erbaugh, 559 has also emerged among Hokkien and Cantonese as well 
as Hakka, a process that she attributes in part to the return visits of Overseas Chinese 
looking for their roots. 560 
The overseas Hakka groups often adjusted to their new environment and to other 
559 Erbaugh, "The Hakka Paradox," 228. 
560 The Hakka ethnic revitalization was well demonstrated by the floods of publications on the Hakka 
history, which were often sponsored by Hakka emigrants from all over the world, and which preserved 
memories of the original Hakka settlements and the history of the Hakkas. 
266 
subethnic groups without losing their subethnic identity. Abner Cohen has called attention 
to similar situations in which immigrant groups, rather than losing their cultural identities 
through integration, have recognized their own traditional customs or developed new 
customs under traditional symbols in an effort to become more distinct. He has called this 
process "ethnic continuity" or "revival."561 
The establishment of Hakka University (Jiaying University ~ $.jc?!t)562 and 
Hakka Research centers563 where none existed before, a vital aspect of the mainland 
China's modernization projects, has been an important catalyst in the continuing 
construction of identity among the Hakkas.564 As a Hakka University was first set up in 
561 Abner Cohen, "Introduction: The Lessons of Ethnicity." In Urban Ethnicity, Abner Cohen, ed., ix-xxiv. 
ASA Monographs, no. 12 (London: Tavistock Publications, 1974), xiv; Erbaugh, "The Hakka Paradox," 
227-229. 
562 Jiaying University, located in Meizhou Prefecture, Guangdong Province, was established in 1985 with 
the approval of the provincial government and the support of overseas Chinese, especially a successful 
Hakka businessman in Taiwan. Accordingly, the Jiaying university is focusing on Hakka culture in their 
outreach to overseas Chinese. See Eriberto P. Lozada, Jr., God Aboveground: Catholic Church, 
Postsocialist State, and Transnational Process in a Chinese Village (Stanford: Stanford University Press, 
2001), 184-185. 
563 Recently many Chinese universities both from the mainland and Taiwan established Hakka Research 
centers with the hope of encouraging academic research, implementing plans related to Hakka culture, 
cooperating with external organizations, promoting creation and preservation of Hakka culture and history. 
Followings are the major Hakka Research Centers: Hakka Research Institute (Jiaying University, 
Guangdong); Hakka Research Society (Shenzhen University, Guangdong); International Hakka Research 
Institute (National Chiao Tung University, Taiwan); Center for Hakka Studies (National Central University, 
Taiwan) 
564 Erbaugh, "The Hakka Paradox," 227-228. 
267 
Meixian, Guangdong, 565 it also became the avenue through which the children of the 
Hakk:as from different areas could come into contact with one another. They not only 
became literate in a standard Hakka language (Meixian Hakka language), they also learn 
a common set of values and shared a common experience. Meizhou prefecture now plays 
that role for the "true Hakka" and the label "Hakka" is regarded with pride throughout the 
Hakka population in the world. The Hakkas accepted the whole story of Luo Xianglin to 
gain a sense of superiority and more and more Hakka speakers have now formed "Hakka 
Associations" and claimed themselves "Hakka."566 
How do we interpret the renewed attractiveness of Hakka or other forms of 
subethnic identities to Chinese? Will new types of mass media (such as the internet) 
intensify the creation of new forms of imagined communities?567 Analysts of global 
trends have noted the divergent tendencies for new communications technologies to both 
foster more universalistic identities, as well as to enhance more localized 
565 Meixian lies in an area generally referred to as the "Hakka heartland." Located in the mountains 
between western Fujian and eastern Guangdong, Meixian exhibited many features that Luo identified as 
being typical of Hakka county. Namely it was hilly, poor, isolated, and handicapped for trade. See Zheng, 
"Guangdong Zhonglu." 
566 Tsung-Tsin Associations throughout the world are the best example of this. See Hu Xizhang, et al. , Kejia 
Fenghua. 
567 Benedict Anderson, Imagined communities: reflections on the origin and spread ofn ationalism (London 
& New York: Verso, 1991). 
268 
identifications. 568 In fact, the attraction to subethnic ties may actually reflect a more 
depoliticized form of Chinese identity, a renewed identification with localized ancestral 
traditions- a sanitized version of a past that has now been safely transcended. The recent 
scholarly interest in global cultural flows that transcend national boundaries has directed 
increasing attention to the far flung populations often associated with these flows, and a 
· new scholarly discourse on diasporic cultures and identities. 569 The wishes of young 
Hakkas living in different non-Chinese settings reassert ethnic bonds through network 
and recover individually one 's cultural roots (i.e. by being eager to study the Hakka 
dialect or history is the most important cohesive element for a subethnic group). 
Participating in the Annual Hakka International conference and chatting through the 
Hakka websites provides important contest for the enhancement of Hakka group 
consciousness.5 70 
We discovered that the subethnic rivalry in the Wuyi had other long-lasting 
effects. Despite the complex pattern of the Hakka migrations, wide distribution of the 
568 S. Hall, ed., Representation: Cultural Representations and Signifying Practices (London: Sage, 1997). 
569 James Clifford, On the Edges ofA nthropology : Interviews (Chicago, University Press Marketing, 2003) 
570 Through the Hakka International Conference and the Internet, the Hakka Diaspora is connecting each 
other. One such Hakka community in cyberspace illustrates how transnational processes shape what it 
meant to be Hakka all over the world, including in the Hakka homeland. 
269 
Hakka people and their varied historical experiences, the sense of a single Hakka identity 
has still remained strong. "Hakka culture" is, however, not constant but subject to 
change.571 That is, the criterion for Hakka identity has changed from dialect to an 
imagined and more abstract ideology of Hakka identity, an ideology that gives a widely 
scattered community a belief in a common Hakka identity. Hakka people today are living 
in many different countries but connected each other through various ways including 
established Hakka user groups on the internet. These Hakkas, in this age of globalization 
and internationalization continue to make their own history. 
571 Howard J. Martin, "The Hakka Ethnic Movement in Taiwan, 1986-1991." In Constable, Guest People, 
176-195. 
270 
BIBLIOGRAPHY 
Abbreviations Used in Footnotes 
C.O. Colonial Office Archives, Public Record Office. London. 
F.O. Foreign Office Archives, Public Record Office. London. 
FZ Fuzhi Jff~ (Prefectural Gazetteer). 
:XZ Xianzhi l'f*'~ (County Gazetteer). 
Anderson, Benedict. Imagined communities: reflections on the origin and spread of 
nationalism. London & New York: Verso, 1991. 
Antony, Robert James. "Pirates, Bandits, and Brotherhoods: A Study of Crime and Law 
In Kwangtung Province, 1796-1839." Ph.D. diss. , University of Hawaii, 1998. 
Averill, Stephen C. "The Shed People and the Opening of the Yangzi Highland," Modern 
China 9.1 (Jan. 1983): 84-126. 
Baker, Hugh. A Chinese Lineage Village: Sheung Shui. London: Frank Case, 1968. 
Barth, Gunther Paul. Bitter strength; a history oft he Chinese in the United States, 1850-
1870. Cambridge, Harvard University Press, 1964. 
Barth, Fredrik. "Introduction." In Ethnic Groups and Boundaries: The Social 
Organization ofC ultural Difference. ed. Fredrik Barth. Oslo: Universitetsforlaget, 
1969. 
Bartlett, Beatrice S. Ch'ing documents in the National Palace Museum Archives. Taibei: 
National Palace Museum, 1975. 
----------. Archival Materials in China on United States History. White Plains, N.Y.: Kraus 
International Publications, 1985. 
----------. Selected articles from Taipei National Palace Museum bulletin, 1972-1979. 
271 
Taibei: National Palace Museum, 1972-79. 
Beattie, Hilary J. , Land and lineage in China: a study ofTung-cheng County, Anhwei, in 
the Ming and Chi"ng dynasties. Cambridge; New York: Cambridge University 
Press, 1979. 
Bian Baodi --tlfffi, Bian Zhijun (Songchen) Zhengshu -ttr~(~§)i&~ . Taibei: 
Wenhai, 1968 Reprint. 
Blake, C. Fred. Ethnic Groups and Social Change in a Hong Kong Market Town. Asian 
Studies at Hawaii, no. 27. Honolulu: University Press of Hawaii, 1981 
Blythe, W. L. The Impact of Chinese Secret Societies in Malaya: A Historical Study. 
London: Oxford University Press, 1969. 
Boardman, Eugene. Christian Influence upon the Taiping Rebellion, 1851-1864. Madison, 
University of Wisconsin Press, 1952. 
Branner, David Prager Branner. Problem In Comparative Chinese Dialectology: the 
Classification ofM iin and Hakka. Berlin and New York: Mouton de Gruyter, 2000. 
Brine, Lindesay. The Taiping Rebellion in China: a narrative of its rise and progress, 
based upon original documents and information obtained in China I by Commander 
Lindesay Brine; with map and plans. London: John Murray, 1862. 
British Foreign Office Archives on file in the Public Record Office (FO 17: General 
Correspondence, F.O. 228: Embassy and Consular Archives, F.O. 405: Confidential 
Prints, F.O. 931: Papers in the Chinese Language, C.O. 129: Hong Kong Dispatches, 
C.O. 273: Static Settlements, Original Correspondence, 1839-1919), London. 
Buoye, Thomas M. "Economic Change and Rural Violence: Homicides Related to 
Disputes over Property Rights in Guangdong During the Eighteenth Century." Peasant 
Studies 17.4 (1990): 233-259. 
----------. Manslaughter, Markets, and Moral Economy: Violent Disputes over Property 
Rights in Eighteenth-century China. Cambridge: Cambridge University Press, 2000. 
Burnett, H. S. and V. V. Hagelstrom. Present Day Political Organization of China. 
Shanghai: Kelly and Walsh, limited, 1912. 
