This book focuses on the roles of law enforcement and the treatment

of criminals in the functioning of traditional Chinese society. More

narrowly, it examines the enlistment methods of the enforcers of the law,
differences between rural and urban law enforcement, changes in law
enforcement over time, and the changing balance between the military
and the civil authorities in law enforcement.
This study explores law enforcement within the context of Sung
society. Professor McKnight describes the group of criminals who were
the core of the habitual criminal group in Sung China: young, unskilled,
and unattached males, just as they are today. In addition, he looks at the
fate of the criminal after capture and conviction, including the various
punishments used by the Sung government.
Cambridge Studies in Chinese History, Literature, and Institutions
General Editor Denis Twitchett

LAW AND ORDER IN SUNG CHINA

 Other books in the series

Glen Dudbridge The Hsi-yu Chi: A Study of Antecedents

to the Sixteenth-Century Chinese Novel
Stephen Fitzgerald China and the Overseas Chinese: A Study
of Peking's Changing Policy 1949-70
Christopher Howe Wage Patterns and Wage Policy in Modern China, 1919-1972
Ray Huang Taxation and Government Finance in Sixteenth-Century Ming China
Diana Lary Region and Nation: The Kwangsi Clique in Chinese Politics, 1925—37
Chi-yun Chen Hsiin Yiieh (A.D. 148-209): The Life and Reflection
of an Early Medieval Confucian
David R. Knechtges The Han Rhapsody: A Study of the Fu
of Yang Hsiung (53 B.G.-A.D. 18)
J. Y. Wong Yeh Ming-ch'en: Viceroy of Liang Kuang (1852-8)
Li-li Ch'en Master Tung's Western Chamber Romance
(Tung hsi-hsiang chu-kung-tiao): a Chinese Chantefable
Donald Holzman Poetry and Politics: The Life and Works
ofJuan Chi (A.D. 210-63)
C. A. Curwen Taiping Rebel: The Deposition of Li Hsiu-Chang
Patricia Buckley Ebrey The Aristocratic Families of Early Imperial China:
A Case Study of the Po-ling Ts'ui Family
Hilary J. Beattie Land and Lineage in China: A Study of Tung-ch'eng County,
Anhwei, in the Ming and Ch'ing Dynasties
William T. Graham, Jr 'The Lament for the South':
Yii Hsin's At Chiang-nan fu
Hans Bielenstein The Bureaucracy of Han Times
Michael J. Godley The Mandarin-Capitalists from Nanyang:
Overseas Chinese Enterprise in the Modernisation of China 1893-1911
Charles Backus The Nan-chao Kingdom and T'ang China's
Southwestern Frontier
A. R. Davis Tao Yiian-ming (A.D. 365-427): His Works and their Meaning
Victor H. Mair Tun-huang Popular Narratives
Ira E. Kasoff The Thought of Chang Tsai (1020-1077)
Ronald C. Egan The Literary Works of Ou-yang Hsiu (1007-1072)
Stanley Weinstein Buddhism under the T'ang
Robert P. Hymes Statesmen and Gentlemen: the Elite of Fu-Chou,
Chang-hsi, in Northern and Southern Sung
David McMullen State and Scholars in T'ang China
Arthur Waldron The Great Wall of China
Hugh R. Clark Community, Trade, and Networks:
Southern Fujian Province from the Third to the Thirteenth Century
Denis Twitchett The Writing of Official History Under the T'ang
J. D. Schmidt Stone Lake: The Poetry of Fan Chengda (1126-1193)
Law and Order
in Sung China

Brian E. McKnight
University of Arizona, Tucson

j | § CAMBRIDGE
;';P UNIVERSITY PRESS
 Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 IRP
40 West 20th Street, New York, NY IOOI 1-4211 USA

10 Stamford Road, Oakleigh, Victoria 3166, Australia

© Cambridge University Press 1992

First published 1992

Printed in the United States of America

Library of Congress Cataloging-in-Publication Data

McKnight, Brian E.
Law and order in Sung China / Brian E. McKnight.
p. cm. - (Cambridge studies in Chinese history, literature,
and institutions)
Includes bibliographical references (p. ) and index.
ISBN 0-521-41121-1 (hardback)
1. Law enforcement - China - History. 2. Corrections - China -
History. 3. China - History-Sung dynasty, 960-1279. 1. Title.
11. Series.
HV8260.A2M35 1992 91 -34373
363.2*0951'09021-dc2o CIP

A catalog record for this book is available from the British Library.

ISNB 0-521 -41121 -1 hardback

 Contents

List of figures, maps, and tables vi

Preface ix
List of abbreviations xiv

i Introduction i
2 The historical context 34
3 Crimes and criminals 67
4 Informal and semiformal agencies of law enforcement 117
5 Formal civil agencies of law enforcement 147
6 The role of the military in law enforcement 191
7 Supervision of law enforcement - the role of the intendants 228
8 Personnel selection 251
9 Urban crime and urban security 283
10 The Sung penal system 321
11 Jails and jailers in the Sung 353
12 Penal registration 385
13 The death penalty 446
14 Modifications of penalties 472
15 Conclusion 507

Glossary 524
Bib liography 533
Index 548
 Figures, maps, and tables

Figures
2.1 Mourning chart 54
2.2 Sung local administrative structure 64
5.1 Chart to be used in preparing inquest reports 156
5.2 Archery practice 182
5.3 Weapons 185
9.1 Kaifeng about 1021: schematic arrangement 295
9.2 Command structure in Kaifeng after the mid-eleventh century 298
10.1 Judicial torture 338
10.2 Judicial torture 339
10.3 Prison instruments 342
10.4 Wearing the cangue 344
11.1 Chinese jail 354
11.2 Tiptoe cell 360
13.1 Beheading 448
13.2 Beheading 449
13.3 Strangulation 450
13.4 Death by slicing 451

Maps
2.1 Sung, Liao, Hsi Hsia, and Koryo borders in the eleventh century 39
2.2 The Southern Sung state and its neighbors by the mid-twelfth century 43
9.1 Plan of Hangchow (Lin-an) 313
12.1 "Distant evil" prefectures 410
12.2 Sramana Island 411
12.3 Sung circuits 432-33
12.4 Transfers of amnestied convicts 434-35
14.1 Early expansion of the heavy-penalty jurisdictions 476
vi
 Figures, maps, and tables

Tables
2.1 Types of Sung laws 62
3.1 Numbers of penal rules 68
3.2 Numbers of bandits reported 113
5.1 Sample reward information 169
5.2 Numbers of archers in selected districts 173
6.1 Numbers of soldiers in selected districts 221
10.1 Sung fetters 348
11.1 Numbers of prisoners and of deaths from disease 373
12.1 Sample figures for prison citadel troops and other military units 397
12.2 Remuneration of soldiers in provincial army units 443
13.1 Numbers of persons sentenced to death 466-67

vn
 Preface

All human associations face some problems of deviance and must re-
spond when they feel threatened. When looking at the responses of any
human group, we can see commonalities, people doing things in the
same way for the same reasons. As the focus of inquiry narrows, the
shared traits of the broader levels are retained, but in addition other
characteristics appear that are specific to the level of inquiry. For
example, all known societies abhor murder, but what sorts of homicides
do they classify as murder? Such differences affect not just the definitions
of crimes but also the law-enforcement responses. The law-enforcement
systems of all formally organized states share some traits that we cannot
find in simple societies. And preindustrial agrarian bureaucratically
organized states, like traditional China and the Byzantine Empire,
again share some law-enforcement problems and responses to problems
peculiar to states of their type.
In moving down through this nested hierarchy of organizational types,
we reach one of the key steps at a level where the social units are distinct
as unique cultural and political entities. What was peculiarly Byzantine,
or peculiarly Chinese, about the perception of law-enforcement problems
and responses and about the ways in which lawbreakers were treated?
These attributes, which distinguish one such political entity from
another, are simply parts of our definition of what it is to be Chinese or
Byzantine or Roman. Such defining attributes evolve over time. The
Chinese language has changed from the time of the Shang dynasty to
today. However, the changes in these fundamental attributes are evolu-
tionary, not revolutionary. Despite the ways in which it has changed, the
Chinese language has remained the Chinese language.
On this level of cultural distinctiveness, the direction of inquiry may
change. Instead of looking synchronically across a set of distinct states,
we can look at the ways in which the institutions, beliefs, and practices
of a single society evolved over time. The practices and beliefs of the
ix
 Preface
Chinese altered in a variety of ways across the centuries. When studying
Chinese beliefs and institutions, historians have often divided their mate-
rials according to the succession of major dynasties. Although this
approach has serious drawbacks as a framework for examining such
areas as economic or intellectual history, it remains useful in the study of
politically determined matters, matters decided at least in part by the
government. For this reason, studies of politically shaped institutions can
profitably concentrate on single dynasties. The Sung dynasty (960-
1279), because of its pivotal position in the evolution of early modern
Chinese institutions and because of the wealth of available nonofficial
sources, is a particularly important and attractive period for study.
Sung history has already attracted the attention of scholars, who
have focused especially on socioeconomic developments, philosophical
changes, and the arts. Relatively little work has been done on the legal
and law-enforcement systems, though these systems clearly affected
other facets of Sung life. A thorough analysis of Sung law-enforcement
and penal systems would find traits from all these different levels of
responses to deviance, some common to all human societies, some to
all states, some to all traditional bureaucratically organized agrarian
empires, and so on. To isolate the different levels, however, and to
describe points of similarity and difference, we would need a host of
studies on similar topics, not only for other periods in Chinese history,
but also for other societies. At our present stage of inquiry, an effort like
this can thus only be a preliminary exploration.
Despite their obvious importance to the working of society, law
enforcement and penal policies have been relatively little studied in
China. Rather, most of the research on legal matters has concentrated on
analysis of legal codes. Such studies are important as foundations for
further work, but they deal at best only with normative situations and
have little to say about actual historical experience. Perhaps because of
the underdeveloped state of the field, some of these studies also tend to
generalize about social phenomena in misleading ways, glossing over
significant differences among different historical periods. Scholars who
write on Chinese history often speak about the "Confucian attitude
toward law" or even the "Chinese attitude toward law," as if there were
a shared view of law held by all Confucians or even all Chinese. Surely
this is not the case. Attitudes toward law differed within Chinese society
from social group to social group and within a single social group from
individual to individual. Furthermore, these attitudes changed over time.
The most we can say is that at any one time the members of a given
social group might have shared a number of core values, a worldview,
and ways of acting, but even then such sharing could never have been
 Preface
complete between any two individuals, nor could their views on the
relevance of those values to specific legal cases have been identical. Such
sharing should probably be thought of as a standard distribution with a
modal commitment to a given position, but with members of the group
distributed around that modal point.
A further problem in studies of Chinese law has been a tendency to
treat attitudes toward law as if they were simply parts of a more general
intellectual position, as if they existed without having reference to real-
world situations. Again, for most of the Sung individuals whose writings
are left to us, this clearly was not so. If we were to identify one charac-
teristic shared by almost all writers whose works are extant, it would
be that they were actively concerned about and participated in the
political as well as the intellectual debates and actions of their time. Like
Marcus Aurelius or Cicero, they were deeply involved in governance and
theorized in terms of their commitment to politics. Because they were
involved with the practical questions of their times, even when they
wrote what appear on the surface to be abstract discussions, they were in
reality discussing current or foreseen worldly problems. Abstract ideas
were used to give coherence and order to their understanding of real-life
data, but these data in turn shaped the actual understanding of these
abstract ideas by burdening them with content. This interrelationship
between abstract ideas and real-life situations is true to some degree
for all societies, but it was especially important in a society that was
culturally resistant to a theory-praxis distinction.
In this study, the problem represented by changes in ideas and prac-
tices over time has been reduced by focusing on a single era, the Sung
dynasty. In providing both analyses of the abstract conceptions under-
lying one large area of law, and descriptions of some of the institutions
used to enforce the laws, I have attempted, where possible, to highlight
the real-world referents that were the (often unarticulated) backdrop
against which men set forth their more abstract discussions of right and
wrong and their suggestions on how to deal with problems of deviance
and disorder. I have concentrated on crime and its punishment. The
study of crime and criminals emphasizes some key aspects of attitudes
toward more general questions of deviance and correctness; the study
of the police and penal systems illustrates both attitudes toward the
meaning and role of punishments and the types of practical constraints
that helped shape Sung responses to deviance.
Because this is the first attempt to deal at length with the problems of
law enforcement and penal policy in the Sung dynasty, it necessarily
cannot be a comprehensive study. In analyzing law-enforcement institu-
tions I have chosen the most important agencies. At one time or another
xi
 Preface
and in one place or another, literally dozens of different kinds of officials
were involved in law enforcement. I will deal here principally with the
three most important formal offices in the police system - the (civilian)
district sheriff (hsien-wei), the (military) patroling inspector (hsun-chien),
and the judicial intendant (t'i tien hsing-yii shih) (a civilian post, though it
was at times filled by military officers). It has been necessary also to
limit the kinds of crimes and criminals examined in any detail. Some
problems, such as the wrongdoings of bureaucratic functionaries, are so
important, so complex in their effects, and so interrelated with political
and economic questions that they deserve a book of their own. Here they
can be touched on only briefly. Furthermore, many of the topics in the
chapter on crime and criminals deserve more extended treatment than
they receive in this book. For the most part, like James Q. Wilson and
Richard Herrnstein, in their important book Crime and Human Nature^ I
have focused on those crimes, such as murder, robbery, rape, and violent
assault, which are seen as crimes in all known societies (though I have
also briefly described a few types of acts, like gambling, which are not
universally seen as criminal). I have also generally avoided any extended
discussion of the trial process in Sung China, except where information
is needed to make the main topics of this work understandable, as there
is already a competent if brief survey of the Sung trial system available
in English.2
In short, my object here is both to portray as clearly as possible the
general patterns of law enforcement and penal policy characteristic of
the Sung period and the ways in which they changed (and the ways in
which they did not), and insofar as it is possible at this early stage of
studying these phenomena, to reveal their ideological foundations and
the interactions of these ideological factors with political, social, and
economic conditions in determining the shape and functioning of these
systems. I am asking who the criminals were, what sorts of acts they are
said to have committed, why (in the opinion of Sung writers) people
violated the laws, and how the people and the government of the Sung
reacted to their deviant behavior.

1 James Q. Wilson and Richard Herrnstein, Crime and Human Nature (New York: Simon &
Schuster, 1985).
2 Ichisada Miyazaki, "The Administration of Justice During the Sung Dynasty," in
Jerome Alan Cohen, R. Randle Edwards, and Fu-mei Chang Chen, eds., Essays on
China's Legal Tradition (Princeton, NJ: Princeton University Press, 1980), pp. 56-76. This
essay is a development of Miyazaki's longer article, "So-Gen jidai no hosei to saidan
kiko" (Law and judicial system in the Sung-Yuan period), Toho Gakuho 24 (1954):
115-225. Other aspects of the Sung trial system are also described in a series of articles
by the late Hsu Dau-lin, published under the title Chung-kuo fa-chih shih lun-chi (Taipei:
Chih-wen ch'u-pan she, 1975).
xii
 Preface
I have tried to make this book accessible to people whose primary area
of interest is not Sung history or even Chinese history. For this reason,
when describing the locations of events, I have used the names of the
modern Chinese provinces when these provinces' boundaries approxi-
mate the boundaries of Sung circuits (as in the case of Fukien) or in
cases in which a province can serve as a sufficient indication of the
general location, even though during the Sung period the region was
divided into several circuits (as in Szechuan). In these cases I have
followed the postal spelling system. To make the book more useful to
students I have also, in most cases, adopted the translations of Chinese
official titles listed in Charles Hucker's A Dictionary of Chinese Officials'
Titles in Imperial China?

As in any work of this sort, I am indebted in innumerable ways to the

scholars and writers of the past and, as the footnotes make clear, to a
number of modern Chinese, Japanese, and Western students of Chinese
history. I am indebted to Professor Robert Hartwell for giving me price
lists that make it possible to provide at least a rough picture of the
practical meaning of rewards offered or salaries paid. The use of such
prices in a historical study is fraught with potential problems, and I
alone am responsible if they are misleading. Other colleagues have sent
me suggestions and information, but I am particularly grateful to those
who have been kind enough to read the manuscript in whole or in part
and give me their suggestions and criticisms. I wish especially to thank
Professors Roger Ames, Harry Ball, Herbert Franke, Kenneth Kipnis,
James T. C. Liu, and Winston Lo, Drs. Charles Benn and Richard
Vuylsteke, and Mr. Lewis Mayo for their advice and counsel.

