Boretz 2011 Gods, Ghosts, and Gangsters (Book)

$9521%25(7=
RITUAL VIOLENCE, MARTIAL ARTS, AND MASCULINITY

ON THE MARGINS OF CHINESE SOCIETY
*2'6
*+2676$1'
*$1*67(56
Gods, Ghosts, and Gangsters
Ritual Violence, Martial Arts, and Masculinity
on the Margins of Chinese Society
avron boretz
University of Hawai‘i Press

Honolulu
© 2011 University of Hawai‘i Press
All rights reserved
Printed in the United States of America
16 15 14 13 12 11 6 5 4 3 2 1
Library of Congress Cataloging-in-Publication Data

Boretz, Avron Albert, 1956-
Gods, ghosts, and gangsters : ritual violence, martial arts, and
masculinity on the margins of Chinese society / Avron Boretz.
p. cm.
Includes bibliographical references and index.
ISBN 978-0-8248-3377-0 (hardcover : alk. paper) —
ISBN 978-0-8248-3491-3 (softcover : alk. paper)
1. Martial arts—China—Religious aspects. 2. Martial arts—
Taiwan—Religious aspects. 3. Violence—China—Religious aspects.
4. Violence—Taiwan—Religious aspects. 5. Masculinity—China.
6. Masculinity—Taiwan. I. Title.
GV1102.7.R44B67 2011
796.815’5 — dc22
2010011684
University of Hawai‘i Press books are printed on acid-free

paper and meet the guidelines for permanence and
durability of the Council on Library Resources.
Designed by Josie Herr

Printed by Sheridan Books, Inc.
Contents
Acknowledgments vii
Note on Translation and Use of Foreign Terms ix
Chapter 1 Introduction 1
Chapter 2 Violence, Honor, and Manhood 21
Chapter 3 Taidong: The Mountains and Beyond 58
Chapter 4 Fire and Fury 87
Chapter 5 Tales from the Jianghu 115
Chapter 6 Wine, Women, and Song 176
Chapter 7 Conclusion: Faces of the Gods 204
Notes 213
Glossary 247
Bibliography 255
Index 269
v
Acknowledgments
This book is the product of countless collaborations and interactions

and would not exist without the participation and contributions of the
people whose lives and stories fill its pages. There is no way to conve-
niently bundle them all into one neatly labeled package, and to list
and thank each of them individually would require an entire chapter.
I will thus single out only a few, though I wish I could give all the men-
tion they deserve: I am particularly indebted to the people of Taidong,
Taiwan, and the villagers of Wuguanzhuang in Yunnan (China), who
allowed me to become, in some measure and for some time, a part
of their respective communities. In Taidong in particular, the devo-
tees of the Zhonghe Gong (Loyal Harmony Temple) and Wuan Gong
(Martial Security Temple), Xuanwu Shrine manager Li Jianzhi, the
residents of Fushan Village, and Ritual Master Lin Jiafu were instru-
mental in bringing this book to life.
For funding fieldwork and supporting write-up at various stages
of the project, I would like to thank the U.S. Department of Educa-
tion Center for International Education (Fulbright Foundation), the
Chiang Ching-kuo Foundation for International Scholarly Exchange,
the American Council of Learned Societies/Social Science Research
Council Joint Committee on Chinese Studies, the China Times Cul-
tural Foundation, the Pacific Cultural Foundation, and the Woodrow
Wilson Foundation. Additional assistance for research travel to China
and Taiwan was provided by three Summer Faculty Research Grants
from Hobart and William Smith Colleges and two Faculty Research
Assignments and a Dean’s Fellowship at the University of Texas at
Austin. I received a small grant as well from the Republic of China
(Taiwan) Council for Cultural Affairs (Wenjian Hui) for ethnographic
video production.
I am also indebted to Hu Taili and Lin Meirong, who generously
sponsored my affiliation with the Institute of Ethnology, Academia
Sinica, during the two two-year stints in Taiwan that bracketed the writ-
ing of this book, first as a research associate (1989 –1991) and then as
a visiting scholar (2004 –2006). I am especially grateful to Lin Meirong
vii
viii acknowledgments
for her gracious invitation to participate in the research and writing

of the Historical Gazetteer of Taidong County, an assignment that enabled
me to return to Taidong for three extended visits between those multi
year stays. My research in Yunnan began in 1994 as part of a multidis-
ciplinary collaborative U.S.-China research project on Buddhism and
popular religion of the Bai, funded by a grant from the Luce Founda-
tion. I am deeply grateful to my friend and principal investigator on
that project, John McRae, for inviting me to join the team as “project
ethnographer.”
I have incurred uncounted intellectual debts on the way to real-
izing this book. From inception to completion, Steven Sangren has
been a conscientious mentor and supportive colleague. His challeng-
ing and enthusiastic engagement continues to motivate and inspire
me, and his theoretical influence will be evident to anyone familiar
with his work. Over the years, Fiorella Allio, Yvonne Chang, Kenneth
Dean, Stephan Feuchtwang, Stevan Harrell, David Holmberg, the late
A. Thomas Kirsch, Murray Rubenstein, G. William Skinner, Robert J.
Smith, Donald Sutton, John Traphagen, Robert Weller, and Angela
Zito have read and commented on drafts of parts of the manuscript
or provided useful feedback during talks and presentations that were
among its precursors. I am grateful to all for their criticisms and sug-
gestions. The weaknesses and errors that remain are entirely the prod-
uct of my own shortcomings.
Some passages in chapter 4 appeared earlier in an article entitled
“Righteous Brothers and Demon Slayers: Subjectivities and Collec-
tive Identities in Taiwanese Temple Processions,” in Religion and the
Formation of Taiwanese Identities, ed. Paul Katz and Murray Rubenstein
(Palgreave Macmillan, 2003). Chapter 5 is a much-revised version of
“Carousing and Masculinity: The Cultural Production of Gender in Tai-
wan,” in Women in the New Taiwan: Gender Roles and Gender Consciousness
in a Changing Society, ed. Catherine Farris et al. (M. E. Sharpe, 2004).
Finally, I wish to thank my family for their faith, patience, and
unwavering support, and especially my father, Ben, who worked
through every chapter with me and saw to it that I pushed through to
the end.
Note on Translation and
Use of Foreign Terms
Fieldwork involved the use of several languages and dialects, includ-

ing Standard Chinese, Southwestern Mandarin, Taiwanese Holo, and
Bai. For Standard Chinese, I use the pinyin system of romanization,
except for persons and places whose customary (non-pinyin) spelling
is already familiar to most readers (for example, “Taipei” and “Chiang
Kai-shek”).
For Holo (the preferred name in Taiwan for the language also
referred to as Hoklo, Taiwanese, Hokkien, Min’nan, Southern Min,
and Amoy dialect), I use the peh-oe-ji system, following the transcription
practice of the Amoy-English Dictionary (Maryknoll Language Service
Center, Taichung, 1976), with the following modifications: I do not
include tone marks or diacritics, and nasalized vowels are indicated
with an asterisk (*) rather than a superscripted “n.” The few Canton-
ese terms are rendered according to the Putonghua Yueyin Zhonghua
Xin Zidian (1982 revised edition).
With the exception of public figures and a few informants who
specifically requested that I use their real names, the names of persons
that appear in this book are all pseudonyms. As far as possible, I have
tried to convey the sense of the original names, to give the reader a
feel for each individual personality, as well as to suggest the relative
formality or familiarity of different situations. All place names are real,
though a few follow local rather than official usage and may be diffi-
cult to locate on most maps.
All translations and transcriptions are by the author unless other-
wise noted.
ix
CHAPTER 1
Introduction
I am a student of the Chinese martial arts. Beyond practice, however,

I have always been fascinated by martial arts culture and history. I first
followed that fascination by reading every book and article I could
track down on the martial arts. I borrowed an old copy of the out-of-
print Secrets of Shaolin Temple Boxing by Robert W. Smith from my col-
lege library and renewed it at least five times. I faithfully subscribed to
enthusiast magazines like Black Belt and Inside Kung Fu.
As my engagement with the practical training deepened, I began
to sense that the very kinetics of the martial arts—the postures, move-
ments, vocalizations, and facial expressions—embodied a sensibility
that was very much “other.” That otherness, naturally, should have
something to do with the origins and long history of the martial arts
in Asia. When I began training in the mid-1970s, most martial arts
schools in North America identified themselves as belonging to one or
another East Asian (Chinese, Japanese, Korean, Okinawan) tradition.
Beginners were required to learn a set of conventional ritual gestures,
expressions, and vocalizations, memorize a few anecdotes about the
founding and history of the style, a bit of “philosophy,” and some ver-
sion of the myth (presented as historical fact) of Bodhidharma and the
Shaolin Temple. Fascinating as all this was, it seemed disjointed and
incomplete; it didn’t explain much about either the practice or the
history of the practice.
Popular martial arts books and magazines did little to fill in the
gaps. This literature was mostly produced by the same practitioners
who consumed it, and it mostly restated and reinforced the existing
preconceptions and prejudices of the American martial arts subcul-
ture. Perhaps even more disappointing, what few references to the
martial arts I could find in works of serious sinology displayed pitifully
little real knowledge or understanding of the martial arts and tended
to minimize their cultural and historical significance (cf. Henning
1999). I thought, and hoped, that someone could do better.1
1
2 gods, ghosts, and gangsters
The present work is partly a response to that challenge. Yet this is

not a book about the Chinese fighting arts themselves. I have left for
other scholars many of the topics that one might hope or expect to find
in a book about the nexus between religion and the Chinese martial
arts. I do not, for example, discuss the important (though occasionally
exaggerated) role of Buddhist monks or Daoist adepts in the develop-
ment and promulgation of martial arts techniques and lore.2 I do not
trace the origins of Chinese combat techniques, nor do I explore the
traditional Chinese (or contemporary American) relationship between
martial arts discipline and spiritual cultivation. The study of the martial
arts in themselves and their co-evolution with other aspects of Chinese
culture is an enormous (and much-neglected) field of research. No
single book, and probably no single scholar can hope to cover it all.3
Yet I hope this book will be received, at least in part, as a contri-
bution to the still nascent field of martial arts scholarship.4 The mar-
tial arts are present here in every chapter, from representations of the
Chinese knight-errant (wuxia) tradition to the ritual and iconographic
styles of Chinese Daoism and popular religion; the narrative and dra-
maturgical conventions of classical vernacular literature and perfor-
mance; and the gestures, expressions, and speech of modern Chinese
and Taiwanese jianghu (underworld) society (lit. “rivers and lakes”; for
an explanation of the term see chapter 2). Interviews with martial arts
masters and years of interaction with martial arts teachers and students
have been critical to the framing of the topic.
Without reliable research and informed commentary on the mar-
tial arts, our knowledge of Chinese society and culture in general is
uneven and incomplete. At the same time—here I am speaking to my
fellow martial arts enthusiasts in particular—our understanding of the
martial arts will remain inadequate until we recognize the ways that
they have evolved in China within a unique cultural frame, persistently
influenced and enriched by (and in turn influencing and enriching) a
variety of practices, such as folk religion and narrative traditions, that
are essential components of what we generally think of as traditional
Chinese culture. Without better knowledge of the cultural context and
historical process, we are left with a partial, even distorted, understand-
ing of our own practice.
A Felicitous Conjunction
I had already been studying Chinese and martial arts for some time
when I first went to live in Taiwan in the early 1980s. One hazy spring
introduction 3
afternoon, as I was making my way home along Xinyi Road (one of

Taipei’s major east-west thoroughfares), I was startled by the nearby
clack and clang of a traditional Chinese drum and gong troupe (guluo-
zhen) . Down the block, near a major intersection, a crowd was forming
as a steady stream of figures spilled onto the sidewalk from a nearby
alley. During my approach, the stream of figures slowly resolved into
a two-column marching formation of perhaps two dozen young men,
all wearing black pants and white T-shirts. Each held a long-handled
weapon, pointed skyward in the shoulder-arms position. Behind
them, an eight-man team carried a canopy-draped carved wood sedan
chair, jouncing it vigorously as they marched. Following the chair was
a loosely packed group of about fifty men and women who, like the
palanquin bearers, were wearing identical baby blue track suits and
matching baseball caps. Finally, bringing up the rear came the drum
and gong band, perched on the back of a small blue flatbed truck.
The marchers halted near the intersection. Three men then ran
ahead to the corner and set off a long, braided string of firecrackers.
Two others waded into the intersection, raising white-gloved hands
and tweeting police whistles, momentarily keeping the backed-up line
of honking taxis and utility trucks at bay. The parade crossed the ave-
nue, heading south, and a wave of traffic immediately surged into the
breach. I decided to follow, losing some time navigating the pedestrian
underpass. Once across the busy avenue, however, I quickly picked
up the sound trail. I finally caught up with them halfway down a side
street, where they had stopped outside a small tailor shop across from
a school playground.
A small folding card table had been set up by the door, loaded with
unopened bundles of incense, fruit, plates of cooked fish and pork,
and a tray of prepared betel nuts (binlang, pin-ng) . As the drums and
gong reached a crescendo, another string of firecrackers blasted the
pavement. On signal, the troupe tightened up its formation, continu-
ing to march in step. The weapon bearers then proceeded, individually
and in pairs, to move through the paces of mock attack and defense
drills. The entire exercise lasted about fifteen minutes.
Although this was my first encounter with such a temple procession,
much of what I saw seemed familiar. The implements carried by the
performers appeared to be classical Chinese martial arts weapons, while
their choreographed drills clearly belonged to some style of southern
Shaolin gongfu (kung fu). The mock combat drills, as well as the pound-
ing drums and salvos of firecrackers, immediately brought to mind the
lion dances and street performances of martial arts I had watched dur-
ing Chinese Lunar New Year celebrations in North America.
On closer inspection, however, I began to notice some curious
differences. What had appeared from a distance to be steel halberds
( ji), spears (qiang) , and broad-bladed halberds (guandao) turned out
instead to be oddly carved and painted wooden replicas. Mounted on
the handles, in place of edged blades and pointed spikes, was a strange
collection of fanciful, intricate shapes; a giant hand gripping a writ-
ing brush, a wreath of intersecting gold and silver rings, and the Chi-
nese characters for “sun” and “moon” embossed on filigreed gold discs
were among the least weapon-like objects. Others were reminiscent of
real weapons: a golden lance, the stars of the Big Dipper embossed in
black across the face; a red-and-gold-painted dragon-head mace. Even
hard steel versions of such odd implements would be of dubious value
on the battlefield. All the same, these “soldiers” carried and flourished
them just as they would handle practical weapons. 5
The quality of the performance also differed from my expecta-
tions: many of the actors, who appeared to be mostly in their teens and
twenties, moved through their assigned routines hesitantly and with
little obvious energy. I had seen junior novices at their first karate tour-
nament with more impressive moves. But then things started to look
up: a stockily built man, clearly older than the others, stepped into
the circle. Onlookers who had been milling around and chatting now
stopped and turned to watch as the man performed a long, complex,
and athletic routine with dual spike-handled battle axes, one of the
most difficult weapons sets in the martial arts repertoire. He moved
effortlessly, his form flawless, his stances balanced, his strikes and
blocks fluid, precise, and powerful. From start to finish he radiated
an intensity and fierce confidence that marked him as an experienced
fighter and martial arts adept. The contrast with the rest of the troupe
was striking and, to me, reassuring.6
Another string of firecrackers crackled, and the procession moved
on. I tagged along as it wound through the maze of streets and alleys,
pausing at intersections and storefronts; at each stop, the “soldiers”
ran through the same routine: a short drill initiated, punctuated,
and concluded with exploding firecrackers. After watching perhaps a
dozen repetitions, I left, fascinated but unsure exactly what to make of
this juxtaposition of religious devotion, symbolic violence, and retail
business.7 I wondered, too, who the performers were and what their
lives were about.
introduction 5
Martial Arts and Martial Cults

Noisy, colorful, chaotically intense ritual processions have been a high-
light of Chinese community life for centuries.8 Yet throughout much
of this history, popular religion, including its sometimes carnivalesque
festivals and processions, has been persistently derided in official writ-
ings as wasteful, superstitious, and disruptive to the social and moral
order. The documentary record is therefore sparse and uneven. True,
most late imperial local gazetteers ( fangzhi) , which are generally rich
in empirical detail, include a chapter on “local customs” ( fengsu) . Mar-
riage and funeral practices, for example, get their due in these texts.
But when it comes to the religion of the common people, the por-
trayals turn vague, the descriptions formulaic. Practices that offend
Confucian sensibilities—including much of the popular ritual reper-
toire—are either ignored or vilified as “heterodox” (xie) and “licen-
tious” (yin) .9
Given the tremendous modern popularity of the Chinese martial
arts, the scarcity of pre-twentieth-century sources also begs explana-
tion. The written record, as noted in the case of popular religion, is
largely an artifact of the tastes, habits, and values of the literate elite.
The literati, especially from the Song dynasty (960 –1279 CE) on,
tended to derogate all that is military and physical (wu) as categorically
inferior to the civil and learned (wen). Military affairs of state had a
secure place in the official annals but were still treated by most literati
as less worthy of account than civil affairs and the pursuit of literary
cultivation (cf. Graff 2002:1–3).
Outside of official annals, however, there is plenty of indirect evi-
dence for the antiquity and durable popularity of the martial arts, even
among the educated elite. It is significant that much of this evidence
ties the martial arts closely to the many narrative genres of popular cul-
ture and to local ritual practice. Traditional Chinese opera and pup-
pet troupes, for example, which feature generals, warriors, and bandits
among their standard roles, are known for their flashy displays of mar-
tial-style acrobatics. Troupes often performed in village temple court-
yards, sponsored by the community to entertain the gods on important
festival occasions. In the late imperial era, episodes drawn from vernac-
ular novels like Journey to the West or the Creation of the Gods, which drew
heavily upon (and subsequently influenced) popular religious beliefs
and imagery, had become among the most popular pieces in these
troupes’ repertoires. Ming and Qing vernacular novels are also among
the rare sources for descriptions of premodern martial arts practice.

Indeed, the development of martial ritual performance styles may
have been derived more from the theatrical versions of fighting arts
than their original, practical forms.10 Furthermore, the popular narra-
tive traditions reflect an understanding of the spirit world that clearly
marks the martial as the embodiment of supernatural power, a view
already fully developed in the religious beliefs of the Han dynasty.11
The written record, while it affords occasional clues to the under-
standing of the historical convergence of popular religion and martial
arts, leaves many basic questions unanswered; on the lives of individual
practitioners, history is virtually silent. But in the streets, temples, and
markets of Taiwan, I found a living and indeed thriving practice. And by
the mid-1980s, some of those traditions had begun to reemerge as well
across the Taiwan Strait in post-Mao China. The documentary record
might be sparse, but the living tradition was clearly rich, observable,
and accessible. I could talk to the practitioners themselves, observe
their routines, and solicit their explanations. This simple empirical
objective led me to anthropology, to fieldwork in both China and Tai-
wan, and (inexorably or not) to the writing of this book.
Changing Fieldscapes
The ethnography that comprises the empirical core of this book is
comparative and multisited. I conducted extended fieldwork in two
locales: Taidong, Taiwan, and Dali, Yunnan. Both are communities of
considerable ethnic and cultural diversity. Taidong is a county town
on Taiwan’s southeast Pacific coast, separated from the main popu-
lation centers by the spine of the Central Mountains. Dali sits in a
broad, high valley along Erhai Lake at the foot of the Cangshan range
in western Yunnan. Both locales are frequently described (by outsid-
ers and locals alike) as remote and home to large numbers of “ethnic
minority” peoples. Yet at the same time, both are intricately linked to
near and distant places through long-standing networks of sojourners
and outmigrants. In recent decades, Taidong and Dali have, in paral-
lel, undergone striking and sometimes jarring transformations, both
aspiring to become world-class tourist havens. In all, this ethnography
draws on nearly eight years in the field between 1988 and 2005. During
that period I visited China and Taiwan numerous times, each visit last-
ing between one month and two years. On a few occasions, I was able
to combine extended stays in Taiwan with short visits to Yunnan.
While Dali was already a well-known tourist destination, the vil-
introduction 7
lages in which I worked offered few amenities and were rarely visited
by outsiders. Most villagers derived little direct benefit from the tour-
ist industry, although the general level of prosperity of their agricul-
tural and household-based manufacturing enterprises was very much
entwined in the global economy. Nevertheless, local identities were
clearly correlated with easily visible territorial boundaries, and the
rhythms of daily life were still very much centered on neighbors, kin,
the local temple, the surrounding fields, and the periodic markets.
In Taiwan, on the other hand, extensive urbanization, a rapid rise
in general prosperity, and rural-urban migration since the 1960s have
transformed the island into a highly mobile society occupying an over-
whelmingly urbanized landscape. There are, however, a few spots in
Taiwan where the economic miracle was postponed by a decade or
two. The small coastal city of Taidong was still a quiet backwater when I
began doing fieldwork there in 1988. But even here, most everyone was
tied into social, business, kinship, and ritual networks that stretched
around and across the island.
Shifting Focus
I started fieldwork in Taidong with a survey of the local temples, visit-
ing those in particular whose principal deity could be considered mar-
tial in character. To my surprise, many of those I met during this stage
had little interest in discussing martial deities, ritual exorcisms, and
spirit soldiers with me. If they did talk about such things, it was often to
try to deflect and neutralize any suspicion that their beliefs might have
something to do with heterodox spirits and disreputable practices.
I was a scholar, they reminded me, so I should just focus on history
and discover for myself the authentic, orthodox tradition. Rather than
pressing the issue, I spent my time visiting temples, attending nightly
séances, and observing religious festivals, trying to learn as much as
I could about martial deities and the myths and rituals attached to
them. In order to make sense of what my informants were saying and
doing and to earn their trust, I had to learn more about the rituals,
symbols, and logics of their everyday lifeworld.
Mastering the details and meanings of ritual practice soon became
as much a means to an end as an end in itself: during evenings spent
over tea and conversation in the halls and courtyards of local temples,
I found myself increasingly drawn to my informants’ enthusiasm for
the lore of deities, miraculous power, and ritual efficacy. The logic and
details of ritual magic; stories of miraculous events; and the invisible
world of gods, demon generals, and wayward ghosts were our favored
medium of exchange, but much more was conveyed in the process.
Here I began to understand individual personalities and learn about
their backgrounds and the ups and downs of friendships and rivalries
both within and beyond each group. As I was granted access to some
of their lives, I began to focus increasingly on observing, documenting,
and participating in conversations, actions, work, and a wide range of
social and business-related activities both in and outside temples.
My new purpose, which became the guiding intention of this book,
was to understand who these men were and what role these severe,
sometimes violent ritual performances played in their lives. Dramatic
occasions, such as rituals and festivals, do not afford much opportunity
to learn the mundane details of peoples’ everyday lives. But as spaces
of emotional intensity, festivals and ritual performances often reveal
otherwise concealed aspects of personality. To understand what devo-
tees seek in the miraculous requires some sense of what their ordinary
lives are like, what they desire in material and emotional terms.
Gods, Ghosts, and Gangsters, then, is about ritual violence and vio-
lent rituals in Chinese and Taiwanese popular religion. It is about the
martial deity temples and shrine-altars that serve as physical, spatial
extensions of shared cultural myths, local histories, and narratives
of personal identity. It is about the lived experience of individuals,
mostly working-class men who are often members of sworn brother-
hoods and local gangs, who perform these rituals during temple fes-
tivals and processions. Finally and most centrally, it is an account of
the ways that these men think of, speak of, and perform themselves as
men. It is an ethnographic study intended to contribute to ongoing
discussions in the study of Chinese ritual and religion, the history and
sociology of the Chinese underworld, and the anthropological study
of masculinity.
Violence and Religion

The term “religious violence” has an immediate and tragic resonance
in the world today. In China, religiously inspired, directed, and/or
justified collective violence has historically been associated with sectar-
ian religions based on revealed prophecies, often of a coming apoca-
lypse. Since their initial appearance at the end of the Han dynasty (206
BCE–220 CE), millenarian sects have occasionally incited rebellion; a
handful have precipitated widespread catastrophe. The Han-era Dao-
ist Yellow Turban uprising, the Buddhist White Lotus of the Ming and
introduction 9
Qing, and the mid-Qing Christian-inspired Taiping rebels are all famil-
iar to even the most casual student of Chinese history.12
Unlike devotion-based Buddhist or sectarian religious communi-
ties, popular religion as I treat it here is characteristically territorial,
deeply rooted in the particularistic and the local. Embedded in and
reflecting mundane daily life, the basic aims of domestic and commu-
nity-centered ritual are to tap the responsive power (lingyan) of gods
and ancestors in order to induce fertility (the reproduction of family
and community and material increase and abundance) and to ensure
a state of security and social harmony (ping’an) (cf. Sangren 2000:170).
As I will explain below, however, symbols of coercive force and destruc-
tive power are widely deployed in the pursuit of local order.
Each incense burner found in every temple, lineage, guild, and
association hall, as well as innumerable household ancestor shrines,
is the material center of an exclusive, semi-autonomous ritual unit.
Yet every such unit is also embedded in a nested hierarchy of local
and translocal ritual and social networks. A market town temple and
the temple cults of nearby satellite villages, for example, constitute
an obvious small-scale hierarchy, with the central place usually in the
dominant position. Nevertheless, the relationship between any two
given nodes in the network depends on a cycle of material and sym-
bolic exchange that is simultaneously symmetrical and reciprocal (cf.
Bateson 1972:61–72)—that is, the outcome of any particular exchange
is theoretically uncertain. Both groups have a stake in alliance, on the
one hand, and in achieving dominance on the other. The climax of
each such cycle occurs during the periodic temple festivals (yingshen
saihui), in which the martial procession troupes of various temples
play a prominent role. Here groups greet and test each other in open
competition. The prospect for conflict simmers constantly; occasion-
ally there are fights. Even when individuals or groups come to blows
and the violence is fiercely real, however, the event is almost always
recalled as an epic battle between supernatural forces rather than a
mundane rumble.
The Ritual Logic of Domination

The most basic use of violence in religious context is to offer blood,
either one’s own or that of another. All Chinese ritual is arguably pre-
mised on the logic of sacrifice, the offering of a prestation to secure
the possibility of a response, reciprocate a favor done, or fulfill an
ascriptive obligation.13 The main constituents of Chinese popular reli-
gion—blood sacrifice, territoriality, the imperial metaphor, a sequence

of periodic festivals and community sacrifices, the efficacy of numinous
spirit (ling), and the continuum of cosmic material energy (qi) — com-
prise a consistent and broadly intelligible symbolic system. Anchoring
this system is a logic of purity and pollution that gives priority, over all
other functions, to defense against and extirpation of invasive spirits
(the malevolent residues of death).
Threatening substances and beings are generally classed as yin —
that is, shadowy, insubstantial, female. Subduing such threats requires
exorcistic power, which works, in turn, through objects and beings
that are relatively more yang (bright, substantial, male). Fire, in par-
ticular, is a powerful agent of purification, both because it epitomizes
yang and because it effects irreversible transformation from visible to
invisible, substantial to insubstantial. But the most unusual aspect of
Chinese exorcism is its distinctive use of the martial idiom. I posit that
the martial, in turn, is a categorically male domain, and martial prow-
ess, the mastery of innate male aggressivity (which implies, even as it
represses and channels, a vigorous sexual potency) is then a sign of
dominant masculinity. Thus, in the simplest terms, for men with few
prospects for conventional social stature and economic stability, vio-
lence, including the performance of martial ritual, becomes a viable
medium for self-production.
This scenario can be reframed in the terms of ritual logic as fol-
lows: (1) ghosts and the remnants of death are extremes of yin, which,
it turns out, is also the female principle; (2) the power to manage and
destroy malevolent ghosts, then, should logically be male (yang) rela-
tive to its object; (3) the relation of yin to yang is asymmetric, with yang
dominant—this is asserted both in ontological and concrete social
terms, such that the cosmic structure is understood as validating the
gendered social hierarchy; (4) the coercive force of physical restraint
and destruction needed to overcome malevolent threats, human and
spiritual, is most directly embodied by the (paradigmatically male)
collective power of military organization but also by the more auton-
omous (and therefore hierarchy-threatening) force of individual
martial prowess; (5) subordinate minions are yin (hence feminized)
relative to their superiors but bridle under the yoke of subordination
and are prone to capricious acts of violence (a characteristic of young
males); (6) those (like high-level military gods or clan patriarchs) who
command the labor and the collective violent potential of well-trained,
introduction 11
inherently fierce and dangerous minions are structurally dominant,

but their power and position are precarious.
The ambiguity suggested in point 6 above embodies a contradic-
tion both vexing and essential to the reproduction of the Chinese
patrilineal family. The father’s role, that is, requires the subordination
of his sons, a role he himself endured and has finally transcended.
Sons, meanwhile, desire autonomy but can achieve it only either by
the unfilial act of breaking away or by biding their time. The intact
extended household represents the continued charisma and potency
of the patriarch; his decline and demise are prefigured in the transfer
of responsibilities and property to sons. This inevitability ensures that
at least some sons will endure their subjugation and aspire to the role
of patriarch—conceiving as their own project the reproduction of the
means of their own (temporary) subjection.
Ritual, Violence, and Masculinity

Fierce-countenanced demon puppets, ritual specialists with their
cobra-headed whips, sword-wielding Daoist priests, and trancing spirit
mediums lacerating their bodies with spikes and blades are among the
most dramatic and distinctive features of Chinese popular religion.
Among practitioners and even in the broader community, it seems nat-
ural and intuitive to associate these baleful icons with the local under-
world of drifters, brotherhoods, and gangs.14 The presence of such
shadowy denizens of the social margin enhances the aura of menace
and mystery for believers and casual observers alike. Popular descrip-
tions of ghosts often describe them as gangsters of the spirit world. At
the same time, the association is not unambiguously pejorative. Rather,
malicious spirits and those who embody them in ritual performance
have long been models for each other in the collective imagination,
and both, perhaps in part because of this very association, continue to
generate a broad, if morbid, fascination.
On the one hand, the exorcistic military metaphor can be seen
as a form of symbolic violence that serves to reproduce hierarchical
structures, coerce submission, and assure the continued dominance
of those in superior positions. The power of violent aggression is actu-
ally generated by the minions laboring in the trenches, but it is imme-
diately claimed as an intrinsic property of the superior. The logic of
martial ritual would then seem to be a classic version of the Marxian
process of alienation and fetishization, though what is produced in
the process is a form of symbolic, not tangible, capital—the patriarch’s

aura of power.
Let us assume, without further argument for the moment, that the
military metaphor does in this way symbolically reinforce and validate
patriarchal domination (of women and younger men). Why, then, are
those most strongly attracted to the practice of martial ritual precisely
those ritually processed into subjugation? They are recruited because
their blood and pain are required by the logic of ritual efficacy. But
what entices them, and what is their reward? The following account
evokes a few key facts with which we can begin to answer that decep-
tively simple question:
[They] burned incense, swore mutual allegiance and became

secret society members. . . . They called each other “elder brother,”
“younger brother,” following the example of the Gelaohui. . . .15
Their basic operating principle was to act wildly and without
regard to reason. Every imperious youth, local strongman and
evil gentry [soon] followed their example, and the district ended
up divided into eight sections, [each controlled by] one of eight
criminal brotherhoods. Each brotherhood had a big boss [da
ye]. . . . They were all despotic and arrogant, but Wan Zhong, head
of the Banqiao gang, went to extremes. Whenever he went out,
he imitated the Governor-General, sitting in a palanquin carried
by eight bearers—what’s more, for this purpose he would com-
mandeer the Guangzun Temple’s gold-throned palanquin, the
one used to carry the gods. Sitting on a tiger skin, his feet resting
on a golden lion, he truly resembled a heavenly deity! Five or six
hundred men, all armed, comprised his retinue. At the head of
the column marched four men bearing censers, exactly like the
rituals for welcoming gods. (Li Yüanbing 1953:3)
Why would “big boss” Wan Zhong choose to represent himself as

a god? Why this particular kind of god? We know very little about Wan
Zhong beyond what can be inferred from this caricature. He was active
in the early 1840s around Banqiao, a market town in the Baoshan
Basin, near Burma in western Yunnan Province; he was head of a local
armed gang organized, like most Chinese underworld associations, as
a secret brotherhood; and he clearly had a flair for the dramatic.16 For
the chronicler who penned this description, Wan Zhong was a stereo-
typical villain. His “despotic and arrogant” ways are amplified in this
introduction 13
impersonation of a military god (only military gods would be carried

about, like Wan Zhong, in an open sedan chair, sitting on a tiger skin,
foot resting on a golden lion, preceded by incense-bearing lictors and
accompanied by a menacing entourage of armed henchmen).
Wan Zhong’s ostentatious display obviously played well to his
intended audience. What, then, made his performance meaningful
and, indeed, effective? The answer seems to embody a paradox, but
it describes what I have come to see as a linkage intrinsic—critical, in
fact—to male identity in modern Chinese society. In the first place,
military deities of the rank that Wan Zhong was imitating are generally
titled “general” ( jiangjun) or “generalissimo” (yuanshuai) . Wan Zhong
was thus claiming to be not merely a bandit chief or boss of a market
town protection racket; riding the palanquin authenticated his role of
commander, with all its implied legitimacy and formal authority.
A powerful man is one who, through possession of an innate
potency, is able to summon and command the labor of others. Such
power is often expressed in terms of territory or “turf,” exemplified
in the above account by Wan Zhong and the other seven (unnamed)
gang bosses. As the anecdote suggests, gang bosses and local strong-
men are “like gods”: they emanate an aura of potency premised on
the power to project or withhold violence rather than the sanction of
bureaucratic authority. Vainglorious Wan Zhong would seem at first
glance to be the antithesis of conventional Chinese manhood, nothing
like the filial sons and loyal subjects who are “content with their lot and
act accordingly” (anfen shouji). Yet in this account, these men who “act
wildly and without regard to reason” have no greater ambition than to
be patriarchs.
It turns out, then, that such claims to power and aggressive rep-
resentations of dominance do not, in fact, violate the mainstream
conventions of Chinese manhood— quite the contrary. They reveal a
social and cultural logic, shared by erudite metropolitan scholars as
much as illiterate frontier gang bosses, in which a dominating mascu-
linity is generated through the hierarchical structuring of potency and
charisma. Simply being in a position of dominance (as in the case of a
family patriarch) automatically affords a certain measure of that cha-
risma, but the greater portion is taken to be a form of intrinsic potency
(often described in Chinese in terms of qi ) manifested as extraordi-
nary personal attributes.
But what of those men on the other side of this relation? Not every
son, nor certainly every junior gang member, can hope to someday
replace the father/boss. Nor is either likely to be satisfied with per-

manent subordination—manhood is bound to the scale of masculine
domination, which may help to explain the tendency of Chinese sons
to divide their patrimony at the earliest possible moment and the well-
documented desire of nearly all Chinese men to “be a boss” (ziji dang
laoban). For men with little chance (or, for some, little desire) to strive
within the sanctioned social and vocational milieu, the best, or only,
option is to set out for the “rivers and lakes” or join the brothers of the
“dark path” (heidao). For a particular subset of these men, martial tem-
ple cults and the practice of martial ritual offer a related attraction:
immediate and direct access to numinous power, by which they can
appropriate, in the frame of public ritual performance, the charisma
and potency otherwise denied.
The Aesthetics of Martial Ritual Performance

We can surmise that Wan Zhong played the part of a stern, magisterial
martial god reasonably well. To effectively personify a god or an official
on a tour of inspection, Wan Zhong would have had to spend most of
his time holding a pose, much like the one described. The two- and
three-dimensional representations of gods and officials familiar to his
audience—the carved images of gods on temple altars, temple paint-
ings and friezes depicting mythic battles, woodblock prints of gods
and immortals—were all static. Ritual constrains the body, prescribing
certain behaviors—postures, gestures, and recursive sequences of pos-
tures and gestures—and proscribing others.
Martial temple processions are a species of dramatic performance,
the aesthetic power of which is inseparable from its perceived ritual
efficacy. Certain intrinsic qualities of martial ritual (the frightening
images of demonic soldiers, for example) automatically generate an
aura of theatricality, a mood equally morbid and festive, constrained
and unleashed. So with the form of the ritual itself: along the proces-
sion route, troupes march in tightly disciplined formation, repeating
their set routines at designated, ritually significant stations. Speaking,
eating, drinking, and smoking, are proscribed; the troupes march
through aches and discomfort, subjected to hours without food or
rest. Then, suddenly, one of the actors begins shaking, foaming at the
mouth, belching, gripped in the throes of a violent possession trance.
These trance episodes are manifestations of miraculous power—
that is, of powerful spirits appearing and intervening in the visible
world. Some play out as intricate narratives in which we can clearly dis-
introduction 15
cern the creative agency of the actor. Others are purely kinetic, wild,
uncontrollable fits of flailing, swinging, and kicking. There is a pur-
poseful aesthetic in play here, too, however: such regressive outbursts
relive the primal frustration of desire and violently assert, in the idiom
of a childhood tantrum, an autonomous male self that transcends
the constraints of assigned social roles—be they demanding father or
obedient son.17 As in Wan Zhong’s performance, the actors oscillate
between two extremes of martial masculinity, austere and restrained
at one end, feral, capricious, and uncontrollable at the other. What
we see, then, is the ritual objectification of the defining emotional
struggle of Chinese male subjectivity, rooted in the conflicted rela-
tions between fathers and sons in the patrilineal family (cf. Sangren
2000:186–223).
This ambiguity is mirrored in the poetic structure of the ritual pro-
cession itself, which oscillates between the constraint of regulated rou-
tines and the spontaneous freedom of possession trance. A successful
ritual outcome requires an authentic enactment. That authenticity, in
turn, is contingent upon proper execution of the ritual prescription
(in Chinese popular religion often uncodified and contended) and
also upon the aesthetic and emotional components of performance.
Of Morals and Margins

The assumed association of secret societies, heterodox religious cults,
and outlaw violence is hardly limited to nineteenth-century Yun-
nan. Rather, it is part and parcel of a stock portrayal, replayed in the
accounts of British colonial officials, martial arts adventure novels, and
innumerable Hong Kong movies; it is also, to some extent, corrobo-
rated by historical and ethnographic evidence. While only a minority
of the men active in martial cults would qualify as true liumang (gang-
sters), they all live in a tough and sometimes violent social milieu, a
world imaginatively characterized in the popular imagination as the
“rivers and lakes.” The romanticized and sensationalized jianghu por-
trayed in novels, television serials, and the newspaper crime pages, in
turn, bears only the faintest resemblance to the facts of daily life. Yet
the real-life denizens of the jianghu are themselves avid consumers of
such media images. They not only partake, however, but also create:
social performances, as well as ritual acting, are implicated in larger
processes of the cultural production not only of shared cultural forms,
but also of identities, selves, and persons.
In other words, the jianghu, as a collective fantasy of the violent
but honorable social other, serves as a symbolic field through which

the ordinary foot soldiers of the dark path create and represent them-
selves, to themselves and to others. The demons and deities that appear
during martial rituals and possession trance performances are arche-
typal jianghu figures, simplified and abstracted collective representa-
tions from which all complicated personal details and idiosyncrasies
of individual character have been stripped away. Publicly embodying
these archetypes, ritual actors turn streets and temple courtyards into
mythic battlegrounds. As I argue below, the actors themselves are living
out fantasies of supernatural power, of knights-errant heroics, as well
as cruel violence. From the perspective of individual performers, then,
ritual enactment is an emotional process (in the technical sense), a
means of self-production and self-presentation. It is not my purpose
to discern the details of that process for individual subjects; emotional
pathways are hidden, of course, and it is beyond my power to untangle
the subtle psychodynamics of others’ experience. I do hypothesize,
however, about the shared narratives and forms of desire visible in the
content of performance itself or that informants suggest in their own
comments and conversations.
Finally, while the jianghu is partly collective fantasy, it also describes
a real social space. It is the marginal world of drifters, outlaws, con art-
ists, thieves, bodyguards, loan sharks and debt collectors, vagabonds,
gamblers, prostitutes, and anyone living by the sword or by wits alone. It
is a highly fluid and competitive hierarchy, a fractious fraternity whose
members indeed address each other as “elder brother” and “younger
brother.” In stark contrast to the romantic fantasy of the outlaw life,
brothers (xiongdi; H. hia*-ti) of the jianghu revealingly describe their
lives as a combination of “play” (wan; H. chhit-tho), and “just passing the
time” (hun). All of this playing and passing of time, however, entails,
besides existential boredom, some kind of productive activity. The
“brothers” are, indeed, a real social presence and integral to the oper-
ation of local economies throughout the Chinese-speaking world.18
Herein lies another contradiction: though they live on the margins,
sometimes violating mainstream norms and making a business out of
human weakness, the brothers of the jianghu have nevertheless become
a vital constituent of modern Chinese culture and society.
In the ritual logic of the imperial state (as well as the guiding ide-
ology of the modern education system), the destructive power of the
martial is subordinated to, and neutralized by, the creative power of
culture (as the written word, pattern, and discourse). Popular religion,
introduction 17
however, inverts this relation, positing martial power as the originating

force and recognizing its necessity in the maintenance of the territorial
and social integrity of the collectivity, from the household and village
to the state itself. Despite the obvious importance of the martial genre,
the historical record is almost silent on its provenance, distribution, or
details of performance. We also know very little about those who sus-
tained its practice, beyond miscellaneous accounts of bandits and reb-
els. That bias has unfortunately been carried over to the modern era.
Today, virtually the only way martial cults and the most violent forms
of martial ritual can find mainstream acceptance is to be identified as
unique, traditional “folk customs.” Yet as in the past, the lives of the
practitioners (ritual actors for the martial troupes are still recruited
mostly from the social margins) are rarely documented in newspaper
articles and folklore books, their voices rarely heard even as their per-
formances are videotaped, broadcast, and shared on YouTube.
Outline of Chapters
The visual symbols, narrative depictions, and ritual performances of
popular religion make tangibly present a myth-world originating in an
indeterminate “ancient” imperial past. Among the most striking and
powerful of these images and actions are those that evoke the violent
and shadowy elements of this myth-world. The martial is ubiquitous in
the iconography and architecture of Chinese temples; in the intona-
tions, gestures, and implements of rituals of purification and exorcism
(which are, in part or full, requisite to nearly every religious occasion,
private and public, minor and major); and in the endless recreations
of images and stories, collective myth rendered in the narrative idiom
of “history.” Yet this collection of fantastic tales, arcane language, and
archaic styles is anything but an inert “tradition” or cultural survival;
it is an axial plane of a living and constantly evolving Chinese cultural
imaginary.
To account for the meaning and power of these representations,
I begin chapter 2 with a hypothetical encounter, a simple act of obei-
sance in a typical temple of popular religion. That the martial is so
palpably present in even the simplest everyday rituals underscores
its central place in Chinese cultural logic. I then present some of the
main historical and literary prototypes that devotees and scholars alike
generally associate with the symbolic field of the martial and examine
the ways violence, aggressivity, physical prowess, and bodily disciplines
(including the traditional martial arts) have come to be associated with
a set of social roles, moral qualities, and modes of self-presentation that

are, in turn, identified as defining attributes of a specifically non-elite
form of Chinese masculinity. Finally, I proceed to the main theme of
this book, which I posit first as a question: what is actually happening
in the act of martial ritual performance? My answer, in brief, is that it is
a dialectical process of cultural production predicated on shared aes-
thetic and empathic conventions that articulates and affirms a shared
ontology of experience and joins scripted, collective expectations to
the creative agency and personal desires of individual actors.
To build my case for this proposition, I begin in chapter 3 by recon-
structing one geographically and historically localized version of this
shared ontology. Specifically, I introduce the first of my two main field
sites, the Taiwanese coastal city of Taidong (and environs), invoking
local (as well as outsiders’) narratives that identify salient attributes of
the landscape and local social history with communal experience and
collective character. I focus in particular on descriptions of remote-
ness, marginality, wildness, and violence featured in my informants’
narratives of place, community, and self and on the ways social perfor-
mance and ritual practice are shaped by and play out the ambivalences
and limitations but also what informants perceive to be the positive
aspects of the objective geophysical, economic, and social conditions
(including the vexed asymmetrical relations between the Han Chinese
and aborigine communities).
In chapter 4 I narrow the focus to two rituals, each unique to a
single locale (Taidong and Dali, my second main field site) and both
characterized by dramatic public displays of male bravado. Both the
ritual of Blasting Handan Ye (Taidong) and the Torch Festival (Dali)
feature the controlled but inherently and conspicuously dangerous
exposure of (male) bodies to fire and incendiary explosives (that is,
firecrackers). Participants risk injury and even death, ostensibly for
the sake of communal fertility (for Handan Ye mostly in the more fig-
urative sense of purely economic increase; in the case of the Torch
Festival, fertility in the more concrete sense of biological as well as
economic production), but also not incidentally in pursuit of individ-
ual affirmation as a particular kind of vigorous, aggressively masculine
man. I argue that the defiant, violent, even rebellious style of macho
aggressivity displayed by the risk takers is actually instrumental to the
community-affirming outcome of the ritual, first of all through the
implied link between aggressive masculinity and male sexual prowess
that is ritually shared across the group through the public “sacrifice” of
introduction 19
the actors. I also describe how both rituals have at times served as sites
of political conflict: both were suppressed by the authorities in the late
twentieth century, reemerging briefly as expressions of autonomous
local identity in the late 1980s, only to be quickly coopted by the vicis-
situdes of globalization and economic development (specifically tour-
ism) since the early 1990s.
Chapter 5 is the ethnographic pivot of this book, a study of the
jianghu and martial ritual performance troupes and the critical impor-
tance of their historical and cultural affinity to the broader efficacy
and meaning of Chinese popular religion. I focus first on the Military
Retainers (Jiajiang), a ritual form unique to Taiwan (though one spe-
cies of a genre once common to most, if not all of “cultural China”).
Starting in the 1980s, the Military Retainers became notorious in their
association with juvenile delinquency, blood spectacle, and other unsa-
vory elements comprising the “dark” (hei) side of popular religion (yet,
paradoxically, in the years of political ferment immediately following
the end of martial law in 1987, they also became emblematic of the
bright “local color” of native Taiwanese tradition). Closely linked to the
Military Retainers, and in recent years increasingly marketed as coor-
dinated performance packages, are groups of possessed, entranced
spirit mediums (tangki), who provide the most dramatic and (for many)
most unnerving ritual spectacle of all. In fits of real or feigned trance,
these mediums vigorously and skillfully strike their heads and backs
with edged and spiked weapons, slice their tongues with knives and ice
saws, and pierce their bodies with a variety of sharp-pointed metal rods
and needles. Chapter 5 features a series of ethnographic descriptions
focused on individual actors in the context of specific events. Here I
endeavor to present the perspective and reconstruct the lived experi-
ence of the performers themselves, these socially and often economi-
cally liminal men who are the dedicated mainstays of, the essential
labor pool for, and (as noted in chapter 3) the ritually indispensable
participants in martial ritual performance troupes.
Chapter 6 is, on the surface, a momentary departure from the
topic of ritual and religion. Rather than follow the thread of ritual per-
formance itself, in this chapter I follow my informants instead, moving
with them into another arena of performance that figures prominently
in their social and economic lives—namely, the nightlife pursuits of
drinking and male carousing. Drinking and carousing are seminal to
the construction of masculine identity in China and Taiwan; for the
denizens of the jianghu, it is an undertaking critical not only to their
identities as individuals and as men, but also to their very livelihoods.

The conventionalized practices (but also the spontaneous and some-
times violent exigencies) of the nightlife, exaggerated and intensified
by alcohol and charged with female sexuality, greatly amplify the per-
formative qualities of behavior and the kinesthetic, bodily nature of
social interaction.
Nightlife activities are thus a perfect venue to observe generic
expressions of male camaraderie, competition and contest, and sex-
uality. But because I am primarily interested in particular subjects—
informants who also happen to be ritual actors and self-described
brothers of the jianghu—these episodes of drinking and carousing can
(and often do) reveal aspects of personality, tendencies of self-perfor-
mance, and candid (or embellished) narratives of self-description that
augment, in important ways, our understanding of their lives and per-
spectives gleaned from other contexts. Again, some of these men, who
obviously have a special affinity for dramatic ritual performance, turn
out to have an especially clear grasp of the poetics of social perfor-
mance as well. Documenting this as well as other aspects of their lives
brings us closer to identifying those aesthetic and empathic conven-
tions, the nexus of individual desire and collective expectation, the
shared ontology of experience.
Finally, in chapter 7, I return to the initial question: what is actually
happening in the act of martial ritual performance? I conclude by pro-
posing that we can explain wu masculinity as a mode of cultural pro-
duction, premised on a conventional aesthetics, through which per-
formers make visible and tangible otherwise diffuse collective moods
and expectations. Through ritual and social performances, these men
present, as qualities intrinsic to their own bodies and biographies, emo-
tions and desires linked to the often vexed nature of being a man in
Chinese society. This vehicle of public performance empowers men to
be agents of self-production, creating themselves through an enactment
of the heroic, tragic, hypersexual, capriciously infantile, or aggressively
violent phases comprising the habitus of Chinese wu masculinity.
What follows, then, is a cultural study of martial ritual and ritual
violence in Chinese popular religion, focused on the men and women
whose lives are entwined with its practice. Before turning to the eth-
nography, however, I begin by locating the martial in the context of a
broader Chinese cultural ontology; defining the cultural, historical,
social, and economic premises of the martial mythos; and explicating
the logic of martial ritual practice.
CHAPTER 2
Violence, Honor, and Manhood
Walk into any shrine or temple just about anywhere in China and you
will be presented with a tableau that differs little from place to place.1
If a devotee of Chinese popular religion from the mountains of Yun-
nan happened to venture into a typical Taiwanese temple, he would
immediately recognize its purpose and, once he got past a few obvious
differences in architecture and decoration, would have little trouble
identifying its main features: the large incense burner outside the main
entrance; the dimly lit open hall within; the elevated altar along the
rearmost wall, on which stand, centered and symmetrically arranged,
a group of miniature carved wooden (or cast metal or porcelain or
even plastic) human-like images wearing the robes, hats, and crowns
and bearing the talismans and weapons of emperors and imperial offi-
cials—the gods, shen, of the Daoist and popular pantheon. While local
worshippers might realize the devotee was “not from around here” as
he knelt and performed a ketou, or “kowtow,” they would still be thor-
oughly familiar with the gesture’s purpose and meaning. His other acts
of devotion would probably elicit no special notice.2 No different from
his Taiwanese counterparts, he would find a bundle of incense sticks
and paper spirit money on a wall to the side of the main entrance.
He would make a contribution for the incense, take the bundle, and
separate the incense from the spirit money. He would then light the
incense from the nearby propane burner provided for the purpose
and would offer incense to the deities following a sequence that differs
only in detail from place to place: the path begins at the large tripod
placed just outside the main door. The first three incense sticks are
offered here, the Emperor of Heaven’s incense burner (tiangong lu) .
Moving back inside, a worshipper stops first at the main altar to pay
respect; make an offering; quietly announce his or her name, place,
and date of birth (according to the lunar calendar); and then perhaps
recite a petition or make a vow. The worshipper then moves to the
subsidiary altar, again offering one (for the Earth God) or three (for
higher-ranked gods) incense sticks at each station. Before returning
21
outside to burn spirit money in the “gold-burning furnace” ( jinlu), the

worshipper offers one last stick at the niche of the Tiger General, the
yin spirit who guards the base of the main altar.3 Finally, the visiting
worshipper would be sure to drop a small cash contribution into the
“merit box” (gongde xiang) on the way out.
In styles of architecture, decoration, and iconography, temples of
Chinese popular religion are as diverse as the communities they serve.
Yet wherever they may be found, such temples can be identified from
a certain set of shared features. Most, for example, are decorated with
paintings, carvings, and terracotta figures of dragons and other auspi-
cious beings from Chinese mythology. Painted panels, some in relief,
depict scenes from popular stories associated with the temple’s patron
deity or another popular deity, such as the capricious child god Nezha
riding his fire wheel and grasping a magic ring, or the hero Wu Song
subduing a tiger; elsewhere, a mounted warrior charges into battle or
grotesque, laughing demons carry the fierce, bearded, ghost-eating
god Zhongkui in his open sedan chair.
Inside the temple, the main altar stands centered against the rear
wall of the main hall. Arrayed across the altar are the figures of the
deities. Male and female, some are seated, majestically enthroned,
calm and impassive. Others stand in menacing postures, brandishing
swords, halberds, or other classical weapons, their faces (red, black, or
painted with geometric patterns) frozen in what may be expressions of
anguished grief or glaring fury; some show faces oddly disfigured and
distorted. Most of these figurines are set on wooden bases, some of
which feature the name and title of the deity.
In temples of gods who attend to the world of the dead, even
stranger images flank the altar. Some are puppet-like heads mounted
on long, pointed rods, their demonically grimacing faces painted
white, red, or bright green, their hair long and disheveled; they wear
military robes but have no bodies. Sitting on a low table in front of
the main array may be another set of figurines, unadorned, carved
heads mounted on wooden or metal spikes a few inches long set into
a small wooden rack. In most places it is a set of five, painted in the
colors (blue, white, red, black, and yellow) of the five directions (east,
west, south, north, and center); elsewhere there may be twenty-four or
thirty-six, roughly carved and lacking names or titles.
The calm, seated deities are usually the highest-ranking, includ-
ing the patron god of the temple at the center, flanked by lower-rank-
ing “colleagues” and subordinates and accompanied by “guest” deities
violence, honor, and manhood 23
(who may be placed at a lower level while the patron deity, presiding as
host, is raised on a platform). To satisfy the needs of protocol, higher-
ranked deities, such as the Emperor of Heaven, may be housed sepa-
rately, in a shrine on an upper floor.
The pantheon of Chinese popular religion, like that of religious
Daoism, with which it is closely linked, echoes the ranked hierarchy
of the old imperial bureaucracy. Temples are metaphorical palaces
or, more accurately, yamen, the multipurpose official buildings that
in imperial times served as residence and courtroom for the local
magistrate, contained offices for the clerks, quarters for guards, and
an armory for weapons and the implements of torture and punish-
ment used to interrogate suspects and punish convicts.4 All of these
functions, and most of these characters, can be found in an average
Chinese folk temple. In the larger temples, even the demonic ghost
catchers are represented with carved images; in smaller or poorer
places a crudely inscribed plaque may be the only object signifying the
presence and identity of the resident spirits.
Two further features of the old imperial bureaucracy give shape to
this folk religious iconography. First, deities are designated as belong-
ing to one of two functional categories (as was the imperial bureau-
cracy for most of its nearly two thousand years of existence)—namely,
the civil (wen) and the military (wu). Second, whether civil or military,
every ranked god has a retinue of underlings: the aforementioned
clerks, runners, attendants, guards, and retainers. Imperial officials
were assigned to posts (in theory, by the emperor himself) outside
their home areas. They served as the emperor’s proxies, stewards of
his authority. To maintain the requisite aura of authority, magistrates
were expected also to maintain an appropriate social distance. To that
end, they deputed the work of administration and enforcement to
their minions.
Imperial officials were appointed out of an empire-wide pool of
successful civil service examination candidates. They were the intel-
lectual as well as political elite of the empire. The clerks, runners, and
other functionaries who worked in the yamen, on the other hand, were
generally commoners recruited from the local population. Unlike the
city or county magistrates they served, they were part of the local com-
munity, usually with a network of personal and kin ties in the area.
They spoke the local dialect and were familiar with local ways. They
could, then, either serve or subvert the magistrate’s interests. Since
magistrates and their subordinates alike were expected to supple-
ment their stipends through gifts and favors, local government was
extremely susceptible to bribery, extortion, and worse. Whether or not
they are historically accurate, popular narratives play on the venality
of such gatekeeping functionaries as a common theme. In novels and
dramas of the time, clerks and runners are typically depicted as unsa-
vory, troublesome characters.5
The runners (yayi) often did inhabit the same social world as local
toughs, bandits, bodyguards, and beggars (Hansson 1996:48). These
foot soldiers, flunkies, and enforcers were ambiguous characters; their
predatory bullying was a headache for imperial officials, yet their brut-
ish and aggressive disposition was exactly what made them useful. Their
excesses were tolerated because they were indispensable tools of impe-
rial administration and justice. Runners took care of unpleasant tasks
like serving summonses and collecting taxes; together with constables
and guards, they handled most of the work associated with policing
and incarceration, from patrolling town streets to capturing criminals
and bandits and handling prisoners. It was their hands that tortured
suspects to extract confessions and coerced witnesses to speak. They
were the instruments of discipline and punishment, from backroom
floggings to public beheadings. Only men of such low status could be
recruited to carry out such disagreeable tasks. And only those with a
taste and talent for violence, cruelty, and physical domination could be
expected to do such work efficiently.
The runners were men with few prospects; furthermore, as noted,
men in all positions of the imperial bureaucracy were poorly paid and
were expected to supplement their meager salaries by one means or
another. Extracting surplus from the local populace did not necessar-
ily entail direct intimidation, but it exploited the principle of reciproc-
ity that so defines Chinese social relations. In practice, the spectrum
ranged from gift-favor exchanges (glossed as li, protocols of “human
feeling,” renqing ) to outright extortion, blackmail, and bribery. Never
theless, such interactions point to the mediating role of yamen run-
ners. Though despised as predatory bullies, if they were provided with
the appropriate “recognition” and reward, they could facilitate (or
impede) grievances and proceedings, bring to light or cover up infrac-
tions and crimes, overlook illegal or unsavory activities, or apply the
letter of the law to their benefactors’ competition.6 Meanwhile, the
relationship between appointed officials and functionaries was also
negotiable, reciprocal, and sometimes delicate. In the worst case, offi-
cials were at the mercy of their minions: without loyal, dependable
subordinates, magistrates, and commanders would have little chance

of real achievement (or of filling their own coffers).
As ritual master Lin Jiafu of Taidong explained to me, a magis-
trate or military official could not catch outlaws and bandits by him-
self, even if he was a skilled martial artist; he needed constables and
soldiers to bring them to judgment, bailiffs to carry out torture and
interrogation, guards to escort them to and from their dungeon cells,
and executioners to carry out sentences. Similarly, gods activate their
potential through the force of their minions, the spirit soldiers and
generals.7
Spirit bailiffs and soldiers are just a step removed, in both sub-
stance and character, from the ghosts they are delegated to control.
Conscripted directly from among the hosts of unworshipped dead,
they resemble outlaws who have accepted the emperor’s amnesty. Not
all are willing recruits; only partially rehabilitated, they retain a mea-
sure of demonic malevolence.8 Like yamen runners, their ambiguous
status requires constant vigilance on the part of the gods (and Daoist
priests and ritual masters) they serve. They can threaten or embarrass
their patrons; at the same time, they insulate them from the pollution
of yin death residues, much as the rowdy runners allowed magistrates
to keep their hands clean. Both are thought to be difficult to con-
trol and must be frequently propitiated (paid off). Their troublesome
character reflects their transitional state along the yin/yang continuum,
intermediate between ghosts and gods. Thus the rituals that invoke
them are invariably conducted outside, never within the threshold of
a temple or home. As I explain in the next section, the coarse, violent
character and ambiguous moral quality of these spirits are more than
a mythologized version of the imperial past; in the cultural idiom of
Chinese religion, they are central to the logic of ritual efficacy.
The Celestial Interface

Most believers probably do not give much thought to the question
of how gods manifest their miraculous power in the human world;
what matters is whether or not they are efficacious in answering the
requests of petitioners. But an ordered, if arcane, logic emerges from
the exegetical discourse and liturgical practice of ritual experts like
Master Lin. The first axiom of this logic is that the visible and invisible
realms are merely phases along a continuum, the realm of qi. Ghosts
(gui) and gods are thus constantly interacting with the living.
The beings of the invisible realm, however, are all spirits of the
dead. The qi of those who die violently or prematurely lingers among

the living, tainted with the residues of decay. These noxious beings,
generically labeled ghosts, are the most dangerous, since they are not
only poisonous but also bear malice toward the living. On the other
hand, the remains and spirits of those cared for by living kin are trans-
formed into ancestors (zuling) , and ghosts who possess extraordinary
power or talent can be redeemed and installed as gods (shenling) . The
dichotomy between gods and ghosts is not strictly one of good and evil:
either can be benign (zheng, upright) or malignant (xie, perverse).
Human misfortune, in this schema, is often the consequence of
an encounter with malicious spirits. Upright gods (zhengshen) , how-
ever, can intervene to prevent or alleviate such misfortune; it is their
responsibility to police the ghosts and other polluting anomalies that
threaten the living order. Here the logic takes a more concrete turn:
driving out malevolent spirits entails a military approach. Even the
most benign rituals commence by summoning spirit soldiers and gen-
erals (tianbing shenjiang) , who are commanded to fight and capture any
malevolent forces that might be present. In the registers ( jue) used by
ritual specialists, for instance, the majority of talismans ( fulu, charms
written on rectangular strips of red or yellow paper) command gods
to send their spirit soldiers and generals to subdue and destroy odious
beings. Talismans are composed as metaphorical bodies, beginning
most often with three v-shaped marks at the top (the “head”) repre-
senting the Three Pure Ones (the transcendental order, or Dao itself ).
Next appears a term meaning “respectfully accepts orders from,” fol-
lowed by the name of the god delegated to the task at hand—perhaps
Guan Gong, Mazu, or the Emperor of the Dark Heavens. The orders
for the talisman’s “mission”—for example, “dismember demons and
extirpate evil airs” or “suppress a dwelling”—are written next, at the
center of the talisman (the “torso”). On some talismans this order is
flanked (or followed) by additional subsidiary spirits such as the “six
ding and six jia generals” and/or the “five thunder deities,” which add
additional force to the god’s command. Further down, near the bot-
tom of the talisman (in an area called the “spleen,” fudan) are more
military symbols associated with exorcism, including the Thirty-six Big
Dipper Generals (Tiangang); beneath these again, at the “foot” of the
talisman ( fujiao) is a crosshatched pattern representing the Seventy-
two Baleful Earth Generals (Disha).9
The gods’ attire mimics the conventional styles and ranking codes
of Song and Ming dynasty officialdom, though not always with per-
fect historical accuracy. Most of what is tangible in popular religious

practice—from ritual implements like double-edged swords and the
magistrate’s gavel to the archaic language of spells and charms—simi-
larly evoke a now distant imperial past. Anthropologist Arthur Wolf
was among the first to offer a sociological explanation for the per-
sistence of imperial styles and archetypes in the symbolic system of
Chinese popular religion. He suggested that the gods were depicted as
imperial bureaucrats because peasants’ most concrete experience of
power was embodied in government officials, who were usually distant
and inaccessible and thus presumably best represented by an abstract,
mythologized image of bureaucratic power (A. Wolf 1974). Of course,
Wolf ’s explanation raises as many questions as it answers; for example,
if the gods were merely a conflation of abstract power with the insti-
tutions and agents of the state, why, then, would peasants—who by
the twentieth century in Taiwan certainly had a clear idea what real
government officials looked like—not update their representations,
preferring instead to render them in this peculiar, archaic form?
The hierarchical order, military chain of command, and represen-
tational logic of the Chinese pantheon are mythologized transforma-
tions, rather than simply copies, of imperial institutions. The beliefs
and rituals of popular religion are more than a vestigial survival of
the historical relations of power in the imperial past. To some extent
they are that, but if only that, it is unlikely that contemporary believers
would find it sufficiently compelling to devote, as many do, so much of
their energy and economic resources to their practice. Of course there
would be little of interest to the scholar or the average believer with-
out the tangible properties of temples, gods, incense, the sensual and
social intensity of festival celebrations, and the fascinating richness of
narrative and ritual traditions. But the persistent appeal of Chinese
popular religion—the traditional temples, their archaically costumed
gods, and the arcane details of the supernatural world—lies just a bit
further beyond the horizon of history and the attraction of tangible
qualities.
Partly in response to Wolf, a number of anthropologists have pro-
posed somewhat more nuanced analyses of the cultural and social log-
ics of power represented in Chinese religion.10 Most agree that the
bureaucratic model of supernatural power cannot by itself explain the
gods’ miraculous power. Yet the “imperial metaphor,” to use Stephan
Feuchtwang’s term, is certainly relevant to our understanding of this
symbolic realm. Drawing on the insights of Feuchtwang, Steven San-
gren, and Angela Zito in particular, I adopt the view here that supernat-
ural power (ling) is premised on a hierarchically layered cosmos, each
layer of which is embedded within hierarchies each of which repro-
duces the whole in microcosm. The miraculous efficacy of gods mani-
fests an immanent power. Their rank, or authority, depends on this
immanent power, not the other way around. Thus we find old, derelict
temples whose deities have ceased manifesting miraculous efficacy and
new sites of great miraculous power identified with ghosts or deities of
low bureaucratic rank. Within their own domains, moreover, deities
are represented as autonomous masters, presiding over the affairs of
their own houses (temples).11 Supernatural power is thus most like
the charismatic power embodied in the archetypal male form of the
emperor, which transcended the mere mechanics of bureaucracy.
Immanence and Self-Transcendence

The immanence of male charisma in the body/person of the emperor
perfectly exemplifies what Steven Sangren has called the Chinese
“patrilineal mode of production of desire” (2000:189). Here, “man-
hood” and “masculinity” are seamlessly merged. Manhood is a status
that must be attained and is socially recognized. A man in this sense is
a moral and social exemplar; he possesses an exclusive qualification to
serve critical ritual functions, which makes him a key agent of patriar-
chal order. Masculinity, on the other hand, is an amalgam of flexible
characteristics, potentials, attributes, qualities, attitudes, and behav-
iors.12 Some of the standard expectations of manhood indicate a mea-
surable degree of masculinity—producing male heirs not the least of
them. On the other hand, manhood is predicated on the control and
repression of masculine proclivities, most particularly the iconically
masculine (but infantile, anti-social) desire for unrestrained physical
and emotional autonomy.
Briefly, Sangren posits that the psychodynamics of the Chinese
patrilineal family give rise to a crucial emotional conflict between
fathers and sons. A man’s social prestige, economic security, and very
identity rest on his ability first of all to marry and produce heirs and
then to preside over an extended household in which even adult sons,
married and with children of their own, remain resident and subordi-
nate. All constituents of the household pool their labor and earnings,
and an increase in family wealth and property represents an extension
of the patriarch’s charisma.
Paradoxically, a patriarch’s moral authority derives from his own
exclusive position as a sacrificer (that is, a son) vis-à-vis his patrilin-

eal ancestors (Zito 1987:363). The effect of this symbolic inversion is
that the father is seen to embody in his person the entire patriline,
from the ancestors (whom he sustains through sacrificial offerings) to
the succeeding generations, his own progeny. The mediating position
of the sacrificer, that is, engenders a kind of ritual omnipotence. His
sons, meanwhile, can either bide their time or rebel, a problem I take
up again below in this chapter.
In addition to underwriting the institution of patriarchy, the Chi-
nese imperium was the “patrilineal mode of production” writ large. In
the performance of rituals of grand sacrifice, the emperor also placed
himself in the role of supplicant, to Heaven and the imperial ances-
tors. Situated in an exclusive mediating position between Heaven and
the visible world, the emperor ritually produced the manifest order
(tianxia, or “all under Heaven”) itself (Zito 1987:363 passim).
Positing the Son of Heaven as patriarch-producer of “all under
Heaven” made plausible as well the myth of the emperor’s superhu-
man masculine prowess.13 Radiating vital yang energy, he was hyper-
bolically blessed with a life span of “ten thousand years” (wan sui) ,
throughout which he would enjoy vigorous good health and the power
to continually produce sons by his multitude of wives and concubines.
The person—the body—of the emperor, then, combined rather
than filtered out the contradictions inherent in being a man in Chinese
society. Here patriarchal authority and filial piety are linked with sex-
ual potency and vitality. It was, finally, the emperor’s role as producer
of the bureaucratic order, his power of mediation between realms,
rather than any characteristic of the bureaucracy itself that seems the
most plausible derivation of ling, the power to mediate between the
transcendental and visible realms.14
The Jianghu
One event that I had not (and could not have) planned or anticipated
marks my passage into the role of true participant-observer. One eve-
ning in February 1990, just before the Lantern Festival (first full moon
of the Lunar New Year), I was at the Loyal Harmony Temple, video
taping the Military Retainers troupe as they trained for their perfor-
mance in the annual festival procession. Halfway through the session,
A-gi appeared and signaled that she wanted to talk to me. I walked
over to where she stood, and we were quickly joined by Mr. Liao, chair
of the temple management committee. They then asked me if I might
perhaps be interested in playing the part of Civil Magistrate (Wenpan)

in the upcoming procession. Trying not to show my elation too obvi-
ously, I immediately agreed. From that moment, I became a partici-
pant in the ritual life of a Taiwanese martial temple cult.
A few minutes later, Yam was already putting me through some
basic drills, teaching me the steps, postures, and gestures; describing
the timing of each of the sub-rituals; and explaining the Civil Magis-
trate’s position and function in the Military Retainers troupe. For the
next two weeks, I spent every evening training in the temple courtyard,
practicing stances, gestures and ritual procedures and, just as impor-
tant, learning the lore and listening to stories of processions past. For
three days before the festival my fellow “generals” and I observed the
required regimen of ritual purification; on the eve of the procession,
we were confined to the temple grounds, eating and sleeping in the
temple dormitory. From this new position—both spatially and socially
closer to the people and actions I was trying to understand—I was
allowed access to conversations, events, and details of practice invisible
or obscure from the conventional observer’s perspective.
The procession itself, which lasts for two days and nights, is a period
of emotional and social intensity, both ordeal and entertainment. Dur-
ing festival time personalities are magnified, bonds grow tighter, and
existing tensions flare into open conflict. After the procession finally
reaches its conclusion on the second night, the temple elders invite
the exhausted procession troupe actors to a feast in the courtyard.
Seven or eight round tables, each seating ten guests, are set up and
laden with dishes of pork, fish, chicken, and seafood. Every table is
provided with a case of chilled Taiwan Beer in half-liter bottles. This
“reward for the troops” (kao jun) invariably involves multiple toasts and
endless rounds of drinking, smoking, and betel chews.
It was already past midnight by the time we began eating and
drinking at this, the first of what would turn out to be many feasts
with the devotees of the Loyal Harmony Temple. After we had emp-
tied our beer glasses a dozen or so times, some of the young toughs
with whom I had just performed began to talk with me directly—first
joking and bantering, then a few more serious exchanges. Until then,
during training sessions and breaks, most had seemed uninterested or
reticent. Apparently, the shared ordeal (and probably the attention
that my participation had attracted in the local community and from
the media) had kindled a new level of rapport between me and my
informants.
The Military Retainers troupe of Taidong Loyal Harmony Temple encounters

miscreant ghosts. The Retainer running toward the camera (Er Ye) has lost his hat,
a sign of demonic attack that usually precipitates a violent possession performance.
All photographs are by the author.
Most of the devotees at Loyal Harmony Temple, and nearly all

the troupe performers, were connected to the jianghu, either making
their living in the shadow economy or closely related to someone who
did. Some were even more deeply invested. They were “brothers of
the dark path” (heidao xiongdi) , small-time gangsters and drifters that
Taiwanese derisively call lomoa* (hereafter lomoa).15 I had thus inadver-
tently gained entrée into the marginal world of Taidong’s jianghu, the
“rivers and lakes.”
No English phrase quite captures the full meaning of the term
jianghu.16 In the cultural imagery of popular novels, dramas, and comic
books, the “rivers and lakes” is a fantasy world of vagabonds and adven-
turers, bandits who rob only the rich, along with assorted con men,
thieves, and gamblers. Whether by choice or fate, these are all charac-
ters alienated from the mainstream and at odds with convention and
often the law as well.17 Wandering among, but somewhat aloof from,
this motley bunch are the swordsmen, fighting monks, and other mar-
tial adepts, whose fighting skill and (for the upright) heroic deeds lead
them to fame within the wulin, the “world of martial fame.” 18
Like most myths, tales of the jianghu oscillate between the famil-
iar and the fantastic. Life in the jianghu parallels the everyday, but
it is never mundane, more an intensification than an exaggeration
of the ordinary. The heroes of the rivers and lakes do not just sit
down for a quick bite; they order piles of meat, washed down with
pails of grain alcohol. And such feasts are not merely celebrations
of prodigious masculine appetites—they are preludes to the chance
meeting of fellow heroes or death- defying adventures. In the Ming
dynasty novel Water Margin,19 Wu Song, one of the narrative’s most
memorable martial heroes, stops at a tavern to eat before heading up
a forested ridge that, he soon learns, is home to a vicious rogue tiger.
This tavern advertises that its wine is so strong that “three bowls and
you can’t make it across the ridge.” Wu Song takes this as a personal
challenge and not only drinks many more than three bowls, but, filled
with drunken courage, also heads up the ridge at dusk, defying all
warnings. As expected, as he reaches the deepest part of the forest,
he is ambushed by the savage tiger. After breaking his staff and seem-
ingly cornered, Wu Song finally subdues and kills the tiger with his
bare hands.
On another occasion, Wu Song stops at a roadside tavern run by the
“Female Yaksha demon” Sun Erniang and her husband Zhang Qing,
who make their living drugging and braining hapless strangers, then
rendering their corpses into filling for steamed meat buns. Streetwise
Wu Song catches the ruse and avoids a grisly death. Pretending to have
been knocked out by the drugs in his wine, Wu Song grabs and sub-
dues the treacherous Sun and keeps Zhang at bay. Overpowered and
outmatched, they beg him for mercy. Wu Song then magnanimously
spares their lives. After the three all recognize each other as fellow
travelers, they become sworn comrades. Killing unwary travelers and
turning them into snacks, in the context of the story, is a desperate way
to make a living. Like the roving fighter-for-hire Wu Song, the pair are
drifters on the rivers and lakes.20 The message here, repeated, as we
shall see, in the narratives of many of those who speak in this book, is
that the jianghu, fraught with danger, insecurity, and moral ambiguity,
is also a shared fate and a common ordeal.
In addition to showcasing again Wu Song’s quick wits and fighting
prowess, this gruesome episode evokes a fundamental anxiety within
Chinese culture: outside the confines of one’s social familiars, the
in-groups of family, kin group, and village, life is precarious. Strang-

ers, from the perspective of villagers, are a threat, likened to ghosts
precisely because they are rootless. But the association is more than
metaphorical: an unknown stranger is a possible threat but more often
an easy victim. Dying alone and far from home, outsiders are literally
transformed into unworshipped, solitary ghosts. Wu Song later rec-
ognizes some of the corpses strung up like ox carcasses to have been
heroes and skilled fighters. As solitary travelers, even these worthy men
became easy victims.
From classical vernacular novels like the Water Margin to the best-
selling works of modern authors like Jin Yong and Gu Long to comic
books, television serials, and the gangster and “kung-fu” genres of Hong
Kong and Taiwan cinema, the violence and romance of the jianghu are
among the most enduring and pervasive motifs in Chinese popular
culture. At the same time, jianghu also designates society’s social and
economic margin and the diverse class of people who live by their wits,
skill, and, sometimes, brutality. In contemporary usage, “rivers and
lakes” is often taken to be synonymous with the outlaw world of the
“dark path”—that is, gangsters, gamblers, and prostitutes. The “dark
path” in this sense is unambiguously a hard-core underworld of pro-
fessional criminals, organized rackets, shady businesses, and crooked
politicians. Bribe-taking officials, on the other hand, are not properly
part of the jianghu, whereas small-time, unlicensed street hawkers, for
instance, are technically outlaws but can in no way be considered hard-
core “dark path elements” (heidao fenzi) .
Embarking on any sort of risky venture, whatever its legality, can
be described metaphorically (though ironically) as “going out into the
rivers and lakes” (chu jianghu) .21 Jianghu thus signifies both the anxiety
and the allure of a life of risk and possibility. The term evokes mobility,
fluidity, and movement, life in a shadow society populated by thieves,
gamblers, prostitutes, highwaymen, itinerant swordsmen, drifters, and
entertainers. It is a world that can exist only beyond the stability and
security of village and family and conventional occupations.
Then again, jianghu is not truly a separate realm at all. Socially
and economically, it plays an important if not always visible role in
Chinese and Taiwanese urban working-class life. As for its denizens,
being “on the path” (daoshang) does sometimes require going out “on
the road.” As sojourners and occasionally as fugitives (not always from
the law), they sometimes lead a vagabond existence far from home.
In story and song, they lament transient lives afflicted with loneliness
and regret. But I would propose that even when sparked by physical
separation and distance, these moods are a culturally specific form of
alienation that is best understood as a metaphorical disconnection, a
psychosocial estrangement from the institutions of mainstream society
(including families and communities).22
Despite these intimations of deviance, I would surmise that for
most Chinese outside of official and intellectual circles (whether on
the “path” or not), the jianghu appears as an unavoidable, if not exactly
“normal,” social fact. Despite historical links and some formal resem-
blances, there are significant differences between the jianghu cultures
of China, Hong Kong, and Taiwan. Some of the differences are owing
to their divergent political and historical environments. In the PRC,
the huge scale of rural-urban migration, oscillating and uneven lev-
els of law enforcement and social control, and discontinuity between
pre-revolution and post-Mao iterations of the jianghu require a greater
sensitivity to historical context.
Hong Kong and Taiwan are similar to each other (and differ from
the mainland) in that, among other things, local gangs have been a
persistent feature of working-class society, without significant interrup-
tion, throughout the modern era. Levels of urbanization are similar,
and in both societies, working-class people—by which I mean specifi-
cally those who have been marginalized by the two-track education
system—are likely to know—as neighbor, customer, or perhaps kin—
someone in the jianghu. In Taiwan and Hong Kong, and increasingly
since the late 1980s in the PRC as well, working-class men and women
seeking a livelihood move relatively easily between the “proper” and
illicit spheres. It is probable that at some point in their lives many men
and women will find themselves faced with the prospect of stepping
into the rivers and lakes. It may be a conscious choice; in many cases,
it is unintended and inadvertent, a venture into the “gray” rather than
the “black” way.
At the far end of the spectrum, the jianghu includes the parasitic
and sometimes violent criminal underworld, from freelancing con art-
ists, gamblers, and fences to organized gangs and rackets; from loan
sharks to debt collectors, human traffickers, bank robbers, and hit
men. In this sense, it certainly serves as a social and symbolic refuge
for those at and beyond the margins of family and community: the
alienated, discontented, habitually violent, and criminal (cf. Ownby
1995:1039). Yet even this world is enmeshed with the everyday, and
only a few cross the line without recourse. From this perspective, the
jianghu is intrinsic to, and in some ways constitutive of, working-class

society, not a separate sphere of deviance.23
Honor and Reciprocity

The jianghu is a culturally specific formation unique to Chinese-speak-
ing communities.24 Its “operational norms” and fictive kinship orga-
nization are typical characteristics of the Chinese cooperative associa-
tion, or hui (Sangren 1984:410). But however fuzzy the divide between
proper, legal cooperative associations and the collectivities of the
jianghu, they are not precisely the same thing. We thus have to look
beyond the simple correspondence of organizational characteristics
to understand what, if anything, is unique to the “rivers and lakes”
context.
I begin with the discursive, or one might say spiritual, attributes
of the jianghu. People throughout the Chinese-speaking world, within
or outside jianghu society, would identify one moral quality above
all—yiqi, the “spirit of honor”—as a defining characteristic of jianghu
culture. Yiqi, which can be glossed as “honor” or “righteous spirit,”
implies a measure of manly courage and generally indicates acts of (or
a spirit of) selfless generosity. An act is deemed honorable and righ-
teous only if there is no ulterior motive or expectation attached. It is
not the repayment of a favor, nor is it intended to incur an obligation
or counter-favor. Honor entails self-limitation; sacrifice; and a post-
ponement of personal interest, need, or desire. Honor here trumps
profit.
Yet honor is nevertheless a form of reciprocity that entails social
and material, as well as moral, obligation. Assuming that all parties
are equally bound to the principle of honor, a righteous act on one’s
behalf raises the stakes of the relationship and incurs an obligation to
respond (and a sanction for failing to respond) at some appropriate
future occasion. In this sense, honor comprises a fundamental param-
eter of certain kinds of relationships that are premised on trust and
most vulnerable to its violation (for example, friendship, collegiality,
contractual agreements of all kinds). Within those relationships, it is a
form of exchangeable symbolic capital.
In myth and fiction, the jianghu is “a realm of freedom, where the
laws of family, society, and state no longer apply” (Shahar 2001:380).
But even the wandering knights-errant of fictional jianghu narratives
are social animals, burdened by their nearly obsessive commitment to
honor and righteousness. They are driven to revenge but also impelled
to forgo personal pleasures and profit, to put themselves in harm’s way

for the sake of those who have shown them the same regard in the
past. The knight errant and freelance assassin live by and for loyalty,
trust, and fellowship. It is precisely this quality that men of the jianghu
are said to hold most sacred; it is honor that excuses their other moral
transgressions and that allows them to be men. Honor in this sense
has a dual nature: righteous and heroic actions are assertions of indi-
vidual agency that implicitly concede the actor’s acknowledgement of
his place in and submission to the social order.
I once asked Siong San, one of my more self-consciously “righ-
teous” underworld acquaintances, “What exactly compels a man to act
with honor?” “Sometimes you have no choice,” he replied, quoting an
old cliché. “In the jianghu, you are not your own master [ren zai jian-
ghu, shen bu you ji ]. When you’re on the spot, there is no way to ‘get
off the stage’ [xia bu liao tai ].” What if you fail to act honorably? “ You
would be ostracized. What does it mean to have sworn brothers if there
is no honor, no righteous spirit?”
The popular image of the “brothers of the rivers and lakes” is an
amalgam of representations ranging from classical narratives like the
Water Margin to media portrayals of Japanese yakuza and Hong Kong
and Taiwanese gangster movies. The typical “brother” is depicted as a
man from the economic margins deeply loyal to his sworn brothers,
for whom face and honor are all that matter. There are two primary
and somewhat contradictory types of gangster (as described/imagined
in the popular media, at least). One is a loyal organization man driven
only to climb the “corporate ladder” (gang hierarchy). His ambition
is to become a wealthy, powerful “big brother” or godfather, the patri-
arch of his fictive lineage. The other is a solitary outlaw, a knight-errant
who prefers to operate outside the constraints of hierarchy.25 Like the
peripatetic mercenary Wu Song or the freelance hit man John (played
by Chow Yun-fat) in the classic 1989 Hong Kong gangster film The
Killer, they live for honor and the fellowship of their comrades but
shun the organizational mainstream.
The “organization man” is perhaps less romanticized than the soli-
tary outlaw, but both types appear frequently as characters in popular
narratives, especially in Taiwan, where resistance to external hege-
mony is central to the island’s historical mythology. The alienated
loner filled with impossible desire and driven to act beyond the law by
frustration is a common character in contemporary popular culture.
In films, television serials, and pop songs about life on the “rivers and
lakes,” as well as in sordid tabloid accounts and published confessions

of famous gangsters, such figures either fade away in anguished loneli-
ness or go out violently in a blaze of glory.
The myth of the gangster (anti-)hero serves as an important part
of the Chinese masculine imagination. It is not, however, its totality.
Rather, the myth comes to life for most men when carousing with their
“brothers of the path,” singing songs full of wanderer’s angst; drink-
ing and gambling without regard to wins and losses; showing loyalty
and honor through generosity and standing with friends in case of
threat; defending family honor; and displaying the proper balance of
social dominance, sexual desire, and emotional control in relations
with women.
Manhood here is thus constituted out of contradictory qualities.
As sons beholden to their fathers (including fictive fathers, those who
stand in positions of power and authority) and elder brothers (includ-
ing fictive brothers), Chinese men are expected to unquestioningly
accept their subordination to their elders’ authority and the collective
interest. The desire to empower themselves must be repressed until
such time as the responsibility of fatherhood and family leadership
falls upon their shoulders. Young men, if they have any prospect at
all of emerging as men of position and prestige, must endure a long
wait, during which the central marker of their social position is that
of subjection and obedience. The frustration this engenders cannot
be expressed openly without directly challenging those who currently
own the claim to authority. The knight-errant or outlaw, then, belongs
to a mythology of resistance and rebellion that represents desires that
could not otherwise be expressed. The archetypes of this mythology
include both the man of prowess—the mature, gifted, and skilled
strategist or warrior—and the mischievous, capricious, rebellious male
child who confounds, embarrasses, and inverts the patriarchal order
by manifesting his individual desire openly and to excess. Competi-
tive, capricious, even violent action fulfills the desire, so clearly repre-
sented in myth, to transcend, if not to overtly break, the bonds of filial
obligation.
Class and Masculinity