Cai Shaoqing ~:'J,~g~p. Zhongguo Jindai Huidangshi yanjiu ~ OO:ili:1~~;t~li}fJi. 
Beijing: Zhonghua shuju, 1987. 
----------. Zhongguo Mimi Shehui ~ OO ~~t±~. Hangzhou: Zliejiang renmin chubanshe, 
1989. 
----------. "Guanyu Tiandihui de Qiyuan Wenti IUJ-=f.7;;Jt!?.fr1¥J~V!Fi:i~~-" Beijing Daxue 
Xuebao :lt~=k~~f!i 1 (1987): 53-64. 
272 
----------. "Lun Taipingtianguo yu Tiandihui guanxi ~ ::k2f::R~ ~J:t-tg~ IUJ~." In Idem. 
Zhongguo jindai huidangshi yanjiu 4i ~ ill:~ ~:i:5'::.wf~. Beijing: Zhonghua shuju, 
1987: 158-171. 
Campbell, George. "Origin and Migrations of the Hakkas." Chinese Recorder 43 (1912). 
Cao Fengzhen • ffel._V[ ed., "Yahu Caoshi zupu ~¥!W-- B~J~if," In Guangdong hongbing 
qiyi shiliao JJ:*1AA~~5':.~. Guangdong sheng wenshi yanjiuguan, Zhongshan 
daxue Ii shi xibian, 851-852. Guangzhou : Guangdong renmin chubanshe, 1992. 
Chang Chung-Ii. The Chinese Gentry Seattle: University of Washington, 1955. 
Chang Hsin-pao. Commissioner Lin and the Opium War. Cambridge: Harvard University 
Press, 1964. 
Char, Andrew. "Social Origins of Ethnicity: The Hakka in Central Kwangtung." Stone 
Lion Review 9 (Harvard University, Spring 1982): 53-78. 
Char, Tin-yuke and Wai Jane. Chinese Historical Sites and Pioneer Families oft he Island 
ofH awaii. Honolulu: University of Hawaii Press, 1983. 
Chen Chengxi ITT!Ul-\ffi!. Guangdong dizhi !J:*:i-m~t. Hong Kong: Tiandi chubanshe, 1978. 
Chen Dianlan ~~~- Gang~heng zhengeji llrrl~tt::IG!c." In Guangdong hongbing qiy i 
shiliao !J:*1AA~~5':,;fS!-. Guangdong sheng wenshi yanjiuguan, Zhongshan daxue 
lishi xibian, 895-947. Guangzhou: Guangdong renmin chubanshe, 1992. 
Chen Ganhua ~ =fit. "Yuexi Lianjiang de Kejiaren ~iffiltUEB~*A-" Kejia yanjiu 
jikan ~ ~ :fi}fJi~fIJ. Vol. 9 (1996): 55-60. 
Chen Han sheng. Landlord and Peasant in China: A Study oft he Agrarian Crisis in South 
China. New York: International Publishers, 1936. 
Chen Kun ~ :!:$ . "Yuedongjiaofeijilue • * j j filfo.\c~." In Guangdong hongbing qiy i 
shiliao /J:*fAA~~Y:~. Guangdong sheng wenshi yanjiuguan, Zhongshan daxue 
lishi xibian, 665-772. Guangzhou : Guangdong renmin chubanshe, 1992. 
"Chen Song shiyou ~tl$El3 ." In Guangdong hongbing qiyi shiliao !J:*1AA~~5':.~. 
Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi xibian, 212-213 . 
Guangzhou : Guangdong renmin chubanshe, 1992. 
Chen Yundong ~ ~tf. Kejiaren ~ * A· Taibei: Lianya chubanshe, 1978. 
Chen Zhengxiang ~IEJr,+. Guangdong Dizhi JJ:*:i-m~ . Hong Kong: Tiandi tu hu 
youxian gongsi, 1978. 
Chen Zhiping ~3z Zfi . Kejia yuanliu xinlun ~*mOjfEffi~. Nanning: Guangxijiaoyu 
chubanshe, 1997. 
Chesneaux, Jean. Secret Societies in China in the Nineteenth and Twentieth Century. Ann 
273 
Arbor: University ofM ichigan Press, 1971. 
----------. Peasant Revolts in China 1840-1949. London: Thames and Hudson, 1973. 
Ch'iang T'ing-fu. "The Government and Co-hong of Canton, 1839," Chinese Social and 
Political Science Review, 15.4 (1932): 602-7. 
China Mail 10 (May 1855), 22 (May 1856), 27 (October 1853). 
Chinese Repository III (1834-5), IV (1836), XII (1843). 
Chixi Xianzhi ?JFfslfvt~. (Gazetteer of Chixi County). 1920. 
Choi Chi-chenug. "Descent Group Unification and Segmentation in the Coastal Area of 
Southern China." Ph.D. diss. , University of Tokyo, 1987. 
Chouban yiwu shimo IOJ.i~frtr~. Beijing, 1930. 
Ch'u T'ung-tsu. Local Government in China under the Ch 'ing. Cambridge: Harvard 
University Press, 1962. 
Clifford, James. On the Edges ofA nthropology : Interviews. Chicago, University Press 
Marketing, 2003. 
Cohen, Abner. "Introduction: The Lessons of Ethnicity." In Urban Ethnicity, ed. Abner 
Cohen, ix-xxiv. ASA Monographs, no. 12. London: Tavistock Publications, 1974. 
Cohen, Myron L. "The Hakka or 'Guest People': Dialect as a Sociocultural Variable in 
Southeastern China." Ethnohistory 15.3 (Summer 1968): 237-92. 
----------. "The Hakka or "Guest People," Dialect as a Sociocultural Variable in Southeast 
China." In Guest People: Hakka Identity in China and Abroad. ed. Nicole Constable, 
36-79. Seattle and London: University of Washington Press, 1996. 
Cohen, Paul A. China and Christianity: The Missionary Movement and the Growth of 
Chinese Antiforeignism, 1860-70. Cambridge: Harvard University Press, 1963. 
Comaroff, John L., "OfTotemism and Ethnicity: Consciousness, Practice and the Signs 
of Inequality." Ethnos 52. 3-4 (1987): 301-23. 
Comber, Leon. Chinese Secret Societies in Malaya; A Survey oft he Triad Society from 
1800 to 1900 . Locust Valley, N. Y., Published for the Association for Asian Studies by 
J. J. Augustin, 1959. 
Constable, Nicole. Christian Souls and Chinese Spirits: A Hakka Community in Hong 
Kong. Berkeley: University of California Press, 1994 
----------. ed. Guest People: Hakka Identity in China and Abroad. Seattle and London: 
University of Washington Press, 1996. 
Consular Report Vol. 23. Washington D.C.: U.S. Government Printing Office, 1889. 
274 
Cooke, George W. China and Lower Bengal: Being the Times Special Correspondence 
from China. London: Warne & Routledge, 1858. 
Costin, W.C. Great Britain and China, 1833-1860. Oxford: The Clarendon press, 1937. 
Cressey, George B. China's Geographical Foundation. New York: McGraw Hill, 1934. 
Croizier, Ralph C. Koxinga and Chinese Nationalism: History, Myth, and the Hero. 
Cambridge: Harvard University Press, 1977. 
Crossley, Pamela Kyle. "Thinking about Ethnicity in Early Modem China," Late Imperial 
China, 11 (1990),pp.1-35. 
----------. Orphan Warriors: Three Manchu Generations and the End oft he Qing World. 
Princeton: Princeton University Press, 1990. 
----------. The Manchus. Oxford and Cambridge, Blackwell, 1997. 
Cui Yingchang W~J§ . Sichuan Fangyanyu Bashu wenhua IZY )ll:1:f s ~ EliJt-ft. 
Chengdu: Sichuan daxue chubanshe, 1996. 
Daoguangchao donghualu 3i:J1t~*~~' 1899. 
Davis, Fei-ling. Primitive Revolutionaries ofC hina: a study ofs ecret societies in the late 
nineteenth century. Honolulu: University Press of Hawaii, 1971 . 
"Darning Douyuanshuai xiwen * SJHm5cgrtJ~X." In Guangdong hongbing qiyi shiliao 
/J::W~~~~~;fSI-. Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi 
xibian, 50-52. Guangzhou : Guangdong renmin chubanshe, 1992. 
Da Qing Gaozong Chun (Qianlong) Huangdi shilu *mr\§J*liH(1iiz:~)~'rf:fJt~. Taibei: 
Hualian Chubanshe, 1964 Reprint. 
Daqing Lichao shilu, Daoguangperiod *m~t!Hl~, 3i:J1t~. Taibei, 1964 Reprint. 
Daqing Lichao shilu, Xianfengperiod *m~~-~' &Z.:'.~. Taibei, 1964 Reprint .. 
Daqing Muzong Yi (Tongzhi) Huangdi shilu mt--l*~(jq)W)~'rf:f'il~. Taibei: Taiwan 
huawen shuju, 1968 Reprint. 
Da Qing Renzong Rui (Jiaqing) huangdi shilu ::krn t* ~ (~el) ~* Ji'.~. Taibei 
Shi : Faxingzhe Huawen shuju, 1968 Reprint. 
Da Qing Shizong Xian (Yongzheng) Huangdi Shilu *mt!t*:l:(*IE)~1iJ'.it~. Taibei: 
Hualian Chubanshe, 1965 reprint. 
Da Qing Xuanzong Cheng (Daoguang) huangdi shilu .:km11r*J»G(3i:J1t)~1i1Jt~ . 
Taibei: Huawen shuju, 1964 Reprint. 
Deqing Zhouzhi ffiefflt=t, 1909. 
275 
DongguanXianzhi W..~~~- 1765. Taibei: Chengwen chubanshe. 1969 Reprint. 
Donghua quanlu (daoguangchao) W..~~~ (313't~)- Taibei, 1963 Reprint. 
Eberhard, Wolfram. Social Mobility in Traditional China. Leiden: E. J. Brill, 1962. 