3 Charles Hucker, A Dictionary of Official Titles in Imperial China (Stanford, CA: Stanford
University Press, 1985).

Xlll
 Abbreviations

CHTK Chou-hsien Vi-kang

CYKC Che-yii kuei-chien
HCP Hsu tzu-chih t'ung-chien ch'ang-pien
HNYL Chien-yen i-lai hsi-nienyao-lu
SHT Sung hsing-t'ung
SHY Sung hui-yao chi-kao
SKCS Ssu-k 'u ch 'iian-shu
SKCSCP Ssu-k }u ch 'uan-shu chen-pen
SPPY Ssu-pu pei-yao
SPTK Ssu-pu ts'ung-k'an
ss
STCLC
Sung shih
Sung ta-chao ling-chi
SYTFCTS Sung-Yuan ti-fang chih ts'ung-shu
TFSL Ch'ing-yuan t'iao-fa shih-lei
TITC Tso-i tzu-chen
TLSI T'ang-lu shu-i
TSCC Ts'ung-shu chi-ch'eng
WHTK Wen-hsien t'ung-k'ao
WTHY Wu-tai hui-yao

XIV
 1
Introduction

Politics and law

This book is about social control. More specifically, it is about certain
law-enforcement and penal institutions, practices, and policies used by
the Chinese of Sung times to support the social order. It describes these
institutions, practices, and policies during the Sung dynasty. It also tries
to illuminate some of the ways in which they were shaped by the ideas
and attitudes of the ruling elite, the fiscal situation of the state, the
nature of the economy, political realities, local geography, bureaucratic
skills, and technology. It seeks to show what functions these institutions,
policies, and practices played in Sung society and why they were needed.
The topic also illuminates, incidentally, the character of the tradi-
tional Chinese state. Western thinkers on the origins and continuation of
states are divided into two opposing camps. One tradition stresses the
integrative nature of the state. From Plato and Aristotle to English
writers of the eighteenth century to some modern sociologists like Talcott
Parsons, these thinkers have emphasized the origins and continuing role
of the state as a positive integrative system. Such thinkers tend to view
stability and order as normal, with conflict an abnormal distortion
of the natural order. The other tradition, more specifically modern,
from Hobbes to Marx to some modern anthropologists like Maurice
Freedman, sees the state as a system born out of violence and coercion.
For such thinkers the state is created from conflict and continues as
a mechanism by which the majority is subjected to external control
through fear or force.
These opposing lines of thinking are deeply embedded in the Western
tradition. The central tradition of Chinese culture - with its belief in a
world of complementarity rather than contradictory opposites, a world
in which change defines stability and stability defines change, a world in
which integration is the organically interrelated counterpart of dis-
1
 Law and order in Sung China
integration, and cooperation is not conceivable without conflict or
conflict without cooperation - is a much more realistic vision of the
human world as it is and has been. This traditional Chinese view of the
world, in which the appearance of disorder and conflict are embedded
in a higher level of integration, made possible a measured response to
disorder that could legitimately be modified to accord with current
conditions.
Law-enforcement and penal institutions are surely quintessential
instruments and symptoms of conflict in society, but as we shall see, the
Sung Chinese envisioned and constructed law-enforcement agencies in
ways which fostered the integration of disparate elements of society.
Furthermore, Sung Chinese penal systems, while embodying the state's
power to outcast wrongdoers from normal society, also were designed to
allow and encourage the reintegration of the reformed into ongoing
community life. People who break widely shared social norms threaten
the common good, but they are still people. The problem is not simply
how to stop their abuses but also how to save them from themselves.
The Sung Chinese reaction to the threat of disorder posed by those
who violated important social norms was conditioned by the interaction
of their beliefs about the nature of people, the level of managerial and
physical technology, the state of economic development, the collective
perception of the lessons of the past and the character of the present,
and their control over resources. Ideology, finances, and technology
interacted to create a constellation of institutions and practices designed
to preserve the social order and the political power of the ruling group.
Some of these institutions and practices were educational, having as
their aim the inculcation of "proper" views and patterns of behavior.
Some were legal, in that they were concerned with determining the
boundaries of permissible behavior and the appropriate treatment of
those who promoted disorder by going beyond those boundaries.
The ways in which a society deals with its deviants reflect both
fundamental intellectual commitments and the historical situation. We
cannot adequately understand the traditional Chinese view of their place
in the world or the evolution of their society without understanding how
those deviants who violated laws were controlled through the organized
use of force. Relatively little has been written on this problem for tradi-
tional China. Some reasons for the paucity of studies are clear. In part it
mirrors and perpetuates the tendency of the traditional Chinese elite to
ignore or minimize the importance of law and the police power in
society. The self-justification and self-image of the elite in traditional
China stressed their moral superiority over the common people. They
systematically downplayed the importance of the formal judicial system
 Introduction

and law enforcement. They were remarkably successful in disguising the

degree to which law and force governed Chinese society in times past.
As a result, among ordinary Chinese today and even among scholars
concerned with China, there is a widespread misunderstanding of
the importance of the judicial system in premodern China. This mis-
apprehension can be overcome by the work of legal scholars, but so little
has been done that for the moment we need to explore limited periods
and special topics within the Chinese tradition in preparation for the
more general works which will make possible a more balanced under-
standing of the patterns of the Chinese past.
The Sung dynasty makes a particularly attractive starting point in
studying how the premodern Chinese dealt with these problems. Because
printing became widespread in the Sung period, a larger quantity and
greater variety of sources have survived from this period than from
earlier eras. In particular, we have more and larger sets of the collected
papers of individuals, more collected government documents, and more
detailed annalistic histories based largely on government archives.
Furthermore, a study of Sung institutions provides a benchmark for
examining the institutions of later dynasties. In some ways the Sung was
very different from later times. In others, the Sung experience helped
shape late imperial institutions, the institutions that existed when the
Chinese empire came into conflict with the Western powers in the
nineteenth century. An exploration of Sung sources can illuminate not
only the particularities of the Sung experience but also the more general
problems that plagued succeeding Chinese dynasties.
In one sense the problems of lawbreaking faced by the Sung Chinese
were universal problems, and the Sung responses therefore share certain
basic characteristics with law-enforcement and penal strategies used
elsewhere. We shall be concerned with the ways in which the specific
responses of the Sung Chinese show how answers to universal problems
are idiosyncratically shaped by a society's particular constellation of
means and goals. Some usages may seem hauntingly familiar because
they are used in our own culture. Others are distinctively Chinese, not in
the sense that they are unknown elsewhere, but that they are grounded
in commitments that were common to the Chinese elite under the
empire. Still other usages are especially characteristic of Sung people.
An examination of the Sung Chinese experience can be of value as an
example of a specific response to general problems. All societies in all
times and all places face problems of disorder. The view that order in
society is normal and disorder abnormal is fundamentally naive. Order
implies stability, but in society, as in nature, the world is constantly in
flux. People have to work very hard to create a recognition of a social
 Law and order in Sung China

order that persists through time. Furthermore, at any given moment in

any given social system, the sharing of values, worldviews, and ways of
acting is never complete. The human problem is to create and recreate,
moment by moment, a sense of continuity and of a shared life sufficiently
convincing that social organization becomes possible. The idea of social
order is an indispensable collective creation, but the idea, though
indispensable, always and everywhere corresponds only in an imperfect
way to real life. "Society," whether in China or elsewhere, is in reality
an aggregation of discrete individuals experiencing a sequence of unique
and unrepeatable moments. People have to learn to see enough con-
tinuity in this flux to enable ongoing social organization.
One key to the successful creation of a sufficiently compelling vision of
order in society is the adequate education of newcomers, whether infants
or significant outsiders. Education in turn is dependent on communica-
tion. Each individual new to the social group must be schooled in its
patterns of communication, verbal and nonverbal. More than this, each
individual must be imbued with sets of ideas, attitudes, beliefs, feelings
that will make possible ongoing if ever faulty communication. To live
together we must be socialized. People learn patterns of behavior at the
primary level that enable them to function reasonably well beyond the
family. An appreciation of this is one of the hallmarks of the early
Confucians and was a widely held belief among later members of the
Chinese ruling elite. Later Confucian statesmen thought in terms of
"educational transformation" (chiao-hua), which meant for them the
molding of people through the power of example and exhortation into
patterns of behavior more like those advocated by the elite. One prime
element in this belief was the conception that people mirror authoritative
exemplars and can learn proper behavior within the family by observing
others. If a person has learned the appropriate patterns of behavior in a
primary-level setting, he or she can then extrapolate the learned patterns
to other people who are filling roles in some sense analogous to the roles
first encountered in the home. A disciple of Confucius in the fourth
century B.C. stated:

It is rare for a man whose character is such that he is good as a son and obedient
as a young man to have the inclination to transgress against his superiors; it is
unheard of for one who has no such inclination to be inclined to start a rebellion.
The gentleman devotes his efforts to the roots, for once the roots are established,
the Way will grow therefrom. Being good as a son and obedient as a young man
is, perhaps, the root of a man's character.1

1 Confucius, The Analects, trans. D. C. Lau (New York: Viking Penguin, 1979), p. 59.
4
 Introduction

Socialization is absolutely necessary, but it never succeeds in mak-

ing one individual like another. Because communication is always
imperfect, because each particular is unique, and because each time and
place is different, teaching and socialization are always imperfect and
incomplete. Although the majority of the members of a given social
system may be in general agreement on some norms, it is important to
remember that such agreement is never total between any two individ-
uals. By definition, individuals differ. Even when they share a given
value, worldview, or way of acting, individuals will weight it differently
and see it differently reflected in the affairs of everyday life. The dif-
ferences may be, in a given instance, too small to affect a decision, but
they nonetheless are there. More than this, a society of any size is
composed of subgroups whose members share certain values, attitudes,
ideas, feelings, and patterns of behavior which are not shared equally by
outsiders. Like individuals, subgroups not only differ in values, views,
and behaviors but also weight differently those that they do share with
the larger society.
The law, in China as elsewhere, is able to function because individuals
are able to place their own views and the views of the subgroups to
which they belong in a hierarchy and to subordinate inevitable dif-
ferences in nuance or detail on specific subjects to higher levels of broad
agreement. Despite their unavoidable disagreements in detail, juries
continue to bring in verdicts of guilty or not guilty, and magistrates
agree on judgments. When people say a society continues, they mean
that the imperfections of communication, failures of socialization, and
the resulting conflicts are insufficient to destroy the basic ground of
agreements among a sufficiently large number of aggregated individuals.
In real life there is much miscommunication and failed teaching. The
order created and recreated by endless labor always threatens to break
down in the face of the failures in socialization and the incommensurable
aims of social subgroups. Forces of disintegration and disorder lurk
behind the facade of stability. Individuals and groups repeatedly violate
norms that are widely shared within society. Worse, some important
and widely shared norms, under some conditions, are in practice
incompatible with other widely shared norms.
The problems are particularly acute when external conditions such as
warfare or economic dislocations put unusual strains on the social order.
The Confucians recognized this from the beginnings of their tradition,
but it was stated with special clarity by the philosopher Mencius (372-
289). After discoursing on the importance of government policies that
support a healthy economy, he remarked that if these policies were
followed,
 Law and order in Sung China

when the people have more grain, more fish and turtles than they can eat, and
more timber than they can use, then in the support of their parents when alive
and in the mourning of them when dead, they will be able to have no regrets
over anything left undone. This is the first step along the kingly way. 2

Basic beliefs
Mencius, and indeed most thinkers in traditional China, believed in
the immanence of order in the world. Correct behavior was behavior
consonant with this immanent order. Man's uniqueness consisted of his
ability to be consciously participatory and creative in his responses to
this immanent order. Man makes himself: Only man is fundamentally a
creature of art rather than instinct. His ability to create appropriate
behavior by performance (/fa) is not, of course, unlimited. 3 The imma-
nent order sets boundaries to appropriate responses, though within these
boundaries men can respond in somewhat different ways. "Normal"
people usually respond within the boundaries; deviance is going beyond
them. For some behaviors the law, fa, defines the boundaries, and the
punishments, hsing* state the potential cost to the individual of exceed-
ing them. But we should not overstate the distinction between these two
concepts of law and punishment. In Chinese the word fa, which is
often translated as "law," also means "model," a word that implies
boundaries and edges, and the word hsing* which is usually translated as
"penalties," is in many classical contexts interchangeable with another
word, hsing,h which means "form," "model," or, as a verb, "to shape."
The Chinese, like other peoples, have created the boundaries in law
through legislation, indicated what particular acts fall within the bound-
aries through adjudication, and attempted to coerce people to staying in
bounds by means of punishments.
In China the behavior of those whose acts accorded with what was
orderly was cheng - "orderly," "straight," "correct." The behavior of
those whose acts did not accord with what was orderly was hsieh - "not
correct," "deviant." The state of the world was determined by people's
behavior, either correct or deviant. An emphasis on the importance of
not being hsieh, "deviant," can be traced to Confucius, who quotes a line
from the Book of Poetry to emphasize the fundamental need not to go
astray and to keep to the correct path. 4 The degree of disorder intro-
2 D. C. Lau, trans., Mencius (Baltimore: Penguin Books, 1970), p. 51.
3 For this insightful understanding of the relationship of men and /i, a I am indebted to
Roger Ames and David L. Hall, Thinking Through Confucius (Albany: State University of
New York Press, 1987).
4 This line from the Book of Poetry, ssu wu hsieh, is translated in radically different ways, but
in all cases the word hsieh itself is not so problematical. It implies leaving or deviating
 Introduction

duced by deviant behavior varied with both the importance of the

deviant individual within the social system and the seriousness of the
misbehavior. According to this view, the misdeeds of the emperor, even
relatively minor ones, might have serious repercussions, whereas mis-
deeds by an ordinary man, though not unimportant, would hardly have
the same impact on the world. Given the Chinese view of the importance
of role modeling in the teaching of behavior, and the key place played by
authority figures in setting standards by the proper performance of their
duties, it is hardly surprising that so much attention was paid to the
moral weight of imperial acts.
The Classical Confucian thinkers differed in their views on how people
knew and why they observed or violated the boundaries of appropriate
behavior. For the Mencian tradition the roots of such knowledge and the
tendency to behave well were innate but could be obscured. For the
tradition of the last of the great Classical Confucian thinkers, Hsiin-tzu
(fl. 298-238), such knowledge was not innate, and people by nature
tended toward evil, but the knowledge of the boundaries could be
learned, and the commitment to stay within them could be instilled.
For Han (and most later) Confucians, people by nature possessed the
potential for recognizing and observing these boundaries, but also the
potential for violating them. This conception of a mixed human nature,
containing the capacity for both proper and improper behavior, was
widely accepted by members of the Sung elite. It was expressed most
coherently by the philosopher Chu Hsi (1130-1200), who was in part
repeating widely shared views when he described people as being com-
posed in part of principle (lih) - by definition the regularity that con-
stituted continuity in the immanent order of the universe - and in
part of what we might call material force (ch'i), which varied in its
purity. 5 Because people's actions were affected by their material force
and because some people's material force was less clear than the mate-
rial force of others, the level of ethical behavior varied. 6
When looking at the problems of deviance and disorder, Sung

from the correct path. See Confucius, The Analects, p. 63; James Legge, trans., The
Analects (Hong Kong: Hong Kong University Press, i960), The Chinese Classics, vol. 1, p.
146; James Legge, trans., The She King or Book of Poetry (Hong Kong: Hong Kong
University Press, i960), The Chinese Classics, vol. 4, p. 613.
5 The appropriate translation for ch'i is a much debated question. In a recent collection of
writings on the philosophy on Chu Hsi, the editors have chosen to offer their readers a
menu of possibilities - material force, ether, vital force, breath. See Wing Tsit-chan, Chu
Hsi and Neo-Confucianism (Honolulu: University of Hawaii Press, 1986), p. 618.
6 These conceptions are reflected in various works by Chu Hsi but are perhaps most
clearly accessible in the records of his conversations with disciples that were gathered
together under the title Chu-tzu yu-lei. See Chu Hsi, Chu-tzu yu'-lei (Taipei: Wen-lu
ch'u-pan she, 1986).
 Law and order in Sung China

Confucians accepted the validity of both the internal roots of evil a la

Hsiin Tzu and the impact of environment a la Mencius. In either case,
appropriate education, which could properly be received only within
Chinese culture, was needed to lead people to good behavior.
For all these thinkers, an educative process, what we would call social-
ization, was necessary. Education was an intellectual process in part, but
always and more importantly, it was moral. The aim of education was to
produce good people. Whether this process uncovered innate goodne°s
or inculcated a penchant for behaving properly in creatures that were by
nature amoral was not of decisive importance. Preserving the process of
education itself was. The learning of good behavior and the reforma-
tion of bad behavior were clearly possibilities. Human behavio was
voluntary and subject to formation and reformation. People could learn
where the boundaries of appropriate behavior were and how to act
within them. If they were provided with appropriate models and had
placed before them the proper mixture of rewards and penalties, the
people could be led toward transformation. The thirteenth-century
official Yuan Fu wanted the government to "reward the village officials
so that the people will be pleased and will mutually encourage one
another, so that customs will gradually be transformed. . . . By rewarding
the good, the evil can be made to feel ashamed." 7 Toward those who
would learn, the authorities could be forgiving; toward the willfully
intransigent, they would be ruthless.

Environmental influences
For Hsiin Tzu, with his belief that human beings were basically amoral
and driven by desires, the problem was how to train them to goodness.
The existence of bad behavior was not an intellectual problem. The
basically optimistic Mencian view, that people were by nature good, had
raised the problem of the origin of evil. Mencius traced evil - the
knowing violation of those norms and behavior patterns that were in
accord with the roots of human nature - to the influence of the environ-
ment. Environmental influences were both short term, embodying
people's reactions to their immediate surroundings and problems, and
long term, reflecting both long continuing relations with individuals and
social systems and continuing physical environmental problems. The
short-term problems included such things as the temporary devastation
of the economy resulting from warfare or natural disasters, or currently
associating with undesirables. Long-term problems included having been

7 Yuan Fu, Meng-chai chi, Ts'ung-shu chi-ch'eng (hereafter cited as TSCQ) ed., 3.33.

8
 Introduction

improperly raised from childhood or growing up in an impoverished

region. Geography molded character. The Sung Chinese firmly believed
that people had regional characters determined by their environment.
Thus a Sung official, writing about Ching-hu South Circuit, could say
that "the terrain is dangerous and the people hard." 8

Types of lawbreakers
In their writings about the problem of current criminal behavior, Sung
thinkers began by dividing lawbreakers into types based on their motiva-
tions. One group consisted of those who were basically good, in that they
had been so socialized that their proclivity for appropriate behavior
would ordinarily be dominant. Under normal conditions they would be
law-abiding, though they lacked that special intensity of moral com-
mitment which kept the superior person moral in adversity. They might
err but would not be evil. In the words of the Southern Sung official and
poet Lu Yu, "When the four quarters are without troubles and the
autumn grain is like a cloud, who would be a bandit?" 9 In hard times,
this line of thought ran, such people could be driven to crime by their
will to survive. The second Sung emperor, T'ai-tsung (r. 976-97),
speaking to his chief minister in the autumn of 985, observed: "The
people are the foundation of the state; the people know food to be their
lifeblood. . . . Recently because of natural disasters in Chiang-nan, con-
ditions are terrible. . . . As a result there is the calamity of refugees and
bandits." 10
This view was the dominant one among the Sung literati when faced
with widespread unrest. Indeed, there are numerous essays by Sung
officials that reflect this widely shared belief. In the words of the official
Fan Chun (1120-50), the people in years of drought and locusts "flee
death and seek for themselves. They pilfer gold and steal provisions.
They know no shame. At worst they gather together to do evil. . . . In
this way they become bandits. They are compelled by circumstance.
Surely they should be soothed and pacified, and their punishments
should be minimized." 11 Many others, such as Su Ch'e (1039-1112),
Ts'ai Hsiang (d. after 1128), Ch'en Yuan (d. 1145), Li Mi-hsiin (1089-

8 Ts'ai Kan, Ting-chai chi, Ch'ang-chou hsien-che i-shu ti-i chi ed., 2.13b. Sung circuits
were units of local supervision, many of which corresponded geographically if not
functionally to the provinces of later dynasties. For more about these circuits, see
Chapter 2 in this book.
9 Lu Yu, Wei-nan wen-chi, Ssu-pu ts'ung-k'an (hereafter cited as SPTK) ed., 13.11b.
10 Li Tao, Hsu tzu-chih t'ung-chien ch'ang-pien (hereafter cited as HCP) (Taipei: Shih-chieh
shu-chu, 1964), 26.3a-b.
11 Fan Chun, Hsiang-ch'i chi, TSCC ed., 15.149.
 Law and order in Sung China
1153), P'eng Kuei-nien (1142-1206), and Liu Yuen (1144-1216), also
appealed to this common wisdom in suggesting answers to banditry. 12
In times of troubles, the government might even be able to forestall
banditry if it dealt expeditiously with these economic roots of unrest. 13
In the Northern Sung, Ch'in Kuan (1049-1100) voiced a traditional
refinement to this analysis, saying that "since antiquity, robbers have
arisen because floods and droughts had occurred, while the taxes are still
fully collected and the labor services vexatious. Therefore the people
become robbers." 14 Ch'in Kuan was following a famous tradition. A
fellow official speaking in the same vein quoted the T'ang emperor T'ai-
tsung (r. 627-50) - seen by some as a paragon among rulers - as stat-
ing that "people become robbers because the taxes are high and the
demands for labor services vexatious. Officials are corrupt.... If the
people have sufficient food and clothing, they will not be robbers." 15 An
almost identical position was taken by Li Kuang (1077-1155), who
wrote that the answer to the constant banditry in Ch'ien and Chi
prefectures in Chiang-nan West Circuit was simply to send in prefects
and district magistrates who would lighten taxes and service burdens. 16
The problem was in part economic. Bad conditions allowed the poten-
tial for evil to surface, exposed the results of improper socialization,
and drove even basically decent people to commit crimes. However,
the impact of such conditions was exacerbated by the inappropriate
responses of incumbent officials. Abusive officials could create the pre-
conditions of lawbreaking even in good times. In bad times their juris-
dictions were often plagued by serious uprisings.
The state was faced with a dilemma: It had to secure resources, but
excessive demands would lead to unrest. During the troubles of the
1040s, the official Yii Ching (1000-64) P u t t n e dilemma neatly: "The
best policy would be to head off trouble by offering tax relief and by
guarding against bandits, but with the current troubles on the border, it
clearly will be hard to reduce taxes." 17 To survive, the state needed
revenues, but it also needed good administrators, men capable and
desirous of balancing the requirements of the government against the

12 Su Ch'e, Luan-ch'eng chi, SPTK ed., 36.12a; Ts'ai Hsiang, Tung-ming chi, Ssu-k'u ch'u'an-
shu chen-pen (hereafter cited as SKCSCP) ed. 1973, 23.9b-10a; Ch'en Yuan, Mo-t'ang
chi, SKCSCP 1972, I4.i9b-22b; Li Mi-hsun, Yun-ch'i chi, SKCSCP ed., g.iobff; P'eng
Kuei-nien, Chih-Vang chi, Ssu-k'u ch'uan-shu (hereafter cited as SKCS) ed., n . 137; Liu
Yueh, Yun-chuang chi, SKCSCP 1971, 7.15a.
13 Yu Ching, Wu-ch'i chi, SKCSCP ed., tsou-i shang 7b-8a.
14 Ch'in Kuan, Huai-hai chi, SPTK ed., 17.4a.
15 Fan Tsu-yii, Fan-t'ai shih-chi, SKCSCP ed., 22.1 ia.
16 Li Kuang, Chuang-chien chi, SKCSCP ed., 12.17b.
17 Yii Ching, Wu-ch'i chi, tsou-i shang 7b.