Much as the meaning of “theory” in scientific research has been mis-
represented and abused in the debate over the teaching of evolution
in American public schools, the term “myth” is often misleadingly
conflated with “falsehood,” “untruth,” or “superstition,” not only in
common parlance, but by many writers with academic credentials as

well. Gong Pengcheng (1987), for example, suggests that while we
praise the knight-errant as a man of righteousness and “liberator of
the oppressed,” we could hardly admire the street toughs, secret soci-
ety members, and gangsters who similarly espouse these virtues. For
Gong, the knight-errant ethos exists only in novels and the popular
imagination: Robin Hood never really existed, or if he did, he was no
Robin Hood. Such stories, he insists, distort the facts of history; they
are mystifying delusions.
Debunking myths, however, explains little about how such myths
came to be or what their value and function in the process of socio-
cultural reproduction might be. The knight-errant is a role model for
criminal gangs and secret societies, his moral code the basis of their
ethos of loyalty, righteousness, recompense, and revenge. This suggests
the importance of such mythology in the transformation and produc-
tion of self and in the collective reproduction of structures of value
(which necessarily takes place through the socialization of individuals).
The gangster identifies with the righteous hero, whose transcendental
moral and positive social value are validated by the apotheosis and
official recognition of martial heroes such as Guan Gong. He thereby
rationalizes his role as legitimate within the established social structure,
rather than recognizing in his alienation from the mainstream a legiti-
mate reason to challenge the hierarchy. Thus in their self-construc-
tion, while their activities may be inherently anti-social, people such as
gangsters claim a share in, and thereby reproduce, the same hierarchy
through which they are systematically subordinated. Their identifica-
tion with the given relations of power is manifested in the structure of
official corruption, a dependence on the maintenance of the status
quo; disruptive as their activities are in fact, organized criminal gangs
tend to be ultra-conservative in their social and political attitudes.
The so-called “martial virtues” are transcendental; their positive
value is unquestioned. The association of these virtues with the mar-
tial hero enables otherwise socially disenfranchised individuals and
groups to evaluate their actions and conceive their intentions in a way
that accords with normative values: “Although they are in the green
wood, in fact they hold loyalty and righteousness in their hearts; they
are upright, straight, and unselfish. It is all because of the oppression
of corrupt officials; they really had no choice” (Gong 1987:7).
Their own subculture is structured according to the same axiom-
atic principles as the larger society and is nested within the larger social
framework. For example, even cults and secret societies whose activities
are clearly counter to mainstream interests publicly style themselves in
terms of universally understood traditional goals and virtues. The code
of honor espoused by secret societies, street toughs, and criminal orga-
nizations—that of recompense (bao’en) and retribution (baochou) — is
rooted in (and expressed in terms of) the Confucian virtues of loyalty
(zhong) and righteousness (yi).
Xiao Zhang was a struggling twenty-three-year-year-old small-timer
when I met him in Taidong in 1988. He often spoke nostalgically about
his glory years as the leader of a teenage protection ring that controlled
the turf around the Hualian train station. The gang had dispersed, as
many do, when the older members (including Xiao Zhang) left for
compulsory military service. When he got back to town, he was no lon-
ger a big shot and had no particular abilities to attract the attention of
the local “big brothers” (da ge —that is, gang bosses). Seeking divine
intervention, he had sworn an oath before the martial god Guan Gong
(in Taiwan the patron deity of both gangsters and policemen), promis-
ing to “succeed” by the age of twenty-eight, or he would sacrifice his
life. “Success” for him meant status and financial security, which in his
eyes were attainable for him, the minimally educated son of mixed
Taiwanese-Amis parents, only by “following the dark path” (zou heidao) .
His ultimate objective in this endeavor, as he explained it, was to fulfill
his filial obligation to his mother, who supported her children work-
ing in a brothel in Taidong. He felt that he had no choice other than
to “get by on the dark path” (zai daoshang hun). To Xiao Zhang, social
and financial success were secondary to honor. He summed up the
ethos of “brotherhood” for me in the following terms: “ You can forget
everything else —for us, all that matters is honor.”
As others have observed, the “promise of access to supernatu-
ral power through the manipulation of spiritual forces” (Ownby
1995:1024) may hold particular appeal to marginalized young men
(cf. DeBernardi 1987; Sutton 1989). As noted, rather than promising a
way to bypass or restructure the social hierarchy through exposing the
arbitrariness of the mechanisms by which it is maintained (a clearly
subversive endeavor), secret societies and “dark” cults offer certain
marginalized individuals and groups a chance to identify with main-
stream images and values, to redefine, for themselves and society at
large, their place in the social hierarchy. Through this activity, such
groups reproduce the hierarchical structure that underlies the Chi-
nese social pattern, even if the structure is self- consciously inverted, as
in some “dark cult” and secret society rituals (DeBernardi 1987; J. S. M.

Ward and Stirling 1925).
The Interpenetration of Wu and Wen

Even a cursory examination of Chinese cultural production suggests
(and native interpretation confirms) that the military or martial (wu)
and the civil or literary (wen) are important constituent categories of
Chinese thought, history, religion, rulership, and everyday social life.
The complementary opposition of wen and wu was critical to the way
the Chinese theorized the emergence and sustenance of Chinese civi-
lization itself (both as culture and as imperium). Each played a part,
wen, as “suasive, civilizing ‘patterns’ ” (Zito 1989:67), and wu, as the
extension and maintenance by military force of a political order that
was claimed to embody those patterns. Chinese rulers, from the early
feudal kings to the last emperor (and perhaps beyond) were expected
to display both wen (as civil/literary) and wu (as military/martial)
qualities and attributes. The ideal emperor was master of both literary
and martial arts, a scholar-poet and a warrior. 26
The representations and behaviors generally implicated within the
category of “the martial” in Chinese culture (1) are associated with
the subclass of those engaged in marginal, “mean” occupations and
(2) are expressive symbols that tangibly and kinetically represent the
power inherent in violence. Because such symbols legitimate the uses
of violence (both individual fighting prowess and the magnified force
of an organized group) outside those sanctioned (and deployed) by
and for the state and because they are, as noted, associated with local
cults and local interests, they are therefore viewed with suspicion, par-
ticularly by the elite classes, who recognize therein a potentially sub-
versive challenge to the hegemony of the state and to their own self-
serving structure of value based on wen.
Zito defines wen as a “process which includes writing, ritual, and
building, [and] should be approached as a body of practices that pro-
duce the king and his imperium, as sensuous embodiments of the
king’s power” (1989:70). The key point here is that wen (and wu) are
productive, transformative processes, not simply static taxonomic catego-
ries. I would contend, then, that we need to examine both wen and wu
as “bod[ies] of practices” and that both produce not only “the king
and his imperium,” but ordinary social subjects and the local cultural
lifeworld as well.
Feuchtwang (1992:17) argues that the “imperial metaphor” of the
Chinese folk pantheon is not simply historical reenactment but rather

appropriates the imperium directly to itself. Local cults and commu-
nities adapt the symbols and structure of the imperium in order to
articulate and manifest the power inherent in their own collective soli-
darity. In taking on the forms of the imperial state, they situate them-
selves within the larger hierarchy. But in reproducing the imperium in
microcosm, they are also asserting a self-contained identity equal to (as
a metonym of) the center. This local production of power, moreover,
is enacted primarily through the symbolic-ritual mode of the martial
(ibid.:161).
From the Confucian perspective, wen is the concrete manifesta-
tion of cosmic principle, li.27 On the other hand, wu signifies the tech-
niques by which efficacy, numinous power (ling) is made manifest. The
conflation of wu with efficacious elevates the status of local collectivi-
ties and cults (by according them an autonomous value and identity,
represented by territorial or cultic deities). It also presents the pos-
sibility and affirms the legitimacy of violence not sanctioned by the
state. From the perspective of the center (the court, the official elite),
such potential was at best subversive in the abstract; at the practical
extreme, it could (and historically sometimes did) become a destabi-
lizing threat to the regime.28 Wu and wen, that is, encapsulate a histori-
cal contest between common and elite structures of value, respectively,
within Chinese society.
As the equation of the written language and civilization itself, wen
formed an exclusive and jealously guarded domain that was the foun-
dation of literate classes’ social prestige and political influence. In this
scheme, the cosmic as well as social order, embodied in the emperor
and the imperial bureaucracy, is founded upon wen. In this view, the
force of arms should be deployed only to extend or defend the civi-
lized order. Unchecked and unleashed, the power entailed in violence
is a constant threat to the civilized order. From the local community
to the macrolevel of state and empire, violent force must be subordi-
nated to those “suasive, civilizing patterns” and strictly controlled.
This insistent focus on patterns and order is a symptom of a char-
acteristic obsession in Chinese culture with order and the correspond-
ing fear of disorder and dissolution. It is particularly pronounced in
the Confucian tradition, dominated by elites who for centuries main-
tained a firm grip on the means of cultural production. One curious
outcome of this elite monopoly on representation is the view, still com-
mon in the West, that premodern China was a model of utopian har-
mony. Beginning with Voltaire and other eighteenth-century writers,

the Confucian classics were often interpreted as literal descriptions
of a timeless, perfectly ordered, and harmonious society governed by
philosophers and sage emperors. This doubly distorted representa-
tion (first by the Confucian elites, then by Western idealists) of Chi-
nese society gradually gained an unshakeable hold on the collective
Western imagination. Thus despite widespread knowledge of China’s
chaotic and violent twentieth-century experience, such fantasies con-
tinued to frame popular imaginings of an ageless China based on an
undying and changeless “tradition.”
Culture and Ideology in the Wen/Wu Opposition

The wen/wu dyad is a fundamental structure of conceptual organiza-
tion in Chinese culture with a tremendously broad scope of applica-
tion. The imperial bureaucracy, of course, was divided into civil and
military branches. The division of other practices and cultural forms
along lines of wen and wu can be explained partly as a metaphorical
transference of this primary institutional structure and partly as a
purely abstract classification of things according to perceived “liter-
ary” and “martial” attributes. For example, both the historical/institu-
tional and abstract/theoretical senses of wen and wu apply in popular
religion as I have described it to this point. In the classical Chinese
theater tradition, all plays, operas, and dramas are classified as either
wen or wu, and conventional character roles, even elements of theater
architecture, are also divided into “civil” and “military” types (B. Ward
1979:28). Imperial sacrifice involved civil and military ritual segments
(Zito 1989:428–436), and Daoist rituals of transformation and renewal
are likewise organized partly according to wen/wu logic (Lagerwey
1987:88, 216 –237).
The immanence of the martial was explicitly represented in Qing
court ritual. Unlike martial ritual in popular religion, the court ver-
sions were muted and restrained, martial only by symbolic allusion.
The martial rituals approved for court performance bore none of
the violent hallmarks of the popular genre: there was no belligerent
brandishing of weapons; no threatening stances and gestures; no sud-
den, forceful movements; no contortions of body and face; no blasting
of objects and men with explosive firecrackers; and no outbursts of
trance-induced rage.
The more colorful aspects of popular religion—spirit possession,
self-mortification, blood sacrifice, and the collective violence of com-
petitive temple festivals—were often derogated by the literati elite

as “licentious” (yin) practices. Daoist priests, Buddhist clergy, and lay
devotees also habitually denigrated popular religion as impure and
degraded. The tangible contrast between popular practice and the sol-
emn formality of Confucian and Buddhist ritual, for example, gave
cause to orthodox Confucians, Daoists, and Buddhists to view the for-
mer with suspicion. For the Confucian literati elite, the term wen itself
defined and encompassed human civilization.29 Popular religion (as
well as the exorcistic rituals and talismans of Zhengyi and Maoshan
Daoism) inverts the Confucian presumption that the martial is sub-
ordinate to the civil and should be called upon only as a last resort.
The ethos of popular, or local, religion is active, brash, and sometimes
violent and valorizes the military and the martial. For orthodox Confu-
cians, martial or military power that is not organized and sanctioned
by the state is illegitimate, subversive, and nearly synonymous with
chaos (luan) itself.
From its first appearance in late Shang dynasty oracle bone divina-
tion texts down to its many colloquial and technical uses in modern Chi-
nese, the term wu has consistently had two related but distinct senses.
The first is a broad ontological category that encompasses all things
having to do with martial skill, combat, self-defense, force, threat, and,
by extension, more or less the entire realm of human aggression. The
second, narrower meaning is derived from the first and denotes insti-
tutions and cultural practices that have a distinctly military function or
martial character.
In either sense, wu is distinct from unconstrained violence; it is the
power of controlling violence with force, to forestall chaos by uproot-
ing or obstructing its causes. Here wu serves only to defend and pro-
tect the legitimate moral/political order against disruptive forces both
within and from outside of society. Some scholars consider this sense
of wu as the central and determining one (Fried 1952:348n). Chinese
philological interpretation of the character wu accords with this inter-
pretation. The ideograph is composed of the elements zhi, to stop or
prevent, and ge, a lance, giving the original sense of the character as
“forcing an attacker wielding a lance to desist” (this is evident only
from forms of the character appearing on early bronze vessels, as the
modern form has been modified such that ge has come to resemble
another weapon, yi, a tethered dart).30
Second, wu functions as the operative link between the unseen
world of Heaven and numinous beings, on the one hand, and the
visible, human world on the other. The moral order of the universe,
Heaven, is personified in the form of deities who concretize and man-
age the specifics of this order. They do not generally intervene in
person, however, but manifest their will (and by extension, the tran-
scendent moral imperative of Heaven) through proxies, primarily the
transformed, redeemed ghosts and demons who command and man
the vast spirit armies. These forces are amplified and reified even more
concretely in the form of human actors who, in certain ritual contexts,
serve as surrogate bodies for both the higher deities and their demonic
spirit generals. Thus the fundamental logic and mechanics of spiri-
tual efficacy (ling, also translated as “magical power” or “numinous
power”) is that of a military hierarchy, and its fundamental actional
and representational mode is that of controlled violence—that is, the
skill and force of arms.
Ling is mediating, transformative power, made tangible through
military symbols and enacted through martial ritual techniques. The
techniques of wu are regarded as particularly efficacious because they
enact this critical, mediative function. To transform one (usually unde-
sirable) state into another (more desirable), it is often necessary to
destroy the first state, to induce disorder for the purpose of achieving
a preferred order.
Martial power enables exception through a temporary disordering
of the hierarchical structures (a liminal moment of chaos); the state
and condition of individuals can be revised, but in the end hierarchy
itself is preserved. Nevertheless, even a single exception (and certainly
a complete inversion) to the ordained process constitutes a breach of
moral and cosmic order. A monopoly on violence has been one defin-
ing feature of nearly every state in recorded history, and China is no
exception. There is thus clearly a threat of sedition inherent in this
desire to disorder and reorder existing relations; subordination within
a hierarchy generates the desire to rise; the institutionalized barriers
that protect those in the upper echelons are inherently vulnerable.
Popular religion, and in particular martial deity cults associated with
the marginal and peripheral elements of society, thus constituted a
low-intensity but constant threat to the enforcement of state power, as
well as a challenge to the philosophical sensibilities of the Confucian
elite. While local cults did not by themselves immediately constitute
any practical danger, they nevertheless represented a desire among
subject people for access to power structurally denied them. Individual
requests for assistance from gods and demons constitute “exceptions,”
not rebellion; yet chaos is entailed even in the possibility of exception

and in the nature of ling -based power.
The founding of the Zhou dynasty (1027–256 BCE) through the
conquest and destruction of the neighboring (and loosely allied) Shang
civilization marks a critical moment in early Chinese history. Among
the early Zhou period’s greatest innovations was the introduction of
the concept of a “Mandate of Heaven” (Tianming), which lends tran-
scendental moral legitimacy to the exercise of political and military
power. Prior to the conquest, the Zhou’s (as well as the Shang’s) claims
to power were based largely on the hierarchy of ancestors—those of
the royal and aristocratic clans, it was understood, were simply more
powerful and efficacious than those of other clans. This innovation of
political ideology, along with superior martial skill and adoption of the
war chariot, significantly accelerated the development and expansion
of Chinese territory and cultural influence.
Kings Wen (the “civil/literary” king) and King Wu (the “martial/
military” king) were the posthumous clan-temple names of the two
great legendary founders of the Zhou dynasty. Wen was the father of
Wu, whose military campaigns brought down the degenerate King
Zhou of Shang. This historical paradigm establishes the asymmetry
of the wen/wu dyad in the Confucian (read: wenren) account. Here,
wen is to wu as father is to son. Or, to cite another classical expression,
“This is the way of King Wen and King Wu: The bow is drawn, the
arrow released” (Wen wu zhi dao, yi zhang yi chi) . The stretched bow is
the “unissued” (weifa) , the inherent, the unrealized. The release of the
arrow is the activation of force, the “becoming,” the “already issued”
(yifa) . It is the dynamic force of action itself. Wen, as “pattern” and
“principle,” then, is the moral will, while wu is the amoral destructive
power that, in the hands of Heaven’s proxies, becomes a mediative
force in the production, reproduction, and (when necessary) restora-
tion of cosmic and moral order.
The Chinese concept of a Heaven-ordained imperium legitimates
the authority of the state vis-à-vis all threats to its stability and hege-
mony, both internal (rebellions, local interests that subvert the author-
ity of the center) and external (invasion, pernicious cultural influences
from abroad). In this version of the cosmic and political order, wu is
predicated on wen. Played out as the history of (Chinese) civilization,
wen is the transcendental principle of hierarchy underlying civilized
order, while wu supplies the means to transform barbarian space into
civilized space.
Among the pet tenets of classical Confucianism is that the moral

example and charisma of an exemplary ruler are sufficient to sustain
the inner integrity of the imperium and to enlighten the “barbarian”
peoples who threaten from without (thus effectively extending the
empire through acculturation rather than conquest). Yet no Confu-
cian was so radical as to suggest that there was no need for the military,
that the martial served no practical purpose. Indeed, Confucius him-
self advocated the cultivation of physical as well as intellectual talents:
archery and charioteering were two of the “six arts” (liu yi) cultivated by
superior men. The superordinate status of wen, then, was always a logi-
cally tenuous proposition. Learned in history, Confucian scholars were
well aware that the Mandate of Heaven passed from the corrupt Shang
to the upright Zhou dynasty only through King Wu’s military conquest.
Potential without dynamic action remains incipient, invisible, unreal.
It is largely due to the influence of Confucius and his intellectual
heirs that history came to be the defining mode of cultural mythology
in China. The Confucian tradition rests on the premise that the cos-
mic order, the Mandate of Heaven, was immanent in the thoughts and
deeds of the ancient sages, recorded in the classics. Those who know
and have a correct and profound understanding of classics can culti-
vate themselves to be exemplary moral beings. In the Confucian view,
then, the knowledge class—scholars, literati, the men of wen —were
thus the best qualified of all to administer the empire and arbitrate
social relations. They were ideal public servants but also the epitome
of civilization itself.
That Confucian scholars felt compelled to endlessly expound and
explicate the moral metaphysics of wen suggests that they were uneasy
about its, and their own, hegemonic position. In Confucian theory,
wen is cosmic pattern, but it is also the set of practices that defines the
literati lifeworld and profession and sets them apart as a class from
the common people (min). In particular, wen signifies everything con-
nected with the written language. Confucianism claims the written
word (and in particular the words of Confucian core texts) as sacred
in itself. The Confucian classics are not only repositories of informa-
tion, but also magically embody the wisdom of the ancient sages. Thus
wen, the manifest principle of Heaven’s moral order, is declared the
exclusive property of the literate, ruling elite.31
From the eighth century until the very end of the last dynasty,
officeholders in the imperial bureaucracy were ideally drawn from a
pool of successful civil service examination (keju) candidates.32 The
chance of rising to the top levels of the official hierarchy, even for
those with the requisite connections and resources, was astronomically
tiny (Elman 1991:14).33 The examination system itself came to symbol-
ize upward mobility because it opened the theoretical possibility that
any peasant’s son could rise to become prime minister of the empire.
In practice, however, the sons of non-elite families were generally
screened out at the beginning of the process (Hartwell 1982:413 pas-
sim). Far from being a conduit of social mobility, the system, based on
Confucian learning and controlled by the literati-dominated bureau-
cracy, served to keep out the riffraff and to shore up the social and
political position of already elite families (Hartwell 1982:419).
To join this small, exclusive club of imperial licentiates was a dif-
ficult and unimaginably prestigious achievement, even for those who
enjoyed the advantage of being born into high-placed families. And the
system naturally gave rise to a mythology of its own. Among the most
celebrated creations of the Chinese literary imagination was the caizi,
a talented young scholar with a bright future ahead of him. Youthful
but refined, he embodied the sexual as well cultural and social desires
of elite men.34 In premodern Chinese literature, the heroes in tales
of romance and the leading men in erotic novels were almost always
caizi or men with caizi-like attributes. Thus for centuries, two comple-
mentary yet competing representations of ideal masculinity have dom-
inated Chinese popular culture: the “amorous scholar” ( fengliu caizi)
and the “valorous hero” (yingxiong haohan) .35
Wen Masculinity and Wu Masculinity

The caizi, as noted, was a man of cultivation and literary talent. As a
member of the scholarly gentry, he already enjoyed a certain measure
of social status. On the threshold of a promising career, he was as yet
unburdened with the responsibilities of official life. Being a man of
literary talent, refined tastes, emotional sensitivity, and extraordinary
good looks, he was, at least in the literary fantasies of the literati them-
selves, the kind of male that beautiful and talented women ( jiaren)
could not resist.36 Beyond fantasy, that China’s classical version of the
debonair playboy should be a studious man of letters contains a grain
of sociological truth: Chinese society has generally been hypergamous
(that is, women generally marry men of higher status). As a member of
the gentry, a young scholar would enjoy a relatively high social status.
A particularly talented scholar with a promising future—that is, pros-
pects for an official career—represents the potential for wealth as well.
In theory, he would be the most eligible of bachelors in late imperial

Chinese society.37
The haohan personifies a non-elite, and in some ways anti-elitist,
Chinese typology of masculine traits. Haohan are physical and violent,
usually adept at the fighting arts. They epitomize the martial and are
sharply contrasted with the caizi. In the abstract, their moral attributes
give no indication of controversy. To the contrary, they exemplify mas-
culine qualities universally prized in Chinese culture such as loyalty
(zhong) , sincerity (xin) , and righteousness (yi) . Moreover, though blunt
and straightforward, they understand and observe social decorum (li)
as much as any Confucian gentleman. After Wu Song shows himself
to be a “righteous, courageous, real man” (yiyong zhen nanzi) by spar-
ing Sun Erniang and Zhang Qing, for example, he addresses Sun as
“elder sister-in-law” (saozi) and politely asks her not to take offense at
his impropriety (Water Margin, chapter 27).
From a Confucian perspective, the haohan’s version of “righteous-
ness” is heterodox and potentially seditious. Associating righteous-
ness with wu (that is, forceful, direct action)—violence and the arts
of violence—opens the way to the subversion of state authority. Vigi-
lante justice—robbing the rich to save the poor, defying law in the
name of “carrying out Heaven’s will” (ti tian xing dao) — infringes on
the exclusive moral and political claims of the imperial order and
its self-appointed representatives (that is, the Confucian scholar-offi-
cials). Tales of haohan decouple the principle of transcendental justice
(zhengyi) from the political status quo. The heroes of the Water Margin,
for example, are victims of corruption and injustice who rebel against
the state apparatus dominated by Gao Jiu, an evil minister. They are
not trying to topple the emperor; he is the true Son of Heaven but
has been duped by Gao Jiu.38 Equally offensive to literati sensibilities
is the linking of righteousness and transcendent justice with physically
arduous, violent bodily practices that were mostly the province of (and
qualification of ) commoners.39
Here, too, rests one of the clearest contrasts between the caizi and
haohan versions of Chinese virility. The caizi is refined and intellec-
tual, but his relations with beautiful, talented women are romantic and
implicitly (and in some stories quite explicitly) sexual. The stereotypi-
cal righteous man of martial prowess, on the other hand, is thoroughly
physical in every way but the sexual. Where haohan have prodigious
appetites for meat, wine, and combat, they are abstemious and circum-
spect when it comes to relations with females. Yet haohan are hardly
asexual: to the contrary, resisting temptation entails the repression,

not the absence, of desire.
Once again, Wu Song provides a good case study. Between the tiger
slaying and meat-bun episodes, there is an equally famous sequence
(spanning chapters 23–26) that begins when Wu, based on his obvious
martial prowess, is hired as a constable by the magistrate of Yang’gu
County. By coincidence, Wu Song runs into his long-lost elder brother,
who is also living in the town and immediately insists that Wu Song
move in with him. At his brother’s house, he meets his new sister-in-
law, Golden Lotus (Pan Jinlian). Lotus is a hard-luck beauty, a former
indentured servant sold by her jealous mistress to the squat, ugly, and
simple Wu the Elder. When Golden Lotus takes a liking to the tall,
rugged, and good-looking Wu Song, he angrily rejects her advances
and moves back to the barracks. Later, when Wu Song leaves town
on an official errand, Golden Lotus takes up with the rich (caizi -like)
dandy Ximen Qing. The adulterous pair poisons Wu the Elder after
he catches them in flagrante delicto (Wu’s inability to stand up for
himself highlights the ongoing implication that he is socially as well as
sexually impotent). When Wu Song returns, he discovers the treach-
ery. Ximen Qing has bribed the magistrate and his underlings, and
Wu’s only recourse is to enforce private justice. He avenges his elder
brother’s murder first by deftly disemboweling Golden Lotus, tearing
out her heart, liver, and entrails with his bare hands and offering them
as a propitiatory blood sacrifice on his brother’s funerary altar.
The extreme violence of this gruesome ritual reveals an impor-
tant feature of Chinese wu masculinity: in the first place, the evident
misogyny (here an endemic distrust of women, rather than “hatred,”
as the term might imply) of wu masculinity is seen to arise from wom-
en’s inherently destabilizing role in the exogamous, virilocal, patrilin-
eal Chinese family. The requirement that women marry outside their
father’s lineage underlies the alienation and derogation of women
in Chinese society. In the extreme, as Pan Jinlian herself exemplifies,
women can be treated as articles of trade, with little or no power over
the terms or outcome of the transaction. The jianghu, both in its fic-
tional representations and social reality, is predominantly a culture of
males organized into brotherhoods, in turn modeled on the Chinese
system of patrilineal kinship. The bonds between such fictive broth-
ers are no less or more secure than those between blood brothers.
There are, of course, critical differences between real households and
jianghu life, where the presence of females is often transitory. Never-
theless, the patrilineally organized household and the male-dominated

jianghu share the tendency to regard women as both highly valuable,
hard-to-obtain commodities and (for that very reason) threats to male
solidarity—the organizing principle and basis for the identity of the
institution itself.
Thus the denigration, subordination, and exploitation of females;
the ethos of physical aggressivity; and the tenuous and competitive
relationship between “brothers” are shared attributes of mainstream
patrilineal families and the marginal, male-dominated jianghu. The
intergenerational tension between fathers and sons is also recapitu-
lated in the essentially “orphaned” structure of brotherhoods and in
a collective disdain for official authority. The performative aspects
of wu masculinity are, furthermore, primarily reserved for the soci-
ety of men—that is, life outside the domestic sphere. In this public,
performative aspect, as well as in its generally suffused but occasion-
ally explosive and dramatic displays of violence, wu masculinity bears
a strong resemblance to styles of manhood common in other large-
scale, strongly patriarchal societies.
Discursive Formations
The anthropological study of masculinity has been based mostly on
studies of Mediterranean, Latin American, Melanesian, and Sub-Saha-
ran African cultures.40 Conversely, much of the recent scholarship on
gender and gender identity in the contemporary Chinese-speaking
world has focused on women.41 A great deal of the anthropological
literature on Chinese society published before the early 1970s focused
on the lifeworlds of Chinese men, but that focus was not always by
design. Due to the structuralist emphasis on formal organization,
much of this work consisted in extensive studies of household, village,
and lineage, yielding a rather selective account of male-dominated
institutions and little about the lived experience of men or women in
Chinese society. Male identity was generally equated with a sequence
of social roles defined by the patriline itself—that is, boys became men
when they married and attained full male social status only when they
fathered sons.
This assessment first of all somewhat overstates the degree to which
Chinese men’s lives are defined by their roles in family, lineage, and
village and constrained by the formal rules of kinship. The logic of con-
ventional expectations, moreover, tells us little about the inner, lived
experience of men in Chinese society, nor can it sufficiently account
for actual forms of everyday behavior. Anthropologists working in Tai-

wan and, more recently, mainland China have interrogated the unex-
amined assumptions of earlier studies and highlighted inherent issues
of power, identity, and ideology. The focus has shifted from institu-
tions understood as reified entities to social and cultural practice and
the ways that practice produces those very institutions. To date, how-
ever, some of the most interesting scholarship on Chinese masculinity
has come from outside the field of anthropology.
In his 2002 book Theorizing Chinese Masculinity, Kam Louie identi-
fies wen and wu as signifying distinct but complementary ideals of Chi-
nese masculinity. The most original passages come in the penultimate
chapter (Louie 2002:141–159), where he argues that actors Bruce Lee
and Chow Yun-fat are internationalized, updated versions of the classi-
cal Chinese wu masculine ideal. According to Louie, Chow, and Bruce
Lee before him, showed the rest of the world that Chinese men can,
contrary to stereotype, be manly, cool, and tough.42
China’s national humiliation at the hands of the Western powers
and Japan in the nineteenth and early twentieth centuries is often
described as a “crisis” of Chinese masculinity. The West’s domination
of China was conceived sexually, implying a collective emasculation.
Louie claims that the global popularity of Hong Kong cinema has alle-
viated, if not completely balanced, the asymmetry. At the very least,
Chinese men no longer need to emulate Western ideals (and idols)
to reclaim their masculinity. They can instead revel in the knowledge
that their own heroes are admired and emulated in the West as well.
The globalization of popular culture markets, according to Louie, has
proved to be more effective in restoring Chinese male self-esteem than
ideology, economic power, or military posturing.
There is irony as well as overstatement in Louie’s position. First of
all, representations of feminized or violently deviant Chinese men per-
sist (an example of the latter is Jet Li’s Wah Sing Ku in Lethal Weapon
4 ). Second, the mechanism of “restoration” itself makes “the West”
the active agent. In this configuration, it is the (implicitly dominant)
West that endows (or withholds) masculinity, not Chinese men them-
selves. The futility of disempowerment is evident as well in the postwar
history of the Chinese-language culture industry. For decades, state-
run studios in the PRC and private companies in Taiwan and Hong
Kong produced untold numbers of films and television serials that
depicted manly, heroic Chinese standing up to the evil West and Japan
and recovering a bit of lost national face (wei zuren zheng yi kou qi) . A
more subtle message, however, lies in the knowledge that these are
self- delusions, mythic representations of frustrated collective desire.
Consolation, that is, entails defeat. Louie’s claim is thus more limited
and problematic than it appears, but that does not entirely obviate his
point. Bruce Lee, as he suggests, was a modern haohan who reshaped
the image of both Chinese and American masculinity. For a couple of
decades at least, Lee was an icon of popular culture, the very embodi-
ment of macho cool, idolized by young men worldwide; his apotheosis,
then, rings far truer than any fantasy of national revenge.
Xueping Zhong’s Masculinity Besieged? (2000) is an intellectual his-
tory and literary critique of angst and anxiety among China’s male
intelligentsia in the post-Mao era. In her analysis of the xungen (root-
seeking) literary movement, she describes these authors’ vivid, “epic”
depictions of an imagined past. Many of these works, she notes, are
mythic narratives about people—Han and minority—living on China’s
peripheries and wild frontiers. Dwelling close to nature, their lives are
full of healthy (but heavily androcentric) sexuality and vital energy.
These writers (and the “fifth generation” directors like Zhang
Yimou and Chen Kaige, whose early films were based on root-seek-
ing novellas) are not merely lamenting the loss of this world and its
attendant machismo; they are proposing a “new beginning” based on
its values (Zhong 2000:168). Zhong sees in this primitivist identifica-
tion of individual and collective masculinity with sexual potency a kind
of post-post-Mao nostalgia for a world without sexual angst or prob-
lematized gender roles. The overtly sexual haohan of these narratives
is certainly a departure from earlier representations. I would maintain,
however, that even here, masculinity is not strictly equivalent to sexual
virility but (illustrated most dramatically in Zhang Y imou’s Judou) is
still very much embedded in patrilineal production.
Wu, Embodiment, and Bodily Practice

It is fairly obvious what the significance of wu is when it is contrasted
with wen, as in the case of “the interrelated process of empire build-
ing” (Zito 1989:58). This contrast extends to the civil-military structure
of the Daoist pantheon, the oral narratives out of which most of the
above-mentioned vernacular novels were condensed—that is, temple
processions, opera performances, and puppet theater; the differentia-
tion into civil and military categories of traditional operas, acting roles,
and even theater architecture (B. Ward 1979:28); the standard prac-
tice of writers of official local gazetteers to report on locales according
to separate civil and military aspects of social and historical categories;

the division of imperial court ritual into civil and military segments
(Zito 1989:428 – 436); and verbal and kinetic expressions of contrast-
ing literary and martial character in spirit-medium séances and Daoist-
supervised rituals of transformation and renewal (Lagerwey 1987:88,
216 –237). All of these bring to life, usually in the context of a real
or imagined historical setting, the interaction of beings and forces of
the visible (youxing) and invisible (wuxing) realms. In turn, both realms
(which overlap but are not coterminous) are organized as civil and
military bureaucracies, each with its respective agents of order and
disorder.
The Chinese martial arts are fundamentally forms of ritualized vio-
lence.43 They originated and evolved as training and drilling exercises
for the professional military.44 These fighting techniques, in turn, are
the source of the postures, gestures, and implements associated with
the martial in every imaginable mode of representation in Chinese
culture. The postures, gestures, and spectacular acrobatic techniques
of martial opera (wuju) are extensions of the martial arts. Finally, the
carved and painted figures of martial deities, dramatic temple festival
performances, and the ritual exorcisms of popular sorcerers and Dao-
ist priests draw on and combine elements of both the “genuine” fight-
ing arts and the theatrical adaptations.
I began this project with a simple premise: Chinese popular reli-
gion and martial arts are historically and symbolically interlinked,
and their juncture is an important, perhaps defining, nexus of Chi-
nese popular culture. I hoped to discover, through ethnographic and
archival research, how and why this connection occurred (and occurs)
and what, if any, particular meaning or efficacy is being attributed to
particular kinetic notes and phrases (postures, movements, as well as
the tangible, visual, and aural elements such as shapes and colors of
weapons, etc.). To answer this series of historical and semiotic ques-
tions, I needed to gain more practical as well as discursive knowledge
of the martial arts and Chinese religion. I quickly discovered, however,
that this relationship, as well as the form and logic of each discrete
practice, rested on a broader, encompassing cultural logic and histori-
cal process. Other scholars of Chinese culture, including many who
have written about the practice and logic of Daoist and popular ritual,
have noted the martial element.45 Yet few if any extant studies reveal
more than a cursory knowledge of martial arts practice, technically or
otherwise.
Conversely, there is an enormous body of popular “kungfu-cen-

tric” documentary literature, written mostly by martial arts enthusiasts
for martial arts enthusiasts. Even the best of this genre (for example,
Frantzis 1998) leave unquestioned and unexamined the received val-
ues and myths of their own martial arts training. We may learn, for
example, that the author’s master belongs to the “true” lineage of
Yang-style taijiquan or Wing Chun kung fu but will find little cultural
insight into why Chinese martial artists are so obviously and endlessly
obsessed with establishing lineage (or what lineage may entail in this
context). On the other side, the scholarly literature suffers from a dif-
ferent sort of parochialism. Louie, for example, offers an erudite and
nuanced argument for the importance of the wen-wu dichotomy as
a structuring category in Chinese culture, but his research is based
entirely on historical and literary narratives, lacking any ethnographic
input. Ward and Stirling (1925) or, more recently, Chu (2000) present
detailed empirical accounts of Chinese gangs, brotherhoods, and
secret societies, yet they do not explore the broader cultural implica-
tions (nor, interestingly, do they describe martial arts practice or its
place in the social and professional milieu of Chinese gangland). Even
those (mostly historians and a few anthropologists) who have written
about secret societies, gangs, ritual violence, or peasant rebellion touch
lightly, if at all, on the subject of martial arts training and practice.
As I suggested in chapter 1, the interrelationship between Chinese
popular religion and martial arts is both obvious and complex. Martial
prowess has, on the one hand, been linked to the moral ideal of the
knight-errant since at least the Warring States period but, for a thou-
sand years or more, also with the supernatural abilities and fantastic
magic powers of Buddhist adepts and Daoist immortals. Popular cus-
toms such as lion dancing are deeply embedded with both the tradi-
tions of technical martial arts training and communal religious prac-
tice. The lion, it turns out, is a strongly yang creature with the power to
drive away or destroy ghosts and other forms of death pollution. The
lithe, forceful, and sometimes suggestive moves of the lion create an
aura of vigorous male energy that is amplified by the cascading explo-
sions of firecrackers that accompany the performance.
The martial arts appear in the ferocious, aggressive thrashings of
mediums possessed by demonic spirits, as well as in the disciplined
maneuvers of performance troupes like the Song Jiang Battalion, a
ritual “militia” that enacts the fighting prowess of the haohan rebels
of the Water Margin. Those who perform and enact these roles have
varying levels of martial arts expertise; some have merely learned the
moves for the purpose of participating, while some who already pos-
sess martial arts and fighting skill seem naturally drawn to the activity.
The techniques of martial performance troupes are unambiguously
derived from the martial arts. The lion dance and the Song Jiang
Battalion can be considered modes of martial arts training in them-
selves. A powerful lion dance or well-coordinated Song Jiang Battalion
drill is, to untrained and expert eyes alike, a convincing, sometimes
awe-inspiring demonstration of the performers’ potential to do real
damage.
The lore of such troupes recalls the incredible feats of the masters
who taught and trained the performers in the past, men known for
their martial skills but often for their physical and sexual prowess and
even miraculous powers as well. In the past, martial procession troupes
were associated with both local territorial temples and the village com-
munal associations that competed in periodic festivals, where ritual
violence sometimes led to brawls and armed feuds. Even today, during
major religious festivals in Taiwan and Hong Kong, it is not unusual for
fights to break out between lion dance teams, rival martial arts schools,
Team of possessed spirit mediums in attack pose (brandishing spike clubs) during a
procession. Note the tattoos on the back of the middle performer.
and feuding temple groups. The violence is made all the more intense
by the fact that many of these rival groups are composed of young
jianghu toughs, for whom the festival venue can be a proving ground
of sorts. But far from finding such violence incompatible with religious
purpose, most participants, organizers, and onlookers will recognize
that the efficacious powers of exorcistic spirits are only made more
manifest, more powerful, and more believable by these wild, violent
reminders of the marginality of the actors, who they really are, and
what they really do.
Life Stories, Ritual Narratives

There is a fine legal as well as sociological line between sworn broth-
erhoods, fraternal associations modeled on extended lineages (some
of which are self- described tang—that is, they meet and organize as
ritual lineages) and organized criminal factions (bangpai) .46 None of
my informants, even those who were active members of gangs of one
sort or another, could reasonably be described as hard-core criminals.
At the furthest extreme were a few informants who worked as enforc-
ers, debt collectors, messengers, and bodyguards for local gang bosses.
Yet the only individuals within my circle of informants who carried
firearms or forcibly restrained others against their will as part of their
vocation were actually working for the other side: police officers and
prison guards.
The gray area between mainstream, legal work and banned, sanc-
tioned, or criminal activity marked the boundaries of my informants’
social world as well as their economic prospects. At the same time, per-
haps making a virtue out of necessity, these men and women claimed
this very world as their own; they saw their own marginality as a core
attribute of their personal and social identities.
Their stories were full of risk, danger, violence, and transgression.
Curiously, though perhaps not surprisingly, many narratives of and
about China’s 1980s “craze” of “leaping into the sea” (xiahai) make
use of similar tropes and allusions. Around this time, xiahai became a
popular term among Chinese intellectuals and party officials who were
“taking the plunge” into the newly opened world of private business.
Those who cast off into the stormy seas of capitalism were giving up the
security and prestige of an official career. The odds were against them,
but the potential rewards were enormous. But the decision to engage
in entrepreneurial activity was regarded by many, especially intellectu-
als who traditionally defined their calling as pure and above the desire
for profit, as inherently self-debasing. Emerging from decades of col-

lectivization, central planning, and Maoist ideology, some among that
first wave of cadres and officials who decided to pursue personal profit
must have struggled with their consciences as well. For them, “leaping
into the sea” would have implied many of the same moral and personal
dilemmas that “embarking on the dark path” has long had for work-
ing-class men and women in Taiwan.
It is perhaps also significant that as entrepreneurship has moved
into the mainstream in the post-socialist PRC, xiahai has gradually
shifted in meaning as well and is now used more often to describe
women who “fall” into prostitution.47 Both “leaping into the sea” and
“entering the [dark] way” (chu dao) imply that one is somehow choos-
ing a morally lesser path. One is thus forced to accept a sense of oneself
as forever tainted. Indeed, when reflecting on life on the dark path,
people often speak in tones of regret. They describe an anguished
choice made in the face of extreme difficulty or sigh with resignation
about what they often see as an inescapable, predestined fate.
In the following pages, I explore the lives (and lifeworlds) of ordi-
nary people living on the margins of Chinese and Taiwanese society in
the late twentieth century.
C hapter 3
Taidong
The Mountains and Beyond
The renovation of a local temple in late imperial China was an event

that brought officials, local gentry, and commoners together in com-
mon cause. Stone steles, inscribed with a testimonial written by a well-
placed patron, were often erected to commemorate temple renova-
tions. The act affirmed the position of the patron, but his lavishing of
attention and resources enhanced the prestige of the temple as well.
The texts of these inscriptions often preserve details of local history
otherwise unaccounted for in official accounts. Yet the texts tend to be
formulaic, couched in clichés of elite officialdom. There are oblique
references to the local and different but few explicit descriptions of
things that might be regarded as backward or deviant and thus reflect
poorly on the author himself.
Villagers, to the extent that they could read and understand the
inscriptions, probably had little objection to this editing out of local
knowledge. The stele was a talisman of imperial authority and auto-
matically conferred prestige, in proportion to the rank and reputa-
tion of the sponsor. It established the locale as a venerable part of the
empire, no matter how remote, and, by extension, gave it a place of
importance in the cosmic imperium. Such validation could have prac-
tical as well as symbolic advantages, and by the late imperial period
village temples and the voluntary associations connected with them
became the focus of efforts by rural communities to join higher-order
power networks and to be favorably represented in official circles
(Duara 1988a; Watson 1985).
Today, temples throughout Taiwan and China feature similar
commemorative inscriptions, most of which imitate the classical style
and rhetoric of those earlier texts.1 At the same time, there has been
increasing interest over the last two decades in local practice, espe-
cially the unusual and exotic. In Taiwan, and to a limited extent also in
mainland China, local customs once considered deviant or backward
58
the mountains and beyond 59
are increasingly celebrated and even advertised. Difang rentong, “local

identity,” has become a familiar term, appearing not only in these lat-
ter-day stone memorial texts, but also in scholarly articles, newspaper
editorials, political speeches, and everyday conversation.
Quaint folksiness has become a mark of the culturally authentic,
turning traditional artifacts and practices, including the rituals and
spaces of popular religion, into marketable commodities. Ultimately,
however, as with the commemorative steles of the late imperial period,
it is the voice of the elites that claims to speak for a community. The
“discovery” of local culture has not, therefore, led to the unqualified
empowerment of local communities or recognition of marginal voices.
Instead, in both Taiwan and China, the power of representation and
means of cultural production remain, as before, arenas of contention
between local/marginal and elite/national interests. There is more
to this contest than pride and face: there is considerable political and
economic capital at stake as well.
The asymmetry of the contest is greatest, and practical ramifica-
tions are most evident, in places where ethnic differences define the
cultural landscape. Thus in the regions that once marked the empire’s
periphery, from the western desert to the southwestern highlands to
the coastal mountains of Taiwan, ethnicity adds an additional layer of
complexity to the struggle. Here, the long and often vexed relation-
ship between the Han and the “barbarian” peoples on the empire’s
periphery continues to evolve. In government policies and the mass
media (for example, in advertising for the tourist industries that now
dominate the development strategies in both Taidong and Dali), the
interests of the center, the dominant Han majority, are clearly ascen-
dant (Schein 2000; Notar 2006).
As with the rebuilding of temples, there are converging as well
as competing interests at stake in this process. Given endemic dispari
ties in access to resources and control of discourse, exploitation seems
inevitable. But whether in the imperial era or the modern state, Repub-
lican or Communist, mainland or Taiwan, local-state engagement has
often taken the form of accommodation, even across ethnic boundar-
ies. Strategic accommodation is evident, for example, in the imperial
practice of recognizing and awarding titles to local deities. Accepting
imperial recognition, marginal communities and deity cults gained
security and prestige but sacrificed autonomy and, most likely, the
power to represent and interpret their own history and practice.
Reciprocal if not equal, this relationship reveals, in another form,
the mystified dynamic of imperial (and, generally, state) power in China:

in its political organization and ritual process, every community, even
on the most distant periphery, was a microcosm of the greater imperial
order. This structure of “embedded hierarchy” helps us understand,
in part, how the Chinese imperial order, “all under Heaven,” could
transcend not only the vast distances and cultural diversity of its popu-
lation, but even political fragmentation and dynastic continuity.
In the abstract, then, remote outposts were as much constituents
of the empire as central cities.2 The state spatially encompassed the
village, yet the village was also a microcosm of the state. The view from
the center mirrored the view from the periphery: the world as a set
of nested hierarchies, the higher-order structures encompassing (as
yang) those of lower orders (yin) (Sangren 1987:14 –16, 160). In the
idiom of popular religion, local gods are both proxies of the higher
order and masters of their own domain. Every community, even the
most remote, then, reproduces this hegemonic Chinese cultural order.
Frontier outposts are furthest from the geographical center of empire,
yet in rehearsing the patterns of civilization (wen) on the frontier, they
recreate the mythical process of order emerging from chaos through
the agency of military force (wu). Despite the propensity to exhibit
cultural deviance, harbor the dregs of society, and even openly rebel,
such places are, in this sense, paradigmatically Chinese. This paradox
is revealed, as I will show, in an ambivalence that characterizes the
sense of local identity in such places.3
Chinese popular religion is premised upon the possibility of access
to magical power, or ling. The object of devotional worship and ritual
practice is to enter into a reciprocal transaction with powerful spirits,
with the hope that they will intervene on one’s behalf; efficacy is mea-
sured by the (im)probability of outcomes (that is, a highly improb-
able outcome indicates a miraculous intervention). The field of action
encompasses every area of human life, and one meets devotees from
every occupation, background, age, gender, and social class. At the
same time, for subjects otherwise excluded from or subordinated
within the structures of institutional power, the prospect of direct
access to magical power holds an even greater attraction. It is not coin-
cidental, then, that the centers of popular religious devotion and prac-
tice are frequently to be found at the territorial and social margins of
the extended community.
The two locales where I conducted most of the research for this
book were, until quite recently, both considered to be at or even
beyond the margin of Chinese civilization. Remote, mountainous, and

ethnically diverse, Taidong and Dali have also both staked their futures
on exotic local cultures and (relatively) unspoiled natural beauty—the
very attributes that once ensured social, political, and economic exclu-
sion. In the remainder of this chapter, I examine the role that place
plays in the formation of local religious belief, ritual practice, and
collective identity. In particular, I consider the attributes of remote
peripheries as they are perceived and described in historical and con-
temporary narratives. I begin in this chapter with a historical ethnogra-
phy of Taidong. Chapter 4 includes a somewhat less detailed account
of Dali.
The Numinous Land

In Chinese cosmology, the landscape is imagined as a living body, and
the body, conversely, a landscape. Following patterned oscillations of
yin and yang, bodies and landscapes interpenetrate, exchanging and
channeling qi, reciprocally generating and transforming one another.4
Individual persons, but also the collective bodies (structures) of Chi-
nese social experience such as patrilines and households, develop in
dynamic juxtaposition to the landscape. The waxing and waning of
life-energies and social fortunes thus articulate a complex interaction
of correlated organic systems. Applications of this etiology include tra-
ditional Chinese medicine, fengshui, and the ritual practice of popular
religion (which consistently references both traditional medicine and
fengshui ).5
The dynamic resonance among landscape, social body, and indi-
vidual bodies, often called “correlative cosmology,” thus forms the the-
oretical basis for sociality. In Chinese terms, that is, human collectivity
is not only based on shared understanding, but also on shared sub-
stance. Social reciprocity and human “feelings” (qing) are not moral
abstractions, but rather the expression of this intercorporeality.
In Taidong, where geography has so directly constrained human
activity, almost every narrative of self and place alludes to the land-
scape—the harshness and remoteness but also the power and beauty
of the terrain. So I begin my account of Taidong with a rough outline
of its geography and human history.
The Unnamed Mountains

Austronesian-speaking people had colonized Taiwan by about 3200
BCE (Chang and Goodenough 1996:38), and for most of the next five
thousand years, they had the island mostly to themselves. Despite Tai
wan’s size and proximity to southeastern China (eighty miles to the
west), the Ryukyu Archipelago (seventy miles to the east), and the
Philippines (ninety-five miles to the south), the island remained essen-
tially invisible to the outside world until the sixteenth century. Even
more surprising, perhaps, given subsequent history, is that the Chi-
nese seemed uncertain, at best, of Taiwan’s existence until late in the
Ming period.6 Anecdotal evidence suggests that Chinese fishermen
and pirate traders probably began making short forays ashore here
sometime during the 1500s, bartering with the aborigines for supplies
and deer hides (Thomson 1968; Hsu 1980a:8–10; Shepherd 1993:8).
Little is known of the earliest sojourners, but when the Dutch arrived
in 1624, they found a small but stable Chinese presence in the form of
“a network of Chinese village traders, rumored to number from 1,000
to 1,500” (Shepherd 1993: 83).
Taiwan changed hands three times during the seventeenth century,
starting with the Dutch (there was also a brief Spanish colonial pres-
ence in the north, near modern Tamsui). The Dutch were expelled
in 1661 by Ming loyalist Zheng Cheng’gong. Zheng died soon after
the conquest, but the regime held out until 1683, when the island
fell to an imperial Qing force led by Admiral Shi Lang, who had once
been an officer in Zheng’s navy. From 1683 until 1895, when the Qing
ceded Taiwan to Japan, Han Chinese migrants steadily expanded and
settled the Taiwan frontier. Despite treacherous ocean currents, unpre-
dictable shifts in immigration policy, and endemic violence in Taiwan
itself, by the nineteenth century these colonists and their island-born
descendents had transformed the western plains and foothills into a
recognizably Chinese landscape of cities, towns, hamlets, and farms
(Shepherd 1993:137ff.; Knapp 1980:56ff.). By 1905, Taiwan’s popu-
lation had grown to over 3 million, the vast majority descendents of
Qing-era migrants from southern Fujian and northern Guangdong.7
According to the 1905 survey, by the turn of the twentieth century the
aborigines comprised less than 3 percent of the total.8
Yet throughout Qing rule, despite their superior numbers and
demonstrated aptitude for risk, Han colonists were effectively prohib-
ited from moving into or across the mountains, limiting them to a
narrow strip of lowlands along the island’s western shore. On official
Qing maps published before 1875, the eastern two-thirds of Taiwan
appears as an undifferentiated blank space, labeled houshan, literally
“the rearward mountains” or, somewhat more poetically, “the moun-
tains beyond.” 9 For most colonists, “the mountains beyond” must have
seemed a reasonable designation.10 From 1722 to 1875, all ethnic Han
were barred from crossing a boundary that zigzagged among the west-
ern foothills (Shepherd 1993:186). In Chinese lore, moreover, “moun-
tains” (shan) are places untouched by culture: they are wild, mysterious
spaces inhabited by wild animals and spirits of the unworshipped dead.
Taiwan’s mountains were thus out of bounds both legally and symboli-
cally. What lay yonder, on the other side, was an even greater mystery.
Houshan thus came to mean the unseen, unknown lands beyond the
high peaks.11 The descriptions conjured in official documents and
extant reports of ordinary people’s views suggest that in late Qing Tai-
wan, houshan was imagined not as a hidden paradise,12 but a primitive
and inhospitable wild space far beyond the pale.13
Taming the Wild Frontier

Taidong lies approximately 160 miles south-southeast of Taipei, just
below the Tropic of Cancer on Taiwan’s Pacific coast. To reach the
coast, however, the proverbial crow would have to climb to 13,000 feet
to cross the crest of the Central Mountains. For centuries, these steep,
rugged mountains, which run north-south down the entire length of
the island, marked the limit of Chinese dominion and the boundary
of colonial settlement. The mountains and east coast were eventually
declared part of the Qing empire (in 1875), but government author-
ity in the area was never more than tenuous and extended no further
than a handful of administrative and military outposts in the valleys
and coastal precincts. The inner mountains remained entirely beyond
Qing civil or military control. In 1895, on the eve of the Japanese take-
over, Taidong was still very much a frontier.
The history and culture of southeastern Taiwan are inseparable
from its dramatic, dynamic geography, dominated by the Central
Mountains, the Coast Range, and the Eastern Longitudinal Valley,
which divides the two mountain chains. The Longitudinal Valley itself
is an active strike-slip fault marking the transform boundary between
the Philippine and Eurasian plates.14
At its northern and southern extents, the valley opens out into
broad alluvial coastal plains. Hualian, at the northern terminus of the
Coast Range, is the east coast’s main port. Taidong City, ninety miles
south of Hualian, is the administrative and commercial hub of the
southern valley and extreme southeast coast. Rising from the estuary of
the Beinan Stream near the foot of Mt. Dulan, the southernmost peak
Chenggong
Chishang
E
TAIWAN
NG
RA
Guanshan
ST
Y
LLE
COA
Map
N
L VA
Area Luye
I
INA
A
ITUD
T
LONG
N
N
EASTERN
U
O
Taidong City
Beinan
M
L
A
R
PACIFIC OCEAN
(PHILIPPINE SEA)
T
Taimali
N
E
C
Mountainous area
0 5 10 mi
Dawu
0 5 10 15 km
Map 3.1. Taidong

of the Coast Range, the town stretches southward along the Pacific
shore (here known as the Philippine Sea on nautical charts) and up
across the gradual incline of the Beinan Plain. To the west, past a lushly
green belt of rice fields, orchards, betel nut groves, and small hamlets,
the front range of the Central Mountains rises precipitously, ascending
along terraces and sharp ridges toward rocky peaks that reach eleva-
tions over eleven thousand feet just a few miles inland. Moving fur-
ther south past the Zhiben hot springs resort and the outpost town
of Taimali, the rugged coastline, reminiscent of California’s Big Sur,
winds down another sixty-five miles to the coral reefs at Eluanbi.
Landscape and Identity

Taidong has often been cast as a remote hinterland, and the land and
people alike have been described as uncultivated, backward, and primi-
tive. In the summer of 1999, I conducted an informal, not quite random
survey among about thirty natives of Taipei and Kaohsiung, Taiwan’s
two largest cities. Those who had been to Taidong, as well as those who
knew it only from hearsay and imagination, all narrated remarkably
similar impressions. Almost everyone interviewed described Taidong as
“backward” and “primitive” but also “unspoiled” and “folksy.” Asked to
describe Taidong people, they mentioned “uneducated,” “naive,” and
“gullible” along with “unpretentious,” “friendly,” and “hospitable.”
In Taidong, local self-description often echoes these same stereo-
typed impressions.15 In their more optimistic moments, however, Tai
dong natives will strike a more upbeat note. After a bit of grumbling
about the inconveniences and drawbacks of being from such a poor,
backward, and faraway place, they point out that Taidong is a place of
“bright mountains and beautiful waters” (shanming shuixiu) , being a
“place of numinous power and exceptional, heroic people” (diling ren-
jie) . Such expressions invoke the principles of fengshui and traditional
Chinese medicine, based on the correlative, organismic paradigm of
traditional Chinese cosmology (see, for instance, Dikotter 1992:22).
In this view, the character of the people is inseparable from the land
that nurtures them. They can be uncultured and rough because Tai
dong is remote and far from civilization’s core, or they can be heroic
and uncorrupted, infused with the energy and spirit—the qi — of the
imposing mountains and pure waters that surround them.
Representations of Taidong as backward, primitive, and mysteri-
ous have had more than mere symbolic or psychological effect on the
area and its inhabitants. In 1637, the Dutch sent two military expedi-
tions to the Beinan area to search for a rumored mountain of gold; 16

centuries later, the Japanese sought to take advantage of the area’s
special topography and climate to acclimate settlers from the home
islands, a staging ground for the ill-fated “Southward March” to the
tropics.17 In the 1950s and 1960s, a canned pineapple export boom
produced a short-lived surge of new settlement, prosperity, and opti-
mism. Then, as suppliers turned to growers and processors closer to
market—in particular Hawai‘i—the boom fizzled, dreams of sudden
wealth evaporated, and many of the new settlers moved, disillusioned,
back across the mountains. The industries that fueled Taiwan’s eco-
nomic miracle never made it to Taidong, and the region faded back
to the periphery.
Despite Japanese-built infrastructure projects and Taiwan’s post-
war economic “miracle,” Taidong remained rural and isolated for most
of the twentieth century. The rugged landscape and sparse population
discouraged investment in manufacturing or processing, mainstays of
the export economy. During the years of martial law and the “White
Terror” (1947–1987), Taidong was known as much for penal colonies
as pineapples. As Taiwan entered a post-industrial, post-martial-law era
in the early 1990s, however, a newly affluent and growing urban middle
class set out in search of fresh air, cultural authenticity, and alternative
styles of recreation. Taidong, preserved by its remoteness, was made
“consumable.” Once “remote” and “backward,” Taidong has become
“exotic” and “unspoiled.” Since the 1990s, private partnerships and
corporate consortia have been developing resorts and building hotels,
hoping to cash in on this expanding (but still capricious) domestic
leisure market.
These new, post-industrial markets rely on the very same percep-
tions that excluded Taidong from the industrial economy. The attri-
butes—wild, tough, primitive, backward, remote—remain the same,
but their meaning and value have been subtly altered. Nearly everyone
I talked to in Taidong in the late 1980s through the mid-1990s, from
county officials to high school teachers to shop owners, farmers, and
even local gangsters invariably invoked the old, negative stereotypes.
In recent years, however, the tone has changed. Taidong’s natural set-
ting is now its greatest asset, its mountains green, its waters lovely; it
is unpolluted rather than lacking industry; uncrowded rather than
empty; relaxed, not stagnant; authentic, not unsophisticated; and so
on. Trading on its own alterity has become, if not the most desirable
strategy, at least a potential source of both identity and income.
The Frontier Era

Communication between Taidong and the rest of the island has always
been difficult. Yet despite its isolation, Taidong is far from inacces-
sible. The archeological record shows that people have been regu-
larly coming to, settling in, and transiting the area by land and sea
for thousands of years.18 The earliest evidence of human activity in
Taiwan (ca. 30,000 BP) was discovered here, and a series of Neolithic
peoples began settling the Beinan Plain around 3500 BCE. Over the
last two millennia, Austronesian-speaking peoples have spread out
across Taiwan, including the mountains and plains of the east coast.
Dutch reports from 1636 describe a well-developed network of aborig-
ine settlements throughout the Taidong area. It is significant that at
a time when Chinese sojourners were already well established in the
aborigine villages of the western plain, the Dutch found no Chinese
presence east of the mountains (Shepherd 1993:39 – 40; Kaneko and
Tsuchida 1982:109–111).
By the end of the nineteenth century, migrants from Fujian and
Guangdong had transformed western Taiwan into a thoroughly Chi-
nese landscape (Knapp 1980). Rice exports were the mainstay of
the island’s economy, but after 1850 worldwide demand for tea and
camphor created a strong incentive to open up the mountains.19 The
imperial court, however, showed little interest in an area it regarded as
“raw barbarian” territory. The mountains marked the edge of civiliza-
tion and thus the empire’s periphery (cf. Harrell 2000:135, diagram).
That all changed, however, when news of the Japanese punitive expe-
dition to southern Taiwan (the so-called “Mudan Incident”) reached
Beijing in the spring of 1874.20 In 1875 the Qing court finally claimed
sovereignty over the mountains and Pacific coast of Taiwan, setting
in motion a chain of events that effectively ended the area’s isolation
from the rest of the island. Ministers then quickly devised a joint civil
and military plan to “civilize” the mountain territories. They would
first open new routes across the mountains, pacify the aborigines, and
finally open to Han cultivators the “wild, undeveloped” aborigine land
of the coastal region.
From Frontier to Colony to Backwater

Early in 1875, General Wu Guangliang and his Flying Tiger brigade
were assigned the responsibility of being the first official Chinese
military expedition to southeastern Taiwan. Starting east from the
Puli Basin in central Taiwan, the expedition cleared a path across the
mountains, reaching the middle Rift Valley near the end of the year.
The following year (1876) the entire east coast from Yilan south to
modern Daren was designated as the Beinan Commandery (Beinan
ting), and a branch of the Bureau of Pacification and Reclamation
(Fukenju) was established to “civilize” (that is, sinify) the aborigines
and open the way for Han settlement.
Until this time only a few thousand Han Chinese—among them
enterprising traders and farmers and desperate fugitives—had crossed
the boundary and tracked up the narrow southeastern littoral or tra-
versed the high mountains into the eastern regions. From the Chi-
nese agrarian perspective, there was plenty of “unused” land but too
few hands to “reclaim” it. The Qing authorities thus began to actively
promote emigration. They offered assistance and subsidies to land-
less peasants from western Taiwan and the Chinese mainland, promis-
ing them land, seed stock, and basic farm implements. Most of these
efforts, however, were ill conceived, poorly executed, and abortive.21
Just getting to Taidong was a tricky and dangerous venture. In the
rainy seasons even the easiest route, the relatively low-altitude trek
across the southern peninsula and up the coast, was frequently impass-
able. Encounters with head-hunting parties were a threat on any of
the high mountain tracks. The approach by sea was no more ideal
than the land routes. Tricky currents and typhoons made the passage
treacherous, but once vessels were off the eastern shore, the steep,
straight, rocky coast made landing difficult. Junks could anchor a few
hundred yards off shore and, assuming calm winds and a gentle swell,
transfer cargo and passengers to smaller, beach-landing craft. But they
had to get under way again quickly, unable to linger unsheltered for
long. It is thus not surprising that so few, including the landless and
desperate, thought of venturing over to the “mountains beyond,” even
under the official sanction and promised protection of the imperial
authorities. Indeed, Chinese officials and European adventurers who
reached the area in the late nineteenth century found only a hand-
ful of sojourning Chinese petty merchants and a few widely scattered
peasant hamlets.22
Despite such perceptions and practical obstacles, Han Chinese set-
tlement began to take root during the latter half of the nineteenth cen-
tury, mostly in mixed communities of Chinese and sinified aborigine
families.23 Dazhuang (in the area of modern Fuli) was one of the larg-
est of these communities. It was established by Kavalan plains aborigi-
nes who had migrated over the mountains during the 1850s; by 1885
several Han (mostly Hakka men) had joined the community. In 1887,
this village was the flashpoint of a violent local uprising (the so-called
“Dazhuang Incident”) that led to the destruction of the nearby Shui-
wei Commandery and the yamen at Beinan. Soon after, the Qing mili-
tary mission was restructured and expanded, yet the number and qual-
ity of imperial troops was never sufficient to consolidate and maintain
control of the area.24 The area of Han settlement actually contracted
following the Dazhuang Incident, as understandably nervous Chinese
colonists either returned west over the mountains or relocated closer
to existing military garrisons (Meng 2002).
The soldiers billeted at Beinan were mostly recruits culled from
the lowest grades of the Qing army and were constantly hobbled by
morale problems. Opium addiction was rife, and tropical diseases took
a heavy toll.25 It is surprising to note, then, that when China ceded
dominion of Taiwan to Japan under the Treaty of Shimonoseki in 1895,
the outnumbered and outgunned Beinan garrison took action against
the arriving Japanese expeditionary forces. This turned out to be the
last organized Qing military action against the Japanese occupation.
The garrison’s bravery was no match for the superior Japanese forces,
and early in 1896 the remnants of the garrison scattered westward in a
disorderly retreat across the mountains. In the end, despite these last-
minute heroics, from beginning to end the Qing presence in Taidong
can only be described as tenuous, violent, and rather brief.
A Model Tropical Colony: Taidong under Japanese Administration