Eitel, E. J. "Ethnographical Sketches of the Hakka Chinese." Notes and Queries on China 
and Japan I.5 (31 May 1867), I.6 (29 June 1867), I. 7 (July 31 , 1867). 
----------. "An Outline History of the Hakkas." China Review 2 (July 1873-June 1874): 
160-64. 
Elvin, Mark J. Elvin "The Last Thousand Years of Chinese History." Modern Asian 
Studies, 4.2 (1970): 97-114. 
Encyclopaedia Britannica, a new survey ofu niversal knowledge. Chicago 
and London: Encyclopaedia Britannica, Inc., 1946. 
Eng, Robert Y. "Institutional and Secondary Landlordism in the Pearl River Delta." 
Modern China, Vol. 12 No. 1 (Jan. 1986). 
Enloe, Cynthia H. Ethnic conflict and political development. Boston: Little, Brown, 1973. 
Enping Xianzhi ~,Zp-~~- 1934. 
Erbaugh, Mary S. "The Secret History of the Hakkas: The Chinese Revolution as Hakka 
Enterprise." China Quarterly 132 (Dec. 1992): 93 7-68. 
----------. "The Hakka Paradox in the People's Republic of China: Exile, Eminence, and 
Public Silence." In Guest People: Hakka Identity in China and Abroad. ed. Nicole 
Constable, 196-231. Seattle and London: University of Washington Press, 1996. 
Esherick, Joseph. Chinese Archives: An Introductory Guide. Berkeley : Institute of East 
Asian Studies, University of California, Berkeley, Center for Chinese Studies, 1996. 
Fairbank, John K. Trade and Diplomacy on the China Coast. Cambridge: Harvard 
University Press, 1953. 
Fan 1-chun. "Long Distance Trade and Market Integration in the Ming-Ch' ing Period, 
1400-1850." Ph.D. diss., Stanford University, 1992. 
Fang Xuejia m~*-Kejia yuanliu tan 'ao ~%~¥nt~~- Guangzhou: Guangdong 
gaodeng j iaoyu chubanshe, 1994. 
Faure, David. "The Lineage as a Cultural Invention: The Case of the Pearl River Delta." 
Modern China 15. l (1989): 4-36. 
----------. "What Made Foshan a Town? The Evolution of Rural-Urban Identities in Ming-
Qing China." Late Imperial China, Vol. 11 , No.2 (Dec. , 1990): 1-31 . 
276 
----------. The structure of Chinese rural society : lineage and village in the eastern New 
Territories, Hong Kong. Hong Kong; New York: Oxford University Press, 1986. 
Faure, David and Helen F. Siu. "Introduction." In Down to Earth: The Territorial Bond in 
South China. Eds. David Faure and Helen F. Siu 1-20. Stanford: Stanford University 
Press, 1995. 
Fay, Peter Ward. The Opium War, 1840-1842. Chapel Hill: University of North Carolina 
Press, 1975. 
Feuerwerker, Albert. Rebellion in Nineteenth-Century China. Ann Arbor: Center for 
Chinese Studies, University of Michigan, 1975. 
-----------------------. Studies in the Economic History ofL ate Imperial China: Handicraft, 
Modern Industry, and the State. Ann Arbor: University of Michigan, Center for 
Chinese Studies, 1995. 
Foshan Zhongyi xiangzhi {~Ll.J ,~,~~~~ - 1830. 
"Poling Tiandihui junji wenfang sishi tiaochen {~~3(:t-t!rf(./J~)(,m o] 4J{!5R~." In 
Shinmatsu no himitsu kessha-Shiryo hen m* O)f£,WMf!f!±-~~ ~ -Sasaki Masaya 
-ftc lz *IE~. Tokyo: Kindai Chugoku Kenkyu Iinkai, 1967. 
Freedman, Maurice. Chinese Lineage and Society: Fukien and Kwangtung. London: 
Athlone, 1966. 
----------. Lineage Organization in Southeast Kwangtung, London: Athlone, 1958. 
----------. ed. Family and Kinship in Chinese Society. Stanford: Stanford University Press, 
1970. 
Friedman, Edward. "The Revolution or Just Another Bloody Cycle: Swatow and the 1911 
Revolution." Journal ofA sian Studies 29.2 (Feb. 1970): 289-308. 
"Fuming zongbing dayuanshuai hong _gaoshi, ~~,tr}, 9; jcjcyrjJ ;~ 'a" 71' ." In Guangdong 
hongbing qiyi shiliao !J:*¥3=t~~~5':~. Guangdong sheng wenshi yanjiuguan, 
Zhongshan daxue lishi xibian, 61-62. Guangzhou : Guangdong renmin chubanshe, 
1992. 
Fu Tongqin {f[q]~. "Ming Qing shiqi di Guangdong shatian BAmB;j;ltf3:J:-tg!J:*¥9,;El3 ." In 
Ming Qing Guangdong shehuijingji xingtai yanjiu !J:*!f!±wrf£~%fl!Hif-f§'E. ed. 
Guangdong History Institute, 65-74. Guangdong: Renmin Chubanshe, 1985. 
Gaoming xian zhi r-\§J BA,(11~~- 1756. 
Geertz, Clifford. "The Integrative Revolution: Primordial Sentiments and Civil Politics in 
the New States." In Old Societies and New States. ed. Geertz, Clifford. Glencoe, Ill.: 
The Free Press, 1963. 
Gladney, Dru C. A1uslim Chinese: Ethnic Nationalism in the People's Republic ofC hina. 
Cambridge: Harvard University, 1991 
277 
Gongzhongdang Jiaqingchao Zouzhe 'gq:im;~~~~m, 1982. Taibei: Guoli Gugong 
Bowuyuan. 
Gongzhongdang Qianlong chao Zouzhe '§q:im;~[t~*m, Nongye lei -~~ 33. 
Taibei: Guoli Gugong Bowuyuan, 1982 Reprint. 
Goodman, Bryna. "The Locality as Microcosm of the Nation?-Native Place Networks 
and Early Urban Nationalism in China" Modern China, Vol. 21 No. 4 (October 1995), 
387-419. 
Guangdong hongbing qiyi shiliao /J:}i!Ut~~~~%;f21-. Guangdong sheng wenshi 
yanjiuguan, Zhongshan daxue lishi xibian. Guangzhou: Guangdong renmin chubanshe, 
1992. 
Guangdongsheng fan~an diaocha zhidaozu r * :ii, 15 ~ 1'"l 11t ffi ~ rn. . ed. Siy iren xuexi 
putonghua shouce !LY e:!.A ~~~:iii 3"-=f-.fflr. Guangzhou: Guangdong renmin 
chubanshe, 1958. 
Guangdongsheng minzhengting JJ:*:ii'~i&ft. ed. Guangdongsheng zhengqu tuce fl* 
:ii'i&~!Milff!J-. Guangzhou: Guangdong sheng ditu chubanshe, 1995. 
Guangdong Tongzhi fl*ITB.~. 1822. 
Guangzhou Fuzhi !J:fl'l.fff~. Taibei chengwen chubanshe, 1966 Reprint. 
Guo Shouhua l6;;~. Kejiayuanliuxinzhi ~~~mt:ffi~. Taibei, 1964. 
Guo Tingyi $6~tJ. Taiping tianguo shishi rizhi i;::4-Jcm%$ B ~t. Taibei: Shangwu 
Printing Co., 1976. 
Hall, S. ed. Representation: Cultural Representations and Signifying Practices. London: 
Sage, 1997. 
Hamberg, Theodore. The Visions ofH ung-siu-tshuen, and Origin oft he K wang-si 
Insurrection. New York: Praeger, 1854. 
Han Maoli ftil;fU. "Songdai Lingnan diqu nongye dili chutan 
5K1~~ffi:ttg~M:~:tm:tLJM~." Lishi dili M!J::t.m~ II (1993): 30-34. 
Harrell, Stevan. ed. Cultural Encounters on China's Ethnic Frontiers. Seattle and 
London: University of Washington Press, 1994 
----------. "Ethnicity, Local Interests, and the State: Yi Communities in Southwest China." 
Comparative Studies in Society and History, 32 (1990): 51-548. 
Hase, P.H. "The Mutual Defense Alliance (Yeuk) of the New Territories." Journal oft he 
Hong Kong Branch oft he Royal Asiatic Society 29 (1989): 384-88. 
Hayes, James. "Rural Society and Economy in Late Ch'ing: A Case Study of the New 
Territories of Hong Kong (Kwangtung)." Ch 'ing-shih Wen-t 'i 3 (1976): 33-71. 
278 
----------. The Rural Communities ofH ong Kong: Studies and Themes. Hong Kong: 
Oxford University Press, 1983. 
----------. The Hong Kong region, 1850-1911 : institutions and leadership in town and 
countryside. Hamden, Conn. : Archon Books, 1977. 
Henry, B.C. Ling-Nam, or Interior Views ofS outhern China. London: S.W. Partridge, 
1886. 
"He Liu deng shilong CQShi Y,Uany!)u fiiJ /\ ~~:i)tiE9¥~ EB ." In Guangdong hongbing 
qiyi shiliao JJt~JJU=~~~~;fS/-. Guangdong sheng wenshi yanjiuguan, Zhongshan 
daxue lishi xibian, 214-215. Guangzhou: Guangdong renrnin chubanshe, 1992. 
HeshanXianzhi nw~;:t. 1754. 
Hirayama Shu Zp-LI.J~. Zhongguo Mimi Shehuishi i:f:imf£,Wifd:~~-Shanghai: 
Commercial Press, 1912. 
Hobsbawm, Eric. "Introduction: Inventing Traditions." In The Invention of Tradition. eds. 
Hobasbawm and Ranger, 11-14. New York: Cambridge University Press. 1983. 
----------. Primitive Rebels: Studies in Archaic Forms ofS ocial Afovement in the 19th and 
20th Centuries. New York & London: W. W. Norton & Company, 1959. 