10
 Introduction

capacity of the people. A stress on the role of bad administration in

causing unrest is one of the most common themes in the critical writings
of Sung officials.18 To critics who charged that the root problem was an
excessively harsh legal system, Fan Chung-yen (989-1052) and Han
Ch'i (1008-75) replied that the problem was not that the legal system
was too harsh but that the administrators were incompetent. If the
administrators were good, then the people would benefit from the legal
system. If the administrators were bad, then the people would suffer.19
Individuals driven by want represented many - perhaps most -
criminals, but not all. There were also those who were by nature re-
calcitrantly evil. In the words of Wang Chih (1127-89):
Bandits arise from three causes - because people are hungry, because people are
thoughtless, or because people are evil. Hungry people are seeking to survive.
Thoughtless people are seeking riches. Evil people are seeking personal advan-
tage {If). The origins thus emerge from what people are seeking to avoid and
from what they want, yet with regard to the ultimate end points of these feelings,
some can be reversed and some cannot. The feelings of the hungry can be
reversed. The feelings of the thoughtless and evil cannot. Why is this so? Hungry
people who become bandits do not have any great desire to do so. They have no
way of living, and so thay are willing to face death. It is not that at heart they do
not value life and fear to die but that because they are pressed by circumstance,
they repress their feelings. They feel that if they are to continue living, they must
be willing to die, and so they become bandits. When placed between living and
dying, who at such a time, except the pure and upright, will sit quietly and wait
to die? Thus when they suffer a year of dearth, they must make shameless plans,
attack and pillage. . . . When it is a good year, they again become peaceful.. . .
Hungry people are to be pitied rather than hated, to be helped rather than killed,
to be grieved over rather than feared. On the other hand, thoughtless people who
covet riches have no self-control. They would seek riches from the Buddha and
still would not consider them enough. They would seek riches from ghosts and
goblins and still would not consider them enough. Therefore they are deluded
heterodox people. The Manichaeans are this sort of people. Trouble is born
when these thoughtless people seek riches without being sated. The evil people,
who seek personal advantage without stopping, are lazy and will not farm. They
are crude and will not work as artisans. .. . Therefore they come together un-
governably, with swords, depending on good fortune to grasp tenfold advantage.
Such people will not even regret dying if they can gain personal advantage.20

18 Compare Ch'in Kuan, who said that one problem was the overemphasis on using
examination graduates. He advocated using the able, including clerks, thereby
following the T'ang practice of letting local officials choose their own subordinates and
sparing local officials corporal punishment for errors. Ch'in Kuan, Huai-hai chi, 17.5b ff.
See also, for example, Wang Shih-p'eng, Mei-ch'i chi, SPTK 2.10b-lib.
19 HCP I 4 i . i 2 b - i 3 b (1043).
20 Wang Chih, Hsueh-shan chi, TSCC ed., 3.25-26. On the Manichaeans in China at this
time, see Chapter 3.
11
 Law and order in Sung China

The Chinese thus made a clear analytical distinction between those who
erred and those who were evil, and they stressed the importance of
treating these two types of lawbreakers in different ways: The errant
were to be helped; the evil were to be suppressed.

Leaders and followers

In part, this distinction between the basically good and the basically evil
was equated with a distinction between leaders and followers. From very
early in the Chinese written tradition, a distinction was drawn between
evil leaders who had to be destroyed and followers who might be saved.
The mid-eleventh-century official Yii Ching remarked that "since
antiquity, when bandits have caused injury, the authorities have always
pardoned the followers while strictly punishing the leaders." 21 The
Southern Sung official Li Kuang, concerned about the unrest which had
followed the invasion of north China by the Jurchen tribesmen who
founded the Chin dynasty (i 115-1232), argued for mercy for some and
extermination for others. He quoted a widely known passage from the
classical Book of Historical Documents:

"When the fire blazes over the ridge of K'uan, the gems burn equally with the
stones. . . . I will thus destroy only the chief criminals and will not punish their
forced followers." . . . When the Sung founder came to the throne he followed this
Way. Therefore his ability to unify the districts and pacify the world stemmed
from winning the people's hearts and nothing else. Since the Chien-k'ang period
(1125), the men of Chin have oppressed the people, who have lost their liveli-
hoods and have no way to support themselves. The weak die in the ditches; the
strong gather together to become bandits. Thus they do not have hearts that hate
the ruler or detest the emperor. Their evildoing stems only from their want.22
In practice, the Sung authorities adopted policies that reflected both
this belief in the importance of the division between leaders and fol-
lowers and their belief that in the case of followers in particular, the
impetus to wrongdoing was want and suffering. This Sung analysis was
admirably summarized by the official Ts'ai K'an, who also articulated
the widespread belief among the traditional elite that judicial abuses
could upset the balance of the natural order symbolized by the harmo-
nious ethers:
I have heard that although the bandit gangs of Kuang-nan West Circuit contain
many men, the key evil leaders and those who arose with them are but a handful.
The rest are simply coerced followers. How important this is! For these latter

21 Yii Ching, Wu-ch'i chi, tsou-i shang 5a. 22 Li Kuang, Chuang-chien chi, 11.12b.
12
 Introduction

are the children of your majesty. They are the stupidly honest and ignorant,
oppressed by hunger and cold. Being credulously deluded, they fall in with the
bandits. I am concerned that if they should inappropriately suffer injury and
death, the harmonious ethers will be harmed. I hope that the Court will inform
the bandits capturing officials that if they are wholly able to spare such men's
lives their merit will be the greater. It is not necessary to exterminate [them]. 23

Incorrigible wrongdoers
Although most men might behave well except when twisted by environ-
mental pressures, there were always some who would violate the social
order even under the best conditions. The appropriate response to the
first sort of lawbreakers was to alleviate the conditions that had led
them to break the laws, to persuade them to change their ways, and to
make possible their reintegration into normal society. The appropriate
response to the second sort of criminals depended not on persuasion but
on control or elimination. Sung officials, like judges everywhere, some-
times came to have a jaundiced view of the nature of many people. In
the thirteenth century Yuan Fu remarked that whenever lawsuits were
heard, it became abundantly clear that many people loved money and
paid no attention to propriety. They were willing to act with force and
valued life lightly, breaking the laws and violating principle. 24 For this
reason the mechanisms for socialization, for creating order, always had
to be accompanied by mechanisms for preventing or punishing deviance.

Political implications
It is important also to recognize the political implications of this view of
deviance. If moderately good men, many of whom had rather limited
opportunities for a proper education, could be driven into doing evil only
by the stark demands of survival, then those who behaved improperly
when not so driven must be basically flawed. Wrongful behavior among
members of the official elite could hardly be excused on the grounds of
immediate physical environmental pressures. The bitterness of Sung
partisan politics in part reflects the underlying feeling that political
opponents, whose views were seen as deviant, were behaving in illicit
ways despite the best physical conditions.
During the late Northern Sung (960-1126), with its bitter political
struggles, the problem of disorder loomed large in the minds of officials.
They were especially concerned by what they saw as deviant behavior by

23 Ts'ai K'an, Ting-chai chi, 1.20a. 24 Yuan Fu, Meng-chai chi, 3.32.

13
 Law and order in Sung China

members of the official elite. The high positions of such men meant that
their misbehavior, which could not be blamed on their personal suffer-
ings, would have serious repercussions on society. As a result, attacks on
individuals and on specific policies often contain more general exposi-
tions on the interrelationship of ethical orientation and social order. In
writing about practical problems, officials phrase their analysis in terms
of the general origins of disorder, without needing to describe the viola-
tions in detail. Sharing a common culture and a set of values, they did
not need to go beyond generalities. The intended audience knew full well
who the objects of attack were and the kinds of wrongs implied. In the
words of the Northern Sung official Sun Sheng:
In the world, good order and disorder stem from the rise and fall of correctness
[cheng] and deviance [hsieh] and on the appropriateness of rewards and penalties.
If the correct and the deviant are mixed together confusedly, this will give birth
to disorder. If right and wrong [shihfei] are not distinguished, then the just Way
will be lost.25
It was crucial that those with power establish policies which recognized
the distinction between the proper and the deviant, kept them distinct,
suppressed the wrong, and practiced the correct.

Responses to deviance
The general response to social disorder can take four forms - persuasion,
prevention, cooptation, and suppression. The persuasive answer consists
of and depends on education. In theory at least, most people can be
raised and socialized so that they will behave properly under most
circumstances. And even if they act improperly, weaning them away
from evildoing and back to propriety may depend heavily on their
previous training. The classic Chinese teaching tools were rites and
music: rites to teach people the proper observance of distinctions in
status and condition, and music to demonstrate the commonality that
ties all men together as parts of a harmonious whole. The preventive
answer involves both attempts to reduce socially destabilizing factors
through such measures as famine relief and attempts to make disorderly
behavior seem inadvisable, by making it more difficult to commit or
by increasing the chances of its being followed by negative sanctions.
Cooptation is buying off those committing the infractions. Only when
these strategies fail is there need for suppression.
The Sung Chinese saw the cooptation strategy as a facet of educa-

25 HCP 445.9a.
14
 Introduction

tional transformation, since they assumed that the wanted men who
surrendered in return for promised rewards had in fact been changed, a
view that reveals the deep Chinese belief that wrongful behavior was
under the voluntary control of the person in question. Thus a bandit
chief who surrendered in return for a commission in the army was seen
as someone who had "renewed himself" (tzu-hsin), at least until his
behavior proved otherwise.
Sung Chinese also saw prevention as one aspect of punishment.
Prevention through building walls or relieving the distressed was clearly
not punitive, but prevention which rested on the deterrent effect of
threatened penalties was understood to be part of the punitive side of
law enforcement. Sung officials spoke of the general approaches used to
control deviance as a combination of transformation - achieved through
the appropriate use of rites and music, which was possible as long as the
state ensured the minimal livelihoods of the people - and the punitive,
which was embodied in penalties. Effective responses to undesirable
behavior demanded an appropriate mixture of the two approaches.
Justifying a change in current bandit suppression policy, a decree from
the mid-eleventh century asked:
Why is it that the transformative power of rites and music are not used and we
simply rely on punishment? Confucius said, "If the rites and music do not
flourish, then the punishments will not hit the mark. If the punishments do not
hit the mark, then the people will not know where to put their hands and feet"
[i.e., will feel apprehensive and not know how to behave].26

Role of the emperor

In the Chinese traditional political and social schema, the highest figure
capable of creating the greatest degree of disorder through deviance, and
at the same time of bringing about the greatest good order through
proper behavior, of dealing in a fundamental way with the struggle
between cheng and hsieh, was the emperor. He was the person chiefly
responsible for providing a model of correct behavior in his own acts and
for distinguishing between the correct and the deviant in the acts of
others. In the basic traditional Chinese political myth, the emperor was
the Son of Heaven, invested by heaven with power because he was the
individual most able to provide for the welfare of humankind. This
welfare was both physical and spiritual. The emperor was a shaper of
economic and social policies, but given the technological limitations of
traditional times, he was more importantly a moral and religious leader.

26 HCP 143.29a.

15
 Law and order in Sung China

He led in part by serving as an exemplar of proper behavior. By dis-

playing his virtue he could transform the world.
This transforming power is at times described almost mystically. Most
Chinese thinkers saw the universe as innately orderly. As long as each
participant played his or her role properly, order would be maintained.
But people are possessed of free will. They can fail to play their roles
properly either because they are ignorant through lack of proper educa-
tion or because they are willfully obstructive. The roles played by dif-
ferent actors in the human drama naturally differ in their importance
and in their potential effects on the achievement of good order. As the
central actor, the emperor had the greatest effects on the universe. His
influence was so great that he could bring about good order simply by
fulfilling his proper role. Confucius reflected this belief when he noted:
"If there was a ruler who achieved order without taking any action, it
was, perhaps, Shun. There was nothing for him to do but hold himself in
a respectful posture and to face due south." 27 On a more pragmatic level
the Chinese recognized early that example, not injunction, was the most
effective way to educate. The emperor set an example, most immediately
for his court and officials. They in their turn were models for those under
them. All officials in traditional China were familiar with Confucius's
comments in the Analects. When asked by a ruler about robbers, the
Master replied, "If you yourself were not a man of desires, no one would
steal even if stealing carried a reward." When asked further about
the value of killing the unprincipled for the sake of the principled, he
answered,
What need is there for you to kill? Just desire the good yourself and the common
people will be good. The virtue of the gentleman is like wind; the virtue of the
small man is like grass. Let the wind blow over the grass and it is sure to bend.28
Ideally, this radiating process of inculcating proper behavior that
stemmed from the emperor would transform all those that it touched,
leading to general good order, marred only by the remnant group of the
truly evil. In reality, however, the state always faced a problem of
endemic disorder.

Pacifying versus suppressing bandits

Because of his mediating position between Heaven and man, the emperor
also possessed to a supreme degree the capacity to cleanse wrongdoers
and to reintegrate them into normal society. Given the emphasis on the

27 Confucius, The Analects, p. 132.

28 Ibid., pp. 115-16; quoted by Fan Tsu-yii, Fan-t'ai shih-chi, 22.11a.

16
 Introduction

division between types of bandits, between the incorrigible and the

salvageable, this power was reflected in a regular policy of summoning
bandits to surrender in return for amnesty. The way in which this
practice worked was described by the twelfth-century official Liao Kang
(1070-1143):
I have taken note of the Court's concern about the bandits in Fukien and its
desire to dispatch officials to soothe and pacify the region. . . . Such officials,
before they enter Fukien, ought first to display notices, recruiting men to enter
the area of banditry to proclaim the amnesty offered by the Court. Thus the
bandits' feelings will be moved, and they can surrender.29
In practice, forgiveness was sometimes an offer extended not just to
followers but more importantly to leaders, who on surrender might not
only be forgiven but even awarded offices and ranks. Suppression by
force was supposed to be a last resort for use against the fundamentally
recalcitrant. The Southern Sung official Wang Hsiang (doctor of letters
degree, 1153) put it this way:
I have heard that there are two basic approaches to getting rid of bandits. For
petty bandits you should seek means whereby to lead them to peace, not ways to
defeat them. For major bandits you should seek the means whereby to defeat
them, not ways to lead them to peace. Why is this? Petty bandits and skulk-
ing thieves form groups of hundreds and thousands, at times because they
have lusted after profit or because they could not endure some petty anger.
You should soothingly lead them toward peace. Thereupon, they can reform
themselves. . . . If they are major bandits, part of a network that penetrates the
prefectures and bestrides the districts, calling together a crowd without being
exhausted and pillaging the good people, these are the sort that the four bar-
barians see as harbingers of our decline. . . . They must be suppressed without a
thought of being merciful.30

Selection of personnel
The emperor had to serve as the supreme exemplar, transforming those
around him. At the same time Chinese commentators were sensible
enough to recognize that it might be more immediately important if he
chose wisely to begin with, picking moral subordinates. The emperor
was the ultimate personnel recruiter. Political struggles over the appoint-
ments of individuals were seen as specific examples of the more general
processes of selecting morally correct subordinates. In theory, the role of
the emperor - his importance in the choice of the good and the moral -
was an article of faith among Sung officials. In the late eleventh century
29 Liao Kang, Kao-shan wen-chi, SPTK ed., 1.23b.
30 Wang Yang, Tung-mou chi, SKCSCP ed., 10.13a.

17
 Law and order in Sung China

the official Fan Tsu-yii remarked that "the virtue of the king is like
Heaven. There is nothing that fails to respond to it. It is like Earth.
There is nothing that it does not sustain. Within the Four Seas everyone
is his child. No place is distant. All are one."31 In practice, for Sung
officials, the emperor exerted his power and showed his commitment to
propriety through his choice of subordinates. According to the late
Northern Sung official Liang Tao (1034-97),
For the virtue of the ruler, nothing is more consequential than knowing men, and
in the ruling of the Court, nothing has precedence over stressing those who are
sagely. Virtue has intelligence as its epitome, and ruling has solid trustworthi-
ness as its root. One who, in a judicious way, is able to distinguish between those
who are deviant and those who are correct and who promotes the worthy and
restricts the deviant, so that what is correct cannot be disordered, may be called
intelligent. 32

The emperor could lead by example, and in theory the resonance of

his exemplary influence would reform the world. As a practical matter,
however, he had to use a large number of subordinates to control an
enormously larger number of ordinary people. He needed methods
to encourage appropriate behavior and to discourage inappropriate
behavior. It was especially important that his subordinate officials
behave appropriately. The emperor had to rely on a cadre of sub-
ordinates whose values and goals, reflected in their administrative
behavior, would support the perpetuation of dynastic power, which in
the final analysis could be endangered if discontent with the government
became too widespread.

Examination system
In pursuit of this goal, traditional Chinese governments developed a
variety of policies and mechanisms. The most renowned, and in some
eras the most important of these mechanisms, was the competitive
civil service examination system. This system was unquestionably the
most sophisticated premodern device for indoctrinating a cadre of sup-
porters. This superb device promised enormous social and political (and
economic) rewards to the successful. For century after century, key social
groups voluntarily and at their own expense subjected their younger
members to a process of training in the values that the state wanted to
31 Fan Tsu-yii, Fan t'ai shih chi, 24.2a.
32 HCP 468.22b. See also the similar sentiments expressed by Liu Ch'ang (1019-68):
"The Way whereby ministers are controlled consists of separating the correct and the
deviant. Correct ministers are used as close advisers; deviant ministers are kept at a
distance." Liu Ch'ang, Kung-shih chi, TSCC ed., 31.374.