In marked contrast to the ambivalent Qing regime, the Japanese recog-
nized the economic potential of the east coast and immediately started
work on a series of infrastructure, land improvement, and settlement
projects. The first stages in the Japanese plan were, quite logically, to
develop and exploit the east coast’s natural resources, improve land
and sea access, and make the area part of Japan’s grand “civilizing”
project (among the first structures built by the Japanese in Taidong
were a hospital and a post and telegraph office, both completed in
1896). By 1905 thousands of acres of swamp and marsh in the middle
Beinan Plain had been drained and transformed into sugarcane plan-
tations.26 More than a dozen Japanese immigrant settlements were
planned, although only two, Asahimura (Xucun) and Shikano (Luye),
made it off the drawing board.
While the colonizers worked rapidly to “civilize” the east coast,
such early projects had only a limited impact on the region. The most
dramatic and effective Japanese development effort was undoubt-
edly the construction of a modern transportation system in the East-
ern Longitudinal Valley and, somewhat later, its integration into the
island-wide network. A road across the high mountains, linking Tai
dong with Pingdong directly to the west, was completed in 1928, and
another, more easily maintained highway (still in use and now known
as the Nanhui Gonglu, or Southern Link Highway) was opened in
1939 (Knapp 1980:186 –195 passim).27 A narrow-gauge rail line serv-
ing the southern valley from Beinan to Guanshan began operation in
1919 and was later connected to the northern valley branch, creating
the Taitō Railroad that linked Taidong and Hualian by 1926.28 Port
facilities at Shinkō (Xin’gang, now Chenggong) were also expanded
in 1932, and the Taitō airfield (now Taidong Fengnin Airport) was
opened in 1938, initially serving as a military transport facility.29
Such improvements in communications, security, and general liv-
ing conditions made Taidong a better-known, more appealing desti-
nation, if not quite a magnet for new settlement. From about 1900,
peasants, laborers, and others began migrating to Taidong in small but
steadily increasing numbers from Tainan, Kaohsiung, and Pingdong.
Yet even this trend was small and relatively short-lived. As Taiwan’s eco-
nomic “miracle” hit full stride in the early 1970s, prospects for better
work, education, and living conditions began luring Taidong’s work-
ing-age men and women back across the mountains to Taipei and
other major cities. Peaking at 250,000 less than a hundred years after
the town had become part of the Qing empire, Taidong’s population
has gone into slow, steady decline.
Because the area lacked modern roads—until the beginning of
the Pacific War, the Japanese transported most goods and personnel
in and out of the southeast region by sea—most of the new settlers and
sojourners came by foot, individually and in small groups, walking the
precarious and often overgrown trails from Hengchun, Pingdong, and
Kaohsiung.30 From the 1920s on, in parallel with the physical transfor-
mation imposed by the Japanese infrastructure and economic devel-
opment, this second wave of Han Chinese migration had a profound
impact on the area.31 Much as earlier Chinese colonists had reshaped
the western plains and hills, these new migrants gradually but irrevo-
cably transformed the physical and cultural landscape of the southeast
coast.
Frontier Imaginary: Modern Taidong Identity

Sojourning and migration have for centuries been looked upon as fea-
sible life options in the southeast coast region of mainland China.32
Yet attachment to the place in which one’s patrilineal ancestors are
buried has for centuries been a central, unquestioned axiom of Chi-
nese male identity. Thinking particularly about the traders, sojourn-
ers, and settlers who set out for Taiwan from the coastal provinces of
Fujian and Guangdong in the eighteenth and nineteenth centuries,
Charles Stafford notes, “Most migrants to Taiwan did not see them-
selves, nor did others see them, as having left home permanently”
(2000:162).33 Indeed, for most first-generation immigrants, even those
seeking refuge from debts, vendettas, and arrest, leaving was premised
upon the possibility of an eventual return. Emigrants to Taiwan, like
others who departed for Southeast Asia and North America, undoubt-
edly harbored dreams of revisiting their mainland homes; most would
have had a compelling desire to be buried alongside their ancestors.
Dying far from home, with no descendents or kin to offer incense and
perform the requisite rituals, they would have been doomed to suffer
the fate of hungry ghosts, wandering interminably without hope of
redemption.
Despite such practical dangers and moral disincentives, the major-
ity of Qing-era migrants to Taiwan did leave home permanently. Once
in Taiwan, lingering attachments to mainland homes were superseded
by new social and material challenges. The Taiwan-born descendents
of those immigrants, moreover, had no memory of the mainland. Sac-
rificing at the graves of their pioneer fathers and grandfathers, these
first- and second-generation Taiwanese completed the transformation
of frontier towns and villages into new, full-fledged ancestral homes.
Even once rooted in Taiwan, however, individuals and families
sometimes chose, or were compelled, to pull up stakes and move again.
Such internal “second-wave” migrations from one rural or semi-rural
place to another were common during the Qing and Japanese periods,
although they had all but ended as a measurable demographic trend
by the early 1960s.34 Rural-to-urban migration (or more generally,
movement of the working-age population from smaller, peripheral
communities to those at the core), however, is a trend that continues,
unabated to this day.
Taidong is one of the few areas settled almost entirely by such
second-wave migrants. Isolated and remote, the island’s southeast-

ern quadrant was not even officially included as a part of Taiwan for
most of the four hundred years that Chinese-speaking peoples have
lived there. Even when finally drawn into the empire toward the end
of Qing rule in the late nineteenth century, Taidong continued to be
regarded as a wild, remote, and dangerous periphery. In the decades
that followed, Taidong was a destination of choice only for a few hardy
entrepreneurs and adventurous souls. Most who headed east across
the mountains did so out of desperation: they were often the youngest
sons of land-starved families, unemployed laborers, or villagers who
had lost farms to floods or confiscation. Others were solitary fugitives
escaping gambling debts, underworld vendettas, and police warrants.
This history of a harsh, remote, and backward frontier, which most
Taidong residents regard as a blessing as much as a curse, persists at
the core of local identity.
The Mystique of the Frontier

In the early days of migration from the mainland, the Qing prohibi-
tion on permanent migration ensured that most of those sojourning in
Taiwan were single men, a factor often cited as one of the root causes
of the island society’s inclination toward violence. The settlement of
the western plain is a story of conflict between groups of rebellious
subaltern men and the state; between Han and aborigine; among the
Han, between dialect groups; and, more recently, among turf-battling
criminal gangs and, most infamously perhaps, among debating mem-
bers of the national legislature.
There are a few locales in Taiwan where local history is so fraught
with violence that it has been welded onto local identity. Xiluo, on
the banks of the broad Xiluo River in Yunlin County, is a large, once
prosperous commercial town famous for martial arts masters; a few
miles south is Huwei, a smaller settlement bordered by a shallow, reed-
choked creek. Local legend asserts that armed highwaymen used to
hide in the reeds and ambush passersby. Huwei’s modern reputation
brands it as the hometown of a particularly daring and ruthless brand
of gangster. A local saying goes, “You might make it across the Xiluo
River, but you’ll never get past the Huwei Creek.”
Encoded within this proverbial warning is a critical point of Chi-
nese social logic: communities stand in complementary but also in
symmetrical opposition (competition) to one another. To sustain its
power to produce and reproduce itself (its yang energy in cosmological
terms; the continuity of its male descent lines in the idiom of kinship),
a community must preserve its territorial and social integrity. That
integrity, in turn, depends first on the masculine character (including
individual fighting prowess) of its men and, second, on the capacity
to mobilize that prowess into a collective force. This force, in turn, is
objectified as an intrinsic quality of the place itself, personified in local
tales of heroic feats and miraculous victories.
Violence and retribution figure prominently in narratives of village
character and identity. The men of Beinan, on the outskirts of Taidong
City, are known for their solidarity and toughness. A-piek, who grew up
there, illustrated the point with an anecdote:
The kids in Beinan stick together. No one dares to come into the
village to make trouble. If someone has a vendetta, or whatever,
and comes into the village to look for someone, the guy being
threatened just gives a shout, and everyone comes running to
back him up. One time, some gangsters from Fengtian beat up a
Beinan kid and smashed his motorcycle. It turned out to be a case
of mistaken identity. The head of the Fengtian village council led
a whole delegation to Beinan to personally apologize for the inci-
dent, a real loss of face. But they knew that if they didn’t do this,
there would be a bloody vendetta between the two villages, and
they knew they’d lose and people would get hurt, maybe killed.
People in Beinan have a saying that goes, “If you come into Beinan
and spit one gob of saliva, we will drown you.”
Any aggression against any member of the group, in other words,

will be met with an overwhelming collective response, whatever the
actual threat. This ethos may very well have grown out of the violent
uncertainties of frontier life; in any event, the very same uncertain-
ties and undertones of violence still pervade the lives of those on the
economic and social margins. In Beinan and other towns and villages
throughout Taiwan, the young “braves” (yong) who defended local ter-
ritory and honor were, and are often, the same gangsters (lomoa) who
extort protection money from local businesses; run gambling, pros-
titution, and smuggling rackets; and control local politics, enforcing
control through intimidation and violence.
Yet despite the clannish closeness of some of its neighborhoods and
villages, until quite recently Taidong’s remoteness made it a favored
destination for those escaping social, legal, and economic troubles
elsewhere. Some were fugitives from justice, but most sought refuge
from debts or personal vendettas. Over the course of my research I
made the acquaintance of quite a few who fit the latter description.
I managed also to collect quite a few stories about an older genera-
tion of solitary, desperate, tough men who crossed the mountains with
nothing but the clothes on their backs. Of those whom I knew person-
ally, most were understandably reluctant to reveal their reasons for
flight. But in the right situation, some of them did talk about their past
and lives often filled with vengeance, bad business deals, and gambling
debts.
A- chan, for instance, was himself an expert martial artist who,
according to his own account, had left Zhanghua to spare his family
the anxiety of repeated threats by those he had defeated or uninten-
tionally intimidated. He kept a low profile, teaching a few students
privately (but enthusiastically) on the condition that they not reveal
the identity of their teacher. Unlike local martial arts teachers who
made a business of it (in traditional fashion, practicing herbal medi-
cine and acupuncture as well as offering instruction in martial arts),
A-chan was reticent about discussing his martial arts skills, which were
quite authentic and very effective.
A-bao was the son of a man nicknamed “Iron Finger” Zheng, whose
martial arts skills and prowess are still part of local lore. In his later life
Zheng kept two wives, with whom he had several children, most of
whom became estranged from him. A-bao described his father, who
was in his fifties when A-bao was born, as a man without a friendly bone
in his body. He had been a fighter and enforcer in Tainan, known
for a vicious temper and masterful knife technique. As a young man,
he had made more enemies than he could handle. Sometime near
the end of the Japanese occupation, he escaped across the mountains,
carrying nothing but some clothes and an extra pair of homemade
straw sandals. Once in Taidong, he started out collecting debts but had
few of the social skills needed to cultivate and manage his own gang;
he never became a “boss” but was feared and respected until he died
under mysterious circumstances in the early 1960s. As in Tainan, he
managed to make plenty of enemies, but none could outwit or out-
fight him (although the implication was that he had ultimately met a
violent end). Once, as he walked along a particularly lonely stretch of
the old mountain coast road between Zhiben and Taimali, two men
tried to crush him by rolling a large boulder down the slope above.
He sensed the vibration and evaded the slide by mere inches. Hiding
by pressing himself under an overhang, he waited for his assailants to

show themselves, thinking he had been killed. After a few minutes, he
heard their shouts. He bounded up the ridge like a cat, cornered his
assailants, and beat them both nearly to death.
A-bao’s father was just another nasty hit man in Tainan, but in
Taidong he became a local legend. On Taidong’s wild periphery his
ruthless disposition fit the landscape; indeed, only in this landscape
did his particular wit and skill find their fullest expression— or at least
this is what he came to represent in his sons’ and others’ memories
and anecdotes.
The legendary tales of “Iron Finger” Zheng are paradigmatic of a
kind of story I often heard told in Taidong. In these stories, the protag-
onist escapes a shady past, never described in much detail. He sets out
across the mountains—again, details like the time of year, the route,
how many days it took, and so on are omitted. The crossing is a criti-
cal element of the tale: Taidong natives are actually quite ambivalent
about the mountains that surround them. When I asked the names of
the peaks that towered over the town, most were able to identify only
Mt. Dulan, known for its sapphires, jade, and poisonous snakes. The
mountains, that is, have no particular personality; there is no need
to name them (other than one or two that play a special role in local
lore—Mt. Dulan, for example, is a place of potential wealth and tan-
gible danger). Taidong is of, but also separate from, the mountains.
Echoing the imperial vision, the mountains are everything civilization
is not: wild, hostile, and undifferentiated.
Being of a place evokes a set of traits, some drawing on the natu-
ral characteristics of the place, others defined by its human history. A
person of a place is immediately associated with everything that people
typically say, think, visualize, joke, and imagine about that place. But
rarely are the characteristics of a place enumerated and qualified; they
are bundled into a single attribute that defines the set. “Brooklyn” or
“Shanghai” describes what is understood to be an intrinsic, defining
quality of things and people from those places. Taidong, among other
things, is a heading on a collective résumé, a preamble to how one
identifies and presents oneself to oneself and to others. The value and
meaning of such identity, however, vary considerably among contexts.
Taidong is “hometown” for most of the informants about whom
I have written here. Some have always lived there. Others, following
a decades-long trend of outmigration, have left to try their fortunes
in one of Taiwan’s major cities but return regularly on important rit-
ual and family occasions. When the question comes up, nearly all of
these men and women identify themselves with the phrase, “I am of
Taidong” (goa Tai-tang ê-ah). That declaration certainly expresses the
pride that tends to imbue any statement about where one is from. But
it is often spoken with an air of defiance, for while Taidong is their
“center on the periphery,” they have been long accustomed to hearing
others derogate their hometown as marginal, remote, and backward.
Taidong’s veritable isolation eased considerably with the arrival
of the Japanese, and, as noted, intra- and inter-county transportation
improved dramatically in the second half of the twentieth century.
Today Taidong is linked to Hualian by two highways, one along the
coast and one through the valley. Southbound road and rail links
(finished in 1992) connect the city to Pingdong. The airport handles
nearly a dozen flights every day from Taipei, and there is ferry service
from the small fishing harbor at Fugang to nearby Green and Orchid
Islands. Finally, the Southern Cross Island Highway, which terminates
a few miles north of the city, connects Taidong with Tainan and Kao
hsiung. Though more a touring road than a logistically useful trans-
port route, this newest of the three cross-island highways (completed
in 1972) crosses some of Taiwan’s highest, most difficult, and most
spectacular mountain terrain.
With the advent of standard-gauge rail links to the north and west,
modern highways, scheduled jet service, and, of course, reliable tele-
phone and Internet connections, it is increasingly difficult to imag-
ine Taidong as an isolated frontier or a remote outpost. The urban
area of Taidong has expanded (and risen vertically) dramatically since
the early 1990s, and once sleepy satellite hamlets in the Beinan Plain
and along the coast to the north and south are being rapidly built up
as well. Nevertheless, it is the frontier that frames Taidong identity.
For locals and outsiders alike, there is an obvious association between
local character and Taidong’s remoteness and harsh environment.
This comes across, for instance, in the way working-class Taidong men
tend to highlight a certain set of valued, specifically masculine, attri-
butes. Whether narrating tales of the frontier or just relating personal
anecdotes, they emphasize toughness, directness, and simplicity while
eschewing guile and cunning (though a quick wit and an ability to take
advantage of a lucky chance are held in high esteem).
Pointing to such qualities certifies a common local identity in the
terms of a collectively held ideal of manhood. It also suggests that such
character is home-grown, not something carried along from the west-

ern plains or inherited from distant ancestors. Such mythologizing
reinforces male prestige while disguising the absolutely critical role
women’s labor plays in producing and maintaining the sources of that
prestige; the ability of a temple to field a well-trained ritual procession
troupe (zhentou) depends on the labor and economic contribution of
temple women, yet it is the men who claim the credit and bank the
symbolic rewards.
In the first official census of the Taidong area taken in 1894, less
than 10 percent of the area’s roughly 20,000 residents were Han Chi-
nese. Fifty years later, in 1944, the last full year of Japanese colonial
rule, the population of Taitō (Taidong) Prefecture had grown to
nearly 100,000; almost half, or close to 50,000, were Han.35 Fifty years
beyond that, in 1994, there were 255,000 people registered in Taidong
County, of whom just under 70 percent were non-aborigines.36 In the
semi-urbanized area that includes Taidong City more than 80 percent
are now identified as Han in household registers, and the percent-
age is even higher in the most fertile sections of the Rift Valley. Yet
in the mountain townships aborigines actually comprise a majority of
the population, in some cases surpassing 90 percent (Taidong County
Government, Bureau of Statistics).37
These statistics, rough as they are, reflect three salient histori-
cal realities. First of all, Taidong was the last county in Taiwan with
a non-Han majority. Second, while both Han and aborigine popula-
tions grew considerably during the Japanese period and for the first
two decades of Republic of China (ROC) administration, the num-
ber and percentage of ethnic Han Chinese grew much faster. Finally,
all categories of the population reached a plateau in the mid-1960s;
since then population growth has stagnated or reversed, due mostly
to outmigration of both Han and aborigines to the larger population
centers of western and northern Taiwan. The ratio of ethnic categories
has remained relatively unchanged for most of the last four decades
as Han migration came to an end and the rates of natural increase
for both Han and aborigine groups approached parity. Furthermore,
while aborigine populations have been eclipsed in the urban core and
some of the lowland agricultural areas (the focus of both Japanese and
Han interest), they are still the majority in both the mountain regions
and mixed settlements (few of which are specifically identified in the
county-level statistics) throughout the valley and the coastal strip.
The New Economy

Isolation has been both a burden and a blessing for Taidong. Con-
strained economic development has, however unintentionally, also
limited urbanization and pollution. The low population density and
prominent aborigine presence give empirical weight to the popular
perception that Taidong is still a very different place, still a “wild, prim-
itive, and backward” territory. In particular, after centuries of being
exploited, outnumbered, culturally overwhelmed, and economically
and politically marginalized, Taidong’s aborigines have been moving
closer to the cultural and economic mainstream. At least part of the
reason for this trend lies in the ideology of “nativization” that has swept
the island since the end of martial law in 1987.
There is also a somewhat more problematic reason for this resur-
gence of aborigine culture. Taidong’s economic development has
come to rely increasingly on tourism over the last two decades, and
aborigine culture has taken center stage in advertising and promo-
tion efforts. In addition to photographic images of aborigine men and
(mostly) women in traditional costume used in tour and hotel adver-
tisements, there has been a spate of upbeat popular songs (in both
Mandarin and Taiwanese) that enumerate the supposed charms of the
area: drinking millet beer, eating wild game, circle-dancing drunkenly
into the night, and so on.
In a 2005 television ad, pop music diva Zhang Huimei (A-mei), a
Puyuma aborigine woman born and raised in Taidong, is shown float-
ing, carefree, down a pristine mountain stream, drinking the adver-
tised brand of bottled iced tea and singing joyously at the top of her
lungs; a middle-class urban Han family camping nearby is startled by
the music ringing through the forest. In the final shot, A-mei joins
the family in a forest clearing. Laughing together, they toast each
other with bottles of the advertised iced tea. Locally, aborigine culture
is being artifactualized and commodified, presented and sold in the
form of museum pieces, handicrafts, and souvenirs; traditional local
festival rituals; and performances of music and dance in diverse ven-
ues, including the National Museum of Prehistory café and the lobby
of the five-star Naruwan Hotel.
There is considerable ambivalence about this publicity in Taidong’s
aborigine communities. On the one hand, many consider such recog-
nition, even if distorting or misrepresenting them and their culture, as
a long-awaited compensation for years of repression and exploitation.
Among working-class Han, however, one can sometimes hear expres-

sions of resentment about the aborigines’ improved social status. For
example, a frequently heard complaint goes that while Han people
work hard but barely scrape by, aborigines are getting rich on sales of
inherited land and unfair taxation policies.38 This is an updated and
somewhat more subtle rendition of the long-lived Han stereotype of
aborigines: they dislike work and have no sense of value (they sell their
land, their patrimony, rather than save and invest, and they party off
the profits, spending away their gains).
The social divide between Han and aborigines persists as well, and
interactions are frequently fraught with distrust and mutual distaste.
One often hears shop owners or stall vendors laughing at or even curs-
ing aborigines from the countryside who have just asked for some prod-
uct they don’t stock or sell or complained about the price of something.
Sometimes the comment is nothing more than a single phrase, “ hoan-
a” (savage), or an ironic “ yuanzhumin” (the politically correct Manda-
rin term for aborigines) and an exchange of knowing looks among the
vendors. Ethnic prejudice in Taidong persists and remains relatively
unaltered in form, despite new political rhetoric, social reconfigura-
tion, and rising prosperity throughout the community.
On the other hand, the new public prestige of aborigine culture
is also understood in the Han community as an effective develop-
ment strategy and economic boon. It is one of the things that makes
Taidong unique and therefore a desirable destination for vacationers
and travelers. The advertising campaigns and popular culture repre-
sentations promote a notion of Taidong as a wild, mysterious, exotic,
and sometimes rough and violent place. This serves to enhance many
informants’ sense of Taidong (and of themselves as men of Taidong)
as inextricably tied to the frontier.
Among those directly responsible for new development as well as
new cultural representations of Taidong is a group of outside inves-
tors who have built hundreds of vacation villas and opened dozens
of spas and leisure resorts. Taidong’s middle class also grew steadily
throughout the 1980s and 1990s. Made up largely of young profession-
als, officials, and entrepreneurs from Taipei and other urban centers,
these new residents (along with the vacationers and tourists) have had
a considerable impact on both the local economy and local culture,
accelerating what some might consider the homogenization and glo-
balization of Taidong. Despite the new general prosperity, the wealth
has not been evenly distributed: as might be expected, the biggest
players have been investors and developers working with influential

local officials and elite businessmen.
The pattern of outmigration that began in the late 1960s has thus
continued unabated. Not everyone leaves, of course; there are oppor-
tunities and prospects for those who stick around. It is a fact, however,
that for most the employment prospects (whether in the shadow econ-
omy or mainstream society) are better elsewhere. Making it in the big
city, moreover, carries far more prestige, especially when one returns
home to Taidong for important occasions like the annual Lantern Fes-
tival procession.
Temple Networks, Ritual Performance, and Turf

The handful of Chinese who settled in Taidong before 1875 congre-
gated in an area of coastal flats south of the Beinan Stream, a place
that came to be called Baosang Village.39 The waterfront neighbor-
hood now called Baosang District was probably the core area of the
original settlement. During the years of Japanese occupation this once
agricultural community was thoroughly urbanized. After the Pacific
War, as the city expanded to the south and west, the old Baosong
area was largely neglected. Even in the mid-1990s, many of its streets
and winding alleys were still lined with dilapidated single-story adobe
and wood houses. This was a largely low-income area, housing mostly
low-skilled laborers, shopkeepers, and tradesmen. Baosang and the
neighboring downtown area are also home to a substantial portion of
Taidong’s jianghu population and its legal and semi-legal commercial
operations. Baosang is, however (or, perhaps, therefore) also the site
of the city’s oldest temples and today boasts the highest concentration
of shrines, temples, and religious specialists in the city. It is, in other
words, Taidong’s ritual center.
The connections and divisions among Taidong’s urban temples
are dynamic. With the exception of the two oldest temples, the rela-
tive prestige and ranking of temples (and gods) is interminably con-
tested. Within temples, schisms create factions that have often broken
off to establish their own shrines, each group offering a competing,
contradictory version of the reasons for the split. Some of these splits
have generated long, bitter feuds that play out on ritual and social
occasions, sometimes escalating into violent confrontations. Other
splits more closely resemble the familiar “division of incense,” wherein
burning incense embers from the mother temple’s main censer are
transferred to the new temple. The god’s magical power, ling, is thus
extended to the new site.
These forms are hardly unique to Taidong: similar patterns of
factional infighting feature prominently in the sociology of Chinese
popular religion (and of the jianghu and of Chinese communities
generally). Two factors, however, make Taidong somewhat of a special
case. In the first place, with one notable exception, none of Taidong’s
urban temples can be considered territorial cults.40 It is a common-
place in Chinese popular religion that when local gods are taken on
“rounds of inspection” on their “birthdays” or other ritual occasions,
they trace the limits of a territorial boundary. On other occasions, they
follow the lines of the ritual hierarchy, visiting “sister” temples as they
move toward the center of the network. In Taidong, however, “tours
of inspection” almost always travel a route of factional alliances. It is
virtually impossible to define the territorial limits of Taidong’s gods;
it is equally difficult to describe the hierarchy of cults because at any
given moment there are multiple, competing claims of primacy; fac-
tions form, dissolve, rise, fall, and shift allegiance.41
It is the nature of popular cults in China that the power of gods
is measured by the number of devotees the god attracts. Especially
in urban areas, devotees move freely from temple to temple, visiting
those deemed to be particularly powerful or petitioning particular
gods on appropriate occasions (the god Wen Chang, patron of exami-
nations, for example, is particularly popular around the time of the
college entrance tests in both China and Taiwan). But rarely does one
find, as is the case in Taidong, a regional temple network structured
almost entirely along the lines of the current map of powerful dei-
ties. Taidong has nothing quite like the “turf-based” ( jiaotou) temples
found elsewhere in Taiwan. Turf, rather, is metaphorical rather than
geographical and is almost entirely the product of the god’s perceived
measure of power. Most of Taidong’s various factions and fraternal
associations—religious, but also occupational, political, commercial,
and so on—are defined by the aura of their power rather than any
objective boundary. We can see this, for example, in the weak and
unstable structure of local gangs. There are no named “triads,” no
franchises or branch halls of the Bamboo Union or Four Seas Gang.
There are certainly no “godfathers.” Rather, gangs coalesce around
charismatic, influential “big brothers,” whose time at the top is usually
quite brief (a few years at most) before they voluntarily retire. To this
day Taidong’s underworld seems marginal even relative to the rest of

Taiwan. One should not make the mistake of taking them lightly (or
suggesting that they are small-timers), but Taidong’s brothers of the
“rivers and lakes” are clearly men who prefer the freedom and flexibil-
ity of self-employment and the relaxed pace of life in the mountains
beyond.42
Separation and Return: Hometown Taidong

If Taidong society is indeed relatively less hierarchical and less regi-
mented than elsewhere in Taiwan or China (and I would argue that it
is), this phenomenon remains to be explained and linked somehow to
the characteristics of the place itself. It could be that Taidongers delib-
erately eschew organizational rigidity or that they merely haven’t been
settled long enough to develop stronger institutions, or it could be due
to some combination of these and perhaps other, unexamined, fac-
tors. One factor may be the unique nature of Taidong’s demographic
history. Taidong is unusual as well for having been colonized almost
entirely by “second-wave” internal migrants. They came from long-set-
tled communities elsewhere in Taiwan, first in a slow but steady trickle
during the 1920s and ’30s and then in a series of small surges during
the first two decades of Guomindang rule. The majority of ethnically
Han adults living in Taidong today are the children or grandchildren
of internal migrants from western Taiwan. Only a very few can trace
their family’s presence back before the Japanese period or claim more
than six generations on the land.
Within just a few years after the last surge of resettlement (ca.
1960), Taiwan’s economy underwent a major transformation. By the
mid-1960s light industry had replaced agriculture as Taiwan’s main
economic engine. Its remoteness now a liability, Taidong began to
lose its working-age population to the cities of western and northern
Taiwan.
Given such a short interval between the arrival of new immigrants
and the beginning of the outmigration trend, it would be reasonable
to expect that current residents have at best a shallow sense of local
identity. What we find, however, is that most who migrated out as
adults continue to identify themselves, in many cases quite strongly,
as Taidongese. Some demonstrate this identification dramatically by
returning to Taidong for the Lunar New Year and to participate in
the annual Lantern Festival temple procession. Without the returned
sojourners, many temples would find it difficult to field a full-scale pro-
cession squad. When these prodigal sons return to their native place,
they enjoy a subtle elevation of status. In their adopted homes they
are striving outsiders; here, they are big men, more so than they could
have become had they stayed.
Here, then, we find a basic flaw in the theory of attachment to
ancestral places; in practice, such attachment is far more proximate
than an analysis of traditional accounts of Chinese patrilineal families
would suggest. Those who are able to establish themselves in a new
community become attached to that community, I would argue, first
through the children they raise to adulthood there. Those children,
in turn, even those who migrate or sojourn elsewhere in turn, are
attached to the home of their parents’ residence. Such attachment,
however, is not purely filial; it is defined as much by the community
of peers and associates as it is by bonds of kinship and abstract moral
obligations to parents and ancestors.
A Fortuitous Arrival
I first visited the remote southeast coast of Taiwan in 1982, traveling
on a local “Golden Horse” bus over what were then partially paved,
gravelly roads that traversed the southernmost spur of the Central
Mountains, then wound north along the headlands, tiny bays, and
steep-sloped beaches of the Pacific shore. The contrast with the rest
of the island, even the most rural sections of central and southern
Taiwan, was remarkable. Here, at a time when much of Taiwan had
become a polluted, crowded, semi-urban landscape of gray concrete
apartment blocks and small factories, Taidong was still a sparsely popu-
lated boondocks, a wide-open space of steep, wooded mountains and
rugged ocean coast, dotted with small towns and hamlets surrounded
by green rice paddies and orchards. Even in Taidong City, the county
seat, some farmers still drove their produce to the morning market in
ox carts.
As idyllic as this seemed to me at the time, for local residents Tai
dong’s pristine environment offered little incentive to remain. There
was little pollution because there were few people and even fewer
factories. At a time when Taiwan’s economy had moved firmly into
the industrial stage, Taidong’s remoteness and isolation were above
all obstacles to progress and development. At a time of growing pros-
perity elsewhere in Taiwan, Taidong was home to a disproportionate
number of charity hospitals and missionary-run relief organizations.
When I returned in the summer of 1987, the winding gravel road
up the coast had been newly graded and paved, and work was under
way on a rail link across the mountains to Pingdong. The largest
employer in town was still the old sugar refinery built by the Japanese
in 1913, but Taiwan was on the verge of changes that even Taidong
could not escape. In 1987, an economic bubble was rising, started by
a huge influx of private foreign exchange, driving a craze of specula-
tion in real estate and the stock market. Throughout the late 1980s
and early 1990s, it seemed everyone in Taiwan was investing, speculat-
ing, and spending like never before. Taidong was no longer the sleepy,
remote outpost I had first stumbled across seven years earlier. But it
still lagged far behind the rest of the island economically: “the moun-
tains beyond” remained well off the beaten track.
It was during this time that domestic tourism in Taiwan began to
take off. People had money to spend and a new sense of leisure that
increasingly included travel, sightseeing, and outdoor activity. Recog-
nizing the trend, the government and local business boosters began
earnestly searching for ways to attract visitors and tap into this new
wealth. Taidong’s isolation had protected it from the scourge of pollu-
tion and overcrowding. What had been obstacles to development now
became valuable, exploitable assets.
Throughout Taiwan—Taidong included—the shadow economy
was growing as fast as any sector from the financial surplus. Money was
flowing into leisure services such as karaoke clubs and video game par-
lors, the sex trade was flourishing, and underground operations like
gambling rackets were running full steam. This was, then, a period of
unprecedented prosperity and even popularity for Taidong’s loosely
organized but highly active underworld.
“ Naruwán! ”
In March 2003, a spokeswoman for then premier You Xikun officially
declared 2004 “Visit Taiwan Year.” During the announcement, she
introduced the campaign’s slogan: “Naruwán! Welcome to Taiwan!”
“Naruwán,” she explained, means “How are you?” in the language
of Taidong’s Amis (Pangcah) aborigines (actually it doesn’t).43 The
cheerful, folksy slogan was part English, part Amis; what it was emphat-
ically not was Chinese. This elision was more than a linguistic trick.
The ruling Democratic Progressive Party was in the midst of a program
to dissociate Taiwan symbolically from “China”—not that many for-
eign tourists would know the difference, much less get the subliminal
message.
For Taidong, the promotion was more than symbolic, however.

Officials and boosters in Taidong had long pinned hopes for a local
“economic miracle” on the tourist trade. In the mid-1990s, the Tai
dong County government circulated a “General Development Plan”
that recommended heavy investment in tourism as the area’s best
available development strategy. The synthesis of a growing Taiwanese
nostalgia for pristine, folksy Taiwan and a new wave of international
tourists made perfect sense to the designers of the international adver-
tising campaign and local Taidong entrepreneurs, even if the rest of
the world never quite caught on.44
In anticipation of “ Visit Taiwan Year,” a long-delayed airport
expansion was hurriedly completed, and new hotels, restaurants, and
resorts rushed to open their doors in time for the hoped-for influx of
tourists. As of 2006, the transformation of Taidong from backwater to
mainstream seemed nearly complete. The open deer fields (hunting
grounds) and rice paddies; the swamps drained for sugarcane planta-
tions; the failed, abandoned Japanese settlements; and the scrubland
converted to pineapple orchards by refugees from Typhoon Karen
have all been divided into lots, with villas and spas starting to crowd
out the few remaining farmsteads. The “uncivilized headhunters” have
been turned into “friendly natives,” greeting visitors at the airport with
a cheerful “Naruwán! Welcome to Taidong!” And the deviant rituals of
superstitious gamblers and violent hooligans have become authentic
frontier folk traditions. Though it is no longer remote or primitive, the
myth of frontier Taidong persists. It sustains the profits of real estate
speculators, resort developers, and package tour operators, but it is
also a source of prideful identity for some who remain on the margins.
“Peh,” A-hian explained, using the Holo version of my Chinese sur-
name, “let me tell you. Taidong gangsters are simple and direct. They
think we’re all wild mountain men,” he laughed, “but ‘the mountains
are green and the waters run clear [shanqing shuixiu].’ Excuse me, we
may be rough, but we get right to the point.”
Let us return at last to the legends of fugitives and outlaws: cross-
ing the mountains, they find not peace but even more danger, albeit
a different kind of danger. While the real experience of the frontier
is gradually receding from living memory, in these tales, Taidong
becomes again ao-soa* (houshan), the mountains beyond. As a space
beyond the margin, touched only lightly by law and civilization, Tai
dong is both beyond and of the mountains. Its heroes became heroes
only in crossing the divide; they survive only because they are them-
selves wild, strong, and resilient enough to make it to the other side.
They are men whose character, or in other words their qi, is compatible
with the place; in crossing the wild periphery, they are further trans-
formed, bringing some of the mountains with them. Jianghushang, “on
the rivers and lakes,” suggests a rootless, treacherous, and peripheral
existence. Working-class people in Taidong— especially but not exclu-
sively the gangsters—have a particularly acute sense of Taidong’s geo-
graphical isolation and economic marginality. Thus they can be only
of Taidong and never return to the other side.
Reveling in these stories of danger and prowess, my Taidong infor-
mants often seemed to imply that they themselves—indeed everyone
of the place (and perhaps also those who live here long enough)—
have absorbed some of that wild frontier spirit as well. But storytelling
is a creative—that is, a productive—act. The teller produces the scene,
but he also plays the medium, in this case manifesting the spirits of
Taidong past into the moment of the present. We revel in the adven-
tures and exploits of these characters but yet are obviously removed
from the danger and deprivation that suffused their lives.
Telling a story, however, achieves only a partial identification. In
the kinetic immediacy of ritual action and spirit possession, men—in
many cases these very same raconteurs—are able to overcome the
aesthetic distance entailed by verbal narrative. In the bodily idiom of
ritual, they can possess directly what eludes them, materially and emo-
tionally, in daily life.
CHAPTER 4
Fire and Fury
T he walled prefectural city of Dali, in western Yunnan Province, rests

in a valley, strategically equidistant from mountains and water. Like
Taidong, it sits on an active fault zone. Like Taidong, a massive, sheer
mountain barrier rises up from the valley, and lower, but equally rug-
ged, peaks protect and isolate it on the north and south. Lying to the
east, again coincidentally like Taidong, is a large body of deep water.1
In Dali, that water is the Erhai—literally “ear-shaped sea”—which cov-
ers most of the valley floor. East of the lake stretches not the empty
vastness of open ocean but an arid expanse of dusty, red limestone
hills. To the west, dominating the landscape at every turn, the massive
Cangshan Mountains rise up to form a steep, continuous barrier for
more than thirty miles. From the base of the mountains a long, narrow
plain, crosscut by a series of fast-moving mountain streams, extends
down to the edge of the lake. This plain is a densely settled patch-
work of rice paddies and cornfields, lakeside hamlets, hilltop villages,
and market towns. To the south, the lake becomes a narrow river and
turns west, flowing through a narrow gap in the mountains toward the
Mekong. Guarding the southern entrance to the basin is the sprawl-
ing city of Xiaguan (lit. lower outpost), which in recent decades has
replaced the old walled town as the area’s commercial and administra-
tive center.
Though remote from the perspective of the Chinese imperium,
the Erhai Basin was for centuries an important hub of cultural, politi-
cal, and commercial activity. From ca. 750 CE until the Mongol vic-
tory of 1254, the Nanzhao and Dali kingdoms successively held sway
over much of what is now Yunnan Province.2 Dali’s strategic advantage
lay not only in its nearly impenetrable mountain fastness; it was also
the junction of four major trade routes that comprised what has been
called the “Southern Silk Road” (see Yang 2004).3
Today the majority of the prefecture’s rural population are ethnic
Bai, one of the PRC’s fifty-six recognized “ethnic minorities.” 4 Lin-
guistic evidence generally supports the Bai claim that their ancestors
87
settled in the region more than two thousand years ago; their asso-
ciation with the founding of the Nanzhao and Dali kingdoms is less
certain, but they were certainly part of the ethnic confederation that
made up both states. Throughout their history, however, the Bai have
not been alone in the Erhai region: close to half a million Y i populate
the upland areas, and Chinese speakers, both Han and Muslim Hui,
dominate the urban districts. Hui villages are also clustered along the
old highway, which winds across the mountains to Burma, less than
one hundred miles to the west.
On top of the geographical distance, political division, and diver-
gent historical experience of the people, my research in Dali differs
from my Taidong work in one very important respect. Taidong and
Dali are both multiethnic regions that were until recently the domain
of non-Chinese peoples, now politically and economically dominated
by the Chinese state and increasingly marginalized in their own land
by a growing Han majority. But in Taidong, my research focused
almost exclusively on the Han; in Dali, I worked in ethnic Bai villages
and towns. Yet despite the differences, I found that the communities
I studied in Taidong and Dali shared basic forms of family and com-
munity organization that are, at some level, common throughout the
historically Chinese-speaking regions. Religion in Dali is in many ways
significantly different from what I had known in Taiwan; yet my Taiwan
experience made it possible to comprehend and to talk with local resi-
dents about Bai religion.
Mainstream Chinese culture began infiltrating Dali no later than
the tenth century and accelerated considerably during the Ming and
Qing periods.5 Moreover, in recent decades the Bai have increas-
ingly pointed to their “Chineseness” to contrast themselves favorably
with both the Muslim Hui and the highland Y i, whom they consider
less “civilized” (that is, sinified). Just as important, the Bai long ago
adopted Chinese rationales and representational forms to theorize
their own patrilineal family system. Indeed most religious practice and
belief among the Bai seems at most a variation on ubiquitous Han
themes: images of deities arrayed hierarchically in village temples;
offerings of food, incense, and spirit money and the burning of writ-
ten petitions; Daoist priests and liturgy; and familiar Buddhist icons,
rituals, and beliefs.
Yunnan and Taiwan were both remote frontier territories that
remained largely unassimilated and uncontrolled until the very end
of the nineteenth century. Conflicts between Chinese-speaking and
Map 4.1. Dali
minority inhabitants and subethnic rivalries (Zhang-Quan and Holo-

Hakka in Taiwan, Han and Hui [among others] in Yunnan) often
turned violent, occasionally escalating into open rebellion. The vicis-
situdes of such reciprocal violence, the yields of corpses, and the rise
and decline of local strongmen were often explained in terms of
miraculous or supernatural power. In Dali, for instance, many of the
gods that are considered efficacious are the souls of men of both sides
who met violent ends during such rebellions.
In Dali as in Taidong, the rugged, threatening qualities of the land
seem to merge with the collective memory of the sometimes violent,
usually vexed relationship among communities divided along cultural,
ethnic, and even class lines. This collective memory finds expression in
the myth narratives and in ritual practices unique to each place. These
rituals, in turn, feature displays of violent masculine bravado by young
men who risk life and limb in hopes of gaining supernatural as well as
social rewards.
Fire and Fertility: The Bai Torch Festival

Every year on the twenty-fifth day of the sixth lunar month, the Bai in
towns and villages throughout Dali Prefecture celebrate a community
ritual called the “Torch Festival.” “Torch festival” (huobajie) is a Chinese
term, coined by outsiders; in Bai the festival is called simply feuh-wa-
vvh, “end of the sixth lunar month.” Since “torch festival” has already
become the generic term and since it accurately describes the center-
piece of the ritual, I will continue to use it.
Traditionally, both Han and Bai villages had two symbolic centers:
a village temple and an old banyan tree, and throughout Dali today
nearly every village has one or both. Bai village temples follow Han
models, resembling imperial-era structures, and to an untutored eye,
there is little that differentiates Bai representational styles or iconogra-
phy from their Han counterparts. There are some salient differences,
however. Bai local patron deities (benzhu), while carved and arrayed
much like those elsewhere, are unique to each village (or at most a
pair or cluster of villages). Although they are “appointed” to office
like the gods of the Chinese celestial bureaucracy, they do not belong
to that pantheon. Other than Buddhas and bodhisattvas, each village
recognizes only its own set of patron deities.6
My research on the Bai Torch Festival took place on four sepa-
rate trips to the field, covering two distinct kinds of locale. In 1995 I
observed the Torch Festival in the upper sector of the western shore
fire and fury 91
of Erhai Lake. I spent the majority of this time in Zhoucheng and

Xizhou, major market towns that figure prominently on the standard
Dali sightseeing circuit.7 The third site, Xiangyang, is a small lakeside
village. In 1996, 2002, and 2005 I worked in and around Y inqiao Town-
ship, which is bounded by the highway to the west and the lake to the
east. In those years I lived in one village—which I will call Wuguan-
zhuang—near the center of the township. During the weeks prior to
the festival, I also visited neighboring villages to interview and observe
ritual masters and craftsmen hired by the Wuguanzhuang sponsors.
The Torch Festival is a community ritual celebrating the birth of
a firstborn—traditionally a firstborn son. The parents and household
heads of firstborns who have been born during the year since the last
festival become sponsors. They are responsible for contributing all the
money and labor required to build a large torch, including the tree
trunk core, the wheat straw wrapping, the bamboo strips that bind it
together, fruit and firecrackers strung along the outside, and the sheng-
dou (described below), which graces the top of the torch. They are also
required to pay for the services of ritual specialists and entertainers
who are hired from outside the village. Young women (except the new
mothers) return to their native villages on festival day. All men, chil-
dren, uxorilocally married women, and older women remain.
The Bai traditionally explain the performance of the ritual in at
least two different ways. One directly concerns agricultural produc-
tion—the festival takes place during a period of relative rest, just
before the summer rice harvest and during a warm, rainy period when
insects are particularly numerous and threaten the ripening grain. The
namesake of the festival—the torch—refers to both the large, central
torch pole and small, pine pitch torches carried by anyone old and
strong enough to handle one. In many villages women and children
run through the fields of ripening rice waving the burning, smoking
torches as they go. Some claim this is just to get rid of insects; several
older villagers, however, suggested that it was a way to send off the
ghosts who had been invited to feast on small packets of sticky rice
dumplings scattered around the fields earlier in the day (see below).
Another of the minor rituals leads to a second explanation. Two
or three days before the ritual, young women and children soak their
fingers in red dye, alluding to a Bai myth of ethnogenesis: the wife
of a lord burned to death by a treacherous king literally scraped his
remains out of the ashes, covering her fingernails in blood. Return-
ing home, she vowed revenge. The ensuing war united the tribes of
all the chiefs burned to death by the evil king, marking the origin of
the Bai as a people. The burning of the torch recalls the conflagration
that marks this creation of the Bai out of the ashes of their patrilineal
ancestors. The myth reveals a central feature of the ritual itself: social
production requires the fertility and labor of women; in the myth, and
in the ritual, we see the recognition that women are the necessary pro-
ductive link in the reproduction of the patriline. Women’s power is
acknowledged—but only as a subordinated power. The ultimate pur-
pose of women’s labor and fertility is to reproduce the lineage into
which they have married.
Two minor rituals performed by women and children deserve men-
tion here. The night before the festival, or in some places on the day of
the festival, a special meal is prepared. Leftovers are packed into leaf
wrappers and placed in one’s own fields. Some say this is to propitiate
ghosts; some say the food was formerly left out as a festival gift for beg-
gars or the porters from Sichuan and Guizhou, outsiders with no local
ties who frequented the area. In fact, the packets in each field were
collected and eaten by children of other village families. As with all
propitiatory offerings to gods and ghosts, once the spirits have eaten
the “essence,” the physical food retains protective, talismanic prop-
erties that are absorbed by consuming the “leftovers.” The practice
seems to have waned and perhaps disappeared in recent years, though
Wuguanzhuang villagers recall making and eating the rice packets as
recently as the late 1980s.
The second ritual, still common, takes place at home as well. Moth-
ers and grandmothers make glutinous rice-flour balls, mixing in herbs
that are believed to cleanse and purify. They rub the bodies of their
young children or grandchildren, concentrating on each vital area,
eliciting cries of protest and giggles in turn. Finally, the rice-flour balls
are inserted into wheat straw bundles, carried to the main square, and
surreptitiously tied into the torch to be burned, carrying away what-
ever baleful airs (shaqi) may have been absorbed from the children’s
bodies.
The torch ritual takes place in the plaza of the village temple.
Throughout the preparations and in the ceremonies and rituals of the
main event, there is a clear sexual division of labor. The work done by
the male sponsors takes place almost entirely in the temple square and
on the temple steps. The men congregate, forming a village unit that
works, eats, and rests together throughout the day. Older men write
couplets, while the younger men do the hard labor of cutting and
fire and fury 93
stripping the tree and transporting it to the village square. They gather
the wheat straw into bundles and spend the better part of a day fasten-
ing them to the tree pole with bamboo strips. This endeavor requires
a set of long levers and four or five men for each tightening motion.
Twelve bamboo bands, one for each of the lunar months (thirteen in
a leap year) hold the torch together. The entire building operation
takes several hours—most of the day if it rains, which it often does at
this time of year.
Female members of the sponsoring families—the new mothers
if they are able, unmarried sisters, and daughters—make social vis-
its, carrying baskets of sweets to share with neighbors and relatives.
Paternal grandmothers perform domestic rituals such as the rice-flour
ball purifications for all the household’s children. Later in the village
square, they cluster along lines of family, friendship, and age cohort.
But throughout the day, women meander around the torch, surrepti-
tiously inserting the bundles of straw that contain incense, firecrack-
ers, and the rice-flour balls that have been rubbed over the bodies of
their children.
The men of Wuguanzhuang raise a torch.

The main event begins as the village head or an invited party offi-
cer reads a commemorative edict, usually in Mandarin, authorizing
the festival as an officially sanctioned event. After the edict is read,
strings of firecrackers attached to the north and south corners of the
torch are set off. At this point the women of the Lotus Pond Society
(Lianchi Hui), elder women of the village who have been singing Bud-
dhist chants most of the day inside the temple, conduct a brief purifi-
cation ritual. Female representatives of each village family then gather
and make obeisance to the torch. The torch is lit from a point near the
top by a village elder using a long bamboo pole. Right after the torch
is ignited, women carrying infants, children, and open umbrellas rush
under the torch and circle it twice, counterclockwise.
The climax of the ritual comes shortly after the torch has been
set alight. A crowd of young men gathers at the south, or yang, cor-
ner of the torch where, near the top, a three-tiered bamboo-and-col-
ored-paper construction has been inserted and hangs over the crowd.
Called a shengdou, “peck and bushel measure,” it is a full-scale model
of the containers used to measure the volume of rice in traditional
markets. Here, the containers are inverted so that bushels of heav-
enly grain will pour ceaselessly into ever-increasing piles. At the top of
the papier-mâché construction is a model of an imperial magistrate’s
seal, wrapped in red cloth, and a miniature spear, bow-and-arrow, or
halberd blade. The characters for a happy and fruitful marriage—the
“double happiness”—are pasted on the bushel measures; in some vil-
lages, figurines of the bodhisattva Guanyin or the son-granting goddess
Zhusheng Niang’niang, plastic baby dolls, or the images of the mis-
chievous monkey god Sun Wukong are mounted inside the measures.
The fire soon burns through the bamboo pole that holds this
pagoda-like apparatus in place. If the pole burns clean through, it falls
to the ground, and the young men scramble to grab it. If the break is
only partial, one or more particularly daring youths will try to climb
the burning torch as the crowd shouts encouragement. The successful
grabber is mobbed and congratulated by his friends, and they head to
his house. Later that evening and in some places for several days, he
will entertain the village men and invited guests.
The crowd in the square now concentrates on the small torches.
Groups of young men compete in throwing handfuls of powdered
pine incense on the flaming torches, sending up loud puffs of fire and
smoke. Anyone is fair game, and being “smoked” is said to purify one
of stale airs (meiqi) . The prime targets, however, are village girls. The
fire and fury 95
courtship play often takes an aggressive form, and the girls sometimes
suffer scarring burns. As the “party” moves out from the village square
to the boundary with neighboring villages, the competition often esca-
lates well beyond horseplay. Every year, the young men of Wuguang-
zhuang “burn” their way down to Daxiang on the west and Hanxi to
the north, meeting the youths of the other villages at the bridges that
mark village boundaries. At first the encounter takes the form of a
suspicious greeting, a probing of possible weakness. Often, however,
these encounters escalate into open challenge, although serious inju-
ries are rare.
In a survey I conducted in 1996, fire battles, or “playing the torches
to the next village,” were one of the most commonly cited reasons for
fights among the young men of neighboring communities. A number
of villages in the area no longer undertake a torch festival; many cite
past violence as the reason. However, as informants in Wuguanzhuang
told me privately, fights between neighboring villages or production
units have always been a regular feature of the festival. There must be
other reasons, they say, why the festival has been discontinued in those
places.
Back in the temple courtyard, the torch begins to burn down; the
singing contest is over. The entertainers have been fed and paid and
have left for home. By midnight most visitors have moved on, and the
locals have mostly returned home to sleep. A few men and women
from the sponsoring families remain to keep watch through the night.
The following morning, the festival committee of family heads meets
in the temple square. The new fathers, usually in their early twenties,
join the family heads in taking down the torch pole and cleaning up.
The men negotiate the proper division of the leftover money, bamboo
strips, and food; the money is distributed by the chief purser to the
other men; the surplus rice and other food is distributed by women.
Before collectivization in the 1950s, only the fathers of firstborn
sons were entitled to act as torch sponsors and builders. Correspond-
ing to this privilege was the penalty imposed on the fathers of firstborn
daughters: they were sent to dig the hole, a shameful task that most
chose to do in the darkness of the moonless night. The conflicts that
arose between the proud sponsors of torches and the shamed diggers
of holes often led to violence and even death. In the 1930s, the father
of one newborn daughter was so distraught at his social failure that he
beat the infant to death and threw her corpse into the newly dug pit.
Many villages tell of less tragic but equally violent clashes between the
torch builders and hole diggers. Since parents can sponsor a torch
burning only at the birth of their first child, the social position of a
father whose firstborn was a girl could be permanently diminished.
Violence and masculinity are linked with fertility and the social
vitality of the community in nearly every symbol and ritual act through-
out the festival. All of these meanings and associations are woven into
the torch itself, the symbolic focus of the event. The torch celebrates,
first of all, the reproductive power of the new fathers, those who have
successfully continued their patrilines, realized their greatest filial obli-
gation, added new life to the community, and thus confirmed them-
selves as full-fledged men.
Such layering of symbolic meaning, however, does not obscure
the obvious—namely, that the reproduction of the patriline and the
renewal of village vitality both depend on the sexual ability of village
men. A direct discussion in such terms would be awkward for most vil-
lagers, who nevertheless address the subject easily through innuendo
and joking. Humor, in fact, is built into the traditional form of the
celebration. After the main ritual is over, as the torch burns in the
background, the villagers gather in the temple courtyard to listen to
hired entertainers. The highlight of their performance is a piece of
extended sung repartee between a “boy” and “girl” (usually both roles
are played by old women), full of sexual innuendo that incites gales of
laughter. The phallic form of the torch might seem an all too obvious
symbol, but again, that meaning is confirmed only through humor. As
one informant put it, those who do not produce sons, who do not ful-
fill their obligation to the lineage, are failed men who endure shame
and social ridicule. He told the story of a Wuguanzhuang native, one
of his grandfather’s generation, who had died without heirs in shame
and disgrace. The person telling me this sad story then joked, “That
poor bastard could never erect his torch.”
The joke, however, was not facetious. The fertilizing power of the
torch comes in part from being sponsored, assembled, and erected by
the village men who have just fathered firstborn sons (and their fathers
and patrikin). The torch represents the unquestioned association of
male fertility with the existence and perpetuation of village society
itself and village society as an amalgamation of patrilineal households.
The ritual thus functions, at another level, to mystify the real relations
of production—that is, to assert the “natural” association of maleness
with community identity, power, and authority.
As a rite of passage, the ritual evokes two stages of masculine iden-
fire and fury 97
tity. For the young men who climb the burning torch and fight to grab
the halberd (or spear or bow and arrow), the act demonstrates their
youthful prowess. Newly married men who have not yet fathered sons,
however, are very much betwixt and between as far as their social iden-
tity is concerned. Participating in the competition focuses on them
as candidates for membership in the community, but candidates
who have yet to prove their value as men. Grabbing the halberd indi-
cates the favor of Heaven, a reward for their strength and daring. Yet
at the same time, it anticipates the birth of a son and future social
responsibility.8
The one who gets the spear celebrates by feasting the village elders
and invited guests on the night of the festival. It is exclusively men that
host and attend these feasts, at which they are also expected to dis-
play their drinking prowess. Women are not allowed on the premises,
though they are sometimes provided with a share of the food outside
the compound walls.
Along with village temples, dragon dances, and exorcisms, the
Torch Festival was banned as a vestige of “feudal superstition” during
Fighting for the shengdou.

the Cultural Revolution and slowly revived throughout the Dali region
during the 1980s. Following fieldwork in the Pearl River Delta in the
1980s, Helen Siu (1989) concluded that the reappearance of once
banned customs and religious rituals in the PRC after 1978 should
not be considered revivals or renewals, arguing that the power struc-
tures and informal organization of local society that gave meaning to
rituals in the past had disappeared. Bereft of their original context,
rituals have taken on completely new meanings in post-Mao China. In
Dali, the situation seems somewhat more complicated. As the contrast
between the torch festivals now celebrated in tourist-oriented market
towns and rural lakeside villages suggests, the extent of recontextual-
ization varies considerably from place to place. At the same time, mod-
ernization and the influence of global commercial culture are clearly
changing both the social context of ritual performance and the forms
of rituals themselves.
But the process goes both ways—local cultures are certainly being
affected by globalizing forces; the Torch Festival has been quite suc-
cessfully touted by local Dali officials as a tourist folk festival, and tour-
ism is the heart and soul of economic development in Dali. But the
power of locality as a marker of identity remains remarkably stable. In
this part of Dali at least, local identity, or rather identification of and
with a locale, has generally been premised on the knowledge that the
local place is a node in an extensive regional and, increasingly, inter-
national network of people and institutions. Local identities, as repre-
sented, for instance, in networks of community organizations, almost
always extend to higher-order market towns and cities, as well as to
higher-order cosmic locales.
Different kinds of violence and different ritual representations of
violence evoke and produce different sorts, or stages, of male iden-
tity. In the rural economy of Wuguanzhuang, violence as a primary
mode of domination is the exclusive province of young men. Control
of younger men and the productive labor of women marks a real man
of prowess. For successful, necessarily middle-aged village leaders, the
control and manipulation of violence serves as an objective mecha-
nism of domination.
The ritual manifests the productive power of male charisma. One
individual rises out of a crowd into a hail of fire and debris. Perform-
ers certainly are motivated by the chance to gain name, honor, and the
admiration of peers—all constitutive elements of subjective masculine
identity. But the ritual process here first glorifies, then subordinates,
fire and fury 99
the power of individual young men to the higher-order institutional

values of male group solidarity and collective power.
The Torch Festival is a ritual of renewal and reproduction struc-
tured very much around the dualistic qualities of male and female, fire
and blood.9 Transposing culturally constructed male and female attri-
butes onto the tangible qualities of ritual object, the torch’s construc-
tion and destruction can also be understood as an ideological repre-
sentation of the sexual hierarchy that inheres in local social structure.
The Torch Festival thus links male fertility, economic production,
local identity, and patrilineal ideology under one very colorful, fiery
umbrella.
The Taidong Lantern Festival

On the fifteenth day of the first lunar month—the first full moon of
the new year—the gods of Taidong’s temples and shrines are carefully
lifted from their altar stands and mounted on carved rosewood palan-
quins for the Lantern Festival. Meanwhile, families and businesses are
busy setting out folding tables covered with offerings of food, incense,
and spirit money for the deities, their demon generals, and the armies
of ghostly soldiers. Beginning in the late morning, the gods are fer-
ried by pickup truck to the homes and businesses of faithful patrons
throughout the city, accompanied by teams of helpers and sometimes
by lion dancers, troupes of costumed spirit soldiers, and entranced
spirit mediums. All afternoon, the staccato pop of firecrackers marks
the hundreds of greetings and sendoffs, each temple attending to its
own.10
At dusk the gods and their entourages head off once more, this
time to the staging area for the evening’s procession. As night falls,
the procession, now several thousand strong, starts moving along the
spiraling route, north-south the first night, east-west the second. The
afternoon’s sporadic detonations now intensify into sustained vol-
leys, reaching deafening crescendos as the parading gods, their heavy
palanquins shouldered by teams of eight men, stop in at every temple,
shrine, and makeshift altar along the way.
During the procession, every few minutes a minor but noisy ritual
is repeated at every temple, shrine, and table-altar along the route.
Preceding each of the gods is a team of four men carrying a miniature
sedan chair (H. kio-a) mounted on two bamboo poles. Approaching
at a run, they pull up at the threshold, greeted by a volley of firecrack-
ers tossed at their feet, and dip the chair one or three times before
rushing back to the procession.11 This simple ritual—the shirtless

carriers, their heads wrapped in keffiyeh-like head towels, the minia-
ture sedan chair, and the shower of exploding firecrackers that greets
each team— can be seen throughout the year at any temple festival
in Taiwan. During Taidong’s Lantern Festival, however, it takes on a
special significance; for old-timers and eager tourists alike, this briefly
repeated scene foreshadows the main event: Blasting Handan Ye, Tai
dong’s “gangster god.”
The God of Easy Come, Easy Go

Handan Ye is the name of a god of obscure identity and character
whose cult, about which little has been written, seems to have propa-
gated mostly around the fringes of Taiwan’s shadow world of itinerant
gamblers and gangsters. Handan Ye’s cult apparently did not spread
through the conventional practice of “dividing incense,” nor was he
part of any larger temple or pilgrimage networks. From the few extant
fragmentary accounts, we know that there were Handan Ye cults in Jiayi
and Zhunan in the 1930s and in Taipei (in the old riverside neighbor-
hoods of Wanhua and Dadaocheng) as late as the 1960s (Feuchtwang
1992:56; Suzuki 1978:313 –315).12 Given the extreme scarcity of docu-
mentation, it is likely that there were never more than a few isolated
Handan Ye cults at any given time. Sometime during the 1950s Handan
Ye and the simple, violent ritual that accompanies his cult appeared in
Taidong, where it developed in relative isolation for about thirty years,
was briefly suppressed by the authorities, and was then revived in the
late 1980s. No one can recall the name of the man who brought Han-
dan Ye across the mountains, other than that he raised ducks under
the Kangle Bridge and left Taidong after turning the god over to a
local gambler nicknamed “Big Hog.” Elsewhere in Taiwan, Handan Ye
had faded out of existence, if not memory, by about 1970.
Even the name and iconography of the god are intriguingly ambig-
uous. Li Jianzhi, who has been the cult’s de facto “master of the censer”
(luzhu) since 1990, has pursued a long campaign to enhance Handan
Ye’s public reputation.13 This public relations effort centers on a drive
to “standardize” the narrative and iconic representations of the god.
The campaign has been partly successful: Handan Ye is now generally
associated with the orthodox deity Zhao Gongming, a Daoist adept
immortalized in the Ming novel Fengshen yanyi (Creation of the Gods),
enfeoffed as the Military God of Wealth. Although known by various
names (for example, Black Tiger Marshal—he is often depicted rid-
fire and fury 101
ing a black tiger), Zhao Gongming’s title in the Daoist pantheon is

“Field Marshal of the Abstruse Altar,” Xuantan Yuanshuai. This is often
shortened to Xuantan Ye, which is pronounced han-tan ia in Taiwanese
Holo. According to this account, then, “Handan Ye” is nothing more
than a misidentification of xuantan ye.
This sort of arcane philology, once the bread and butter of Confu-
cian scholars of the “evidentiary inquiry” (kaozheng) school, remains
an important form of legitimating discourse in popular religion (and
among folklorists). For Li Jianzhi, it has been particularly important
to emphasize the orthodox identity of the god, given the once dubi-
ous status of the cult and its infamous ritual. I take up Li’s interest in
Handan Ye’s “standardization” again in the next section. My purpose
here is actually to dispute his version, which erases much of the sym-
bolic information encoded in the popular stories and unconventional
depictions of the god prior to 1990.
The orthodox God of Wealth, Zhao Xuantan, is a “black-faced”
god (o-bian-e); Taidong’s Handan Ye has a multicolored face (hoe-bian-
e). He doesn’t ride a black tiger; in his right hand he wields a nine-
sectioned club; in his left he carries not the trademark gold ingot of
the God of Wealth but a Heavenly Seal. The miniature figurine that
adorns the back of the sedan chair during the procession, a copy of the
earliest Taidong version, has a face patterned with radiating red and
white stripes. This echoes one of the origin stories that was the expla-
nation of choice among Handan Ye devotees before 1990, essentially a
modified version of the myth of the ancient archer Hou Yi. In the time
of the sage emperor Yao, so the myth goes, there were ten suns that
were scorching the earth. For the sake of the suffering people, Hou
Yi, an expert archer, shot down nine of them.14 Eight became plague
gods; the one unaccounted for is Handan Ye. Having lost his fire, he
is constantly cold. Blasting him with fireworks warms his heart and
moves him to lend his power to enriching the blasting patron.
Since its revival in 1990, Blasting Handan Ye has become the cen-
terpiece of Taidong’s annual Lantern Festival procession. In form and
conception, the ritual is simple: a volunteer, usually a young or middle-
aged member of a local gang, stands astride the god’s palanquin, a
simple wicker chair with a miniature carved figure of the god strapped
to the top of the seat back. The chair is supported on a pair of white-
and-red candy striped bamboo poles, carried by eight bearers. (The
carriers are not volunteers expressing their devotion and allegiance
to the god but a hired crew of Amis aborigine laborers.) “Handan Ye”
rides stripped to the waist, wearing only a pair of red shorts and a red
head scarf bound with a yellow strip of cotton. Goggles and a spray of
willow fronds grasped in the right hand are his only safety equipment.
At designated points—or whenever a patron makes a sufficiently gen-
erous offering—the crowd is moved back, a space cleared, and bundles
of live fireworks are stacked around the perimeter. When the signal is
given, the team begins to march in circles, carrying Handan Ye. As
he circles the performance space, he is bombarded with firecrackers
tossed from the sidelines or strung from a bamboo pole and thrust
against his body. Enduring salvo after salvo, the “flesh body Handan
Ye” (roushen Handan Ye) does his best to bat away the blasts near his
face but makes no effort to protect the rest of his exposed body. At the
end of the round, which can last for up to ten minutes of continuous
blasting, he steps from the chair and is surrounded by the crowd eager
to diagnose the severity of his burns. By midnight, downtown Taidong
is carpeted with spent firecrackers and the air saturated with a smoky
fog of incense, burned spirit paper, and pyrotechnic residues.
The violence of the ritual and the social marginality of the partici-
pants led to an official ban on the ritual for five years, from 1984 until
1989. It was revived, however, in the midst of a wave of popular interest
in local “folk customs,” especially strange ones. It was soon clear to all
parties that the ritual, performed as part of the city-wide Lantern Fes-
tival procession, was a tremendous tourist draw.
Handan Ye in parade display mode.

fire and fury 103
It is significant that the cult’s devotees have been mostly gamblers

and members of local gangster society. The ritual of Blasting Handan
Ye is intimately tied in with the self-conscious values of gangster subcul-
ture and identity. Getting up on the sedan chair and risking life, limb,
and eyesight is an act of pure daring. The possessed spirit mediums
who pierce and flail themselves while in trance are thought to feel no
pain; their bodies and senses have, theoretically, been given over to the
possessing god. Being unwilled, their self-mutilations are not violations
of the prohibition against damaging the body one owes to one’s par-
Blasting Handan Ye.

ents. Handan Ye, however, enjoys no such exemption. He deliberately

and unflinchingly subjects himself to the pain, danger, and injury. The
sacrifice releases him from the debt he owes the community. As he
stands down, he hands out cigarettes and betel nut to his comrades,
displaying an air of total nonchalance, while the burns covering his
body are treated with antibiotic spray.
The entire performance evokes the stories of two mythical charac-
ters, both deities beloved throughout China but who have particular
meaning for (and enormous popularity among) the “brothers of the
dark path.” Like the capricious immortal monkey god Sun Wukong,
the gangster who stands on the chair is punished with fire for defying
Heaven. But instead of being reduced to ashes, he emerges purified
and hardened; he now possesses a “pure yang body.” And like the dis-
obedient patricidal child god Nezha, by “returning his flesh and bones
to his parents,” he is freed of his filial debt (and guilt), releasing the
potential for wreaking even greater havoc (cf. Sangren 2000:202). One
additional layer of myth completes the picture: as A-hiang, an expert
in these matters pointed out to me, Handan Ye is iconographically
identical to the deity Wang Tianjun. Wang is the personal bodyguard
of the Jade Emperor. Though he is of low rank, he answers only and
directly to his master, the Emperor of Heaven. He can thus act with
impunity, and, indeed, he is known for his arrogance and capricious
nature. Here, then, the myth, iconography, and ritual logic of the prac-
tice converge. Handan Ye is a god of profits, windfalls, unholy luck,
and extortion. He is indeed Taidong’s lomoa sin, the gangster god.
Burning Ambiguities
After rotating among members of Taidong’s gangster subculture for
forty years, the cult of Handan Ye ended up in the capable hands of Mr.
Li Jianzhi, a reformed gangster who recently realized a lifelong ambi-
tion, winning election to the Taidong County legislature. Li’s drive and
strategic navigation of the local political landscape have much to do
with the public acceptance of Handan Ye as a characteristic Taidong
folk custom. Yet to successfully sustain Handan Ye’s popularity, he has
to perform a delicate balancing act, seeking legitimacy and notoriety
at the same time. On the one hand, to obtain official sanction he must
downplay the heterodox elements of the ritual—the physical violence,
the questionable origin of the deity, the ritual’s association with gang-
ster machismo, and its lack of narrative structure. On the other hand,
he must emphasize the more notorious and heterodox elements of the
fire and fury 105
cult in order to sustain and increase public interest. The Taiwanese

and Japanese producers who have featured the ritual on various recent
TV news programs are clearly more interested in the spectacular, het-
erodox elements of the cult than in Li’s orations on filial piety.
The struggle over control of Handan Ye in the context of govern-
ment-sponsored religious festivals has a number of ramifications. On
the surface, the county government’s motives ostensibly stem from a
distrust of those whom it sees as marginal and violent members of soci-
ety, and efforts to control the cult and its activities have generally been
made in the police rhetoric of public safety. I would argue, however,
that the form of the ritual lies at the heart of Li’s dilemma and pro-
duces profound discomfort among the enforcers of patriarchal order.
Unlike Daoist exorcisms, unlike the martial demonstrations of lion
dancing and military procession troupes, the ritual has no clear begin-
ning and no clear end. Unlike possession trances, there is no narrative
progression, no creative variation, no miracles, no drama. A volunteer
mounts the sedan chair, the carriers hoist the rig onto their shoul-
ders, the signal is sounded, and Handan Ye begins a long, continuous
orbit around the performance space, enduring blasts and burns until
he can stand no more. When one man is done, another ascends the
chair, and the process begins again. Like the rootless liminality of life
in the jianghu, the ritual is a harsh trial by fire with no tangible reward
beyond the pride of survival. The lived experience of many of those
who ride the chair is hardly enviable—not a few have served long sen-
tences in prison for felony offenses, and at least two recent perform-
ers have died premature, violent deaths (one by suicide, one gunned
down in revenge). But as “living Handan,” they revel for a moment in
their own violent wildness, holding an audience in morbid thrall.
From the perspective of the logic of martial symbols, Handan Ye’s
power to create wealth derives from the link between fire and the
transformation of qi (the dynamic substratum of matter) through sac-
rifice, here played out in dramatic fashion in the burning and blasting
of the living body of the surrogate Handan Ye. Looking at the ritual
from a more sociological point of view, however, we can see that the
violence of underworld subculture is here ritually shared by the com-
munity through the sacrifice of the “gangster” on the sedan chair. In
this way the community sees a way to bypass conventional limitations
and channels (as gangsters routinely do) to generate and redistrib-
ute material wealth. The personality of power, however, is capricious:
transformation entails both creation and destruction, curing or mak-
ing ill, extending life and cutting it short, engendering prosperity or

casting into poverty. Thus the man of prowess—here represented by a
mythologized gangster—must also be propitiated to keep his (and the
collectivity’s) destructive propensities in check. Yet it is the dangerous
and heterodox quality of the ritual of Blasting Handan Ye, the his-
torical connection with local gangster subculture, and most of all the
exaggerated performance of culturally valorized traits of masculinity
that invite the public fervor of the performers, the curiosity of tourists,
and the anxiety of the local authorities.
In this age of mass media and culture tourism, the fame of Handan
Ye has spread across the island. In 1996, Handan Ye was the featured
topic of the National Performing Arts Season, and a professionally
produced documentary film about the ritual and some of its devotees
was released in 2006. Handan Ye has become an icon of Taidong’s new
identity through its reinvention as a relic of the old, wild frontier town.
In a way probably unintended by the authorities, the compact they
have made with Li and Handan Ye authorizes the violence of the ritual
and raises the heretofore marginalized cult members to the status of
honored sons of Taidong.
Ritual Danger and Social Production