Ho Ping-ti. Studies on the Population ofC hina, 1368-1953. Cambridge: Harvard 
University Press, 1959. 
Honig, Emily, Creating Chinese Ethnicity: Subei People In Shanghai, 1850-1980. New 
Haven and London: Yale University Press, 1992 
Hsiao Kung-chuan Rural China: Imperial Control in the Nineteenth Century. Seattle: 
University of Washington Press, 1960. 
Hsieh Kuo-ching. "Removal of the Coastal Population in Early Tsing Period." Chinese 
Social and Political Science Review 15 (1932): 559-96. 
Hsieh T'ing-yu. "Origins and Migrations of the Hakkas." Chinese Social and Political 
Science Review 13 (220), 1929: 208-28 
Hiseh, Winston. "Triads, Salt Smugglers; and Local Uprisings: Observations on the 
Social and Economic Background of the Waichow Revolution of 1911." In Popular 
Movements and Secret Societies in China: 1840-1950, ed. Jean Chesneaux, 145-164. 
Stanford: Stanford University Press, 1972. 
Hobsbawm, Eric. "Introduction: Inventing Traditions." In The Invention of Tradition. eds. 
Hobasbawm and Ranger. New York: Cambridge University Press. 
Hu Chunfan Mli:$Jl, Hua Yu~~. Huang Shiqing ~tttm, and Wen Ji X.9R. "Shilun 
Qing _qianqi di juanrnian zhengze ~ ~ mfi DVJ :ttH ~ 5E !Sdft" Qingshi yanji uj i 
trr~W-t~ll[t!: 3. Chengdu: Sichuan Renmin Chubanshe, 1984; 150-65. 
279 
Hu Hsien-chin. The Common Descent Group in China and its Functions. New York, 
Johnson Reprint Corp., 1948. 
Hu Xizhang iS)lfiB·!t et al. Kejiafenghua ~*Jxt.f.-t. Guangzhou: Guangdong renmin 
chubanshe, 1997. 
Hu Zhusheng tJ:l ~~. Qingdai Hongmenshi w.r1~~ i'l .9,:. Shenyang: Liaoning renmin 
chubanshe, 1996. 
Hu'a*n*g: !T:it~l2gg~ui. "J t~tt. "Shijiu shigi zhon~ede Guangdong Tiandihui +. 71 t!t~c i:p pt t8 InXueshu yanjiu ~:fJi}fJ,:. No. 1. Guangdong renmin chubanshe, 1963. 
Huang Xianfan Jtf,W,I-1. Zhuangzu tongshi )J-±Jj~;iffl._9.:. Nanning: Guangxi minzu 
chubanshe, 1988. 
Huangqing zouyi ~m~~- comp. Qinchuanjushi ~JI I@-±. Taibei: Wenhai 
chubanshe, 1967. 
Huizhou Fuzhi !Ul'liff :t. 1881. Taibei: Chengwen chubanshe, 1966 Reprint. 
Hummel, Arthur W. ed. Eminent Chinese oft he Ch 'ing Period (1644-1912) (Washington: 
United States Government Printing Office, 1943-44. 
Huntington, Ellsworth. Civilization and Climate. New Haven: Yale University Press, 
1924. 
Isaacs, Harold R. "Basic Group Identity: The Idols of the Tribe." In Ethnicity: Theory and 
Experience. eds. Glazer, Nathan and Daniel P. Moynihan. Cambridge, Mass.: Harvard 
University Press, 1975. 
Isajiw W.W. "Definitions of Ethnicity." Ethnicity 1 (1974): 111-24. 
Jen Yu-wen. The Taiping Revolutionary Movement. New Haven: Yale University Press, 
1972. 
----------. Taiping Tianguo quanshi °j(.:iJZ*00~-9.:. Hong Kong: Jianshi mengjin shudian, 
1962. 
Jiang Bingzhao ~;lrgilJ . "Lun kejia yu dangdi minzu zhu guanxi" 
~~* ~ 'i' :1-m~~ ~ ~ ~. Xiarnen daxue xuebao, zhe she ban MF i*~~¥~ fg-ffd:!t& . 
1992.4 (Oct. 1992): 106-12. 
Jiang Taixin IT:kffi and Duan Xueyu ~~:f. "Lun Qingdai qianqi tudi kentuo dui 
shehuijingji fazhan di yingxiang ~mt~frJ:ltJ ± :imTf6f·Hct:iU~Utf~l~:t-lQ~~-" 
Zhongguo jingjishi yanjiu q=i ~*~~ 7: 1irf 5\: 1 (1996). 
Jiang Z1:1_xuan ~Wt§.~ and Fang Zhiqin 1];t~ et al. Guangdong Tong Shi. Gudai 
/Jt~ims!::. ~~- Guangzhou: Guangdong gao dengjiao yu chubanshe, 1996. 
----------. Jianming Guangdong Shi 1m SJl!J:*.'t: -Guangzhou: Guangdong renmin 
280 
chubanshe, 1993. 
Jiaying Zhouzhi ~ ffl f l·I~ - 1898. 
Jin Yudi ½fm~ ed. Taiping Tianguo shiliao )cZp-7:~~~- Beijing: Zhonghua shuju, 
1953. 
Johnson, Graham. "Family Strategies and Economic Transportation in Rural China: Some 
Evidence from the Pearl River Delta." In Chinese Families in the Post-Mao Era. eds. 
Deborah Davis and Stevan Harrell. Berkeley: California University Press, 1993. 
Junjichu zouzhe lufu •~~J.H!Hfci~t Nongminyundonglei jl.5;:;iitiJ~. Beijing: Fist 
Historical Archives. 
Junjichutang zouzhe lufu •~~m~iU~m. Taibei: Guoli Gugong Bowuyuan, 1982 
Reprint. 
Kaiping Xianzhi 00+~~-1823. 
Keyes, Charles F. "The Dialectics of Ethnic Change." In Ethnic Change. ed Charles F. 
Keyes. Seattle: University of Washington Press, 1981. 
Kiang, Clyde Y. The Hakka Odyssey & Their Taiwan Homeland. Elgin: Allegheny Press, 
1992. 
Kuhn, Philip A. "Origins of the Tai ping Vision: Cross-Cultural Dimensions of a Chinese 
Rebellion." Comparative Studies in Society and History 19, no. 3. 
----------. Rebellion and its Enemies in Late Imperial China, militarization and social 
structure, 1796-1864. Cambridge: Cambridge University Press, 1970. 
----------. "The T'uan-lien Local Defense System at the time of the Taiping Rebellion." 
Harvard Journal ofA siatic Studies 27 (1967). 
----------. "The Taiping Rebellion." In The Cambridge History ofC hina, Volume 10, Late 
Ch 'ing, 1800-1911, Part I. ed. John K. Fairbank. Cambridge: Cambridge University 
Press, 1978. 
Krone. "A notice of the Sanon District. ' China Branch, Royal Asiatic Society, 
Transactions. (1859). 
Laai Yi-faai. "The Part Played by the Pirates of Kwangtung and Kwangsi Provinces in the 
T'ai-p'ing Insurrection." Ph.D. diss., University of California, Berkeley. 1950. 
----------. "River Strategy: a place of the Tai ping's Military Development," Orie ns 5 
(1952). 
Lamley, Harry J. "Subethnic Rivalry in the Ch'ing Period." In The Anthropology of 
Taiwanese Society. eds. Emily M. Ahem and Hill Gates, 282-318. Stanford: Stanford 
University Press, 1981. 
281 
----------. "Hsieh-tou: The Pathology of Violence in Southeastern China." Chi 'ing-shih 
wen-t 'i 3.7 (1977): 1-39. 
----------. "Lineage and Surname Feuds in Southern Fukien and Eastern Kwangtung 
Under the Ch'ing." In Orthodoxy in Late Imperial China. ed. Kwang-Ching Liu, 255-
78. Albany: State University of New York Press, 1990. 
----------. "Lineage Feuding in Southern Fujian and Eastern Guangdong Under Qing 
Rule." In Violence in China: Essays in Culture and Counter culture. eds. Jonathan N . 
Lipman and Stevan Harrell, 27-64. New York: State University of New York Press, 
1990. 
Landsberger, Henry A. Rural Protest: Peasant Movements and Social Change. N.Y.: 
Harper & Row Publishers, Inc. 1973. 
Lane-Poole, Stanley. The life ofS ir Harry Parkes, K.C.B. , GC.MG, sometime Her 
Majesty's minister to China and Japan. Vol. I. London, New York: Macmillian and co., 
1894. 
Lang Jingxiao ~~~ W. "Zhongguo nanfang xiedou z*hi: nyu~anUyi:n ji jizuzhi q=i~ffiJJ~IMlz~lzs!N.JUJl~" Dongfangzazhi 30 no. 19 (1933): 81-96. 
----------. "Qingdai Yuedong xiedou shishi frHt -1' *~I M1 j: ff ." Lingnan xuebao 
~ffi ~¥Ii 4.2 (1935). 
Lary, Diana. Region and Nation: the Kwangsi Clique in Chinese Politics, 1925-193 7. 
Cambridge: Cambridge University Press, 1975. 
Lau, Chunfat. The Decline oft he General Hakka Accent in Hong Kong: A Comparison of 
"Old-Style " and "New-Style " as spoken by the indigenous inhabitants. Milnchen: 
Lincom Europa, 2000. 
Lavely, William and R. Bin Wong. "Revising the Malthusian Narrative: The Comparative 
Study of Population Dynamics in Late Imperial China." Journal ofA sian Studies 57.3 
(August 1998): 714-48. 
Lechler, Rudolph. "The Hakka Chinese." Chinese Recorder and Missionary Journal. Vol. 
9 (Sep-Oct. , 1878): 352-359. 
Lee, James. "Migration and Expansion in Chinese History." In Human Migration. eds. 