18
 Introduction

promote. The state itself did not have to bear the bulk of the costs of
this training. Although state schools did exist in the Sung period, and
perhaps played a more crucial role in the late Northern Sung than in any
other comparable era of traditional Chinese history, for the most part the
costs of education were borne by private individuals. For a trifling
expenditure the state was able to mold educational aims to its own uses.
Successive dynasties adopted and sponsored the promotion of values and
views supportive of a stable, hierarchical, and generally authoritarian
political order. By espousing views initially developed by certain groups
within the literati class and by granting political powers to those certified
to be adept at discoursing in terms of these views, the authorities were
able to promote the spread, among the literati and to a remarkable
degree among other groups in Chinese society, of an ideology whose
values buttressed the imperial system. The result was the creation of
a ruling class that acted within a shared universe of discourse, the
character of which had been heavily influenced, by if not determined, by
the needs of the central authorities.33 Those who were able to demon-
strate their mastery of this type of discourse were granted degrees by
the state, which certified them as qualified for high government office.
During the Sung dynasty such men thoroughly dominated the policy-
making levels of the government.34
The state set before the society a goal and reward for a certain style of
education. The relatively well-to-do, who bore the greatest economic
costs of education and so could set its tone and agenda, by and large
wanted to see at least some of their offspring enter government service.
In pursuit of the goal of entering the charmed circle of the civil service
elite, they chose to educate their sons in those texts and traditions that
might enable success in government. They, of course, recognized that
success in the examinations was not an everyday event and that often
several generations might pass without any family member's being
successful. Indeed, no family member might ever succeed at the final
stages of the examination process. Still, the prize was so great that a
continuing investment in pursuing it seemed worthwhile. Furthermore,
sons who were able to enter government service by other routes needed
to be able to associate in a competent way with the government elite
dominated by examination graduates. Even sons who would never
33 On the examination system, see John Chaflee, The Thorny Gate of Learning in Sung China
(Cambridge: Cambridge University Press, 1985); Thomas Lee, Government Education and
Examinations in Sung China (Hong Kong: Chinese University Press, 1985); and Miyazaki
Ichisada, China's Examination Hell, trans. Conrad Shirokauer (New York: Weatherhill,
1976).
34 See Brian E. McKnight, "Administrators of Hangchow Under the Northern Sung,"
Harvard Journal of Asiatic Studies 30 (1970): 185-211.

19
 Law and order in Sung China
succeed at the lower levels of the process needed to be trained in the
proper traditions and values if they were to be accepted as members of
the literati class, so that they could deal socially with members of
the government. For all these reasons the examination system had an
enormous impact, far beyond the ranks of those who achieved degrees.
Chinese education from the primary level upward was concerned with
inculcating a shared range of values and view of the world. In assessing
the role to be played by those, even the educated, who had not passed
through the examination system, the problem for the state was the
practical impossibility of evaluating objectively the degree to which the
appropriate values had been absorbed. However, given the curriculum,
the authorities could hope for some level of appropriate training among
the educated.

Control of the commoners

Not so for the illiterate masses. The state authorities could and did use
various devices to try to promote traditional values supportive of the
status quo, but they were always chary of their degree of success.
Various dynasties tried to stimulate the inculcation of desired values and
attitudes by rewarding individuals who displayed exemplary behavior,
by encouraging institutions like the community ritual assemblies, or by
posting moral exhortations or having edifying precepts read aloud to the
people. However, from very early in Chinese history the records reflect a
recognition that some people had internalized the desired values, views,
and behaviors either insufficiently or not at all. Some groups might fail
to be properly socialized because they were not part of the Chinese social
world. Such people, whether internal non-Chinese enclaves or foreigners
beyond the borders, became a problem for the military if they acted in
ways that threatened the state's power. But because the Chinese state
was the core of the civilized world, problems from non-Chinese sources
were peripheral. In the Chinese phrase, they were diseases of the skin,
whereas internal Chinese enemies, those supposed to be under state
control, could become diseases of the vitals. 35 Such enemies needed to be
identified and, if they could not be reformed, had to be isolated and
neutralized.

Role of law
A firm faith in the importance of appropriate education through the
provision of models of proper behavior was coupled in the minds of the
35 Compare Wang Shih-p'eng, Mei-ch'i chi, 2 . 1 0 b - n b .

20
 Introduction

Sung bureaucrats with the recognition that law enforcement also was
necessary. As the Southern Sung official Lou Yueh (i 137-1213) put it:
Confucius said, "Guide them by edicts, keep them in line with punishments, and
the common people will stay out of trouble but have no sense of shame. Guide
them by virtue, keep them in line with rites (/ia), and they will, basides having a
sense of shame, reform themselves."36 However, Mencius said, "A ruler should
take advantage of times of peace to explain government and punishments to
the people."37 Confucius considered the government and punishments to be
secondary. Mencius considered the government and punishments to be primary.
Men sometimes wonder about this, but really there is only a single Way.
Confucius did not think that government and punishments could be dispensed
with. Mencius did not think that virtue and rites could be forgotten. If govern-
ment and punishments are not cultivated, there will be no state. When they have
been established, transformation can occur.38

Chinese rulers sought to maintain power over and make use of groups
which did not always share their interests. After the rise of the examina-
tion system, the most heavily indoctrinated group consisted of examina-
tion graduates, but even their ethos was by no means always consonant
with imperial desires. Furthermore, examination graduates as officials
did not always live up to their professional ideals. The second major
group was composed of those numerous officials, a majority at all times,
who did not enter the civil service through the examinations. Their
commitments were more problematic. Even more so were the loyalty and
support of the locally influential elites. Probably most of their male
members had been exposed to Confucian education, but this certainly
did not ensure their loyalty to or compliance with the government's
wishes. Other important groups, especially the military and the common
people, were even less trustworthy.
In trying to deal with this situation - the inadequacy of educational
transformation alone to ensure compliance with imperial desires - the
emperors employed the strategies and ideas of another influential group
of classical thinkers, the fa-chia, or legalists. Written law began early in
China. 39 Long after the rise of written law, a group of thinkers began
systematically to analyze the role and uses of such law in society and
government. In the chaotic conditions of the Warring States period

36 Confucius, The Analects, p. 63.

37 My rendering basically follows the translation of Lau, Mencius, p. 81.
38 Lou Yuen, Kung-k'uei chi, TSCC ed., 20.311. Confucius's opinion is often referred to in
Sung memorials stressing the importance of educational transformation.
39 For a discussion of the early origins of written law in China, see Herrlee Glessner Creel,
"Legal Institutions and Procedure During the Chou Dynasty," in Cohen et al., eds.,
Essays on China's Legal Tradition pp. 26fT.

21
 Law and order in Sung China

(403-221), these thinkers espoused the use of realpolitik in relations

among states and the use of law in the internal control of bureaucracies
and commoners. Their proposals covered a variety of approaches to
control, two of which are of special interest to us here. First, the
legalists promoted, if they did not originate, the idea of organizing
the common people into small groups that were to be held mutually
responsible for the actions of any member. This cellular system of organ-
ization, aimed at both suppressing the deviant behavior of members and
isolating outside troublemakers, continued to be used by many Chinese
regimes down to the present. T h e Sung dynasty was no exception.
Second, the legalists advocated the reasoned use of written law as a
control device.
During the dynasties after the fall of the Ch'in dynasty ( 2 2 1 - 7 ) ,
legalism was denigrated and seen as the root of the abuses attributed
to the Ch'in. However, men of Confucian persuasion continued to be
interested in the law. Indeed, during the H a n dynasty, the most im-
portant teachers of the law were Confucian scholars. 4 0 During later
dynasties, in seeking ways to acknowledge the fundamental utility of law
while remaining critical of the use of law in practice, Chinese thinkers
developed a mythos about the evolution of the law. During the late
eleventh century the official Liu Chih summarized this conventional
wisdom:

The laws [fa] are the Great Commands of all under Heaven. When the Former
Kings systematized the law, their intention was to cause men to reform them-
selves [tzu-hsin], so that they might avoid penalties, and to make crime difficult.
Therefore the laws were extremely simple and extremely correct, and yet they
were adequate to fulfill the principle of Heaven and Earth. When in later times
men systematized the laws, they were merely fearful that the guilty might
occasionally escape. Therefore, their laws were numerous, extensive, and closely
meshed so that the people did not know where they might put their hands and
feet.41

Properly understood and used, the law could be a valuable tool for
helping people become better. Improperly understood and abused, it
could cause increased disorder.
Laws and their enforcement are questions of boundaries. Legislation
creates the boundaries by defining them, stating what acts are within
and what acts are outside the boundaries. Adjudication clarifies the
boundaries by determining whether specific actions or classes of actions

40 On the relations between the Confucians and legal thought in this period, see Ch'ii
T'ung-tsu, Law and Society in Traditional China (Leiden: Mouton, 1965), esp. chap. 6.
41 Liu Chih, Chung-su chi, TSCC ed., 6.81; HCP 373.2a.
22
 Introduction

fall within the general boundaries set by legislation. Law enforcement

provides the mechanisms for enforcing the state's commitment to its
laws, and penal policy is the menu of possible responses the state might
make if the boundaries are violated.
Legislation, adjudication, and law enforcement play these same roles
in different societies. In different societies many of the functions of laws
so created, interpreted, and enforced are similar. If we want to under-
stand how these laws worked in practice, their actual social functions, we
must examine how they were related to the promotion of state goals,
especially social and political stability. Like most laws in most times and
places, most Chinese laws were concerned with preserving the political
power of the ruling group, both directly and by promoting the interests
of other groups who would in turn support the rulers. Rules directly
supporting imperial power included not only laws concerned with sub-
version or sedition but also an enormous number of administrative rules
whose purpose was to assure those at the top that the political apparatus
would operate in ways that would preserve the power of the rulers.
Administrative rules preserve authority partly by injecting pre-
dictability into government and government-influenced operations.
Predictability permits superiors to control subordinates, by defining the
subordinates' powers and their degrees of discretion. Predictability is
also one value that the ruling group trades to those outside the state
apparatus in return for fiscal contributions, for passive acquiescence,
or for active support. To those outside the state apparatus, the pre-
dictability of state-enforced rules allows the efficient organization of their
own affairs.
Another large set of rules promotes stability by inducing large groups
in society to support the current political arrangements. Some of these
rules are designed to support the interests of influential social groups; in
the traditional Chinese case this meant, most importantly, the literati
families that were not themselves currently involved in officialdom but
that were major landowners and influential in the affairs of their home
areas.
Another set of rules was concerned with the enforcement of those
social norms that were shared widely, though not universally, within
the society, by most peasants, merchants, artisans, and urban working
people, as well as landlords and members of the political elite. One
of the goods provided to society by the political machinery was this
enforcement of widely shared social norms. Here is the domain of what
we often call "the law": Thou shalt not kill. Thou shalt not commit
adultery. Thou shalt not steal. Thou shalt not bear false witness against
thy neighbor. Thou shalt not covet thy neighbor's house. Thou shalt not
23
 Law and order in Sung China

covet thy neighbor's wife, nor his manservant, nor his ox, nor his ass,
nor any thing that is thy neighbor's. At this point the moral leadership of
the emperor coincided with the strictures of the codes. He was the
ultimate locus of the sovereign power which sought to create a morally
proper society by exemplifying propriety and punishing misbehavior.

Reward and punishment

In seeking to promote obedience to these various sorts of laws, the
traditional Chinese fell back on another of the devices advocated by the
legalists. The legalists' view of human nature was more simpleminded
than that of the Confucians: They felt that most people were dominated
by self-interest, which in practice meant that their actions would only
adventitiously be in conformity with the desires of the state. Most people
had to be encouraged to behave properly by means of rewards and
discouraged from behaving badly by means of severe penalties.
Acceptance of this idea — that positive and negative sanctions are the
key of guiding behavior — was not, of course, necessarily in conflict with
the belief that most people have within them the basic stuff of goodness.
Most members of the Chinese elite in Sung times had been socialized to
accept both the Mencian dictum that every man by nature contained
the roots of goodness and the conflicting, but not contrary, belief that
the composition of men also supported a tendency toward deviant
behavior.42 This formulation had not been clearly articulated at the
beginning of the Sung dynasty and, indeed, is missing from the works of
the earliest of those whom we classify as Neo-Confucians. However, in
the works of Ch'eng Yi (1033-1107) the concept of li,h the pattern which
runs through all things and joins man to the universe, is given the
central position among Confucian concepts. This new formulation was
the foundation on which Chu Hsi built his ideas about human nature
which were to remain orthodox until the end of the empire. As used
by the Ch'eng-Chu school, lih is always good, perfect, and unchanging.
If it is fundamental to man, as to all else, what then can explain
deviance? The key here is the concept of ch'i, material force, the stuff
through which the disembodied lih takes shape and form. Ch'i can be
pure or muddied, and so the behavior of things can also be pure or
muddied.
Therefore, when generating men and other things, there is a difference of coarse
and fine. Speaking in terms of the ch'i as one, both men and other things are
42 For a discussion of this problem, see A. C. Graham, "What Was New in the Ch'eng-
Chu Theory of Human Nature?" in Wing-tsit Chan (ed.), Chu Hsi and Neo-Confucianism,
pp. 138-58.

24
 Introduction

generated by receiving this ch 'i. Speaking in terms of the coarse and the fine, men
get ch'i that is well adjusted and permeable, and other things ch'i that is ill
adjusted and impeding. Because man alone gets the well adjusted, the li b in
permeating him is nowhere impeded; because other things get the ill adjusted,
the lih being impeded, they know nothing. . . . But speaking in terms of man's
own endowment, here too there is a difference of dull and bright, clear and
murky. Therefore the highest in knowledge, who know from birth, are so con-
stituted that their ch'i is clear, bright, pure, choice, and without a trace of the
dull and murky. As for the next in degree, who are second to those who know
from birth, they have to learn before they know, have to practice before they
attain.43
The Sung official Hua Chen (i051-after 1110) indicates the way in
which a Confucian belief in the possibilities of goodness in men could be
combined with an acceptance of the necessity of using a legalist policy of
rewards and punishments to deal with the majority:
The elite man [shih]e acts from righteousness, and his motivation derives from
this. He does not act with his eye on benefits. He does not smell rewards and feel
thereafter encouraged. He does not become mired in unrighteousness. He does
not see injury and avoid it. He does not wait to be penalized and afterward stand
corrected. The rules of reward and punishment may be engraved, but they are
not used. Such men are thus. . . . Middle-grade men, seeing their own good, are
moved to action. Seeing what will harm them, they are thereupon corrected and
refrain from bad behavior. Therefore even the Sage Emperors Yao and Shun
[mythical rulers whose traditional reign dates are 2357 to 2205] could not govern
without rewards and punishments. Indeed, if you have rewards without punish-
ments, then impropriety will gradually come to be tolerated. If you have punish-
ments without rewards, then through stinginess goodness will be lost. If rewards
are not awarded for goodness, then the people will harbor doubts and will not
feel admiration. If punishments are not carried out against the guilty, then those
below will be suspicious and yet not stand forewarned.44
And in the mid-thirteenth century the official Yuan Fu (doctorate of
letters degree 1214) remarked:
Whenever we hear legal cases, we can see that the people love money and for-
get about righteousness. They rely on forcefulness and make light of life. They
violate the law and transgress against principles. Certainly we must be cau-

43 Chu Hsi, Chu-tzu jyu-lei, 4.10b. Chu Hsi was the Sung Confucian thinker who explored
this problem most thoroughly. He and like-minded Neo-Confucians did arrive at a
widely accepted solution. His thoughts on this problem are reflected with especial
clarity in the collection of his supposed discussions with disciples. These discussions,
though not as reliable as the works written by Chu Hsi himself as sources for evaluating
his philosophical ideas, are valuable here because of their attempts to deal with
concrete problems and examples of behavior.
44 Hua Chen, Yun-ch'i chi, SKCSCP ed., 17.8a.

25
 Law and order in Sung China

tiously apprehensive, thoughtful morning and evening on how we may use

customs to influence such people. By rewarding the good, we can cause the evil
to feel shame. Men's hearts are not fundamentally evil. The principle of Heaven
is not basically unclear. 45

In the 1040s, when the state was faced with disorders associated with
the border wars in the northwest, the Sung statesman Sung Ch'i (998-
1061), borrowing from the legalist text Han Fei Tzu, stated with great
simplicity the basic position of most of the Sung elite: To control the
people, the Son of Heaven had only two handles, "punishments and
rewards, that is all." 46 Sung's contemporary YCi Ching (1000-64) noted
that "the Court controls the world through the system of rewards and
punishments. Nowadays, the empire being large and the officials lax, the
people gather together and become bandits. Punishments do not suffice
to stop them. This is because the system of rewards and punishments
has not been practiced." 47 A century later the official Wang Chih (1127-
89) wrote that he "had made a study of the methods of opposing and
controlling robbers that have been in use historically. Although they
are detailed and numerous, the important points do not go beyond
using punishments to serve as warnings, and rewards to serve as
enticements." 48
This stress on rewards and punishments is so widespread as to be a
fundamental article of faith. The Southern Sung official Wu Ching
(1125-83), who had himself seen the problem firsthand while serving
as a district sheriff and a district magistrate, commented that if the
authorities wanted to capture bandits, there was only one answer, the
effective use of rewards and punishments. 49 "Why are there so many
troublesome bandits?" the great Confucian scholar-bureaucrat Ou-yang
Hsiu (1007-72) asked in the middle of the eleventh century, and
answered his own question: "Because the state is ill prepared, and
rewards and punishments for the responsible officials are not enforced." 50

45 Yuan Fu, Meng-chai chi, 3.32.

46 Sung Ch'i, Ching-wen chi, TSCC ed., 25.355.
47 HCP 141.16a (1043). A few months later he rephrased the same sentiments (HCP
150.5a). See also Chao Shan-k'uo, Ying-chai tsa-chu, SKCSCP 1975 ed., 1.2b; Hsu
Ying-ling (doctorate of letters 1224), Tung-chien chi, SKCSCP ed., 8.ia~4b; T'ang
Keng, Mei-shan T'ang hsien-sheng wen-chi, SKCSCP 1976, 1.15a; Ch'en Ch'un, Pei-ch'i
ta-ch'tian chi, SKCSCP 1973, 45-5a-b; Ts'ai Hsiang, Tuan-ming chi, SKCSCP ed.,
9.i9b-2ia.
48 Wang Chih, Hsiieh-shan chi, 3.24.
49 Wu Ching, Chu-chou chi, SKCSCP 1973, 8.3b~4a. See also Wang Chih-tao (1093-
1169), Hsiang-shan chi, SKCSCP ed., 22.5a, who said that "the problem today is not the
border raiders or the bandits; it is that the rewards and punishments are
inappropriate."
50 HCP I4i.i4b-i5a.

26
 Introduction

The use of rewards and punishments was universal, running through

all aspects of Sung governmental operations, and could affect anyone
inside or outside the government, but Ou-yang Hsiu quite properly
stresses its relevance to officials. Without appropriate and timely official
actions, the whole edifice of government would collapse. Many if not
most of the important policy decisions of Sung rulers were aimed at
establishing a system that would place the right officials in the right
positions and also would motivate them to act appropriately. An elab-
orate system of rewards and punishments was used as a basic device for
motivating their behavior.
The rewards were potentially of considerable significance to the
officials, because they might affect their advancement within the civil
service; the penalties were mitigated by a system of official privileges.
These privileges, which varied with rank, assured active officials and
some other persons closely tied to the government that in many instances
they could avoid the regular system of penalties. There were, however,
circumstances and classes of crimes that might bar them from using their
privileged status.