The Bai Torch Festival and the Blasting of Handan Ye exploit the dra-
matic and tangible qualities of fire and explosion to focus attention on
a sacrificial victim, in both cases a young man who has volunteered to
put himself in harm’s way. He implicitly challenges the onlookers, who
in turn seem to anticipate the specter of disfigurement or injury by
fire or fall. The sneer of victory as the successful torch climber plants
the spear and swears an oath to the village god and the nonchalant
grin as the living but scorched Handan Ye dismounts from the palan-
quin are palpable and unmistakable declarations of triumph over the
protagonist’s private fears and the secret guilt of the crowd. In that
moment, it is not hard to understand what drives them to it. As oth-
erwise low-status, socially marginal males, through this act of symbolic
intimidation they are suddenly feared and admired by the crowd and
acknowledged by their peers.
Both rituals, however, require large expenditures, extensive plan-
ning; well-managed organization of labor; and the integration of mul-
tiple layers of interest, desire, and recognition across an entire com-
munity. The satisfaction of one subject’s narcissistic desire, therefore,
is insufficient to explain the ritual’s origin or purpose. Of course we
fire and fury 107
also want to understand, beyond momentary emotional gratification,

what could motivate someone to undergo such an ordeal. The key, I
think, lies in the way both rituals simultaneously engender and con-
strain the expression of violent aggression, which is staged, celebrated,
and consumed as drama, spectacle, and entertainment but is also con-
trolled and, once the ritual is formally concluded, subdued.
In expressing the desire for completeness, the ritual subjects prom-
ise the sponsors and spectators that they possess a masculine vitality
sufficient to endure but ultimately be transformed by the ordeal. In
this sense, the performance gives expression to the violent, yang, mas-
culine energy of these young men. But ultimately it harnesses that
energy to the productive and protective needs of the community.15
Those needs are somewhat different in each context. In the case
of the Bai torch festival, young men race the fire and fight their way up
the torch to grab the prize—the spear, talisman of male fertility. The
performers are desperate to realize their manhood and take their place
in the community; their failure would disrupt not only their own self-
esteem but also the emotional stability of the household (in another
era, and even today, economic status permitting, the solution might be
to take a concubine), the continuity of the patriline, and the social and
economic vitality of the village, which depends on reproductive labor
to replenish and sustain itself as a viable community. This message, and
function, plays out as well in the contests of challenge and aggressive
courtship that take place on the village outskirts. These fire battles,
which on occasion escalate into brawls among neighboring villages,
identify individual male sexual aggressivity with the shared need to
protect the territorial integrity of the community. It thus contrasts with
yet mirrors the discontinued tradition of tree stealing, the organized
raiding by stealth of another village. In those raids, the men crossed
the village’s protective boundary quickly and deliberately. They tres-
passed the territories of neutral villages to reach their objective and,
relying on surprise, stealth, protective talismans, and the assistance of
demonic spirits invoked for the occasion, identified, cut, and hauled
back across hostile ground a thirty-foot, half-ton tree trunk.16
The culprits would be known only after the fact; the act of revenge
would be incomplete unless the victims had an intimation of who had
violated them, stolen their property, and diminished their face. The
raiding teams proved themselves stalwart and efficient; waking to dis-
cover the violation, the victims had to suffer the sting of having failed
to defend territory and property. Until such time as they could exact
revenge and balance the ledger, they would be lesser men than the
raiders. The victims’ loss and the aggressors’ gain were only inciden-
tally measurable in economic terms. The tree stolen by the raiders was
a talisman of fertility, a portion of the victims’ productive (masculine)
vitality.
Like the (now discontinued) practice of forcing fathers of first-
born girls to dig the torch pit, the midnight stealth of the tree-steal-
ing mission was crucial to both the ethos and purpose of the act. The
insistence that the victims bear their shame in (open) secret made
their emasculation complete. Any village father of a firstborn daugh-
ter who publicly challenged the practice would have suffered the dou-
ble misfortune of losing his moral face (lian) as well as his social face
(mianzi) .17 Similarly, the men of a village that protested its loss of a
tree would have become a laughingstock. Ostensibly, the stealth factor
was intended to give the victims a way out (taijie ; lit. a stage step —that
is, the exit from an embarrassing scene). But conflicts still occurred,
aggravated (as the story of the distraught father suggests) by the expo-
sure of a shameful secret.
Besides the climbing of the burning torch, the most dramatic
scene of the Bai Torch Festival plays out as the hand-torch fire battles
that begin in the temple plaza move out to the village margins. Here,
the action is not planned but spontaneous and random; the actors
not a select crew of experienced men but mostly uninitiated teenag-
ers; the drama not concealed in the blackness of a moonless night but
brightly illuminated by the torches and powder flashes, every move
and the identity of every participant in full view. The main torch is a
symbolic distillation of tremendous creative effort, physical labor, orga-
nizational aptitude, and (not least) economic capacity. Its construc-
tion manifests the productive power of the community; its burning is
a conspicuous sacrifice that amplifies and generalizes that productive
energy. The Torch Festival ritually produces both men and the forms
of masculinity proper to each phase of male subjectivity in Bai village
society: the sacrifice of the new fathers, whose productive capacity and
transformation into real men is celebrated by the entire event; the
performance of the builders; the childless young newlyweds who fight
for the spear; and finally the playful fire battles (with their intimations
of danger, displays of male aggression as both courtship gambits and
male contest). These ritualized performances of an aggressive mascu-
linity generate an excess of male vitality. This is not to say, of course,
that Bai society valorizes impulsive violence and uncontrolled aggres-
fire and fury 109
sion, which are undesirable and unacceptable in the context of nor-

mal, everyday affairs. But for Bai villagers, aggression is also a natural
expression of male vitality. In this view, male aggression is disruptive
and anarchic unless brought under control. Once controlled, it can be
harnessed to the collective interest, most important the reproduction
of the patriline.
The Bai Torch Festival comprises a multiplicity of rituals, some in
sequence, some simultaneous, some public, some performed inside
the walls of the family compound (and, in the past, some conducted
in stealth and secrecy). It has elements of carnival—boundaries are
relaxed; everyday rules suspended; individual energies given over to
carousing, revelry, and indulgence and moments of deep ritual solem-
nity as well. It is, on the one hand, a miscellaneous assortment of leg-
ends, symbols, rituals, and customs, a mishmash of Han and Bai religion
and contemporary popular culture. Yet at the core is a single purpose,
a consistent logic of ritual production. The ritual both celebrates and
activates the fertility of crops, domestic animals, and families, thus
establishing the linked interdependence of masculine identity, the
continued viability of the community, and patrilineal production.
In Taidong, the ritual of Blasting Handan Ye is but one, albeit a
defining, element of the larger festival. It is the entire Lantern Fes-
tival, rather, that stands for comparison with the Bai case. Taidong’s
annual, two-day Lantern Festival procession is in many ways a typical
Chinese temple festival, a “contesting conference of the gods” (ying-
shen saihui) .18 But because it takes place at the First Prime, Yuanxiao, it
takes on additional significant qualities. The first full moon is the con-
clusion and climax of the Lunar New Year period, the Spring Festival.
New Year feasts and rituals (which kick off with preparations weeks in
advance) purify, exorcise, and invigorate the body-self, as well as the
many levels of the social body (household, extended family, commu-
nity). They are intended to magnify the efficacy of productive labor,
anticipating increased fertility, physical vitality, and material surplus
(some, of course, hope for a miraculous windfall).
Held at the peak of this auspicious period—the full moon is the
most yang of the Spring Festival’s fifteen days—the Lantern Festival,
like the Torch Festival, is a carnivalesque blend of fire and fertility.
Until China’s turn to a more puritanical culture during the Song
dynasty, it was generally a time of relative sexual license. In the liter-
ary imagination of the Ming and Qing, if not in social fact, the Lan-
tern Festival continued to be a pretext for stories of illicit romance
and elopement. Today, this scandalous element survives mostly in the

form of children’s games and deracinated customs. Riddle challenges,
now sponsored by local culture bureaus and television stations, may be
transforms of the suggestive “stomping songs” (ta’ge) that were once
a common feature of the celebration. Indeed, the custom of riddling
seems to have originated during the prudish Song dynasty. Until then,
open courtship was permitted on this night, and singing suggestive
duets was a customary feature of the celebration (see Hartman 1995:16
passim). Coincidentally or not, the Torch Festival similarly features
the singing of suggestive (even ribald) courtship songs. Encouraging
courtship and relaxing everyday strictures on young couples’ behavior
are practically as well as symbolically effective ways to ensure biological
continuity and increase.
One would be extremely hard-pressed, however, to find any such
suggestive overtones in modern Lantern Festival customs. Neverthe-
less, fertility remains firmly embedded in the event’s ritual code. But
whereas “stomping songs” and Bai courtship duets are initiated and
dominated by female performers, the ritual performances of the Tai
dong Lantern Festival are characteristically martial and male.19 The
forms of verbal and physical play are unambiguously masculine and
aggressive (cf. Gilmore 1987).
The ritual of Blasting Handan Ye epitomizes the martial masculine
ethos of the procession and serves as its most dramatic expression.
There is considerable ambivalence, however, in the way the practice is
perceived by the local public and depicted in the national media. The
community’s association of the ritual with the local underworld and
the undeniable fact that most of the young men who ride the sedan
chair have criminal records lend a somewhat sinister implication to
the performance. The performance, as many of the “good citizens”
of Taidong seem to fear, is not purely symbolic but is actually an act
of extortion in the guise of a ritual. Many times I was told (mostly by
people who had no direct personal knowledge of the practice) that
the gangsters who ran the Handan Ye cult used the event to threaten
local businesses. In this view, the troupe would pick a business it felt it
could squeeze and start performing. Unless a sufficiently well-stuffed
red envelope was forthcoming, the owner could be sure of another
visit, this time with no ritual pretensions, a few days later. Whether or
not that was ever the case, I cannot say. But over several years of direct,
extended participation with the Handan Ye cult and its organizers, I
never saw any evidence of such a practice. In fact, although the cult is
fire and fury 111
certainly connected to local gamblers and gangster society, it is not in

itself a gang or a brotherhood in any real sense. There are but a few
regular devotees, and men in the retinue gathered for the procession,
including the volunteers who ride the chair, do not socialize, work, or
worship together as a group in any regular context.
The extortion story has lost currency over the last two decades, as
ordinary Taidong residents have come to acknowledge the ritual as
a genuine expression of Taidong’s “elemental,” “primitive,” and vio-
lent masculinity. At the same time, and despite the heroic efforts of Li
Jianzhi, the devotees of Handan Ye are welcomed only in their roles as
ritual performers. During the rest of the year, they are still feared and
marginalized gangsters.
Blasting Handan Ye is clearly a practice introduced by Han Chi-
nese, one that is thoroughly encompassed within the frame of Chinese
popular religion. Yet both the earlier criticism and recent promotion
of the ritual as a unique and defining expression of local Taidong cul-
ture imply that it embodies “barbaric” and “primitive” attributes, which
stem from both its mountain-bound remoteness and close contact with
aborigine culture.
Postscript
In 1984, the police commissioner of Taidong City issued an edict ban-
ning Blasting Handan Ye as a “threat to public security”; the Torch
Festival was suppressed in the 1960s as part of the general attack on tra-
dition during the Cultural Revolution. Both rituals were later revived
through local grassroots efforts, only to be transformed again as local
authorities, now promoting ethno-tourism, recognized them as poten-
tial commercial assets. The impact of globalization has affected not
only the meanings attached to the rituals, but technical aspects of the
performance as well. Tourists now pay to stand on Handan Ye’s chair,
although they are pelted with water balloons rather than firecrackers;
and in many places around the Erhai Basin much of the Torch Festival
day is devoted to folk dancing displays by village women, staged and
directed by male urban videographers who edit these “authentic,” col-
orful cultural performances onto video CDs (which they sell to tourists
as well as to the villagers).
Blasting Handan Ye was once performed mostly on demand, at the
doors of Taidong’s brothels, gambling dens, and underground the-
aters. Now the ritual is performed according to a published schedule,
off the procession route, in large, accessible public spaces. The hot-
test ticket in Taidong on the first night of the Lantern Festival is just
in front of the downtown McDonald’s, where Handan Ye puts on a
two-hour show. Every year now, the firecrackers get more explosive,
and the expected “chair time” for each volunteer is extended. The
après-chair examination is now explicitly part of the program, avidly
recorded on video and cell phone camera by those lucky enough to
get close.
In Dali, the torches burn much faster now, aided by dry kindling
woven into the straw. In some places accelerants like kerosene are
added to the mix. As recently as 2002, the torch at Wuguanzhuang
burned all night; in 2005, it had been completely consumed within an
hour of lighting. Trivial as that might seem, it reflects a very different
intentionality. As the Torch Festival takes place during a rainy season,
the straw is often wet, needing multiple attempts to light and hours
to burn off. The ritual takes place, in fact, during an agricultural lull,
just after the final weeding and just before the harvest of the summer
rice crop. The leisurely pace of the burning, the all-night vigil, indeed
the very fact that the straw was soaked by rain, were all meaningful ele-
ments of the ceremony that no longer fit. The ritual is in no danger of
disappearing again, but the context has clearly changed.
The revival of traditional folk religious customs and practices
since the 1980s has followed different trajectories in different parts
of mainland China. In the Pearl River Delta, Helen Siu (1989) saw
reconstituted traditions functioning as exotic anomalies, amusements,
or at best post-socialist venues for social and political maneuvering.
Kenneth Dean (1993:99–117), whose research examines Daoism and
popular religion in north-central Fujian (like the Pearl River Delta,
part of the coastal prosperity belt, but an area where more local prac-
tices apparently went “underground” rather than disappearing alto-
gether during the Mao years) found a very different situation. There
the revival of processions of the gods in the late 1980s brought local
residents into direct conflict with the authorities. Compared with the
groups Siu observed, practitioners in the Putian-Xianyou region of
Fujian seemed to have a greater investment in, and deeper knowledge
of, the religious concepts and practices they sought to revive.
In Dali, I would argue that the dilemma faced by local villagers
is not unlike that faced by the Handan Ye cult: out on the periph-
ery, the full weight of Communist anti-religious policy and ideology
was not felt until the Cultural Revolution (whereas, for example, in
eastern China religious activity effectively ended with the Great Leap
fire and fury 113
Forward, almost a decade earlier). The interruption, then, spanned

less than a generation. At the point of revival, not only was the form
of the ritual clearly remembered, but also the symbolic context was, at
least for older residents, still relevant and meaningful. But in places
like Wuguanzhuang, agriculture has been steadily decreasing as a
focus of labor and a source of income for villagers. In the 1990s, they
set up village enterprises, turning living rooms into workshops. For
several years the whole village of Wuguanzhuang became an assembly
line for leather dress shoes, with families specializing in various phases
of manufacture, packaging, and distribution. Time was at a premium,
but even then, most villagers remained at home most of the year.
Since about 2002, however, households have turned to other
sources of income. Many villagers work as migrant laborers or run
small businesses in local towns and cities. They come back for impor-
tant holidays if they can, but the competition is harsh, and the rituals
and symbols of fertility seem increasingly quaint and irrelevant. All the
same, young men still scramble up the burning torch and spar over the
shengdou with as much passion as ever. Village society is still very much
premised on, and male social identity embedded in, the “patrilineal
mode of production.”
There are really two distinct versions of the Bai Torch Festival in
the Erhai Basin: one is organized, financed, and carried out by villag-
ers for villagers (and the deities of the local benzhu temple) in places
like Wuguanzhuang. The other version, which takes place in the pub-
lic squares of the larger towns, caters more obviously to foreign and
Chinese tourists’ desire for “ethnic color.” In those settings, tourists
now often outnumber local participants.
The contrast is not absolute, however. Even the “tourist” version
holds interest and significance for local inhabitants; in the villages,
even without outside spectators, the desire to be simultaneously mod-
ern and authentic has had a palpable effect on the evolution of the
festival. Old-fashioned forms of entertainment that once accompa-
nied the ritual, such as the hired comedic teams of old women singing
ribald songs, have been replaced by teams of village women, young
and old, performing standardized folk dances to loudly amplified
pop songs. And both town and village forms reveal a growing ethnic
self-consciousness expressed, paradoxically, through the performance
(and consumption) of a commodified, externally generated version of
Bai culture and identity.
Similarly, Blasting Handan Ye can be viewed as nothing more than
a show for tourists. At the same time, however, as a new Taidong iden-
tity reconciled to its wild, mountain persona has emerged, the blast-
ing ritual enjoys a level of local interest and support that could never
have been achieved in the pre-tourist era. And Blasting Handan Ye
clearly continues to serve its long-established function as a rite of pas-
sage for the more desperate or daring young men of Taidong’s jianghu
subculture. Handan Ye’s prestige as a symbol of gangster masculinity
has, if anything, been amplified by its new high profile. As Black Bear,
one informant, said to me, snorting derisively, “Every year, you see all
these small-timers running all over town showing off their Handan Ye
T-shirts. All everyone wants is a Handan Ye T-shirt. But they’d piss their
pants if they actually had to get up on the chair!”
CHAPTER 5
Tales from the Jianghu
T he Taidong Lantern Festival procession winds a tightening spiral

through the city’s Japanese-era grid of downtown streets and alleys.
The center of the spiral, the procession’s endpoint, is the courtyard of
the city’s oldest and largest temple, the Tianhou Gong. As the gods and
their retinues arrive, they are ushered, two by two, into the spacious
courtyard to pay obeisance to the temple’s patron, the goddess Mazu.1
Each troupe tries to upstage the others in energetic displays of ritual
zeal. Groups of young, barefoot trance performers, called “sorcerer’s
troupes” (H. hoat- a tin), act out wild, animated possessions, lacerating
head and back with spike balls (ci qiu) , swords, and ice saws; the martial
troupes execute complex, spectacular routines; and finally, the gods
themselves are carried forward in their carved hardwood sedan chairs,
the bearers bouncing and bobbing through a continuous barrage of
exploding firecrackers. After the visiting god salutes Mazu, the troupes
next in line, who have been waiting impatiently to make their own
grand entrance, begin to push their way in. Their performances con-
cluded, the troupes at the front file out into the street through the side
gates. Clear of the crowds and commotion, assistants wipe the blood
and sweat from detranced mediums, and the whole entourage begins
a quiet, exhausted trek back to the home temple.
Returning to their familiar confines, the nearly spent performers
must regroup one more time for the festival’s final act: the homecom-
ing of the patron god. At martial deity temples fielding troupes of
spirit soldiers, the ritual is especially dramatic. Summoning the last of
their energy and inspiration, the performers line up and run through
one more sequence of drills and stances. Fanning out to the sides of
the temple courtyard, they clear a path and stand guard as scores of
devotees queue up to crawl, one by one, under the god’s palanquin
(H. ng kio - khah). Over the two days of the procession, the god in his
palanquin has accepted thousands of sacrificial offerings, been hailed
by dozens of other gods, and vanquished innumerable demons and
ghosts. At this moment, then, the god, the conveyance, and all the sub-
115
sidiary deities and objects carried on it, have acquired an unusual but
temporary bounty of efficacious power. By crawling underneath, devo-
tees may absorb some of this momentarily enhanced power, hoping
for protection against ill fortune and the attacks of baleful forces.
At last, men of the temple’s core group gather to unbind and lift
the god’s image from the sedan chair, pass it through the smoke of the
main incense burner, and ceremoniously reinstall it on the main altar.
The assembly follows the god into the main hall to make a final offer-
ing, as the spirit soldiers retire to a side chapel, where they are ritually
“decommissioned.” They shed their costumes, wash off the face paint,
and change back into street clothes.
One by one, they drift back into the courtyard to join the oth-
ers on benches or at the round tables set out for the feast to come.
There is rice, meat, vegetables, soups, and chilled, if not quite cold,
liter bottles of beer. Like the biweekly propitiatory feeding of ghostly
soldiers and generals, this feast is a rewarding of the troupes, set out-
side in an open, public space. In graphic confirmation of the allusion,
many of the ritual actors sport vivid tattoos of demon heads, tigers, and
dragons, badges of toughness and comradeship. These telltale marks
identify the bearers as daoshang xiongdi, “brothers of the [dark] way.” 2
For much of Chinese history, tattoos were marks of deviance, asso-
ciated with outcast, subaltern status (C. Reed 2000:361). Facial tattoos
were stipulated as punishment for certain crimes, a stigma that perma-
nently marked a convict as a social pariah. Often tattooing was com-
bined with exile or military conscription. But within Chinese military
culture, tattoos came to mark membership in a fraternity of marginal-
ized men that included hardened warriors and righteous hero-bandits,
as well as bullying government runners and street toughs (ibid.:368
passim). Men from the margins of society were already the most likely
to find themselves in the military’s lower ranks; the jianghu was, in
turn, a natural refuge for demobilized soldiers. Circulating through
the imperial military system, among prison guards and state enforc-
ers, rural bandits and urban gangs, tattoos became a defining mark of
jianghu subculture, along with such less tangible signifiers as fighting
prowess and honor.
It should come as no surprise, then, that tattoos are both common
and positively envalued in the jianghu of postwar Hong Kong, Taiwan,
and overseas Chinese communities.3 In Taiwan in particular, the pain-
ful, extended ordeal of a full-body tattoo has long been a requirement
for initiation into the fraternity of the dark path.4 During the years
tales from the jianghu 117
of Taiwan’s draconian anti-hooligan laws, which made visible tattoos

sufficient reason for classification (and incarceration), tattooing was
rarely anything other than a declaration of jianghu membership.
Secret societies, “triads,” “tongs,” and their esoteric, dramatic initi-
ation rituals have long captivated both the Chinese and Western popu-
lar imagination.5 The solemn blood oaths, esoteric martial symbolism,
and classic rites de passage structure of secret society initiations would
seem to be the perfect place to begin an anthropological examina-
tion of ritual, violence, and masculinity on the margins of Chinese
society.6 But throughout the Chinese-speaking world, initiations sealed
by oaths of blood, secret hand signals, and esoteric code words are far
more common in media portrayals of jianghu life than in the jianghu
life itself. The contemporary initiation ceremony, when performed at
all, is exceptionally brief and perfunctory. The incense hall (xiangtang)
altars and requisite ritual paraphernalia are generally improvised and
set up in an empty apartment or the back of a restaurant. While what
does take place is largely a pro forma ceremony, the fact that the ritual
is still performed at all is noteworthy.
As I noted in chapter 2, the configuration of Taiwan’s jianghu is
relatively diffuse. From early in the postwar period, Taiwan’s under-
world has been effectively comprised of two “species,” divided more
or less along mainlander-native lines. Neighborhood-based organiza-
tions called jiaotou (H. kak-thau; lit. “corner heads”) are mostly based
in ethnically Taiwanese areas. Traditionally, members have been local
men who have an allegiance to the town or neighborhood as well as
to their faction boss. On the other hand, mainlander gangs have been
generally less territory-bound. In the late 1950s, youth gangs began to
spring up in the so-called “military dependents villages” ( juancun).7
They characteristically justified their purpose, often through a set of
regulations, in terms that melded Confucian and jianghu morality
(for example, “parents first, brothers second”). They also adopted the
nomenclature and structure of Qing-era secret societies.
Most of these ad hoc sodalities were fluid and short-lived, but a few
of the stronger and more stable gangs expanded through winning turf
wars and occasionally absorbing other groups. A few of these larger
gangs evolved considerably during the 1970s and ’80s, transforming
themselves into urban syndicates and operating sophisticated rackets,
backed up with intimidating muscle and firepower. Some extended
their reach regionally and globally, either directly or through alliances
in Hong Kong, Southeast Asia, Japan, the United States, Latin America,
and (increasingly since the early 1990s) the PRC. The largest and best-
organized of these groups, such as the Four Seas League (Sihai Bang),
the Pine Union League (Songlian Bang), and, most infamously, the
Bamboo Union League (Zhulian Bang), have been identified as inter-
national crime cartels.8
These organizations follow what might be called a generic Chinese
secret society model of organization. They are comprised of branches
called tangkou, tang being the metaphorical ancestral altar before
which all are initiated and that bonds them as fictive kin. While each
tangkou is nominally a branch of the main organization, the “organiza-
tions” are really diffuse networks of semi-autonomous cells, with little,
if any, high-level command structure, common property, or shared
function.9 At the same time, the segmented branch structure, the cat-
egorical logic of branch naming, and the nominal reverence for secret
society traditions suggest that a desire for higher-order coherence and
permanence plays some motivational role even in the most modern,
sophisticated versions of the triad.
Ownby (1995:1023 passim) argues that late imperial secret societies
should be understood as examples or extensions of Chinese popular
religion. Secret society beliefs and practices were, I would concur, not
only consistent with Chinese popular religion, but also deeply embed-
ded and invested in existing syncretic traditions and local temple net-
works (cf. ter Haar 1997:224 –230). The contemporary case, however,
suggests a more complex and less assured connection between the
jianghu and popular religion. In the first place, the esoteric rituals and
codes through which fraternal associations once invested members and
created encompassing lifeworlds play a much-reduced role in Taiwan’s
modern gangland (or China’s or Hong Kong’s, for that matter).10 Yet
a diminished emphasis on ritual does not necessarily entail the lack
of a religious dimension. The tattoos that feature so prominently in
Taiwanese jianghu culture are a perfect example: the vigorous, violent,
fearsome dragons, yakshas, and other mythical and martial symbols
and allusions inscribed on the body identify the bearer as an aspirant
to martial masculinity and imbue him with a power that is both tan-
gible (visual effect) and transcendent.
The drawn-out process of full-body tattooing is more than an
application of paint to a canvas. Men undergo a trial of physical pain,
patience, and endurance; they experience a dissolution and recreation
of personal identity, literally sporting a new skin. But these subjects
also undergo transformation into a different species of social object.
Here, two kinds of “fetishization” take place: first, the power of the
images of destructive and heroic figures, gods, and demons is trans-
fused directly into the subject (cf. C. Reed 2000:371). Second, as the
tattooing covertly transforms the uninitiated into a brother of the dark
path, it also overtly transforms the brother-body into an efficacious
object. From the perspective of an observer, an initiated brother is
a metonymic embodiment of jianghu sodality. This aesthetic relation,
then, produces both individual cultural subjects (men of the jianghu)
and the cultural lifeworlds (including the archetypes and ideals of
masculinity) they inhabit. But the tattoos also imbue the bearer with a
measure of instrumental power, realized in the fear, admiration, fasci-
nation, or even disgust that his appearance can evoke.
In the following sections, I present first a series of ethno-biogra-
phies intended to elucidate this process of ritual production—not of
tattooing, but of jianghu subjects, and reflexively of the jianghu through
the lived experience, practical action, and self-descriptions of some of
those subjects. I venture to explain, first of all, how and why men of the
jianghu should be particularly attracted to (and welcomed in) martial
cults and martial ritual performance troupes. Here I first posit one pos-
sible answer: if we limit the sense of “ jianghu” to organized sodalities
like “secret societies” or “triads,” it seems clear that the religious inspi-
ration that motivated recruitment and served as the foundational pre-
text for earlier secret societies has diminished, shifted, or disappeared.
I suggest, instead, that modern gangs and brotherhoods are not func-
tionally continuous with late-imperial secret societies. They are not
required to fulfill the needs and envelop the lives of members in the
same way as earlier societies, guilds, and brotherhoods once did.
For men living in the chaotic, poverty-riven regions of nineteenth-
century Fujian, Taiwan, Henan, Yunnan, or the colonial enclaves and
Chinatowns of Southeast Asia and North America, such organizations
took on multiple social and spiritual functions. What we find today,
in contrast, is a division of institutional functions between fraternal
sodalities and temple cults. The demonological practice and ideology
of popular religion are operationally focused in temples and cult-spon-
sored ritual. Yet the source of membership and ritual labor in the most
martial and heterodox of these cults is still the “dark societies,” for
some of the same sociological and personal reasons that men of the
jianghu found security, meaning, and solace in joining secret societies
of an earlier era. And the very logic of exorcism—what ter Haar calls
“demonology”—almost entails their participation. These capricious,
violent martial outcasts are often compared to ghosts; it is common

sense among believers that hunhun (“riffraff,” juvenile delinquents)
and lomoa (older, true gangsters) should be chosen to personify the
ghost catchers and spirit soldiers.
The symbiotic relationship between exorcistic, martial cults and
jianghu society has a deep and continuous history. Beggars, bandits,
and gangsters have long been recruited as ritual surrogates and spirit
mediums, for example. Plague processions in nineteenth-century Xia
men (Amoy) (De Groot 1892–1910) and in Wenzhou and Hangzhou
(Katz 1995:143ff.) featured just such outcast ritual actors, and the
late-nineteenth-century Feast of Universal Salvation in northern Tai-
wan included a particularly violent ritual called “robbing the hungry
ghosts” (Weller 1987:74ff.). Temple processions are competitive and
physical, and local toughs associated with rival factions (or recruited
as mercenaries for the occasion) often play an important role (Choi
1995:120; Dean 1993; Katz 1995).
The religious and ideological character of jianghu organizations
has shifted considerably over the last few decades. Whether or not this
has caused or is about to cause a shift in the relationship between orga-
nized brotherhoods and local temple activity begs scrutiny. Temple
festivals are increasingly being coopted by political and business inter-
ests. Does this “folkification” and commercialization of temple festivals
entail the deracination of popular religion? As noted above, even the
most organized gangs and fraternal associations no longer stress the
initiation rituals, origin myths, or esoteric knowledge so important to
their late-imperial predecessors. Do “good brothers” join martial cults
to compensate for the increasingly secular nature of the gangs, or is
participating in martial festival performance troupes merely another
opportunity to “get by”? 11
The connection between martial procession troupes and the jian-
ghu is most evident today in Taiwan. I therefore begin with a focus on
Chinese communities outside the mainland (Taiwan and Hong Kong
in particular) and consider the PRC from a somewhat different angle,
presenting my ethnographic cases from the mainland toward the end
of the chapter.
The Jianghu and Territorial Cults

In the decades after World War II, the ethnic division between “Tai-
wanese” and “mainlanders” marked a structural as well as social divide
within Taiwan’s jianghu . Taiwanese who chose the “dark path” were
less likely to aspire to membership in organized gangs (hun bangpai) .
Rather, they tended to join informal, turf-based street gangs or to
move among various “venture squads” (zuhe) , small, temporary crews
of “drifters” recruited by local bosses for short periods (or to under-
take a particularly unpleasant or risky task). Within this fluid milieu,
tightly structured, stable gang organizations were, and are, somewhat
of a rarity.12
Turf-based gangs occupy a place somewhere between the orga-
nized gangs and the venture squads. They are community-level sodali-
ties and see themselves as serving protective, political, economic,
social, as well as ritual functions. They may simply serve as neighbor-
hood associations, and many territorial temples in southern Taiwan
are still commonly referred to as jiaotoumiao (H. kak-thau biu). Yet for
this very reason, jiaotou leaders are not uncommonly influential neigh-
borhood bosses and big brothers, and the term now refers most often
to small, turf-based street gangs.
Throughout Taiwan, the turf-based gangs have a direct historical
and functional relationship with local deity cults and martial ritual
performance troupes. The very tradition of these troupes, in fact, is
historically linked to (though not, as some maintain, merely a relic
or symbolic echo of ) the village militias, self-protection associations,
and local strongmen that proliferated in Taiwan’s premodern fron-
tier culture.13 Although they might have a separate corporate identity
as wuguan (lit. armory or military training hall), self-protection asso-
ciations typically met and trained on the premises of a local temple,
where they also stored their weapons, lion masks, and other ritual
paraphernalia.14
Self-defense (individual fighting technique and group drills) and
martial performance easily combined with the ongoing ritual practice
of the temple—after all, the men and boys who trained there were
already part of the temple community and often deeply involved in
its ritual activity.15 The lions, Song Jiang Battalions, and spirit soldiers
of each armory represented their communities during periodic tem-
ple festivals, vying to outdo each other in demonstrations of martial
prowess and masculine vigor. This competitive spirit, mostly friendly,
sometimes not, still prevails at festival processions in Taiwan, Hong
Kong, and areas of the mainland where traditional rituals have pub-
licly reemerged.16
Martial Ritual Procession Troupes

Ritual processions metaphorically combine the tours of emperors and
imperial officials, the bandit-hunting posses, and the routine patrols
of state marshals and bailiffs. Along the route, which traces the juris-
diction of the presiding god, any ghost or demon that has chosen to
hide or stand ground instead of wisely leaving for other parts is hunted
down, captured, and destroyed. The authority and command devolve
from the god (who himself or herself may be a civil or military god),
but the actual work of exorcism (in the name of the god and of cosmic
order) is carried out by spirit generals and soldiers, who “appear” in
tangible form as martial performance troupes, or wuzhen.
With the exception of the giant “frame puppets” (H. tai-sin ang-
a) that tower over temple processions in Taiwan, Fujian, and Hoklo
communities in Hong Kong, martial troupes are composed simply of
costumed actors with faces painted in the style of Chinese opera per-
formers. As befits warriors and military officials, they carry a variety of
weapons and implements used in the capture, intimidation, torture,
punishment, and dispatching of wandering ghosts. Some implements
are authentic replicas of swords, spears, and halberds or traditional
tools of capture, restraint, and torture; some are whimsical renderings
of powerful spirit weapons like yin-yang clubs or ghost-quelling mudra
spears. The spirit generals march in formation, breaking out into cho-
reographed battle arrays at critical moments, striking poses, and mov-
ing through routines derived from traditional martial arts forms. Lion
dancers and a few of the other traditional troupes may even be based
in martial arts schools rather than temples; some, like the Song Jiang
Battalion, feature extended demonstrations of fighting skills in addi-
tion to the more stylized moves.
Blood Sacrifice
Many groups feature one or more entranced spirit mediums. Stripped
to the waist, wearing embroidered vests and/or red sashes tied at the
waist, these mediums also wield weapons, here called “precious magic
tools” ( fabao). These weapons, however, have real blades and spikes,
and they cut and pierce not only the invisible ghosts but the medium’s
own body. In recent years, the “sorcerer troupes” noted above, made
up entirely of young mediums who specialize in blood spectacle, have
become the main (but controversial) attraction at many festivals. In
the most spectacular performances, dozens of blood-spattered young
men, many sporting the telltale tattoos of gangs and brotherhoods;

some chewing betel, smoking, or casting glances to their girlfriends
on the sidelines, parade aggressively through the crowd, wielding their
weapons in threatening and photogenic attitudes, stopping frequently
to pose for pictures and video shots.
Although spirit possession and self-mortification are probably
as old as Chinese religion itself, there is considerable disagreement
among devotees, festival organizers, and ritual specialists in Taiwan
over the meaning and legitimacy of these “sorcerer troupes.” Accord-
ing to tradition, when they enter trance, mediums are understood to
lose all awareness as their bodies, consciousness, and senses are taken
over by the possessing spirit. Most of these actors, however, make lit-
tle pretense of being in trance. Their managers explain their zeal as
inspired by the gods, even if they are not directly possessed. Indeed,
although “sorcerer troupes” are now unwelcome at government-sanc-
tioned events, many sponsors and spectators accept this explanation.
From this perspective, even “false trance” performances are sincere
expressions of religious devotion that validate the gods’ incredible,
miraculous power. Despite the restrictions, the troupes are still quite
popular throughout central and southern Taiwan.
A-tiong and Pik-khut-a were fellow Taidong emigrants who had set
up shop in the suburbs of Taizhong. A-tiong had left Taidong as a teen-
ager, trained as a ritual master in the Maoshan tradition, and eventu-
ally opened his own shopfront shrine and folk healing clinic. His sworn
brother, Pik-khut-a, is a gambler, former gang member, and shrine
master who manages a Military Retainers troupe. Pik-khut-a’s is a pro-
fessional troupe, hired to perform for other temples to augment their
image and face during pilgrimages and festivals. In the early 1990s, he
also organized a sorcerer troupe. This procession ensemble was fea-
tured in a popular docutainment video distributed through depart-
ment stores and night market stalls throughout the island in 1991.17
A-tiong had high regard for Pik-khut-a’s band of performers but
noted as well that they were a difficult bunch to handle: “Those ‘little
ghosts’ are really fearless [gan si]; they’ve got plenty of pluck [zhen gou
yong ]. But a lot of them do drugs [amphetamines] and steal. Pik-khut-a
smashes their stuff and beats them if he finds out, but they keep doing
it. But that’s why the gods chose them to do this work.”
“You should have been here earlier today,” A-tiong continued; “we
really put on a show. There was blood everywhere. We were all covered
with blood; it was great. Everyone was afraid to come near us. They
[sponsors] sure got their money’s worth. Of course, they have to pay us
enough and also cover transportation and food and drink and provide
a place to rest.” Sponsors paid extra for guaranteed blood spectacle,
he told me. He went on to discuss the fine points of mortification tech-
nique: the thickness and alloy composition of the cheek-piercing rods,
whether to do a quick puncture or slowly pierce the flesh, the effect of
insertion through only one side or passing through the entire mouth,
and so on. Competition, he explained, was intense, and every group
In the 1990s, creative, spectacular, and sometimes macabre styles of

ritual mortification appeared as temples and performance troupes
competed for attention and revenue.
had to have a trick or a technique that made it stand out from the rest.
“Most of the other troupes, you know, they have nothing except more
blood. So we have to put on that kind of show, too, but of course the
patron god inspires our boys to be creative, so we get hired more often.”
A-tiong spoke sometimes as if there was no pretense of real trance; at
other times, he implied that the performances were completely authen-
tic and “orthodox” displays of medium possession, guided entirely by
the gods who had descended into the bodies of the actors.
An assistant pins spirit money onto the steel skewer piercing the
medium’s cheek; the rods of the Banners of the Five Garrisons
puncture the skin of the medium’s lower back. Note the gang tattoo.
In recent years the Taiwan media (along with education officials,

a few sociologists and criminologists, and at least one feature film-
maker—Chang Tso-chi, in his 1997 film Ah-chung) have portrayed the
Military Retainers and sorcerer troupes as breeding grounds for devi-
ance, crime, and delinquency. I would suggest, however, that these
young men, denied access to mainstream modes of achievement and
mobility, are attempting to produce themselves as autonomous sub-
Thoroughly modern medium, striking the back and hairline with a

pair of spiked clubs, protected by a black Maoshan sorcerer’s banner.
Mediums train to strike skin surfaces with high capillary density,
which bleed easily with minimum penetration, then heal quickly.
jects in their own terms. They are empowering themselves in a way

that none of the “legitimate” forms of activity available or known to
them possibly can. Their violent trance performances attest not only
to the power of the gods, but also to their own character. By flaunt-
ing their strength, endurance, dexterity with weapons, and sheer
toughness, these very marginal young men demonstrate a dangerous
unpredictability. In these seemingly uncontrolled moments of violent
trance, they are both powerfully autonomous and (especially in the
case of sorcerer troupe performers, barefoot, stripped to the waist,
and bleeding profusely) painfully vulnerable. Either way, they seem
to have an intuitive, if unreflective, understanding of the aesthetics
of performance: in satisfying the audience’s vicarious desire for pain
and blood, reveling in the spotlight, they can momentarily fulfill their
own narcissistic desire. The spectators’ awed fascination validates the
performance and valorizes the performer. The power of the act lies
not merely in the spectacle of blood, but also in the nonchalance with
which the victim wounds himself, bleeds, and wounds himself again.
Here, too, the effect depends on the marginal status of the protago-
nists as much as the extreme violence of the performance.
A-peh
A-peh was, for a time (in the early 1990s), the coach and master
teacher of the Military Retainers troupe at Eastern Prosperity Hall,
Pik-khut-a’s shrine in a suburb of Taizhong. He trained his protégés
in trancing, self-mortification, and a bit of martial arts as well. During
processions he personally took center stage as the medium for Wenfu
Qiansui, a demonic plague god, wenshen (one of the “Royal Lords,” or
Wang Ye) and the shrine’s patron deity. Young, charismatic, proficient
in martial arts, and recently married, A-peh seemed full of self-confi-
dence and social savvy. While Pik-khut-a handled the everyday affairs
of the temple and did business with associates of his own generation,
A-peh organized his own subgroup within the temple, a group of about
twenty sworn brothers in their late teens and early twenties. Most of
them were young gang runners disparagingly referred to as “bamboo
partridges” (tek-ke-a).18 All had obligations to other brotherhoods, and
some, as mentioned, had drug habits that could interfere with their
commitment to the troupe. Even with a full team of backup perform-
ers, Pik-khut-a and A-peh sometimes found it difficult to assemble the
full complement of Military Retainers and self-mortifiers. The first
time I met A-peh was just such an occasion.
I had arrived at the temple at dawn, as preparations were already

under way for the day’s activities. Pik-khut-a’s troupe had been hired
by the local Mazu temple to join its pilgrimage procession, a major
event that was to take us to Zhunan, Douliu, and then back to Tai-
zhong all in one day. Arrangements had been made well in advance,
and much face (and money) was at stake. At 7:30, as time for depar-
ture neared, two of the troupe members had still not shown up. When
they finally appeared, exhausted and crashing from an all-night crystal
meth binge, A-peh blew up. He punched out a plate glass window,
showering one of the troupe members—Gu-thau, an innocent party,
as it turned out—with glass shards. His face cut and bleeding, Gu-thau
offered no complaint but went quietly to the back to have his cuts
treated by Pik-khut-a’s wife. Pik-khut-a looked surprisingly calm and
unconcerned by the violent outburst and its unfortunate outcome: “It
doesn’t matter,” he explained to me matter of factly; “they’re sworn
brothers.”
Having finally calmed down, A-peh went himself to begin the labo-
rious process of painting the faces of the Military Retainers. Once fully
decorated and attired, the Retainers took their place at the long table
set up in front of the main altar. Seated on packets of spirit money set
on plastic stools, their ritual implements arrayed on the table before
them, the troupe sat silently in formation. Pik-khut-a, A-piang, and
some others began to chant the mantra of the Ten Great Deities (the
Retainers), accompanied by drums and gongs. After about twenty min-
utes, two of the Retainers became possessed and had to be exorcised.
Finally, the divination blocks were thrown to inquire whether or
not the deities had in fact descended. When their presence was con-
firmed by three consecutive positive throws, the chant came to an end.
Pik-khut-a then read a petition to Heaven, asking for the assistance of
Great Emperor Liu and his contingent of Retainers for carrying out
the tasks that lay before us that day. A paper talisman was burned and
the ashes crumbled into a bowl of water, from which each Retainer
drank in turn, recreating the ritual of sworn brotherhood. Each pair,
beginning with the lowly Civil and Military Yamen Messengers and
continuing to the high-ranking Civil and Military Magistrates, came to
the front and worshipped with incense, recreating the Retainers’ oath
of allegiance to Great Emperor Liu. The ritual of obeisance complete,
the Retainers were commanded to stand again. At this command, the
actors now embodied the deities, without the usual progression into
possession trance. Passing their implements of exorcism over the main
censer as they exited, they filed out of the temple in pairs. Crossing a
burning pyre of spirit money, the Retainers fell into formation and
began their ritual steps, making a circuit of the alley and returning to
make a final bow of obeisance to the shrine’s patron deity before set-
ting out on the day’s appointed mission.
After a two-hour ride in the back of a flatbed truck, we reached
Zhunan. On debarkation, the initial rituals of setting up formation
and presenting arms were repeated, with the image of Liubu set on
a cart substituting for the temple altar.19 Then the long, slow proces-
sion to the Zhunan Mazu Temple began. Since this was a pilgrimage
on “foreign” territory, for a temple affiliated with a different lineage
and a different deity, there were no house or business exorcisms. A few
“spontaneous” possessions took place at crossroads and other places
of ritual danger, but not until the troupe—now part of a long, snak-
ing parade—approached within a few blocks of the temple did the
number and intensity of possessions start to peak. It was at the point,
where the CT V (one of the national television networks) cameras first
appeared, that A-peh became possessed. This was, it turned out, an
officially designated “Tourism Folk Festival,” and the media were out
in force.
The possession began when A-peh started to stare vacantly ahead
of him, immobile. Soon he began to shake, and Pik-khut-a and others
rushed over to remove his shirt and tie a strip of red cloth around his
waist. Suddenly, A-peh broke into a crisp, fast, and powerful martial
arts form. The god who had descended was the shrine’s chief patron,
Wenfu Qiansui, who in life had been a failed military exam candidate.
With the god’s presence confirmed, the censer was brought over, and
“Wenfu Qiansui” (A-peh) was offered incense. A-peh grabbed the cen-
ser and shoveled handfuls of ash and burning embers into his mouth,
smearing his face and torso until he was entirely covered with gray ash
and black cinders. He tied another red cloth around his head, tying
tubes of paper talismans upright at the corners. After marching a few
yards using the “dipper step,” he stopped again and gestured to the
assistants, who brought him a long bronze needle. Concealed behind
the black talisman flag of the Mt. Lu branch of ritual magic, A-peh
methodically inserted the point of the needle into and through his
own left cheek.
Inserting knife blades and steel needles through the cheek and
tongue are time-honored techniques among Chinese spirit mediums.20
With the cameras rolling, A-peh, however, did not stop at simply pierc-
ing the cheek. He took the needle and bent it into a loop and stuck
the sharp end again through his right cheek. He then called for two
spike balls, which he had fastened to his heels. Finally, he called for a
sawfish blade (pai jian) . Thus equipped, he alternately walked in ritual
step and rode the god’s palanquin, gesturing commands to stop and
start the procession as he deemed appropriate.
The Military Retainers perform a variety of group maneuvers,
called by such names as “Eight Trigrams Formation” (Bagua Zhen),
Old school medium, skillfully wielding a swordfish bill club;

a traditional medium “partakes” of the spirit money offered the
possessing deity by dancing around and through the bonfire.
“Lotus Bloom Formation” (Lianhua Zhen), and “Opening the Four

Gates” (Kai Simen). Unlike the routine patrols and the lightning
attack charge used to exorcise and purify houses, these outdoor drills
are semi-choreographed performances invoked mostly at moments of
ritual climax. In Zhunan, the main performance venue was the Mazu
Temple courtyard, which was packed that day with devotees, tourists,
and T V news cameramen. The Military Retainers entered the court-
yard first and began a long, energetically performed drill formation—
Opening the Four Gates. On this day the performance went on for
almost twenty minutes as the troupe elaborated the traditional moves
with acrobatic spins, acted out the story of Elder Master and Second
Master, and fell one by one into trance as the cameras rolled.21
Having paid their respects to Mazu at the Emperor of Heaven’s
censer, the Retainers doubled back and set up a gauntlet, lining up
at the sides of the courtyard to stand guard as the two patron gods,
Wenfu Qiansui and Great Emperor Liu, made their entrance. Six self-
mortifying “sorcerers” preceded the gods’ palanquin through the gate.
Stripped to the waist, they began beating themselves on the back and
forehead with nail clubs (tong gun) , spike balls, and other traditional
and nontraditional implements of self-mortification. A-hiang began to
cut his tongue with a now popular but very nontraditional ice saw,
while Han-chi-a sliced his tongue with a classic weapon, a moon ax
blade (yue fu). The cameras captured it all from every angle.
Finally A-peh entered, spike balls still firmly bound to his bare feet.
As the din of firecrackers began a long crescendo, he strode up (still
walking the dipper step) to meet the host temple’s entranced medium,
who had come forward to greet the arriving deity. The two communed
animatedly for a minute or so. A-peh then called for a nail club and
began to spike himself in the back over and over again. Finally, the two
entranced shamans turned and marched, side by side, past the main
censer and into the temple. The crowd—and the cameras—followed.
Inside, A-peh went up to the main altar. He started pounding and writ-
ing with his fist on a table set up for just this purpose. Pik-khut-a and
a representative of the Mazu congregation stood by, interpreting the
god’s written communication.
A-peh then stood back and removed the spike balls from his feet
and the metal rod from his mouth. He stood up on the altar table,
gave out a shriek, somersaulted backward, and collapsed into the arms
of two waiting helpers. This signaled the departure of the deity and
the end of the possession. Clips of A-peh in trance and the Military
Retainers’ performance were included in that evening’s local evening

television news, which featured an extended report on the day’s events
at the temple.
I saw A-peh in trance several times over a period of three years.
While he occasionally added or deleted one or another detail, the “nar-
rative” and style of his trance performance were remarkably uniform.
His transition always began with convulsions, violent face contortions,
and foaming at the mouth, followed immediately by a fast, powerful,
and precisely executed martial arts set. At this point assistants would
rush over to remove his shirt and tie a red bandanna around his head
and the medium’s apron around his waist. As he was fed sandalwood
incense smoke, he would invariably grab handfuls of burning incense
and smear his face with soot. He would then call for his trademark cop-
per rod and insert it in his cheeks himself and have spike balls tied to
the soles of his feet. He would almost always alternate walking with rid-
ing on the palanquin, and his trances usually ended with a backward
somersault off the altar. Unlike Pik-khut-a and many of the troupe mem-
bers, A-peh remained aloof and silent whenever I tried to discuss his
trances and his career with him, and I observed that he rarely joined in
discussions of religious practice or cosmology with his older colleagues.
But in his coaching of the Military Retainers, his demand for allegiance
from the group, his “righteous” response to drinking challenges and
toasts at feasts, and his spectacular trance performances, he showed
himself to be an adept and confident handler of people and a savvy
manipulator of public attention. Hardly shy or socially inept, then, A-
peh used silence to create an aura of mystery around his entranced
self. This mystery enhanced the authenticity of his performance, yet,
paradoxically, it also served to fuse A-peh’s everyday self and his trance
persona into a recognizably consistent character.
The Military Retainers

The Eight (sometimes Ten) Military Retainers (ba/shi jiajiang; H. pat/
sip ka-chiong) is a martial ritual procession troupe style that appeared in
southern Taiwan at the end of the nineteenth century. Over the course
of the twentieth century Military Retainers troupes, mostly in southern
Taiwan, evolved an elaborate and dramatic ritual repertoire, mythol-
ogy, and social role. Once practiced primarily in the areas around
and to the south of Jiayi, since the 1970s the Military Retainers have
become perhaps the best known and most widely practiced martial
performance style on the island’s temple festival scene.22
The Retainers are one of several styles that belong to the genre
of “theatrical” procession troupes. They wear costumes and their
faces and bodies are painted along the lines of traditional Chinese
folk opera (although the style of the costumes and the particular face
painting patterns emulate the iconography of temple gods rather than
the conventions of theatrical roles).23 Along the routes of major pro-
cessions, the Retainers march in synchronized formation, stopping
along the way to exorcise any death pollution and ghosts that they
encounter. Such exorcisms often occur on highways near blind curves,
dangerous street intersections, unguarded canal bridges, and other
places where fatal accidents and suicides may have left the dangerous,
polluting residues of death. They also charge into houses and busi-
nesses (at the owners’ invitation, of course) to purify the premises of
any unclean things that may be lurking and causing ill health, dishar-
mony, or financial difficulties within the household.
The Retainers maintain an austere silence as they march, and they
are forbidden to eat, smoke, drink, or chew betel while in character.
Non-temple personnel and females generally are kept out of their
ranks.24 Occasionally, however, Military Retainers may be directly pos-
Dongxing Shrine’s Military Retainers perform the Opening the Four Gates formation.
sessed by their patron deities, going into a sometimes violent trance.25

Like mediums, Military Retainers troupe members are required to
abstain from sex and avoid certain foods (especially beef ) for at least
three days prior to the ritual. But unlike mediums, who are under-
stood to embody divinity only when in the throes of possession trance,
the Retainers become fused with their patron spirits through a ritual
of separation, remaining sacralized (that is, ritually off limits to all but
the temple’s own ritual assistants and senior members) throughout
the procession. Having been once encostumed as a Retainer endows
an actor with long-term supernatural protection. The mask of face
paint that he wears of a particular deity, for instance, is said to inhere
on the inner surface of the face of the actor for the rest of his life.26
The actors, then, are “possessed” by the deities whom they represent
as soon as they are costumed and ritually separated. While Retainers
may be possessed during processions, trance is therefore not a neces-
sary mark of sacred presence. The identification of the actor and the
deity is enhanced, of course, if the actor shows proficiency in the ritual
moves and becomes possessed at critical moments.
During the 1980s, as noted, the Military Retainers (along with the
above-mentioned sorcerer troupes) gained an unsavory reputation
among the general public after education officials and the press por-
trayed the troupes as bands of juvenile delinquents, associating them
with violent motorcycle racing gangs (biaochezu) and gangsters (cf.
Wang Yali 2001; Zhuang 2004). The reality, of course, was far more
complex, but the Military Retainers are indeed very much part of the
martial ritual performance tradition that draws its personnel, tech-
niques, and cultural ethos from the “rivers and lakes.” 27 The troupes
of the 1980s and ’90s were a perfect nexus of wu masculinity, jianghu
lifestyle, martial cults, and the Taiwanese working-class economy.
An Inside Story
Beginning in 1990, as mentioned in chapter 2, I was privileged to par-
ticipate in one of Taidong’s best-known Military Retainers troupes,
playing the role of the Civil Magistrate for Loyal Harmony Temple. On
several occasions, I found myself “working” opposite A-kiat-a, who soon
became one of my most enthusiastic informants. A-kiat-a’s identifica-
tion with one role in the Military Retainers troupe, the Military Magis-
trate, has defined his public persona for more than twenty-five years.
While many of his peers have become embroiled in the wrangling
and power intrigues of temple politics, A-kiat-a’s quiet, self-effacing
manner has enabled him to remain mostly disengaged from the con-
troversy and personal rancor. The Military Magistrate, like A-kiat-a
himself, stays in the background most of the time. But the Military
Magistrate is an enforcer, and as such, he is expected to act forcefully
and demonstratively when the situation calls for it.
During the first night of the 1990 Lantern Festival procession, as
the Military Retainers reached a storefront shrine along Zhonghua
The author (seated right) as Civil Magistrate during a rare break in the action on the
second night of the 1990 Taidong Lantern Festival Procession.
Road, A-kiat-a seemed agitated. A ritualist carrying a censer came out

of the shrine to greet the troupe and offer incense to the Retainers.28
As the front ranks finished imbibing their share of incense smoke and
proceeded on to acknowledge the shrine’s deity, A-kiat-a, marching
in place, fell into a full-fledged trance. When his turn came, A-kiat-a,
his head shaking rhythmically from side to side, his face fixed in the
magisterial frown of a stern and angry deity, leapt forward, grabbed the
censer, ate several mouthfuls of burning incense, and kicked the cen-
ser bearer to the ground. As A-kiat-a coiled into an attack posture and
hovered threateningly over the still supine ritualist, the possessed spirit
medium of this temple came out and, moving slowly through the rou-
tines of a martial arts form, offered half-hearted, symbolic resistance.
Within a few seconds, a friend of A-kiat-a’s, a ritual specialist from
Taizhong, jumped in and recited a spell. As he channeled his spell-
empowered qi through a two-fingered thrust of his right hand, A-kiat-a
fell back into the arms of two waiting assistants and withdrew from the
trance. A hurried explanation was offered the offended temple’s rep-
resentatives, the incident passed, and the procession moved on.
Inscribed within both prescribed ritual and the spontaneous, “in-
character” possession modes of Military Retainers performance, we
can see a competitive, violent side of Taiwanese social reality. When a
Retainers troupe encounters a similar troupe from another temple, for
instance, each member must hide his face, for meeting the gaze of his
counterpart in the other troupe would compel him to fight. Although
real fights rarely occur, one can often see troupe members become
suddenly possessed and wildly agitated when such an encounter hap-
pens. The Military Retainers, my informants often enjoyed telling me
sotto voce, are “celestial hooligans,” enforcers and hit men working for
Wang Ye and the plague gods Wufu Dadi. Stories of fistfights between
rival temples in years past are thus part of the lore of every established
Military Retainers troupe in Taidong.
This was the first time I had witnessed an act of real violence in the
context of ritual, and naturally I was anxious to discover the cause of
the incident. In conversations a few days later with A-gi and other tem-
ple regulars, I learned that the storefront temple on Zhonghua Road
had been founded about ten years earlier by a former member of the
Loyal Harmony Temple who had departed angrily over some unspeci-
fied disagreement. A-kiat-a’s violent rebuff was clearly intended to
reprimand the renegade temple members for their transgression and
make it clear that once they left the fold, they had lost the sanction of
the deity and had no further access to his numinous power or to the
advantages of belonging to his fold, including the security of alliance
with the mother temple group.
This performance was clearly meant for those few temple elders
for whom the schism would have significance, most of whom were
present that evening. Few of the younger members of the troupe were
aware that the two temples had a history of antagonism. Yet A-kiat-a’s
outburst was still meaningful for them, but in a very different way. As
soon as he struck the incense bearer, the other members of the troupe
responded immediately by halting their forward progress and clos-
ing ranks, ready for action if necessary. What that action might be no
one would say, but another incident involving A-kiat-a revealed one
possibility.
The long-established Heavenly Official Temple is one of the few
temples in Taidong that can be described as a “territorial cult.” Nearly
all of its devotees are residents of the neighborhoods immediately adja-
A Military Retainer is restrained and “detranced” at the end of a violent and

aggressive episode of possession by the “Monkey General.”
cent to the temple, and both the temple leadership and its procession
performers are comprised of local young men—among them many
“toughs” and gang members—hailing from these neighborhoods.
A-kiat-a’s temple, although its congregation is more diverse than the
Heavenly Official Temple’s, was also associated with one group of
“sworn brothers” for most of the 1950s and ’60s. Although that group
is now largely dispersed, the hostile, competitive relationship engen-
dered by those “turf wars” has become a reified given of both temples’
“processional identities.”
Found in Chenghuang (City God or God of Walls and Moats)
temples in Taiwan, the generals Xie and Fan—Seventh and Eighth
Masters—serve Chenghuang by escorting across the Naihe Bridge the
souls of the recently deceased and any wandering ghosts they may cap-
ture. The highlight of the Heavenly Official Temple’s procession is the
performance of the two large wood-frame effigies of these fearsome
demigods.
On the morning of the second day of the 1990 Lantern Festival
procession, the Military Retainers had been invited to purify a small
dry-goods shop just down the street from the Heavenly Official Tem-
ple. As it happened, the Heavenly Official Temple’s procession troupe
was only starting its morning rounds and had just exited the temple
gate and turned the corner. At that moment, most of the Military
Retainers had already entered the shop and were in the process of
purifying the premises. In this ritual, called “securing [lit. subduing]
the residence” (zhen zhai; H. chin-theh) the Civil and Military Magis-
trates do not enter the premises but stand guard outside. A-kiat-a was
thus in position to see the Heavenly Official Temple’s Eighth Master
as the other group approached. He became possessed suddenly, and
striding angrily toward Eighth Master, he confronted the wood-frame
demon, shaking his feather fan threateningly in its face. As the rest
of the troupe members finished the purification and exited the shop,
they saw the commotion and headed into the fray. Seeing the situation
getting out of hand, procession leaders from both groups ran in to sep-
arate the troupes. After applying a talisman to bring A-kiat-a out of his
trance, some of our group guided the now listless A-kiat-a back toward
the shop, while others exchanged greetings and explanations with the
Heavenly Official Temple leaders. The procession disappeared down
the street, and A-kiat-a was released from his trance.
In both of these incidents, A-kiat-a’s violent trance performance
was a declaration of his temple’s place in the local temple hierarchy,
an assertion of group cohesion in the face of social disintegration and

intergroup rivalry. Yet it was also a creative and productive act emanat-
ing from and reflecting upon one individual. In the first case, A-kiat-a
was performing primarily for the inner circle in a situation whose out-
come was hardly in doubt. The storefront temple would have been at a
severe disadvantage, both in terms of manpower and the moral assess-
ment of the community. There was little real risk, although had anyone
in the other group been moved to rage in response to the humiliation,
the situation could have degenerated quickly. In the second incident,
however, the other group outnumbered our own; the two groups had
a history of violent confrontation that included a number of armed
brawls, and we were on the other group’s turf. A-kiat-a’s challenge in
this case was truly a risky move, yet it had perhaps even more value
just for that reason. As some temple insiders (including some of the
troupe members) later described it, terrified and overwhelmed, the
Heavenly Official Temple’s Eighth Master acknowledged A-kiat-a’s
Military Magistrate as his teacher, affirming, at least in this interpreta-
tion of events, the superior position of Loyal Harmony Temple. Were
A-kiat-a acting on his own, this would have been a foolhardy show of
bravado, but understood as the actions of a powerful and determined
deity, temple members saw it instead as a higher-order sanctioning of
their own version of the local temple hierarchy. The fact of possession
testifies to A-kiat-a’s special relationship to the temple’s supernatural
patrons—to Wang Ye and to his own character, the Military Magistrate.
The contents of his trances, furthermore, allow him to create a special
role for himself within the group, that of a qualified arbiter of social
and moral issues pertaining to temple mythohistory. In both incidents,
A-kiat-a was stepping up to fight for his group, but he was also demon-
strating his intimate knowledge of temple history and showing off his
martial arts prowess.
Such spectacular acts, by making mythohistory tangible and vis-
ible, are critical events in the production of both the collective and the
individual actors. Here, an aesthetic moment must be understood as a
kind of quantum event: the performer improvises on a set of changes
known to all the players and listeners; at the same time, the players and
listeners become a collective unit in their intersubjective realization of
commonality in the very act of providing the ground for the player to
improvise upon. In this sense, the spontaneous possession trance of
the ritual actor is an improvisation on a set of changes, within a set of
parameters comprising the discourse of imaginable possibilities. The
actor creates himself by focusing the energy of the collective upon

himself, while at the same time producing the collective, which, in this
moment, depends on the charisma of the actor to reproduce itself as
a self-identified collectivity.
A-kiat-a’s actions would be unacceptable if seen as willed or vol-
untary, but if the agent of the gesture is a deity, then the actions take
on a completely different meaning, as does A-kiat-a’s life and person
out of character. The act manifests and affirms a shared collective
narrative, as well as identifying A-kiat-a as an essential element of the
drama. By eliciting the collective recognition of a shared narrative,
members of both groups experience a moment of definition. A-kiat-a’s
spontaneous act immediately reproduces and defines the two groups
along their lines of connection and antagonism. A-kiat-a’s own group
certainly recognizes the creative style of the actor himself, as conver-
sations with temple insiders have confirmed on many occasions. Yet
at the same time, they must misrecognize the agency of the act; simi-
larly, A-kiat-a cannot but affirm (and I have no reason to think that he
doesn’t believe with complete conviction) that his outbursts are the
intentional acts of his patron deity working through him as a possessed
medium.
A-kiat-a lives, in some measure, through his identification with
this persona, which comes to be through creative, communicative
acts within the temple milieu. The temple group, in turn, comes to
be through a process of inversion whereby the individuals that com-
prise the group produce the group through the labor of their prac-
tice—practice that emerges from participation in and is limited by the
expectations of the collectivity.
A-Cai and Big Brother Kang

Over the course of a two - or three-day temple procession, performers
occasionally rotate in and out of the Military Retainers troupes. Some-
times this turnover is planned. For instance, during annual festivals like
Taidong’s Lantern Festival procession, the performance conditions for
the daytime “patrols” are particularly demanding: even in February,
afternoon temperatures can reach into the mid-80s. The rituals are
repetitive, performed for individual households rather than crowds of
onlookers and fellow enthusiasts, and the excitement and intensity of
the evening events is noticeably lacking. Performing during the heat
of the afternoon requires considerably more stamina and carries less
prestige than the more important and cooler evening sessions. For
these reasons, the more experienced performers are usually spared

the daytime ordeal. Instead, new recruits and younger performers are
pressed into service during the day. The troupe returns to the temple
at the end of the patrol, ideally just as the evening meal is set out in the
back room. After eating and resting, the older, higher-status “generals”
begin preparations for the night patrol, a time of intense competition
for prestige among the participating groups.
Yet no matter how well planned the rotation schedule, unplanned
substitutions are almost inevitable. Over the course of a day-long march
it is not unusual for performers to suffer some minor but incapacitat-
ing injury; leg cramps, pulled muscles, sprains, fatigue, and dehydra-
tion are the most common ailments. As evening approaches, passions
run high, and troupe members fall into possession trances with greater
frequency. Particularly powerful and violent possessions can provoke
injury or fatigue so severe that a performer simply cannot continue. In
such cases, a substitute actor must be found and prepared as quickly
and discreetly as possible.
Finding a willing individual with sufficient training and experi-
ence in a particular role is the first task; this is rarely a problem for
the professional troupes, who can usually muster at least a couple of
“understudies,” even when traveling away from their home turf. But
for voluntary troupes, even those who field “rookies” during the day
and “old generals” at night, it may be difficult to locate a qualified
performer at the critical moment.
One way or the other, after someone is found to shoulder the task,
the next step is to get him ready and insert him into the procession.
Getting a performer dressed, painted, and ritually prepared takes con-
siderable time, especially for those playing the core roles in troupes like
the Military Retainers that feature elaborate theatrical face painting.
Moreover, a full “theater chest” is a major expense: costumes are cus-
tom-ordered, and an entire set, including headgear and implements,
can cost thousands of dollars; with no exception of which I am aware,
military temple procession troupes, whether voluntary or professional,
own but one set of costumes and implements.
When a substitution does become necessary, the troupe rests while
the rest of the procession continues. If at home, the injured performer,
substitute, face painter, and sometimes a ritual expert are rushed back
to the temple to make the switch. If too far from home base, every-
thing is done on the spot or (preferably) in the recesses of a friendly
sister temple nearby.
I got to know A-cai during just such an unplanned troupe substitu-

tion during Loyal Harmony Temple’s 1991 pilgrimage to Dong’gang
for the triennial plague boat burning festival. Late on the afternoon
of the second day (this procession lasts three days) A-liong, who was
performing Gan Ye, one of the critical Front Four, went into trance
in front of a small earth god shrine. This shrine stood across a small
road that ran along the bank of a drainage canal. At the end of his
trance, A-liong came up lame, an injury apparently instigated by the
malicious soul of an old woman who had drowned herself in the canal.
Ten minutes of rubbing and liniment failed to improve the situation,
and it became apparent to everyone that A-liong was out of the game
for the day.
A-tiong, O-ke, and A-beng, who were handling the logistics that
day, debated what to do. There were five Retainers “kids” (ka-chiong
kin’na) not in costume that afternoon, but three had already gone off
with A-chhiu in his pickup truck to replenish the temple’s food and
water supplies. Neither of the remaining two had any experience play-
ing the demanding role of Gan Ye.
The troupe pulled out of the procession line while A-beng and
I headed to Gongxing Tang (one of Loyal Harmony’s sister temples
in Dong’gang), hoping to find someone who could jump in and take
over as Gan Ye. When we arrived, we found only the caretaker. But just
as we were about to leave, two young men rode up on a scooter and dis-
mounted. A-beng recognized the passenger as A-cai, who had grown
up in Taidong and trained as a Military Retainer a few years before.
After a quick exchange, A-cai agreed to help out. His driver, given
directions, raced to pick up A-liong (the injured Gan Ye) and bring
him back to the temple. Meanwhile A-beng began painting A-cai’s face.
By the time A-liong arrived (riding on the back of the scooter in full
Gan Ye regalia), A-cai was ready to take over. As soon as the costume
transfer was done, A-beng “initiated” A-cai at the temple altar and sent
him off again on the scooter to join the procession.
A-cai was a native of Taidong who had graduated from a local
vocational high school. One of his uncles was a devotee of a small
Mazu temple that had close relations with Loyal Harmony Temple.
There were few young people at the Mazu temple, so A-cai would often
accompany his uncle to Loyal Harmony Temple in the evenings, where
he got to know some of the temple’s local “gang,” finally becoming a
sworn brother of the then teenaged cohort. Despite occasionally join-
ing the Loyal Harmony brothers on their evening forays, A-cai was a
“good student” in his teachers’ eyes. Even more unusual for the time
and for the present company, he attended an English cram school a
couple of evenings a week and had an obvious talent, if not enthusi-
asm, for learning foreign languages.
During his military service, A-cai was stationed at the Zuoying Naval
Base, just north of Gaoxiong. During that time he met Big Brother
Kang, a friend of his uncle through the Mazu pilgrimage network.
Kang, who ran a string of small brothels near the navy base, had orga-
nized a Military Retainers troupe out of his own pocket and paid for
the construction of a small side shrine for them at the Xinglong Hall,
a local Mazu temple. With some Military Retainers experience, A-cai
was a welcome visitor to the shrine on his off-duty days.
When A-cai returned to Taidong after finishing his military service,
he found he had “outgrown” his hometown. As he later told me, “The
place is so small and backward, there was nothing for me to do here.
I would go to Gaoxiong every couple of weeks, and after a while, Big
Brother Kang asked me if I wanted to work with him.” So A-cai moved
back to Zuoying, helping Kang with his business and spending time at
the shrine when he could.
For Big Brother Kang, A-cai’s participation in his Military Retain-
ers troupe was an important part of their relationship. When he first
arrived, A-cai had already had a couple years of training and willingly
passed his knowledge on to the others (at that point, the troupe had
no full-time teacher or coach). Most important, however, Kang saw A-
cai’s interest in religion and ritual practice as a sign of his character,
and as a “disciple” at the temple hall as well as a junior sworn brother
in his gang, A-cai was obligated to Kang more durably than many of
his other minions.
I once dropped in on Big Brother Kang on my way back to Taidong
from Taipei. Even though it was only late afternoon and I was anxious
to get moving (I had a long, dark drive on truck-infested mountain
roads ahead of me), Kang insisted I stick around for dinner. Such invi-
tations almost always entailed some hearty drinking, which meant that
if I accepted, I would have to stick around until the next day. The con-
versation was going well, however, and I decided to take advantage of
Kang’s hospitality. That evening, after a few rounds of food and beer,
Kang, who usually played pretty close to the vest, began to praise his
protégé A-cai. “This kid is my best ‘little brother,’ better than these
Gaoxiong kids. Taidong is a simple place; the kids from there aren’t so
‘complicated’ like the city kids. And A-cai, he’s tough, and he’s loyal.”
Later he confessed he had once had a minor problem with A-cai when
one of his “girls” took a liking to the young man. The relationship
caused friction between A-cai and his comrades but especially irked
Kang, who was obviously ambivalent about A-cai’s sexual impropriety.
“Of course, he’s a man, a fighter and a player. I finally had to move
him over to enforcement. But he understands loyalty and honor. You
can’t let a woman cause a split among brothers. Past is past; you have
to know how to let it go. A strong man knows how to take it or leave it”
(ti de qi, fang de xia) .
Honor above All

Honor (yiqi) for the brothers of the path often means risking one’s own
hide to stick up for comrades. As Brother Qiu explained it, “Friends
are different at different stages of your life. In middle age, it’s the
ones who understand what you say, even without saying it. But in your
teens, it’s all about honor.” For the brothers, honor (yiqi) also entails
generosity with money. Yet cash flow is a focus of anxiety for men in the
jianghu, given their unstable sources of income and the expectation
that whenever flush, a man should be generous with his sworn broth-
ers. Outlaw masculinity is tied to “face” as much as if not more than are
“civilian” modes of masculinity. For instance, owning a BMW or Mer-
cedes (the brands of choice for gangsters in Taiwan since the 1980s)
advertises status, ability, and prowess. But “face” is actually more than
surface. A BMW owner who doesn’t act like one—that is, who does not
spend his cash and treat his associates with the generosity expected of
someone who owns a hot, fast, luxury sedan—will soon be unmasked.
Face is “given” and “lost,” negotiated through the give and take of
social performance. The obligations of social reciprocity, that is, are
the central arena for the performance of masculinity. A man of prow-
ess is bighearted and generous; stinginess is by itself sufficient cause
for ostracism. Then again, in the jianghu generosity is a relative, not an
absolute, moral value: it makes sense to be generous only if and when
you have the means.
Gau-a, for example, became a big shot in Taidong’s jianghu when
he became one of the most successful local bookmakers at the height
of the Dajiale (Everybody’s Happy) numbers racket craze of the late
1980s.29 Then, over the course of a few weeks, betting his own funds
on the numbers, he lost everything. He explained his misfortune as a
combination of fate and inescapable jianghu imperatives: “ You know,
because of that word yiqi I ended up exchanging my new white Mer-
cedes for a used black bicycle. Only those who have been ‘brothers of
the way’ can understand how I feel. I lost half a billion; what face do I
have around here anymore?”
Yet implicit in his lament was another message: his commitment to
that basic principle of gangster society, honor and reciprocity, showed
him to be a real man despite his bad luck. Explaining his losses as a
fault of his own upright moral sensibility, Gau-a was at least able to
claim his status as a true wu-masculine man. He had lost his social face
(mianzi) but not his moral face (lian) . In the end, he proclaimed him-
self free of regret:
Easy come, easy go [qiancai a, jin de kuai, qü de ye kuai]. I’m leaving

Taidong, going up to Taipei. I have a relative who has connec-
tions at the Gong’guan wholesale market; I can get started selling
fruit. As long as I work hard and keep my dignity, I can get back
in the game someday. I’m still standing; no one can say I’m not a
righteous man, a good brother of the way. When I got into this,
I already knew the path was a “sea of blood, a precipitous cliff”;
who is there to blame? It’s fate; you just have to start over; if you’re
a righteous man, if you have self-respect, people will also respect
you. If you lose self-respect and honor, what else is there?
Pik-khut-a
Pik-khut-a is the “hall master” (tangzhu, thng-chu) of a small storefront
plague god shrine in the suburbs of Taizhong. Although this temple
was “divided off ” from a temple near Gaoxiong, Pik-khut-a takes his
retinue of Military Retainers and sorcerer troupe back to Taidong each
year to join Loyal Harmony Temple’s Lantern Festival procession.
Pik-khut-a’s father was a demobilized mainlander soldier from
Guangdong; his mother was Taiwanese. During his twenties he was a
member of an organized mainlander gang, a fact attested to by the
dragon tattoos on his arms and back. However, he spoke Holo most of
the time and nearly all his associates were Holo-speaking Taiwanese, as
were all his assistants and members of his temple procession troupe. As
something of an outsider on the inside and perhaps due in some mea-
sure to his mainlander gang experience, Pik-khut-a was especially con-
scious of jianghu lore and the code of honor. Written into the compact
of friendship among sworn brothers, as exemplified by the practice of
Pik-khut-a and his associates, is an imperative to constantly challenge
and test each others’ generosity, loyalty, and appetites. The most popu-
lar venues for such competition are restaurants, karaoke clubs (KTVs),
and festival banquets.
It is not just the ability to hold one’s liquor that is at stake in these
almost nightly competitions, but the willingness to continue drink-
ing at your hosts’ and friends’ invitation even after going far beyond
your limit. Having a “good time” is not the point. During an evening
out with one’s “brothers” and associates, status is subtly affirmed or
reassigned. Business is often the subtext and sometimes the pretext
of these evenings, but it is rarely discussed at these times, particularly
when it involves activities on or beyond the verge of legality. Through-
out, the host seeks to build his relationships with his “brothers” (or
clients and potential clients) by establishing himself as a man of heroic
ways, an exemplar of “loyalty and reciprocity.”
One day in May 1991, Pik-khut-a went up to Taipei to lend a hand
at Gongrong Temple as its members joined the annual Mazu proces-
sion in nearby Xindian. When he arrived, accompanied by an assis-
tant, it was clear he was still suffering the aftermath of the previous
night’s carousing. Late in the morning, as we walked together behind
the god’s sedan chair, Pik-khut-a explained, “Last night we were drink-
ing XO; that client is really overzealous. Liubu got me; he grabbed me.
I really fear getting possessed by Liubu; he made me throw it all up.
Whoo, I’m exhausted today.” “Liubu got hold of you?” I asked. “ Yup.
I threw the whole thing up.” “ You’ve still got energy and spirit today!”
“Well, I have to; that’s the way it has to be.”
The procession lasted from early morning until late in the evening
of a clear, hot, and unusually dry early summer day. Pik-khut-a worked
especially hard, carrying the “tools of interrogation and punishment”
(xingju) , to lead the Ten Military Retainers through their gruelingly
repetitive house-exorcising rituals all along the route. The role of hall
master requires marching in step and carrying a bamboo or wooden
yoke laden with miniature replicas of the implements of torture found
in the late-imperial state. At ritually important points along the proces-
sion route and at the houses of supplicants who request the exorcism
of their premises, he leads the Military Retainers through a series of
dramatic ritual actions. Doing this properly and efficaciously requires
a long series of low horse- and bow-stance (mabu, gongbu) steps. This
entails martial arts expertise, strength, flexibility, and endurance, and
Pik-khut-a, with his extensive martial arts background, is one of the
best I have seen performing this role. Later that day, as the proces-
sion neared the intersection of two main roads, Pik-khut-a, nearly
exhausted by his exertions, was suddenly and violently possessed by his

patron deity, Wenfu Qiansui.30
Well after nightfall the group returned to the home temple. After
drinking far more than required at the celebratory feast that followed,
Pik-khut-a stood up and announced his departure for Taizhong. As
he prepared to leave, one of the temple committee representatives
offered him a “red envelope” filled with N T$10,000. Laughing and
appearing surprised, he refused to accept the money, pushing the
envelope away repeatedly with a show of distaste.
The dynamics of this exchange are complex. Pik-khut-a’s liveli-
hood actually depended on the demonstrated supernatural efficacy of
the deity in his storefront shrine, and he maintained a professional
Military Retainers troupe that hired out for cash without apology. How-
ever, as his frequent testimonies to the miraculous power of the deity
attested, he strongly believed that his own fortunes were bound up
with Wenfu Qiansui (often referred to with the honorific Wang Ye), his
patron god. According to Pik-khut-a, his success or failure in business,
which at the time included both wholesale soft drink distribution and
gambling, depended entirely on the graces of the deity. At the same
time, he pointed out (presumably without irony) that Wang Ye’s power
depended in turn on the temple’s notoriety and on the number of dev-
otees who came to burn incense, enter into reciprocal obligation with
the deity, and join the temple’s pilgrimages and processions. Pik-khut-a
had been very successful at drawing a strong and consistent following
into his storefront cult, and his Military Retainers and spirit mediums,
who enjoyed a growing reputation for spectacle, were certainly a point
of attraction for many believers. Involved as a teenager in a more insidi-
ous form of brotherhood, his upper body bore the trademark dragon
tattoos. His business activities and connections extended in many
directions and occasionally crossed the lines of legality, although he
was decidedly not engaged in any violent or coercive criminality.
All this was well known to the Gongrong Temple committee mem-
bers, who saw themselves, in contrast to the “professional” religionists
like Pik-khut-a, as humble, devout sacrificers. They were mostly small
businessmen engaged in leather goods manufacturing, and, as they
liked to emphasize, gained no financial profit from their religious
exertions. Quite the contrary: they put up their own funds to sponsor
feasts, processions, and other activities, and they took no outside con-
tributions, not even for successful cures rendered through shamanic
healing sessions.
Pik-khut-a had to work hard to defuse his negative public image.