William H. McNeil and Ruth S. Adams, 20-47. Bloomington: Indiana University Press, 
1978. 
Leong, S. T. "The Hakka Chinese ofLingnan: Ethnicity and Social Change in Modem 
China." In Ideal and Reality: Social and Political Change in Modern China, 1860-
1949. eds. David Pong and Edmund S. K. Fung, 287-326. Lanham: University Press of 
America, 1985. 
----------. Migration and Ethnicity in Chinese History: Hakkas, Pengmin, and Their 
Neighbors. Stanford: Stanford University Press, 1997 
282 
Liang Fangzhong ~1Jfrr. ed. Zhongguo lidai renkou, tiandi, tianfu tongji i=p OOJnf~A 
D, EB t-t!!, EB it#Ht. Shanghai: Shanghai renmin chubanshe, 1980. 
Li Hua **· "QiJ?-gchao qianqi Guangdong de shan~e___yu shangren 
m~JlWJMJJJIH1'Jj1,:ij~~ j1,:ijA." Xueshu Yanjiu ~ ~{ft~, No, 2 (1982): 39-44. 
Li Mo $it "Lun kejia de xingchengji minzu ronghe" ~~*B"JMnlt:&t-t~~:l!,g-. 
Zhongguoshiyanjiu i:p~;!::fiff~. (1993.4): 117-29. 
Li Wenzhi $)(~, et al. comps. Zhongguojindai nongyeshi ziliao i:p~lli:~.l:~~~;pj-. 
Beijing: Sanlian Shudian, 1957. 
Liao, David C. E. The Unresponsive: Resistant or Neglected?: The Hakka Chinese in 
Taiwan Illustrate a Common Missions Problem. Chicago: Moody Press, 1972. 
Liang Rencai ~t;f3. Guangdong Jingji dili JJJlU~~t-tg~. Beijing: New Science 
Publishers, 1956. 
Lin Daquan #ii~. "Keshuo ~~." In Chayang sanjia wenchao "fF-:~ - *XiY. ed. 
Wen Tingjing. Taibei: Wenhai chubanshe, 1966. 
Lin Zexu #!'!.IJ~. "Liangguang zougaoffiJJ:~~ -" In Lin wenzhonggong zhengshu 
#Jt,~,~i&~. Lin Zexu ;l%!'!,IJ~ . Taibei: Taiwan Shangwu Yinshuguan, 1968 Reprint. 
Lindley, Augustus F. Ti-ping Tien-Kwoh: The History oft he Ti-ping Revolution, Vol. I. 
London: Day and Son, Ltd., 1866. 
Little, Daniel and Joseph W. Esherick. "Testing the Testers: A Reply to Barbara Sands 
and Ramon Myers's Critique of G. William Skinner's Regional Systems Approach to 
China," Journal ofA sian Studies 48, no. I (Feb., 1989): 90-99. 
Liu Pi~g xU-1¥, B~i yiw_<;,_ng de z~anz_hen,_g: -!if!nfen_g Tongzhi nianHa!l Guangdong Tuke 
da x1edou yary1u ~~£8"]~~ . fil.f =F°-IEJ~~lsJ1*±~*~4:0Jfo/E. 
Beijing: Shangwu yinshuguan, 2003. 
Liu Weizhang ~ff;~- "Taiping Tianguo Gemingshi Qingchao Guangdong Caizheng 'fr. 
f::Kli.:iJ-®B1mlA,*1Mllili." Chap. in Taiping Tianguo yu Jindai Zhongguo'fr.5f=K 
001::j ili: f~ i=p 00 . Guangzhou: Guangdong renrnin chubanshe, 1993. 
Liu Zhiwei. "Lineage on the Sands: the Case of Shawan." In Down to Earth: The 
Territorial Bond in South China. eds. David Faure and Helen F. Siu, 21-43. Stanford: 
Stanford University Press, I 995. 
Lo, Wan. "Communal Strife in Mid-Nineteenth Century Kwangtung: The Establishment 
ofCh'ih-ch'i." Papers on China 19 (Harvard University, Dec. 1965): 85-119 
LonJ}~~i ff~i&~, "yue~i duanlian jilue,_m 1' iz!i lll~fll~ ff." In Jingdetang wenji 
t,:cffi¥X~. Ta1be1: Y1wen, 1970 Repnnt. 
Lozada, Eriberto P. Lozada, Jr. God Aboveground: Catholic Church, Postsocialist State, 
283 
and Transnational Process in a Chinese Village. Stanford: Stanford University Press, 
2001. 
Lu Baoqian ~-=f. Lun wan-qing Liangguang de Tiandihui zhengquan 
~Bi~ffi!J:Et9xtfuifi&tl. Taibei: Zhongyang yanjiujindaishi yanjiusuo, 1975. 
Lu,M ingzI:iong .§ ~ q:i. "Hanzu naJ?._qian p Lingnan Baiyue diqu de zaoqi__kaifa 
~:n~ffi:i!t~~1¥fs•:l:tE. li s'-J.fp.w=J ~/H§t." Zhongguoshiyanjiu q:i~~{tt~, No. 4 
(1984): 17-24. 
"Lil Zigui gongci 8-=f ti{!UiiJ." In Guangdong hongbing qiyi shiliao JJ:*#t~~~;t;fSI-. 
Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi xibian, 108-110. 
Guangzhou : Guangdong renmin chubanshe, 1992. 
Luo Baoshan !%'.i:~. "Guangdong Hongbing qiyi luelun 1*¥A~~x ~it." In 
Lingnan Wenshi ~ffiX~. Guangzhou: Guangdongsheng wenshi yanjiuguan, 1983 : 
82-89. 
-------:--· ",Guangdo~g !iandihui qiyi qij~~ zh?ngwai fandongpai de _g_o~jie ;~ft-th~ 
~ X M lsJ ~ jr &_ZyJ j~ 8'-J'ZJ ~ ." In Tazpmgtzanguo xuekan :t( f ::K III ~flj. Be1Jmg: 
Zhonghua shuju, 1983: 416-429. 
----------. "T~iping tianguo shiji Guangdong Tiandihui qiyi shulue 
:k2P-x~~M!J:*xtfuWl~~~1Ht." Zhongshan Daxue xuebao q:i Ll.J ::;kijHJft 4-1 
(1981 ): 54-62. 
Luo Ergang mfflffl. Tiandihui Wenxianlu x rfu ifXIH!fe. Shanghai : Zhengzhong shuju, 
1943. 
----------. Taiping Tianguo Shigang :k2P-x~~ffl. Shanghai, Shanghai shudian, 1937. 
----------. Taiping tianguo shiliao kaoshiji :k2P-x~~~~ff~. Beijing: Shanlian, 1956. 
Luo Xianglin m~t-t, comp. Kejia shiliao huipian ~*~t.2J-!Ii~. Hongkong: Zhongguo 
xueshe, 1965. 
----------. Kejiayanjiu daolun ~*Tiff~~~- Xinning: Xishan shucang, 1933 
----------. "K~~j_ifa fyiu*a nliu kao" ~*~mt:~. In Kejia yanjiu ~*Tiff~ . eds. Zhang Weidong and Wang Hongyou ::E#tt:, 3-77. Shanghai: Tongji daxue 
chubanshe, 1989. 
----------. Hong Kong and its External Communications before 1842. Hong Kong: 
Institute of Chinese Culture, 1963. 
----------. "Guangdong minzu gailun '* ~ :n~ffi it. " Minsu zhoukan ~f§~fU 63. 
Special issue, 1929. 
Nakagawa Manabu g:iJ II ~ . "Chugoku kyakka shi kenkyu no shin doko" 
q:i~~*~Tiff~ O) if] [o] . Hitotsubashi ronso ---;fl~:@ 77.4 (1977): 69-81 
284 
Lutz, Jessie G. and Lutz, Rolland Ray. Hakka Chinese Confront Protestant 
Christianity,1850-1900: With the Autobiographies ofE ight Hakka Christians, and 
commentary. Armonk & London: M.E. Shape, 1998. 
Maeda Katsutaro WT EB .,W~j~. "Shindai no Kanton ni okeru nomin toso no kiban 
frr ~ }J:*f:: ii tt G Jl ~-4,~ (7)~~-" Toyo gekuho *?'$~¥Ii, 51 (March 1969): 1-38. 
Ma*i B~in~g-ju"n  ~*i~. "Heshan Maishi zupu ji yutu jishi lunlue ~ LU ~ ft:lit~& ~ 00 i-ei In Guangdong hongbing qiyi shiliao 11:*~:Ait9~9:tl-. Guangdong sheng 
wenshi yanjiuguan, Zhongshan daxue lishi xibian, 1014-1024. Guangzhou : 
Guangdong renmin chubanshe, 1992. 
Mantaro J. Hashimoto. The Hakka Dialect: a Linguistic Study of its Phonology, Syntax 
and Lexicon. Cambridge: Cambridge University Press, 1973. 
Marks, Robert B. Tigers, Rice, Silk, and Silt: Environment and Economy in Late Imperial 
South China. Cambridge: Cambridge University Press, 1998. 
----------. "Rice Prices, Food Supply, and Market Structure in Eighteenth-Century China." 
Late Imperial China 12.2 (December 1991): 64-116. 
Martin, Howard J. "The Hakka Ethnic Movement in Taiwan, 1986-1991." In Guest 
People: Hakka Identity in China and Abroad. ed. Nicole Constable, 176-195. Seattle and 
London: University of Washington Press, 1996. · 
Matsuda Yoshiro ~EB~ j ~- "Minmatsu Shinsho Kanton Juko-deruta no Shaden Kaihatsu 
Shihai no Keisei Katei F!ll~~t]JJL:*~~'.:If'' ;t,, i (7)¥YE8ml§:6 C ~~M!(7) :;z:~cIDnx:~ 
fJi: ." Shakai Keizai Shigaku Jft±~f~~9:~ Vol. 46 (1981) no. 6: 59-66. 