Rewards and the state's fiscal situation

The ability to use rewards and punishments is intimately connected with
the fiscal strength of the state. The unwritten first commandment of any
government is preserve thyself. To preserve themselves, governments
must mobilize resources sufficient to resist attempts to overthrow them.
These resources are used to reward supporters, to mollify the dis-
contented, to prevent those not mollified from becoming dangerous, and
to destroy or at least to fend off those who do become dangerous. Within
limits set by revenues, technology, ideology, organizational sophistica-
tion, and environmental constraints, governments seek to maximize the
number and power of their supporters, minimize the number and power
of their inveterate opponents, and neutralize those in between. To do so
they must manipulate both values and valuables, both ideas and wealth.
The use of values involves governments in instilling, or in encouraging
and supporting others who instill, values supportive of the regime's
continued power. The use of wealth includes both financing the means
for suppressing enemies and purchasing the allegiance or at least the
toleration of the majority of society.
If all people could have been properly socialized, so that they always
acted in accord with the immanent order of the world, good order
would have reigned. Because the political order was a natural part of
this natural order, people would also have absorbed beliefs, values,
27
 Law and order in Sung China

behaviors, and views of the world that supported the imperial system.
The state would have had no opponents, so there would have been no
need for the less savory side of government, that which bribes and per-
secutes. But because all of the people were in reality not fully inculcated
with the traits supportive of this system, the use of wealth to subvert,
control, or destroy opponents was vital to the continuation of govern-
ment power.
Like other premodern states, the Sung government was able to secure
only a small fraction of the gross national product for its use. Even if the
authorities had wanted to increase their revenues, the technological basis
for greatly increasing the proportion of the social product flowing to the
government did not exist. The Sung empire was a remarkably effective
fiscal organization by traditional standards. It apparently was able to
secure a considerably larger share of the economic product of society
than were other Chinese dynasties. Still, by modern standards it was
very limited in the proportion of national income that it could secure
and distribute. The total social output had to be shared by the pri-
mary creators - who at worst had to be kept above the level of bare
subsistence - by the rentier class, by the local political machinery, and
by the central government.
Moreover, the ability to gain and distribute wealth was limited
not just by technical and financial factors but also by the very break-
downs in the government-supporting social order that made this use of
wealth necessary in the first place. Many of those charged with securing
resources for the state's use were themselves inadequately committed to
its interests. Given the opportunity, they would divert to their own uses
those revenues supposedly destined for the state. Therefore the state had
to depend on, and within limits set by its capacities and resources seek to
promote, the appropriate socialization of the people under its jurisdic-
tion, but at the same time it had to prepare to defend its authority (and
its taxing abilities) through the use of force.

Law enforcement
Dealing with criminals, particularly with those who are incorrigible,
raised the classic conundrum of law enforcement: How are we to find out
that the law has been broken, and having found out, how are we to get a
sufficient number of effectively led, adequately armed, and properly
motivated men to the needed spot at the critical moment, at a cost we
can afford, given other demands on our resources? The statement of the
problem of law enforcement is deceptively simple. The practice of law
enforcement is very complex, because the components are interrelated
28
 Introduction

and interdependent. The actual functioning of the law-enforcement

system depends on both general conditions and the particular culture,
place, and time.
The authorities want to bring to bear just enough force to accomplish
their goals. Too little will not succeed in suppressing the disorder. Too
much will not only be economically wasteful but also may entail further
problems. Getting this right amount of force to the needed place at the
right time entails a whole host of problems. The answers to these
problems depend in part on the technologies of transportation, com-
munication, and armaments. How can we get accurate information
about violations of the laws in a timely manner so that we can respond
effectively? Who will give us the needed information? And what will
move them to do so? To whom should such informers report? Having
found out about crimes, how do we get orders to the appropriate law-
enforcement agents? When these agents have been given their orders,
how do they get to the proper place? And having gotten there, what
weapons do they use to suppress the violations? How many agents will
be needed? Who should lead them? And perhaps most crucial of all, how
are we to motivate both law-enforcement leaders and their subordinates
so that they will use their control over force in proper ways for the
ends that we seek? Finally, how are we supposed to pay for all this
activity, without diverting to it resources that are needed for other social
purposes?
A weakness in any link of this chain of actions can jeopardize the
whole enterprise. Ordinary people may know of violations and not tell.
State-mandated mutual security groups may be informed but not pass
on the information to the officials. The officials may be aware of the
problem but not prosecute. 51 When finally called in, law-enforcement
agents may be unwilling or unable to respond or may respond in exces-
sive or dangerous ways.
All parts of the general problem statement are important, but two
have precedence. First, the law-enforcement system must be fiscally
tolerable. It will do us no good to understand what is needed if what is
needed costs more than can be afforded. The ability of the authorities to
squeeze revenues out of the populace has an upper limit. To go beyond
it is to risk both the long-run impoverishment of society and, more
immediately, the revolt of those being squeezed by the government. The
Sung period, especially the Northern Sung, was fortunately an era of

51 See, for example, a memorial by the official Chou Tzu-chih. In a report on the frequent
practice of infanticide, he noted that "the local security organizations know but do not
report. The officials learn of it but do not prosecute." Chou Tzu-chih, T'ai-ts'ang t'i-mi
chi, SKCSGP 1971, 49.3a.

29
 Law and order in Sung China

rapid economic growth. Despite this, the government had serious fiscal
problems by the middle of the eleventh century.
The proportion of revenues which a government can devote to law-
enforcement organizations also has a ceiling set by competing demands
for funds that the state simply cannot ignore. In China, as in most
civilizations, a substantial amount of the revenue was spent on external
defense. During the Sung this might run to half the total budget. Large
amounts of revenues also went to support the bureaucracy and the
Court, leaving relatively little for other purposes. The state also had to
devote some revenue to raising still more revenues and some to generat-
ing or retaining the support of people not part of the formal government,
by promoting the interests of various influential groups.
Furthermore, even in a mythical world in which the state apparatus
had access to limitless resources, the civil elements in its structure would
set limits on the resources given to the professional users of force,
whether military or police, if only to protect themselves. The problem of
who will guard the guardians is a universal one. Intelligent civilian
politicians recognize that an excessively powerful military or police
establishment is a constant danger. The Sung rulers were especially
sensitive to this issue, because they felt that the disasters of the late
T'ang (618-907) and the Five Dynasties (907-60) had stemmed from
the military's possession of too much power.

Civil and military

The effective use of rewards and punishments was central to motivating
officials and also to motivating and controlling the people. Two principal
types of agencies, the civil and the military, embodied the state's poten-
tial for rewarding or punishing. The question of the appropriate rela-
tionship between the civil and military authorities in maintaining order
goes to the heart of deeply rooted traditional Chinese attitudes toward
the state, its government, and the appropriate tools to be used in
governance. The basic Chinese political myth, enshrined in the Classics
and developed by the great Confucian thinkers, held up as a model
those rulers who were able to transform men primarily by example and
benevolence and who resorted to force sparingly, with complete justice,
and as a last resort.
Force was necessary. Even the Sage Kings of antiquity had used it on
occasion. Some individuals and groups who refused to accept their
appropriate place in an orderly world were large and radical enough in
their defiance of propriety to require suppression by violence. But the
victories won through violence were felt to be more fragile than those
30
 Introduction

that followed the conversion of the recalcitrant into good subjects by

means of persuasion and education.
The balance between military and civil which corresponded partly,
but not wholly, to the distinction between the use of force and the use of
gentler measures was always delicate. If the central government moved
too far toward emphasizing the civil, it risked weakening its ability to
withstand violence from within and without. If, as in the late T'ang and
the Five Dynasties, the military came to dominate politics and the
central authority lost its control over large regions, it became incapable
of promoting in those areas the kind of educational and persuasive
socialization that was needed for the long-term stability of the central-
ized political order.
To deal with the immediate problem - the decentralization of power
during the preceding era — and to respond to the overweening role of the
military in the governments of his predecessors, the founder of the Sung,
T'ai-tsu (r. 960-76), strengthened the civil side of local government.
He and his successors set out to solve certain pressing practical prob-
lems. The policies they established reflect both their assessments of the
immediate situation and their fundamental beliefs about the roots and
nature of deviance, as well as about the appropriate roles of the civil and
the military.
The Chinese viewed (and continue to view) deviance as a continuum
measured by the degree of the threat to the existing order of society.
At the low end of the continuum, at which the deviance does not
constitute an immediate threat to the social order, the proper response is
persuasive and educational. First within the family and then at the level
of the village, disputes and deviance are handled through discussion and
persuasion if possible, but by some application of force if necessary. This
level is largely the responsibility of social associates.
As the degree of threat increases, the state begins to apply measured
force through the police function of the civilian government. Even at this
level there is a preference for the mediation of disputes. In modern
America the police are frequently involved in the informal mediation of
disputes; in Sung China the police apparatus would not normally have
been involved in such mediation, which was the preserve of nongovern-
mental groups (though in contemporary China the police may indeed be
involved). When informal mediation by extragovernmental agents fails
to resolve the problem, the next step is often the courtroom. However, in
traditional China, cases which did arrive at the lowest levels of the
hierarchy of courts were often sent back by the presiding magistrates
with an order that they be settled out of court. If mediation fails and the
threat of disorder is serious, formal police action becomes necessary.
31
 Law and order in Sung China

Finally, when the threat is too immediate and too extreme for the
police agents of the civil authorities, it becomes the responsibility of the
military.
In general, this sort of response is hardly unique to China, ancient or
modern. However, the particular forms it took in traditional China were
conditioned by both the given conditions of a premodern peasant society
and the cultural inheritance of the Chinese tradition. In promoting law
enforcement, one obvious response to the technological limitations of
premodern communications and transportation would be to disperse
agents for enforcing order at many points within the area to be con-
trolled. Ideally, this could be done in such a way that no spot was too far
removed from the forces of law and order to be helped by them in need.
However, if these agents were to be effective, they had to be kept in such
concentrations that they would have sufficient collective force to respond
to probable threats. And if they were to be so constituted that they could
devote their time largely to enforcing order, they would have to be
remunerated.
In some times and places the central authorities dealt with these
dilemmas by allowing the local authorities to own the means of admin-
istration, a feudal system. The Chinese authorities in imperial times
chose a different course. They allowed the local authorities to deduct a
proportion of the tax revenues for their own use, including the support of
a police apparatus, while reserving for the central authorities enough
funds to permit the creation of a military force sufficient to prevent
rebellion by local authorities or malcontents.
Allowing local authorities to retain income lowered the handling costs
that would have been entailed in a system of central distribution, but
it carried within itself the risk of the growth of local power centers.
Centralizing revenues would have reduced the dangers of local resistance
to the center but would have greatly increased the cost of administering
the tax system. The long-lived Chinese dynasties generally dealt with
this problem by playing a delicate balancing game, allowing the local
administrative units to retain some income but directing much of it to
the central authority.

Control of local functionaries

This pattern of income distribution highlights one of the central political
problems of traditional empires, how to ensure adequate levels of per-
formance by their functionaries, especially their local functionaries, who
were serving in posts far from the watchful eye of the central govern-
ment. At the same time, local organizations had to be effective if the
32
 Introduction

central government was to remain healthy and yet under control, so that
they could not threaten the center. Local officials might be interested in
increasing their own personal power, or they might act in collusion with
influential groups among the commoner populations in their jurisdic-
tions, against the interests of the centralized state.
Traditional Chinese states responded creatively to this dilemma.
The response involved attempts both to instill internal controls in
the individuals in question and to detect and suppress bureaucratic
deviance. The examination system, with its assumption that knowledge
of the Confucian Classics was a surrogate measure of ethical commit-
ment, is the outstanding example of the internal emphasis. The elaborate
bureaucratic organization created by traditional Chinese dynasties
provided the mechanism for exercising external controls on behavior.
This bureaucratic pattern served the central authorities' control pur-
poses in several ways. By specifying functions it made possible the
objective evaluation of performance. An official who began to abuse his
role by going beyond his specified functions thus could be stopped early.
It set minimum standards of performance so that ineffective officials
could be detected. Such a centralized bureaucracy also was designed to
tie the long-term interests of the bureaucrats to the health of the central
apparatus, by integrating a system of rewards for loyal service. Finally,
the government established surveillance agencies that oversaw the work
of subordinate personnel, to detect wrongful or ineffective perform-
ance, and to make recommendations and evaluations for rewards and
punishments.
This system, like all such systems, can be viewed as an information-
processing system. At the bottom the local functionaries generated a
mass of detailed information about conditions in their area and sent this
up through the hierarchy. Those at the top evaluated this information
and sent down a smaller stream of more general guidelines for behavior,
as well as specific orders for dealing with individual situations. There
was always a tension between the desire of the center for predictability
and its need to generalize on the basis of what it perceived as common
problems and conditions, and the perception by the local authorities of
the distinctiveness of local circumstance and their desire for a degree of
autonomy that would allow them to respond to unusual situations. The
traditional Chinese imperial state responded to this dilemma of core
versus periphery by creating general guidelines, while in practice allow-
ing a degree of local flexibility in their enforcement.

33
 2
The historical context

Historical situation
In some fundamental ways traditional Chinese society was extra-
ordinarily stable. At no time is this more striking than during the Sung
dynasty. This was a time of startling technological innovations, of great
creativity in the arts, sciences, and philosophy, a period during which
the basic claim to political legitimacy was successfully challenged for
centuries by the continuing presence of powerful Sino-foreign states to
the north, a period of major population shifts and growth, and a period
of enormous increase in both the numbers of urbanized areas and the
urban character of the cities. And yet the structures of Chinese society,
its basic values and attitudes, survived this kaleidoscope of intellectual
and economic change, foreign competition, and scientific advance.
Why? What were the strengths of the underlying social structures which
allowed Chinese society to ride out centuries of radical change with
relatively little alteration in the basic commitment to what it was to be
Chinese? Certainly one element was the strength of those structures
in Chinese society that were concerned with inducting new members
(whether the young or significant outsiders) into the central traditions of
the society. This crucial aspect of social control, the pattern of socializing
people into being "Chinese," is beyond the scope of this book. The
subject of this work, the other side of social control, which buttressed
socializing agencies by placing the weight of the law and law enforce-
ment behind them, can be understood only against the background of
the problems faced by the people of the Sung and the changes with
which they had to wrestle.
The foremost task faced by the Sung founder and his immediate
successors was the reassertion of central political power. From the
middle of the eighth century the power of the central government in
many local areas had begun to decline. Despite the partial recovery of
34
 The historical context

power during the late eighth and early ninth centuries, the T'ang (618-
907) central government never recaptured full control of all of the
empire. The rebellion of Huang Ch'ao in the 870s effectively destroyed
T'ang central control, though the government dragged on in name for
several decades before formally ending in the early tenth century. The
most important agents in weakening the Court were the military com-
missioners, the major military commanders stationed in north China,
each of whom held concurrent control over the civil administration in
the key prefecture under his jurisdiction. During the closing decades of
the T'ang, these military commissioners usurped power in their jurisdic-
tions. This process of political decentralization continued during the
early Five Dynasties (907-60). However, by the middle of the tenth
century, despite the appearance of continuing discord, the rulers of the
Five Dynasties states had begun to develop policies to recapture power.
By the Later Chou period (951-60) the basic tools for reasserting con-
trol were already in existence. Only his untimely death prevented the
Later Chou emperor Shih-tsung (r. 954-59) from completing the process
of recentralization. When in 960 Chao K'uang-yin usurped the throne
and founded the Sung dynasty, the stage thus was set for the building of
a strong central government.
The Sung founder, Chao K'uang-yin (r. 960-76), and his immediate
successors on the Sung throne had as their overriding priority the re-
assertion of central authority, against the other states occupying what
had been T'ang territory and against the military commissioners who
still retained considerable internal power. Early Sung policies governing
law and order have to be seen as one aspect of a larger policy aimed at
reestablishing full central government control in local areas. From the
late T'ang the military commissioners had controlled the principal local
government functions in their jurisdictions, including the collection of
taxes, defense, and the maintenance of law and order. l On the local level
they had inherited the old T'ang structure under which the key local
officials were the prefects and district magistrates, and the actual main-
tenance of local order was the responsibility of a civil service subordinate
of the magistrate called the sheriff (hsien-wei). In subverting this inherited
system, the military commissioners retained most of the T'ang offices in
vestigial form but appointed their own personal followers to the positions
of real power, as garrison troop commanders (chen-chiang) in the local
garrison towns (chen).

1 On criminal justice and law enforcement at the end of the T'ang and during the Five
Dynasties, see Muronaga Yoshizo, "Godai gunbatsu no keigoku kiko to setsudo
saibanken"(The standard jurisdiction and system of criminal law of the military cliques
in the Five Dynasties) Kyushu daigaku bungakubu, Toyo shusaku 28 (1965): 61-75.

35
 Law and order in Sung China

During the late T'ang and Five Dynasties periods, the garrison towns
evolved from military outposts located at strategic points into multi-
functional urban centers. Because they were often located at nodes of
transportation such as fords, bridges, river mouths, or passes, these
garrison towns became increasingly important as economic centers. At
times, they rivaled the older subprefectural seats. The military com-
missioners extended their control into the local areas by dispatching
retainers appointed as garrison troop commanders (chen-chiang) to over-
see these towns and the surrounding countryside. These commanders,
often close personal associates of the military commissioner, ruled their
areas using dispatched units of the commissioner's army. They fulfilled
all of the functions of the district administrators: collecting taxes; seizing,
trying, and punishing criminals; and providing military protection of
their domains. 2
Early in his reign the Sung founder began to reemphasize the role of
the regular civil system. In 962 he ordered the establishing of a district
sheriff (hsien-wei) to maintain order in the district, including its rural
areas. 3 The Sung office of hsien-wei was thus to be similar in function to
the office of hsien-wei as it had existed from the Han period to the time of
the Emperor Wen of the Sui (r. 590—604), and not to the office of that
name as reformed by Emperor Wen and as continued in use to the end
of the T'ang. 4 The garrison troop commanders retained control only
over disorders in the garrison's environs. 5

2 On garrison towns and their commanders, see Hino Kaisaburo, "Godai chinsho ko" (A
study of the garrison commanders of the Five Dynasties), Toyo gakuho 25 (February
1938): 54—85; Hino Kaisaburo, "Todai hanchin no bakko to chinsho" (The dominance
of the military commissioners in the T'ang, and the garrison commanders), Toyo gakuho
27 (February 1940): 153-212; Umehara Kaoru, "Sodai chiho shotoshi no ichimen, chin
no hensen o chushin to shite" (One aspect of the small urban markets of the Sung
Dynasty - with emphasis on the transformation of the garrison town), Shirin 41 (1958):
475-91; Sudo Yoshiyuki, "Godai no setsudoshi no shihai taisei" (The system of control
of the regional commandants of the Five Dynasties), Shigaku zasshi 61 (April 1952): 1-42
and (June 1952): 20-39; Edmund Worthy, "The Founding of Sung China: 950-1000"
(Ph. diss., Princeton University, 1976), esp. chap. 6.
3 HCP 3.123b-14a. See also Hsu Sung, ed., Sung huiyao chi-kao (hereafter cited as SHY)
(Taipei: Shih-chieh shu-chu, 1964), chih-kuan 48.92a. These two works contain by far
the largest body of material on the post of sheriff and so have provided much of the
information for my study. In the following notes, when figures appear in parentheses
after source citations, they give the Western-style year date of the memorial or order. See
also Ma Tuan-lin, Wen-hsien t'ung-k'ao (hereafter cited as WHTK) (Taipei: Kuo-hsiieh
chi-pen ts'ung-shu edition, 1968), 63.1467. The post of sheriff itself first appeared in the
Ch'in dynasty. Pan Ku, Han shu, Po-na ed., 19 shang, mentions several oflices associated
with law enforcement whose titles included the term wei and which began in the Ch'in.
4 On the hsien-wei in the T'ang and earlier, see Tonami Mamoru, "Todai no ken-jo" (The
hsien-wei of the T'ang dynasty), Shirin 57 (1974): 79-105.
5 HCP 3.13b-14a. A few years later the tie between the military commissioners and the
garrison troop commanders was broken when the military commissioners were ordered no

36
 The historical context

To appreciate the impact of this order, we need to keep in mind the

existing local political situation. As the Sung regime assumed control
over areas in south China, it replaced the officials and officers of the
preceding states with its own personnel. Since the south was largely free
of military commissioners, this meant in effect that local control was
immediately assumed not simply by the Sung regime but by the Sung
civil authorities. In north China, however, there were some forty military
commissioners. Although the powers wielded by the remaining com-
missioners were much weaker than they had been earlier, they still were
able to exercise considerable influence over local affairs. Thus, for the
south the impact of the decree of 962 was principally to give the newly
appointed district magistrates a mandated civil service subordinate to
control police duties. In the north this decree also subverted the powers
of the military commissioners by greatly reducing the powers of their
subordinates, the garrison troop commanders.
The early Sung emperors also used other policies and devices to
diminish the influence of the military to manageable proportions. In a
famous incident the Sung founder began this process by withdrawing
power from the hands of his leading generals, pensioning them off
handsomely, and seeing to the welfare of their descendants. 6 The Sung
government also instituted a system under which officers and men did
not serve together for extended periods, to preclude the growth of strong
bonds of solidarity. These policies, plus a system of rotating commands
and a skillful use of the law, succeeded to a striking degree. Of all the
long-lived dynasties in Chinese history, the Sung dynasty is the only one
never threatened by internal unrest among the military.