He had come, he wanted the Gongrong Temple group to understand,
through an obligation of service to the deity. It was an implicit over-
ture of friendship and cooperation between his temple and theirs. His
refusal of cash compensation for his labor obligated the Gongrong
Temple members to payment in kind (that is, lending a hand and giv-
ing face) sometime in the future. Their reluctance to enter into such a
bargain suggests that they might have been somewhat wary of Pik-khut-
a and his gangster reputation. By firmly and absolutely rejecting the
offer of compensation, Pik-khut-a demonstrated, to himself and the
Gongrong group, that he was, after all, a righteous man and a devoted
servant of the deity, not a hoodlum trying to exploit Xingfu Qiansui’s
(Gongrong Temple’s patron deity) good reputation.
Pik-khut-a’s cavalier refusal of the money can be seen as an act
through which he defined himself as an individual person as well as
a social subject. Without guile—for I never detected any cynicism or
conscious deceit in his self-presentation—Pik-khut-a strove to be a man
of integrity in the tradition of the Chinese knight-errant. His refusal to
accept payment defined him as a righteous brother, but it also marked
a trajectory of desire that obviously diverged from the Gongrong Tem-
ple group. The offer, followed by the refusal, affirmed, rather than
deflected, the group’s view of him as an outsider and member of the
criminal underworld and of themselves as upwardly mobile but hon-
est businessmen. While each party had its own perspective, this inter-
change demonstrated a collective subjectivity constituted, in part, by a
shared set of mythohistorical narratives.
Prodigal and Reluctantly Filial Sons

At Taidong’s Loyal Harmony Temple, a devotee who has transgressed
may expiate his sins by performing a public act of “self-rectification”
(zizheng) , a ritual that, like the imperial magistrate’s tribunal, combines
confession and penitence. He must first proceed to the side temple,
the shrine to the Military Retainers. He throws divination blocks to
ascertain the severity of the punishment, then takes hold of one of the
lengths of split bamboo that are provided and beats himself with the
requisite number of strokes.
Stories of those who had violated some moral precept or other and
been moved to perform this public act of penitence were part of tem-
ple lore. Every so often, one of the older devotees would point out the
split staves to me and relate one or another of these instructive cases.
Some I heard repeated several times; one young man, directed by the
temple management committee to deposit a sum of “incense money”
collected during the visit of a pilgrimage group from Hualian, lost the
entire sum gambling the very same evening. He rushed to the temple
and, in an act of “heroic” contrition, beat himself so violently that the
bamboo cracked (this talisman is preserved behind the altar, attest-
ing to the piety of this devotee and to the scope of the deity’s power).
Another case recalls how A-beng, then in his early twenties and one
of the central actors in Loyal Harmony’s Military Retainers troupe,
had stopped into a local brothel the day before a ritual performance
(violating the obligation to remain pure for at least three days before-
hand). Suddenly he was seized by an invisible power that hauled him
back to the temple and forced him to his knees in front of the altar
to the Emperors of the Five Blessings. Several different informants
told me slightly different versions of the story, some adding humorous
embellishments (such as describing the sight of A-beng tripping over
his own half-pulled-up trousers as he came running in from the street),
but most were in agreement about what happened next. He crawled to
the side chapel, where the bamboo staves were kept, grabbed one in
his right hand, and began beating himself sonorously on the head and
back until, after perhaps ten minutes of “merciless punishment,” two
temple elders rushed over and restrained him.
While such stories are rarely told by the transgressor himself, nei-
ther are they denied, for they attest to the close personal relation—
sometimes explicitly described in terms of the filial relations between
fathers and sons—between the deity and his devotee. If A-beng hap-
pened to be present when the story was being told (and it was usu-
ally his presence that elicited the telling), he would simply listen from
a distance, expressionless, chewing his customary wad of betel. The
popularity of the story reflected the popularity of the person and indi-
cated his central and continued relevance as a core member of the
temple group. Moreover, being identified as a mischievous character
alludes to the rebellious masculinity of A-beng’s younger days, clearly
an image that he prefers within his own circle of sworn brothers.
In his fifties, A-beng was still known as a man who had little con-
trol over his appetites, and his peers frequently teased him about his
love of drink and the obvious pleasure he took in food (he is short
and stocky and sports an impressive, and oft-noted, beer belly). This
teasing, as well as A-beng’s patient if weary response to it, reaffirms
A-beng’s special relationship with both the patron deity and his fel-
low devotees. In the story of A-beng’s self-rectification, he is a favored,

though wayward, son of the deity. In punishing himself (although in
theory it is the power of the deity, not his own agency at work), he per-
forms simultaneously the roles of stern punishing father and contrite
filial son. In taking preemptive action, A-beng effectively precluded
the imposition of possible further sanctions by any other hand.31 The
retelling of the anecdote, however, implies that there is a collective
sense that A-beng’s capricious and irresponsible tendencies are not
completely benign and must be held in check. A-beng seems to have
accepted this arrangement: well into middle age, he hasn’t stopped
acting up, nor has the teasing ceased.
Lau-hoe-a
My friendship with Lau-hoe-a was based partly on our mutual interest
in the lore of popular religion but even more so on our shared enthu-
siasm for the martial arts. We would often discuss technique, trading
anecdotes or even engaging in light sparring play. A lively, animated
orator, he habitually punctuated his stories with dramatic facial expres-
sions and demonstrative gestures. His mannerisms and carriage would
reveal to any experienced practitioner of the martial arts that he was a
well-trained fighter.
Lau-hoe-a’s fanatical devotion to martial arts is one of his favorite
themes. When he was younger, he would put on his cloth-soled gongfu
shoes, take a light lunch and a canteen, and board the early morn-
ing train to Guanshan. From there he would walk the forty kilome-
ters back to Taidong along the provincial highway, shadow boxing and
punching and kicking trees and signposts along the way. He liked to
emphasize that he wore out over twenty shoe soles this way. He often
tested his skill in the streets and used his fighting skill and intelligence
to establish a reputation among the Taidong gangsters as someone
who could be both counted on in a fight and trusted with important
business.
Lau-hoe-a is an unusually self-reflective and candid speaker who
has an almost psychoanalytic understanding of his own actions and
motives (although he tends to frame his confessions in the idiom of
conventional, broadly Confucian morality). During quieter moments,
sometimes in the early evening over tea in the alley outside ritual mas-
ter A-hok’s shrine, or at 3 a.m., grabbing a snack at a downtown street
stall after he closed his pachinko parlor, he would wistfully describe his
personal transformation from a violent and thoughtless hoodlum into
a responsible husband, exemplary father, filial son, good citizen, and

devoted disciple of his patron deity, the Military God of Wealth.
As a teenager, he was angry and violent and frequently picked
fights. One day after school, he was walking along a dirt path when
two Amis men on bicycles almost ran him down. He chased them,
and despite the fact that there were two of them and they were broad,
strong field laborers (Lau-hoe-a himself sports a slight, wiry build),
he chased them down, knocked them off their bicycles, and trounced
them so badly he was afraid he had killed them. As Lau-hoe-a told me
on another occasion, much of his anger came from his hatred of his
own gangster father, whom he had watched beat and then abandon his
mother. Yet when his father went on trial for stabbing his live-in girl-
friend in a fit of jealousy, Lau-hoe-a prayed to his patron god to inter-
vene. The judge on the case had a reputation for being severe, but in
the end, his father got a very light, suspended sentence. In such stories,
Lau-hoe-a turned confession into a testimonial to his favored relation
with his patron god and a proclamation of his own moral straightness.
He was a violent, rebellious young man who overcame his selfish anger
and was transformed into a filial son (even if his transformation was
mostly visible only to himself—he prayed for his father’s life but after-
ward continued to fight and get into trouble with the law).
At the back of Lau-hoe-a’s pachinko parlor was a “VIP Room”
where he had set up an electronic horse racing game. This was essen-
tially a gambling table, made technically legal by the substitution of
“prizes” (such as cartons of cigarettes, which could be resold) for cash
winnings. After one particularly lucrative late-night session, when the
last (and finally broke) patrons had left, Lau-hoe-a invited me and his
two assistants out for seafood and fried noodles. During the meal, Lau-
hoe-a outlined his plans for the future, which included setting up a
fund for the families of prisoners. He explained: when he was eighteen,
serving a two-year sentence in the reformatory (for nothing more than
his reputation, his underworld associates, and his tattoos), he had to
face his weeping mother when she came to visit him. His sense of filial
duty shamed him into vowing to lead a proper life when he got out (a
vow he eventually honored a dozen or so years later).
A second vow to reform himself came after Lau-hoe-a had observed
others endure the suffering of visiting day. He began to empathize
with the mothers, wives, and young children who broke down in tears
over their incarcerated sons, husbands, and fathers. Should he ever
succeed financially, he told me, he wanted to help out such families,
not just to ease the pain of the innocents, but also to give ex-convicts a
real chance to turn their lives around after their release. He had been
able to revise his own fate, he explained, partly thanks to his patron
god’s intervention but also because he had more discipline and opti-
mism than most who turned to the “dark path.” Moral virtue, he often
reminded me, was rewarded with wealth and good fortune. Just as defi-
ance and violence had defined Lau-hoe-a’s masculinity as a youth, his
conscious, public rejection of violence and expressions of social con-
cern now signaled his desire for a mainstream and exemplary kind of
manhood.
Loyal Disciple, Devoted Father

Yam has been part of the Loyal Harmony Temple’s inner core since
childhood. His father was one of the founders, and Yam was the first to
play the role of Civil Magistrate when the temple premiered its Military
Retainers troupe in the mid-1960s. The young men who manned that
troupe formed a tightly knit cohort, both in and out of costume. Their
intense possession displays became a fixture of the annual Lantern
Festival procession, while their aggressive responses to challenges and
insults helped establish the temple’s reputation as a locus of somewhat
sinister numinous power (and a hangout for gangsters). The members
of that first temple cohort later moved into a variety of jobs and profes-
sions, but they all remained closely associated with the temple. Some
became ritual specialists of one sort or another; and most were part of
or socially associated with the jianghu. By the early 1990s, Yam was a
core member of the temple’s managing committee, a popular figure
in Taidong’s temple circles. He was also making a name for himself in
local underground business circles.
Yam’s arrival was almost always greeted with nods, smiles, and a
palpable rise in the collective mood. He projected qualities that typed
him as a man of the jianghu: he was unpretentious, jovial, and gener-
ous with his friends; he had an intuitive grasp of honor, finding ways
to give others face rather than scheming for his own advantage. He
was a hard drinker and inveterate betel chewer, yet at the same time
he was deeply devoted to his wife and four sons, giving him an aura
of dependability that complemented rather than belied his party boy
habits.
Yam bore one difficult burden, however: his second son, Jiafu,
had suffered acute kidney failure at the age of five, an event that had
framed Yam’s life ever since:
A few weeks after Jiafu got sick, Fifth Lord Xing [Xingfu Wuqian
sui, Loyal Harmony Temple’s patron deity] descended [during a
séance] and revealed that he was being punished for an unpaid
filial debt owed his parents from a previous life. Wang Ye [that
is, Lord Xing] knew Jiafu’s fate, but he had intervened and bar-
gained to extend his life. We didn’t know any of this. But just
before Jiafu fell ill, Lord Xing had been ordered on a mission
by the Jade Emperor and was away from his post when Jiafu’s
case came up again. Fortunately he returned before the sentence
could be fully carried out, and he was able to save Jiafu’s life.
(Conversation with A-luo, Yam’s wife, August 3, 1999)
Lord Xing’s ability to intercede in the matter of Jiafu’s fate came

not from his bureaucratic function as a plague god but through the
fear he instills, even in other gods. The god’s numinous power entails
a terrifying force: in life Lord Xing was a military man who died a
premature, violent death together with his sworn brothers.32 He is
characterized in image and narrative as an austere and strict master
who rewards loyalty but metes out harsh punishments to violators.
He is quick to anger and can be impetuous to the point of defying
Heaven; temple members point out that the god was once severely
reprimanded (some say exiled from the temple for twelve years) by
the Jade Emperor for redirecting a typhoon that had threatened an
important temple ceremony, resulting in dozens of deaths in Hualian,
to the north (see below).
It is Lord Xing’s command over a stable of demonic underlings
that makes him particularly fearsome. These ghostly servants—the
Military Retainers, Generals Chen (the “crocodile” general) and Mu
(wood), Ghost Eater, and others—are demons that the god has sub-
dued and redeemed. As my informants repeated in many discussions
on the subject, the god’s greatest power rests not in the direct control
of his minions but rather in the very tenuous nature of that control.
Once unleashed by his order in the name of Heaven, they tend to
“return to their original nature” and can wreak violent havoc on their
hapless victims.
Like most devotees at Loyal Harmony Temple, Yam and A-luo
zealously proclaimed their temple’s adherence to orthodoxy and the
austere majesty of their patron. Yet in narratives detailing the unseen
realm, they often described the gods as if they were figures from the
jianghu. (Indeed, in such narratives Lord Xing operates much like a
gang boss, rationally managing and directing violent forces through

the power of his own charisma.) In this realm the greatest power is in
the hands of beings who mediate order and chaos, yang and yin. Their
power represents not the corporate order of bureaucracy but rather
the fact of that order’s vulnerability. In this view, the cosmic order
has been entrusted to morally ambiguous beings like Lord Xing (who
“enforce the Way in the name of Heaven”—ti tian xing dao). Their
success depends on the vigilance of the demonic underlings he has
deputized to suppress disorder and check disincorporation.
Lord Xing directed Yam to sponsor a ritual to repay the spectral
parents and thus relieve Jiafu of his karmic debt.33 The god announced
that he had done his part to manipulate Jiafu’s fate as a favor to a
devoted disciple, but it was now up to Yam to pay for both the expiat-
ing ritual and the required (and expensive) dialysis treatments. Yam
promised to increase his devotion and work harder at his business;
Lord Xing then divined that Yam’s business fortunes would improve,
but he would encounter unnamed obstacles along the way.
While some of my informants peppered their daily conversation
with testimonials, Yam rarely talked about his patron god’s miraculous
power. When he did broach the subject, he spoke with passion and
conviction; he had no doubt that only Lord Xing’s intervention could
have saved his son’s life. He pointed to the trajectory of his business
fortunes as a case in point: before his son’s illness, he was dissolute,
drinking, gambling, and carousing nearly every night. At that time
he ran a “wild chicken” taxi service, ferrying carloads of passengers
between Taidong and Gaoxiong. While he rarely had trouble finding
customers, the money was never very good. But after his son’s crisis
and his renewed oath of loyalty to Lord Xing, he had a sudden run
of success. His taxi business mysteriously picked up, and his gambling
fortunes changed as well. He was able to pay off his debts within a year
and made enough to start up his own numbers operation. His drink-
ing prowess now became an asset rather than a weakness as he culti-
vated business connections. To Yam these were all clear signs of Lord
Xing’s miraculous intervention on his behalf.
Optimistic by nature, Yam harnessed his interpretation of events
to transform anxiety and despair into confidence and entrepreneurial
energy. In fact Yam worked extraordinarily hard to develop his new
business, although perhaps the most arduous aspect of the process was
the almost nightly series of social obligations (yingchou) , eating, drink-
ing, and carousing with his clients, his protectors on both sides of the
law, and current and potential business partners.
Yam’s Story
Yam paid his taxes, but most of his income came through his backroom
enterprises. His main business was gambling, but as with many such
operations, he extended credit to clients and provided certain finan-
cial services to other shadow economy ventures. Many of his associates
were brothers of the path, divided more or less evenly (in my observa-
tion, at least) among those who, like Yam, had been deeply involved
since childhood in the world of popular religion; others who simply
happened to be interested in the prospect of access to supernatural
power (but who had heard of Yam through jianghu connections); and
those who were purely business associates.
Most of Yam’s working evenings started out with a quick meal at
home. Around half past seven, the whole family would ride over to the
temple on two or three scooters. Yam would greet whoever was hang-
ing out under the awning, drinking tea or playing cards, and then go
inside to greet his “boss,” the patron deity Xingfu Wang Ye, lighting
a bundle of incense sticks and making the rounds. As was the cus-
tom at Loyal Harmony, he would finish up at the spirit soldiers’ sen-
try box in the outer courtyard. He would then go back inside, unlock
the “incense desk,” and check the day’s accounts. As fellow devotees
drifted in, he would hold court in the temple office, discussing temple
business or just shooting the breeze with fellow committee officers,
elders, and whoever happened to stop by.
On a late spring night during this period of Yam’s ascendancy, I
happened to arrive at Loyal Harmony just as Yam and a small crowd of
men and women were exiting the temple. Yam saw me and shouted,
“Hey, we’re going to see Tudi Gong!” 34 He invited me to go along,
and I squeezed into the back seat of A-beng’s white Chrysler. Another
group, most of whose members I had never seen before, followed in a
black BMW. We stopped at a roadside stall in Beinan to buy some fried
snacks and then headed up the valley toward Guanshan.
We drove west and north for about twenty minutes. Just before
Chulu, we turned east off the highway onto a narrow concrete road
that wound around the back of an abandoned restaurant, across a few
acres of cane field, and then began to ascend the low, unstable hills
that overlie the Rift Valley fault. The pavement ended and the grade
got noticeably steeper after two or three traverses. We leveled out in a

hanging valley, then started to ascend again, surrounded by betel palm
groves. After one more hairpin turn, A-beng stopped and announced
that we had arrived. We got out and trudged up a steep, narrow dirt
path toward what looked like a farmhouse nestled in the brush above
the road.
When we arrived, what we found instead was a small, spare, one-
room earth god shrine lit by a single kerosene lamp. A-gi, Baldy Chen,
and a few of the other Loyal Harmony regulars stood outside, each hold-
ing a bunch of burning incense sticks. Inside the bare concrete room a
stout, youngish medium stood shirtless at the altar, leaning on a peach-
wood pestle. A-gi told me they had already been there for almost an
hour and had gone through three bunches of incense trying to induce
the deity to descend. A-beng went in and lit another large bunch of
incense and handed out three sticks each to the new arrivals. We stood,
expectantly, for another half hour; finally, the medium began to twitch
and grunt; everyone moved forward. False alarm—he went still again.
Then, a few minutes later, it started again, the low groans growing more
intense, face contorting into a painful grimace. Suddenly, he stepped
back from the altar, balanced on his right leg, left foot raised slightly off
the ground; he brandished the pestle like a club in his right hand, his
left arm, elbow out in front of his chest, hand palm up, fingers curled
in. He began to speak, throat constricted, voice breaking, in pure vowel
syllables that Baldy Chen (the most accomplished fashi, “ritual master,”
among those gathered) attempted to translate. The medium’s expres-
sion became severe, his body shook, his voice rose—Baldy Chen’s inter-
pretation, it seemed, was missing the mark. He slammed the pestle
down onto the altar table and held the position, not moving at all for at
least a minute. Then he began to move the pestle in circles and zigzags
across the table surface, raising up and striking again at the end of each
tracing. Baldy Chen and the others crowded in to watch; several in the
group began to shout out words, numbers, waiting for confirmation
from the god. Instead, the medium repeated the same trace five times,
shouted, and fell straight back, caught just in time by A-beng and Yam.
Everyone bowed in the direction of the earth god’s image, planted his
incense in the burner, and withdrew as Yam and Baldy Chen attended
to the now detranced medium.
The group milled around for a bit, some of us wandering back into
the shrine to see if the tracings had been incised into the table or not
and watching the incense sticks as they turned to ash, some holding
their shape briefly before disintegrating into the censer. The shapes of
the ashes as they burned down were a visual code by which the spirits
consuming the incense might communicate with the devotees—but
were they Arabic or Chinese numerals? or characters? or Roman let-
ters? Apparently, no one got it right. None of the numbers came up on
the next night’s Hong Kong lottery drawing, and no one at Loyal Har-
mony Temple ever mentioned the shrine or its medium to me again.
As we returned to the cars, Yam signaled to me to ride in the BMW
instead of A-beng’s Chrysler. As we headed back to town, I learned that
we had come up to give Yam’s guest, Black Snake, a taste of wild Tai
dong. Snake, the owner of the BMW, was one of the very few hard-core
gangsters with whom I had extended contact during my fieldwork. He
was the hall master of a small branch of a well-known Taipei-based
gang, and he was thinking of investing in Yam’s numbers operation.
Snake had come down from Taipei for a three-day “vacation” to
evaluate Yam’s outfit firsthand. Somewhat disappointed by the scale
of this small-town operation, Snake would commit to only a one-shot
deal (the arrangement involved sanitizing profits). But along the way,
he and Yam hit it off as friends, and he was enjoying his “vacation” so
much that once business had been taken care of, he decided to stick
around and “play” for a few more days. The midnight quest for a ming-
pai—a winning lottery number—was part of the entertainment.35
On the way back to town, Yam started comparing the younger gen-
eration of mediums to the “old generals.” He told the story of A-lim’s
father, known as a very powerful sorcerer and efficacious medium. But
he loved to gamble, especially at dice. During the war, he was drafted
into the Japanese Army and sent to fight in Southeast Asia. Once, in
the middle of a battle, a local spirit appeared and told him the pla-
toon must immediately retreat at least one hundred yards. He told
his commanding officer about this. The Japanese officer was skeptical
but decided to withdraw anyway. Just as the troops finished their pull-
back, a five-hundred-pound bomb dropped and exploded right on the
spot where they had been just moments before, blasting an enormous
crater; all the trees within fifty yards were riddled with shrapnel. After
this, the officer always consulted him before issuing orders to advance
or retreat in battle.
After the war, A-lim’s father returned to Taidong. He was poor,
scratching out a living selling glutinous rice dumplings from the back
of a bicycle. Whenever he had any cash, he would gamble it all away.
At that time, he became friends with another medium, O-thau’s father,
who worshipped a plague deity, Xingfu Qiansui. A-lim’s father made

a vow that he would serve this god if his gambling fortunes improved.
The next day, he started a winning streak and was able to pay off all
his debts. A series of miracles ensued—a woman given up for dead was
cured, trucks suddenly started losing tire treads just down the road
from A-hian’s (another disciple) repair shop—and people started
showing up at a makeshift shrine to the god to burn incense and leave
donations. Within a few months, O-thau’s father and A-lim’s father
had enough to buy land and start building a real temple—Loyal
Harmony.36
Unfortunately, on the day the new temple was to be dedicated,
a typhoon threatened to strike Taidong. Just before making landfall,
the typhoon changed course and turned north. Hundreds of people
in the Hualian area were drowned or buried in landslides. Not long
after, the deity’s image disappeared. Many people believed that A-lim’s
father, the gambling medium, had sold the image to a fellow gambler,
an itinerant who had left town around the same time.
Yam pointed out that local sorcerers visited the valley we had just
left— called “Monkey Throat Gulch”—to recruit wandering ghosts,
especially as spirit assistants. A-lim’s father had many rivals among the
local ritual masters, one of whom was notorious for his collection of
Monkey Throat Gulch ghosts, whom he deployed for his own selfish,
destructive purposes. One day after the image’s disappearance, A-lim’s
father was walking past this sorcerer’s home when he was attacked by
a cloud of black qi. In the past, the god had protected his loyal dis-
ciple, catching the black qi before it could harm him, but this time, his
patron gone, the sorcerer finally succeeded in killing him.37
A-lim swore that he would clear his father’s name. He learned,
from another god, that the disappearance of the image was punish-
ment for the god’s having changed the typhoon’s course. By exercis-
ing his magical power for selfish purposes and causing the deaths of
innocent people, the deity had broken heavenly laws. Moreover, dur-
ing this time he had accepted spirit money that had been intended for
another god. For these infractions, the Jade Emperor had exiled him
for twelve years.
Exactly twelve years after the disappearance of the image, A-lim
was working in the shipbuilding yards at Dong’gang as a journeyman
carpenter. One afternoon, he got on his Yelang 125 motorcycle and
started to ride, in a semi-trance, toward Tainan. Traveling local roads,
without a map or any conscious sense of where he was headed, he
somehow found the most direct route to Xinying, in Tainan County,

arriving before sunset. Awakening from his trance, he heard a high-
pitched child’s voice summoning him, crying out, “We’re all here cel-
ebrating Wuqiansui’s birthday!” He followed the voice (which was that
of the child god Taizi Ye, later to become A-lim’s “master”) to a nearby
farmhouse. Inside the empty farmhouse, he found the long-lost image
of his father’s patron god. He picked it up, strapped it to the rack of
his motorbike, and sped away.
I had heard this story before from A-gi and others—it was a well-
known episode in Loyal Harmony Temple’s mythohistory, though
rarely a story told to outsiders. Snake seemed interested enough, but
by the time we got back to town, his thoughts were elsewhere. For
the moment, he’d had enough of malevolent ghosts and mysterious
mountain wilds. “Let’s go sing a few songs,” he suggested to Yam, and
we headed straight for the Golden Dragon KT V.
Mainland Rambles
The evidence presented so far should leave little doubt that, empiri-
cally speaking, the martial ritual performance traditions of Chinese
popular religion are tightly linked with that storied corner of Chinese
society (and imagination) known as the jianghu. There is sufficient (if
uneven) documentary support for the claim that this connection has
a deep-rooted and enduring history. The ethnographic cases I have
presented from Taiwan document the ways contemporary actors find
meaning in, yet at the same time continuously reinterpret and refash-
ion, the handed-down tradition. For them, as for most working-class
men and women in Taiwan, the association between martial ritual and
jianghu identity, between the demonic nature of martial deities and
the men living precarious lives who portray them, is evident, simple,
and logical.
On the mainland, however, that association is far from evident,
simple, or logical. Life in mainland China has changed in fundamen-
tal and dramatic ways since 1949.38 During the Mao years—sporadically
until about 1957 and relentlessly thereafter—popular religion and
associated practices were labeled “feudal superstition” and targeted
for suppression.39 During those decades, many of the basic cultural
assumptions and socioeconomic mechanisms that had sustained Han
Chinese popular religion for centuries were broadly, fundamentally,
and stridently challenged. Even today, religion (especially local, grass-
roots religion) remains a touchy subject in the PRC, despite the advent
of a new market economy, the rejection of Mao-era radicalism, and a

broad relaxation of controls in the social and cultural spheres.
Yet since the 1980s, as we have seen, many popular ritual traditions
have been effectively revived, at least at the local level.40 Religion is
thriving in the PRC and has again become an important, if still con-
troversial and highly regulated, feature of Chinese culture.41 The geog-
raphy of this emerging process, however, displays a clear asymmetry:
while the resurgence of religious devotion and practice includes most
of the country, there is an easily observable disparity in the level of
public ritual activity between urban and rural areas.42 Festival proces-
sions are effectively prohibited in the cities. In the PRC, they are now
exclusively limited to small towns and rural villages.43
Villages on both sides of Erhai Lake in Dali hold annual festivals to
honor their local deities. These festivals typically feature processions
that assemble at and eventually return to the local temple square,
marking the village boundaries and highlighting points of spiritual
power along the way (sacred trees, shrines to the Mountain God, and
sites where the vestiges of death and tragedy must be excised). The
processions are held in conjunction with all-day activities in the temple
itself, including sutra chanting and collective and individual offering
of sacrifices, culminating in a communal feast. The processions gener-
ally include performance troupes of one sort or another, mostly young
and middle-aged female folk dancers, older male musicians, and sutra-
chanting grandmothers. But troupes of a martial character do appear
here and there: dragon and lion dancers and teams of “lictors” accom-
panying a god’s palanquin are common features of many village fes-
tival processions in Dali and elsewhere. On the surface at least, it is
not difficult to identify characteristics, even specific ritual artifacts and
gestures, shared with similar small-scale events in Taiwan and ethnic
Chinese communities outside the mainland.
In the processions I observed in the Dali area in the mid- and late
1990s, a small number of troupe actors and organizers were older men
who had participated in such local processions before the early 1960s,
when all popular religious activity in the area was banned. There were
also a few young men, motivated partly by the pressure of social obli-
gation, as we saw in the case of the Torch Festival. Many, however,
especially those subjected to wearing costumes and face paint, were
elementary-school-aged boys and girls.44 Aside from a symbolic allu-
sion or two, such as reports that the food packets placed in the fields
during the Torch Festival were once intended for itinerant strangers,
there was nothing substantial to link these rural festival procession

performances with jianghu society or mythology.
Then again, I have occasionally encountered martial arts practi-
tioners in the PRC who are knowledgeable, even passionate, about
the spiritual value and ritual purpose of their disciplines. I have met
lion dancers in Guangzhou and itinerant Buddhist monks sojourn-
ing, meditating, and training in remote mountain temples who invoke
both the supernatural and the mythology of the jianghu when talk-
ing to me about their lives and practice. These are, however, isolated
exceptions and do not represent a widespread cultural phenomenon.
In other words, if we search deliberately and carefully enough, it is
certainly possible to discover a connection between religion and the
marginal subcultures of the jianghu, with its ethos of wu masculinity.
The connection between the two, however, once (and in Taiwan still)
taken for granted is here weak and discontinuous and goes largely
unrecognized in its own social milieu.
In the PRC, I have sometimes heard informants describe super-
natural manifestations and testify to the power of deities, martial or
otherwise. But I cannot recall having ever sensed the kind of intimacy
and personal familiarity that often characterize Taiwanese devotees’
descriptions of their patron gods. In the following, concluding set of
“tales from the jianghu,” then, there are no procession troupes or spirit
mediums and only an occasional private description of miraculous
intervention.
A Band of Student Brothers

When I first met him, Er Ge (Second Brother) was a twenty-one-year-
old student at an engineering college in Kunming. I met him as one
of a group of friends, mostly classmates, introduced by Lao Fang, an
old classmate of mine in the United States. When Lao Fang heard I
was going to be in Kunming (his hometown), he suggested I look up
his brother, who was still living there and taking care of their elderly
parents.
While doing fieldwork in Dali, I occasionally traveled to Kunming
to visit the research unit with which I was nominally affiliated. Dur-
ing one visit, bracketed by two official meetings and not much else in
between, I decided to give Lao Fang’s brother a call. I invited Alex—he
asked me to address him using his English name—to meet me at a
local tea shop. He brought his fiancée, Xiao Peng, a student at the
above-mentioned engineering college. The three of us spent most of
the afternoon enjoying tea and rambling conversation. The next day,
Alex returned the favor, inviting me to join him and Xiao Peng for a
hot pot dinner at a restaurant near the city gymnasium.
Alex met me at the corner of a winding side street. The restaurant
was tucked in the elbow of yet another branching street, not an easy
place to find. We arrived after a ten-minute walk. Given the obscure
location, the restaurant was surprisingly fancy, new and clean, with
glass picture windows at the front, a large central dining room, and
several private rooms along the side. The group was already seated at
one of the large, round “eight immortals” tables in a large, well-lit side
room. Alex introduced his friends (who were, more properly, Xiao
Peng’s classmates). Not counting Alex and Xiao Peng, there were five
men and three women who, I learned, often socialized together. Alex,
the only one not studying at the college, joined them now and then for
special occasions.
I was seated next to Er Ge. Noticing a bottle of baijiu (distilled
grain liquor) at his elbow, I politely advised the group that I had some
important business the next day and apologized for not being able to
drink much. Er Ge turned to me and quietly advised me, in turn, that
I was among friends and shouldn’t feel the need to explain. Today,
however (he continued), was Qinghai’s birthday, and they were there
that evening to celebrate.45 Protocol thus required that I at least share
a few glasses to recognize the auspicious occasion properly. “Here,” he
said, gesturing around the table, “we all talk righteousness [ jiang yiqi].”
With the addition of that phrase, I inferred that Er Ge’s friendly admo-
nition contained something of a veiled warning: not drinking implied
a rejection of friendship offered, so it would be best—consequences
not spelled out—for me to join the festivities as a full participant.
As we waited for the hot pot to boil, the shot glasses were filled and
the toasting began. After three quick rounds, the hot pot was ready,
and we started to eat. In the end, I managed to keep up with Er Ge,
Qinghai, A-dong, and the other brothers, finishing off first one and
then a second bottle of baijiu. I left before the party ended, taking
advantage of Alex and Xiao Peng’s early departure. But it seemed I
had passed the test: before I left for Dali later in the week, Alex called
to invite me to meet with Er Ge one more time.
We met at a coffee shop near Yunnan University. This time, only
Er Ge, Alex, Xiao Peng, and A-dong attended. Er Ge addressed me
respectfully now as “Elder Brother,” but the atmosphere was noticeably
different from the previous meeting. Befitting the venue perhaps, we
chatted in the mode of educated sophisticates rather than bantering

as brothers of the rivers and lakes. Er Ge talked a bit about his own
background, growing up in Sichuan in a military family, being the first
in his family to attend college, with all the attendant hopes and pres-
sures that entailed. He mentioned that he had recently become wor-
ried and anxious about the future, not entirely sure what he was going
to do once he graduated, unsure that he had the ability to make it in
the real world as a planner or engineer. The confidence and arrogant
bravado he had displayed at the previous occasion—on his own turf,
as it were—seemed to have evaporated.
Toward the end of August, I made another short visit to Kunming.
In addition to my research and other social obligations, I made sure to
check in with Alex. Alex told me that Er Ge and the others were laying
low at the moment—there had been an “incident,” which he described
to me over the phone.46 Xiao Peng had signed up for a summer fine
arts class given at the auxiliary campus of the college. The instructor
was a well-known professor, hired for the summer from an elite univer-
sity in Shanghai. A few weeks into the course, the professor started “act-
ing indecorously” ( feili) toward some of the female students, including
Xiao Peng. At the end of one evening class, while evaluating her work,
he surreptitiously moved his hand onto her thigh. She sat there until
he moved on, then packed up her things and left the building. Back at
her dorm, deeply upset, she reported the incident to her roommates;
the next day, Er Ge got wind of the situation and decided something
had to be done.
Er Ge and his band of four immediately began planning punitive
action. They had Xiao Peng ask the professor to meet her privately in
the usual classroom to look over her drawings. Delighted and unsus-
pecting, he agreed. Meanwhile, the five men prepared kitchen cleav-
ers, wrapping the blades several times around with thin cotton cloth
and hiding them under their jackets. Xiao Peng herself stayed behind
as the five headed off to the auxiliary campus. As luck would have
it, the building watchman was nowhere to be seen, and the route to
the classroom seemed clear and unmonitored. A-dong was stationed
downstairs as a lookout, while the other four headed up to the third
floor.
The instructor was waiting in the classroom, as arranged. Er Ge
entered and asked his name; surprised and annoyed at the intrusion,
he suggested that Er Ge must be in the wrong place and should leave
immediately. Er Ge asked again—this was serious business, and there
was no room for mistaken identity. “Are you Professor Xu?” he asked.
The instructor looked startled; at that point, he noticed the other
three young men at the door. “ What is this? What the hell is going on?
I’m going to call security.” Er Ge approached him. “ You’re a famous
professor, but you come here and rudely abuse our female classmates.
You have disgraced yourself and insulted our honor.” With that, Er Ge
pulled out his cleaver, and the four of them piled on, beating him on
the legs, arms, and head with the wrapped blades.
As soon as the blows started to rain down, however, the victim
started to shout, “I have a heart condition, I’m going to die! Don’t
kill me!” He slumped against the wall, apparently slipping into uncon-
sciousness. Er Ge may have panicked at that point—they had come
to administer justice, but the plan called only for a bit of roughing
up and intimidation, not murder. He stuffed his cleaver back under
his jacket—the others did the same—and started to drag the sagging
professor out of the room. “Better get him out of here.” They hauled
him downstairs. Once outside the building, they signaled A-dong. “Get
a taxi! We need to send this bastard to the hospital.” A-dong went out
the nearby side gate and hailed a passing taxi as the group hurried to
sneak the “body” out into the street without being noticed (somehow
no one had yet noticed or reported the disturbance). They tossed their
softly groaning burden into the back seat of the taxi and slammed
the door shut. Er Ge threw a twenty-yuan note into the front seat and
instructed the driver to head for the nearest hospital. Without waiting
for a reply, they took off, first walking quickly but quietly to the other
side of the auxiliary campus, discarding their cleavers in a water-filled
ditch along the way. They exited by another side gate and took another
taxi back to their dorm.
In the end, Er Ge reported, the teacher never made it to the hospi-
tal. As soon as the avenging crew had left, he sat up and redirected the
driver to his lodging place. He left Kunming the next day, although Er
Ge didn’t know how he explained his sudden departure to the school.
The next summer I returned to Yunnan for another field visit.
Heading up to Kunming in mid-July, I decided to look up Er Ge and
his group again. I gave Alex and Xiao Peng a call. They contacted
A-dong, who was moonlighting for the summer at a provincial gov-
ernment planning office in Kunming. The next day, A-dong tracked
down Er Ge and Qinghai, and we arranged to meet for dinner the next
evening at a small restaurant in Alex’s neighborhood.
Er Ge was wearing his habitual black cloth jacket, collar turned
up; he seemed particularly morose that evening. After ordering a few

dishes, we started working through a case of beer. After we had fin-
ished off a half dozen bottles, Qinghai announced that the group had
recently been involved in another “incident.” This time, it turned out,
it was something of an old-fashioned vendetta that had escalated into
a full-scale street battle. Until that moment, I had not thought of Er
Ge and his band of engineering classmate brothers as a proper gang,
but I was now starting to reevaluate. A-dong, Er Ge, and Qinghai nar-
rated bits and pieces of the incident over the course of the evening.
The upshot of the story was that Er Ge, unemployed (and thus bored
and probably depressed) for the summer, had headed to a basement
arcade near the college late one night to kill time playing video games.
A couple of neighborhood teenagers dropped in a bit later. Impa-
tiently waiting for a free console, they started to harass him, trying to
get him off the machine so one of them could play. Words and glances
were exchanged, and the two youths left. Half an hour later, they reap-
peared, along with three older boys. They waited at the door as Er Ge
paid his bill and started for home. His dead fish stare and coolness
under threat kept the group from getting too close; they didn’t dare
attack him then and there, but they followed him for several blocks,
taunting and threatening him along the way.
Er Ge alerted the two members of the band who were in Kunming
for the summer, and again they prepared their weapons and a plan of
attack. Er Ge described the two offenders and sent A-dong to scout the
arcade. A-dong emerged, signaling that the two teenagers were inside.
It turned out that the whole gang of perhaps seven or eight neighbor-
hood kids was hanging out in the basement that night.
The narrative continued, but I was not taking notes, and at some
point that I can now only vaguely recall, we switched from beer to
baijiu. I did manage to catch the gist and eventual outcome of the
incident, however: the three of them ambushed one of the youths
when he emerged from the arcade to urinate. He shouted for help,
and the whole gang poured out into the street. The teenage gang had
no blades, but one wielded a sawed-off pool cue, and the others had
similar club-style weapons. Outmanned, the three fought along the
sidewalk, looking for some daylight to make their escape. Along the
way, they thought they probably broke a couple of bones, left a few
permanent scars, and dropped at least three of the gang before getting
away. This time, they headed for A-dong’s room to throw any pursu-
ers off the trail. Other than a few bruises, they all got off relatively
unscathed, and they celebrated in A-dong’s room the rest of the night,
finishing off a bottle of Wuliangye liquor for good measure.
What most surprised me about this gang of engineering students
was that it existed at all. While I am confident that the PRC is not
teeming with college-educated street avengers, I am aware that Er Ge’s
gang is not unique. At the same time, it is not irrelevant that all the
brotherhoods of which I have knowledge—although Er Ge’s band is
the only one I have documented directly—are composed of men with
rural or working-class backgrounds, often with a family connection to
the military. This is not a case of deviance, but I would still venture that
a sense of outsiderhood, of marginality, and perhaps inferiority rela-
tive to others in their cohort, plays a part in drawing these young men
together into brotherhoods in the first place.
Through my contact with Alex, I was able to follow Er Ge, Qing
hai, and A-dong’s path after they graduated the next year. All were
able to find employment related to their training. Based on this admit-
tedly tiny sample, I would propose that very few, if any, of these young
men choose a career that entails violence, sanctioned or not. It is not
violence per se, then, that defines the group or individual masculine
identity among this group. Their identification with jianghu values
and action is transient, focused on the moral, social, and performative
dimensions of the knight-errant myth. In terms of livelihood and social
mobility, even though the employment picture for college graduates
in the mainland has appeared increasingly bleak in recent years, they
still have choices that most who tread the dark path do not.
They are not in it for the money. Their adventures have no obvi-
ous practical utility. What counted for Er Ge and his group was the
mutual loyalty and obligatory generosity of brotherhood, the heroic
defending of female honor, the taking of righteous revenge, and the
camaraderie of drunken feasting. They did not collect debts, extort
protection money from vendors, run gambling dens or loan sharking
rackets, or take what wasn’t theirs. Invoking every cliché of the knight-
errant myth, they played a band of righteous hero-bandits, each an
avatar of heroic wu masculinity.
Hauling Cargo
A-zhuan was definitely in it for the money, though the money was not
always very good. Son of a Hui father and a Bai mother, A-zhuan drove
a heavy-duty ten-wheeled Dongfang truck, hauling logs from the high
mountains of eastern Tibet and northwestern Yunnan to Kunming
and carrying goods back to the county and local market towns along
the reverse route. In the mid-1990s, the mountain roads A-Zhuan
mostly traveled were narrow, winding, and steep. They mostly lacked
guardrails, signage, and markings, and, north from Lijiang, only the
stretches close to the major towns were paved.
I met A-zhuan through Lao Li, who was working at the time as a
driver for our research team in Dali. Like Lao Li, A-zhuan had learned
the trade of driving, maintaining, and fixing cars in Xiaguan in the
late 1980s. Before that, he had a somewhat checkered past, having
done time for a robbery in which the victim was killed. The killing was
apparently at the hands of A-zhuan’s accomplice; being minors at the
time of the crime, both received relatively light sentences but served
their time in an adult prison.
The competition among truckers is fierce in China, particularly
in remote, mountainous areas like Yunnan, where trucks are the main
and sometimes only available means of moving goods. A-zhuan was
thus on the road most of the time, stopping in Xiaguan now and then
if he needed some maintenance or perhaps a short rest. I had a chance
to talk with him just a few times in the summer of 1996, with a chance
meeting and quick follow-up in 2002. Fortunately, A-zhuan liked to
talk about his life, even the less savory elements of his past and present
occupations.
Despite, or maybe because of, his early turn to violence and crime
and the price he paid—doing time but also ending up with very few
options in life—A-zhuan did not seem a violent person at all. He was
particularly jovial and friendly, in fact, characteristics that served him
well in a profession where friends and allies were critical to business
success and, in the high passes and long, remote stretches on the pla-
teau, could mean the difference between life and death.
“It’s a dangerous business,” he told me the first time we met, “but I
like being my own boss. No one tells me what to do. I just have to stay
on top of the news, know where to pick up the good loads.” I had gone
with Lao Li to get a late-night snack, and we had stopped at a line of
food stalls on the outskirts of Xiaguan, where taxi and truck drivers
liked to congregate in the late hours. Lao Li noticed his old friend,
who just happened to pull up and get down from his truck a few feet
from where we were sitting. He greeted A-zhuan and called him over.
We sat together for half an hour or so, just enough to eat a few bites
and exchange a couple of stories.
About a month later, Lao Li invited a couple of us from the research
group to his house for dinner. It turned out that A-zhuan was in town
for a couple of days, waiting for a part—something about his brake
lines—so he was at the table when I arrived. Lao Li didn’t drink, but
he provided beer for his guests, and the rest of us were happy to take
advantage of the hospitality.
After dinner, we sat in Lao Li’s living room and listened to him
reminisce about how he and A-zhuan had gotten their start as driv-
ers; relaxed and in private surroundings, A-zhuan then picked up the
thread, talking about the before and after and his life on the road in
the new era of individual business (getihu) . “I got turned the wrong way
as a youth,” he said, “but I got straightened out. My mother, she’s a
devout Buddhist. That’s what saved me.”
The bodhisattva Guanyin watches over me now. I keep her image

on my dashboard; she’s saved my life at least a dozen times. One
time near Taizishan [Kawakarpo; at 22,110 feet, the highest peak
in Yunnan], I was coming around a bend when I saw a bright
light. I was almost blinded, so I had to stop. I got out to see what it
was, but when I got out of the truck, I couldn’t see a thing. Then I
heard a huge rumble and crash just ahead. The roadbed had just
given way less than fifty meters from where I had stopped, and the
two trucks in front of me had driven over the precipice; everyone
died in that wreck. I couldn’t figure out what the bright light was,
but now I know—Guanyin had warned me, kept me from driving
over the edge. I went to a nearby Tibetan temple the next morn-
ing and gave the lamas half my gas money, to appease the ghosts.
A-zhuan’s ritual devotion was not limited to Guanyin or Buddhism

in general. Like many citizens of Xiaguan, he made occasional visits to
the General’s Cave ( Jiangjun Dong), a large, popular shrine perched
below the last peak of the Cangshan range, guarding the narrow pass
that divides the Erhai Basin from the watershed of the Mekong (Lan-
cang) River to the west. The shrine serves as the unofficial city god
temple of Xiaguan. The main side chapel is devoted to the Black Tiger
General, Zhao Gongming, the Military God of Wealth. Most “incense
guests” who visit the temple will stop at this altar. A-zhuan was particu-
larly impressed with the power of this deity, as he told me the last time
I saw him, in 2002.
I had run into Lao Li by chance in Xiaguan. Lao Li had started
driving a metered taxi by then and was waiting for a fare at one of the
main downtown intersections. It had been a few years since we’d seen
each other, though, so he told me to hop in and we’d go get a bite to
eat, like old times. We headed out to his favorite line of stalls—the
same place we had met A-zhuan the first time six years earlier. By sheer
luck, as we pulled up, Lao Li noticed A-zhuan’s truck parked down the
road, loading up with water. We looked around, and sure enough, A-
zhuan was sitting at a table with a few of his colleagues, talking shop.
When he saw us, he called us over to join him. We sat down as A-zhuan
held court. He looked around the table and started:
Last year, let me tell you, I had terrible luck. My truck was always
breaking down; once I even took a wrong turn just outside of Kun-
ming—that had never happened before—and got stuck in a field
for four hours and missed my load. I was ready to give this whole
thing up. I borrowed 10,000 yuan [renminbi], and I was getting
behind every month. You know, A-bing, he started taking on loads
from Burma—you know what I mean [note: transporting heroin].
I started thinking maybe that’s the way to go. I was going to talk to
that guy, what’s his name, he offered me something before. Then
my truck broke down again, right down from the Number One
bridge, you know, the one next to the village below the General’s
Cave. Lao San was with me; I was fed up; I told him, stay here and
watch the truck. I don’t know why, I walked up to the temple; usu-
ally it takes an hour, [but] I got there in twenty minutes. I really
just wanted to think, you know? But I bought some spirit money
and incense, some candles, and bowed to the God of Wealth. I
made a vow: if I could pay back the loan and double it in a month,
I would come back every trip to Xiaguan and make an offering.
When I got back to the truck, it was running—Lao San just said it
started up by itself. The next day, I got a load, and then my luck
turned. I made 20,000 that month, and it kept going, maybe not
that great, but I was making a profit again.
He shook his head. “Mysterious.” The conversation turned to shop

talk. Lao Li was starting to get anxious about his idle taxi meter and
suggested that I had some business elsewhere and was waiting for a
ride back to town. A-zhuan ignored him and continued:
You know, we truckers help each other out. But sometimes there
are fights over loads or getting in line for fuel, that kind of thing.
Then you need your brothers, the ones who stick with you. If
people know you can fight, that you’re tough but honest, they’ll
stick with you. Without friends you won’t survive on the road.
The ones you’ve been through stuff with, not just the ones who
eat and drink with you when times are good and everyone’s mak-
ing money. You have to watch out for those; they’ll spike your
tires while you’re not watching, and you’ll roll over—in a ditch if
you’re lucky. Real friends are only the ones with whom you have
no profit-and-loss relationship [ lihai guanxi].
He explained to us that some backstabbing like this had just hap-

pened to someone he knew up north, and that’s what the group had
been talking about when we showed up. “Driving is a man’s profes-
sion,” he continued. “You’re always on the move; you have to have a
feel for the road and know the machine. Come on, let’s have a couple
of glasses. . . .”
The trucking business in Yunnan is largely a collection of small-
time entrepreneurs, a shadow world of sometimes fuzzy legality, occa-
sional violence, and cut-throat competition. In many ways, it is no
different from the rest of the universe of small entrepreneurs in the
PRC. But perhaps because, as A-zhuan points out, it is a world of move-
ment, most (though by no means all) long-haul truck drivers are men
who find an affirmation of their own masculinity in the very act of driv-
ing (though certainly the gain and loss of competition plays into that
equation as well). Though they are businessmen, many of the drivers
are marginal characters, some of them ex-cons like A-zhuan. They live
a dangerous, grim, and unsettled life, yet one that offers refuge and a
livelihood that suits (and perhaps nurtures) their rootless marginality.
The heroes and vagabonds of the literary jianghu hired a boat now and
then, but they mostly traveled the roads.
The Xiaguan Tiger

For much of the summer of 1996, I lived on the top floor of a six-story
apartment block on the grounds of a medical college in Xiaguan, a
small city in western Yunnan. Taking a break at home from fieldwork
one afternoon, I noticed a strange scraping and tapping at the front
door. Sitting on the sofa facing the door, I watched as a thin metal file
appeared, probing and sawing above the old-fashioned spring-bolt lock
that secured the entrance. Pondering for a moment, I realized that the
wisest course of action would be to interrupt the attempted break-in
before the intruder could breach the thin wooden door. I grabbed a
length of steel pipe that had been propped up in the kitchen, then
instinctively banged on the table and shouted: I thought it would be
best to send a warning, rather than confronting a cornered, desper-
ate, and probably armed (at least with a sharp metal file) person or
persons. The file disappeared, and I heard footsteps retreating down
the stairs. I opened the door, ready to strike if necessary, just in time
to see a man in green khaki pants bounding away from the landing.
Assured that he—actually they—were in retreat, I put the pipe aside
and quietly followed them down. I caught up with the two somewhat
bedraggled men, both in their late twenties or early thirties, just as
they were exiting the block. Now in an open public space, I challenged
them and asked what they had been doing in the building. The taller,
possibly older one, mumbled that they were going to look for a friend
in another unit. They were clearly eager to get away and started to
move toward the rear campus gate.
As they moved west, I ran east to the Public Security sentry post
at the main gate. I quickly explained to the uniformed guard on duty
what had happened. He ducked into the guardhouse, emerging a few
seconds later with two middle-aged officers in plain clothes. One was
stocky and broad-shouldered and sported a military-style crew cut; the
other was tall, thin, and well dressed. He wore aviator sunglasses and
an old-fashioned compact revolver holstered at his left hip.
I pointed out the two fugitives, who were just disappearing into
the warren of alleys that bordered the west wall of the campus. The
three of us headed off in pursuit, but by the time we reached the gate,
they had vanished. The tall, thin, older plainclothesman signaled the
stockier one—I heard him addressed as Lao Hu—and told me to stay
put as they moved off into the old neighborhood maze. I waited near
the gate; no more than ten minutes later, Lao Hu emerged from a side
street, half-dragging the man I had seen on the landing. “Come on,”
he said to me, hauling the stumbling would-be burglar. As we made
our way back to the guardhouse, Lao Hu stopped every twenty paces
or so to curse his prisoner, pulling the man’s head back by the hair and
twisting his arm to an impossible angle.
When we finally arrived back at the main gate, Lao Hu handcuffed
the suspect to the guardhouse fence, hands behind his back and in full
public view. He ordered one of the uniformed sentries to stand guard
and invited me to take a seat on the old sofa in the guardhouse office.
A few minutes later, the thin, older officer appeared with the second
prisoner, who was quickly hustled through the office and into a small
storeroom at the back. For ethical as well as aesthetic reasons, I will
not describe the methods used by Lao Hu, his senior colleague, and
several young, uniformed guards to “interrogate” the two suspects. It
is enough to note that this situation was especially sensitive and highly
charged and that the arresting officers were fuming. The intruders
had caused them to lose face, having somehow slipped onto the cam-
pus on their watch. To make matters worse, the victim was a foreigner.
Retribution for this insult, then, was both swift and remorseless.47
I contemplated for a moment what might have happened to me
had the burglars managed to break in while I was in the apartment.
Bound, helpless, and miserable, they were clearly no longer a threat
to me, but I could find no satisfaction in their suffering. So I thanked
the officers and left.
Two evenings later, as I passed the guardhouse, I noticed Lao Hu
and one of the younger guards walking their bicycles down the drive-
way, also on the way out. I greeted them and struck up a conversation. I
was hoping to learn more about the two perpetrators—who they were,
what had happened to them. The guards had just finished their shift,
so I invited them to join me for a snack at one of the roast lamb ven-
dors near the gate. Lao Hu seemed wary at first but finally agreed to
join me for “a glass or two.” He led the way to one of the stalls and
ordered up a dozen or so skewers of barbecued beef and mutton, a
plate of marinated bean curd, and six bottles of beer. The three of
us sat down on plastic stools around one of three folding card tables
arrayed on the sidewalk next to the vendor’s cart.
The beer came first, and we opened two bottles. I immediately
offered the first toast, thanking Lao Hu and his colleagues for their
trouble and praising their swift and efficient police work. Lao Hu
quickly responded with a return toast, welcoming me to China. I then
toasted his junior colleague—his family name was Chen—who also
returned the favor. Before we could refill the glasses for a fifth round,
the food arrived at the table.
Lao Hu dropped two skewers of meat on a plastic plate and pushed
it across to me. “Eat; this is the best barbecue on the block.” He called
the vendor, who was preparing fresh skewers in anticipation of the
coming night’s business, to join us for a glass. The vendor wiped his
hands and walked over. The junior officer poured him a glass, which
he immediately downed. “This foreigner speaks Chinese,” Lao Hu
informed him. “Ah, well, you know,” the owner said, still facing Lao Hu
but apparently addressing me, “this ‘tiger’ is a great and famous police
officer in Xiaguan!” At first I wasn’t sure if I had understood—to an
outsider, even one somewhat familiar with the local dialect, “Old Hu”
and “tiger” sound almost exactly the same.48 Officer Chen chimed in:
“My elder brother here, see, we call him ‘Old Hu,’ but he’s so fierce
everyone else around here assumes his name is ‘tiger.’ No one’s seen a
tiger in Xiaguan for thirty years; here’s the only one.” Lao Hu grunted;
I offered another toast. We finished off a plate of meat, and the beer
was already gone. Lao Hu seemed to be enjoying this end-of-the-day
moment of relaxation and was in no hurry to leave, so I ordered more
beer. I asked about the two thieves.
“Worthless vagrants; they’re both heroin addicts from Eryuan [one
of the poorer rural areas in the prefecture], trying to steal money for
drugs. They’ll probably be sent to a drying-out facility [ jiedusuo], but
they’ll be back on the street in a year or two. That kind, they can’t
change their ‘drug nature,’ but they’ve learned never to show up
around here again, you can rest assured.”
“When they hear my elder brother ‘tiger’ is around, these guys
won’t dare to come around. He’s good to us,” Officer Chen pointed
out, toasting his superior again, “but merciless when there is work to be
done.” Lao Hu frowned and grunted again, then tossed back another
glass. “You collared that guy so fast the other day,” I said to him. “And
watching you deal with the suspects, I’d say you are probably a mar-
tial arts master.” This comment seemed to grab Lao Hu’s interest; his
rather dour expression had hardly changed up to this point, but sud-
denly he became animated. “Fighting for me is as natural as holding
a rice bowl and chopsticks. I got my formal training, though, when I
was in the infantry,” he said. “I was a boxer in the army. But it doesn’t
matter if you’re in the ring or on the job; it’s always about being quick
and not worrying about getting hit. Hold nothing back.” He turned
around toward the street, reached out, and lifted up his old-fashioned
bicycle in his left hand. “How many forty-year-olds can do that? Keep
sharp and stay tough. This is our responsibility because we are here
to serve the people.” I was about to probe for more details about his
boxing experience and life in the army, but Officer Chen, the alcohol
starting to take effect, and perhaps led on by my question, cut in. “This
is a real hero, a real man [yingxiong haohan]. All of us younger brothers
try to learn from him,” he said, pointing to Lao Hu.
After a few seconds of awkward silence, I asked Lao Hu where he was
from. “I’m native-born Xiaguan,” he said, “but my father came down
here in the 1950s from Shandong.” Officer Chen chimed in again:

“So Elder Brother Tiger can drink us all under the table. These Shan-
dongese are really tough; they drink grain alcohol like boiled water.”
Grateful for the moment that we were drinking Dali Beer rather than
120 proof grain alcohol, I toasted Lao Hu one more time and started
to get up to pay the bill. Officer Chen jumped in front of me. “You’re
a foreign guest,” he said. “But I want to thank Elder Brother Tiger for
his help the other day,” I protested. “No big deal—just part of the daily
routine, nothing more.” He dropped a few small bills on top of the
vendor’s cart; the owner gathered them up without counting; he was
happy to have the business but knew Lao Hu’s reputation and seemed
relieved that we were moving on. “Be careful around here,” Lao Hu
grunted and gave me a hard stare as he mounted his bicycle. “Better
go home early.”
The obvious point of intersection between the narratives of mainland

and Taiwan informants is the social and psychological alienation expe-
rienced by marginal men in Chinese society, rooted in the frustration
of desire inherent in the patrilineal Chinese family but deeply exac-
erbated (more so in the mainland than in Taiwan or Hong Kong) by
exclusionary national educational, economic, and political structures
that create a vast underclass (or rather underclasses—rural, urban, and
rural-urban migrants). Such disempowerment motivated an earlier
generation to revolution, but no such scenario emerges in my infor-
mants’ narratives. Struggling at the lowest levels of the bureaucracy
(like the yamen runners of yore) or scratching a living in some corner
of China’s vast shadow economy, they adapt by taking refuge among
their peers, representing their manhood during rounds of feasting
and drinking. They perform a heroic masculinity through acts of gen-
erosity and loyalty and, if threatened, a demonstrated willingness to
sacrifice body and life. A small but conspicuous minority—here rep-
resented by Lao Hu—rely on aggression and physical domination to
define themselves as men and as the key to their livelihood as well.
The shadow economy in China, as in Taiwan, includes sectors that
unambiguously belong to the “dark path”: drug smugglers and human
traffickers are perhaps its most maleficent agents, corrupt officials
its most powerful and insidious. Here one also finds small, transient
groups that resemble Taiwan’s “venture squads,” joining up to steal
motorcycles for parts or distribute a haul of black market cigarettes.
There are high-powered bosses, often allied with the aforementioned
corrupt officials, running major rackets—protection, prostitution,

loan sharking, debt collecting, gambling, and so on. As in Taiwan,
however—even more so, as the stakes are higher and the risks greater
on the mainland—I made no effort to identify or interview anyone
involved in the illicit and frequently violent world of organized crime.
In Taiwan, my interest in military deities and personal participation in
martial rituals led to conversations, and sometimes long-term rapport,
with individuals that I would not otherwise have sought out and who
would otherwise not have welcomed my company. This did not and
probably could not have happened in the mainland.49
Based on the firsthand information I was able to obtain, which
is at least supported by anecdotal reports in the Chinese press, the
“footsoldiers” at the lowest rungs of the gang ladder are, not surpris-
ingly, mostly unemployed or underemployed laborers or moonlight-
ing freelancers. As in Taiwan and Hong Kong, many are men who
find themselves at a crossroads and have no intention of pursuing a
criminal career. They may have no intention of joining a gang but will
participate in a “venture” to avenge an insult to a friend or to their
peer group, or they may need quick cash and be willing to perform
some unpleasant task for a few yuan or even a pack of cigarettes. This
diffuse, often fractious, subculture of those whose livelihood depends
on the shadow economy includes violent, professional criminals but
also itinerant laborers, unlicensed vendors and taxi drivers, and other
“entrepreneurs,” many of them laid-off (xiagang) workers scratching
out a living on the social and economic margins. Participating in a ven-
ture or joining a gang promises little financial reward, but it provides
a stage for, and prompts a performance of wu-masculine selfhood that
would be immediately recognized by Taiwanese sorcerer troupe actors
or the young men who ride Handan Ye’s chair.
In Taiwan and in rural villages in the PRC, I was able to cultivate
sustained, trusting relationships with my informants. My accounts of
mainland urban lives are based on interactions of somewhat briefer
duration. What I report here, however, is at least an accurate record
of those interactions, many of which began and were sustained sim-
ply through sharing a meal, often an evening or late-night snack that
included repeated toasts with glasses of beer or locally produced,
distilled grain liquor. The close and durable interconnection among
drinking, camaraderie, and masculinity in Chinese culture is the sub-
ject of the next chapter.
CHAPTER 6
Wine, Women, and Song
The fragrance of willow wafts through the tavern.

A singsong girl of Suzhou presses a fresh brew and urges the
guests to drink.
My drinking buddies from Jinling have come to see me off ;
Hesitating to leave, we have another round.
Let me ask you: can the great river, flowing east, compare with
the undying affection between parting friends?
— Li, Bai, “Parting at a Jinling Tavern” 1
For most men in Chinese society, joining friends and associates for
a night on the town can be as much obligation as indulgence. Both
senses are captured in the two phrases most often used to describe such
socializing: yingchou, “reciprocal entertainment,” and he jiu, “drinking”
or “carousing.” Alcohol, in fact, is crucial to Chinese commensality; its
importance in both religious and social practice can hardly be over-
stated. From solemn official banquets to raucous village feasts, from
the cultured meetings of literati poets to the backstreet revels of gam-
blers and gangsters, the meanings and manners of alcohol consump-
tion have been among the most widespread, stable, and enduring fea-
tures of Chinese culture.
Yingchou is a term so plainly idiomatic that most Chinese refer-
ences simply gloss its common, familiar usage as “social intercourse
in general.” But the word implies something much more specific: it is
action predicated on response and reciprocity. Ying carries the sense of
“responsively/actively attend to.” Chou originally referred to the third
round of drinking during a formal gathering or feast. According to
the protocols of the feudal Western Zhou period (ca. 1050–771 BCE),
a host (in the superior position) poured the first round (xian) for his
guests, leaving his own cup unfilled until the guest(s) reciprocated
in the second round (cu) . For the third, or chou, round, the host first
filled and raised his own cup, then entreated his guests to drink freely
and without restraint.
The first round establishes the host’s role as master of the pro-
176
wine, women, and song 177
ceedings (and of the alcohol itself, which is his to offer); the second,
reciprocating round confirms the hierarchy and establishes the bond
between host and guest. By serving his guests and leaving his own cup
unfilled, the host leaves open the possibility that his generosity will
not be reciprocated. The implication is that his toast is a true act of
generosity, with no strings attached. The guests are fully aware of the
obligation to reciprocate, of course, but this is not mere protocol or
pretense. The host’s offering is a true gift in the sense that it only cre-
ates the conditions for the creation of bond but does not entail it. The
ritual is thus a social drama. The protocol limits the range of possible
outcomes and consequences but assures none of them.2
The third round marks the transition from formal ritual to infor-
mal sociality (although the second round “completes” the exchange,
it does not fully encompass the relationship thus established, and thus
a third, “exiting” or “returning” round is required). Anyone, host or
guest, could invite or entreat anyone else to drink from this point on.
Whether or not contemporary practice derives from the three-
thousand-year-old customs of feudal chiefs and warriors is not impor-
tant (at least in the present context). The value of comparison here lies
in being able to see the spare logic of the ancient sequence; as it turns
out, modern conventions strongly echo this same logic: the host first
urges his guests to drink (zhu quan ke) , guests then respectfully toast
their host (ke jing zhu) , and finally guests toast each other continuously
throughout the rest of the feast (ke xiang quan) (cf. Guo 1989:235).
Thus socializing (including public feasting and male carousing)
in Chinese society is comprehensible only if we can grasp the dynam-
ics of reciprocal obligation. Yingchou in contemporary usage refers
either to social interaction in the abstract or to the cycles of gifting
and feasting through which informal social networks are produced.
Such networks—those “based on personal, particularistic bonds and
taking the form of a network of variable dyadic relations” (DeGlop-
per 1995:38)—have been described most extensively in recent eth-
nographic literature on “guanxi networks” in the PRC (Huang Shu-
min 1998:143ff.; Kipnis 1997; Yan 1996; M. Yang 1994) and Taiwan
(DeGlopper 1995; Weller 1987).3
The sorts of reciprocity that characterize social interactions serve
to reproduce particular sorts of hierarchical relationships. In the cases
I cite here, the hierarchy is neither fixed nor explicitly denoted but
rather fluid and negotiable. Competition is implicit in these acts of
greeting, entertaining, contest, and struggle—whether formal, as in
the mutual greeting and testing of boundaries and skill by temple pro-
cession troupes, or mundane, as in negotiating or acknowledging rela-
tive status in competitive drinking.
This competitive ethos underlies the creative negotiation that takes
place in the midst of these rituals. As a simultaneously private and
public contest for position and “face” (mianzi) , the danger of conflict
both within and among groups simmers perpetually just below boiling
point. Alcohol can severely exacerbate the mood, especially when a
preexisting personal or political vendetta is added to the mix, and the
party sometimes ends violently, even tragically. Discord and violence,
therefore, are always possible, even if in practice they are avoided or
resolved more often than they are allowed to emerge and escalate.
Yingchou includes social activities of any scale; carousing, on the
other hand, is almost always done in small groups. Until at least the
early 1990s, working-class men in Taiwan often referred to carous-
ing as a special form of “recreating,” chhit-tho.4 I have identified seven
tangible ingredients essential to carousing: eating, drinking, sexual
play, singing, betel chewing, gambling, and smoking; all are constitu-
ent features of carousing in Taiwan (and, with the exception of betel
chewing, in China as well). While I touch upon all of these elements,
my focus here is on two of these activities—namely, drinking and sing-
ing. I choose to examine these elements in particular because they
afford two very different, though complementary, accounts of Chinese
masculinity.
Most working-class men in urban China and Taiwan socialize with
friends and comrades in the evenings, but everyday gatherings mostly
center around sharing tea, cigarettes, and conversation rather than
alcohol and female companionship. These get-togethers typically take
place in semi-public spaces: benches in a temple courtyard, plastic
armchairs set out on a sidewalk or alley, or folding chairs around a
card table in a friend’s living room. The early arrivals sit, while late-
comers, except high-status guests, may stand or squat. If the venue is
a home or business, the host usually provides the tea, while arriving
guests casually offer around cigarettes and, in Taiwan, betel as well.
In public spaces, the host-guest status is fluid, but new arrivals usu-
ally come armed with something to share around. Though women are
often present, they do not participate in the smoking, chewing, or ban-
ter; the relaxed postures, informal dress (shorts, T-shirts, and flip-flops
are not inappropriate), easy joking, and cursing wordplay express a
distinctly masculine sociality.5
For the small entrepreneurs, politicians, and “players” (chhit-tho-

lang) of the jianghu, however, such evening sessions are frequently just
a prelude to a more expensive mode of play where the level of con-
test is far more intense, the symbolic stakes far higher, and the risks
(including the potential for physical violence) far greater. The tests of
drinking prowess, the paid company of (at least implicitly available)
females, exhibitionistic singing, and, not least, the conspicuous cost of
the entertainment become highly charged representations of sexual as
well as social power. Every action accrues symbolic significance: a night
of carousing becomes a ritual drama.
Neon-framed nightclubs, karaoke bars, and spas are the “temples”
of East Asian urban night life. In an evening of carousing, typically
groups will change venues several times. This roving is characteristic of
both procession troupes and bands of comrades out on the town. The
events that take place at each venue are like the sequence of rituals
and possession trances at this temple or that shrine along the proces-
sion route. They can be meaningful performances, remembered in
the next day’s narration, becoming in some cases the collective mythic
lore of the group as a group. The street, in both cases, is a liminal
space to be transited, peacefully it is hoped; occasionally one encoun-
ters ghosts (or cops or a car wreck or a knife fight). Reaching the next
meaningful place, one crosses from the brightness of the doorway into
the well-lit lobby, then into a dark and mysterious interior filled with
images, sounds, and smells that (with apologies to Clifford Geertz)
“establish powerful, pervasive, and long-lasting moods and motiva-
tions in men” (Geertz 1973:90).
Carousing is usually a matter of public social play and accordingly
happens at a public venue.6 Evening entertainment is a ubiquitous and
highly visible business in Taiwan; it is a creative mainstay of popular
culture. The variety, scale, and creative design of entertainment estab-
lishments is dizzying. In reform-era China, the signs are less garish,
the architecture more subdued than in Taiwan, but going out—for
a shared meal or a night of all-out carousing—is just as important a
social and economic activity.7
Here I am concerned with a certain subset of these establish-
ments—namely, those whose main business is to satisfy the needs of
small groups of carousing men. Generally these establishments are
enclosed spaces like KT V clubs, piano bars, teahouses, and restaurants.
The public nature of the activity fosters a certain social permeability.
In chance encounters new bonds and connections can be spontane-
ously created or redefined, thus changing the composition, size, and

reach of the in-group over the course of even a single evening. The
public nature of carousing also offers the opportunity for conspicuous
performances—of social savvy, intelligence, loyalty, “righteousness,”
etc.—that serve to create or enhance a position within the group and
a reputation beyond it. Yet reciprocal entertaining and the carousing
that accompanies it is almost invariably a practice of small groups.
Carousing creates intimacy within the group and fosters a strong, if
temporary, resistance to the intrusions of uninvited others.
Karaoke as a Cultural System