Mazumdar, Sucheta. "A History of the Sugar Industry in China: The Political Economy 
of a Cash Crops in Guangdong, 1644-1834." Ph.D. diss. , UCLA, 1984. 
Meadows, Thomas T. The Chinese and Their Rebellions. London: Smith, Elder, 1856. 
Meskill, Johanna Menzel. A Chinese pioneer family: the Lins of Wu-Jeng, Taiwan, 1729-
1895. Princeton, N.J.: Princeton University Press, 1979. 
Miao Hongji ~~~- Zhujiang Sanjiaozhou Shitu Ziyuan I51;:i[_ fir3 rfH7..K ±~V!. 
Guangzhou: Zhongshan Daxue Chubanshe, 1988. 
Michael, Franz. "Regionalism in Nineteenth-Century China," introduction to Stanley 
Spector, Li Hung-chang and the Huai Army Seattle: University of Washington Press, 
1964. 
Michael, Franz with Chang Chung-Ii. The Taiping Rebellion: History and Documents, Vol. 
1, History. Seattle: University of Washington Press, 1966. 
Milne, William. "Some Account of a Secret Association in China, Entitled the Triad 
Society." Transactions oft he Royal Asiatic Society of Great Britain and Ireland 1 
(1827). 
Morgan, W. P. Triad Societies in Hong Kong. Hong Kong: Government Press, 1960. 
285 
Murray, Dian H., in collaboration with Qin Baoqi. The Origins oft he Tiandihui the 
Chinese Triads in Legend and History. Stanford: Stanford University Press, 1974. 
----------. Pirates oft he South China Coast, 1790-1810. Stanford: Stanford University 
Press, 1987. 
NanhaiXianzhi ffi?'B.H%;t. 1872. Taibei: Chengwen chubanshe. 1967 Reprint. 
Naquin Susan. Millenarian Rebellion in China : The Eight Trigrams Uprising of 1813. 
New Haven: Yale University Press, 1976 
----------------. Shantung Rebellion: The Wang Lun Uprising of 1774. New Haven: Yale 
University Press, 1981 . 
Naquin, Susan and Evelyn Rawski. Chinese Society in Eighteenth Century. New Haven: 
Yale University Press, 1987. 
Niida Noboru t:#EB~. Chugoku no noson kazoku i:pm(l)J.IJ1~t~. Tokyo, Tokyo 
Daigaku, 1952. 
Nishikawa Kikuko il!f5'i;::g~ -:F- . "Juntoku danren so~_yokuno seitatsu Jll&ffi~~t~m!J(l) 
,9X.ll." Toyobunka kenkyusho kiyo }R¥$ X1{3i}fyt]i)Ttc~105 (1988). 
----------. "Qingdai Zhujiang Sanjiaozhou Shatiankao m1~Et:u.= ;iEJ V·H¥0 ' E8 ~." In 
Lingnan Wenshi itffiX9:. tran. Cao Leishi ~~E, 11-14. Gungdongsheng Wenshi 
Yanjiuguan, 1983. 
North China Herald, No. 301 (may 3, 1856), and No. 359 (June 13, 1857). 
Overland China Mail. Hong Kong, 1848-1909. 
Ownby, David. "The Ethnic Feud in Qing Taiwan: What is This Violence Business, 
Anyway? An Interpretation of the 1782 Zhang-Quan Xiedou," Late Imperial China, 11 
(1990), pp. 75-98 
----------. Brotherhoods and Secret Societies in Early Mid-Qing China: The Formation of 
A Tradition. Stanford: Stanford University Press, 1996. 
----------. "Communal Violence in Eighteenth Century Southeast China: The Background 
to the Lin Shuangwen Uprising of 1787." Ph.D. diss. , Harvard University, 1989. 
----------. "The Heaven and Earth Society as Popular Religion," in The Journal ofA sian 
Studies 54, no. 4 (Nov. 1995): 1023-1046. 
Oxfelld, Ellen. "Still 'Guest People': The Reproduction of Hakka Identity in Calcutta, 
India." In Guest People: Hakka Identity in China and Abroad. ed. Nicole Constable, 
149-175. Seattle and London: University of Washington Press, 1996. 
Pan, Lyn, Paul De Angelis, ed. Sons oft he Yellow Emperor: A History oft he Chinese 
Diaspora. Kodansha America Inc. , 1994 
286 
PanyuXianzhi .ffi: ~1~ ~ - 1871. Taibei: Chengwen chubanshe. 1967 Reprint. 
Parliament Papers 1859, Session 2, v. 23 
Pasternak, Burton. Kinship and Community in Two Villages. Stanford: Stanford Univ. 
Press, 1972. 
Peng Yuxin ~ffiffi. Qingdai tudi kaiken shi m1~±:1:-&ml~ ~- Beijing: Nongye 
Chubanshe, 1990. 
Perkins, Dwight H. Agricultural Development in China, 1368-1968. Chicago: Aldine 
Publishing Company, 1969. 
Perry, Elizabeth J. Rebels and Revolutionaries in North China, 1845-1945. Stanford: 
Stanford University Press, 1980. 
Peterson, Anya, Ethnic Identity: Strategies ofD iversity. Bloomington: Indiana University 
Press, 1982. 
Piton, Ch. "On the Origin and History of the Hakkas." China Review, 2 (July 1873-June 
1874): 222-26. 
Polachek, James M. The Inner Opium War. Cambridge: Harvard University Press, 1992. 
Pong, David. A Critical Guide to the Kwangtung Provincial Archives: Deposited at the 
Public Record of Office ofL ondon. Cambridge and London: Harvard University Press, 
1975. 
----------. "The K wangtung Provincial Archives at the Public Record Office of London: A 
Progress Report," Journal ofA sian Studies, vol. 28, Issue 1 (Nov. 1968). 
Potter, Jack. "Land and Lineage in Traditional China." In Family and Kinship in Chinese 
Society. ed. Maurice Freedman. Stanford: Stanford University Press, 1970. 
Powers, Marshall K. "Chinese Coolie Migration to Cuba." Ph.D. Dissertation. University 
of Florida, 1953. 
Purdue, Peter. Exhausting the Earth: State and Peasant in Hunan, 15 00-185 0. 
Cambridge: Harvard University, Council on East Asian Studies, 1987. 
Qin Baoqi ~jff~ and Liu Meizhen t U~f&. eds. Tiandihui. J(J-!QW{ 7 vols. Beijing: The 
Number One Historical Archive, 1981-89 
----------. Qing Qianqi Tiandihui Yanjiu fflr ~]Wl/jJ(J1Q Wf lirf~. Beijing: Renmin daxue 
chubanshe, 1988. 
Qinding Daqing huidian shili $XJE::k:fflr~ #!PJ{7'J. Taibei: Xinwenfeng, 1899. 
Qingshi fflrsl:. (History of the Qing Dynasty). Taibei: Huawen shuju, 1961 Reprint. 
287 
"Qing Wenzong _Xianfeng Shilu ~X* ~ -=F~3K-" In Guangdong hongbing qiyi shiliao 
JJ:*~~~~5':.~- Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi 
xibian, 1137-1202. Guangzhou: Guangdong renmin chubanshe, 1992. 
Qu Dajun lffi*:rey· Guangdong xinyu ll*WT~. Beijing: Zhonghua Shuju, 1985. 
Qujiang XZ EEll lCW*';G;, 1875. 
Rawski, Evelyn S. Agricultural Change and the Peasant Economy ofS outh China. 
Cambridge, Harvard University Press, 1972. 
Renhua XZ t: 1l:W*';G;, 1883. 
Rhoads, Edward. China 's Republican Revolution: The Case ofK wangtung, 1895-1913. 
Cambridge: Harvard University Press, 1975. 
Roberts, J. A.G. "The Hakka-Punti War." Ph.D. diss., Oxford University, 1969. 
Roger, Banister T. A History oft he External Trade ofC hina, 1834-81. Shanghai: 
Inspector General of Chinese Customs, 1931. 
Ruddle, Kenneth and Gongfu Zhong. Integrated Agriculture-Aquaculture in South China: 
The Dike-Pond System oft he Zhujiang Delta. Cambridge: Cambridge University Press, 
1988. 
Sands, Barbara and Ramon H. Myers. "The Spatial Approach to Chinese History: A 
Test," Journal ofA sian Studies 45, no. 4 (August, 1986): 721-43. 
Sasaki Masaya 1tr: 7 * IE~- Shinmatsu no himitsu kessha-Shiryo hen ~ * <7) ;f:£,ffi Mifrr.-
:tt~ • · Tokyo: Kindai Chugoku Kenkyu Iinkai, 1967. 
----------. "J untoku ken kyoshiri to Tokai jurokusha Jl!Jl ~ l1r*' ~B* $ c: *# l}-j--~ fj; ." Kindai 
Chugoku kenkyu ;Ji~q:i~iiff~ 3 (1959): 161-232. 
-----------. "Kampo yonen Kanton Tenchikai no haran ~_:_izg~)J:*:;}(t!P;~<l) &~L" 
Kindai Chugoku Kenkyu Senta iho ;Ji~q=i~1tf~k::,, ~ ~¥~, 2.3 (1963): 1-17. 
Scarth, John. Twelve Years in China: The People, the Rebels, and the Mandarins . 
Edinburgh: T. Constable and co., 1860. 
Schoppa, R. Keith. Chinese Elites and Political Change : Zhejiang Province in the Early 
Twentieth Century. Cambridge, Mass. : Harvard University Press, 1982. 
Schlegel, Gustave. Thian Ti Hwui: The Hung League or Heaven-Earth League. Batavia: 
Lange and Co., 1866. 
Schurhammer, Georg. Francis Xavier, His Life, His Times , Vol. 4. Rome: Jesuit 
Historical Institute, 1973-1982. 
Shaozhou Fuzhi fill fi'I ffJ ;G;, 1843. 
288 
Shepherd, John Robert. Statecraft and Political Economy on the Taiwan Frontier. 