Foreign problems
This treatment of the military, though it did respond effectively to
serious internal problems, actually increased the external dangers
threatening the state. When the Sung founder, Chao K'uang-yin, who
had been serving as the "controller general before the palace," displaced
the last boy ruler of the Later Chou and ascended the throne, he
succeeded to control over only a part of the North China Plain. The

longer to appoint their followers to these posts. Such posts were to be filled only by men
from the prefectural staffs. See SHY, chih-kuan 48.92a. We do have reports, however,
that the commanders continued to seize "village bandits" in the encampments in Hsin,
Li, and K'uei prefectures and to hear complaints. See HCP 61.2b. They also continued
to hear cases and punish offenders, at least as late as 1013. See HCP 80.9b and also
SHY, chih-kuan 48.92a.
6 There are various descriptions of this famous series of incidents in secondary works. See,
for example, Worthy, "The Founding of Sung China."

37
 Law and order in Sung China

regions beyond his control held both opportunities and dangers. All of
south China was occupied by a group of long-lived states called the Ten
Kingdoms. To the north the Sung position was threatened by a powerful
state called Northern Han located in what is today Shensi Province. To
the northwest the Sung forces faced a state dominated by the Tibetan
group called the Tanguts, a state that in the eleventh century was to
assume the name of Western Hsia. To the north and northeast the Sung
territory was bounded by the Khitan-dominated Liao state.
The Tangut state, a cultured Buddhist state with an economy based
on irrigation agriculture, was primarily interested in self-preservation.
Despite the propaganda of Sung officials, it was a threat to Sung power
only when the Sung leaders sought to destroy it. 7 Indeed, the first
Sung-Tangut war began in 982 when the Sung attempted to bring the
western state into subjection, and it continued with some interruptions
into the eleventh century. In 1006, during the reign of the third Sung
emperor, Chen-tsung (r. 997-1022), this war was settled by a treaty
which inaugurated an era of peace lasting until the 1040s.
The territory controlled by the Western Hsia was a temptation to the
Sung leaders but not a threat. During this same era, however, a truly
dangerous threat existed to the north and northeast. There the Sino-
foreign state of Liao, dominated by the Khitan tribes, controlled one of
the most formidable armies of its age. 8 The Liao state, although its
center of power lay to the north of the present Great Wall, also held
territory south of the wall, centering on the area of present-day Beijing,
and governed a population that included almost 2.5 million Chinese. 9
The Liao forces were an acute danger to the Sung state because no
natural defensive barriers lay between their area of control and the
Sung capital at Kaifeng (see Map 2.1). The authors of the Liao History
estimated that the Liao military machine numbered more than 100,000
cavalrymen. 10 In an age of cavalry superiority (in appropriate territory),
this was a truly formidable force. Defense of the Sung capital thus
required the use of very large standing armies, to be placed across the
path of a possible Liao advance. The second emperor had attempted to
drive the Liao from their foothold within the Great Wall. The third

7 Evgenij Kychanov, "Les Guerres entre les Sung du Nord et le Hsi-Hsia," Etudes Song
(The Hague: Mouton, 1971), 1st series, no. 2, pp. 103-18.
8 Karl A. Wittfogel and Feng Chia-sheng, History of Chinese Society: Liao (Philadelphia:
American Philosophical Society, 1949). For further information on Sung-Liao relations,
see especially Jing-shen Tao, Two Sons of Heaven (Tucson: University of Arizona Press,
1988).
9 Wittfogel and Feng, History, p. 58.
10 Ibid., p. 515.

38
 The historical context

Scale 1:15,000.000

a
» State boundary

Map 2.1. Sung, Liao, Hsi Hsia, and Koryo borders in the eleventh century.
Based on Albert Herrman, An Historical Atlas of China, new ed. (Chicago: Aldine,
1966), p. 35.

39
 Law and order in Sung China

emperor, Chen-tsung, also involved the Sung armies in wars against the
Liao but in the end was forced to conclude a humiliating peace.

Political history
This third emperor, Chen-tsung (r. 997-1022), was the first Sung ruler
to have been born in the palace. 11 Conscientious but in his later years
mentally unstable, Chen-tsung fortunately was able to rely on the judg-
ment of the statesman Wang Tan (957—1017), who served as a coun-
cillor of state for seventeen years (eleven of them as chief councillor).
Chen-tsung, embarrassed by the demeaning peace treaty concluded with
the Liao dynasty in 1004, fell under the spell of the official Wang Ch'in-
jo (962-1025), who encouraged the emperor to perform a series of
elaborate rituals as a way of reasserting Sung values. 12 Early in 1008
when a message from "Heaven," in the form of a yellow silk scroll, was
found hanging from a roof tile of a gate of the imperial palace, the
emperor was prompted to begin an increasingly expensive series of
religious observances. Despite the more and more bizarre behavior of
Chen-tsung during his later years (which the cautious Wang Tan did not
oppose), his reign and the early years of the reign of his successor Jen-
tsung (r. 1022-63) were a prosperous and peaceful era.
Jen-tsung was the longest-reigning Sung monarch. During the first
eleven years of his reign, he was under the tutelage of the empress-
dowager, nee Liu. After her death he himself held power for another
thirty years. Impetuous in his youth, he became more cautious with age
but ruled without great firmness and with limited understanding. His
reign is a watershed, marked by renewed warfare and the growth of
serious fiscal problems. In 1038 when the Western Hsia ruler took the
title of emperor, the Sung armies renewed hostilities. Four years of
warfare was followed by a peace in which both sides compromised. This
war had cost the Sung state dearly in terms of the level of internal peace
within the empire and the fiscal health of the state. These problems
encouraged a cadre of thoughtful officials to continue their rethinking of
the meaning of the Confucian heritage and to propose reforms of the
state based on their new perceptions.
The reform movement was largely unsuccessful but set the stage

11 The interested reader can find more information on most of the individuals mentioned
in the general description of Sung developments in Herbert Franke, ed., Sung Biographies
(Wiesbaden: Franz Steiner Verlag, 1976).
12 For a description and a sympathetic interpretation of these incidents, see Suzanne E.
Cahill, "Taoism at the Sung Court: The Heavenly Text Affair of 1008," Bulletin of
Sung-Yuan Studies, no. 16 (1980): 23-45.

40
 The historical context

for more radical reforms later in the dynasty. Jen-tsung's successor,

Ying-tsung, was sickly and reigned for only four years, from 1064 to
1067. Like Jen-tsung, he relied heavily on distinguished officials of long
service. In 1067 Ying-tsung was succeeded by an emperor of a very
different cast of mind. Shen-tsung (r. 1067—85) was an energetic and
opinionated emperor. He inherited an empire faced with an accumula-
tion of internal fiscal problems; he dreamed of a strengthened state that
could recapture the lost territories then under foreign control. Recogniz-
ing that the conservative policy of more of the same would never be able
to solve the problems facing the state, he threw his support behind a
program of radical reforms shaped by the statesman Wang An-shih
(1021-86). In a variety of ways these reforms were aimed at creating a
fiscally healthy state, a prosperous population, and an effective but less
costly military machine.
Despite a period of reaction during the minority of Shen-tsung's suc-
cessor, Che-tsung (r. 1085-1100), the reforms inaugurated under Shen-
tsung were continued in part and were reinstituted when Che-tsung
took personal control of the administration upon his coming of age.
Unfortunately, in the political struggles which marked the era of reform
and counterreform, the earlier policy of treating political enemies with
some care and respect gave way to vicious factional infighting. This
struggle left a heritage of bitterness that was to help shape internal
politics for the remainder of the dynasty.
Che-tsung died in 1100. His place as ruler was taken by a man who, at
least in legend, was far more concerned with aesthetics and artwork than
with governing the empire. During most of his reign this emperor, Hui-
tsung (r. 1100—26), kept as his chief minister a man who initiated a
series of institutional reforms in the tradition of Wang An-shih while
mercilessly persecuting those who opposed him.
Although the settlement with the Western Hsia in the 1040s had
promised an era of peace, Sung leaders were never able to resist for long
the temptation to try to exploit signs of foreign weakness. They were
at peace for a generation, but then, during the later decades of the
Northern Sung period, there were repeated clashes with the Western
Hsia. More ominously, some Sung leaders continued to think about the
reconquest of territories controlled by the Liao. Then during the reign of
Hui-tsung, when a new tribal group, the Jurchen, became increasingly
powerful in the region of present-day Manchuria, the Sung leaders were
tempted into forging an alliance aimed at destroying the Liao. 13 With

13 On the Chin state, see Jing-shen Tao, The Jurchen in Twelfth Century China (Seattle:
University of Washington Press, 1976); Jing-shen Tao, "The Influence of Jurchen Rule
41
 Law and order in Sung China

the help of the Sung military, the Jurchen drove the Liao out of north
China and established the Chin dynasty, but the Jurchen armies did not
stop at the old Liao southern border. Emperor Hui-tsung, faced with the
Jurchen invasion, abdicated the throne, but both he and the young man
who had taken his place as emperor were captured by the invading
armies, ending the period known as the Northern Sung. A Sung prince,
fortuitously absent from the capital when it was captured, took the
throne as Emperor Kao-tsung (r. 1127-62), the first ruler of the period
known as the Southern Sung. For some years the fortunes of the Sung
state were precarious. In the late 1120s, Sung armies were driven south
of the Yangtze River. At one point the new emperor barely escaped
capture by fleeing to sea. Only after a series of successful Sung counter-
attacks in the 1130s did the frontier stabilize, along the Huai River (see
Map 2.2). In the treaties that punctuated these years of warfare, the
Sung emperors recognized their inferior position and agreed to pay a
sizable annual tribute to the northern state. 14 For the next two decades
relations between the two states were relatively amicable. Until 1155 the
Sung Court was dominated by the powerful official Ch'in Kuei (1090-
1155), who had built his career on accommodation with the Chin.
Even after his death and subsequent vilification by his long-suppressed
bureaucratic enemies, the peace remained in effect, until shattered
briefly by the ambitions of the Chin leader, Prince Hai-ling (r. 1150-
61), who had seized the Chin throne through intrigue and assassination.
The aging emperor Kao-tsung, weary of the burden of office he had
carried for so long, abdicated at this point in favor of his heir, who
reigned as Emperor Hsiao-tsung (r. 1162-89). Hsiao-tsung was one of
the most effective of the Sung emperors. He carefully kept ultimate
power in his own hands, in part by frequently changing his chief min-
isters. Within a few years of his accession a new peace treaty was drawn
up, inaugurating four decades of peace along the northern borders, a
peace which lasted until the abortive attempt to invade the Chin state
during the first decade of the thirteenth century. The Sung defeat was
followed by two more decades of peace, but a peace made problematic
by the advent in the 1220s of a new foreign tribal power, the Mongols.

on Chinese Political Institutions," Journal of Asian Studies 30.1 (1970): 121-30; and
Jing-shen Tao, "Political Recruitment in the Chin Dynasty," Journal of the American
Oriental Society 94.1 (1974): 24-39.
14 For the early treaties signed, see Herbert Franke, "Treaties between Sung and Chin" in
Etudes Song, 1st series, no. 1 (The Hague: Mouton, 1970), pp. 55-84. For the politics of
this era, see Richard L. Davis, Court and Family in Sung China, g6o-i2yg (Durham: Duke
University Press, 1986), esp. pp. 46 fT.

42
 The historical context

Map 2.2. The Southern Sung state and its neighbors by the mid-twelfth century

The Sung Court found it very difficult to create a workable policy to

respond to the altered power situation created by the rise of this new
tribal group. 15 A belated and probably foredoomed attempt to exploit
Chin weakness by allying with the Mongols only set the stage for
Mongol attacks against the Sung territories. From the time when the
Mongols were able to destroy the Chin state in the late 1220s until the
12 70s, the Mongols fought an intermittent war against the Southern
Sung, which ended only with the final collapse of the Sung dynasty in
1279.
Despite these occasional wars, the Sung period overall was an era of
peace and prosperity. It was a time of great economic growth and
change and of technical advances. This era of growth and development,

15 See Charles Peterson, "First Sung Reactions to the Mongol Invasion of the North,
1211-1217," in John W. Haeger, ed., Crisis and Prosperity in Sung China (Tucson:
University of Arizona Press, 1975), pp. 215-55.

43
 Law and order in Sung China

which had begun in the T'ang and continued through the Northern
Sung, did not begin to slacken in pace until the twelfth century.

Money and credit

During this remarkable period of more than two centuries, from the
late T'ang into the mid Southern Sung, China developed a thoroughly
monetized economy, with the invention and spread of the use of paper
money, the widespread use of a variety of other paper instruments for
commerce, an enormous increase in minted coinage, and the use of other
media of exchange, including precious metals and silk.16 These increases
were truly astonishing. The major type of coinage in imperial China was
a round bronze coin with a square hole in the center. A string holding
iooo such coins was a standard accounting unit. Already by 997 the
Sung government was annually minting 800,000 strings of coins, two and
a half times the largest output of the T'ang. By 1085 this had risen to
over 6 million strings, the greatest annual output ever under the Chinese
empire. It is estimated that over 200 billion coins were cast during
the Northern Sung. In addition, the use of silver increased greatly, as
reflected in increases in government collection of silver in taxes, from
884,000 ounces in 1021 to almost 3 million ounces in 1075, and a
staggering 18 million ounces in 1120.17
The most noteworthy monetary innovation was the creation of a
government system of paper money. The initial invention of paper
money was stimulated by the peculiar coinage situation in Szechuan.18
In that region, iron rather than bronze was used for coinage. Because
per unit of value such coins were much heavier than bronze coins,
merchants were greatly inconvenienced. In the late T'ang, therefore,
merchants in the region began using as money the certificates of deposit
that they received from the deposit shops at which they had left money
or goods. The early Sung authorities awarded a monopoly on the issuing
of these certificates to a small set of shops and then in the 1120s
took over the system and issued the world's first official paper money.
Although there were periods during which the government incautiously
issued money without sufficient backing, leading to serious inflation, by

16 L. S. Yang, A Short History of Money and Credit in China (Cambridge, MA: Harvard
University Press, 1952); Robert Hartwell, "The Evolution of the Early Sung Monetary
System," Journal of the American Oriental Society 87 (1967): 280-89; and Mark Elvin, The
Pattern of the Chinese Past (Stanford, CA: Stanford University Press, 1973).
17 These figures are taken from Peter Golas, "Sung Financial Administration,"
unpublished manuscript, n.d.
18 Cristoph Schifferli, "Le systeme monetaire au Sichuan vers la fin du Xe siecle," T'oung
Pao 72 (1986): 269-90.

44
 The historical context

and large the Sung system was kept under careful control, which helped
enormously in the growth of Sung commerce.
Internally the rise of commerce changed the face of Sung China. The
goods gathered and distributed by itinerant merchants had in the past
been largely limited to luxury goods for the rich. During the Sung period
they came to include daily necessities, which were sold not just to the
rich but also to the majority of the population. Farmers, most of them no
longer self-sufficient subsistence producers, became tied into a wider
market and so were subject to its price fluctuations. As producers they
supplied the urban trade; as consumers they also regularly purchased
necessities.19
Externally, from the beginning of the dynasty, the Sung authorities
enacted policies designed to stimulate overseas commerce. The Sung
founder took care to assure merchants of his protection, to fix tax
assessments, and to formulate commercial regulations. The Sung leaders
soon began establishing officials charged with overseeing and encourag-
ing foreign trade. Court officials were sent on missions to southeast
Asian countries to encourage their traders to come to China. Money
generated by foreign trade thus came to play a significant role in the
government's total income, especially in the Southern Sung. Perhaps
more importantly in the long run, this new orientation toward the sea
contributed to the rise, for the first time in Chinese history, of a powerful
blue-water navy and to a displacement of south and southwest Asian
merchants by Chinese merchants sailing on Chinese ships as the prin-
cipal commercial figures in east and southeast Asian waters. 20

Demographic changes
The rise of a money economy and the commercialization of Chinese life
were accompanied by some remarkable demographic changes. The
whole rural demographic pattern had changed from the late T'ang into
the Sung, with the rise of a vast number of small market towns, which
served as markets for rural products and as suppliers of rural needs.
19 For more on these changes, see especially Yoshinobu Shiba, "Urbanization and the
Development of Markets in the Lower Yangtze Valley," in Haeger, ed., Crisis and
Prosperity, pp. 1 3 - 4 8 ; Yoshinobu Shiba, Commerce and Society in Sung China, trans. Mark
Elvin (Ann Arbor: University of Michigan Center for Chinese Studies, 1970);
Yoshinobu Shiba, "The Commercialization of Farm Products in the Sung Period," Ada
Asiatica, December 1970, pp. 77-96; Lawrence J. C. Ma, Commercial Development and
Urban Change in Sung China (g6o-i2?g), Michigan Geographical Publication no. 6 (Ann
Arbor: University of Michigan Department of Geography, 1971).
20 For a classic description of this process, which still holds good a generation after it was
written, see Jung-pang Lo, "Maritime Commerce and Its Relation to the Sung Navy,"
Journal of the Economic and Social History of the Orient 12 (January 1969): 57—101.