Karaoke, particularly the “follow the bouncing ball” video machine that
has become the hallmark of modern karaoke technology, has unques-
tionably brought about important changes in carousing behavior
throughout East Asia. Karaoke quite literally automates the entertain-
ment, putting control of the performance into the hands of the audi-
ence (in the form of a remote control console) and turns the guests
into auto- (that is, self-) entertainers. Nevertheless, despite important
changes in aesthetics, there is a notable continuity in the “poetics” of
carousing from the pre-karaoke days to the present. Modern carousing
consists of modes of interaction and self-presentation that were very
much part of Qing-era practice and quite likely have been developing
since the medieval period in China. Musical performance and drink-
ing games (hua quan) were essential features of the literati carousing
tradition, for instance. More recently (and locally), customers who fre-
quented the nightclubs and brothels of Beitou (a hot springs resort
in the volcanic hills north of Taipei) in the 1940s and 1950s famously
sang along with the itinerant musicians who stopped in on their nightly
circuits.
Although the association of drinking, singing, and showing off
in front of friends may be close to a cultural universal, the karaoke
phenomenon has specific cultural and historical roots in the Japanese
bubble economy of the 1980s. The first karaoke bars to open in Taipei
followed the Japanese model—in fact, they were initially installed in
bars that catered to (male) Japanese sex tourists and businessmen.8
Once planted in Taiwanese cultural soil, however, Japanese karaoke
began to undergo a process of localization. By the mid-1980s karaoke
clubs featuring large rooms, dance floors, and modest light shows were
popping up along Linsen and Zhongshan roads, at the time Taipei’s
main entertainment district. By 1988, the sparse, simple, cozy hole-in-
the-wall clubs had mostly disappeared, replaced by lavishly decorated

nightclubs with gaudy neoclassical facades and modern, cosmopolitan
interiors. They had also spread well beyond the entertainment districts
of the urban centers and tourist resorts to smaller cities and market
towns throughout the island.
At their most innocent, karaoke clubs are spaces for conventional
expressions of comradeship and courtship. There are certainly many
“family-oriented” karaoke clubs, but until quite recently, these occu-
pied a small niche of the Taiwanese KTV business. At the other end
Taiwan’s karaoke craze began in the 1980s in the Japanese-style

members-only clubs and hostess bars of Taipei’s “Nine Alleys”
nightlife district.
of the spectrum, some backstreet KTV parlors are little more than
brothels that provide a place for customers to ritually drink and croon
before heading off with their hostess companions to the nearest hotel
or guest house.
Most karaoke clubs in Taiwan (and increasingly in the PRC as well)
fall somewhere in between G-rated family entertainment centers and
commercial sex establishments. Moreover, as primary sites of male-cen-
tered carousing, these clubs are not simply purveyors of sex as a directly
exchangeable commodity. Rather, karaoke clubs and similar entertain-
ment establishments in Taiwan function as part of a more complex cul-
tural process in which sex (most often as the anticipation and/or aura
of sex) is one signifier among many comprising the symbolic matrix
through which status, honor, and manhood are presented and con-
tested. Each of the elements of successful entertainment, alone and in
combination, provide a public, actional context for the expression and
production of male subjectivity. In turn, these actional contexts are
predicated on a particular ethos of masculine identity and behavior.
Entertainment in rural and working-class communities tends to be
less formal than official affairs (though there are certain, inviolable
protocols that must be followed in both). This is true both at occa-
sional social events like temple festivals and wedding feasts and at fixed
venues like nightclubs and KT V parlors. Ironically, in the fluid jianghu
subculture, it is the combination of formal and informal—a mastery
of formal etiquette with a manly display of appetite for drink—that
is most highly esteemed. For many on the “rivers and lakes,” indul-
gence and obligation are nearly indistinguishable. Both are intrinsic
elements of everyday life.
Both of these roles— of self-mastery and dominance and of self-
indulgence and rebelliousness— can be performed by one and the
same social actor during a single carousing event. For example, A-lang,
an office worker in Taizhong and part of a group of high school class-
mates who had become sworn brothers, seemed to particularly enjoy
going beyond the limits of propriety in his interactions with hostesses.
On several occasions I saw him almost get bounced from the premises
of four or five different KT V parlors when, either really drunk or feign-
ing sloppy intoxication, he would very inappropriately and repeatedly
grope a hostess. Yet even in this state of infantile indulgence, he was
also the first to notice when others were losing control. He would
intervene if things got too dangerous. If an exchange of words with
another group began to escalate, for instance, A-lang, playing the part
of eldest, most worldly, and most responsible “sibling,” would step in

to calm the storm. If one of his friends was too drunk to drive, he
would hide that friend’s car or scooter keys and offer him a ride home.
Far from objecting to such protective intervention as unmanly—what
might be derisively labeled ke-pho (Holo) or popomamade (Mandarin),
literally “acting like an old housewife”— the group expected A-lang, as
the eldest among sworn brothers, to take control in these situations
and protect his juniors. His friends did indeed respect and respond
to his authority as their senior, even as they tolerated his occasional
infantile regressions.
This changing of roles is perfectly understandable and perceived
as normal at the level of practice, even if the roles or qualities per-
formed seem logically or structurally incompatible. Thus when drunk
and in the company of friends who were in control of the situation,
A-lang could act playfully aggressive toward females. A-lang played out
his own sexual insecurities through the combative sexual symmetry
between men and women, but he did so safely within the idiom of
what is understood by all to be a complementary, interdependent rela-
tionship between male customers and female hostesses. Later on, if
and when his “younger brothers” lost control, the protective isolation
from public (including family) consequences disappeared, and A-lang
could move quickly into the role (again, complementary relative to
his brothers) of protector and authority figure. A-lang’s behaviors,
appearing in sequence and under particular circumstances, did not
entail any sort of contradiction and did not fix a permanent identity in
their minds, nor did one negate or delegitimize the other.
On the Town
An evening of carousing typically involves visits to several venues, with a
fairly predictable, though generally unplanned, pattern of movement
among the various types of establishments. For groups of friends and
associates who meet regularly, there is often a default meeting loca-
tion, often a restaurant. Many evenings of carousing begin unplanned
at the default location; here, the cast assembles, and the evening’s
drama takes shape. But at first, the group is fluid and the roles of host
and guests remain flexible.
After a few rounds of drinks and food— during which even casual
drop-ins from other tables are greeted with a full glass and a pair of
chopsticks—the party begins to take shape. The decision to move on
depends, in part, on the hour and the group’s stage of intoxication. If
the conditions are right—that is, if there is a critical mass with the will,
energy, time, and cash to continue—the conversation turns to choos-
ing the next location. Whether it turns out to be a nightclub, piano
bar, or someplace else, it is nearly certain that there will be some form
of public singing, probably in the form of karaoke. This form of enter-
tainment, in which the partiers perform for each other rather than
being serenaded by professional musicians, sets modern carousing
apart from past practice. Yet the long tradition of courtesan perform-
ers continues in the form of senior hostesses, who are expected to pos-
sess both superior looks and skills and to perform a karaoke song or
two for wealthy (and favored) customers and important guests.
The following accounts are based on my own recollections of
events, corroborated through additional observations and conversa-
tions with the participants. The actors here, even those in their early
twenties, were all fully conversant in the rituals and protocols of social
drinking and carousing. While some of the carousing events I describe
were little more than recreation and social play, every one of the par-
ticipants considered it to be an indispensable part of “performing per-
sonhood” (zuoren) . At other places and times, the expertise gained in
such “practice sessions” would pay off in more practical ways. Carous-
ing, especially for men in the jianghu, continues to be an important
part of their livelihood and entails significant investments of time,
money, and energy.
The first thing that struck me when I was first invited to participate
in nights of carousing in Taiwan was the absolute nature of gender
and age divisions. A carousing group is almost always exclusively made
up of men who are, more often than not, close in age, social posi-
tion, and economic status. Groups are fluid: participants filter in and
out of them both on a given evening and from occasion to occasion.
Once constituted on a given evening, a group may move from place to
place. Most party locales provide the company of female hostesses as
part of their service, yet the women generally do not become part of
the carousing party. If the group moves on as a group, it is only (and
usually all) the men that move. Hostesses are necessary participants in
carousing, of course, but their participation is limited by their desig-
nated role, which is to create an atmosphere of sexual possibility.
I spent the fall of 1989 in Taizhong (Taiwan’s third-largest city),
a guest of the Lim household, at the time a stem family of four gen-
erations. Grandpa Lim was a rags-to-riches entrepreneur, then in his
eighties and fully retired, who had made his fortune in the shoe export
business; he had broad connections in both “white” and “black” paths,

and his excellent connections were a cornerstone of the family’s ongo-
ing business activities. His eldest son, Lim Kian-ti, was the practical
head of the household and in business perhaps even more astute and
hard-nosed than his father. Since taking over, he had diversified into
real estate, banking, and the restaurant business and had done quite
well at all three. Lim Kian-ti himself had three sons: the eldest, who was
married and in turn the father of two young sons, lived at home and
managed the restaurant and entertainment side of things; the young-
est son, A-kun, worked as a “field officer” for another of the family’s
ventures, a cooperative lending society. The middle son, whose room
I occupied during my stay in their home, was studying in the United
States, working toward an MBA at a well-known Texas university.
Taizhong has a reputation as a party town, and my time with the
Lim family became, in part, a crash course in Taiwanese-style carous-
ing, drinking, and karaoke singing. A-kun and his drinking buddies, a
cohort of twenty-something, middle- to upper-income entrepreneurs,
went out carousing at least three times a week and during this period
generously included me in their social circle.
A-kun’s grandfather and father had close connections in the local
underworld, but the family’s investments and business operations were
now all of the legal, tax-paying variety. A-kun himself got no closer to
the “dark path” than his occasional work during election campaigns
(vote buying was an unfortunate fact of political life in Taiwan at the
time). Among his friends and colleagues, including those who often
joined us for nights on the town, a few had family or business con-
nections with local jianghu circles, but none were, properly speaking,
gangsters or brothers of the rivers and lakes. Nevertheless, they were
dedicated and serious carousers and, as a group, had enough com-
bined status and “face” to feel secure as they negotiated Taizhong’s
underworld-dominated nightlife.
When I left Taizhong a few months later, I came away with a basic
knowledge of Taiwanese drinking practices and protocols, a useful
carousing vocabulary, a performing repertoire of Mandarin and Tai-
wanese pop songs, and a much-improved capacity for alcohol. In the
course of my fieldwork in Taiwan and China, this combination turned
out to be an invaluable tool in developing rapport with informants.
Many of the brothers of the jianghu otherwise had little reason or incli-
nation to talk with an inquiring foreign anthropologist.
Another reason I choose to begin with a description of carousing
in Taizhong is that the city has a popular reputation as the birthplace

of many of Taiwan’s best-known (and most notorious) fads. The ubiq-
uitous barber shop brothel, “bubble tea” stands, and wild-looking beer
pubs (pijiu wu) that would not be out of place on the Las Vegas Strip
are all considered to have originated there.9 Over the last two decades,
Taizhong has grown at a faster pace than any other major city in Tai-
wan, due (as the popular wisdom declares and city boosters proclaim)
to the greater availability of space, the climate, the centralized loca-
tion, and excellent transportation infrastructure. Service businesses
have thrived, and following the city’s reputation for creativity, many
new forms and variations have appeared, particularly in the area of the
“special professions” (tezhong hangye) —that is, the sex trade. Indeed, in
the summer of 2000 the manager of a large KT V in Taizhong’s new-
est entertainment district (in the southwestern quadrant of the city)
proudly claimed that “Taizhong boasts more hostess clubs [ jiudian]
than any other city in Southeast Asia.”
Wine
A typical carousing episode started at about seven thirty at the Pirate
Ship (Haidao Chuan), a beer house built in the shape of an eigh-
teenth-century European sailing vessel. The owner’s son belonged
to this group, so the restaurant was a frequent starting place for an
evening of carousing. By nine, the table was crowded with plates of
seafood and beer mugs in various states of emptiness. Cigarettes and
betel, offered around in turn by each individual, punctuated the
rounds of eating and drinking. There was no formal ritual of host fill-
ing the guest’s cup; rather, waiters brought full mugs of beer to the
table. But the ritual principles of reciprocity and sociality were care-
fully observed as every gulp of beer, preferably at least a half mug at a
time, came in response to the urging of one or another of the group.
Sometimes someone toasted a particular friend; sometimes another
exhorted the group to drink as one. As drinking came to predomi-
nate over eating, a few rounds of drinking games began.10 This sort of
drinking in a spirit of friendly competition among comrades or busi-
ness associates can be seen at any restaurant table just about anywhere
at any time of day throughout China and Taiwan. When drinking with
friends, one affirms the mutual bond by drinking with abandon. To do
otherwise would be to imply that the present company are not “true
friends.” 11 Drinking on cue, without hesitation, conveys a shared plea-
sure of intoxication. Moreover, continuing to acknowledge a host or
guest even when enduring the discomfort of too much alcohol affirms
the importance of the relationship. A refusal to drink is a rejection, a
slight. Among friends, and when women are present, such behavior is
often met with mock collective indignation, derision, and a “fine” of
“drinking penalty wine,” he fajiu. In a situation where the participants
are not all familiars or where existing tensions have been disinhibited
by intoxication, such slights, perceived or intended, can and often do
induce angry, even violent, responses.12
This volatility simmers constantly beneath the social surface in
Chinese working- class society; it is almost tangible in the disinhibited
intensity of jianghu carousing. Here “play” is the pretext for the deadly
serious business of negotiating identity and status. The currency of this
negotiation is “face.” Face is not purely symbolic capital but embodies
the total social value of an individual (cf. Kipnis 1995). In the height-
ened intensity of carousing play, stimulated and disinhibited by alco-
hol, face lost, or not given, can trigger sudden explosions of rage that
escalate into group violence of almost inexplicable ferocity. More than
once at the late-night seafood stalls near the old Taidong train station
I have been startled by the sound of shouts and crashing dishes and
turned to watch four or five men trying to restrain another, hell-bent
on tearing into everything and everyone he can get his hands on. If
he manages to momentarily fight free, punches are thrown and tables
overturned, and the party spills out onto the street. The tussle rarely
lasts more than a few minutes (the popular wisdom in Taiwan is that
you can fight for ten minutes but no longer; the police will usually
arrive at that point), but unless tempers cool, unless face is restored,
almost surely there will be reciprocation—revenge—within minutes or
hours, as the aggrieved party calls in his allies.
Back, however, to the Pirate Ship. At about ten o’clock, the sugges-
tion was made that we retire to another location. A number of alterna-
tives were mentioned—we decided to check out the Pyramid Club, a
new KT V parlor just down the block. In this case, there was no need to
drive. We ambled across one side street, up the stairs, and through the
fake bronze door. We told the hostess who greeted us that we were all
shareholders of the Pirate Ship Restaurant, which of course she knew
well. One of the group asked directly for Assistant Manager So-and-So.
Our credentials were good enough to put us at the top of the wait-
ing list, and soon we were led from the atrium through a low-lighted
black marble central bar area to a private room opening onto the sec-
ond floor balcony. The room was quite typical of this sort of establish-
ment: a carpeted space about twelve by twenty feet, a large-screen T V

at one end, plush leather sofas on three sides, a low glass tea table in
the center. The entire wall dividing the room from the corridor was
made of frosted glass etched with half-clad female figures in a style that
vaguely evokes classical Rome or nineteenth-century Paris. We took
our seats, close enough to reach for proffered cigarettes and betel and
far enough apart to accommodate the young female “drinking part-
ners” (peijiu) who arrived a few minutes later.
Busboys appeared periodically to wipe up the tables and serve
drinks and food; my friends addressed them (according to modern
KT V custom) as shaoye (which has about the same feel as the French
“garçon”), in imitation of customers in late-imperial teahouses and
wine shops. The order for a bottle of Hennessey XO cognac was
given indirectly, however. This was done through a young woman who
arrived before the others, exchanging business cards with each of us in
turn. The woman who performs this greeting and intermediary role is
an “assistant manager” ( fuli). Assistant managers are often a bit older,
more experienced, and sometimes better educated than those work-
ing as “drinking partners.” The management gives face to the custom-
ers by sending in a high-status hostess to greet the customers, while
the fuli can assess the guests for the establishment and, in some cases,
for herself. Thus low-status males provide basic amenities, while the
higher-priced and more specialized transactions (the drinks and host-
esses) are handled, discreetly, by a female intermediary. This protocol
is observed more strictly in places with high-class pretensions, but the
pattern obtains pretty much across the whole spectrum of clubs.
The cognac arrived, and the assistant manager poured and toasted
each of us in turn. After our glasses were refilled, two more women
arrived. They took their seats by two of the revelers and invited them
to drink. When the conversation ran out, the drinking games began.
The men challenged each other, as earlier at the beer hall, but here
the main interaction was between the men and the women. Soon the
songbooks came out and numbers were chosen—perusing the song
lists and choosing songs together with a “drinking partner” creates an
opportunity for intimacy, or at least the illusion of intimacy.
The men generally try to maintain a show of reserve and sobri-
ety— demonstrated, in part, by superior skill in the drinking games.
As the evening wears on, those less skilled or on a losing skid end up
drinking more and more “penalty wine” and have to work harder to
maintain the appearance of holding their liquor. On this occasion,
every round of brandy was “dry cup” (bottoms up). After all the beer
consumed an hour earlier, even the best drinkers in the group were
feeling the effects.
After another hour or so, however, we decided that this was not to
be the final event of the evening. Once the bill had been settled (and
the competition for the honor of playing host and paying it had been
resolved), we were escorted to the door by our “assistant manager” and
three of the hostesses. After the good-byes, we stopped at the stand of
a sausage vendor who was parked just outside to take advantage of the
KT V traffic. These sausage vendors often sport a bowl and dice, and
customers can play for a freebie. A-lang was feeling lucky that evening
and offered to treat us all—figuring that he could win a few rounds (he
won two and paid for the rest). As we gathered around to watch the
action, others offered up cigarettes and betel.
The group was going mobile now. We decided to go to the God
of Wealth Club. Motorcycles were left behind and we piled into three
cars. Following A-hian, we sped through the back streets of the Tai-
zhong Harbor Road area, finally pulling up at the God of Wealth
Club. We parked on a side street and walked around the corner to the
entrance.
Women
Some of the group often frequented the God of Wealth Club, so there
was no need to ask for such-and-such a manager. Two hostesses who
knew some of the group led us directly to one of the choice rooms,
right by the piano bar. We were joined almost immediately by five or
six young women, all in various stages of intoxication—it was already
well into the evening, and they had all been through quite a few tables
already. That, it seems, is part of the strategy—the last table of the night
will have a better chance at seduction. The women working in these
clubs are not directly selling sexual favors for cash. In fact, that would
completely undermine the pretext of the ritual—that the customer will
attract and, if sufficiently endowed with male potency, seduce one or
more of the courtesans.13 The challenge of the game, the uncertainty
of the outcome, provides a field of competition among men negoti-
ating identity and establishing degrees of prowess and power. It is a
stage upon which one can perform one’s masculinity. One pays direct
attention here primarily to one’s friends, sworn brothers, comrades—
a man among men, it would be unmanly to allow oneself to be overly
distracted by the feminine charms of the “drinking partners.” Such
unmanly behavior is subject to derision—in putting one’s own sex-

ual desire above male solidarity (zhong se qing you) , one is displaying a
pathetic weakness (such a display is derided as weak and unmasculine
perhaps because it is thought to derive from a man’s infantile attach-
ment to his mother). At the same time, the women engage the men
in sexual banter, joking suggestively about each man’s sexual prowess.
As closing time nears, this drinking session ends like many others—at
least one of the group offers to drive a couple of the women back to
their dormitory. We all say good night, and a few of us head off to get
a midnight snack and sober up.
The evening is not yet over, however. On the way to eat, one of the
group suggests going to a twenty-four-hour teahouse in the downtown
area. This place is where the die-hards—the after-hours crowd—hang
out. Here the smoking, betel chewing, and drinking continue—alcohol
along with the tea—and of course here too there are women to keep
us company and provide the aura of sexual promise. Here, the women
are a bit older, less elegantly dressed, and, perhaps given the lateness of
the hour, less than enthusiastic about the arrival of yet another group
of half-drunk carousers. In this establishment the women are some-
what euphemistically called “tea masters” (chayi shi); one of the group
suggests that they are easier to seduce than the higher-class drinking
partners.
Although it is a Taiwanese-style (that is, hybrid Chinese/ Japanese)
teahouse equipped with the proper tables, tatami rooms, and all the
utensils and equipment necessary for brewing and serving tea, we
order rice wine and continue to drink, accompanied by three or four
“tea masters.” After a few half-hearted attempts at conversation, the
men and women begin to play drinking games. After a few rounds, as
the company becomes even more tired and drunk, the men start to
chafe and argue among themselves, and one goes off to get sick in the
bathroom. The women don’t seem surprised by any of this, and one by
one, they slip off to join another party. One, an older hostess, remains,
and takes control of the table. When it becomes clear that the men are
not going to order any more wine or food and are about at the end of
their evening, she calls in a busboy for hot towels, then convinces the
men to order some rather expensive tea to sober up ( jie jiu; lit. “undo
the alcohol”). After preparing and serving the tea, she herds the group
to the exit (after prodding one of the crew to help his semi-comatose
comrade to his feet). Later a few of the group will go off alone or in
pairs to a brothel; the others will head home to sleep it off.
The next day, members of the group will get together again at the
Pirate Ship and discuss the charms of the various women involved in
the previous evening’s revelry. At some point one of the group who
had taken his leave at the God of Wealth Club insinuates a successful
seduction of one of the hostesses. While I have no reliable data on the
sexual relationships between hostesses and customers, it is apparent
from observing several groups of carousing men over several years that
“seduction” is achieved far less frequently than claimed or implied in
the next day’s recounting. Conversely, when occasionally one of the
group did become involved in a steady sexual relationship with a host-
ess, he tended not to talk about it. Rather, it would be obvious when a
particular hostess began to treat him with special affection and kept
her distance from the other men or when the couple began to exhibit
signs of “domesticity” in public. If the hostess was an assistant man-
ager at one of the more expensive clubs, the relationship could indeed
enhance the man’s status, assuming he was able to meet his social and
financial obligations as the girl’s sponsor. These obligations include
paying for frequent parties at the club. As credit is often extended
to regular customers, such relationships often end when a customer
overextends himself. Losing his credit (and credibility), a customer
who can’t pay his bar tab loses face but also causes “his” hostess to lose
face and prestige in the club.14
And Song
Such relatively genteel carousing as I found in Taizhong was in sharp
contrast to what my fieldwork in Taidong turned up. The difference
had much to do with the nature of my fieldwork in Taidong, as well
as with the quite different socioeconomic environment of that city.
Most of the episodes of carousing in which I participated during my
time in Taidong were connected in some way to the network of tem-
ples, shrines, and ritual specialists that were the “real” subjects of my
research there. My contacts in Taidong were mostly working-class men,
often engaged in marginal professions that occasionally included ille-
gal activities, most commonly (but not exclusively) gambling.
In the late 1980s and early ’90s, most of the entertainment busi-
nesses in Taidong were still quite rustic and “proletarian.” 15 A few
proper nightclubs and KTVs had appeared, but they were still outnum-
bered by “tea rooms” (chashi) and similar “front” establishments of the
sort that had by then all but disappeared from Taiwan’s larger cities.
At the temple feasts and the seedy downtown nightclubs, the drinks
of choice in Taidong were usually beer or Shaohsing wine. Even the

few who could afford XO brandy or imported whisky had little taste for
it and preferred the more dilute beverages. The hostesses in the work-
ing-class clubs of Taidong were often teenage girls from aboriginal vil-
lages or unmarried mothers forced into this line of work by economic
necessity. Others were virtual bond slaves, paying off gambling and
business debts—their own or, more often, those of an unlucky parent
or male sibling. Consequently, there were often more explicit negotia-
tions of payment for sexual favors here than in the high-class clubs.
When treating an important guest, especially a high-ranking potential
business partner (often with some status or at least connections to the
upper echelons of one or another gang), the host would also usually
provide him with the company of a prostitute after the drinking was
done. Conversely, the sexual innuendo in conversation and physical
interaction between the men and hostesses tended to be more sub-
dued than in the big-city clubs.
Neither the competitive drinking games nor the offering of ciga-
rettes and betel varies much from place to place in Taiwan. Nor does
there seem to be much local variation in the choice of karaoke songs,
although the ethnic origin (Mandarin- or Holo-speaking, for instance)
and educational background of a particular participant make some
difference in the songs he or she prefers to perform. One practice that
distinguishes Taidong men from their compatriots to the west is their
preference for a more “primitive” style of betel chewing. In most of
Taiwan, the betel nuts are cut open lengthwise and stuffed with a red
lime compound that includes herbal medicine, pepper, and other sub-
stances. In Taidong, however, most prefer the larger (cheaper) betel,
which is prepared by simply being wrapped in a leaf smeared with pure
white lime. A bit of the stem is often left on to show freshness and is
only removed (by biting it off ) just before chewing.
When treating guests from Taipei, Kaohsiung, and other cities to
the west, Taidong hosts often profess their preference for the rougher,
plainer, stronger, and presumably more masculine white lime betel,
while politely providing their guests with the more expensive, tender,
red lime betel (in this context, foreign, weaker, feminine). In part,
Taidong men derive their particular form of manhood by representing
themselves as hardier, rougher, and tougher than their more “civilized”
and urbane but weaker comrades from the other side of the moun-
tains. Most people in western Taiwan, if they think about Taidong at
all, imagine it as wild, distant, mountainous, and full of matrilineal
aboriginal tribes whose women are supposed to be particularly beauti-

ful while the men are weak and childlike. It is against this feminized
version of the frontier that men in Taidong—particularly those of Han
Chinese ethnicity—present these very same qualities of wildness and
primitivity as masculine and heroic. For them, Taidong’s backward-
ness is at once an embarrassment and a point of pride, the source of an
innate toughness and manliness lacking in their more “civilized” com-
patriots to the west. This tension between what is imagined as the out-
side world’s contempt and a local mythos of toughness gives Taidong’s
carousing—not to mention its everyday conversation—a very special
flavor.
The competitive nature of the drinking here was therefore, if
anything, fiercer than I observed in Taizhong. On many occasions I
was witness to evening-long business negotiations that involved con-
stant competitive drinking. At one session hosted by A-beng, the guest
of honor was a local gambling operator and temple devotee, a “big
brother” (that is, a high-ranking gangster) from Taipei.16 A-beng was
so intent on giving face to his guest that he didn’t even excuse himself
and step outside to vomit—he simply puked into a spittoon and, barely
missing a beat, joined the next toast. Far from being an exhibition of
physical weakness or social subservience, this was very much a demon-
stration of the real spirit of honor. By putting his moral obligation to
his guest above his own physical misery—and doing it in such a casual,
unhesitating way—A-beng acted heroically, not only giving face to his
guest, but also confirming both his own and Taidong men’s reputation
for stamina, fearlessness, and—not least—hospitality.
Carousing sessions in Taidong (and elsewhere) are sometimes
planned events. More commonly, however, they are spontaneous exten-
sions of get-togethers and celebrations that begin in late afternoon or
early evening. In Taidong I was often invited by temple devotees and
their associates or sworn brothers for rounds of carousing.17 Twice a
month (at the new and full moon) and after every festival or major
temple ritual, for instance, the spirit soldiers are rewarded with offer-
ings of raw and cooked food, fruits, alcohol, cigarettes, betel, and other
items. Once the spirit soldiers have partaken of their share and have
been sent off with a gift of spirit money, the devotees themselves feast
on the “leftovers,” the food of which the spirit soldiers partook only the
invisible essence, leaving the material substance. Since the devotees—
particularly the men of the temple—are understood to be in some
sense soldiers of the deity as well, they are rewarded too. Here they start
the ritual as sacrificers and conclude as receivers. The participants are

expected to indulge their appetites and enjoy themselves without res-
ervation, and the mood is generally open and raucous. Drinking goes
on throughout the meal and continues after everyone has (over)eaten.
Some then retire to another table or, in cooler weather, to a side room
of the temple to play cards, dominoes, or mahjong. At this point, if one
or another of the men is in an especially good mood and/or has had
a good month, he may propose moving the party to another location.
The carousing party (usually in smaller groups, divided by cohort)18
then moves off in cars and on scooters, leaving the women, children,
and a few of the older men at the temple to continue their gambling
and cleaning up.
On one warm Taidong evening in mid-1991 at the Loyal Harmony
Temple, the instigator of the carousing party was A-liong, a member
of the cohort then in their early thirties. A-liong was a core temple
member. His mother had been the beneficiary of one of the patron
deity’s first local miracles in the early 1950s, and family and temple his-
tory had become inextricably interwoven. Several of the carved deity
images that graced his home shrine had been “divided off” from the
temple’s altars. Among them was a figure of Da Ye, “Elder Lord,” the
tall, pale demon-bailiff who, with his squat, black sworn brother Er Ye,
“Second Lord,” escorts the souls of the dead to the halls of judgment.
A-liong had played the part of Da Ye in the Loyal Harmony Temple’s
Military Retainers troupe longer than anyone else. In his teens, like
many of those in this poor, run-down neighborhood, A-liong had been
in a gang and had been tattooed across the chest, shoulders, and upper
arms. After several years of sporadic employment, gambling debts, and
unpaid bar tabs and with a wife, two young children, and an arthritis-
ridden widowed mother to support, A-liong went over to the other
side. After testing into the civil service as a prison guard, he had the
tattoos surgically removed.
As a public servant, A-liong earned barely enough to keep the fam-
ily fed and clothed, and there was little if anything left over for hosting
reciprocal entertainment activities. However, A-liong was both natu-
rally generous and deeply conscious of his obligations as a core temple
devotee. Loath to accept the low status to which his financial situation
relegated him, he often went to great lengths to present himself as a
man of means and connections. One way he did this was to host par-
ties at establishments where one or another acquaintance was working
in some mid-level position, using the connection to establish enough
credit for a night or two of carousing with his friends. Nightlife spots
come and go, but Taidong is not a big city, and A-liong’s reputation
often preceded him. Through his temple connections and position as
a low-level law enforcement officer, he had managed to get most of his
past debts forgiven or deferred. On this particular night, he was feel-
ing flush and offered to take his friends out on a carousing spree.
Starting at a basement karaoke club (no private rooms) near the
Taidong railway station, the group occupied two small round tables,
leaving two or three seats at each open for hostesses or friends who
might drop in from other parties. We ordered two bottles of Shaohsing
wine and some snacks. A few very unenthusiastic hostesses joined the
group, two at a time, for a few minutes each. One of the temple group,
A-hong, had a wonderfully deep bass voice and took the microphone
for a couple of slow ballads. His performance drew attention all across
the room, and he was loudly applauded; a “big brother” at another
table toasted him once for each song. A bit later, A-liong went to sing
his own current favorite, “Red Light Pier” (Ang-teng Be-thau), a song
of love and regret with a tango beat; only our group paid attention
this time. After a few more rounds of drinking and singing, A-liong
suggested moving on. He negotiated the bill (later I learned he had
some trouble extending his credit and had to get help from others in
the group, much to their annoyance), and we headed back up to street
level.
From here, we rode to a place A-liong wanted to “introduce” to
us, a new “shrimp fishing pool” (diao xia chang) an acquaintance of his
had opened nearby. Like many such places, quite popular for a time
in the late 1980s and early 1990s, the pool was a concrete basin built
in an open lot under a temporary corrugated aluminum roof. On one
side of the pool, next to a single propane-fueled burner, stood two or
three racks of vegetables and seafood. Nearby were a few glass tanks of
fish and shrimp and a beer cooler. A few bottles of black market liquor
and packs of domestic and foreign cigarettes were displayed on a shelf
behind the cashier’s counter. No one was much in the mood to fish for
shrimp, so we ordered two dozen tiger prawns from the tank and bar-
bequed “self-service”-style on one of several hibachis scattered around
the “picnic area” next to the pool.
After we had eaten, the owner came over and joined us for a round
of beer; he thanked A-liong for coming and encouraged us all to drop
in whenever we had some leisure time, proclaiming the myriad joys of
shrimp fishing and letting us in on his shrimp-pool business philoso-
phy: reasonable cost, guaranteed catches, biggest and freshest shrimp

in town, and a clean, family atmosphere. Later on A-hong confided to
me that he thought A-liong took us there primarily because he could
treat us on the cheap. Business was slow, and the owner was anxious to
add to his customer base. This, of course, was a one-time event unless
A-liong could keep the new customers coming.
There were now six members of the group left. Having talked him-
self into the mood for more carousing, A-liong invited the owner to
join us at a newly opened “teashop.” The Azure Shores Teashop and
KT V was on the other side of town, but everyone seemed to be up for
the trip. The club was in one of the typical older store-front shop build-
ings near downtown; long and narrow, the reception area consisted of
a small sofa along one side and a counter that ran halfway along the
facing wall. At the far end of the room was a small, wood-veneer door
with a small, one-way mirrored pane right at eye level. Beyond the door
was a corridor and four or five small, partitioned karaoke rooms.
We were seated in one of the rooms near the front and ordered a
bottle of Shaohsing, a dish of dried sour plums (to reduce the potency
of the alcohol), and two packs of imported “555” cigarettes. Four host-
esses joined us almost immediately (business was plainly not booming
that night) and introduced themselves in the standard fashion, toast-
ing each of us in turn. Their gestures betrayed a studied awkwardness,
suggesting that they were probably new to this business and still “in
training.” As they raised their glasses, they asked each of us in turn
how we would like to be addressed.19 After small talk, I asked about
the girls’ hometowns and learned that the hostesses were all aborigine
girls in their teens and twenties, primarily Amis from towns and vil-
lages between Taidong and Hualian. Most had been in the city only a
month or two. Most spoke little Taiwanese other than a few epithets
and jokes they had picked up from the mama-san and customers.
One of the hostesses in our cubicle that night was an Amis girl
from Hualian who had just arrived a few days earlier, “introduced” by
a relative. It was her birthday and she had already been celebrating.
A-liong joked suggestively with her, but after a few rounds of wine, she
stopped talking to him and sat sullenly by herself at the end of the sofa.
After a few minutes, she turned back to the group and began toasting
everyone in turn, tossing down a full glass of wine at every round. A-
liong, apparently feeling attacked by her cold-shoulder protest and
drinking challenge, leaned over, put a microphone in her lap, and
insisted that she sing a love duet with him. This was the last straw for
her. Thoroughly drunk and angry, she threw down the microphone
(breaking it and probably incurring a deduction from her pay) and
stormed out of the room, pointing and cursing at A-liong as she left.
Two of the girls, apologizing, followed her out. A-liong stood up, curs-
ing and motioning for us to leave. Before we could make a move, the
mama-san knocked and entered; two tough-looking male employees
waited outside the door.
The mama-san, a somewhat plump woman perhaps in her mid- or
late forties, her hair permed, wearing a lime green gauzy dress and
white high heels, apologized to the group, looking at each of us in
turn. She criticized the girl, calling her hoan-a, the condescending
Taiwanese epithet for aborigines. She offered to pay the hostess fees
for the girl and buy us another bottle of Shaohsing. After a few tense
moments, A-liong relaxed and the conversation turned to joking and
making light of the incident. The mama-san called in one of the male
employees, who offered cigarettes and betel all around while the other
one went off for the wine. We took our seats again, and the mama-san
personally opened the bottle, poured the drinks, and toasted A-liong,
then each of us in turn.
When we had finished about half the bottle, A-liong politely but
coolly thanked the mama-san and indicated that we had business else-
where. A waiter brought in the bill; A-liong looked it over with a frown
and quickly disputed two of the hostess fees. He had lost face; not fin-
ishing the wine implied a partial rejection of the mama-san’s compen-
sation. Implying that the bill was padded was also a kind of retaliation.
But the mama-san, having already lost plenty on this group, was in no
mood to back off. The argument went on for several minutes; in the
end, A-liong assented. With obvious irritation, he emptied his pockets
and paid the bill. It had been a thoroughly unsatisfying and unsuccess-
ful episode for A-liong: slighted and embarrassed by the girl, he had
expected far more in compensation (a stronger show of respect and a
free pass for the evening). The final blow, however, was the frustration
in his own inability to resolve the situation in a way that maintained
either his social face or his male pride.
By now it was after midnight. We left the Azure Shores and headed
back in the direction of the temple, where some of the group had left
their scooters. On the way, A-liong, unwilling to let the evening end on
a sour note, decided to drop by the house of a friend he called A-kian.
We pulled into a small alley behind one of the older temples in the
north-central area of town. At the time, the neighborhood was still dot-
ted with one-story Japanese-era wood-and-stucco houses. We stopped

in front of a door near the corner of the alley, and A-liong knocked and
called for A-kian. A-kian came to the door, apparently not unhappy to
have visitors drop by unannounced in the middle of the night. The oth-
ers went on after a quick greeting, but A-liong and I were persuaded
to stick around for some tea. We shed our shoes and entered a small
living room furnished with rattan chairs, a carved teak tea table, and a
small altar. A-kian boiled up the water in an adjustable electric kettle
popular at the time among tea aficionados. He added some tea from
the open canister on the table, explaining that this was high-altitude
spring tea from a private stock of “competition tea” given to him by a
tea plantation owner in Nantou. After a round or two, we struck up a
conversation about spirit mediums and martial arts, shared interests
of all present. A-kian, I learned, was a “wanderer,” a low-level gangster
from Huwei in central Taiwan who had slipped across the mountains to
Taidong a few years before to escape one of the periodic, highly publi-
cized police “sweeps.” After things cooled down, he brought his family
over and decided to stay in Taidong, although he frequently returned
to Huwei and his work involved connections in both places.
Soft-spoken and serious, A-kian, who was in his mid-thirties at the
time, felt that the younger brothers today didn’t understand the ethics
of honor and reciprocity that made the underworld a way of life. He
explained:
In the old days, a lot of vendettas began as slights or misunder-

standings during drinking sessions in nightclubs. Things might
get bad if there was tension between the followers of rival “big
brothers,” and sometimes people ended up in the hospital. There
was always retribution, but it was in proportion—you got one of
ours, we get one of yours. And for things to get violent, it had to
be really bad, like refusing to acknowledge a toast or invitation
to drink not just once, but three times; I mean, nowadays, these
gangs of kids go after some innocent bystander and cut off his
arm because they don’t like the way he looked at them. They have
no concept of honor. I don’t bother going to the nightclubs much
anymore; I prefer drinking at home. But sometimes, of course, I
have to go for the sake of reciprocity [yingchou, here implying a
social obligation]. My wife doesn’t mind; she knows I won’t act
irresponsibly, and my friends are righteous men, even if we’re all
of the rivers and lakes [ jianghu shang de].
Performing Masculinity: Carousing as Ritual Contest

Men drinking together in China and Taiwan bring to the table a spe-
cific set of expectations and behaviors. The enjoyment of drinking
certainly includes the pleasure of intoxication and the experience of
camaraderie. However, the dominant expectation is not a pleasant
physical sensation but rather the trial of appetite, capacity, and social
ability.
The drinking and feasting practices of the ancient Chinese nobil-
ity are reproduced, in a fashion, in the rounds of reciprocal feasting
and entertaining that dominate social and political life for most men
in China and Taiwan. In the liminal space of drunken play—that is to
say, carousing—the characteristic qualities of manhood are produced.
Here the internal moral quality of “righteous honor” (yi) is material-
ized (actualized as yiqi ) in an enactment of prowess that calls up the
collective representations of an epic mythworld. True righteousness is
manifested not in the mechanical reproduction of rules in ritual, but
rather in the spontaneous evocation of a righteous spirit that displays
a mastery of social propriety and a sense —that is, a deeply intuitive
understanding and consequent embodiment— of righteousness or
other markers of masculine character (cf. Bourdieu 1977:10 –15).
To this point, I have stressed the importance of drinking perfor-
mance, the force of competition that drives such performance, and,
most important, the social efficacy built into drinking ritual. On the
last point, for instance, we can see the self-mortifying act of reciprocat-
ing toasts to the point of misery as a sacrifice that establishes a hierar-
chical bond (the sacrificer in the subordinate position). Indeed, the
“challenge” to drink takes a very specific ritual form: raising one’s glass
with the right hand, tipped in the direction of the one challenged, and
making eye contact. In responding, one raises one’s glass in return,
touching the bottom of the glass (unless one intends to demonstrate
or make a claim to social superiority) with the fingers of the left hand.
The “challenge” is called “a respectful offering” ( jing) , recalling the
fact that both host and guest—both superior and subordinate—recip-
rocally toast one another. Relative position in this transaction is enacted
in the timing of the toast and the attitude of the response. The act of
offering merely establishes or confirms a connection, the nature of
which is exterior to the toast itself.
The competitive element in Chinese social drinking is quite
explicit. The qualifications and protocols of social drinking are, reveal-
ingly, expressed in military terms: drinking requires both courage ( jiu

dan) and capacity ( jiuliang); one participates in a “drinking battle”
( jiuzhan) , drinks as a penalty/punishment ( fajiu) , and engages in
“drinking duels” (doujiu). In Taiwan, drinking as social contest is called
pinjiu, “all out” drinking.20 In most settings, such “combat” is good
natured, a form of mutual pleasure that creates or affirms a bond of
friendship and trust between friends or colleagues. This is certainly
the mood one finds among more educated, professional, middle-class
carousers (who increasingly go out in mixed, rather than exclusively
male, groups).21 Yet a challenge to drink can entail rather more com-
plex intentions and consequences. In Taiwan’s “rivers and lakes,” a
first meeting (whether arranged or by chance) often takes the form of
a contest, sometimes exuberant, sometimes aggressive. Social drinking
is also a favored opportunity for newly introduced members to estab-
lish their credentials.
The challenge to drink is not entirely benign, however. Refusing to
answer a challenge risks not only disgrace, but, on occasion, even vio-
lent retribution. Not to drain one’s glass when challenged may indicate
that one has reached the limit of one’s courage, but it is almost always
framed as the refusal of an offered gift. Failing to reciprocate, then,
implies a personal rejection. At times, it may be a deliberate move, cal-
culated to diminish a rival. But if that intention fails or if the refusal to
drink is motivated by fear or physical discomfort, it is ultimately seen as
a selfish, “unrighteous” act (bu jiang yi) . In that case, even if the rejec-
tor suffers no retribution from the rejected, it is he who usually ends
up diminished in reputation and face.
Song, Reconsidered
What, then, of karaoke? How are drinking and singing integrated
within the whole of carousing practice? Karaoke singing, unlike drink-
ing, seems not at all a competitive performance.22 Everyone takes a
turn—experiencing the vulnerability of the spotlight, the individual
sense of contributing one’s due, and the sense of collectivity that sub-
sumes this individual vulnerability to the protective support of the
group. In Taiwan, karaoke singers, encouraged and, one might say,
nurtured in the presence of their fellow carousers, seem remarkably
unself-conscious, especially when compared to Americans in similar
situations of potential public embarrassment. In inland cities of the
PRC, where some rather novel karaoke permutations appeared in the
1990s—street corner and curbside karaoke, for example—the level of
unself-consciousness can reach startling levels, as utterly out-of-tune

crooners keep the neighbors awake into the wee hours.
Singing here is a “performance” only in the most general sense; it
is not primarily for providing entertainment or demonstrating talent.
Rather, the enjoyment of public singing, for both the performer and
fellow participants, derives, in large measure, from the social mutual-
ity entailed in the act. As each singer takes the microphone, he enters
a space of emotion, drawn in some measure from the sentiments of
the song, drawing his fellow carousers in as he sings. This effect is, of
course, enhanced by a talented singer, but talent is largely beside the
point. The singer demonstrates trust in the group, putting aside his
own sense of vulnerability.
Singing is thus a shared enjoyment. Whether or not an individ-
ual actually enjoys singing is also beside the point. There is a public
expectation that everyone wants to sing and that the singing is both a
personal pleasure and a prestation offered to the group. The singer
professes trust in his comrades by his willingness to expose his emo-
tional vulnerability. Not to sing is to withhold oneself, to demonstrate
a lack of trust in one’s fellows, thus withdrawing from the group.
While there are certainly many other, equally important, reasons that
karaoke singing has become so essential to male carousing all across
East Asia, I suggest that effective social practice in Taiwanese society
requires just such an experience of emotional solidarity, one that tran-
scends—even as it legitimizes—the calculations and contests of recip-
rocal entertainment.
Reciprocal entertainment activities are essential as fields of pro-
ductive action for men engaged in transacting business, political
power, and/or personal prestige. Carousing, in particular, creates a
space in which men can enact themselves as both cultural subjects and
individual actors within their social milieu. Success in these endeav-
ors—not simply drinking and seduction, but also in the way one uses
one’s capacity for drink and sex to one’s social advantage—is a mark of
being a real man, a “man of prowess.” The particular forms of sociality
and reciprocity played out here are gendered—masculine—although
they depend directly and ineluctably on the labor of (particular kinds
of ) women.
In her study of “corporate masculinity” in Japan, Anne Allison
describes styles of “ritualistic male behavior,” prominently involving
“alcohol, women, and sexual play,” that serve to create “bonds of fel-
lowship” among men of all classes and backgrounds (1994: esp. 151–
167). The subjectivities created and enacted in the world of Japanese

corporate-sponsored nightlife ultimately produce a particular kind of
male ethos wherein the embodiment of masculinity entails “being a
male who can pay a female to service him” (1994:204). The perfor-
mance of this ethos in carousing on the company tab both reifies the
Japanese corporate hierarchy and serves to reproduce it (198 –202).
In Taiwan and the PRC, carousing practice binds individual desire
to the service of reproducing structures of domination, including cor-
porate and political institutions. Partly for historical reasons, carousing
in Taiwan in particular bears more than a superficial resemblance to
Japanese corporate nightlife practice: the ubiquity of alcohol, sexually
available hostesses, drinking games and “alcohol battles,” and karaoke,
to name only the most obvious common features. At a higher level of
abstraction, however, comparison yields as many differences as com-
monalities. Subtle differences (as in the example of economic organi-
zation above) reveal important cultural particularities that cannot be
ignored.
For instance, the institutional organization reflected in and served
by the practice of carousing is quite different in Japan and Taiwan.
Japan’s economy has long been dominated by large corporations
whose economic and social reach has had a profound influence on
culture, including concepts and practices that produce gendered iden-
tities throughout the society (both within and beyond the corporate
workplace). Unlike Japan, however, social and economic development
in postwar Taiwan has favored smaller, family-scale enterprises rather
than mega-corporations. One result of this convergence of this small-
enterprise-oriented mode of economic development and the localist
tendency in Taiwanese social organization is that the entertainment
economy came to depend mostly (at least through the early 1990s)
on smaller, local groups of consociates than on corporate expense
accounts. Clearly, the cultural, historical, and economic distance
between Japan and Taiwan, despite the impact of the pre-1945 Japa-
nese colonial presence and continued cultural and economic cross-
flows, is considerable.
Nevertheless, Allison’s description of “corporate masculinity” in
contemporary Japan sheds important light on carousing practice in
China and Taiwan. Entertainment that involves social interaction in
public venues is central to the reproduction of the structures and
ethos of a web of ideologies of dominators—in particular, by serving
as a forum for the production and amplification of both individual
identities and interpersonal relationships consistent with the patrilin-

eal social structure and the patriarchal ideology that defines the limits
of meaningful action and subjectivity within that social structure. Spe-
cifically, ritual competition establishes and amplifies both individual
identities within cohorts and the hierarchical structure and behavioral
norms of the cohort itself.
Camaraderie can be both egalitarian and stratified. Moral qualities
of character, such as selflessness and generosity, bear no intrinsic rela-
tion to hierarchy. Yet the communicative performance of such quali-
ties depends entirely on collective agreement of an individual’s place
on the social grid. Who provides the cigarettes and betel to whom, who
lights whose cigarette, who elicits immediate response and an empty
cup when toasting, who gets how much of the group’s attention when
singing, and who gets the attention, and perhaps the sexual favors,
of the working girls (and who pays this and the drinking and singing
bill) are all part of a process of negotiation, production, and domi-
nation—directly of the women and subordinate men and, indirectly,
of all those traditionally dominated within the patrilineally modeled
social institutions of Chinese society.
CHAPTER 7
Conclusion
Faces of the Gods
“ Young men should not study the Water Margin; old men should not
read the Three Kingdoms” (Shao bu du Shuihu, lao bu kan Sanguo) . Reflect-
ing on this aphorism, kung-fu master and raconteur Cheng Jiamiao
explained that from these two great Chinese classical vernacular nov-
els, and in their effects on men at different stages of life, one can dis-
cern the trajectory of a man’s life. Since young men are naturally hot-
tempered, the Water Margin’s stories of injustice, adventure, rebellion,
and violence would encourage them to be insubordinate and violent;
if one has avoided the dangers of a life of violence and safely reached
old age, delving then into the Three Kingdoms (a historical novel about
unbridled ambition, political intrigue, strategy, and deceit) could only
incite frustration and despair since at that point one lacks both the
time and vitality needed to fulfill the aspirations described therein.
But the real meaning of the exhortation is precisely the opposite: it is
because the two texts so vividly glorify the desires and passions of youth
and middle age, respectively, that young men love to read the Water
Margin and older men the Three Kingdoms.
In the Confucian order, the capricious child must be transformed
into an obedient son. As Zito notes, “The father whose sons obey him
do so because he has given them the example of his own filial obedi-
ence. They are yin to him as he is yin to his own father” (1997:212).1
This simple abstraction of opposing contrasts also renders a stereotype
of Chinese male subjectivity: be on your guard, for a filial son, bristling
under his father’s domination, can turn into an impulsive rebel at any
moment; and even the wisest father, fearing a loss of control and the
waning of his own vitality, can easily lapse into imperiousness. In this
scenario, a power struggle between generations would seem an intrinsic
cost borne by the patrilineal Chinese family. While not necessarily acted
out in every case, it is nevertheless a universal and dominant (though
often subconscious) feature of Chinese men’s emotional experience.
204
faces of the gods 205
In the lived reality of Chinese and Taiwanese working-class soci-

ety, there is considerably more room for the capricious, rebellious
elements of the masculine persona. We can see this in the social man-
ner of the young farmers and laborers in lakeside Bai villages and in
streetwise young Taiwanese men, gathered with their comrades for a
temple celebration, during the downtimes of a procession or just hang-
ing around waiting for something to happen; their interactions consist
in endless teasing, sexual joking, relentless cursing, and mock fight-
ing, dishing it out and taking it in equal measure. There is a continu-
ous play of performance, paradoxically mixing defiant bravado with a
sentimentality and mutual intimacy that, to those unfamiliar with the
ethos of Chinese friendship and brotherhood, seems utterly incongru-
ous. Even in challenge, there is a closing in, a merging of social space
that suffuses the moods and obligations of comradeship with individ-
ual interest and identity: cigarettes are distributed without ceremony,
scooters are borrowed and shared without a second mention.
Generosity and selfless identification with the group manifest the
spirit of honor. But honor has another face, that of righteous anger
and vengeance. Honor within the sodality maintains a boundary of
exclusivity. But even here, friendships and group cohesion are always
tenuous, the danger of selfish instrumentality—being “sold out” (chu
mai) — ever present. The tensions among brothers are amplified in
intergroup rivalries. In-group exclusivity entails a natural suspicion
of rival factions, which see each other first of all as competitors (for
turf, profits, reputation). This relationship is itself dichotomous, as
alliances are formed, only to dissolve, for sentimental or instrumental
reasons, into blood feuds.
The comradeship as well as the fighting—with fists, staves, knives,
machetes, or guns—is mostly the domain of those young men exhorted
not to read the Water Margin. The bosses, the “big brothers,” men from
their late twenties to early fifties (who, one infers, are allowed in the
prime of life to read both the Water Margin and the Three Kingdoms),
gather and direct (but rarely control completely) this individual and
collective violence. The role of elder would seem to require a far
greater measure of reserve and control, but not all “big brothers” act
like distinguished patriarchs, at least not all the time. A certain playful-
ness, as well as a potential for violence, extends to middle-aged pat-
terns of sociality among working-class men. This is most evident in,
but not exclusive to, men of lower status within jianghu society. At the
beer halls and seafood restaurants of working-class towns and neigh-
borhoods throughout Taiwan, it is not unusual to see a group of men

in their thirties and forties grappling together as one of their party
loses it at the end of a night of feasting and drinking, smashing dishes
and kicking over tables, his friends holding him back as he tries to grab
for a glass beer bottle, a table leg, or even the knife or gun in his belt.
Chinese male subjectivity, it turns out, is more than a binary struc-
ture, the crossing of a threshold from filial (or prodigal) son to wise
(or imperious) patriarch. Fathers are secondary (or absent) figures in
the Water Margin and the Three Kingdoms; instead, comradeship, face,
and honor are everything. Brother Qiu explained: “Friends are differ-
ent at different stages of your life. In your teens, it’s all about honor,
but when you reach middle age, friends are the ones who understand
what you mean, even without saying it.”
Realizing a measure of autonomy does not necessarily require a
son to be disobedient, nor does it necessarily diminish the social face
and emotional security of a father. Yet somehow, a degree of ambiv-
alence seems inevitable. Brother Qiu’s father was a powerful and
respected boss in a part of Taipei County known as a gangland haven.
He once hoped Brother Qiu, his youngest son, would go to college,
perhaps study abroad, and bring to the family the elite credentials
that he himself could not claim. He was thus understandably dismayed
when Qiu decided to be a professional fighter. Yet when Qiu headed
to Japan to train and compete, his father called on his business con-
nections there (many in the local underworld) to ensure that his son’s
career stayed on track. Qiu, for his part, was determined to escape his
father’s shadow. Yet after he returned to Taiwan, set up his own busi-
ness, and married, he found it impossible not to draw on his father’s
social account. All the same, Qiu said, “Out in the world, you depend
on your friends.”
For the wanderers and prodigal sons of Taiwan’s jianghu, fathers
are an often problematic presence, a source of fear, resentment, and
anger. This is especially true for those whose fathers were themselves
travelers on the rivers and lakes—as illustrated here, for example, in
the cases of Xiao Zhang (chapter 2), A-bao (chapter 3), and Lau-hoe-a
(chapter 5). The anger and frustration generated by (or at least articu-
lated in) the father-son conflict stands in sharp contrast to the “soft”
emotions and unquestioned loyalty these same men expressed when
speaking of their mothers. Repentant gangster and successful busi-
nessman Lau-hoe-a’s vexed relationship with his father may have had
complex origins, but he framed his anger in terms of filial loyalty to
his victimized mother. In his testimonials of personal redemption, he

often highlighted the pain he felt at seeing his mother’s grief (when
she came to visit him in prison) as the beginning of his own moral trans-
formation. In the Water Margin the most important and sentimental
bonds, other than those among real and fictive brothers, are between
certain key heroes—including the bandit chief Song Jiang—and their
mothers.2 The biographies of “righteous” bandits often highlight their
protagonists’ basic moral goodness by pointing out how devoted they
were to their mothers. Accounts of the early twentieth-century bandit
Bai Lang point out that he joined the jianghu brotherhood only after
obtaining his mother’s permission (Perry 1983:364). As Ye Qitian, the
“King of Taiwan Pop” and “brother of the path” in his own right, sings,
At the ends of the earth, there’s no one to lean on; I’m just a prodigal son.
Though I left home for the sake of the future,
I have no idea when I can make it; my prospects are hazy.
Forgive me, oh please forgive me for being so unfilial, my dear mama.3
The ethos of jianghu life permits only a very narrow range of emo-
tional expression in everyday contexts. Not only are some emotions
effectively proscribed (as inappropriate or unmasculine), but the
manner and form of expression are severely limited as well. Violent
modes of ritual performance (and ritualized modes of violence—why
else would real-life gangsters choose to “go out in a blaze of glory”?)—
and alcohol-disinhibited carousing dramatically expand the range of
expressible emotions.
Yet within these arenas of masculine performance, only certain
kinds of affect are appropriate, and only certain modes of expres-
sion are permissible (and usefully communicative). Righteous anger,
anguished grief, or a furious response to a face-diminishing insult are
unambiguously manifested in the facial expressions, gestures, pos-
tures, and even violent aggression that comprise the “durable disposi-
tions” of wu masculinity. Guilt and grief are equally valid masculine
affects, but they cannot be resolved passively (such as by crying in one’s
beer); they must be objectified (and transferred) through a passion-
ate violence or, perhaps, channeled through a song. Because they are
strongly kinetic, unmistakably bodily (as opposed to, say, textual) prac-
tices, ritualized violence and ritualized drinking (and singing) enable
the transformation of threatening, potentially debilitating, affects into
personally and socially generative energy.4
Here I extend Pierre Bourdieu’s notion of “field” to include not

only things like professions and social identities, but also worldviews
(including belief systems) that, like the more restricted sense of “field,”
are similarly constituted through a dialectical interconnection with a
particular habitus. Thus the “field” here is a worldview that includes
(as a necessary though not necessarily pervasive constituent) as a
contextualizing framework the logic of Chinese popular religious
demonology; by “habitus” here I mean not only the discrete, “sacred”
realm of ritual or ritual that takes place specifically in temples or other
spaces ontologically separated from the everyday, but rather the diffuse
complex of bodily (including neural, intellectual, “mental”) habits
through which those worlds are infused into bodily and everyday space
and in fact become an unconscious interpretive filter on the world.
For instance, one may become accustomed to turning one’s gaze in
a particular direction in hearing some unexpected sound that raises
suspicions of the presence of ghosts, let’s say; the envisioned “fact”
and conditioned response includes physiological reactions such as
raised blood pressure, increased pulse rate, and so on, which lead to
secondary symptoms such as a momentary sense of breathing difficulty,
which refers to the behavior of predatory ghosts at certain times of day,
at which point one mentally references the correlative logic of daytime
and yin-yang continuum that reinforces (or reassures) the suspicion. It is
probably impossible (and uneconomical) for any individual researcher
to document such a mechanism beyond this kind of microlevel
example. But it is precisely at this level of unconscious operation that
the subjective and objective merge, the habitus generated within the
logical boundaries of the field, and the field reproduced through the
operations of the habitus.
To explain human behavior, including the emotional and psycho-
logical dimensions of experience and decision, we must account for
the communicative, representational, and aesthetic aspects of ritual
and social action. My aim has been, first, to discover what actors intend
to signify and what effects they desire or anticipate; second, to exam-
ine their techniques of communication, taking note of how individu-
als make use of both “scripted” and spontaneous behaviors (phrases,
manners, gestures, expressions); and finally, to discern something of
actors’ internal dialogues and affective states. I have not simply let my
informants speak for themselves, which could yield only a fragmented
and probably incoherent account; to simply transcribe their verbal
narratives and self-presentations would do as much a disservice to their
communicative intentions as to the enterprise of ethnography. Rather,

I have endeavored to faithfully document and reconstruct individu-
als as living persons, not mere “personalities” or disembodied verbal
traces. I have foregrounded the transactional nature of participant-
observation fieldwork, endeavoring at the same time to transcend the
limitations of the solo ethnographer’s perspective by learning what I
could of informants’ life milieu, their families and circles of friends
and associates, their general moods and temperament, and their hab-
its and quirks.
I have also sought to uncover the salient social and historical layers
of events that featured in my informants’ conversations and narrations.
At times, I was confronted with more information than I wanted about
matters that did not seem important at the time but whose significance
I have subsequently come to appreciate.5 No fact is obviously irrele-
vant as one tries to reconstruct the complex, largely hidden, dynam-
ics of ritual and social actions and interactions. But eventually some
facts emerge as more salient than others. What’s more, the facts one
sees as significant and carefully documents in the heat of the moment
are not always the ones that shed the most light on the experience in
retrospect.
I draw one final theoretical inference: affect (in the sense of emo-
tion, sentiment, mood) is the core of performance aesthetics; the
social efficacy (that is, productive function) of ritual practice and
expressive performance is premised on the bodily nature of emotion;
affect is, then, the very medium of collective experience. Only in this
way can the self, as subject, emerge “in the field of the Other” (Lacan
1978:208)—that is, in the subjective consciousness of oneself as an
object in the consciousness of an Other; or, as Žižek explains it, “Imag-
inary identification is identification with the image . . . representing
‘what we would like to be’ ” (1989:105). Ritual performance, however
scripted and rehearsed, is very much a space of creative agency. The
violence of A-kiat-a’s trance attack on the censer bearer, Handan Ye’s
defiant sacrifice, the Bai torch climbers and tree stealers—all are “iden-
tifying” or, perhaps better put, creatively and dynamically producing “the
image representing what they would like to be.” Nor are those who
partake in the performance merely passive spectators, but they are at
once product (as a collective, “intercorporeal” subject) and producer,
the objectifying “field of the Other,” the doxic “I” coterminous with
the ideological limit of collectivity.
Ritual performance achieves its effect through both symbolic allu-
sion and conditioned response. The kinetic expression of certain pow-

erful emotions creates a kind of collective body, an “intercorporeality.”
Though conventionalized and naturalized in unique ways within dif-
ferent cultures, the facial expressions and aggressive gestures of grief,
anguish, and furious anger shared by the iconographic representations
of Chinese gods, entranced spirit mediums, emoting songsters, and
enraged street toughs are, I would argue, immediately, intuitively, accu-
rately, and indeed intercorporeally interpreted by most human beings.
Emotions are not mere psychobiological facts that precede or tran-
scend culture, but in the case of the primary passions, there are core
features, including modes of expression (both static, as in a particu-
lar facial expression or posture, or moving, as in striking the chest or
jaw in grieving desperation, say) that are so common and immediately
intuitable as to be considered transcultural, if not precultural. Culture
does not produce the primary passions or at least some of the body
responses/expressions, but it certainly provides different interpretive
contexts, and thus through feedback emphasizing one or another
particular nuance or mode of expression (as in the historically and
morally central “martial righteous anger” pose of Chinese gods and
mediums) it creates a self-sustaining, culturally independent discourse
that continually feeds on instances of that emotion while channeling it
into a particular, bounded range of meaning and possibility specific to
a particular cultural milieu.
Taiwanese spirit mediums signal that they are about to enter trance
with a series of conventional signs. The beginning of the trance is
often marked by a display of extreme discomfort that typically includes
retching, belching, and even vomiting. The medium’s face and body
then begin to contort and twitch. At this time he may also vocalize a
series of inarticulate grunts and sighs, as if resisting as well as suffer-
ing the approach of the possessing spirit. During this transition, as the
medium loses control of his body and consciousness, the most obvi-
ous expressions (of the body, face, and voice) are those of pain and
struggle. During these few moments, the medium’s body, facial expres-
sion, and voice become fully “molded” into those of the possessing
spirit. Before the medium/spirit begins to speak to those assembled or
to move or act in any way directed specifically at the audience or the
situation, in this brief, transient set of motions he establishes clearly
that his body is genuinely and completely inhabited and controlled by
an Other. If the possessing spirit is a martial (or a generally fearsome
and awesome) one such as Wang Ye, the medium’s face contorts into
Demonic spirits are conventionally portrayed with distorted features and expressions of
profound grief and rage, evoking their ghost-like existence and tragic, usually violent
deaths. A minor, ghost-eating “general,” Loyal Harmony Temple, Taidong, Taiwan.
a deep, scowling, glaring frown. Quite unlike a mask, the medium’s

scowling face is not frozen but transitions between a scowl, glare, or
sneer. The medium’s expression conveys the agony of a painful strug-
gle, an inward-, then outward-directed anger. This kind of emoting is
far out of the ordinary in normal social interaction. Both the intensity
of affect and the passion of expression indicate unambiguously that
the spirit has taken possession of the medium’s body.
By embodying, not merely mimicking, the “image” of what they
would like to be — demonic, heroic, violent, capricious, austere, com-
pletely self-possessed—ritual actors dissolve the aesthetic distinc-
tion between subject and object, role and instance. This is the very
object of wu masculinity: to embody (self-)productive, (self-)transfor-
mative, efficacious power. Therein must lie a great—but ultimately
silent—pleasure.
Notes
Chapter 1 Introduction
1. Established in 1991, the Journal of Asian Martial Arts has since filled a
space somewhere between the pulp magazines and peer-reviewed journals. In
intent at least, it represents a small but important step toward a serious martial
arts scholarship.
2. The Shaolin Temple, of course, holds a special place in the myth and
lore of Chinese martial arts: in the popular account, the Shaolin Temple and
the Daoist Mt. Wudang were both the source and epitome of the Chinese
martial arts. Some readers, then, might be surprised that the Shaolin Temple,
or indeed any element of Chinese Buddhist tradition, is not featured more
prominently here. The Shaolin Temple was indeed home to a venerable tradi-
tion of fighting monks, and they most certainly played a role in the evolution
of the Chinese martial arts both within and beyond the monastery walls (Sha-
har 2001). It was primarily imperial officers and discharged soldiers, however,
in considerable numbers and across the broad range of social and economic
activities, who popularized and transmitted the martial arts to the broader
lay/civilian population. The techniques of the Chinese fighting arts developed
over a period of centuries, mostly in the crucible of military combat training
(even some records of Shaolin martial activity link the temple to historical
military campaigns), and it is the power of the emperor’s armies and the prow-
ess of fighting masters that captured the popular imagination and became
enshrined—literally—as, among other things, the military gods of the popular
pantheon and the exorcistic martial techniques of Daoist ritual magic.
3. Among other worthy areas of inquiry, for example, there is a complex
and fraught relationship among religion, the martial arts, and Asian national-
isms during the colonial, war, and postwar/postcolonial eras.
4. By “nascent field,” I do not mean to imply that no serious scholarship
has been published or is currently being done. I am merely emphasizing that
while the martial arts are very much a part of mainstream Chinese (and now
global) culture, most research on the martial arts in Western languages (with
the notable exception of Meir Shahar’s work and a few of Stanley Henning’s
articles) focuses on and has been motivated by issues and interests internal to
the martial arts community itself.
5. Many temple cults feature similar arsenals of spirit weapons, but the
set described here is unique to temples dedicated to the goddess Mazu, one
213
214 notes to pages 4–6
of the most popular and important deities in Taiwan, southeastern coastal

China, and many overseas emigrant Chinese communities. Mazu’s “birthday,”
the twenty-third day of the third lunar month, marks the beginning of one
of the most active phases of the ritual calendar in Taiwan, Hong Kong, and
elsewhere in Mazu’s “realm.”
6. My initial (mis)understanding of the ritual was that it was a folksy ver-
sion of a martial arts demonstration. The meaning I, as a practitioner, initially
read into the performance had little to do with the intentionality of the actors
or the sponsors: I was viewing the event with a trained and critical eye, but my
evaluation had little to do with what I was actually seeing. The crowd’s obvi-
ous appreciation of the one highly skilled performer made it clear that skill
and aesthetics were not irrelevant, but they were secondary to other, symbolic
dimensions (the first priority, for example, is to field a fully manned team,
even at the expense of lower-quality performances).
7. I use the term “symbolic violence” here in its simplest, most generic
sense to suggest that the performance of martial arts postures and gestures
always communicates a threat of coercive force, even in formalized ritual rou-
tines like the ones described here (in fact, to be considered “rituals” rather
than mere performances entails “real” violence, given that these actions are,
in theory, directed at confining, torturing, and killing real “victims”—that is,
malevolent ghosts).
8. Festival processions similar in broad outline to what we see today
appear in thirteenth-century records ( Von Glahn 1991:670; Szonyi 1997:118 –
119). The use of parading shamans and actors masquerading as demon-quell-
ing gods in public exorcism spectacles, called nuo, is even more ancient.
9. The “local customs” chapters of Ming, Qing, and early Republican-era
gazetteers invariably include some details about local marriage and funeral
customs but are otherwise structured around a generic catalogue of obsolete
practices derived from the Confucian classics (for example, the largely extinct
“capping ceremonies” [ guan li] marking the transition to manhood). Many of
the entries, therefore, simply take the form “Ritual X has not been practiced
for some time.” Practices such as spirit possession and healing are mostly char-
acterized as unscrupulous con games aimed at fleecing ignorant, desperate
villagers. These accounts are highly formulaic, and other than indicating that
this or that variety of practice may be found in the area, they rarely provide
information about real local conditions. They also tend to use archaic literary
categories (for example, labeling as wu all spirit mediums, shamans, and ritu-
alists) in place of local terms that distinguish such roles and vary widely from
region to region. This linguistic practice has allowed these authors to present
what was, and is, an empirically diverse ethnographic reality as an epistemic
unity that validated the Confucian (and imperial) vision.
10. There is also a body of discourse that belongs to the highly politi-
cized and factional martial arts profession. It is largely devoted to debating
the merits of a particular technique, style, or school and to internecine battles
over legitimacy and authenticity. The concern with authenticity is manifested
through interminable genealogical disputes—who was whose teacher, who
transmitted what technique to whom, how the ancestral teacher (perhaps a
notes to pages 6 – 11 215
wayward Buddhist monk of the Shaolin Temple or a Daoist hermit from Mt.
Wudang) acquired his secret insights from watching animals fight, and so
on. Legitimacy, on the other hand, begins with a declared devotion to the
preservation and promotion of tradition, culture, patriotism, and morality.
In modern and contemporary schools, this is most evident in the rules or
“philosophy” impressed upon new students. These are variously presented
as Buddhist, Daoist, or Confucian in origin but consist mostly of a handful
of well-known and well-worn popular formulae preaching loyalty to superi-
ors and obedience to parents, using one’s skill only for self-defense and to
protect the innocent, and the like. Martial arts lineages and associations also
emphasize such patriotic and nationalistic themes as the preservation of tra-
ditional Chinese culture and the “self-strengthening” value of martial arts
training.
11. The association of supernatural power with martial prowess and mili-
tary rank was, for example, already a prominent feature in the cosmology of
the Celestial Master sect (Tianshi Dao, also known as the Five Pecks of Rice
Movement), active in Sichuan in the late first and second centuries CE. See,
for example, Chen Guofu 1963.
12. Millenarian eschatology is itself often rife with violent imagery, but
the sectarians’ place in the history of Chinese religions does not begin and
end with armed violence. Devotees of the Queen Mother of the West during
the Han dynasty (206 BCE–220 CE) were, for example, among the first to
articulate a concept of personal salvation. They were also the first popular
movements to break down the boundaries of clan and village to create com-
munities based on shared belief and devotion rather than particularistic local
ties; in both regards, early millenarian sects prepared the field, as it were, for
the arrival and assimilation of Buddhism (and more recently, Christianity).
13. Remnants of sacrificial offerings are found in abundance in late
Neolithic sites in northern China; human victims have been found at a few
Neolithic sites as well. The advent of animal (and human) sacrifice parallels
the appearance of fortified towns, a tremendous increase in the kinds and
quantities of symbolic wealth such as finely worked jade ornaments, and a con-
centration of that wealth in the hands of what was likely an emerging warrior
aristocracy (Chang Kwang-chih 1986). Human funerary sacrifices, a defining
feature of Shang aristocratic burials, were replaced starting in the Western
Zhou with inanimate substitutes. I am not suggesting that these ancient cus-
toms are directly ancestral to modern practices; nevertheless, the offering of
sacrifice has been a consistent feature of every region and era of Chinese
civilization. Moreover, from Shang oracle bone texts down to the present, sac-
rifice has been consistently described as a form of prestation that expedites
and mediates an ongoing, reciprocal relationship between the living and the
dead. Cf. Girard 1977; Hyde 1983.
14. Stephan Feuchtwang, for instance, records that people in Mountain-
street, a village in northern Taiwan “likened gui [homeless ghosts] to gangs of
bandits or to the young men who swore brotherhood and ran rackets, gangs
which also formed the militia who acted as the gods’ retinue in procession”
(1992:108).
15. Gelaohui, literally “association of big brothers,” is a generic label for

the secret societies that formed during the late Qing period.
16. I came across the passage while searching for historical accounts of
inter-village feuds in Dali, where I was then doing fieldwork. Although this
vignette has only an incidental connection to the rest of the book, I was struck
by the way it condenses many of the book’s main themes.
17. These performances evoke, if not intentionally, similar episodes in
myth narratives about the mischievous and patricidal child god Nezha and the
insubordinate immortal monkey god Sun Wukong. See also Sangren 2000:200
passim and chapter 4 below.
18. In Taiwan, for example, the shadow economy accounts for nearly a fifth
of the total GDP, and much of this productive activity is generated by jianghu-
related enterprises, both legal and otherwise. Some of this activity represents
practices that transgress moral as well as legal boundaries: gambling operators
often extend credit (at high interest); those who guard and stand lookout often
double as debt collectors and enforcers, resolving conflicts through intimida-
tion and violence. On the other hand, it is a stretch to describe most local
brotherhoods as “gangs.” Many of the brothers have been simply channeled
into the vocational track in Taiwan’s, or Hong Kong’s, or China’s rather merci-
less education systems and find themselves with few appealing prospects in the
legitimate work world. Many young Chinese and Taiwanese working-class men
thus see life on the “rivers and lakes” as a viable occupational choice.
Chapter 2 Violence, Honor, and Manhood