Stanford: Stanford University Press, 1993. 
Shi Naian. The Marshes ofM ount Liang : A New Translation oft he Shuihu zhuan, John 
and Alex Dent-Young, trans. Hong Kong: Chinese University Press, 1994. 
Shih, Vincent Y. The Taiping Ideology: Its Sources, Interpretations, and Influences. 
Seattle and London: University of Washington Press, 1967. 
Shils, Edward. "Primordial, Personal, Sacred and Civil Ties." British Journal of 
Sociology 8 (1957): 130-45. 
Shirokogoroff, S. M. Anthropology ofE astern China and Kwangtung Province. Royal 
Asiatic Society, North China Branch, Extra Vol. IV: Shanghai, 1925. 
Shunde Xianzhi }lw:{~~;s. 1929. Taibei: Chengwen chubanshe, 1966 Reprint. 
"Shundexian huifeilan duanyuanvou Jlw:i ~~if!~ IL Yiffi ~ 83 ." In Shinmatsu no himitsu 
kessha-Shiryo hen frlr* O)i!,ffi*tirrr-:Gt~ ii. Sasaki Masaya {ti: 1-.z *IE~-
Tokyo: Kindai Chugoku Kenkyii Iinkai, 1967. 
Sihui Xianzhi IZ:9i!ff~;=s. 1896. Guangzhou: Guangdong Renmin chubanshe, 1996 
Reprint. 
Skinner, G. William. "Cities and the Hierarchy of Local Systems." In City in Late 
Imperial China. ed. G. W. Skinner, 275-352. Stanford: Stanford University Press, 1976. 
----------. "Chinese Peasants and the Closed Community: An Open and Shut Case." 
Comparative Studies in Society and History, 13 (1971): 270-281. 
----------. "Regional Urbanization in Nineteenth Century China." In City in Late Imperial 
China, ed. G. W. Skinner, 212-49. Stanford: Stanford University Press, 1976. 
Smith, Anthony D. The Ethnic Origins ofN ations. Oxford: Basil Blackwell, 1986. 
So, Alvin Y. The South China Silk District: Local Historical Transformation and World-
System Theory. New York: State University of New York Press, 1986. 
Soda Yo ffl EB!$ . "Kokinko }IrtJ~." Toyoshi kenkyu *1$ ~ wf~, 38.4 (March 1980). 
Spence, Jonathan D. The Taiping Vision ofC hristian China. Waco, TX: Baylor University, 
1996. 
----------. God 's Chinese Son: The Taiping Heavenly Kingdom of Hong Xiuquan. New 
York: W.W. Norton & Company, Inc. , 1996. 
Stauffer, Milton T. ed. Christian Occupation ofC hina. Shanghai: China Continuation 
Committee 1922. 
Struve, Lynn A. The Southern Ming, 1644-1662. New Haven, Conn.: Yale University 
Press, 1984. 
289 
Su Fengwen f/iJfj..)( . Gufei zonglu B1~t~~-Taibei: Xiangsheng chubanshe, 1975 
Reprint. 
----------. "Pinggui jilue Zfia;2. ~ ." In Guangdong hongbing qiyi shiliao }J:].f[~;A;~~ 
~fl. Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi xibian, 772-813. 
Guangzhou : Guangdong renmin chubanshe, 1992. 
Taishanxian dang'anguan ~ Wi!-f13~ffi'. ed. Taishan dilizhi ii'Li.J:tl!ffJ.;'t. Taishan: 
Zhongguo Taishanxian weiyichuanbu, 1984. 
Tan Xin ?!{ ,, . ed. Siy ifengguang IZ.9 E.00.:3/6. Hong Kong: Wanye Chubanshe, n.d. 
Tan Zuen i!JUL~. "Xinhuijingbian shilue ffi~*rf'~i,R~." In Guangdong hongbing qiy i 
shiliao }J:].f[~;A;~~~fl. Guangdong sheng wenshi yanjiuguan, Zhongshan daxue 
lishi xibian. Guangzhou : Guangdong renmin chubanshe, 1992. 
Tang Shen m~ and Li Lon__g_qian $~M. "Mi~qin:& Guangdong Jingji Zuowu de 
Zhongzhiji Yiyi ~ ~!Jf3f[~ ~ft 4me{]l! *1Lbt ~~" In Mingqing Guangdong Shehui 
JingjiXingtai Yanjiu ~ ~!JfJ:IH±wr~~~-li}f J'E. Guangzhou: Guangdong Renmin 
Chubanshe, 1985, 1-21. 
"Tao Daqing xiwen !1::kfrf~X." " In Guangdong hongbing qiy i shiliao !J:].f[~;A;~~ 
~fl. Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi xibian, 38-40. 
Guangzhou : Guangdong renmin chubanshe, 1992. 
Teng Ssu-yu. New Light on the History oft he Taiping Rebellion. New York: Russell & 
Russell, 1966. 
Terrill, Ross. 800,000,000: The Real China. Boston: Little, Brown and Co. , 1971. 
The Chinese and Japanese Repository 3 (1865). 
The Taiping Revolution. Beijing: Foreign Language Press, 1976. 
Thernstrom, Stephen. ed. , Harvard Encyclopedia ofA merican Ethnic Groups. 
Cambridge: Harvard University Press, 1981. 
T'ien Ju-kang. The Chinese Sarawak: A Study ofS ocial Structure. London: L.S.E. 
Monographs on Social Anthropology No. 12, 1956. 
Tilly, Charles, Louise, and Richard. The Rebellious Century, 1830-1930. Cambridge: 
Harvard University Press, 1975. 
Toland, Judith. "Introduction: Dialogue of Self and Other: Ethnicity and the Statehood 
Building Process." In Ethnicity and the State, ed., Toland, Judith D. New Brunswick, 
N.J.: Tansaction Publishers, 1993. 
Tregear, T. R. Hong Kong Gazetteer to the Land Utilization Map ofH ong Kong and the 
New Territories, with China and English Names. Hong Kong: Hong Kong University 
Press, 1958. 
290 
Vermeer, E. B. "Introduction: Historical Background and Major Issues." In Development 
and Decline ofF ukien Province in the Jih and 18th Centuries. ed. E. B. Vermeer, 5-34. 
E. J. Brill, 1990. 
Vos, George de. "Ethnic Pluralism: Conflict and Accommodation." In Ethnic Identity: 
Cultural Continuities and Change. eds. George de Vos and Lola Romanucci-Ross. 
Chicago: University of Chicago Press, 1975. 
Wakeman, Frederic, Jr. "The Secret Societies ofKwangtung, 1800-1856" In Popular 
Movements and Secret Societies in China: 1840-1950. ed. Jean Chesneaux. Stanford: 
Stanford University Press, 1972. 
----------. "'High Ch'ing: 1683-1839." In Modern East Asia: Essays in Interpretation. ed. 
J.B. Crowley, 1-28. New York: Harcourt, Brace and World, 1970. 
----------. Strangers at the Gate. Berkeley and L.A.: Univ. of California Press, 1966. 
----------. "The Canton Trade and the Opium War." In The Cambridge History ofC hina, 
Vol. 10. ed. Denis Twitchett and John K. Fairbank, 1978-1987. Cambridge: Cambridge 
University Press, 1980. 
Wang Tianj_i_ang .3:.:K~, "19shiji xiaban de zhongguo di mimi shehui 19tltti:.r :¥B'-Ji:p 
00±-&MWf±~," Lishi yanjiu }Jj .t,:U]fo/'[. No. 2. Beijing, 1963. 
Wang Xin .3:./i. Wangzhuang wugong weiji .:Dl±fEtJ}JJi~, 1892. 
Wang Yeh-chien. Land Taxation in Imperial China, 1750-19JJ. Cambridge: Harvard 
University Press, 1973. 
Watson, James. "Chinese Kinship Reconsidered: Anthropological Approaches on 
Historical Research." China Quarterly 92 (1982): 589-627. 
Wei Yuan ~~- Shengwuji ~jt~c,. Shanghai: Shanghai Guji Chubanshe, 1995 reprint. 
Weller, Robert P. Resistance, Chaos and Control in China: Taiping Rebels, Taiwanese 
Ghosts and Tiananmen. Seattle: University of Washington Press, 1994. 
Wiens, Harold J. Chinas March Toward the Tropics: a discussion oft he southward 
penetration ofC hina's culture, peoples, and political control in relation to the non-
Han-Chinese peoples ofs outh China and in the perspective ofh istorical and cultural 
geography. Hamden, Conn.: The Shoe String Press, 1954. 
----------. Han Chinese Expansion in South China. Hamden, Conn.: The Shoe String 
Press, 1967. 
Williams Brackette. "A Class Act: Anthropology and the Race to Nation across Ethnic 
Terrain." Annual Review ofA nthropology 18 (1989): 401-4. 
Wolf, Eric R. Peasants. Englewood Cliffs: Prentice-Hall, 1966. 
291 
----------. Peasant Wars oft he Twentieth Century. New York, Harper & Row, 1969. 
Wong, J. Y. Yeh Ming-ch'en: Viceroy ofL iang Kuang 1852-8. Cambridge [Eng.]; New 
York : Cambridge University Press, 1976. 
----------. Deadly Dreams: Opium, Imperialism and the Arrow War (1856-1860) in China. 
Cambridge [U.K.]; New York: Cambridge University Press, 1996. 
----------. Peasant Wars oft he Twentieth Century. New York, Harper & Row, 1969. 
Wright, Stanley. The Origin and Development oft he Chinese Customs Service, 1843-
1911 An Historical Outline. Shanghai, 1939. 
Wu Bingheng ~ 3'Hij. "Xinfeng shouchengji shilue fj§ fl ~ Mic~~." In Guangdong 
hongbing qiyi shiliao }J:}f[{AA~ ~~:M. Guangdong sheng wenshi yanjiuguan, 
Zhongshan daxue lishi xibian, 1125-1136. Guangzhou: Guangdong renrnin chubanshe, 
1992. 