45
 Law and order in Sung China

Many of these towns began as "grass markets," small periodic markets

that were sometimes far from the district or prefectural cities. These
markets attracted teahouses, shops that sold daily necessities, and a
growing population. Eventually they often attracted the attention of the
government, which would establish tax collection offices there. 21
These changes in the distribution of the population also seem to
have been accompanied by a continuous overall growth in population.
Although the available statistics bristle with problems, it still seems
quite possible that the Chinese population doubled between the middle
of the eighth century and n o o , reaching a total of about ioo million in
the latter year. 22

Industrial development
In addition, great regional markets developed and, with them, sig-
nificant industrial enterprises employing large work forces of hired
laborers. Many of these enterprises, in which ownership was separated
from management, manufactured goods for mass distribution. 23 In the
words of Robert Hartwell:
By the time of Wang An-shih (1021-86), alum making, salt processing, quick-
silver and cinnabar production, shipbuilding, papermaking, and printing were
all businesses in which the scale of operation and the absolute level of physical
output were greater than was common in any other national economy before the
last decades of the eighteenth century. But progress in the extraction and refining
of metallic ores was even more astonishing. In 1067, 331,500 tons of iron, copper,
lead, and tin were fabricated into articles to be used in construction, production,
and destruction.24

Urbanization
This industrial transformation was accompanied by an unprecedented
increase in urbanization. The most economically advanced region of

21 Kato Shigeshi, Shina keizaishi kosho (Studies in Chinese economic history) (Tokyo: Toyo
bunko, 1952), vol. 1, pp. 299-421.
22 Hans Bielenstein, Chinese Historical Demography, A.D. 2-1982 (Stockholm: Museum of Far
Eastern Antiquities, 1987), vol. 59; and Ping-ti Ho, "An Estimate of the Total
Population of Sung and Chin China," Etudes Song, 1st series, no. 1 (The Hague:
Mouton, 1970): 33-53.
23 Robert Hartwell, "A Revolution in the Chinese Iron and Coal Industries," Journal of
Asian Studies 21 (1962); and Robert Hartwell, "A Cycle of Economic Change,''' Journal of
the Economic and Social History of the Orient 10 (July 1967): 162-59.
24 Robert Hartwell, "Markets, Technology, and the Structure of Enterprise in the
Development of the Eleventh-Century Chinese Iron and Steel Industry," Journal of
Economic History 26 (March 1966): 32.

46
 The historical context

China, the section of the delta south of the Yangtze River known as
Chiang-nan, was probably the most highly urbanized area in the world
before the rise of industrial Europe in the late eighteenth and early
nineteenth centuries. The few cities of T'ang times had been predomi-
nantly governmental in character, planned in form, and kept under
tight security control. Internally these T'ang cities had been divided by
interior walls. Travel on the open thoroughfares was forbidden during
the curfew hours of the night. The state attempted, with some success, to
exercise control over the market systems that supplied these cities.
Sung cities were radically different. Gone were the interior neighbor-
hood walls with their police posts at the gates. Gone for the most part
were the restrictive zoning rules and the curfews. Wide open, bustling,
expanding outside the old walls and reaching into the surrounding
countryside, Sung cities were a wonder of their times. Their people -
merchants, artisans, entrepreneurs, entertainers, soldiers, foreigners -
jostled one another in the crowded streets, 25 in which a new culture was
being born. Indeed, the traveler Marco Polo, trying to describe the great
city of Hangchou to his fellow Europeans, was continually concerned
that the marvels he described would be discounted as fabulous. 26
This urbanization, reflected in the rise of great cities and the enor-
mous increase in the number of market towns, is perhaps the single most
striking trait of Sung demographic change. Still, in order to understand
the problems of law and order in the Sung, we should keep in mind that
even after this unprecedented urban growth, there were large sections of
Sung China that were still quite wild and sparsely populated country-
side, where tigers haunted the forests and bandits could roam almost at
will.

Sung style of life

Surely it is no accident that in these centuries there were changes in the
ways in which the Chinese lived their lives. Before this period the
Chinese knelt on grass mats, like the Japanese. By the Sung they were
sitting on chairs. This led to other changes. The Chinese began to use
higher tables. They no longer used mats on their floors, which now were
often made of stone or tile. People could wear their shoes inside the
house. Easily doffed slippers gave way to boots. If we look at Sung
paintings, even those from the early years of the dynasty, we can see men

25 Jacques Gernet, Daily Life in China on the Eve of the Mongol Invasions (Stanford, CA:
Stanford University Press, 1970), gives a good picture of these great cities.
26 Marco Polo, The Travels of Marco Polo, trans. Henry Yule (London: Murray, 1921), vol.
2, chaps. 76, 77.

47
 Law and order in Sung China

wearing such boots, sitting on chairs and eating and drinking from high
tables. 27
The austere and somewhat heavy architecture of the earlier periods
becomes lighter and taller. 28 In the Sung the straight roof lines charac-
teristic of earlier buildings, at least in north China, gave way to grace-
fully curved eaves. The emphasis on lightness is also reflected in the
Sung penchant for building soaring pagodas. In addition, the restrained
color scheme of the past began to disappear when the Sung artisans
solved the puzzle of glazing colored roof and floor tiles. Their golden
roofs and green tiles began to pave the way toward the garishly painted
red, green, blue, and gold buildings of later times. Then at some time in
the Five Dynasties the gruesome custom of binding women's feet began.
Girls had bandages tied around their feet, bending their toes over toward
their insteps, so that eventually their arches broke. During the extended
period before their feet became painless, girls suffered acutely. Often
flesh sloughed away from the instep and one or more toes dropped off.
The origin of this custom and the reasons for its widespread acceptance
remain unclear. Still, it does seem to be connected with a greater degree
of subordination of women in the Chinese elite classes. Until the end of
the traditional empire, and even beyond, generations of Chinese mothers
inflicted the burden of bound feet on their daughters. 29

The arts
The Sung dynasty was also a period of changes in the arts. New poetic
forms became the rage. Based on the patterns of popular melodies, these
songs were freer in prosody than the older styles of poetry had been. 30
Sung poets also took the older forms in new directions. Sung poetry
in the old style continues the earlier tradition of expressing certain
emotions but now came to reflect intellect as well. Many poems were
narrative in character, dealt with the minutiae of daily life, or expounded
philosophical issues.31 Among the newer forms, the Sung writers carried
furthest the development of songs. These originally were entertainment

27 For this development, see C. P. Fitzgerald, Barbarian Beds: The Origins of the Chinese Chair
(New York: Barnes, 1965).
28 For Chinese architecture in this era, see Lawrence Sickman and Alexander Soper, The
Art and Architecture of China (Baltimore: Penguin Books, 1956), chap. 35.
29 Howard S. Levy, Footbinding: The History of a Curious Erotic Custom (New York: Rawls,
1966).
30 James Hightower, Topics in Chinese Literature (Cambridge, MA: Harvard University
Press, 1953), esp. chaps. 13, 14.
31 Kojiro Yoshikawa, An Introduction to Sung Poetry (Cambridge, MA: Harvard University
Press, 1967).

48
 The historical context

songs, words set to already existing tunes, and were intended to be sung
by women. Along with a new prosody deriving from its lines of unequal
length, the song form as developed in the Sung introduced the theme of
romantic love into the realm of poetic discourse.
In the world of the visual arts the Sung contribution was distinctive.
Its most renowned figures were literati who turned to calligraphy and
painting. Of course, in times past there had been many men who
dabbled to some degree in the worlds of both the visual and the written
arts, but in the Sung there is for the first time the presentation of such
balanced participation as an ideal. In contrast to earlier times, the ideal
of the Sung man of letters was to be skilled also in the visual arts.
This aesthetic approach to the building of a proper Confucian life is
epitomized by the figure of Su Shih. Poet, calligrapher, and writer of fine
prose, he was also one of the leading statesmen of his age. Indeed, in his
lifetime he was a leading figure in a group of Sung thinkers who stressed
the positive role of aesthetic sensibility in the pursuit of the ideal of a
Confucian gentleman. Only in the following century were the views of
this group decisively defeated by the advocates of the philosophical and
moralistic movement known as Neo-Confucianism.32
Sung pottery, too, reached levels of aesthetic refinement that many feel
were never attained again, despite later improvements in technical skills.
Indeed, a major Japanese intellectual historian has taken Sung poetry,
calligraphy, and painting, and most especially ceramics, as the key
to explaining changes in Sung Confucian thought. The central search
among concerned Sung thinkers was for a synthesis of insight and
expression that could be revealed in the arts as an outward expression of
an internal transformation of the self. The objects of art should be the
coming together of learned forms and inner vision and should stand in
some sense for the self, which was also the product of learning and
vision.33

Theater
An urbanized culture with urbane tastes set the stage for the explosive
growth in the popularity of the drama that marked the succeeding era of
Mongol ascendancy. In his description of life in the Sung capital, Nai
Te-weng wrote about the plays called tsa-chu:

32 On Su Shih, see Peter Bol, "Culture and the Way in Eleventh Century China" (Ph.D.
diss., Princeton University, 1982).
33 Okada Takehiko, So-Min tetsugaku no honshitsu (The essence of Sung and Ming
philosophy) (Tokyo: Mokujisha, 1984).

49
 Law and order in Sung China

In the tsa-chu the actor director is senior, and there are always four or five people
in one performance, first performing a section of the usual, familiar things, called
"charm sections," then performing the "tsa-chu proper," also known as "the two
sections." The actor-director role directs the general settings, the playleader role
gives the specific directions, the clown role does the fooling, the jester role the
quipping, and sometimes another actor is added to play the "mandarin role." . . .
In general their performances are nothing but farcical comedy about matters old
and new.34

The great popularity of such performances in the Sung is shown in the

comment of Meng Yiian-lao describing the life of the Northern Sung
capital, that in three of the city's amusement parks there were some
fifty stages or theaters. The largest was said to hold several thousand
people. 35

Technology and science

The Sung dynasty was also an era of technological and scientific ad-
vance. Extensive use was made of explosives in construction. New tech-
niques were employed for deep mining. Copper became so valuable that
technicians designed systems for its hydrometallurgical extraction from
mine waters. For the first time China became a sea power, with its blue-
water navy becoming especially important after the loss of the north in
the 1120S. An overall admiralty was established in 1132, and the Sung
navy grew to a force of hundreds of ships manned by thousands of men
and armed with trebuchets that could hurl gunpowder missiles. 36
Advances in ship construction, such as watertight bulkheads, and the
use of the magnetic compass in navigation (definitely in use by 1070),
made the Chinese ships of the day doubtless the safest in the world. The
text which mentions the use of the magnetic compass in navigation
before the end of the eleventh century also mentions the use of astro-
nomical navigation, soundings, and studies of samples of the sea bottom
as navigational aids. Also, because of the type of sail that they used,
Chinese ships were particularly effective in taking advantage of winds
coming from almost any quarter. 37

34 Nai Te-weng, Tu-ch'eng chi-sheng, in Meng Yiian-lao, Tung-ching meng-hua lu, wai ssu-
chung; quoted by William Dolby, "Early Chinese Plays and Theatre," in Colin
Mackerras, ed., Chinese Theatre from Its Origins to the Present Day (Honolulu: University of
Hawaii Press, 1983), p. 20.
35 Dolby, "Early Chinese Plays," p. 25.
36 Joseph Needham, Science and Civilization in China (Cambridge: Cambridge University
Press, 1965), vol. 4, no. 3, p. 476. See also Jung-pang Lo, "The Emergence of China as
a Sea Power," Far Eastern Quarterly 14 (1955): 489-503.
37 Needham, Science and Civilization, vol. 4, no. 3, pp. 563, 602.

50
 The historical context

Explosive weapons began to be used in warfare: From at least the

eleventh century, Chinese armies were using flame-throwing weapons.
Soon they developed grenades and bombs and by the thirteenth century
were using firearms and cannon. 38
There also were advances in mathematics. Most notably in the last
half of the thirteenth century, a brilliant group of algebraists worked
independently on such problems as indeterminate analysis, the computa-
tion of complex areas and volumes, alligation problems, and arithmetical
progressions.39 These examples could be almost endlessly multiplied.
The Sung dynasty was one of the great eras of scholarly and scientific
curiosity in Chinese history, its writings containing numerous treatises
on archeology, architecture, and other scholarly and scientific topics.

Printing
One development with great long-term consequences was the rise of
printing. Printing, invented in the T'ang and further developed during
the Five Dynasties, reached its full potential only under the Sung.40 By
making possible the wide dissemination of approved versions of the
Classics, it set the stage for a revival of widespread interest in the
Confucian canon; by lowering the price of books, it laid the foundations
for a great expansion of literacy. The first printing of the Confucian
Classics, undertaken by the official Feng Tao during the Five Dynasties,
was completed only a few years before the founding of the Sung. Early in
the Sung, scholars working at the Directorate of Education printed
the Five Classics with Commentaries (in 988) and the Nine Classics with
Commentaries (in 1001). Printing of Buddhist writings was also supported
by the government. The Sung founder in 971 ordered the printing of the
Buddhist canon, a task not completed until early in the next decade. The
text ran to five thousand volumes.41 These were only the largest projects
in an era that also saw the printing of several histories, dictionaries, and
other such works. Both block printing and movable type were in use. A

38 L. C. Goodrich and Feng Chia-sheng, "The Early Development of Firearms in China,"

his 36 (1946): 114-23; and Wang Ling, "On the Invention and Use of Gunpowder and
Firearms in China," his 37 (1947): 160-78.
39 Needham, Science and Civilization, vol. 3, pp. 38 ft0.
40 For information on printing, see Thomas F. Carter, The Invention of Printing and Its Spread
Westward, revised by L. C. Goodrich (New York: Ronald Press, 1955); Paul Pelliot, Les
debuts de Vimprimerie en Chine (Paris: Imprimerie nationale, 1953); L. C. Goodrich, "The
Development of Printing in China and Its Effects on the Renaissance Under the Sung
Dynasty," Journal of the Hong Kong Branch of the Royal Asiatic Society 3 (1963): 36-43.
41 Kenneth C. S. Ch'en, Buddhism in China: A Historical Survey (Princeton, NJ: Princeton
University Press, 1964), p. 375.

51
 Law and order in Sung China

regular book industry grew up, catering to the vastly increased number
of literate people.

Sung intellectual life

During the Sung dynasty China's intellectual life was transformed. The
stress on the examination system as a route to office under the Sung
prompted an increase in education geared to earning degrees. Because
passing the examinations depended heavily on a knowledge of the
Confucian Classics and commentaries, a growing pool of men devoted
themselves to studying these works. Though it would be a mistake to
underestimate the importance of Buddhism and Taoism to the male
members of the Chinese elite in the Sung, nonetheless it is fair to say
that Confucian values and attitudes became more and more important.
During the reign of the fourth Sung emperor, Jen-tsung, these develop-
ments began to bear fruit. A group of scholar-officials arose who adopted
a different approach to the Classics: They began to emphasize the
importance of directly understanding their basic message, thereby
bypassing the learned commentaries of the Han and T'ang. The insights
gained in this way were to be applied to the reformation of social and
political life. The last four decades of the Northern Sung, when this
movement blossomed, were one of the most creative eras in the history of
Chinese thought. The ideas discussed in these decades were further
refined during the Southern Sung, until they provided the agenda and
corpus that were to dominate the thinking of Chinese leaders into the
seventeenth century.
This new interest in the classical heritage raised with renewed sharp-
ness the continuing tensions in China between duty to self and duty to
society. From its beginnings Confucianism had strongly emphasized the
idea of service to society. Indeed, this is one defining characteristic of the
Confucian persuasion. The ideal is superbly expressed in the epigram of
the statesman Fan Chung-yen, who stated that "a scholar should be the
first to become concerned with the world's problems and the last to
rejoice in its happiness." 42 Such an emphasis on service is characteristic
of most of those involved in the Confucian revival at this time. However,
Confucianism also underscored the need for considering oneself. It
was possible to be too selfless, like the thinker Mo-tzu, of whom the
Confucian Mencius had said that "Mo Tzu advocates love without
discrimination. If by shaving his head and showing his heels he could
42 Translation in James T. C. Liu, "An Early Sung Reformer: Fan Chung-yen," in John
K. Fairbank, ed., Chinese Thought and Institutions (Chicago: University of Chicago Press,
1957), P- " I -
52
 The historical context

benefit the Empire, he would do it." 43 Certainly, individuals have social

responsibilities, but these should not lead them to abuse their bodies,
which are a sacred gift from the ancestors. To disfigure oneself so that
one looks like a convict was beyond the reasonable.
A balance was needed between service to self and service to society,
but the very idea of self was ambiguous. In one sense, self in the Chinese
context was not the isolated individual but, rather, the individual as a
creature defined by interrelationships with others. Such relationships
could be based on friendship, shared work or residence, or, most import-
antly, ties of kinship through blood or marriage. The closeness of these
kinship ties were precisely defined in terms of "mourning degrees,"
which were reflected in the various mourning rituals owed to different
relatives. The closer the relative was, the longer the mourning period
and the more elaborate the rituals should be (see Figure 2.1). The "self"
was thus seen in terms of others. To preserve and promote this self was
not just to preserve and promote ego, but also to bring honor to one's
ancestors and to protect one's descendants.
At the minimum, such a self encompassed duties and ties to progen-
itors and descendants. The corporate character of the core kinship group
in traditional times could be seen in customs, mores, and laws. Before
the establishment of the empire in the third century B.C., and to a lesser
degree thereafter, this corporate character meant a duty to avenge
wronged relatives. This duty is closely bound up with the practice of
joint adjudication and punishment. In preimperial times the prudent
victor in one of the innumerable conflicts of that age often exterminated
all the relatives of the loser that his power could reach. Even after the
beginning of the empire, such joint adjudication and punishment were
retained for certain classes of crimes. Those involved through association
might have their patrimonies confiscated or be severely punished, even
enslaved by the state or executed. Such joint adjudication affected
certain related groups, most commonly people living in the same house-
hold and close relatives. Under some circumstances it might be extended
to other members of certain types of organizations such as state-mandated
mutual security and surveillance groups, a clear indication of the felt
importance of corporate identities. Given the existence of such corporate
groupings, and the identification of the self in terms of ties to other
members of the group, the possibilities for conflict between these cor-
porate, especially family interests and the public interest, are obvious.
The tensions between Confucian social commitment and concern for
the self was complicated by a new emphasis on the importance of self-

43 Lau, trans., Mencius, p. 188.

53
 b brother fffm ffff For female relatives the variations in mourning depend on whether
d daughter 2d they are still living at home or have married out.
f father
m mother
For sons, grandsons, and their wives the variations
s sister fffs ffm fff fffbw fffb
depend on whether the son or grandson involved
S son 5 2c 5
is eldest or not.
w wife
Sw son's wife fffbd ffs fm ff ffbw ffb fffbSw fffbS
ff father's father 5 or none 4 or 5 2b 4 5
and so on

fffbSd ffbd fs m f fbw fb ffbSw ffbS fffbSSw fffbbS

5 or none 4 oi 5 2b or 3 2a 1 2b 4 5

fffbSSd ffbSd fbd s bw b fbSw fbS ffbSS fffbSSS

EGO
5 or none 4 or 5 3 or 4 2b or 3 4 2b 3 4 5

ffbSSd fbSd dbd Sw S bSw bS fbSSw fbSS ffbSSS

5 or none 4 or 5 2b or 3 2b or 3 1 or 2b 4 2b 5 4 5

fbSSd bSd SSw SS bSSw bSS fbSS

5 or none 4 or 5 3 or 5 2b or 3 5 4 5

fbSSSd SSS bSSS

5 or none 5 5

This chart is based on the assumption that ego is

male. There would be some relatively minor vari- SSSS
ations if ego were female. 5

Figure 2.1. Mourning chart. From Yuan tien-chang. Reproduced by permission from Brian E. McKnight, "Song Legal Privileges,"
Journal of the American Oriental Society 105 (1985): 105. The numbers 1-5 indicate mourning degrees. Numbers followed by letters
indicate subdivisions within mourning degrees.
 The historical context
discovery and the self-improvement of the individual. Many men who
were committed to the Confucian vision wanted to make themselves
exemplary individuals. Certainly there could be strong commitments to
promoting the welfare of relatives and the society at large; to this was
added a desire to mold the individual personality like a work of art, so
that one might become an exemplary gentleman (chun-tzu) or even a
sage.
Until the middle of the eleventh century this movement was present
only in embryo. A century and a half of avid pursuit of and debate about
Confucian goals and the ways to reach them began in the 1040s. During
this period of 150 years, a variety of answers were proposed. One group,
especially prominent during the Northern Sung, saw artistic creation as
the proper avenue to the Confucian realization of the Way. Its leader,
Su Shih, "made himself important because he concerned himself with
standards for culture and cultural activities at a time when literati
believed that culture was central to the organization of national life
and the foundation upon which literati maintained themselves as the
political, social and intellectual elite of China." 44
This artistic orientation was soon challenged by men who emphasized
the primacy of moral and ethical self-cultivation. Despite differences
among some of the Neo-Confucian thinkers of the late eleventh and
twelfth centuries, they did share a sense of the central importance of the
moral cultivation of the self. Chu Hsi (1130-1200), who synthesized the
ideas of many of his predecessors into a philosophy that was to become
and remain orthodox in much of East Asia until the twentieth century,
wrote:
Confucius spoke of "the conquest of self and return to propriety." The Doctrine of
the Mean says, "Advance toward equilibrium and harmony," or again "Prize the
virtuous nature and pursue the path of inquiry and study." . . . Man's nature is
originally clear, but it is like a pearl immersed in impure water, where its luster
cannot be seen. Being removed from the dirty water, however, it becomes
lustrous of itself as before. If each person could himself realize that it is human
desire that causes this obscuring, this would bring enlightenment. It is on this
point alone that all one's efforts must be concentrated.45
This group of thinkers redirected the focus of the interests of the Chinese
elite.