1. In ethnic Chinese communities outside the People’s Republic of China
(PRC)—Hong Kong, Macao, Taiwan, Southeast Asia, and elsewhere—although
at times subject to varying degrees of official regulation and diverse non-Chi-
nese cultural influences, Chinese popular religion has been practiced without
the kind of violent disruption that occurred in mainland China after 1949.
From the early years of the PRC, the Communist Party increasingly regarded
popular religion, including its sacred objects, buildings, and ritual practices,
as “feudal superstition” ( fengjian mixin) , an obnoxious remnant of pre-revolu-
tionary society. During the Great Leap Forward (1958 –1960) and the Cultural
Revolution (1966–1976) nearly every structure and object connected with
religion in mainland China was destroyed, damaged, or—in the case of many
churches, temples, and monasteries—gutted and converted to other uses. In
the post-Mao era of opening and reform, despite the continuing use of dispar-
aging rhetoric and sporadic “anti-superstition” campaigns, many once banned
practices have reappeared, and many temples have been rebuilt or restored
(especially, but not exclusively, in rural communities).
2. In southwestern China, among other regions, the kowtow, literally
“knocking the head” (a half prostration from the kneeling position), is a com-
mon gesture of obeisance in Buddhist, Daoist, as well as popular temples. In
Taiwan, however, it is generally limited to Buddhist practice and moments of
particular solemnity in other ritual contexts.
3. In addition to regional variations, there are even differences among
neighboring temples—for example, in Taiwan’s Royal Lords (Wang Ye) tem-
notes to pages 23–28 217
ples, one must also make an offering at the flagpole and “guardhouse” erected
to one side of the temple courtyard. The sequence as described, however, is
generic enough to be recognized by practitioners most anywhere.
4. Among the characteristics borrowed from imperial institutional archi-
tecture are the stone lions, one male, one female, that guard the entrance of
most large popular temples.
5. Reed’s research on the late Qing dynasty bureaucracy suggests that the
reputation of such low-level functionaries was not entirely deserved (B. Reed
2000). The common stereotype, however, is fully congruent with (and likely
contributed to) the way the demonic minions of the gods are represented in
the iconography and narratives of popular religion.
6. There may seem to be little difference between yamen runners and
corrupt officials of any era or culture. In Chinese society, however, the prin-
ciple of social reciprocity, enacted through exchanges of gifts and favors and
displays of generosity, is absolutely central to the creation and maintenance
of social relations (whether between equals, as friends and kin, or across the
levels of hierarchy), creating a somewhat looser standard for what might oth-
erwise be considered corruption. For a discussion of factional politics and
corruption in contemporary Taiwan, see Chin 2003; see also Jeffrey Martin’s
work on the culture and practice of policing in Taiwan, especially the subtlety
of the relationship among police, local factions, and operations of the shadow
economy (Martin 2006).
7. Lin Jiafu, personal communication.
8. In addition to rehabilitated bandits, Ming and Qing law reserved the
possibility of assigning former convicts to serve as assistants to yamen runners
(Hansson 1996:49).
9. Not every talisman contains all of these elements, and the crosshatched
cipher at the bottom has a number of different purposes. As esoteric devices,
written talismans can and are interpreted and explained in divergent, some-
times conflicting, ways by different ritual experts. My description synthesizes
the explanations of three ritual masters and is generally consistent with the
references, scholarly as well as popular, that I have consulted over the years.
10. The most persuasive accounts appear in the works of Feuchtwang
(1976, 1992), Sangren (1987), Weller (1987), and Zito (1996).
11. Feuchtwang (1992:192) points out that the local militias that partici-
pated in Daoist ritual processions also served local gentry acting to maintain
the imperial order but that these same bands were also associated with crimi-
nal activity and factional violence. Historically, rebel armies were made up
of coalitions of such militias (Lipman and Harrell 1990:65–81; Kuhn 1970).
Most such armies were led by individuals whose experience as local warlords
inspired ambitions to expand the scope of their power (Meskill 1979; Kuhn
1970; Perry 1980), although only one such local leader ever succeeded in
establishing an imperial dynasty (that is, Zhu Yuanzhang, founder of the
Ming). In counterpoint to Feuchtwang’s argument, however, recall that most
sectarian movements tended to model their own organization on familiar
imperial institutions. They were thus substitutes, not alternatives, as it were,
that ended up reproducing rather than subverting the imperial order.
12. The distinction I make here between manhood and masculinity can
easily be framed in Chinese terms as the complementary opposition of xing,
“nature” or “disposition,” and ge, or “standard,” “pattern.”
13. My point is premised on Angela Zito’s insight into the logic of impe-
rial sacrifice, in which the emperor, as a “body of yang power,” ritually cre-
ated the imperium and himself as its creator, there being no possibility of an
emperor without the imperium, which could not exist without the creative
power of the emperor (Zito 1997:150).
14. A man needs potency and vitality to fulfill his filial duty to produce
heirs. Fatherhood, in turn, is the non-negotiable foundation upon which a
son bases his claim to succeed his own father as patriarch. Yet here lies one of
the greatest challenges to the integrity of the Chinese household: sons desire
autonomy, which usually entails the division of household property and the
end to a pooling of family labor. Fathers correspondingly strive to control
their sons and resist household division, as family division signals a weaken-
ing of their productive powers, a contraction of social prestige (face), and
ultimately an intimation of their own mortality (Sangren 2000:199). Thus a
vigorous patriarch, as the emperor is expected to be, precludes the possibility
of his sons’ autonomy by both his physical vitality and his ability to postpone
the division of household property.
15. The Hokkien term lomoa* is probably, though not unquestionably,
cognate with the standard Chinese liumang , literally (now archaic) “rootless
commoners,” “itinerants.” (The Hokkien version is written with characters
pronounced lu man in standard Chinese, meaning a kind of eel. There is a
southern Fujianese folktale that purports to explain why “eel” came to mean
“hooligan,” but it is a bit far-fetched and probably a case of reverse etymol-
ogy.) Only in the late imperial period did liumang largely lose its earlier, more
general meaning and take on the narrower and clearly pejorative sense of
“hoodlum,” “hooligan,” “gangster.” In addition to the colloquial meaning,
liumang is currently a legal category in both the PRC and Taiwan (see, for
example, Taiwan’s series of “anti-hooligan” laws [ jianxiao liumang tiaolie], the
latest version amended and passed in 2005). The official definitions, though
extensive, are nevertheless vague, giving license to occasional police “sweeps”
of “undesirables.” Liumang and lomoa* are unequivocally derogatory but are
occasionally used ironically or humorously as self-designations. Within jianghu
circles, men refer to themselves and comrades as “brothers of the [dark] way”
(daoshang xiongdi) or “brothers of the rivers and lakes” ( jianghu xiongdi) . Holo-
speaking Taiwanese “brothers” sometimes describe themselves as chhit-tho lang
or chhit-tho-a, literally “players” or “men of leisure.”
16. The origin of the term “rivers and lakes” lies in the watery topography
of central China, dominated by the Yangtze River and its tributaries. Along
and between the rivers are thousands of lakes and marshes, interconnected by
a dense web of manmade canals. As China’s population shifted south after the
Tang dynasty, this vast network of waterways and the related development of
coastal trade facilitated the growth of a new mercantile economy. This in turn
made sojourning a viable life strategy, favored particularly by men and women
in the lower economic echelons. The rivers and lakes thus came to represent
the inherent risks and rewards shared by all of those who found themselves
frequently, or permanently, outside the boundaries of conventional in-groups,
including high-stakes merchants, as well as peddlers, highwaymen, and sing-
song girls(among others).
17. Those in the jianghu have included nearly all of those who have lived
an itinerant life, including wandering actors and storytellers. Some of these
were classified along with prostitutes and others as “mean people,” jianmin.
Thus while the denizens of the “rivers and lakes” traditionally did not lead
a life on the “dark path” or have a criminal lifestyle, they were depicted as
being rootless and as practitioners of morally questionable (even if popular
and legal) occupations. This sense of jianghu persists and is even expanded in
contemporary usage; entrepreneurs and those in service trades that involve
selling one’s skills and talents—cooks, for example, and even piano teach-
ers—may somewhat ironically refer to themselves as being “of the jianghu.” At
the same time, the most common uses of the term in contemporary discourse
refer either to the mythic jianghu of literature and film or that of criminal
enterprise.
18. Wulin literally translates as “forest of martial [heroes],” evoking the
gallantry and skill of the knight-errant tradition. At the same time, it echoes
the rulin, or “forest of scholars”—that is, the venal, corrupt world of imperial
officialdom—suggesting that both careerist literati and martial adepts shared
an obsession with fame and reputation.
19. Shuihu zhuan, also translated as Outlaws of the Marsh and All Men Are
Brothers, is an episodic account of the 108 righteous, if sometimes hypervio-
lent, outlaws of Mt. Liang. It is generally considered one of the greatest novels
of late imperial classical vernacular literature, as well as being the prototypical
work of the Chinese knight-errant literary genre.
20. Sun was the daughter of a bandit; Zhang was a gardener in a Buddhist
monastery who, having unintentionally killed a monk and burned down the
temple, became a highwayman in the same area as Sun’s father. The elder Sun
caught and subdued Zhang, forcing him to marry his daughter in exchange
for his life.
21. Similar phrases have similarly broad currency: chudao, or “entering
the way,” can be used by Daoist disciples as well as actors and prostitutes. Xia-
hai, or “leaping into the sea,” was a popular expression during the 1980s and
’90s in mainland China that described the high risks faced by officials and
intellectuals who chose to leave behind secure, prestigious, but low-paying
positions for the risks, and potential rewards, of private business.
22. Reports by social workers in Taiwan suggest that violent juvenile
offenders are likely to have vexed relations with family members (Zhang Mei-
feng, personal communication; see also Liu 2004). The data derived from a
relationship among researchers, the courts and police, social workers, and
research subjects, however, are not comparable to the ethnographic material
I present here.
23. This may be a universal feature of the subaltern reaches of urban soci-
ety. See Winslow’s (2001) study of bouncer culture in the northern England
rust belt, for example.
24. Taiwan gangs are more of a cultural hybrid, having evolved under
the influence of Qing-era secret societies, clan feuds and subethnic violence,
strongman-led militias, Japanese yakuza traditions (including the practice of
full-body tattoos and the use of samurai swords in assassinations and revenge
attacks), and, more recently, displaced mainland crime syndicates associated
with the Chiang Kai-shek regime (especially military intelligence) and styles of
teenage rebellion inspired by American and Japanese popular culture.
25. The knight-errant appears in the “Song of the Knight-Errant” (Xia
ke xing), by the peripatetic Tang dynasty poet Li Bai (Li Po, 701–762), and
celebrates men of action and martial heroism. The last couplet, quoted below,
compares the valorous, self-effacing spirit of knights-errant to the vanity of
those who grow decrepit searching for the secret of longevity. Translation by
the author.
Dying bravely, without regret, the knight-errant’s bones are fragrant;

he stands proudly among history’s heroes.
Who, then, would waste precious time locked in the study, poring over useless
books until his hair turns white?
26. Some emperors deliberately cultivated a literary persona, emphasiz-

ing wen over wu. But for most subjects (and in the idiom of popular religion),
the emperor was primarily an abstraction rather than a personality. Thus the
predilections of a particular emperor as an individual had little to do with the
general conflation of “emperor” with a kind of transcendental power in the
popular conception that, if anything, was weighted toward the martial rather
than the civil-literary.
27. Confucian ritual has both wen and wu forms, but this correlation
effectively “civilizes”wu. Confucian rituals that allude to wu are in themselves
as genteel as the wen forms, thus making tangible the elite structure of value
that subordinates wu to wen; wen is valorized, wu is derogated. This subtle,
abstract conceptual distinction is ideological in that it affirms and sustains
the relative social, cultural, and political prestige of the educated literati and
official elites (the “people of wen,” wenren).
28. There are dozens, perhaps hundreds, of examples in Chinese history
of what might be called an “ideology of ling,” motivating, or at least provid-
ing the pretext for, resistance and rebellion against state authority. Beginning
with the Yellow Turbans in the third century to the White Lotus, Taiping, and
Boxer rebellions in the late imperial period, most peasant uprisings, large
and small, either began as local cults or sectarian movements or appropriated
religious eschatology and symbolism to define and legitimate their actions. In
Taiwan, the Zhu Yigui, Wu Fusheng, and Lin Shuangwen rebellions against
Qing rule and the 1915 Xilai’an uprising against the Japanese all fit the lat-
ter model. Cf. Chen Guofu 1963:330 –342; Esherick 1987:38 –95; Kuhn 1970;
Lipman in Lipman and Harrell 1990:65–81; Murray and Qin 1994; Naquin
1976; Ownby 1995; Shek in Lipman and Harrell 1990:87–114.
29. Ter Haar (1998) makes the case that this refined passivity of the lite-
rati was a deliberate subterfuge. He cites documentary evidence that elite men
did not hesitate to employ violence when it suited their purposes. In debunk-
ing the myth, however, ter Haar actually underscores the contrast between
literati wen and commoner wu masculine ideals.
30. See, for example, the entry for wu in the Zhengzhong xingyinyi zonghe
zidian (Gao Shufan 1984).
31. Confucianism was not the totality of premodern China’s intellectual
universe, nor can we assume that all members of the knowledge class were
philosophically inclined thinkers. My point here is rather that the intellectual
elite maintained their political leverage and social privilege in part through
their domination of cultural production. The most important residue of Con-
fucian ideology in the modern era is not the influence of Confucian ideas per
se, but rather the intellectual elite’s continued sense of entitlement, which in
mainland China has survived the denigration of Confucianism and the crises
of the Mao years and adapted to the completely new conditions of technoc-
racy and marketization.
32. There were other routes to gaining office throughout imperial his-
tory, including appointment by the emperor. The most common way to bypass
the examination was to purchase an official degree, sometimes through brib-
ery, but at times as a legitimate practice sanctioned as state policy. This route,
however, implicitly acknowledged the examination system even as it under-
mined and corrupted it.
33. As participation increased, success rates dropped. According to
Elman, an examinee’s chances of reaching and passing the triennial metro-
politan examination (the prerequisite degree for an imperial appointment)
in 1850 were about one in six thousand (1991:14).
34. Caizi narratives played out the more material and romantic desires
of elite men, rather than the lofty principles of Confucian sagehood. While
the Confucian gentleman ( junzi) remained a model for ethical behavior, only
a few men in each generation could expect to join the “sages and worthies”
enshrined in the official Confucian temple; the fengliu caizi , on the other
hand, was deeply embedded in the social and personal realities of the lite-
rati lifeworld. Along the same lines, the haohan was neither the exclusive nor
the most exalted representation of wu masculinity (that honor goes to such
“great” military men of Chinese mythohistory as Guan Yu and Zhang Fei) but
one that epitomized familiar and attainable qualities.
35. Outside of novels and dramas, no one calls another caizi or haohan
except humorously or sarcastically.
36. It is worth pointing out, without venturing an explanation, that caizi
were often described as having soft, feminine features rather than “rugged”
good looks.
37. Hypergamy was not universal, however. Uxorilocal marriage (resi-
dence in the wife’s household) in China is most commonly a way for families
with daughters but no sons to ensure the continuity of the patriline. It usually
involves lower-status men who agree to assign at least one male child to the
wife’s father’s patrilineage. In extreme cases, a man gives up his own patri-
line completely and takes on his father-in-law’s surname, effectively becoming
both son-in-law and adopted son. Hartwell (1982:419–420) notes an unex-
pectedly high frequency of uxorilocal marriage in elite gentry families and

documents cases where newly wealthy but low-status families married out sons
uxorilocally in order to gain elite status and privilege—including access to the
civil service examination cycle.
38. The implication is that the rebels are the rightful sons of the emperor
and are acting on his and Heaven’s behalf. The pretext of the novel is that
the 108 heroes are earthly manifestations of the Thirty-six Big Dipper Gener-
als and Seventy-two Baleful Earth Generals—that is, sent by Heaven to rescue
the empire (and emperor) from the clutches of corrupt officials, the worst of
whom are incarnated demons.
39. Whereas Confucius identified physical culture and martial arts as
essential to the cultivation of aristocratic virtue, the martial arts as they devel-
oped in the later empire were primarily associated with the professional mili-
tary or with religious orders (such as the famous Shaolin “fighting monks”),
groups that were, if not completely marginal, at least seen as ignorant of or
threats to the civilized imperatives of the literati class.
40. See in particular Brandes 1980; Driessen 1983; Gilmore 1987, 1990;
Gutmann 1996; Heald 1999; Herzfeld 1985; Lancaster 1994. Gilmore (1990)
includes a brief discussion of Chinese society, but it is oversimplified and mis-
informed. Among recent studies of Chinese masculinity and male identity, a
few, particularly works by Sangren and Zito (whom I have cited extensively
throughout this chapter) have yielded persuasive theoretical insights but are
not primarily ethnographic studies. Though not quite a trend yet, there have
been some ethnographic studies that focus on masculinity (mainly in the PRC);
see, for example, Chen Hongtu 2002 and Brownell and Wasserstrom 2002.
41. Cf. Honig and Hershatter 1988; M. Wolf 1972, 1985; Schein 2000;
Yang 1999.
42. Louie argues that Bruce Lee and Chow Yun-fat created cool, super-
masculine personae by invoking specifically wu-masculine moods and ges-
tures. Xueping Zhong (2000:46), however, contends that Hong Kong film
heroes have been little more than Chinese versions of Sylvester Stallone and
Takakura Ken. Chow’s global appeal, from this angle, is less an adaptation of
Chinese wu than a mastery of Hollywood cool.
43. I am speaking here of the “traditional” martial arts rather than the
crowd-pleasing mélange of martial arts, acrobatics, gymnastics, and ballet
known as wushu.
44. The Chinese martial arts were most likely not created, contrary to
myth, as a way to improve the endurance of Buddhist monks so they could
handle the rigors of meditation. However, as Meir Shahar (2001) has demon-
strated, there was a firmly established tradition of fighting arts in the Chinese
Buddhist monastic community that certainly played a role in the evolution of
the martial arts as we know them today.
45. See, for instance, Lagerwey 1987 (esp. 216–237); Saso 1978 (esp. chs.
4 and 6); and Strickmann and Faure 2002.
46. In Taiwan, anti-hooligan laws formerly allowed the authorities to
arrest and hold adults without formal charges or trial for up to three years.
Many of my acquaintances spent time in “custodial reform camps” for very
minor infractions. These laws were only recently amended to curb the obvious
violation of international human rights standards. In the PRC, the laws (and
legal enforcement) still await revision.
47. This may actually come closer to an earlier usage of the term xiahai
(ca. 1900), which referred to amateur entertainers who decided to turn pro-
fessional (note from anonymous reviewer). Given the symbolic power of water
to cleanse and purify, logic suggests that “leaping into the sea” would refer
mainly to the risk of adventure. It is curious, then, that the term seems to have
been used almost exclusively to indicate the decision to embark on a morally
ambiguous path.
Chapter 3 Taidong
1. Ideally such inscriptions are carved into a marble plaque with gold leaf
lettering and mounted in a frame in front of the temple; more commonly,
however, they are simply written in black paint on a red background on an
available internal temple wall.
2. While the nationalistic rhetoric of “sacred” territorial integrity is a
recent creation, the concept of an organically integrated cosmic order under-
lying the political and social order is an obvious and deeply rooted legacy.
3. G. William Skinner’s analysis of marketing systems in late imperial
China demonstrated the structural correlation as well as functional integra-
tion of core and periphery (Skinner 1974). Sangren, expanding on Skinner,
examined the ritual organization of a local marketing system in northern Tai-
wan. He found first that the local temple network and festival cycle subsumed
the temples, neighborhoods, and other associative groups within a hierarchy
of place (paralleling the administrative ranking of the celestial bureaucracy).
But following the logic of ling (as the power of mediating between lower-order
and encompassing levels), he noted that collectivities can also assert, in sym-
bolic terms, a separate territorial and social integrity: “Because ling creates
order by differentiating outsiders from insiders (yin from yang), community
at any particular level can be asserted only by ignoring more encompassing
levels (at least in cult rituals)” (1987:219).
4. This vision of a resonant relationship between microcosm and macro-
cosm is one of the most influential and enduring innovations of pre-Buddhist
Chinese philosophy and has come to form the structuring logic and explana-
tory idiom for a diverse array of discourses and practices, from Chinese medi-
cine (Sivin 1995) and divination to Daoist ritual (Lagerwey 1987; Schipper
1993) and imperial sacrifice (Zito 1997).
5. Fengshui, literally “wind and water,” is a group of related practices
devoted to measuring and manipulating the circulation and concentration
of qi through the natural and man-made environment. Fengshui thus shares
a theoretical foundation with traditional Chinese medicine and the martial
arts, which work to measure and manipulate qi within the body. Until quite
recently, fengshui was used almost exclusively to locate and orient auspicious
sites for “yin dwellings” (yinzhai, graves, where bodies of ancestors “reside”)
and “yang dwellings” (yangzhai, houses of the living). There are two distinct
methods or “schools” of fengshui: the xingshi, or “landforms,” approach rec-
ognizes and interprets patterns and relations in natural features of the land-
scape (hills, hollows, streams, ponds, and so on); the liqi, or “calculating qi,”
approach, measures qi using the geomantic compass (luopan) and computes
the complex relationship among spatial and temporal coordinates.
6. Several intriguing descriptions in pre-Ming Chinese texts may refer to
Taiwan, to the Ryukyus, to both, or to neither. The most frequently cited pas-
sages are found in Wang Dayuan’s Dao yi zhi lue and Zhao Rugua’s Zhu fan zhi.
Hsu (1980: esp. 3–16) examines these as well as a number of less well-known
texts.
7. Immigration from the mainland effectively ended at the start of the
Japanese occupation. Given natural rates of increase, the 1905 estimate of just
over 3 million suggests that Taiwan’s population at the end of Qing rule was
approximately 2.7 million.
8. Aborigines, both “cooked” (plains, assimilated; shufan) and “raw”
(mountain; shengfan) were probably undercounted in the 1905 census. Even
after compensation for such errors, however, the aborigine population could
still not have been much more than 3 percent of the total ( Japan, Provi-
sional Bureau of Census Investigation of Formosa 1909), even assuming (as
Melissa Brown [2004:140–149] argues) that the rate of intermarriage between
Han and plains aborigines during the seventeenth and eighteenth centuries
was considerably higher than other scholars (notably Shepherd 1993) have
acknowledged.
9. Portuguese and Spanish mariners noted the existence and location
of Taiwan (apparently without venturing a landing) in the mid-sixteenth cen-
tury, and Taiwan appeared on European maps from 1554 on (beginning with
the world map of Portuguese cartographer Lopo Homem). However, the east
coast and mountain regions remained largely uncharted until the nineteenth
century.
10. While the forbidding terrain and “raw” aborigines’ penchant for tak-
ing heads as trophies dissuaded most Han from venturing into the prohibited
zone, both Chinese and aborigines on both sides had reason to cross the line.
Officials, charged with controlling an unruly and dispersed population, even-
tually decided to fence off the wilderness. Beginning in 1722, a network of
trenches and berms called “earth oxen” (tu niu) was constructed to keep Han
adventurers out of aborigine territory and to discourage the “raw” aborigines
from raiding lowland settlements. The official interdiction lasted until 1875.
A comprehensive account of these policies and their repercussions is found in
Shepherd (1993: esp. 178 –307).
11. Houshan is often translated as “beyond the mountains,” which is both
historically and grammatically inaccurate. The hou in houshan modifies shan:
hou means “beyond” or “after”; shan is “mountains” or “mountainous region.”
Houshan is, then, a category of mountains. Conversely, compound terms in
which shan precedes a locational adjective — e.g., shan’nei, “the mountain inte-
rior”; shanjiao, “mountain base (lit. mountain foot); and shanding, “mountain
heights” or “mountaintops”— indicate places on, in, or of mountains. Flexible
as classical Chinese grammar can be, this difference in word order is not triv-
ial. Houshan is a conventional late-imperial reference to mountain peripheries

throughout the empire, not only Taiwan.
12. One early Qing account describes the southern mountains as full of
hobgoblins and demons, sprites, and apparitions. They are home to “animal-
faced” people with beak-like mouths who lived in “nests and hollows.” Such
tropes of the freakish and exotic date back to the Han dynasty text Shanhai-
jing, the Classic of Mountains and Seas. The use of such clichéd formulas helped
maintain the cultural order discursively, while relieving the scholar-official
authors from the unwelcome prospect of real frontier duty.
13. Until 1874, the empire’s boundary had been estimated as a zone that
marked the vanishing horizon of Chinese cultural influence rather than a
precisely charted and posted line. Even that most famous of Chinese bound-
ary markers, the Great Wall, marked the general area of the periphery (which
sometimes extended well beyond the gates), not an exact dividing line. This
logic is graphically evident in court-sponsored maps drafted before 1875 and
contrasts starkly with the premises of contemporary European maps, which
often aimed to reify otherwise tenuous territorial claims.
14. The Longitudinal Valley is sometimes called the Eastern Taiwan
Rift Valley. It is not, in fact, a rift (where two tectonic plates are moving away
from one another) but rather the surface expression of the collision between
between the Philippine Sea Plate and the Eurasian Plate. The valley region
includes at least two major north-south faults, the Central Range Fault and
the Longitudinal Valley Fault, which yield frequent moderate and strong
earthquakes and occasional major-magnitude events (greater than m7.0),
including two m7.3 quakes in 1951. The shoreline and Coast Range rest on
the northwestern “lip” of the Philippine plate. The Central Mountains are
part of the Eurasian plate. Further to the north, the Philippine plate subducts
northward beneath the Ryukyuan Arc. The complex interaction of multiple
tectonic plates makes Taiwan one of the most seismically active places in the
world (see Chou et al. 2006; Wu et al. 2006).
15. Regional stereotypes have been part of Chinese folk ethnology for
centuries. While less developed than for provincial and north-south divisions
in mainland China, stereotyping according to subregion in Taiwan is a com-
mon topic in casual conversation.
16. While the Dutch never found their gold mountain, the middle and
lower reaches of the Beinan Stream, which flows through the Eastern Longi-
tudinal Valley, as well as a few of its tributaries, have yielded small quantities of
gold in the form of dust, flakes, and nuggets (Fang and Yu 1998:69–70).
17. The Southward March policy (nanjin zhengce; J. nanshin seisaku) was
promulgated in 1936 as part of a broader policy to “imperialize” (huangmin-
hua; J. kominka)—that is, fully Japanify—Taiwan. The plan called for establish-
ing new Japanese agricultural settlements in central and southwestern Taiwan
and restoring previously attempted (but failed) settlements in the east coast
region. The purpose, as with earlier colonial immigration projects, was pri-
marily to relieve overpopulation on the Japanese main islands. This was the
last and most extensive of several intermittent Japanese efforts to colonize
226 notes to page 67
parts of Taiwan dating to 1906, when plans were first drafted for a number of
settlements in the Taidong area.
18. Stone flake tools excavated at Eight Immortals Cave (Baxian Dong)
near Changbin, about fifty miles north of Taidong City, have been dated to
the late Paleolithic, about 30,000 BP, and are the earliest evidence of a human
presence in Taiwan. More recently—from roughly 5500 BP through the
beginning of the Common Era—a succession of Neolithic peoples occupied
and abandoned the area, leaving behind a dramatic material record. At sites
scattered across the Beinan Plain, the southern Rift Valley, and the narrow
strip along the coast, archeologists have turned up thousands of stone, pot-
tery, and jade artifacts. Most spectacular, however, are the standing megaliths,
stone cellars, and sarcophagi of the Beinan Culture, which reached a level of
aesthetic and technological sophistication comparable to contemporary Neo-
lithic groups in northern China and Southeast Asia (cf. Chang 1969; Pearson
and Underhill 1987).
19. The microclimates of the lower and mid-level mountain elevations in
northern and central Taiwan are ideal for tea cultivation. The camphor tree,
Cinnamomum camphora, found mostly in the middle elevations throughout the
Central Mountains, is native to Taiwan.
20. In 1871 a ship (accounts differ on whether the vessel was a cargo or
fishing craft) from Miyako Island, in the western Ryukyu Archipelago, about
halfway between Okinawa and Taiwan, wrecked on the reefs near modern
Manzhou. After coming ashore, fifty-four of the sixty-six surviving crew mem-
bers were massacred by local Mudan (Paiwan) aborigines. The new Japanese
leadership had already conceived an interest in Taiwan as a site of colonial
expansion and was anxious to test the island’s defenses as well as the inten-
tions of the Qing court. Japan also sought to countermand China’s claims
to the Ryukyus, which had been based on the Ryukyuan Kingdom’s tribu-
tary relations with China since mid-Ming. A request for reparations on behalf
of the murdered fishermen was rebuffed because, as the official in charge
explained, “the raw barbarians [the Paiwan] are beyond the pale of civiliza-
tion” (shengfan zhi zhiyu huawai)—that is, outside Qing jurisdiction. The Japa-
nese responded by forcing the issue militarily. In the spring of 1874 a Japa-
nese punitive expedition of 3,600 troops landed and encamped near modern
Fangliao. Despite their tactical success in the ensuing campaign against the
Paiwan, the invasion force was severely depleted by disease: almost six hun-
dred Japanese men and officers had succumbed to malaria and encephalitis
by the end of 1874. The demoralized expeditionary force was ordered home
early in 1875 but not before the Japanese had wrought important concessions
from the Chinese: the Qing court was forced to recognize the Japanese claim
to the Ryukyus and pay an indemnity to the victims of the 1871 massacre. The
impunity with which the Japanese had landed and operated throughout the
campaign shocked the imperial authorities and impelled them to take a num-
ber of steps, including the building of battlements at Hengchun (cf. Wang
Yuanzhi 1959). Even more significant, the Qing court was finally compelled to
abandon its traditional paradigm of “cultural empire.” Thenceforth all official
maps represented the mountains and east coast as Chinese-administered ter-
ritory. Ironically, the Qing claim on eastern Taiwan lasted only twenty years,
until the signing of the Treaty of Shimoneseki, which ceded Taiwan to Japan
“in perpetuity.”
21. Perhaps the most far-fetched of these attempts, and one most cer-
tainly doomed to fail, involved shipping in approximately two thousand Hakka
contract laborers from Chaozhou, in Guangdong Province (Hu 1952:43). Hu
suggests that since “these laborers were, it is said, mostly wastrels and wander-
ers, they did not take to cultivating the land for a living” (ibid.). On the other
hand, many “Cantonese” (i.e., Hakka-speaking) colonists settled in the area
quite early—for instance, George Taylor (quoted in Liu 1992:95–96), claims
to have run across a village of eleven Hakka households, on land “rented”
from a local Paiwan tribe, probably between modern Shangwu and Dawu.
Since Hakka settlers in the Longitudinal Valley mostly migrated south from
Yilan and those along the southern coast probably came across from the Mei-
nong area, it is unlikely that the Hakka presence in Taidong owes much to this
earlier “seeding” of Hakka laborers from Guangdong.
22. Much of what is known about the Chinese settlements around Tai
dong before 1880 comes from brief descriptions by English lighthouse keeper
George Taylor and Presbyterian minister Hugh Ritchie (Liu 1992). Hu
Chuan’s Handbook (1952) documents conditions in the mid-1890s. It is now
generally accepted that splinter groups of sinified Pingpu aborigines, who
moved into Hualian from the western plains in the late eighteenth century,
were the first to introduce wet rice agriculture as well as other Chinese life-
ways to the area. Some of these Pingpu aborigines (Siraya and Kavalan tribes)
later moved south and settled in the central part of the Longitudinal Valley,
near modern-day Yuli, and on the seacoast at Fengbin. Nevertheless, it is not
impossible that small numbers of Han Chinese sojourned and even settled in
the area as early as 1800, trading with and possibly living in aborigine commu-
nities along the coast. One intriguing but uncorroborated source describes
the ordeal of a Japanese officer named Honsuke, who was shipwrecked along
the southeast coast at a place called Dagangkou (big harbor) in 1803 along
with nine other sailors. As the superior, he was sheltered in the house of a
“Cantonese [Hakka?] merchant.” Apparently the marginally better food and
living conditions in the Chinese house were enough to keep him alive, while
his shipmates, housed in aborigine dwellings, all died of disease.
23. “[In the twelfth month of the first year of the Guangxu reign period]
Feng’gang villager Lin Zan recruited sixty peasants and undertook the devel-
opment of the uncultivated land at Balangyu [near modern Dawu], in [the
district of Beinan]; sinified aborigine Pan Qinyüan recruited sixty peasants
and undertook the development of the uncultivated land east of Dapotou. . . .
Baosang villager Chen Yunqing recruited fifty peasants [and?] assimilated
aborigines and undertook the development of the uncultivated area at Lijiliji”
(Hu 1952:37a–b).
24. The Beinan Commandery headquarters was in the area of what is
now downtown Taidong City. The main military outpost was in the Longitudi-
nal Valley to the north, at Shuiwei (now Ruisui, in Hualian County). During
the Dazhuang Uprising, both were attacked and destroyed by the rebels. All
imperial personnel were moved to Beinan after the military reasserted control
(Hu 1952:1). Taidong was subsequently elevated from a commandery to an
autonomous prefecture (zhilizhou) .
25. Hu Chuan provides a vivid description of the conditions at the
Beinan garrison just after his arrival in Taidong. Just how bad morale and dis-
cipline actually were at the time is evident from his account: in target practice
the Beinan station troupes hit barely 30 percent (1952:9a), and malaria and
encephalitis were rife. Within a month of his arrival two of Hu’s chief officers
had died of tropical fever. Hu also found that of the troops under his direct
command and in his jurisdiction, only one hundred or so were not addicted
to opium (1952:10b). To put the troops in order, Hu personally encouraged
them to give up the habit and claimed there were only thirty that were unable
or unwilling to try. He then requested a naval transport to ship out all those
who were still addicted (13a). Perhaps doubting the practicality of the exer-
cise, he later revised his order to allow those whose addiction was “weak” to
take medicine identified only as jieyanwan, or “kick-the-habit pills” (19b),
while those whose addiction was “profound” would be discharged (13b). He
does not record the final success rate; in any event, Hu, the first, last, and only
commandant of the Taidong Autonomous Prefecture left his post in 1895 as
part of the general evacuation of all Qing imperial government personnel.
Ironically, having survived the harsh conditions of Taidong, he fell ill and died
shortly after his return to the mainland.
26. The Japanese-built Taidong sugar refinery was in continuous opera-
tion until 1996. Davidson (1903) confirms that the cultivation and processing
of sugarcane in the Beinan Plain was already in full swing less than ten years
after the beginning of Japanese occupation. Taidong University geographer
Li Yufen (personal communication) has gathered cartographic and other evi-
dence that suggests that most of the lower Beinan Plain, formed by the unpre-
dictably shifting alluvial deltas of three major streams, was at best marginally
cultivable until the Japanese saw the area’s potential for cane cultivation and
built a series of flood control and drainage works.
27. Until the early 1930s, the sea route remained Taidong’s fastest and
most reliable connection to the outside world. Even so, because of the lack of
a natural harbor, even small seagoing vessels anchored offshore, while passen-
gers and cargo were transferred to and from shore by sampan.
28. This narrow-gauge line served the east coast as an independent
branch until mid-1982, when work on a standard-gauge replacement line was
completed. The route was then linked, via the Su’ao-Hualian section, to the
rest of the island’s rail network. The Southern Link line, connecting Taidong
with Pingdong, opened in 1992, making it possible for the first time to make
a complete circuit of the island by train.
29. Despite Taidong’s isolation and relative obscurity, the airfield, mod-
est harbor facilities, and sugar refinery were noted on U.S. reconnaissance
maps and became sporadic targets of Allied bombing prior to the Philippine
and Okinawa campaigns in World War II.
30. The Japanese initially considered the southeast as a prime site for
resettlement from the Japanese islands. Most attempts to settle Japanese farm-
ers in Taidong ended in failure, primarily due to the difficult climate and
endemic diseases. Two small village projects, those at Xucun and Luye, did
finally succeed, but the Japanese population of Taitō Prefecture, including
settlers, officials, and military, only barely exceeded five thousand, or about 12
percent of the population in 1944.
31. This was a second wave of migration in two senses. It was the second
wave of migration into the Taidong area, the first being those few peddlers
and peasants who had moved in at the end of the nineteenth century under
the Qing administration. In a more significant sense, however, this was a sec-
ond wave vis-à-vis the new migrants themselves, who were mostly descendents
(some only of the second or third generation) of the immigrants from Fujian
and Guangdong who had settled and populated western Taiwan over the
course of the eighteenth and nineteenth centuries.
32. A pattern of Chinese emigration to the nanyang (the “south seas”)
dates to at least the thirteenth century, and by the early nineteenth century
sojourning overseas had become an established economic option in the vil-
lages of coastal Fujian and Guangdong. This tradition continues today, as
manual laborers and sex workers from Fujian have become the largest undoc-
umented group in Taiwan, with northern and southern Fujian about equally
represented in statistics compiled between 1991 and 1999.
33. The literary term guxiang can mean “ancestral home” but more com-
monly alludes to the place one’s parents reside, where one was born and
raised. Zuji and jiguan, on the other hand, are formal literary and techni-
cal terms respectively that expressly indicate the place of one’s patrilineal
ancestors, whether or not one has ever actually lived there. The notion that
one belongs to one’s ancestral home (even if that is not the place of birth or
residence) is prominent in Confucian and Confucian-derived models of the
patrilineal family and conspicuous in a variety of practices and representa-
tions. It was standard bureaucratic practice in Republican China and until
quite recently in Taiwan, for example, to list a person’s ancestral origin as well
as place of residence on all official documents. I would argue, however, that
in cases where residence is three or fewer generations deep, most Chinese
would consider their “hometown” to be the place they were born and raised,
the home of their parents and grandparents, rather than a place (if different)
associated with more distant ancestors.
34. The last wave of migration to Taidong occurred following the disas-
trous floods of August 7, 1959 (triggered by Typhoon Ellen), which washed
away whole villages and destroyed thousands of acres of farmland in west-cen-
tral Taiwan.
35. There are no reliable statistics on fertility and morbidity for the area
during this period, but it is quite likely that the aborigine population grew
almost entirely by natural increase (and even then at a very low rate due to
Japanese relocation policies and to lack of access to health care; an increase
in alcohol addiction; and the forced drafting of aborigine men into hard-
labor crews and, somewhat later, the Japanese Army). The Han population,
on the other hand, was persistently augmented by new immigration, as well

as a higher rate of natural increase due in part to lower infant mortality and
better access to health care.
36. The three census reports cited here, by a Qing dynasty civil official
(1894), the Japanese colonial administration (1944), and the Taidong County
Government (1994) respectively, use somewhat different terms of ethnic
classification. For example, the 1894 census classifies all Han Chinese as min
(i.e., full-fledged common subjects of the emperor), whereas the Japanese
report further distinguishes between ethnic Han born in Taiwan and those
recently arrived or still claiming Chinese citizenship. Population surveys and
household registration records after 1945 distinguish categories of aborigines
(“mountain-dwelling” and “plains-dwelling”) but do not specifically identify
members of non-minority groups. However, given this caveat, the categories
are roughly comparable, and there is sufficient basis at least to demonstrate
such general, long-term trends.
37. Figures are from Taidong County Government Web sites. See http://
www.taitung.gov.tw/tw/CP/1289/history04.aspx and http://www.taitung.gov
.tw/statistics/.
38. In fact for most aborigine communities in Taidong, the standard of
living is measurably below the county average and well below the measures for
the non-aborigine population. Nearly 2 percent of households, mostly aborig-
ine, live below the official national poverty line. On the other hand, per capita
income in Taidong for all groups is well below the all-Taiwan average. As of
1995, nearly 40 percent of the eligible working population derived most of its
income from agriculture, fishing, hunting, and/or animal husbandry. Of the
remaining 60 percent, a significant number were in fact living and working
elsewhere, although still officially resident in the county (1995 census, Tai
dong County Government archives, at http://www.taitung.gov.tw/statistics/).
39. Baosang is pronounced po-song in Holo, from Amis posog, “place of
small hillocks.”
40. The one clear exception is the Heavenly Official Hall, which is prob-
ably the oldest extant neighborhood-based cult in Taidong. The temple’s core
members are all longtime residents of the surrounding neighborhood. The
weak territorial identification is partly a matter of urban-rural difference, as in
many of the villages and towns surrounding Taidong (some of which, like the
Beinan community noted above, have since been administratively incorpo-
rated into Taidong City but still maintain a separate local identity) the god’s
authority is described as extending precisely to the community’s territorial
limit. There are, of course, a number of Earth God shrines scattered through-
out Taidong, but here they do not generally function as rallying points for
local factions or operate as local social centers.
41. For example, the ritual of “pitching the outer five camps” (an wai
wuying), common in the local territorial cults of central and southern Tai-
wan, is rarely performed in Taidong. In “pitching the outer five camps,” a god
(represented by his possessed medium) travels the periphery of his domain,
planting bamboo stakes—setting up garrisons of spirit soldiers—at particu-
larly vulnerable points.
42. As with every other occupation and profession, most Taidong gang-
sters with career ambitions look elsewhere. That is not to say, however, that
Taidong’s is a kinder, gentler underworld. On the other hand, the smaller
scale of operations and the fluidity of alliances does create a less outwardly
belligerent atmosphere. Egos clash but rarely with guns, knives, or clubs. A
meeting of local underworld figures that I attended in 1991, for example,
was more reminiscent of a fraternity get-together than a parley between rival
crime bosses.
43. Apparently no one bothered to consult a native speaker of Amis:
naruwán has no semantic meaning. It is a syllable string used in Amis song
phrasing indicating a lighthearted, cheerful mood.
44. The main targets of the 2004 promotion were Japanese travelers, for
decades Taiwan’s most dependable source of tourist revenue. Indeed, with
an eye both to developing a domestic tourism base and increasing Taiwan’s
appeal for Japanese visitors, marketers and boosters have tried to emulate the
“Discover Japan” and “Exotic Japan” campaigns of the 1970s and ’80s, which
similarly packaged and promoted cultural authenticity (see Ivy 1995:29–65).
The Taidong development plan, for example, suggested that Taidong should
reference Japan’s experience in developing hot springs resorts.
Chapter 4 Fire and Fury
1. Taidong appears in the historical record no earlier than the fragmen-
tary seventeenth-century Dutch accounts; until the late nineteenth century
it remained an obscure and rarely mentioned backwater. Dali, on the other
hand, is well known from Chinese and local records extending back to the
Tang dynasty. It was, if anything, a more central place, even from the sinocen-
tric perspective of imperial history, center of the Nanzhao Kingdom (fl. ca.
650–937 CE) and the even more powerful Dali Kingdom from 937 until the
Mongol invasion of 1253, after which Dali was gradually incorporated into
the Chinese imperium. Quite a bit of this deep local history finds expression
in Bai mythology and religious practice. For example, the chief deity of the
General’s Grotto Temple ( Jiangjun Dong) above Xiaguan is General Li Mi, a
high-ranking general sent by the Tang court who died attempting to escape
the massacre of his army by local forces in 754 CE. The Bai, especially local
elites, often point to this long history of interaction with Han Chinese civiliza-
tion as something that sets them apart from other minorities: they are more
sinified (hanhua), they say, implying that they are more “civilized” than other
minorities, adopting the cultural hierarchy that puts Han Chinese civilization
at the top and center. This tendency to understate their distance from the
Han and emphasize difference from other ethnic minorities is un-ironically
combined with the representation of Bai culture as colorful and “ethnic” and
of mountain-walled Dali as remote, pristine, and scenic. The promotion of
Dali as a tourist destination, particularly targeting Chinese tourists, is a colla-
tion of these seemingly contradictory themes. In this sense, an account of the
history and geography of Dali is no less, and in many ways more, deserving of
detailed description than that of Taidong. I take the liberty of using compari-
son with Taidong as a descriptive shortcut, in part for efficiency’s sake.
2. The army Genghis Khan sent to attack Yunnan in 1254 was mostly
made up of Muslim conscripts and mercenaries from Mongol-occupied parts
of Central Asia. These troops are considered to be the founding ancestors
of Yunnan’s Hui community. Significant Han migration, on the other hand,
dates only from the mid-Ming.
3. Used mostly for local trade (and smuggling) in tea, horses, and opium
during the late imperial and Republican eras, these roads, widened, graded,
and consolidated, gained new strategic importance (albeit briefly) as the so-
called Burma Road during World War II.
4. There is disagreement among linguists on the origins and affiliation of
the Bai language. Some contend that it is a Tibeto-Burman language (cf. Mati-
soff 2001). The other prevalent theory argues that Bai is a branch of Sinitic,
albeit one not directly related to the modern Chinese dialects. In either case,
all acknowledge that Bai has been strongly influenced by Chinese, including
heavy borrowing of vocabulary.
5. Much of the evidence for this assertion points to the adoption of Chi-
nese cultural elements by local elites—the writing system foremost, but also
symbolic and religious forms. It is more difficult to pinpoint or measure early
Chinese cultural influence among peasants and commoners, but there is no
question that Han culture is very much embedded in the Bai’s kinship system,
religious practices, and language.
6. Folklorists, both Han and Bai, have popularized the Chinese term ben-
zhu (lit. lord of the locality), but there is no generic category for these territo-
rial deities in the Bai language.
7. Xizhou is the market town called “West Town” by Francis L. K. Hsü in
his now classic ethnography Under the Ancestors’ Shadow (1949).
8. In some cases, the young men who wait or climb the torch are proxies
for a married but childless male friend or agnate, to whom they give the spoils
of the contest if successful. In some villages it is the successful grabber who
provides the feast, in others it is the recipient of the spear, in anticipation of
his coming good fortune.
9. Unlike Blasting Handan Ye, where the blasting wounds bleed, or the
more common slashing and spiking by entranced spirit mediums, the blood
that appears during the Torch Festival is neither real nor the focus of the spec-
tacle. It is, however, apparent, as well as unambiguously a symbol of female
reproductive power: the red-dyed fingernails of the young women.
10. The first Lantern Festival procession in Taidong took place no earlier
than 1920 and until the 1950s was a fairly small-scale observance. The proces-
sion grew considerably in size, importance, and intensity in the early postwar
years as migrants from southern and central Taiwan settled in and around the
town and founded dozens of new temple cults.
11. One dip for most households; three dips for temples and the altars
of important temple members, politicians, and other influential members of
the community.
12. I am aware of reports in Taipei Japanese-language newspapers (ca.
mid-1930s, dates uncertain) describing the Handan Ye ritual as practiced in
Taipei’s Dadaocheng district (Mr. Cai Wu-huang, personal communication).
A similar Lantern Festival custom, called “running Buddha” (zou fo) was appar-
ently common in some areas during the same period. Such practices were
probably known to some of the early Taidong practitioners, though possibly
only as hearsay or hazy rumors. It seems clear, however, that the cult and its
ritual were not directly transmitted to Taidong but rather recreated in local
form.
13. Information about Handan Ye, including the identity of its current
leader, is readily available. In print, in documentary film and television pro-
grams, and on the Web site of the Taidong County Government, Mr. Li’s name
and role are directly and publicly associated with Handan Ye—indeed with his
consent and encouragement. Mr. Li has given his consent to use his name in
this book, and he is fully aware of the context and content of my descriptions,
citations, and quotations. Mr. Li is therefore the prominent exception to my
use of pseudonyms for individuals appearing in this text.
14. The earliest version of the myth of Hou Y i (Archer Y i) shooting
down nine suns appears in the second-century BCE text the Huainanzi (Allan
1981:302).
15. Another “explanation” of the Blasting Handan Ye ritual used to cir-
culate among longtime Taidong residents: local gangsters spend their time
exploiting the people, but they know it’s wrong and feel guilty. Offering them-
selves for “sacrifice” once a year allows them to expiate this guilt. There is, I
think, something revealing, if not quite accurate, in this story. As the embodi-
ment of the common desire for easy gain, Handan Ye represents an embar-
rassing social and psychological fact. The story is a way for its tellers to dis-
tance themselves from their own secret guilt, objectifying and projecting their
own illicit desire onto the bodies of the sacrificed gangsters. As a collective
process, however, we can see a symbolic, largely unconscious, accommodation
between the protagonists and the “good citizens” who depict these “flesh body
Handan Ye” as lacking ethics but yet morally conscious (hence their felt need
for expiation). Li Jianzhi’s rhetoric of redemption can be understood as an
attempt to repossess the morality of Handan Ye; yet in subscribing to and iden-
tifying Handan Ye with such values, Li is participating in the production of a
mainstream moral ethos that recognizes the meaning and power of Handan
Ye only in denial and derogation.
16. Note that they would probably not invoke the assistance of their own
territorial deity, whose jurisdiction did not apply and who, like their fellow
villagers generally, would never in the light of day publicly commend such an
act, even if secretly delighted.
17. See Yan (1996:167) for the distinction between lian as “moral face”
and mianzi as “social face.”
18. The term yingshen saihui, while usually glossed simply as “temple
festival,” literally means a “competitive gathering for welcoming the gods.”
Temples compete for reputation, but the festivals are understood also as con-
tests among the gods themselves: the performance of a temple’s troupe is a
measure of its patron god’s magical power.
19. There are, of course, female performers, but they are mostly passive.
One recent addition to the Taidong Lantern Festival procession, for instance,
is an enormous float featuring dozens of dolled up xian’nü, the fairy atten-

dants of the goddess Mazu who wave their brightly colored fans seductively
and toss handfuls of candy into the crowd. Elsewhere in Taiwan one finds
ribald, humorous performance troupes made up of female impersonators,
but these are not among Taidong’s local customs.
Chapter 5 Tales from the Jianghu
1. Devotees of other temples do not begrudge bestowing this honor on
the goddess, but reflecting the factional (and fluid) nature of community
politics, some measure of resentment follows the attention garnered by the
Tianhou Gong and its temple management committee.
2. “Way,” dao (tao), is a philosophical/technical term of ancient and broad
usage that defies easy translation or precise definition. In simple terms, it indi-
cates a body of esoteric knowledge and techniques shared among a group of
initiates (or members of a fictive lineage). It can refer to any profession or
vocation and, by extension, the collectivity of initiates who share and practice
that knowledge.
3. The 1949 Communist revolution seems to have effectively put an end
to the “feudal” practice of tattooing. It is possible, of course, that it may soon
reappear not as a jianghu custom but as a fashion statement among trend-set-
ting, globalized, urban middle-class youth. The provenance of jianghu-related
tattoos has thus been limited to Chinese-speaking areas outside the PRC for
most of the last several decades, notably Taiwan and Hong Kong.
4. To some extent a matter of personal aesthetics (unlike in Japan, where
the designs usually indicate a specific yakuza clan affiliation), Taiwan’s gang-
sters sport a mélange of Chinese and Japanese designs. Japanese full-body
tattoos frequently extend to the wrists and ankles, whereas tattoos covering
forearm and calf are much rarer in Taiwan; Taiwanese-style tattoos usually
reach only to the elbows and knees. One practical reason for the difference is
climate: tattoos are usually concealed in public, especially given the possibility
of arrest under Taiwan’s anti-hooligan laws, and covering the whole body in
Taiwan’s sweltering summer heat would be impractical.
5. Beyond tabloid news and pulp fiction, the Western-language litera-
ture on secret societies can be divided into five somewhat distinct categories:
(1) Journalistic studies of contemporary (post-1980) Asia-based international
crime syndicates (see, for example, Booth 1999, Kaplan and Dubro 1986, and
Lintner 2002); (2) Pre-1960 accounts of initiations and related triad rituals
by British police and colonial administrators in Singapore, Malaya, and Hong
Kong—the most widely cited being Ward and Stirling 1925; a somewhat less
well-known work is Comber 1959; (3) Recent sociological studies of organized
crime, gangs, and corruption in Hong Kong and Taiwan (see especially Chin
2003 and Chu 2000 for contemporary descriptions and historical sketches
of local gangs and organized crime in Taiwan and Hong Kong respectively);
(4) Analyses of (mostly) archival research by anthropologists, beginning with
Maurice Freedman, of secret societies in Singapore, Hong Kong, and other
Chinese communities in Southeast Asia. Many of the early anthropological
studies treat secret societies and gangs primarily as extraordinary instances of
the generic Chinese voluntary association (hui or huiguan) (Freedman 1960;

Skinner 1958; but see also Sangren 1984). In his brief treatment of secret soci-
ety ritual, for example, Freedman reaches a conventional British structuralist
conclusion: the “creation of a sworn brotherhood and the repetition of solemn
oaths served to bind men together in ritually sanctioned solidarity” (1960:38);
(5) Recent historical research on the Heaven and Earth Society (Tiandihui)
and related groups that thrived during the Ming, Qing, and Republican eras
both in China and in overseas communities (Murray and Qin 1994; Ownby
1995, 1996; ter Haar 1998; see also Chesneaux 1971). Elsewhere, while there
are many excellent historical studies of secret society and sectarian-led rebel-
lions and uprisings (in Chinese as well as Western languages—see, for exam-
ple, Perry 1980), few highlight the ritual practices and mythologies of such
movements and are thus of less interest in the current context.
6. See, for instance, Ward and Stirling 1925, Comber 1959, Freedman
1960, and ter Haar 1998.
7. Now nearly all dispersed or slated for demolition, these developments
were built, often poorly and hastily, to house the families of officers and sol-
diers who had fled to Taiwan with Chiang Kai-shek in 1949. They were walled
off from the surrounding communities, making tangible the de facto social
segregation that divided mainlanders and native Taiwanese in the years of
martial law.
8. A fourth major syndicate, the Heaven’s Way Alliance (Tiandao Meng),
follows the mainlander gang pattern, but leaders and members are almost
all native Taiwanese. This is not a surprising development, as the alliance was
formed by a group of local “big brothers” who were incarcerated together in
the Taipei County Jail during Operation Cleansweep (Yiqing Zhuan’an) in
1985. By this time the mainlander gangs had expanded their scope of opera-
tions across the island as well as overseas, suggesting the advantages of their
modus operandi over the more fragmented and localist neighborhood-based
approach.
9. The Bamboo Union maintained a “feudal” model, in which the branch
halls were nominally subordinate to the founding group, for more than a
decade. However, while achieving probably the greatest degree of centraliza-
tion of any underground society in Taiwan, the Bamboo Union never had a
direct chain of command, nor was the core group able to effectively keep the
various subgroups from competing, fighting, or even breaking away.
10. Hong Kong triads are also “neither a centrally structured nor an
unorganized entity, but loose cartels consisting of numerous autonomous
gangs who adopt a similar organizational structure and rituals to bind their
members together” (Chu 2000:39).
11. For obvious reasons, there has been little in the way of participant
observation fieldwork on the religious element in jianghu life and practice
(with the notable exception of DeBernardi 1987). Jianghu characters and
stories, however, have appeared frequently in the ethnography of Taiwanese
and Hong Kong society, prominently so in descriptions of local temple festi-
vals; see, for instance Choi 1995; Jordan 1972:48–59; M. Wolf 1968:45–58).
However, just as studies of imperial-era secret societies rely on official records
236 notes to page 121
of confessions and trials of captured society members, much of the informa-

tion available on contemporary jianghu life in China, Hong Kong, and Tai-
wan comes in the form of police case records or, at best, research interviews
with prisoners and convicts. Both the intentionality behind the writing of such
reports and the bias toward deviance and delinquency on the part of the tar-
get population render this archive at best marginally useful as a source of
information about the “normal” aspects of gang or jianghu life. Cf. Cai 2000
and Liu 2004.
12. Taidong’s jianghu is a somewhat amorphous combination of the latter
two types. In Taidong there are no established turf-based gangs such as one
finds in older communities elsewhere in Taiwan, but some neighborhoods
and villages (e.g., Beinan, discussed in chapter 3 above) have a strong and
persistent local solidarity that has supported turf-based gangs that double as
neighborhood “militias.” The term zuhe (which I gloss as “venture squad”) is
found in Taiwanese police reports and criminological research and is not a
self-description used by the groups themselves.
13. See, for instance, Feuchtwang 1992:108; Lin Meirong et al. 1997; Shi
Wanshou 1983. The historical relationship between self-defense teams and rit-
ual performance troupes seems intuitively obvious, but I would argue that the
ritual and functional forms were practiced in tandem, rather than one being a
derivation of the other. Evidence for this view includes the fact that “symbolic”
militias—i.e., costumed martial ritual troupes—have a long history on the
mainland, particularly in Fujian (Katz 1995:164; Sutton 2003:220–233), and
that martial arts and ritual forms are frequently taught together. For example,
the lion dance, which is the most common and well known of all Chinese mar-
tial ritual procession forms, has long been a trademark of both temple-based
and private training halls throughout the Chinese-speaking world.
14. The term wuguan is difficult to translate, as it encompasses and evokes
a range of meanings that many readers unfamiliar with the Chinese context
might consider categorically distinct. For example, wuguan can mean both a
martial arts association and the hall or space in which it meets and practices.
Moreover, the “martial arts association” referred to here could be one for
training in the fighting arts (as one might expect) or one where ritual perfor-
mance (for example, lion dance) is practiced or even an association of musi-
cians who gather to play certain kinds of traditional music belonging to the
“martial” rather than the “civil” category. In the present case, self-protection
associations used the term wuguan to evoke the meaning of a hall where weap-
ons are stored and where men are trained in their use. However, associations
rarely had their own separate space but met in the village temple or a fam-
ily courtyard. Groups referred to themselves as wuguan, whereas the physical
space (with or without walls and a roof ) where the group habitually gathered
would only be called wuguan when the group was actually present.
15. If a village had a reputation for strong martial arts, the teachers who
trained the local braves might be from the village itself. Otherwise, they would
be hired from a nearby town. For remote, poorer communities, finding some-
one both qualified and willing to live and teach for extended periods could be
quite difficult. Such communities often had to settle for less reputable (even
disreputable) men for the purpose.
16. Dean’s (1993) descriptions of ritual processions in Fujian in the 1980s
and my own observations of intervillage “god-welcoming” rituals in rural Yun-
nan in the mid-1990s also support the notion that an ethos of competition
that consistently verges on, and occasionally spills over into, intergroup or
intervillage violence is not unique to Taiwan and, in fact, is an expected and
meaning-generating characteristic of temple processions; hence the tradi-
tional term, yingshen saihui, “competitive gatherings for welcoming the gods,”
continues to serve as an accurate and relevant description. This seems to hold
even where local authorities have been working to coopt, control, and manip-
ulate ritual activity, though less so where ritual performances have been fully
incorporated into a commercial tourist economy.
17. Huguo youmin: Ba jiajiang, Jin’ou Broadcasting Company, Zhanghua,
Taiwan, 1990.
18. There are many slang terms, most of them already obsolete, that
might be glossed as “hood,” “tough,” “teddy boy,” and so on (note the simi-
lar variety—and equally rapid obsolescence—of equivalent slang terms in the
English language).
19. Liubu refers to the third of the Emperors of the Five Blessings, Liu
Wenda, a plague god who has incredible martial skill and magic power. He
subdued the “eight demons of Mt. Song” (in some versions Mt. Lu), who were
so malevolent and powerful that even the Jade Emperor was said to fear them.
When he spared their lives, they responded (in good jianghu fashion) by
swearing eternal fealty to him and were transformed into the Military Retain-
ers, whom he now commands as a crack squad of ghost catchers.
20. Cf. De Groot (1892–1910: esp. book 4) for an account of similar prac-
tices in Xiamen during the last century.
21. Elder Master and Second Master is the tragic story of the sworn broth-
ers Xie and Fan, whose souls, as Seventh Master and Eighth Master, serve the
God of Walls and Moats, as well as being integral members of Wufu Dadi’s
retinue of Military Retainers.
22. “Jiajiang” has been glossed a number of different ways, including Sut-
ton’s term, “Infernal Generals.” “Military Retainers,” however, is a more literal
translation and more accurately describes their functional role as personal
servants and bodyguards of high-ranking officials.
23. The best visual presentation of the Military Retainers as practiced in
the Jiayi and Tainan areas today is Lü 2002. Sutton’s (2003) ethnohistorical
study of the Jiajiang is the best and most detailed account in any language of
the evolution of Military Retainers and their contemporary ritual practice.
A number of video documentaries of varying quality have been produced in
recent years, though none are currently in distribution.
24. Such rules are observed with varying degrees of rigor, and some tra-
ditionalists complain that today’s Retainers no longer respect or understand
the regulations and taboos.
25. Some informants actually refer to Military Retainers as mediums,
while others argue that there is a categorical distinction owing in part to

very different modes of training and function, as well as the difference noted
here—namely, that mediums can be identified with their patron deity only
when actually in trance, whereas Military Retainers, in trance or not, are in
a constant state of semi-possession. Military Retainers, moreover, derive their
efficacy from working as a unit, whereas possessed mediums perform an indi-
vidual role, either in processions or at altarside healing/oracle sessions.
26. This provides the actor with some measure of protection against
unclean things that might want to cause him harm. Having once been a
Retainer, as the stock phrase has it, “Big tragedies are reduced to minor inci-
dents, and minor incidents are avoided altogether.” And as with mediums,
there is an unspoken assumption that those with a predilection to contact
with the unseen realm have “one foot in the grave,” as it were (those prone to
possession trance are supposed to have lighter “eight characters” [bazi] than
most mortals). Performing as a Retainer, however, adds time to one’s parents’
life span (tian shou), justifying it therefore as a filial act.
27. Today, however, some troupes are manned not by high school drop-
outs and young gang members but by college students who have learned the
moves and drill formations in “culture”-oriented physical education classes.
28. Incense is offered to any seen or unseen manifestation of super-
natural power, from the wrongfully injured ghosts at accident sites to temple
images to possessed mediums. The offering of incense to Military Retainers
actors, whether visibly possessed or not, is a sign that (during the procession)
they metonymically embody the unseen exorcistic power. The Retainers are
thus “possessed” primarily through the power of ritual magic. This is unusual
in the sense that such magic is used by ritual masters to manipulate super-
natural power, even to channel it through their bodies, but never do they
become embodiments of deities per se. On the other side, mediums are seen
to embody deities or ghosts only when clearly manifesting the signs of posses-
sion trance. The Military Retainers appropriate the conventions of scripted,
costumed traditional opera, but because they also act spontaneously, they take
aesthetic representation one step further, to the complete coextension of sign
and referent.
29. Dajiale was a gambling craze that swept Taiwan during the mid- and
late 1980s. In practice, it was not one single operation but a series of numbers
rackets where people made small bets with relatively low odds (thus ensuring
a larger number of winners, though the odds were still in favor of the house/
bookmakers). The numbers were originally based on the last two digits of
the winning ticket of the Taiwan provincial “patriotic” lottery (aiguocaijuan),
but when that lottery was canceled in 1987, other sources were used, nota-
bly the Hong Kong Mark Six lottery (liuhecai; Cantonese lok hop choi). Not
coincidentally, the numbers craze rose and fell along with the 1980s Taiwan
bubble economy (stock market bubble, foreign exchange bubble, and real
estate bubble).
30. As noted, crossroads are the sites of frequent accidental deaths and
are thus considered places of pollution and danger. Wang Ye and the Military
Retainers are specifically agents of exorcism, and spirit mediums frequently
notes to pages 150 – 157 239
become possessed at such places in order to “transact the affairs [of Heaven]”
(ban shi)—that is, capture and eradicate the ghosts of accident victims that
may populate the area.
31. In the telling of the anecdote, there is no suggestion that A-beng’s
transgression had been or was about to be discovered. A skeptic, however,
might imagine that this very detail might have been deliberately elided from
the narrative, turning the event into a case of supernatural rather than human
judgment.
32. Lord Xing (Xing Wang Ye or Xingfu Qiansui) belongs to a category
of plague deities (the “lords,” wang ye or qiansui ye) in southern Fujian and Tai-
wan. Like the related Great Emperors of the Five Blessings (Wufu Dadi), the
lords are said to have been disappointed scholars who gathered together, swore
an oath of brotherhood, and sacrificed their lives for the sake of saving the
local populace. One peculiar characteristic of these cults is that their patron
deity is often a “plurality”—that is, a group of gods (usually an odd prime
number between three and seven) who are referred to as a single entity and
worshipped collectively (although they are represented by multiple images
arrayed in order of seniority on the altars of cult temples). Xingfu Qiansui
(or Xing Wang Ye) is the collective name of seven deities, sworn brothers in
life who (through their oath) share the surname Xing, the family name of the
eldest “brother.” The patron deity of the Loyal Harmony Temple is not the
“plurality” of seven but rather the fifth “brother,” usually called “Five Thou-
sand Years,” Wuqiansui. When devotees among my informants speak of Lord
Xing, especially in the context of miracle narratives and testimonials, they are
usually referring to the individual patron deity.
33. As McCreery (1990:12–13) describes it, an illness explained as due
to debts owed parents from a previous life is implicitly caused directly by the
angry ghosts themselves. A-luo’s explanation, on the other hand, suggests
that Jiafu’s illness was determined mechanically (according to the popularly
understood laws of karma) by his unfilial behavior in a past life, his fate adju-
dicated in bureaucratic terms (“his case came up for review”) and mediated
through the extrajudicial power of the god Lord Xing. Here, too, however,
the debt must be paid and the parental ghosts propitiated as a prerequisite
for recovery.
34. Tudi Gong is the generic name for the low-level earth or locality gods
whose shrines and temples dot the countryside and old neighborhoods of Tai-
wan and, until the 1950s, much of mainland China as well. During the Dajiale
numbers game craze, many players sought out earth god shrines in remote,
lonely places, where, it was assumed, the ghosts of the unworshipped dead
were particularly numerous. Such earth gods, then, would likely have access to
“inside” information about illegal and immoral matters and could be enticed
to divulge what they might know about (for example) what number would pay
off in the next round. This would be useful information for both individual
gamblers and policy shop operators.
35. The term mingpai, literally a “bright tablet,” first appeared in the late
1980s. The term referred specifically to a winning lottery number revealed in
advance through divination or by a god (usually a god or demon of low rank)
through a possessed spirit medium. A derivative meaning—that is, any illicitly

obtained but reliable information (especially on the stock market)—has since
entered common usage in Taiwan. In Mandarin (but not Holo) it is hom-
onymous with the term for “famous name brand”; thus the slang phrase you
mingpai jiu you mingpai — in other words, if you have reliable connections, you
will get rich and can enjoy an ostentatious lifestyle, wearing, eating, drinking,
and driving famous brands.
36. For the believers, all of these sudden manifestations were signs that
this god’s ling had become particularly powerful and was concentrated in
the local area. According to the typical narrative sequence of cult develop-
ment, the god was purposefully bestowing favors on these believers with the
expectation that they would reciprocate once they had enough resources by
building him a temple and carving a new image. They would, in other words,
help the deity to take form, to become more present in the visible world—to
move more fully from the yin to the yang realm. This would enable the god to
extend his power and establish greater authority, which would in turn ensure
even greater benefits and status for his worldly followers.
37. Such “magic battles,” frequently depicted in classical vernacular nov-
els of fantasy and martial arts, are a favorite theme of films and television seri-
als about ritual masters, martial artists, and Daoist hermits. The battle between
Jiang Ziya (and his “hired gun” Lu Ya) and Zhao Gongming in Fengshen yanyi
(Investiture of the Gods) is one locus classicus of the genre (Wenhua Tushu
edition, 334 –341).
38. This isn’t to say that dramatic social and cultural change has occurred
only since 1949 or that such change has been driven only by the policies and
practices of the Communist Party. Popular religion and the complex of under-
lying meanings and assumptions that have sustained it have come under suspi-
cion and attack from imperial, Republican, and Communist officials in equal
measure. Obviously, however, the scope and outcomes of such critiques have
varied widely.
39. There have been numerous “anti-superstition” campaigns over the
years in the PRC, but they have not all been equally effective, nor have they
necessarily targeted a single category of beliefs and practices. As Mark Selden
(1995:24ff.) points out, suppression on a broad scale began only around 1957,
although it dramatically gained steam during the ideological struggles and
mass mobilizations of the following two decades. However, many remote and
peripheral areas to the south and southwest—rural Dali, for example — did not
feel the full impact of anti-superstition policy until the Cultural Revolution.
40. I agree here with Siu (1989) that the reappearance of some prac-
tices in the post-Mao era is not a simple recovery of the past. Yet at least in
some of the communities studied so far in Yunnan, Fujian, and elsewhere, it
is clear that some of the context, such as a shared belief in efficacious power,
has survived. For example, the festival processions described by Dean (1993)
in Fujian in the late 1980s, along with some of those I observed in villages
in the Dali area in the mid- and late 1990s, attest to the reemergence in the
reform era of a customary, territorially defined local identity that Mao and his
associates found so vexing ( Vogel 1969). It is true that in many cases temple
notes to pages 160 –163 241
activities such as processions are barely tolerated by the authorities, or they

may be promoted as “folk traditions” amenable to official interests, such as a
local investment in ethno-tourism (cf. Schein 2000). At the same time, it is sig-
nificant that at least some assertions of local identity and cultural authenticity
should specifically invoke local deities and a belief in the practical (including
economic) efficacy of religious rituals.
41. Christianity and Buddhism are the fastest-growing religions in the
PRC. Local or popular religion has actually played a relatively small role in
this process.
42. Some city temples have been restored and, in recent years, have even
enjoyed a measure of official recognition. But most such temples are closely
monitored, if not directly managed by one or another city- or provincial-level
bureau. Ritual services are generally provided by clergy trained in govern-
ment-sanctioned seminaries, and activity is restricted to the sites themselves.
There is little, if any, involvement or input from the local community, and
even the popular gods to which the temple was originally dedicated are often
replaced by Daoist or Buddhist deities.
43. The task of administering religious organizations, buildings, and
activities varies from place to place and situation to situation. The Bureau of
Tourism often regulates physical sites, including site rules, entrance fees, and
so on, while the Bureau of Religion is more focused on religious assemblies
and organizations and on determining which ones qualify as proper religion
(guifan zongjiao) and which should be considered heterodox (xiejiao) and thus
prohibited.
44. Some but not all of these children are substituting in roles that would
traditionally have been assigned to older participants. In fact, children are
essential to many procession formations. In Taiwan, the Emperors of the Five
Poisons (Wu Du Da Di) and the Thirteen Imperial Bodyguards (Shisan Tai-
bao) are Military Retainer–style formations performed exclusively by children;
for many years the Tian Hou Temple’s entry in the Taidong Yuanxiao pro-
cession has been a large float seating about two dozen elaborately costumed
young women and girls portraying Mazu and her retinue. And in Hong Kong,
costumed boys and girls secured on wire frames are carried along the pro-
cession route of the annual Cheung Chau Bun Festival. However, in Taiwan
and Hong Kong (and in pre-1949 mainland China), while martial procession
troupes may include and even feature children, most of the actors are older,
mostly male, and drawn from the social margins. The troupes in the Dali area
are therefore categorically distinct and, to some extent, discontinuous with
pre-1960s local tradition.
45. Qinghai was another of the male students, seated across from me;
although ethnic Han, he had grown up in Xining, capital of Qinghai Province,
which comprises the rugged, arid, and sparsely populated northern half of
the Tibetan Plateau. As I got to know the group better, I understood that the
nickname referred to his exceptional physical strength, innocent demeanor,
and fierce determination both in his studies and in fights, qualities the group
whimsically associated with his wild and rugged place of origin.
46. Alex’s description was rather brief. I was later able to corroborate his
account, and the account presented here includes details revealed in later
conversations I had about the incident with Er Ge and others.
47. In my judgment, the guards were not simply acting out their frustra-
tion or incipient sadism. Although they may have technically violated written
Public Security Bureau (PSB) policy, they were following what is in practice
more or less standard procedure for low-level, front-line law enforcement
workers in the PRC. At the same time, a penchant for violence can also be
reason for dismissal from the PSB. Since Hu Jintao, China’s president and
general secretary, announced a concerted campaign against crime and cor-
ruption, there have been several reported cases of such former PSB officers
arrested as gang enforcers.
48. The difference between the two sounds is more subtle in Xiaguan
dialect than in Mandarin. In Mandarin the distinction is purely tonal, whereas
in Xiaguan dialect the tone contours are very close, while there is a slight dif-
ference in aspirating the initial sound of the second syllable ( fu).
49. Only one researcher that I am aware of to date has attempted to con-
tact gang members or gain access to police files during fieldwork in main-
land China (Chen An 2005). Unsurprisingly, he was not successful in either
endeavor (80 – 81), and his report relies entirely on what I would consider
dubious, or at best uncorroborable journalistic sources (81– 83). My com-
ments here are drawn instead on general observation and conversations with
a variety of individuals in the PRC over many years, including, but not limited
to, taxi drivers, local academics, and retired PSB officers.
Chapter 6 Wine, Women, and Song
1. “Jinling Jiu Si Liu Bie” (Parting at a Jinling Tavern, ca. 750) is one of
the many verses by the great Tang dynasty poet Li Bai (701–762) composed on
themes of drinking and fellowship. Translation by the author.
2. Here it may be instructive to consider other uses of chou. In classi-
cal texts, chou may signify superiors rewarding inferiors, as in choulao, a lord
rewarding the efforts of his retainers. Or it may indicate inferiors’ supplica-
tion of superiors in recompense for favors (by gods: choushen; by gods or non-
supernatural superiors: chouyuan)—with the implication that the supplicant
will receive or has received some benefit in return. These usages articulate a
field of social interaction that (1) is based on an asynchronous exchange of
gifts and/or favors, and (2) constitutes social relationships as a hierarchy. Here
it might be objected that the terms most often used to refer to entertainments
among friends and associates—chouzuo and yingchou—carry no implication of
hierarchy, only reciprocity. Nevertheless, both etymological analysis and eth-
nographic evidence support the argument that hosted, reciprocal entertain-
ment in Taiwan entails competition for status, even among recognized social
equals and members of the same in-group. Both codified rules of etiquette
and the spontaneous acts and expressions that define modes of speech and
action during these sessions strongly suggest that status competition is one, if
not the primary, pretext for yingchou.
3. The terms guanxi and li are not used as broadly in Taiwan as they are
in the PRC. In Taiwan one regularly hears the terms renqing (human feeling),
ganqing (emotional bonds/connections), and yiqi in the context of sociality

and interpersonal obligations. Nevertheless, whether in Taiwan or the PRC,
one sees gift exchange and reciprocal entertaining fulfilling both emotional
and instrumental purposes (cf. Gold et al. 2002; M. Yang 1994; see also Hsing
1997).
4. This is a special usage, of course, as chhit-tho in everyday conversation
refers to recreation of all sorts. One is actually most likely to hear the term
when taking leave at the house of a friend or acquaintance: -U-ieng lai chhit-tho
corresponds to the Mandarin You kong lai wan, “Come by and relax whenever
you’re free.”
5. In Taiwan, an aggressive style of oratory is considered properly mas-
culine. It is common practice, for instance, to pepper one’s narrative with
loud and frequent voicing of the verb kan (fuck) or its extension, the “three
character classic.” Another distinctly male habit is to add emphasis after mak-
ing a point by uttering the syllable “ho *,” tilting the head slightly upward,
leaving the mouth open and pointing the chin forward, and looking straight
at the listener—an expression of defiant challenge. These and other tech-
niques of masculine performance in everyday interaction are reminiscent of
masculine socializing styles found in other patriarchal warm-weather cultures
of the Mediterranean and Latin America; see, for example, Driessen 1983 and
Herzfeld 1985 (esp. 123 –162).
6. While private houses are frequently used as sites for reciprocal enter-
taining in rural communities, most such activity takes place in the open court-
yard or in the “public interior” (tang or ting), clearly distinguished from the
private, domestic interior, which is off limits to all but kin and the most inti-
mate friends during public events.
7. In cosmopolitan metropolises like Taipei, Beijing, Hong Kong, and
Shanghai, there is a visible trend toward more Western styles of drinking and
socializing. Here, just as one can find a KFC next door to a dumpling shop,
bars, pubs, and dance clubs share the block with karaoke lounges, hostess
bars, and saunas. Sophisticated young men and women (some are “sea tur-
tles” [haigui] who have returned from abroad, perhaps with an MBA in hand)
drink beer from the bottle or sip martinis and schmooze; college-age friends
might spend weekend nights dancing in basement discos, drinking vodka and
Red Bull. Where this trend may lead no one can predict, but the practices
described in this chapter are in no danger of disappearing. A visit to any sec-
ond- or third-tier city—that is, just about anywhere outside of Shanghai and
Beijing—should put any such doubt quickly and convincingly to rest. I sus-
pect, too, that most of those young sophisticates from Shanghai and Beijing
know how to offer and respond to a “challenge” toast, play at least some tra-
ditional drinking games, and greatly prefer going out with a group of friends
or colleagues to dropping in alone to the neighborhood pub (which doesn’t
exist in China, strictly speaking).
8. According to the popular legend, the karaoke machine was invented
sometime in the 1970s by a bar owner in the Japanese port city of Kobe. When
his roving instrumentalist failed to show one night, he attached a mixer to
the boom box, popped in a cassette of popular songs, and figured out how
to filter out the voice track. Thus kara — empty — ōkesutora — orchestra—was

born. The Chinese term is a transliteration of the Japanese kara, rendered in
Chinese characters (ka la), and the substitution of the upper case Roman let-
ters “OK” is for ōke (the abbreviated form of ōkesutora ).
9. It is interesting to note that the Kansai region, which includes Kobe,
the birthplace of karaoke, has a reputation in Japan for being the source of
the same kinds of popular culture trends for which Taizhong is known in
Taiwan.
10. These range from the simple and ubiquitous “rock, paper, scissors”
hand games to long, complex tongue twisters. The loser or one who trips over
his words must drink penalty wine, which, of course, leads to slower reflexes
and impaired speech and more penalties.
11. The association between drinking and camaraderie is aptly illustrated
by the Chinese idiomatic expression, “When drinking with a true friend, even
a thousand rounds isn’t enough.”
12. In the late 1980s, as martial law was lifted in Taiwan and restrictions
on travel to the mainland were gradually relaxed, among the myriad Chinese
products that made their way onto the Taiwan market in increasing volume
were People’s Liberation Army surplus Type 54 “Black Star” semi-automatic
handguns. The easy availability of cheap, lethal firepower led to what was, by
Taiwan standards, an epidemic of shooting incidents, many of them famously
inside and outside nightclubs and restaurants throughout Taiwan. According
to news reports, many such incidents were the result of ongoing disagree-
ments or vendettas among rival gangs; some, however, were spontaneous
explosions of temper over rejected toasts or perceived insults that in the past
would have been resolved with fists or knives.
13. The selling of “intimations of sex” rather than sex per se is reminis-
cent of similar establishments in Japan. See Allison (1994: esp. 132–133).
14. This link among economic, social, and sexual prowess, as well as the
mutuality of the status-producing value of the relationship between hostesses/
courtesans and customers within their separate but intersecting social milieus
is hardly unique to Taiwan or to the contemporary period (see, for instance,
Hershatter 1997 and Mann 1997 for accounts of the social world and culture
of the entertainment and sex trade in Republican-era Shanghai and late-impe-
rial China respectively). My examples here, rather, highlight the distinction
between two modes of representation that comprise part of the dialectic of
gender performance and identity that gives carousing its productive function:
the recounting of sexual exploits is part of a sometimes evaluative, sometimes
creative, discourse that precedes and follows carousing events, whereas the set
of public acts and attitudes, both verbal and non-verbal, expressed in the case
of an ongoing sexual relationship, is demonstrative (or “performative,” to use
more familiar jargon).
15. This has since changed, as the development of Taidong’s nightlife
paralleled a dramatic economic expansion in the mid- and late 1990s.
16. I wasn’t privy to exactly what was at stake that evening, although
clearly A-beng was making a pitch for some kind of financial cooperation with
the Taipei group. Toward the end of that very long night, a very drunk, thin,
middle-aged man whom I hadn’t seen before and who had remained rather
quiet throughout the negotiations explained to me, soto voce, that he was an
accountant at such-and-such a bank and worked for the operation by launder-
ing money in such-and-such a way, earning a commission for each batch of
funds.
17. These occasionally included gamblers and a few gangsters of a more
professional stripe.
18. I was more often invited by those my age or a little younger, but after
the 1991 Yüanxiao procession, at which point I had gotten to know the temple
leaders better, I was also sometimes asked to partake in the drinking parties of
the cohort who were then in their mid-forties. It was in this group, for whom
carousing was an essential feature of their economic lives, that I saw the most
intense competition in drinking.
19. The customary phrase “How should I address you?,” Zenme chenghu,
is more formal than the more casual “What’s your name?” (Ni jiao shenme
mingzi; H. Li kio sim-mih mia*). This serves a dual purpose: the women, who,
despite being “hostesses,” are in the subordinate, inferior position, defer to
the customer as to the level of formality; there is the further, unarticulated
assumption that customers prefer to use nicknames or pseudonyms (nearly all
the women use “stage names,” yiming). If the customer replies with his family
name (xing ) , however, hostesses usually respond by calling him “Big Brother
X” rather than “Mr. X,” a subtle shift from formality to (dependent, vulner-
able, subordinate) intimacy.
20. Drinking “protocols” ( jiuling) and “regulations of the goblet” (shang
zheng) , which describe proper drinking etiquette and provide instructions
and rules for drinking games, date from no later than the Tang dynasty (Guo
1989:185; Huang Shu-min and Hsu 2000:71–73).
21. David Schak (n.d. and personal communication) points out that
when expatriate Taiwanese entrepreneurs and managers working in China
gather to socialize, they may drink intensely but not competitively. I would
still maintain, however, that the contest does not necessarily take the form of a
symmetrical competition but entails rather meeting the challenge of expecta-
tions. Some groups are certainly less exacting than others in this regard, and
desirable qualities probably vary from one group to another. I would imagine,
for example, that among sojourning businessmen in the potentially hostile
environment of southern China’s export manufacturing zones, qualities like
reliability and quick-wittedness may be valued more highly than more macho
attributes such as daring or physical stamina.
22. Here I am speaking of singing as a leisure activity; the karaoke singing
competitions often sponsored by restaurants, clubs, and organizations belong
to a separate category of activity.
Chapter 7 Conclusion
1. That is, all stages of Chinese male personhood—filial son, patriarch,
and ancestor—are premised on subordinating reproduction (the production
of an other) to self-production. Only by producing sons can one become a
truly filial son, and only such a filial son is qualified to sacrifice to ancestors,
with the expectation that one will take the place of one’s own father as the
household’s dominant male. Thus the patriarch embodies sexual potency
(although that perception may be compromised when that position has been
achieved through other means, such as adoption).
2. Other notably filial characters in the Water Margin include Weapons
Instructor Wang Jin and “Cloud-Entering Dragon” Gongsun Sheng. It is inter-
esting that (conversely), “Nine-Dragoned” Shi Jin’s obsession with martial arts
drove his mother to distraction (and an early grave), according to his worried
but tolerant father.
3. “Tianya liulang’er” (A Prodigal Son at the Ends of the Earth), per-
formed by Ye Qitian; recorded/released 1986; composer/author unknown.
4. Songs, like rituals, can be scrutinized as scripts, but it is only in the
instances of performance and experience (which can merely be privately
reading them to oneself and indulging the emotional response and observing
the echoes of memory and allusion) that we can observe their productive effi-
cacy. Violent aggression serves as a powerful mode of transformative agency
in many cultures. Michelle Rosaldo, for example, observed that “In severing
and tossing human heads, Ilongot men recount, they could relieve hearts
burdened with the ‘weight’ of insult, envy, pain, and grief; and in discarding
‘heavy’ thoughts, they could achieve an ‘anger’ that yields ‘energy,’ makes shy
and burdened youths ‘the same’ or equal to their peers, and ‘lightens’ both
their footsteps and the feelings in their hearts” (1983:137).
5. For example, in exploring what happened when A-kiat-a, in trance
kicked an unfortunate censer bearer from a rival temple to the ground, it
turned out that there was a long history of tension and confrontation between
the two participating groups; that there was no personal enmity between the
two protagonists, who didn’t actually know each other, and little chance of
premeditation in the attack; that A-kiat-a had trained in the martial arts since
he was a child and had a mild speech defect but in everyday life was anything
but violent, known as a model family man; and so on.
Glossary
aiguocaijuan (Taiwan provincial “patriotic” lottery [pre-1987]) 愛國彩卷

an wai wuying (pitching the outer five camps) 安外五營
ao-soa* (the mountains beyond) (Holo) 後山
Asahimura (Xucun) 旭村
baijiu (clear distilled liquor) 白酒
ba jiajiang (eight/ten military retainers) 八家將
ban shi (take care of [ritual] business) 辦事
baochou (exact revenge) 報酬
bao’en (requite a favor or blessing) 報恩
Baoshan 保山
Beinan 卑南
Beinan ting 卑南廳
benzhu (lord of the locality [Bai religion]) 本主
bu jiang yi (lacking a sense of righteousness and honor) 不講義
caizi (talented young scholar) 才子
chashi (“tea room”) 茶室
chayi shi (“tea master”) 茶藝師
Cheng’gong (Xin’gang, Shinkō) 成功 (新港)
chou (to reward, compensate) 酬
chu dao (to begin a career [esp. in entertainment], make one’s debut) 出道
chu jianghu (to begin a career [esp. in illicit or risky trade/business])
出江湖
cu (the second round or toast of a feudal feast [archaic]) 醋
da ge (elder brother, boss) 大哥
Dajiale 大家樂
Dali 大理
dao (way, path, “the Way”) 道
daoshang xiongdi (brothers of/on the [dark] path or way) 道上兄弟
Dazhuang 大庄
247
248 glossary
diao xia chang (shrimp fishing pool) 釣蝦場

difang rentong (local identity) 地方認同
diling renjie (a place of numinous power and exceptional people) 地靈人傑
ding (an able-bodied man) 丁
disha (baleful spirit of the earth) 地煞
doujiu (drinking contest) 鬥酒
fajiu (wine penalty) 罰酒
fangzhi (local gazetteer) 方志
fashi (ritual master; also polite term for Buddhist clergy) 法師
feili (molest) 非禮
fengliu caizi (talented and romantic scholar) 風流才子
Fengshen yanyi 封神演義
fengshui 風水
Fengtian 豐田
fudan (the “spleen,” or middle section of a paper talisman) 符膽
fujiao (the “foot,” or bottom section of a paper talisman) 符腳
Fukenju (Bureau of Pacification and Reclamation) 撫墾局
Fuli 富里
fuli (assistant manager [abbrev.]) 副理
fulu (charm, talisman) 符籙
ge (dart, lance) 戈
ge (pattern) 格
getihu (private business, entrepreneur [PRC]) 個體戶
gongde xiang (collection box) 功德箱
gongfu (time, effort; colloquial term for the martial arts, kung fu) 功夫
guandao (broad-bladed halberd, signature weapon of Guan Gong) 關刀
Guan Gong 關公
guanxi (social network, web of influence, “connections”) 關係
gui (ghost, demon) 鬼
guifan zongjiao (“model” or “proper” religion) 規範宗教
Gu Long 古龍
guluozhen (traditional percussion troupe) 鼓鑼陣
Handan Ye 寒單爺
hanhua (sinified, assimilated to Chinese culture) 漢化
haohan (real man, good fellow, hero) 好漢
heidao (the dark way, gangland, crime and corruption) 黑道
heidao fenzi (gangsters [rhetorical, derogatory]) 黑道分子
glossary 249
he jiu (Holo lim-chiu) (drink alcohol, go drinking) 喝酒 (飲酒)

hoan-a (Holo) (aborigines, barbarians [derogatory]) 番仔
hoat-a tin (Holo) (sorcerer’s troupe) 法仔陣
hoe-bian-e (Holo) (multicolored face [mask, face paint]) 花面的
houshan (the mountains beyond) 後山
hua quan (drinking/hand guessing games) 划拳
hui (association, guild) 會
huiguan (association or guild hall) 會館
hun (to get by, live by one’s wits) 混
hunhun (small-timer, street tough) 混混
huobajie (torch festival) 火把節
Huwei 虎尾
ji (halberd) 戟
jia (first of the Ten Heavenly Stems; armor) 甲
jiajiang (retainer, Military Retainers) 家將
jianghu (the underworld, “rivers and lakes”) 江湖
jianghu shang de (of the underworld) 江湖上的
jianghu xiongdi (brothers of the rivers and lakes) 江湖兄弟
Jiangjun Dong 將軍洞
Jiansu Liumang Tiaolie (Anti-Hooligan Act [Taiwan]) 檢肅流氓條例
jiaotou (turf-based; local gang boss) 角頭
jiaren (beautiful woman [literary]) 佳人
jiedusuo (drug rehabilitation center) 解毒所
jie jiu (sober up) 解酒
jieyanwan (kick-the-opium-habit pills) 戒菸丸
jiguan (ancestral [native] place) 籍貫
jing (offer respect) 敬
jinlu (gold- [paper] burning furnace) 金爐
Jin Yong 金庸
jiudan (drinking courage) 酒膽
jiudian (wine shop [premodern]) 酒店
jiuliang (capacity for alcohol) 酒量
jiuling (drinking protocols) 酒令
jiuzhan (drinking contest) 酒戰
juancun (military dependents’ villages [Taiwan]) 眷村
jue (secret manual) 訣
ka-chiong kin’na (Holo) (Military Retainer “kids”) 家將囝仔
250 glossary
kak-thau (Holo) (turf-based; local gang boss) 角頭

ka la OK (karaoke) 卡拉OK
kao jun (rewarding the troops) 犒軍
kaozheng (philology; the Qing Neo-Confucian “evidentiary research”
school) 考證
karaoke ( Japanese) カラオケ(空オーケストラ)
ke jing zhu (guests toast the host) 客敬主
keju (imperial civil service examination system) 科舉
ke-pho (Holo) (nosy, interfering) 家婆
ketou (kowtow) 嗑頭
ke xiang quan (guests urge each other [to drink]) 客相勸
kio-a (Holo) (miniature sedan chair) 轎仔
Lao Hu (Old Hu) 老胡
laohu/laofu (tiger) 老虎
li (pattern, logic; to bring order) 理
li (gifts; decorum; reciprocity) 禮
lihai guanxi (a relationship based on profit and loss [contrasts with kinship
and friendship]) 利害關係
ling (numinous power of supernatural beings) 靈
lingyan (responsive to petitioners’ requests [gods]) 靈驗
liqi (calculating and manipulating qi [school of fengshui]) 理氣
liuhecai (Mark Six lottery [Hong Kong]) 六合彩
liumang (gangster, hooligan) 流氓
liu yi (the Confucian six arts) 六藝
lok hop choi (Cantonese) (Mark Six lottery [Hong Kong]) 六合彩
lomoa* (Holo) (gangster, hooligan) 鱸鰻
lomoa* sin (Holo) (gangster god) 鱸鰻神
luan (chaos, disorder) 亂
luopan ( fengshui compass) 羅盤
Luye 鹿野
luzhu (incense master) 爐主
meiqi (bad luck caused by baleful influences) 霉氣
mianzi (“face”; social capital or reputation) 面子
min (common people) 民
mingpai (accurate numbers tip from a supernatural being; winning big
[gambling]) 明牌
Nanhui Gonglu (Southern Link Highway) 南迴公路
glossary 251
nanyang (the south seas; Southeast Asia) 南洋