Wu Zhaoqing ~?l~ma nd Hao Zhiqing Mmm. Zhongguo Banghuishi q=i~Mir~. 
Taibei: Wenzhi chubanshe, 1996. 
Xianfengchao chou ban yi wu shi mo ~fllJ1~ ¥]!if~ !9Hi't~- Beijing, 1928 Reprint. 
Xiao Yishan ffi - LIJ. Jindai Mimi Shehui Shiliao i!i1~f.i?W:f±~ _tJ4 Vol. I. Changsha: 
Yuelu shu she Hunan sheng xin hua shu dian fa xing, 1986. 
----------. Qingdai tongshi m~ ~5e, Vol. I Taibei: Taiwan shangwu yinshuguan, 1962. 
Xiao Guojian ffiOO -ru!. Shenzhen diqujiazufazhan ~#11±-t!!IR~ ~.btli:. Hong Kong: 
Xianchao Shushi, 1992. 
XiangshanXianzhi ~ LlJ fvt,;::t . 1923. Taibei : Chengwen chubanshe. 1967 Reprint. 
Xinning Xianzhi ffi$fvt,Jt . 1893. 
Xinning xiangtu dili ffi$~B±:l:t!!:ff. Taishan: n.p., 1900. 
Xinyi XZ ffi1[fvt,Jt . Guangzhou: Guangdong renmin chubanshe, 1993 Reprint. 
Xunzhou Fuzhi ~j-l'l}f-JJt . 1874. 
Xuxiu nanhai Xianzhi i( {~ lfJ ;'BJ fvt, ;::t. 1872 
Xuxiu Guangzhou Fuzhi il {~!J:j-i'l!fJ Jt . 1759. 
Xuxu ~~- Lingnan gujinlu ihtWi--S-4'-JK. Guangzhou: Guangdong renrnin chubanshe, 
1992. 
Xu Xuzeng ~ Af!, if, "Fenghu Zaji _:_ ¥i!iJJ iHc." In Kejia shiliao huipian ~*5!:.;fSl-@E~. 
comp. Luo Xianglin, 279-99. Hongkong: Zhongguo xueshe, 1965. 
Yang, C. K. "Some Preliminary Patterns of Mass Actions in Nineteenth-Century China." 
292 
In Conflict and Control in Late Imperial China, eds. Frederic Wakeman Jr. and 
Carolyn Grant. Berkeley: University of California Press, 1975. 
----------. Chinese Village, Taitou, Shantung Province. New York: Columbia University 
Press, 1945. 
Yeh-chien Wang. "The Secular Trend of Prices during the Ch'ing Period (1644-1911 )." 
Journal oft he Institute of Chinese Studies oft he Chinese University ofH ong Kong 5 .2 
(Dec., 1972): 347-68. 
Yen Ching-hwang. Coolies and Mandarins: Chinas Protection of Overseas Chinese 
during the Late Ch 'ing Period (1851-1911) . Singapore: Singapore University Press, 
1985. 
Ye Xianen llf ~ Ji. "Ming Qing Zhujiang Sanjiaozhou tudi zhidu, Zongzu yu shangyehua 
1!13m~IT-Jf:11fii±:l:&ffi~N, *Mc--5i1ffi .illdt." Zhongguo Wenhua YanjiusuoXuebao i:p 
OOX{tliJfJi?fr~t~ 6. Guangzhou, 1997. 
Yong Kweng ~II. Sohak dongjumki iffi~:JU!Wf~c. Kwon, HeechuL tran. Seoul: 
Ulyumoonhwasa, 1974. 
Yongzheng zhupi yuzhi ~ IE ~!,ttlt~ \§' . Taibei: Wenhai chubanshe, 1965 Reprint. 
Yoo Jangkun %?-J-2. Kundae Joong'Kukyi Bimil Kyulsa -2t:B ~~ .9-l 1:ll W~ ;l Seoul: 
Koryowon, 1996. 
----------. "Yangkwangjiyuckyijayunkwa inmoon hwankyung ffi~:l-'t!rti.9-l § ~ 4 
AX~~," Joongkookyunkoo (Fall, 1993): 26-43. · 
Yu/in zhouzhi IHtfl·l;:t. 1894. Taibei: Chengwen chubanshe. 1967 Reprint 
Zhang_fengian ~•~- "Kejia minxi zhi yanhua 'f;.*.a:;~z~1r.." Taiwan Wenxian 
§~XIX. 13.4 (Dec., 1962). 
Zhang Kuamngiy¥u, *~*1-t¥. "Lun Kejiahua de xingcheng it~*itEr-J%,6.x;." Qinghua Xuebao Vol. 25.2 (1995): 161-180. 
Zhang Zhidong ~Z~. Zhang Wenxiang Gong Quanji ~XII~~- 1886. Taibei: 
Wenhai, 1963 Reprint. 
----------. Zhang Zhidong quanji * Z~~~- Shijiachuang: Hebei renmin chubanshe, 
1998. 
Zhao Yuanyi~_g. ~of~, "Hongbingjishi rIA*c$." In Hongjinjun qiyi ziliaoji fJ: rtJ'.llr 
~~~~~ (Guangdong Jindaishi Ziliao), 94-144. Guangzhou : Guangdongsheng 
Zhongshan tushuguan, 1992. 
Zelin, Madeline. The Magistrates Tale. Berkeley: University of California Press, 1984. 
Zengcheng Xianzhi :ttf~l'i%;:t (Gazetteer of Zengcheng County). 1754. 
293 
Zheng Dehua* 1 3~*-"Shijiou shiji mo Taishan qiaoxiang de xingchengji__g_i pouxi + n tt!J~ tr LL! ,flt ~3 i¥J M $".. l]z .g-utl T ." Guangzhou huaqiao yanji u IJ: 1-1'1¥ ~Tiff~ 4 
(1986). 
----------. "Guangdong Zhonglu Tuke Xiedou Yanjiu /J:*i:p~±~fflxl~wf~." Ph.D. 
diss. , University of Hong Kong, 1989. 
Zheng Dehua and Chef!g Luoxi $"..ffiffi. Taishan qiaoxiang yu Xinning tile 
ir Lll fit~3~ffi$fflb'ie-. Guangzhou: Sanlian shudian, 1991. 
"Zhen'geshi biji tt-x;~gg-i,2,." In Guangdong hongbing qiyi shiliao ~*#U=~JE9~1:*4-
Guangdong sheng wenshi yanjiuguan, Zhongshan daxue lishi xibian, 853-862. 
Guangzhou : Guangdong renmin chubanshe, 1992. 
Zhen~iexin ~ {mn.Ji. "Dachengguo de fanqing qiyi "j( ]Jx, 00 E8 &m !E9x."S hixue y uekan 
1:=r ~ flJ. Henan renmin chubanshe (Dec., 1958): 14-19. 
Zheng_Tianxiang ffl3J;:ff. Zhujiang Sanjiaozhou Jingji Dili Wangluo flt:l.I _ fi3V·IU~~:t-t!?. 
~~~- Guangzhou: Zhongshan University Press, 1991. 
Zhong Zhenwei fl~il ."Chen Kai, Li Wenmao lingdaode honajinjun qiyi ~Jf , c$.X 
;1@1-@ a{J h rfJ ~ JEQX." Zhongxue lishijiaoxue 9=1~JJ5 .'£~=r 9. Guangzhou: 
Guangdong renmin chubanshe, 1958: 16-18. 
Zhou Rijian ./E] B~ - Xinfengfangyanzhi ffi-$ Jj ~~-Guangzhou: Guangdong gaodeng 
jiaoyu chubanshe, 1990. 
Zhou Yumin ~¥f ~. "Zhongguo huidang went~anjiu shuping i:p 00i?:.Jt f6] ~:ri}fy';-:j£ 
·Vf." In Zhongguo huidangshi yanjiuhui i:p 00 A)tj::riJfJtf? ed. Huidangshi y anjiu f? 
.1t1::riJfy';-: . Shanghai: Dueling Chubanshe, 1984. 
----------. Zhongguo banghuishi i:p OOfflf? 1:. Shanghai: Shanghai renmin chubanshe, 
1993. 
Zhuang Jifa #fa~. Qingdai Mimi Huidangshi Yanjiu mrtf£,ffif{;I;~ Tiff~ . Taibei: 
Wenshi zhexue chubanshe, 1994. 
----------. "Qingdai rnimi huidang de tantao mrtf£,ffifl:l:1¥J~H," Zhongguo lishi 
xuehui shixuejikan i:pmM~~f{~~~fU, 16 (July 1984). 
----------. Qing Gaozong Shiquan Wugong Yanjiu rn~*+:i:fttJ}J:ri}fy';-: Taibei: Gugong 
bowuyuan, 1982. 
----------. "Qingdai Min-Yue Diqu de Renkou Liudong yu Mimi Huidang de Fazhan 
mrt00~:J:tg ~i¥J A rJ mt:i!J~f£,ffifl:i:1¥J~~ -" Jindai Zhongguo Chuqi Lishi Yantao 
Huilun Wenji ili:rt i:p mfJJMM~wfg-;j"fl~X~. Taibei: Academia Sinica, 1990. 
----------. Qingdai Tiandihui Yuanliugao mf\'.;J;::J:tgf{ l.W:mt~. Taibei: Shilin, 1982. 
----------. "Qingdai mimi huidang de tantao mrtf£,ffifllti¥J~g-;j"," Zhongguo lishi 
xuehui shixuejikan i:pmfff~~f{~~~fU, 16 (July 1984): 153-83 
294 
Zo, Kil Young. Chinese Emigration into the United States: 1850-1880. New York: Arno 
Press, 1978. 
Zhupi zouzhe ~JHJ:t~~' Nongminyundonglei ,IJ:i::;3.iib~. Beijing: Fist Historical 
Archives.