Nezha (Nuozha) 哪吒
ng kio-khah (Holo) (squeeze under the god’s palanquin) 倰轎腳
o-bian-e (Holo) (black-faced [mask, face paint]) 烏面的
pat/sip ka-chiong (Holo) (eight/ten military retainers) 八／什家將
peijiu (female drinking partner; bar hostess) 陪酒
pinjiu (all-out drinking) 拼酒
popomamade (fussy, foolishly sentimental) 婆婆媽媽的
qi (circulating cosmic force, vital energy) 氣
qiang (spear) 槍
qiansui ye (thousand-year lords—Wang Ye) 千歲爺
qing (feeling, emotion) 情
renqing (protocols of human feeling; mutual favors and obligations) 人情
Ren zai jianghu, shen bu you ji (In the jianghu, you are not your own master)
人在江湖身不由己
roushen Handan Ye (“flesh body” Handan Ye) 肉身寒單爺
Sanguo yanyi 三國演義
saozi (sister-in-law [elder brother’s wife]) 嫂子
sha (baleful spirits) 煞
shang zheng (regulations of the goblet—traditional Chinese drinking
etiquette) 觴政
shanming shuixiu (bright mountains and limpid waters; beautiful natural
scenery) 山明水秀
Shao bu du Shuihu, lao bu kan Sanguo ( Young men should not study the
Water Margin; old men should not read the Three Kingdoms)
少不讀水滸老不看三國
shaoye (young fellow, waiter) 少爺
shen (god, deity) 神
shengdou (peck and bushel measure [Torch Festival]) 升斗
shenling (numinous spirits—gods) 神靈
Shikano (Luye) 鹿野
Shisan Taibao 十三太保
shufan (“cooked” [assimilated] aborigines) 熟番
Shuihu zhuan 水滸傳
Shuiwei 水尾
Song Jiang Zhen 宋江陣
Taidong 台東
taijiquan (tai-chi [martial art]) 太極拳
252 glossary
tai sin ang-a (Holo) (frame puppets) 大神尪仔

tang (lineage hall; branch of secret society) 堂
tangki (Holo) (spirit medium, shaman) 童乩
tianbing shenjiang (spirit soldiers and generals) 天兵神將
Tiangang (Big Dipper Generals) 天罡
tiangong lu (Emperor of Heaven’s censer) 天公爐
tian shou (extend [parents’] life span) 添壽
tianxia (all under heaven; the social/political world) 天下
ting (a frontier commandery, military administrative region [Qing]) 廳
ti tian xing dao (carrying out the will of Heaven) 替天行道
Wang Ye 王爺
wan sui (“ten thousand years”; the emperor) 萬歲
weifa (unissued; potential [metaphysics]) 未發
wei zuren zheng yi kou qi (make good for the sake of the clan)
為族人爭一口氣
wen (the civil, literary; written language) 文
wenpan (civil magistrate) 文判
wenren (literatus) 文人
wen wu zhi dao, yi zhang yi chi (the way of King Wen [civil] and King Wu
[military]: the bow is drawn, the arrow is released) 文武之道一張以弛
Wing Chun (Cantonese) 詠春
wu (the martial, military) 武
wu (shaman [archaic]) 巫
Wufu Dadi (Emperors of the Five Blessings) 五福大帝
wuguan (armory, martial arts training hall/association) 武館
Wuguanzhuang 五官莊
wuju (martial opera) 武劇
wulin (the competitive space of reputation in the Chinese martial arts
world) 武林
wupan (military magistrate) 武判
Wu Song 武松
wuxia (knight-errant) 武俠
wuxing (invisible, insubstantial) 無形
wuzhen (martial ritual procession troupe) 武陣
xia bu liao tai (no way out of an embarassing situation) 下不了台
xiagang (be laid off from state-owned enterprise [PRC]) 下崗
Xiaguan 下關
glossary 253
xiahai (“leap into the sea”; to give up a secure job in hopes of getting rich
in business) 下海
Xia ke xing 俠客行
xian (contribute, offer up; contribution) 獻
xie (evil, heterodox) 邪
xiejiao (heterodox religion) 邪教
Xiluo 西螺
xin (trust; to believe in) 信
xing (nature, sex) 性
Xingfu Qiansui 邢府千歲
xingshi (form; landform school of fengshui) 形式
Xizhou 喜洲
Xucun 旭村
xungen (root-seeking) 尋根
yamen (magistrate’s office and residence [imperial period]) 衙門
yang (south-facing, sun-bathed; male principle; realm of the living) 陽
yangzhai (dwellings of the yang realm; houses of the living [ fengshui ]) 陽宅
yayi (yamen runner, bailiff ) 衙役
yi (tethered dart) 弋
yi (honor, righteousness) 義
yifa (already issued; realized [metaphysics]) 已發
yiming (stage name) 藝名
yin (north-facing, in shadow; female principle; realm of the dead) 陰
yin (licentious) 淫
ying (response; to respond) 應
yingchou (reciprocal entertaining; social obligations) 應酬
yingshen saihui (temple festival) 迎神賽會
Yinqiao 銀橋
yingxiong haohan (heroic man, man of prowess) 英雄好漢
yinzhai (dwellings of the yin realm—that is, graves [ fengshui ]) 陰宅
yiqi (honor, righteousness) 義氣
yiqing zhuan’an (police campaign targeting organized crime [Taiwan,
1984]) 一清專案
yiyong zhen nanzi (a righteous, brave, and real man) 義勇真男子
you mingpai jiu you mingpai (with a good tip [on the numbers] one can buy
brand-name luxuries) 有明牌就有名牌
254 glossary
youxing (visible, substantial) 有形

yuanzhumin ([Taiwan] aborigines; lit. original inhabitants) 原住民
zai daoshang hun (getting by “on the path”; living an idle life) 在道上混
Zhao Gongming 趙公明
zheng (upright, authentic, orthodox) 正
zhengshen (upright/orthodox deities) 正神
zhengyi (true honor, righteousness) 正義
zhentou (ritual procession troupe; drill formation [military, ritual]) 陣頭
zhi (to stop, hinder, restrain) 止
Zhiben 知本
zhilizhou (autonomous region [Qing period]) 直隸州
zhong (loyalty) 忠
Zhongkui 鍾馗
zhong se qing you (obsessed with sex [and thus] ignoring friends) 重色輕友
Zhoucheng 周城
zhu quan ke (host urges guests [to drink]) 主勸客
zou fo (“run the Buddha”; ritual of carrying the image of the God of
Wealth through a barrage of firecrackers) 走佛
zuhe (to organize; “venture squad”) 組合
zuji (ancestral place of origin) 祖籍
zuling (spirits of ancestors) 祖靈
Bibliography
Allan, Sarah. 1981. “Sons of Suns: Myth and Totemism in Early China.” Bulle
tin of the School of Oriental and African Studies (University of London) 44
(2): 290 –326.
Allison, Anne. 1994. Nightwork: Sexuality, Pleasure, and Corporate Masculinity in a
Tokyo Hostess Club. Chicago: University of Chicago Press.
Atwill, David G. 2003. “Blinkered Visions: Islamic Identity, Hui Ethnicity, and
the Panthay Rebellion in Southwest China, 1856–1873.” Journal of Asian
Studies 62 (4): 1079 –1108.
Bai Junxiong 白俊雄. 1971. Guoshu bingqi zatan 國術兵器雜談 (Essay on [Chi-
nese] Martial Techniques). Taipei: Hualian Chubanshe 華聯出版社.
Bateson, Gregory. 1972. Steps to an Ecology of Mind: Collected Essays in Anthropol-
ogy, Psychiatry, Evolution, and Epistemology. San Francisco: Chandler.
Booth, Martin. 1999. The Dragon Syndicates: The Global Phenomenon of the Triads.
New York: Carroll and Graf Publishers.
Boretz, Avron. 1995. “Martial Gods and Magic Swords: Identity, Myth, and Vio-
lence in Chinese Popular Religion.” Journal of Popular Culture 29 (Special
Issue: Anthropology and Popular Culture): 93–109.
———. 白安睿. 2001. “Minsu quyi” 民俗曲藝 (Folk Music and Performance
Troupes). In Taidong Xianshi: Hanzu Pian 台東縣史：漢族篇 (Histori-
cal Gazetteer of Taidong County: The Han Chinese), ed. Lin Meirong
林美容, 113–161. Taidong: Taidong Xian Zhengfu 台東縣政府.
———. 2002. “Carousing and Masculinity: The Cultural Production of Gen-
der in Taiwan.” In Women in the New Taiwan: Gender Roles and Gender Con-
sciousness in a Changing Society, ed. Catherine Farris, Anru Lee, and Mur-
ray Rubinstein. Armonk, N.Y.: M. E. Sharpe
———. 2003. “Righteous Brothers and Demon Slayers: Subjectivities and Col-
lective Identities in Taiwanese Temple Processions.” In Religion, and the
Formation of Taiwanese Identities, ed. Paul R. Katz and Murray Rubinstein.
New York: Palgrave MacMillan.
———. N.d. “Globalizing Local Identities: Urban Migrants and the New Tai-
wan Avant-Gardes.”
Bosco, Joseph. 1992. “Taiwan Factions: Guanxi, Patronage and the State in
Local Politics.” Ethnology 31 (2): 157–183.
Bourdieu, Pierre. 1977. Outline of a Theory of Practice. Cambridge: Cambridge
University Press.
Brandes, Stanley. 1980. Metaphors of Masculinity: Sex and Status in Andalusian
Folklore. Philadelphia: University of Pennsylvania Press.
255
256 bibliography
Brown, Melissa. 2004. Is Taiwan Chinese? The Impact of Culture, Power, and Migra-
tion on Changing Identities. Berkeley: University of California Press.
Brownell, Susan, and Jeffrey N. Wasserstrom. 2002. Chinese Femininities, Chinese
Masculinities: A Reader. Berkeley: University of California Press.
Cai Huimin 蔡惠敏. 2000. “Qingshaonian canyu jiajiangtuan de beijing ji ying
xiang: Y i guanjiangshou shaonian wei zhu” 青少年參與家將團的背景及
影響: 以官將首少年為主 (Backgrounds and Influences of Youth Who
Participate in Jiajiang Troupes: Focus on Guanjiangshou Youth). Mas-
ter’s thesis, Department of Social Work, Soochow University (Taiwan)
東吳大學社會工作系.
Chang, Kwang-chih. 1969. Fengpitou, Tapenkeng, and the Prehistory of Taiwan.
New Haven: Yale University Publications in Anthropology, No. 73.
———. 1986. The Archaeology of Ancient China. New Haven: Yale University
Press.
Chang, Kwang-chih, and Ward H. Goodenough. 1996. “Archeology of South-
eastern China and Its Bearing on the Austronesian Homeland.” In Pre-
historic Settlement of the Pacific, ed. Ward H. Goodenough, 28–35. Philadel-
phia: American Philosophical Society.
Chen An. 2005. “Secret Societies and Organized Crime in Contemporary
China.” Modern Asian Studies 39 (1): 77–107.
Chen Guofu 陳國符. 1963. Daozang yuanliu kao 道藏源流考 (Investigations
into the Origins of the Daoist Canon). Beijing: Zhonghua Shuju.
Chen Hongtu. 2002. “Guanying yimincun yu dong Taiwan de shuili kaifa
(1909–1946)” 官營移民村與東台灣的水利開發 (1909–1946) (Official
Immigrant Villages and the Irrigation Development of Eastern Taiwan
(1909–1946)). Dong Taiwan yanjiu 東台灣研究 ( Journal of Eastern Tai-
wan Studies) 7:135–163.
Chen Pingyuan 陳平原. 1992. Qiangu wenren xiake meng 千古文人俠客夢 (The
Literatus’ Ancient Knight-Errant Dream). Taipei: Maitian (reprint 1995).
Chesneaux, Jean. 1971. Secret Societies in China in the Nineteenth and Twentieth
Centuries. Ann Arbor: University of Michigan Press.
Chin, Ko-lin. 2003. Heijin: Organized Crime, Business, and Politics in Taiwan.
Armonk, N.Y.: M. E. Sharpe.
Choi, Chi-cheung. 1995. “Reinforcing Ethnicity: The Jiao Festival in Cheung
Chau.” In Down to Earth: The Territorial Bond in South China, ed. David
Faure and Helen Siu, 104 –122. Stanford, Calif.: Stanford University
Press
Chou, H., B. Guo, S. Hung, L. Chiao, D. Zhao, and Y. Wu. 2006. “The Taiwan-
Ryukyu Subduction-Collision Complex: Folding of a Viscoelastic Slab
and the Double Seismic Zone.” Journal of Geophysical Research 111.
Chu, Y iu Kong. 2000. The Triads as Business. London: Routledge.
Cohen, Myron. 1976. House United, House Divided: The Chinese Family in Taiwan.
New York: Columbia University Press.
Comber, Leon. 1959. Chinese Secret Societies in Malaya: A Survey of the Triad Soci-
ety from 1800 to 1900. Locust Valley, N.Y.: Published for the Association
for Asian Studies by J. J. Augustin.
Davidson, James Wheeler. 1903. The Island of Formosa. London: Macmillan.
bibliography 257
Davis, Adrian. 2000. “Fraternity and Fratricide in Late Imperial China.” Ameri-
can Historical Review 105 (5): 1630–1640.
Dean, Kenneth. 1993. Taoist Ritual and Popular Cults of Southeast China. Prince
ton, N.J.: Princeton University Press.
DeBernardi, Jean. 1987. “The God of War and the Vagabond Buddha.” Modern
China 13 (3): 310–332.
DeGlopper, Donald. 1995. Lukang: Commerce and Community in a Chinese City.
Albany: State University of New York Press.
De Groot, J. J. M. 1892–1910. The Religious System of China. Leiden: E. J. Brill.
Dell’Orto, Alessandro. 2002. Place and Spirit in Taiwan: Tudi Gong in the Stories,
Strategies and Memories of Everyday Life. Richmond: Curzon.
Dikotter, Frank. 1992. The Discourse of Race in Modern China. London: C. Hurst.
Driessen, Henk. 1983. “Male Sociability and Rituals of Masculinity in Rural
Andalusia.” Anthropological Quarterly 56 (3): 125 –133.
Duara, Prasenjit. 1988a. Culture, Power, and the State: Rural North China, 1900 –
1942. Stanford, Calif.: Stanford University Press.
———. 1988b. “Superscribing Symbols: The Myth of Guandi, Chinese God of
War.” Journal of Asian Studies 47 (4): 778 –795.
———. 1999. “Local Worlds: The Poetics and Politics of the Native Place in
Modern China.” In Huang and Hsu, Imagining China, 161–200.
Durkheim, Emile. 1965. The Elementary Forms of the Religious Life. New York:
Free Press.
Elliott, Alan J. A. 1955. Chinese Spirit-Medium Cults in Singapore. London: Royal
Anthropological Institute.
Elman, Benjamin. 1991. “Social, Political, and Cultural Reproduction via Civil
Service Examinations in Late Imperial China.” Journal of Asian Studies
50 (1): 7–28.
Elvin, Mark. 1985. “Between the Earth and Heaven: Conceptions of the Self in
China.” In The Category of the Person: Anthropology, Philosophy, History, ed.
Michael Carrithers et al., 156 –189. Cambridge: Cambridge University
Press.
Esherick, Joseph. 1987. The Origins of the Boxer Uprising. Berkeley: University of
California Press.
Evans-Pritchard, E. E. 1953. “Nuer Spear Symbolism.” Anthropological Quarterly
26 (1): 1–19.
Fang Jian’neng 方建能, and Yu Bingsheng 余炳盛. 1998. Taiwan zhongyang
shanmai jinkuangchuang. 臺灣中央山脈金礦床 (Gold Seams in Taiwan’s
Central Mountains). Taipei: Taiwan Shengli Bowuguan 台灣省立博物館
(Taiwan Provincial Museum).
Feuchtwang, Stephan. 1976. “School-Temple and City God.” In The City in
Late Imperial China, ed. G. William Skinner, 581–608. Stanford, Calif.:
Stanford University Press.
———. 1992. The Imperial Metaphor: Popular Religion in China. London:
Routledge.
Frantzis, Bruce Kumar. 1998. The Power of Internal Martial Arts: Combat Secrets of
Ba Gua, Tai Chi, and Hsing-i. Berkeley, Calif.: North Atlantic Books.
Freedman, Maurice. 1960. “Immigrants and Associations: Chinese in Nine-
258 bibliography
teenth-Century Singapore.” Comparative Studies in Society and History 3

(1): 25 – 48.
Fried, Morton H. 1952. “Military Status in Chinese Society.” American Journal
of Sociology 57 (4): 347–357.
Gallant, Thomas W. 2000. “Honor, Masculinity, and Ritual Knife Fighting in
Nineteenth-Century Greece.” American Historical Review 105 (2): xvi,
359–382.
Gallin, Bernard. 1966. Hsin Hsing, Taiwan: A Chinese Village in Change. Berke-
ley: University of California Press.
Gallin, Bernard, and Rita Gallin. 1977. “Sociopolitical Power and Sworn
Brother Groups in Chinese Society: A Taiwanese Case.” In The Anthropol-
ogy of Power: Ethnographic Studies from Asia, Oceania, and the New World, ed.
R. N. Adams and R. D. Fogelson. New York: Academic Press.
Gao Chongdao 高崇道. 1986. Zhong’guo wuxue cidian. 中國武學辭典 (Diction-
ary of Chinese Military Studies). Taipei: Changchun Shufang 長春書房.
Gao Shufan 高樹藩, ed. 1984. Zhengzhong xingyinyi zonghe zidian 正中形音義
總合字典 (Comprehensive Etymological Dictionary of Chinese Charac-
ters, 2nd Taiwan edition). Taipei: Zhengzhong Shujü.
Gates, Hill. 1987. Chinese Working-Class Lives: Getting by in Taiwan. Ithaca, N.Y.:
Cornell University Press.
Geertz, Clifford. 1973. The Interpretation of Cultures: Selected Essays. New York:
Basic Books.
Gilmore, David D. 1987. Aggression and Community: Paradoxes of Andalusian Cul-
ture. New Haven: Yale University Press.
———. 1990. Manhood in the Making: Cultural Concepts of Masculinity. New
Haven: Yale University Press.
Girard, René. 1977. Violence and the Sacred. Baltimore: Johns Hopkins Univer-
sity Press.
Goffman, Erving. 1959. The Presentation of Self in Everyday Life. Garden City,
N.Y.: Doubleday.
Gold, Thomas, Doug Guthrie, et al., eds. 2002. Social Connections in China:
Institutions, Culture, and the Changing Nature of Guanxi. Cambridge: Cam-
bridge University Press.
Gong Pengcheng 龔鵬程. 1987. Daxia. 大俠 (The Great Knight-Errant). Tai-
pei: Jinguan Chubanshe 錦冠出版社.
Goodman, David S. G. 2002. “Structuring Local Identity: Nation, Province
and County in Shanxi during the 1990s.” China Quarterly 172:837–862.
Graff, David A. 2002. Medieval Chinese Warfare, 300–900. London: Routledge.
Granet, Marcel. 1975. The Religion of the Chinese People. New York: Harper and
Row.
Guo Panxi 郭泮溪. 1989. Zhong’guo yinjiu xisu. 中国饮酒習俗 (Chinese Drink-
ing Customs). Xi’an: Shaanxi Renmin Chubanshe 陝西人民出版社
(Shaanxi People’s Publishers).
Gutmann, Matthew. 1996. The Meanings of Macho: Being a Man in Mexico City.
Berkeley: University of California Press.
Habermas, Jürgen. 1984. The Theory of Communicative Action. Boston: Beacon
Press.
bibliography 259
Hansson, Anders. 1996. Chinese Outcasts: Discrimination and Emancipation in

Late Imperial China. Leiden: E. J. Brill.
Harrell, Stevan. 2000. “The Role of the Periphery in Chinese Nationalism.” In
Huang and Hsu, Imagining China, 133 –160.
Harrell, Stevan, and Junjie Huang. 1994. Cultural Change in Postwar Taiwan.
Boulder, Colo.: Westview Press.
Hartman, Charles. 1995. “Stomping Songs: Word and Image.” Chinese Litera-
ture: Essays, Articles, Reviews (CLEAR) 17:1– 49.
Hartwell, Robert M. 1982. “Demographic, Political, and Social Transforma-
tions of China, 750–1550.” Harvard Journal of Asiatic Studies 42 (2):
365 – 442.
Heald, Suzette. 1999. Manhood and Morality: Violence and Ritual in Gisu Society.
London: Routledge.
Henning, Stanley. 1999. “Academia Encounters the Chinese Martial Arts.”
China Review International 6 (2): 319–332.
Hershatter, Gail. 1997. Dangerous Pleasures: Prostitution and Modernity in Twenti-
eth-Century Shanghai. Berkeley: University of California Press.
Herzfeld, Michael. 1985. The Poetics of Manhood: Contest and Identity in a Cretan
Mountain Village. Princeton, N.J.: Princeton University Press.
Ho, Ping-ti. 1962. The Ladder of Success in Imperial China: Aspects of Social Mobil-
ity, 1368–1911. New York: Columbia University Press.
Honig, Emily, and Gail Hershatter. 1988. Personal Voices: Chinese Women in the
1980’s. Stanford, Calif.: Stanford University Press.
Hsing, You-tien. 1997. “Building Guanxi across the Straits: Taiwanese Capital
and Local Chinese Bureaucrats.” In Underground Empires: The Cultural
Politics of Modern Chinese Transnationalism, ed. Aihwa Ong and Donald
Nonini, 143–164. New York: Routledge.
Hsu, Francis L. K. 1949. Under the Ancestors’ Shadow. London: Lund Hum
phries.
Hsu Wen-hsiung. 1980. “From Aboriginal Island to Chinese Frontier: The
Development of Taiwan before 1683.” In China’s Island Frontier: Studies in
the Historical Geography of Taiwan, ed. Ronald G. Knapp, 3–29. Honolulu:
University of Hawai‘i Press.
Hu Chuan 胡傳. 1952 (1894). Taidong caifang xiuzhice 台東採訪修志冊 (Revised
Gazetteer of Taidong Prefecture). Taidong: Taidong Xian Zhengfu 台東
縣政府 (Taidong County Government).
Huang, Martin W. 2003. “From Caizi to Yingxiong: Imagining Masculinities
in Two Qing Novels, ‘Yesou puyan’ and ‘Sanfen meng quan zhuan’ .”
Chinese Literature: Essays, Articles, Reviews (CLEAR) 25:59 –98.
Huang, Shu-min. 1998. The Spiral Road: Change in a Chinese Village through the
Eyes of a Communist Party Leader. Boulder, Colo.: Westview Press.
Huang, Shu-min, and Cheng-kuang Hsu. 2000. Imagining China: Regional Divi-
sion and National Unity. Taipei: Institute of Ethnology, Academia Sinica.
Huang Wenbo 黃文博. 1997. “Handan Ye chuxun, taidongren zhapao” 寒單
爺出巡，台東人炸炮 (Handan Ye Goes on Patrol, Taidong’ers Blast with
Firecrackers). Xiangcheng shenghuo zazhi 鄉城生活雜誌 (Town and Coun-
try Life Magazine) 37:33 – 40.
260 bibliography
Huang Xuanxing 黃炫星. 1997. Taiwan de gudao. 台灣的古道 (Old Roads of

Taiwan). Taizhong: Taiwan Sheng Zhengfu Xinwenchu 台灣省政府新聞
處 (Taiwan Provincial Government Information Office).
Hyde, Lewis. 1983. The Gift: Imagination and the Erotic Life of Property. New York:
Vintage Books.
Ivy, Marilyn. 1995. Discourses of the Vanishing: Modernity, Phantasm, Japan. Chi-
cago: University of Chicago Press.
Japan, Provisional Bureau of Census Investigation of Formosa. 1909. The Special
Population Census of Formosa, 1905. Tokyo: Imperial Printing Bureau.
Jefferson, Tony 1994. “Theorising Masculine Subjectivity.” In Just Boys Doing
Business? Men, Masculinities and Crime, ed. Tim Newburn and Elizabeth
A. Stanko, 10 –31. New York: Routledge.
Jordan, David K. 1972. Gods, Ghosts, and Ancestors: The Folk Religion of a Taiwan-
ese Village. Berkeley: University of California Press.
Jordan, David K., and Daniel L. Overmyer. 1986. The Flying Phoenix: Aspects of Chi-
nese Sectarianism in Taiwan. Princeton, N.J.: Princeton University Press.
Kaneko, Erika, and Shigeru Tsuchida. 1982. “Puyuma Ethnohistory and Lin-
guistics: A Review Article.” Asian Folklore Studies 41 (1): 109–117.
Kaplan, David E., and Alec Dubro. 1986. Yakuza: The Explosive Account of Japan’s
Criminal Underworld. Reading, Mass.: Addison-Wesley.
Kataoka, Iwao 片岡嚴. 1981 (1921). Taiwan fengsu zhi. 台灣風俗志 (Record of
the Customs of Taiwan). Taipei: Dali Chubanshe 大立出版社.
Katz, Paul R. 1987. “Demons or Deities? The ‘Wangye’ of Taiwan.” Asian Folk-
lore Studies 46 (2): 197–215.
———. 1995. Demon Hordes and Burning Boats: The Cult of Marshal Wen in Late
Imperial Chekiang. Albany: State University of New York Press.
Kipnis, Andrew B. 1997. Producing Guanxi: Sentiment, Self, and Subculture in a
North China Village. Durham, N.C.: Duke University Press.
Kleinman, Arthur. 1980. Patients and Healers in the Context of Culture: An Explora-
tion of the Borderland between Anthropology, Medicine, and Psychiatry. Berke-
ley: University of California Press.
Knapp, Ronald G. 1980. “Settlement and Frontier Land Tenure.” In China’s
Island Frontier: The Historical Geography of Taiwan, ed. Ronald G. Knapp,
55–68. Honolulu: University of Hawai‘i Press.
Kuhn, Philip A. 1970. Rebellion and Its Enemies in Late Imperial China: Militariza-
tion and Social Structure, 1796 –1864. Cambridge, Mass.: Harvard Univer-
sity Press.
Lacan, Jacques. 1977. Écrits: A Selection. New York: Norton.
———. 1978. The Four Fundamental Concepts of Psychoanalysis. New York: Nor
ton.
Lagerwey, John. 1987. Taoist Ritual in Chinese Society and History. New York:
Collier Macmillan.
Lancaster, Roger. 1994. Life Is Hard: Machismo, Danger, and the Intimacy of Power
in Nicaragua. Berkeley: University of California Press.
Lewis, Mark Edward. 1990. Sanctioned Violence in Early China. Albany: State Uni-
versity of New York Press.
Li Yüanbing 李元丙. 1953. “Yongchang fu Baoshan xian Han-Hui hudou ji
bibliography 261
Du Wenxiu shixing geming zhi yüanqi” (Han-Hui Armed Feuding in

Baoshan County, Yongchang Prefecture, and the Origins of Du Wen
xiu’s Revolutionary Practice). In 白壽彝 Huimin Qiyi 回民起義 (Hui
Uprisings), ed. Bai Shouyi. Shanghai: Shenzhou Guoguangshe 神州國
光社.
Li Zailuan 李載鸞. 1980. Fujian hequan miyao 福建鶴拳秘要 (Essential Secrets
of Fujian Crane Boxing). Taipei: Hualian Chubanshe 華聯出版社.
Lin Ching-Yuan 林慶元. 1992. “Shijiu shiji taiwan dongbu diqu de kaifa”
(The Reclamation of Eastern Taiwan during the Nineteenth Century).
Danjiang shixue 淡江史學 (Tamkang University Journal of History):
4:141–172.
Lin Meirong et al. 林美容. 1997. Zhanghua xian quguan yu wuguan 彰化縣曲
館與武館. (Music and Martial Arts Associations of Zhanghua County).
Zhanghua, Taiwan: Zhanghua Xianli Wenhua Zhongxin 彰化縣立文化
中心 (Zhanghua County Cultural Center).
Lintner, Bertil. 2002. Blood Brothers: Crime, Business and Politics in Asia. Crows
Nest, NSW, Australia: Allen and Unwin.
Lipman, Jonathan Neaman, and Stevan Harrell. 1990. Violence in China: Essays
in Culture and Counterculture. Albany: State University of New York Press.
Liu, James J. Y. 1967. The Chinese Knight-Errant. London: Routledge and Kegan
Paul.
Liu Kexiang 劉克襄. 1992. Houshan tanxian: Shijiu shiji waiguoren zai Taiwan
donghai’an de luxing. 後山探險: 十九世紀外國人在台灣東海岸的旅行
(Adventure in the Mountains Beyond: Foreigners’ Travels on the East
Coast of Taiwan in the Nineteenth Century). Taipei: Zili Wanbao Wen-
hua Chubanshe 自立晚報文化出版社 (Independence Evening Post Cul-
ture Press).
Liu Yanglei 劉揚磊. 2004. “Fei duhuixing bangpai zuhe yu zuzhi fanzui xian
xiang zhi yanjiu: Yi Yunlinxian wei li” 非都會型幫派組合與組織犯罪現
象之研究: 以雲林縣為例 (Research on the Criminalogical Phenomenon
of the Organization and Structure of Non-Urban-Type Gangs: The Case
of Yunlin County). Master’s thesis, National (Taiwan) Central Police
University.
Liu Zhiwan 劉枝萬. 1983. Taiwan minjian xinyang lunji 臺灣民間信仰論集 (Col-
lected Essays on Taiwanese Popular Religion). Taipei: Lianjing Chuban
Shiye Gongsi 聯經出版事業公司.
Louie, Kam. 2002. Theorizing Chinese Masculinity. Cambridge: Cambridge Uni-
versity Press.
Lü Jiangming 呂江銘. 2002. Jiajiang 家將 (Military Retainers). Taipei: Luse
Luxing Wenjiao Jijinhui 綠色旅行文教基金會 (Green Travel Education
Foundation).
Lu Xixing 陸西星 (attrib). 1990. Fengshen Yanyi 封神演藝. Taipei: Wenhua
Tushu 文化圖書公司.
Mann, Susan. 1997. Precious Records: Women in China’s Long Eighteenth Century.
Stanford, Calif.: Stanford University Press.
———. 2000. “The Male Bond in Chinese History and Culture.” American His-
torical Review 105 (5): 1600 –1614.
262 bibliography
Martin, Jeffrey. 2006. “Keeping the Peace in Regime Change: Police Work in
Taiwan.” Ph.D. dissertation, University of Chicago.
Matisoff, James A. 2001. “On the Genetic Position of Bai within Tibeto-Bur-
man.” Paper presented at the 34th International Conference on Sino-
Tibetan Languages and Linguistics, Yunnan Nationalities Institute,
Kunming. Available at http://stedt.berkeley.edu/pdf/JAM/Bai.pdf.
McCreery, John L. 1990. “Why Don’t We See Some Real Money Here? Offer-
ings in Chinese Religion.” Journal of Chinese Religions 18:1–24.
———. 1995. “Negotiating with Demons: The Uses of Magical Language.”
American Ethnologist 22 (1): 144 –164.
McIsaac, Lee. 2000. “ ‘Righteous Fraternities’ and Honorable Men: Sworn
Brotherhoods in Wartime Chongqing.” American Historical Review 105
(5): 1641–1655.
Meng Xianghan 孟祥瀚. 1997. Taidong xianshi kaituo pian 台東縣史開拓篇 (His-
torical Gazetteer of Taidong County: The Pioneer Era). Taidong, Taiwan:
Taidong Xian Zhengfu 台東縣政府 (Taidong County Government).
———. 2002. “Qingdai Cheng’guang’ao de Tuoken yu Fazhan” 清代成廣
澳的拓墾與發展 (Settlement and Development of Cheng’guang’ao
[Cheng’gong] in the Qing Period). Xingda renwen xuebao 興大人文學報
(Chung Hsing University Humanities Journal) 32:879–918.
Meskill, Johanna. 1979. A Chinese Pioneer Family: The Lins of Wu-Feng, Taiwan,
1729–1895. Princeton, N.J.: Princeton University Press.
Murray, Dian H., and Qin Baoqi. 1994. The Origins of the Tiandihui: The Chinese
Triads in Legend and History. Stanford, Calif.: Stanford University Press.
Myers, John T. 1975. “A Hong Kong Spirit-Medium Temple.” Journal of the
Hong Kong Branch of the Royal Asiatic Society 15:16 –27.
Naquin, Susan. 1976. Millenarian Rebellion in China: The Eight Trigrams Uprising
of 1813. New Haven, Conn.: Yale University Press.
Notar, Beth E. 2006. Displacing Desire: Travel and Popular Culture in China. Hono
lulu: University of Hawai‘i Press.
Obeyesekere, Gananath. 1981. Medusa’s Hair: An Essay on Personal Symbols and
Religious Experience. Chicago: University of Chicago Press.
———. 1990. The Work of Culture: Symbolic Transformation in Psychoanalysis and
Anthropology. Chicago: University of Chicago Press.
Ownby, David. 1995. “The Heaven and Earth Society as Popular Religion.”
Journal of Asian Studies 54 (4): 1023 –1046.
———. 1996. Brotherhoods and Secret Societies in Early and Mid-Qing China: The
Formation of a Tradition. Stanford, Calif.: Stanford University Press.
Oxfeld, Ellen. 1992. “Individualism, Holism, and the Market Mentality: Notes
on the Recollections of a Chinese Entrepreneur.” Cultural Anthropology
7 (3): 267–300.
Pearson, Richard J., and Anne Underhill. 1987. “The Chinese Neolithic:
Recent Trends.” American Anthropologist 89 (4): 807–822
Perry, Elizabeth J. 1980. Rebels and Revolutionaries in North China, 1845–1945.
Stanford, Calif.: Stanford University Press.
———. 1983. “Social Banditry Revisited: The Case of Bai Lang, a Chinese
Brigand.” Modern China 9 (3): 355 –382.
bibliography 263
Pyke, Karen D. 1996. “Class-Based Masculinities: The Interdependence of Gen-

der, Class, and Interpersonal Power.” Gender and Society 10 (5): 527–549.
Qiu Dezai 仇德哉. 1985. Taiwan miaoshen daquan. 台灣廟神大全 (Complete
Compendium of the Temples and Gods of Taiwan). Taipei: Qiu Dezai
仇德哉.
Reed, Bradley W. 2000. Talons and Teeth: County Clerks and Runners in the Qing
Dynasty. Stanford, Calif.: Stanford University Press.
Reed, Carrie E. 2000. “Tattoo in Early China.” Journal of the American Oriental
Society 120 (3): 360–376.
Rosaldo, Michelle Z. 1983. “The Shame of Headhunters and the Autonomy of
Self.” Ethos 11 (3): 135 –151.
Rouget, Gilbert. 1985. Music and Trance: A Theory of the Relations between Music
and Possession. Chicago: University of Chicago Press.
Ruhlen, Merritt. 1987. A Guide to the World’s Languages. Stanford, Calif.: Stan-
ford University Press.
Ruhlmann, Robert. 1960. “Traditional Heroes in Chinese Popular Fiction.” In
The Confucian Persuasion, ed. Arthur F. Wright, 141–176. Stanford, Calif.:
Stanford University Press
Sangren, P. Steven. 1983. “Female Gender in Chinese Religious Symbols: Kuan
Yin, Ma Tsu, and the ‘Eternal Mother.’ ” Signs 9 (1): 4 –25.
———. 1984. “Traditional Chinese Corporations: Beyond Kinship.” Journal of
Asian Studies 43 (3): 391–415.
———. 1987. History and Magical Power in a Chinese Community. Stanford, Calif.:
———. 2000. Chinese Sociologics: An Anthropological Account of the Role of Alien-
ation in Social Reproduction. London: Athlone.
Saso, Michael R. 1978. The Teachings of Taoist Master Chuang. New Haven,
Conn.: Yale University Press.
Schak, David C. N.d. “Drinking, Establishing Trust, and Doing Business in
China and Taiwan.”
Scheff, Thomas J. 1977. “The Distancing of Emotion in Ritual.” Current Anthro-
pology 18 (3): 483 –505.
Schein, Louisa. 2000. Minority Rules: The Miao and the Feminine in China’s Cul-
tural Politics. Durham, N.C.: Duke University Press
Schieffelin, Edward L. 1976. The Sorrow of the Lonely and the Burning of the Danc-
ers. New York: St. Martin’s Press.
Schipper, Kristoffer. M. 1993. The Taoist Body. Berkeley: University of Califor-
nia Press.
Schneider, Friedrich, and Dominik H. Enste. 2000. “Shadow Economies:
Size, Causes, and Consequences.” Journal of Economic Literature 38 (1):
77–114.
Seaman, Gary. 1978. Temple Organization in a Chinese Village. Taipei: Chinese
Association for Folklore.
———. 1981. “The Sexual Politics of Karmic Retribution.” In Anthropology of
Taiwanese Society, ed. E. Martin and H. Gates, 381–396. Stanford, Calif.:
264 bibliography
Selden, Mark. 1995. “ Yan’an Communism Reconsidered.” Modern China 21

(1): 8 – 44.
Shahar, Meir. 2001. “Ming Period Evidence of Shaolin Martial Practice.” Har-
vard Journal of Asiatic Studies 61 (2): 359–413.
———. 2008. The Shaolin Monastery: History, Religion, and the Chinese Martial
Arts. Honolulu: University of Hawai‘i Press.
Shahar, Meir, and Robert P. Weller. 1996. Unruly Gods: Divinity and Society in
China. Honolulu: University of Hawai‘i Press.
Shepherd, John Robert. 1993. Statecraft and Political Economy on the Taiwan
Frontier, 1600 –1800. Stanford, Calif.: Stanford University Press.
Shi Nai’an 施耐庵. 1972. Shuihu zhuan 水滸傳 (Water Margin). Taipei: San-
min Shuju 三民書局.
Shi Wanshou 史萬壽. 1983. “Jiajiang tuan: Tianren heyi de xunbu zuzhi” 家將
團－天人合一的巡捕組織 (Military Retainers Troupes: A Transcenden-
tal Police Organization). Tainan wenhua 台南文化 22:47– 65.
Siu, Helen F. 1989. “Recycling Rituals.” In Unofficial China: Popular Culture and
Thought in the People’s Republic, ed. Perry Link, 121–137. Boulder, Colo.:
Westview Press.
Sivin, Nathan. 1995. Medicine, Philosophy, and Religion in Ancient China: Researches
and Reflections. Aldershot, Hants: Variorum.
Skinner, G. William. 1958. Leadership and Power in the Chinese Community of
Thailand. Ithaca, N.Y.: Published for the Association for Asian Studies by
Cornell University Press.
———. 1974. Marketing and Social Structure in Rural China. Ann Arbor: Associa-
tion for Asian Studies.
Smith, Robert W. 1964. Secrets of Shaolin Temple Boxing. Tokyo: Charles E.
Tuttle
———. 1974. Chinese Boxing: Masters and Methods. Tokyo: Kodansha Interna-
tional.
Spiro, Melford E. 1996. “Narcissus in Asia.” Ethos 24 (1): 165 –191.
Stafford, Charles. 2000. Separation and Reunion in Modern China. Cambridge:
Cambridge University Press.
Strickmann, Michel, and Bernard Faure. 2002. Chinese Magical Medicine. Stan-
ford, Calif.: Stanford University Press.
Sutton, Donald S. 1989. “Ritual Trance and the Social Order: The Persistence
of Taiwanese Shamanism.” In Social History and Issues in Human Con-
sciousness, ed. Andrew E. Barnes and Peter N. Stearns, 108–127. New
York: New York University Press.
———. 1990. “Ritual Drama and Moral Order: Interpreting the Gods’ Festival
Troupes of Southern Taiwan.” Journal of Asian Studies 49 (3): 535–554.
———. 2003. Steps of Perfection: Exorcistic Performers and Chinese Religion in Twen-
tieth-Century Taiwan. Cambridge, Mass.: Harvard University Asia Center.
Suzuki Seiichiro 鈴木清一. 1978. Taiwan jiuguan xisu xinyang 臺灣舊貫習俗
信仰 (Old Customs, Practices, and Beliefs of Taiwan). Taipei: Zhongwen
Tushu Gongsi 眾文圖書公司.
Szonyi, Michael. 1997. “The Illusion of Standardizing the Gods: The Cult of
bibliography 265
the Five Emperors in Late Imperial China.” Journal of Asian Studies 56

(1): 113–135.
Taidong Xian Houshan Wenhua Gongzuo Xiehui 台東縣後山文化工作協會.
1996. Taidong xian simiao zhuanji 台東縣寺廟專輯 (Compendium of the
Temples of Taidong County). Taidong: Taidong Xianli Wenhua Zhong
xin 台東縣立文化中心 (Taidong County Cultural Center).
Taidong Xian Wenxian Weiyuanhui 台東縣文獻委員會 (Taidong County
Archive Committee). 1964a. Jinbu zhong de Taidong 進步中的台東 (Pro-
gressive Taidong). Taidong: Taidong Xian Wenxian Weiyuanhui, 台東縣
文獻委員會.
———. 1964b. Taidong xianzhi 台東縣志 (Gazetteer of Taidong County). Tai
dong: Taidong Xian Zhengfu 台東縣政府 (Taidong County Govern-
ment Office).
Taidong Xian Zhengfu 台東縣政府 (Taiwan County Government Office).
1983. “Taidong xian zhengti fazhan guan’guang gangyao” 台東縣整體
發展觀光綱要 (Outline Plan for the Full Scale Development of Tourism
in Taidong County). Taidong: Taidong Xian Zhengfu.
Taylor, George. 1888. “A Ramble through Southern Formosa.” China Review
16:137–161.
ter Haar, Barend J. 1998. Ritual and Mythology of the Chinese Triads: Creating an
Identity. Leiden: E. J. Brill.
———. 2000. “Rethinking ‘Violence’ in Chinese Culture” In Meanings of
Violence: A Cross Cultural Perspective, ed. Goran and Jon Abbink Aijmer,
123 –140. Oxford: Berg
Thompson, Laurence G. 1968. “The Junk Passage across the Taiwan Strait: Two
Early Chinese Accounts.” Harvard Journal of Asiatic Studies 28:170 –194.
Turner, Terence S. 1977. “Transformation, Hierarchy, and Transcendence: A
Reformulation of Van Gennep’s Model of the Structure of Rites de Pas-
sage.” In Secular Ritual, ed. Sally F. Moore and Barbara Myerhoff, 53 –70.
Assen (Netherlands): Van Gorcum.
———. 1980. “The Social Skin.” In Not Work Alone, ed. Jeremy Cherfas and
Roger Lewin, 112–140. London: Temple Smith.
———. 1991. “ ‘We Are Parrots,’ ‘Twins Are Birds’: ‘Play of Tropes’ as Opera-
tional Structure.” In Beyond Metaphor: The Theory of Tropes in Anthropology,
ed. James W. Fernandez, 121–158. Stanford, Calif.: Stanford University
Press.
———. 1995a. “Representation, Collaboration, and Mediation in Contempo-
rary Ethnographic and Indigenous Media.” Visual Anthropology Review 11
(2): 102–106.
———. 1995b. “Social Body and Embodied Subject: Bodiliness, Subjectivity,
and Sociality among the Kayapo.” Cultural Anthropology 10 (2): 143–170.
Turner, Victor Witter. 1977. The Ritual Process: Structure and Anti-Structure.
Ithaca, N.Y.: Cornell University Press.
Vogel, Ezra. 1969. “Land Reform in Kwangtung 1951–53: Central Control and
Localism.” China Quarterly 38:27–62.
Von Glahn, Richard. 1991. “The Enchantment of Wealth: The God Wutong
266 bibliography
in the Social History of Jiangnan.” Harvard Journal of Asiatic Studies 51

(2): 651–714.
Wacquant, Loic J. D. 1995. “The Pugilistic Point of View: How Boxers Think
and Feel about Their Trade.” Theory and Society 24 (4): 489–535.
Wang Yali 王雅莉. 2001. “Yue gui? Zhuanwei: Qingshaonian canyu ba jiajiang
tuanti zhi fenxi” 越軌？轉位：青少年參與八家將團體之分析 (Step-
ping out of Line? Dislocation: Analysis of Teenagers’ Participation in
Eight Military Retainers Troupes). Master’s thesis, Social Development
Research Institute, Shih Hsin University 世新大學, Taipei.
Wang Yuanzhi 王元稚. 1959. Jiaxu gongdu chaocun 甲戌公牘鈔存 (Public
Records of the Jiaxu Year). Taipei: Taiwan Yinhang 台灣銀行.
Ward, Barbara E. 1979. “Not Merely Players: Drama, Art and Ritual in Tradi-
tional China.” Man 14 (1): 18 –39.
Ward, J. S. M., and W. G. Stirling. 1925. The Hung Society or the Society of Heaven
and Earth. London: Baskerville Press.
Watson, James. 1985. “Standardizing the Gods: The Promotion of Tian Hou
(‘Empress of Heaven’) along the South China Coast, 960–1960.” In Pop-
ular Culture in Late Imperial China, ed. David Johnson, Andrew J. Nathan,
and Evelyn Rawski, 292–324. Berkeley: University of California Press.
Weller, Robert P. 1987. Unities and Diversities in Chinese Religion. Seattle: Univer-
sity of Washington Press.
Winn, J. K. 1994. “Not by Rule of Law: Mediating State-Society Relations in
Taiwan through the Underground Economy.” In The Other Taiwan, ed.
Murray Rubinstein. Armonk, N.Y.: M. E. Sharpe.
Winslow, Simon. 2001. Badfellas: Crime, Tradition and New Masculinities. Oxford:
Berg.
Wolf, Arthur P. 1974. “Gods, Ghosts, and Ancestors.” In Religion and Ritual in
Chinese Society, ed. Arthur P. Wolf, 131–182. Stanford, Calif.: Stanford
University Press.
Wolf, Arthur P., and Huang Shu-min. 1980. Marriage and Adoption in China,
1845–1945. Stanford, Calif.: Stanford University Press.
Wolf, Margery. 1968. The House of Lim: A Study of a Chinese Farm Family. New
York: Appleton-Century-Crofts.
———. 1972. Women and the Family in Rural Taiwan. Stanford, Calif.: Stanford
University Press.
———. 1985. Revolution Postponed: Women in Contemporary China. Stanford,
Calif.: Stanford University Press
Wu Tengda 吳騰達. 1996. Handan Ye yanjiu. 寒單爺研究 (Research on Handan
Ye). Taidong: Taidong Xianli Wenhua Zhongxin 台東縣立文化中心.
Wu Yih-min et al. 2006. “Seismogenic Structure in a Tectonic Suture Zone:
With New Constraints from 2006 Mw6.1 Taitung Earthquake.” Geophysi-
cal Research Letters 33, L22305.
Yan, Yun-xiang. 1996. The Flow of Gifts: Reciprocity and Social Networks in a Chinese
Village. Stanford, Calif.: Stanford University Press.
Yang, Bin. 2004. “Horses, Silver, and Cowries: Yunnan in Global Perspective.”
Journal of World History 15 (3): 281–322.
bibliography 267
Yang, Mayfair Mei-hui. 1994. Gifts, Favors, and Banquets: The Art of Social Rela-
tionships in China. Ithaca, N.Y.: Cornell University Press.
———. 1999. “From Gender Erasure to Gender Difference: State Feminism,
Consumer Sexuality, and a Feminist Public Sphere in China.” In Women
and the Public Sphere in Transnational China, ed. Mayfair Yang, 35 – 67.
New York: Routledge.
Yang Xiandian, Du Yijian, and Zhang Xilu 杨宪典杜乙简张锡祿 1988. “Dali
Baizu jieri shenghui diaocha” 大理白族节日盛会调查 (Investigations on
Dali Bai Holidays and Festivals). In Baizu shehui lishi diaocha 白族社会历
史调查 (Investigations in Bai Social History), ed. Yunnan Sheng Bianji
Zu 云南省编辑组 (Yunnan Provincial Editorial Office), 136–181. Kun-
ming: Yunnan Renmin Chubanshe 云南人民出版社 (Yunnan People’s
Publishers).
Yang Zhengye 杨政业. 1994. Baizu benzhu wenhua 白族本主文化 (Bai Benzhu
[tutelary deity] Culture). Kunming: Yunnan Renmin Chubanshe 云南
人民出版社 (Yunnan People’s Publishers).
Yun Youke 雲游客. 1988. Jianghu congtan 江湖叢談 (Collected Talks on the
Itinerant Life). Beijing: Zhongguo Quyi.
Zhang Xilu 张锡祿. 1993. “Lun baizu gulao shengdian ‘huoba jie’ ” 论白族古
老盛典 “火把节” (A Discussion of the ‘Torch Festival,’ an Ancient Bai
Ceremony). In Nanzhao Dali shilun wenji 南诏。大理史论文集 (Collected
Essays on the Historiography of the Nanzhao and Dali Kingdoms), ed.
Zhang Xu 张旭, 342–361. Kunming: Yunnan Minzu Chubanshe 云南民
族出版社 (Yunnan Nationalities Publishers).
Zhang Zhenyue 張振岳. 1994. Taiwan houshan fengtu ji 台灣後山風土記
(Record of the Customs of Taiwan’s Eastern Mountain Hinterland). Tai-
pei: Taiyuan Chubanshe 臺原出版社.
Zheng Quanxuan 鄭全玄. 1995. Taidong pingyuan de yimin tuoken yu juluo 台
東平原的移民拓墾與聚落 (Migrant Clearing and Settlement of the Tai
dong Plain). Taidong: Dong Taiwan Yanjiu Hui 東台灣研究會 (Eastern
Taiwan Studies Association).
Zhong, Xueping. 2000. Masculinity Besieged? Issues of Modernity and Male Sub-
jectivity in Chinese Literature of the Late Twentieth Century. Durham, N.C.:
Duke University Press.
Zhonghua Shuju Xianggang Fenju 中華書局香港分局. 1982. (Putonghua yue-
yin) Zhonghua Xin Zidian (普通話粵音) 中華新字典 (Standard Chinese/
Cantonese Pronunciation New Zhonghua Dictionary). Hong Kong:
Zhonghua Shuju Xianggang Fenju 中華書局香港分局.
Zhuang Cuixia 莊翠霞. 2004. “You zou yu xuexiao yu jiajiangtuan zhijian de
qingshaonian: Yi dongbu guozhong xuesheng wei li” 遊走於學校與家將
團之間的青少年: 以東部國中學生為例 (Youth Drifting between School
and Jiajiang Troupes: A Case Study of Eastern [Taiwan] Middle School
Students). Master’s thesis, Institute of Local Culture, Hualian Normal
College 花蓮師範學院, 鄉土文化研究所.
Zito, Angela. 1987. “City Gods, Filiality, and Hegemony in Late Imperial
China.” Modern China 13 (3): 333–371.
268 bibliography
———. 1989. “Grand Sacrifice as Text/Performance: Ritual and Writing in

Eighteenth-Century China.” Ph.D. dissertation, University of Chicago.
———. 1996. “City Gods and Their Magistrates.” In Religions of China in Prac-
tice, ed. Donald S. Lopez, Jr., 72–81. Princeton, N.J.: Princeton Univer-
sity Press.
———. 1997. Of Body and Brush: Grand Sacrifice as Text/Performance in Eigh-
teenth-Century China. Chicago: University of Chicago Press.
Žižek, Slavoj. 1989. The Sublime Object of Ideology. New York: Verso Books.
Index
aggression: human, 11, 43, 73, 107, 119, 120, 123, 166, 216n18. See also
246n4; male, 10, 108, 109, 174, 207. secret societies
See also violence brothers of the dark path, 14, 31, 104,
Ah-chung (film), 126 116, 119. See also heidao
Allison, Anne, 201–202 Buddhism: and popular religion, 43
anthropological study of masculinity, Buddhist monks: and martial arts, 2,
50, 117 54, 161, 213n2, 222n44; rituals, 88,
anti-hooligan law (Taiwan), 117, 94, 216n2
218n15, 222n46, 234n4
Austronesian peoples, 61, 67 caizi, 47– 49, 221n34, 221n36
Cangshan Range ( Yunnan), 6, 87, 168
Bai people, 87–88; ethnic identity, carousing, male, 19, 20, 109, 154,
91–92, 113; religion, 88, 90–91, 106, 176–203; and Japanese corporate
109 masculinity, 201–202; and patriar-
Bamboo Union League, 81, 118, chy, 202–203; as social reproduc-
235n9 tion, 202
Baosang (Posong), 80, 227n23, 230n39 Celestial Master sect, 215n11
Baoshan ( Yunnan), 12 Central Mountains (Taiwan), 6, 63, 65,
Beinan, 66, 70, 73, 155, 227n23; 83, 225n14
Commandery, 68, 227n24; Cul- Chang Tso-chi (director), 126
ture, 226n18; Plain, 65, 67, 69, 76, charisma: male, 28, 46, 98
228n26; Stream, 63, 80, 225n16; Chen Kaige (film director), 52
yamen, 69 chhit-tho, 16, 178
benzhu (Bai local patron deities), 90, Chinese opera, 52, 122, 133; martial
113, 232n6 opera (wuju), 53
betel: and carousing, 178; chewing, Chinese patrilineal family, 11, 15, 28,
30, 123, 133, 149, 152, 186, 189; as 88, 174, 204; and native place, 83;
offering to gods and ghosts, 3; and women’s position in, 49
role in social custom, 104, 203 Chow Yun-fat, 36, 51, 222n42
Blasting Handan Ye (ritual), 18, 100– Christianity: and millenarian sects, 9,
106, 109–111, 114, 232n9, 233n15 215n12; in post-Mao China, 241n41
Bordieu, Pierre, 208 City God, 138, 168
Boxer Rebellion, 220n28 civil (wen), 5, 52–53; and military (wu),
brotherhoods, 8, 11, 12, 49–50, 54, 56, 23, 40
269
270 index
Civil Magistrate (Military Retainers), ethnography, 6, 209. See also partici-

30, 134, 152 pant- observation fieldwork
civil service examination (keju), 23, exorcism, 7, 17, 26; conducted by
46 Military Retainers, 133, 146,
Confucianism, 46, 221n31 238n30; Daoist, 105; logic of, 119;
Confucius, 46; and martial arts, military symbols in, 10, 26, 53, 122
222n39
Creation of the Gods (novel), 5. See also face (mianzi ), 108, 128, 145, 148, 152;
Fengshen Yanyi and jianghu masculinity, 36, 144;
cultural production, 18, 20, 41, 59 and violence, 73, 107
Cultural Revolution, 98, 111, 112, Fengshen Yanyi (novel), 100, 240n37
216n1, 240n39 fengshui, 61, 65, 223n5
fertility: community, 9, 18; female, 92;
Dajiale (1980s Taiwan illegal numbers male, 96, 99, 107; and violence, 96
racket), 144, 238n29, 239n34 fighting arts, Chinese, 2, 48, 53,
Dali, 6, 59, 61, 87–90, 216n16, 231n1, 213n2. See also martial arts,
240n39; and Bai Torch Festival, 18, Chinese
90 –99, 160, 240n40, 241n44 filial piety, 37, 151, 153, 206; and
Daoism, 2, 23, 43, 112, 219n21; and Chinese male identity, 96, 245n1;
martial arts, 216n1 and patriarchy, 29, 104, 204, 206,
Daoist: exorcism, 105; immortals, 54, 218n14. See also loyalty
101; pantheon, 21, 52, 101, 241n42; fire: and fertility, 90 –92, 109; and
priests, 25, 43, 88; ritual, 42, 53, transformation of qi, 105; and yang
213n2, 217n11, 223n4 (male) energy, 10
“dark path,” 33, 39, 57, 121, 152, 166, frame puppets (tai-sin ang-a), 122
174, 185, 219n17. See also heidao Fujian (province), 62, 67, 112, 119,
Dazhuang Incident (Uprising), 69, 122, 218n15, 229n31, 236n13,
227n24 237n16, 239n32, 240n40
dragon dance, 97, 160
drinking: and carousing, 19, 176 –177; gambling, 151, 155, 175, 178, 191,
and Chinese masculine identity, 216n16
19 –20, 146, 174, 176 –178; military gangs, 11, 34, 54, 116, 119 –123,
terminology in, 200; prowess, 97, 220n24; and gangsters in Taidong,
154; and reciprocity, 132 81–82, 236n12; mainlander (Tai-
Dutch: accounts of Taidong, 231n1; wan), 117, 235n8. See also brother-
expeditions to southeastern Taiwan, hoods; secret societies
65 – 67, 225n16; in Taiwan, 62 gangster god (Handan Ye), 100–101
gangster subculture, 103 –106
Earth God, 21; shrines, 142, 156, Gelaohui, 12, 216n15
230n40, 239n34. See also Tudi ghosts (gui ), 25 –26; marginal men
Gong compared to, 120, 215n14;
Eastern Longitudinal Valley (Taiwan), redeemed, 44
63, 70, 225n14 globalization: and masculinity in
entrepreneurs: in the jianghu, 179, Chinese popular culture, 51,
219n17; reform-era PRC, 56 –57, 111–112; and ritual change, 19
170, 175; in Taidong, 72, 79, 85 Golden Lotus (Pan Jinlian), 49
Erhai Lake, 6, 87, 91, 160, 168 Great Leap Forward, 113, 216n1
index 271
Gu Long (author), 33 karaoke, 180 –185, 200 –202

Guan Gong (god), 39 Killer, The (movie), 36
guanxi, 177, 242n3 knight-errant, Chinese, 2, 37, 54, 148,
Guanyin, bodhisattva, 94, 168 166, 219n18, 220n25
Handan Ye, 100 –104; and cultural Lantern Festival ( Yuanxiao), 29;

tourism in Taidong, 105 –106, Taidong temple procession, 80,
111–114; iconography of, 101, 104. 99–100, 115, 135–136, 233n19
See also Blasting Handan Ye Lee, Bruce, 51–52, 222n42
haohan, 48, 52, 221n34. See also ying ling, 10, 28 –29, 41, 44 – 45, 60, 81,
xiong haohan 223n3
Heaven and Earth Society (Tiandihui), lion dance, 55, 236n13
235n5 liumang, 15, 218n15. See also lomoa
heidao, 14, 39. See also “dark path” lomoa, 31, 73, 120, 218n15
Hong Kong: cinema, 51; jianghu life Lord Xing (patron deity of Loyal
in, 235n11; popular religion, Harmony Temple), 153 –154,
216n1; temple processions, 239n32
241n44; triads, 235n10 Lotus Pond Society (Lianchi Hui), 94
honor, 35–39, 98, 116, 144 –145, 152, Louie, Kam, 51
182, 198, 199, 205–206. See also Loyal Harmony Temple (Taidong),
yiqi 29–31, 134, 139, 142, 148 –149, 194,
houshan, 62–63, 85, 224n11 239n32
Hsu, Francis L. K., 232n7 loyalty, 36, 38, 48, 144, 146, 180, 206
Hu Chuan, 228n25
Hui (Chinese Muslims), 88, 90 male identity, Chinese, 13, 50, 71;
hui (voluntary association), 35, 234n5 and violence, 98. See also male
Huwei, 72, 198 subjectivity
male subjectivity, Chinese, 15, 108,
imperial metaphor, 10, 27, 40 182, 204, 206
intercorporeality, 210 male vitality, 108 –109. See also
Investiture of the Gods (novel). See Feng- masculine vitality
shen Yanyi man of prowess, 37, 98, 106, 144, 201
Mandate of Heaven, 45 – 46
Jade Emperor, 104, 153, 158 manhood, Chinese, 13 –14, 28, 37, 50,
Japanese Expeditionary Force (Taiwan, 76, 152, 174, 182, 199, 218n12
1874), 226n20 Maoshan Daoism, 43, 123
Japanese occupation (of Taiwan), 69, martial (wu), 6, 10, 16, 20, 40 – 43, 53,
74, 80, 224n7 220n26
jianghu, 2, 15 –16, 29–35, 116 –117; martial arts, Chinese, 1– 6, 17, 54, 74;
carousing in, 179–182; and martial and Chinese popular religion, 5 – 6,
ritual, 118 –120 53, 122, 129, 146, 150
jiaotou, 81, 117, 121 martial arts, Okinawan, 1
Jin Yong (author), 33 martial deities, 14, 53, 159
Journal of Asian Martial Arts, 213n1 martial ritual procession troupes, 30,
Journey to the West (novel), 5 122, 132
juancun (military dependents’ masculine vitality, 107
villages), 117 masculinity, “crisis” of Chinese, 51
272 index
Mazu, 26, 115, 131, 213n5 root-seeking literary movement

migration to Taidong, second wave, (xungen), 52
70 –71, 82 Ryukyu Archipelago, 62
military (wu), 5, 12, 23; deities, 13,
175; and logic of exorcism, 11 sacrifice, 9, 105, 108, 215n13; blood,
Military God of Wealth, 100, 151, 168 10, 42, 49
Military Retainers, 29 –30, 123, Sangren, P. Steven, 28
127–139, 149, 153, 194, 237n19 secret societies, 12, 15, 38 – 40,
millenarian sects, 8, 215n12 117–119; initiation rituals of, 117,
mingpai (winning lottery number), 120
157 self-mortification, 42, 123, 127, 131
Mudan Incident (1871), 67 sexual prowess, male, 18, 55
shadow economy: Chinese, 174 –175;
Nanzhao (kingdom), 87, 231n1 Taiwan, 31, 80, 84, 216n18
“naruwán,” 84, 85, 231n43 Shanhaijing (Classic of Mountains and
Nezha (deity), 22, 104, 216n17 Seas), 225n12
nuo, 214n8 Shaolin: gongfu, 3; Temple, 213n2
shengdou, 91, 94, 113
Operation Cleansweep ( Yiqing shrimp fishing pool, 195 –196
Zhuan’an), 235n8 Shuihu Zhuan (novel). See Water
Outlaws of the Marsh. See Water Margin
Margin sorcerer’s troupes, 122–123, 126 –127,
Ownby, David, 118 145
Southern Link Highway, 70
participant-observation fieldwork, Southern Silk Road, 87
29–30, 162, 209 Southward March ( Japanese coloniz-
patriarch, 11–13, 28, 36, 206, 218n14 ing strategy), 66, 225n17
patriarchy, 28 –29, 37, 50, 105, 203 spirit generals, 44, 122
Pingpu aborigines, 227n22 spirit mediums, 11, 19, 99, 103, 120,
possession trance, 14 –15, 103, 115, 122, 129, 210
122–123, 127–128, 131–138, spirit soldiers, 7, 25 –26, 115, 116,
158 –159, 179, 210 120; rewarding, 193. See also spirit
generals
qi, 10, 13, 25, 61, 65, 86, 105, 136, 158, subethnic violence, 90, 220n24
223n5 Sun Wukong, 94, 104, 216n17
symbolic violence, 4, 11, 214n7
rebellion: against state authority,
220n28 Taidong, 6, 7, 18, 58, 61, 65–70, 101
religious violence, 8 taijiquan (martial arts style), 54
Rift Valley. See Eastern Longitudinal Taimali, 65
Valley talismans, written ( fulu), 26, 43, 128
righteousness, 35, 38, 39, 48, 162, 180, tangki. See spirit mediums
199 tattoo, 116, 118, 145, 147, 220n24
ritual production, 119 Taylor, George, 227nn21–22
rivers and lakes. See jianghu “tea rooms” (chashi), 191
Romance of the Three Kingdoms (novel). temple processions, 3, 14, 52, 120, 122
See Three Kingdoms Three Kingdoms (novel), 204, 206
index 273
Torch Festival, Bai, 18, 90 –99, 108, wuguan, 121, 236n14

111–114, 160 wulin, 32, 219n18
trance. See possession trance wushu, 222n43
Treaty of Shimonoseki, 69 wuxia. See knight-errant, Chinese
triads. See secret societies wuzhen. See martial ritual procession
Tudi Gong, 155, 239n34 troupes
village militias, 121, 215n14, 220n24 xiagang (laid-off workers), 175

violence: fertility and, 96; and male xiahai, 56 –57, 219n21
identity, 98; religious, 8; subethnic, Xiluo, 72
90, 220n24; symbolic, 4, 11, 214n7
yamen runners, 23 –25, 174, 217n6
Water Margin (novel), 32–33, 36, 48, Yellow Turbans (sect), 220n28
54, 204, 219n19 yin and yang, 61
wen. See civil yingchou, 154, 176 –177, 198
Wen, King, 45 yingxiong haohan, 47, 173
White Lotus Society, 8, 220n28 yiqi, 35, 144, 162, 199
Wing Chun, 54 Yunnan, 87, 90, 119, 164, 166; Muslim
Wolf, Arthur, 27 presence in, 232n2; reform-era
wu. See martial revival of ritual practice, 240n40;
Wu, King, 45 trucking business in, 170
wu masculinity, 20, 47–50, 134, 161,
166, 207, 211, 221n34 Zhang Yimou (film director), 52
Wu Song, 22, 32–33, 36, 48 – 49 Zhao Gongming, 101, 168
Wudang style (martial arts), 213n2 Zhongkui (deity), 22
Wufu Dadi (Great Emperors of the Zito, Angela, 28, 40, 204
Five Blessings), 136, 239n32 Žižek, Slavoj, 209
About the Author
Avron Boretz received his Ph.D. in anthropology from Cornell Univer-

sity. He is currently Program Director at the United Board, based in
Hong Kong. Prior to joining the United Board, he served for ten years
as a member of the Asian Studies faculty at the University of Texas at
Austin.

Boretz 2011 Gods, Ghosts, and Gangsters (Book)

Uploaded by

Copyright:

Available Formats

Boretz 2011 Gods, Ghosts, and Gangsters (Book)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Boretz 2011 Gods, Ghosts, and Gangsters (Book)

Uploaded by

Copyright:

Available Formats

$9521%25(7=

RITUAL VIOLENCE, MARTIAL ARTS, AND MASCULINITY

University of Hawai‘i Press

16 15 14 13 12 11 6 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data

University of Hawai‘i Press books are printed on acid-free

Designed by Josie Herr

Note on Translation and Use of Foreign Terms ix

Chapter 2 Violence, Honor, and Manhood 21

Chapter 3 Taidong: The Mountains and Beyond 58

Chapter 4 Fire and Fury 87

Chapter 5 Tales from the Jianghu 115

Chapter 6 Wine, Women, and Song 176

Chapter 7 Conclusion: Faces of the Gods 204

This book is the product of countless collaborations and interactions

for her gracious invitation to participate in the research and writing

Fieldwork involved the use of several languages and dialects, includ-

I am a student of the Chinese martial arts. Beyond practice, however,

The present work is partly a response to that challenge. Yet this is

afternoon, as I was making my way home along Xinyi Road (one of

Martial Arts and Martial Cults

the rare sources for descriptions of premodern martial arts practice.

Violence and Religion

The Ritual Logic of Domination

gion—blood sacrifice, territoriality, the imperial metaphor, a sequence

inherently fierce and dangerous minions are structurally dominant,

Ritual, Violence, and Masculinity

the process is a form of symbolic, not tangible, capital—the patriarch’s

[They] burned incense, swore mutual allegiance and became

Why would “big boss” Wan Zhong choose to represent himself as

impersonation of a military god (only military gods would be carried

replace the father/boss. Nor is either likely to be satisfied with per-

The Aesthetics of Martial Ritual Performance

Of Morals and Margins

but honorable social other, serves as a symbolic field through which

however, inverts this relation, positing martial power as the originating

a set of social roles, moral qualities, and modes of self-presentation that

identities as individuals and as men, but also to their very livelihoods.

Violence, Honor, and Manhood

outside to burn spirit money in the “gold-burning furnace” ( jinlu), the

subordinates, magistrates, and commanders would have little chance

The Celestial Interface

dead. The qi of those who die violently or prematurely lingers among

fect historical accuracy. Most of what is tangible in popular religious

Immanence and Self-Transcendence

exclusive position as a sacrificer (that is, a son) vis-à-vis his patrilin-

perhaps be interested in playing the part of Civil Magistrate (Wenpan)

The Military Retainers troupe of Taidong Loyal Harmony Temple encounters

Most of the devotees at Loyal Harmony Temple, and nearly all

in-groups of family, kin group, and village, life is precarious. Strang-

jianghu is intrinsic to, and in some ways constitutive of, working-class

Honor and Reciprocity

to forgo personal pleasures and profit, to put themselves in harm’s way

lakes,” as well as in sordid tabloid accounts and published confessions

Class and Masculinity

common parlance, but by many writers with academic credentials as

in some “dark cult” and secret society rituals (DeBernardi 1987; J. S. M.

The Interpenetration of Wu and Wen

Chinese folk pantheon is not simply historical reenactment but rather