44 Bol, "Culture and the Way," conclusion, p. 1.

45 Chu Hsi, Chu-tzu yu-lei, 12.8. Translation from Fung Yu-lan, A History of Chinese
Philosophy (Princeton, NJ: Princeton University Press, 1952), vol. 2, pp. 559-60.

55
 Law and order in Sung China

Political and social changes

Such intellectual changes were interwoven with a momentous trans-
formation in the character of the Chinese political and social elite.
Although there are disagreements about the details of this transforma-
tion, it is widely recognized that between the T'ang and the Sung the
nature of the ruling class changed.46 Until the late T'ang, political power
was predominantly in the hands of a relatively small group of families,
which intermarried with one another and dominated life at the Court. In
the provinces a larger group of regional elite families reproduced this
pattern on a smaller scale.47
This situation began to change in the late T'ang, especially after the
rebellion of Huang Ch'ao in the 870s. New powerful regional military
lords arose, who, in the ensuing Five Dynasties period (907—60), domi-
nated politics.48 It was in this Five Dynasties era that the older aristo-
cratic families finally lost their position of power. Military men, often
from humble backgrounds, used other military men as subordinates in
key posts, and even the civilian advisers that they employed came less
and less frequently from the old families. When the Sung founder and his
successor succeeded in reducing the influence of the military on politics,
the civilian cadre of officials was in a position to emerge as the decisive
group in the political equation. 49
Although the details of the changes are not wholly clear, it does seem
that between the T'ang and the Sung periods there was a significant
transformation of the family system. With the decline of the older aristo-
cracy, the older practices for ensuring the continuity of the elite families'
fortunes across the generations changed in their emphasis and in their
particulars, if not in their general character. In traditional Chinese
families the patrimony was divided equally among sons after the death of
the father, with unmarried daughters receiving smaller amounts (usually
in the form of dower portions). Because of this, it was necessary to
work continually if the family's economic position was to be preserved.
Furthermore, as the group of relatives grew larger over the generations,

46 David Johnson, "The Last Years of a Great Clan: The Li Family of Chao Chun in Late
T'ang and Early Sung," Harvard Journal of Asiatic Studies 37 (1977): 5-102; Robert
Hartwell, "Demographic, Political, and Social Transformation of China, 750-1550,"
Harvard Journal of Asiatic Studies 42 (1982): 365-442; and Robert Hymes, Statesman and
Gentleman (Cambridge: Cambridge University Press, 1986).
47 Denis Twitchett, "The Composition of the T'ang Ruling Class: New Evidence from
Tunhuang," in Arthur Wright and Denis Twitchett, eds., Perspectives on the T'ang (New
Haven, CT: Yale University Press, 1973), pp. 47-87.
48 Wang Gung-wu, The Structure of Power in North China During the Five Dynasties (Stanford,
CA: Stanford University Press, 1963).
49 Worthy, "The Founding of Sung China."

56
 The historical context

affective ties declined in strength. By the Sung period, elite families no

longer had the overweening sense of status pride that, despite these
centrifugal tendencies, had tied together members of the great T'ang
choronyms (groups of relatives tied together by both surname and
identification with the family ancestral home, e.g. the Kennedys of
Hyannisport). How were kin beyond those most closely related to be
held together?
Institutionally, the most revealing example of a device for encouraging
kin solidarity was the endowed clan estate. Such estates, administered by
managers, produced revenues that were supposed to be used for projects
and activities that would promote the kin group, such as shared ritual
feasts, schools for the children, and relief for poorer members. The
prototype of these estates was that founded by the great eleventh-century
statesman Fan Chung-yen (989-1052). Himself an orphan in childhood,
he eventually sought out his natal relatives and endowed an estate as a
means of encouraging their solidarity. That these estates frequently did
not work as intended by their founders — that the rich and powerful
among the kin tended to usurp control - points up the very problem of
lack of kin solidarity which the estates were established to remedy.
Nonetheless, they stand as symbols of a renewed emphasis on the long-
term preservation of kin power and solidarity. 50
The families that tended over time to dominate such enterprises were
the same group of relatively well-to-do landowning families who domi-
nated local life.51 The basis of the influence of these families was usually
landownership, the income from which they used to reinforce their role
in local life by taking leading roles in local charitable activities, organiza-
tion of local self-defense forces, construction projects, and other local
activities. Ideally, a family in this group wanted to pass on its patrimony
and occasionally to place a member in the civil service elite through the
competitive civil service examinations. Although graduates of these
examinations were never numerically dominant in the Sung civil service,
they were overwhelmingly important on the higher policymaking levels
of government. It has been suggested that during the Northern Sung,
this official elite was dominated by a group of families who had close
connections by marriage. These families sought to preserve their family
fortunes by obtaining national offices for their members. 52 In the late
Northern Sung and the Southern Sung, when gaining national office

50 On these estates, see Denis Twitchett, "The Fan Clan's Charitable Estate, 1050-
1760," in David S. Nivison and Arthur F. Wright, eds., Confucianism in Action (Stanford,
CA: Stanford University Press, 1959), pp. 97-133.
51 On the local elites, see Hymes, Statesman and Gentleman.
52 Hartwell, "Demographic, Political, and Social Transformations of China."

57
 Law and order in Sung China

through the examination system became more difficult, and the offices so
gained less secure, Sung elite families began to focus their concerns more
on their home areas, intermarrying with their neighbors and concentrat-
ing to a greater degree on local concerns.
The increased competition in the examination halls, which made
ordinary examination degrees harder to earn, coincided with an increase
in the use of less regular channels of recruitment and more importantly
with an increase in the importance of irregularly selected officials. It also
coincided with the spread of a new interpretation of the Confucian
message, which emphasized the importance of education as a path to
personal salvation through the cultivation of the self, rather than as a
route to success in the examinations. This interpretation was only one
among a number of new interpretations of the Confucian message that
competed for the allegiance of educated men in the Sung. However,
under the social and political conditions of the Southern Sung, in its
formulation by the philosopher Chu Hsi (i 130-1200), this interpretation
became more and more influential. In the decades after Chu Hsi's death,
it became the dominant strain of thought among the educated. It
appealed to some people because it validated education for personal
development and a withdrawal from political involvement; it appealed to
the government because of its stress on obedience and loyalty.

Government administration
The Sung regime also reconfirmed the traditional Chinese imperial
attitude that governing was a matter for officials. People who were not
members of the government were not supposed to concern themselves
about political questions. Furthermore, in the Sung period the reproduc-
tion, distribution, or possession of laws or edicts by nonofficials was
illegal. Sung commoners were not citizens in the Western sense. They
were subjects. But although the basic relationship between rulers and
ruled did not change, the governmental institutions through which
power was exercised continued to evolve.
The change most relevant to the processes of the law was the trend
toward the centralization of power, combined with a continuing imperial
respect for bureaucratic opinion. Centralization of power was reflected
in the organization of the central offices, the relation of the center to
local areas, and the role of the ruler. The main change in the way that
the central government was organized was the gradual increase in the
powers of its chief ministers. During the first century of Sung rule
the most important central government organs were the Secretariat-
Chancellery, which controlled government administration (except for
58
 The historical context

military and fiscal administration), the Finance Commission, which con-

trolled fiscal and economic affairs, and the Bureau of Military Affairs,
which planned and directed the national defense. T h e principal policy
g r o u p under this arrangement was the Council of State, a body com-
posed of a varying n u m b e r of councillors. It was headed by chief coun-
cillors, who most often n u m b e r e d two, though sometimes there was only
one a n d occasionally as m a n y as three. These chief councillors were the
heads of the Secretariat-Chancellery. T h e y were assisted by men who
were the heads of the Bureau of Military Affairs and by their own
assistants from within the Secretariat-Chancellery. This group met
regularly to discuss affairs that had only their importance in common. 5 3
T h e n , during the reform era under Shen-tsung, the powers of the chief
ministers were expanded. Control over financial policy and then over
personnel policy became more centralized. In the words of J a m e s Liu,

Another aspect of further centralization was the concentration of power in the

executive branch at the very top of the government organization . . . the executive
department no longer observed the former division of jurisdiction; it now had
direct control over the Finance Commission, and without its approval no regula-
tions pertaining to state finance could be put into effect.54

T h i s increase in the powers of key ministers was carried further during

the Southern Sung, when during military emergencies, the chief minister
came to hold concurrently the position of head of the Bureau of Military
Affairs or its equivalent, thereby giving him control over all vital areas.
T h e three top organs at court, the Imperial Secretariat, the Imperial
Chancellery, and the D e p a r t m e n t of State Affairs - which in the Northern
Sung had been kept separate as a part of a deliberate balancing of
powers - in the Southern Sung were consolidated under the control of
the chief minister. Furthermore, he came to have control over m a n y
cases of bureaucratic promotion and demotion, giving him a powerful
device for controlling his subordinates.
At the same time it was during the Sung that the trend toward an
increasingly autocratic emperor emerged. This change is usually traced
to the change in the social origins of the officials. During earlier times
the officials were themselves usually members of politically powerful
families, and the emperor was often only the first a m o n g equals. D u r i n g
the Sung, when the officials came to be d r a w n from a much larger
g r o u p of less politically powerful families, the emperors had relatively

53 For a good brief description of this system, see E. A. Kracke, Jr., Civil Service in Early
Sung China (Cambridge, MA: Harvard University Press, 1953), chap. 3.
54 James T. C. Liu, Reform in Sung China (Cambridge, MA: Harvard University Press,
1968), pp. 86-87.

59
 Law and order in Sung China

greater power. Although during the Sung itself the emperors treated
their officials with respect, the growing gulf between emperor and
bureaucrat was already being demonstrated in Court ritual. Whereas the
T'ang grand councillors had sat with their sovereign to discuss policy,
Sung Court officials stood at attention. The emperor's potential power
was also less hindered by the power of his relatives. Imperial relatives,
with the exception of some empresses-dowager, were generally less
influential than they had been in earlier times. Until quite late in the
dynasty, members of the imperial Chao clan were kept from participat-
ing in the powerful offices of empire, and though by the time the empire
fell to the Mongols there were many Chao family members in office, and
a few had risen to high office, they in no way exercised an amount of
power analogous to that of the imperial relatives in such dynasties as the
Han and T'ang.
Despite the planting during the Sung of the seeds of later autocratic
rule, this dynasty is known for the degree of respect paid by the em-
perors to their ministers. Perhaps this stems from the high quality of
many leading officials; no doubt it also derives from the example of
respect set by the founding emperors and from a sensitivity to the
importance of imperial cooperation with the bureaucracy. This is not to
say that emperors were not sometimes at odds with the general opinion
of their officials. Emperors might resist the climate of opinion for con-
siderable periods of time, but in the end they usually surrendered. 55 On
several occasions they also backed down when officials remonstrated
with them for making exceptions to the laws and precedents. 56 And
periods like the reign of Hui-tsung, when the emperor often intervened to
decide legal cases by decree, were looked on as unfortunate aberrations
from generally accepted imperial practice.
During the Southern Sung this desire for a cooperative regime was
carried even further. James Liu characterized the political style of this
era as one of accommodative politics, in which the emperors preferred to
gain their ends by indirection, rarely allowing confrontational politics.
Instead, they pigeonholed unwelcome recommendations, asked those
involved to reconsider in search of some grounds for agreement, or

55 For example, Emperor Jen-tsung's retention of his chief minister Ch'en Chih-chung
despite a continuing attack by censors and others, for almost six months. But in the end
Jen-tsung gave in to the weight of opinion. See Brian E. McKnight, "Crime in High
Places: The Case of Ch'en Chih-chung," in Kinugawa Tsuyoshi (ed.), Liu Tzu-chien
hakase soshi kenkyu (Studies in honor of Dr. Liu Tzu-chien) (Kyoto: Dohosha, 1990), pp.
499- 515-
56 This happened repeatedly when emperors wanted to exclude specific individuals from
the benefits of amnesties. Officials argued successfully that this would undermine public
confidence in the sovereign's word.
60
 The historical context

openly agreed with suggestions but did nothing about them. Only as a
last resort would the emperors move strongly against unwelcome critics.
Even then they sometimes did so by subtly soliciting criticism of the
critics, feigning reluctance as they supported such reports. 57

Role of law
O n e characteristic of this bureaucratic system was an elaborately devel-
oped system of laws, both proscriptive and prescriptive. 5 8 T h e Sung
administration had inherited its body of proscriptive laws from the
T ' a n g , which in turn had built on the developments of preceding dyn-
asties. T h e key rules were enacted as lu, or statutes, and were included in
a published code. T h e earliest Sung Code, issued in the early 960s, was
largely a copy of the T ' a n g Code of 737, with the addition of some laws
created after 737. In its thirty chapters it set down the principal crimes
and specified the penalties that were to be attached to them.
More important for controlling the functionaries, including the
legal functionaries, was the enormous elaboration of prescriptive rules,
specifying who was to do what, how, when, and where (see Table 2.1).
During the Sung these prescriptive rules were divided into three types.
T h e broadest were the ling (ordinances), which were general procedural
rules for the conduct of government business. T h e Sung Code compiled
in the early thirteenth century (which preserves rules from much earlier)
contains many examples of ordinances. An example pertaining to the
penal system offers some idea of the character of such rules:

All those who are to be convoyed so that they can be penally registered [for
labor] in the army, upon the arrival of the date that marks the start of winter,
may be temporarily detained for labor at their place of judgment or in the
prefecture at which they have arrived. They should be given stipends. When the
second month arrives, send them on. When the circumstances of their crimes are
serious, or they are to be registered in Kuang-nan and they have already entered
the circuit where they are to be registered, do not apply this ordinance. 59

T h e second type of Sung prescriptive rules were called ko (regula-

tions). These rules were concerned (during the Sung) with determining

57 See James T. C. Liu, China Turning Inward: Intellectual and Political Changes in the Early
Twelfth Century (Cambridge, MA: Harvard University Press, 1987).
58 For a more detailed description of the pattern of Sung laws, see Brian E. McKnight,
"From Statute to Precedent: An Introduction to Sung Law and Its Transformation," in
Brian E. McKnight, ed., Law and the State in Traditional East Asia: Six Studies on the Sources
of East Asian Law (Honolulu: University of Hawaii Press, 1987).
59 Hsieh Shen-fu, Ch'ing-yuan Viao-fa shih-lei (hereafter cited as TFSL) (Tokyo: Koten
kenkyu kai, 1968), p. 523.

61
 Table 2.1. Types of Sung laws

Category Character

Lu (statutes) Fundamental rules governing some aspects of

governmental organization, legal procedures, and what
we would consider crimes. All "criminal" in the sense
that the rules take the form, "If you do (or fail to do)
act A, you will be liable for punishment B."
Ling (ordinances) Basic procedural rules. These rules governed many
aspects of governmental activities. They were
prescriptive rather than proscriptive. Ordinarily the
texts of such rules did not specify the penalty that
would follow infraction. (These penalties were set by a
separate rule, and were very light.)
Ko (regulations) In the Sung these rules were concerned with questions of
quantity or amount. They determined such things as
the complement of clerks in an office, the amounts of
rewards, the lengths of mourning leaves that officials
should observe for categories of relatives, and so on.
Shift* (specifications) Excluding the handful of rules in this category which
the Sung administration had taken over from earlier
dynasties, most specifications determined the
specific characteristics of certain materials used in
communication. Thus they determined the formats of
official documents, but they also determined the
specific characteristics of objects involved in ritual
communication. In Chinese thinking this meant that
specifications fixed the nature and dimensions of
imperial buildings, since such buildings were ritually
significant.
Ch'ih (edicts) During the Sung the use of this term is sometimes
ambiguous. It is used simply to mean an edict, any
edict, and thus in terms of content might describe
rulings which would fall in any one of the categories
listed above. However, during the eleventh century the
term also came to be applied specifically to rulings
which had the character of statutes, that is to say
rulings which specified a particular punishment for
doing or failing to do a particular action.
Lf (precedents) Precedents became increasingly important in this era,
These precedents were the responses to specific cases
or issues which were promulgated by the central
administration and which nominally indicated imperial
decisions. Such rulings were more and more often cited
by Sung officials in determining cases, and eventually
were compiled into regular sets of laws. Because the
word indicates their origin rather than their content,
precedents could resemble in content any of the above
sorts of substantive categories of rules.

62
 The historical context

levels or quantity. T h a t is, regulations determined such questions as the

personnel quotas of government units, pay scales, the amounts of money
or goods to be used in rewarding the meritorious, and the lengths of
mourning leaves for officials.
T h e third type of prescriptive rule, a type peculiar to the Sung as a
formal category of law, was called shih* (specifications). These rules
determined the specifications of aspects of systems of communications.
They thus encompassed a wide spectrum of activities which the Sung
Chinese conceptualized as communications patterns, including all
aspects of ritual behavior as well as more m u n d a n e matters like docu-
mentary formats. 6 0
Some idea of the vast size of these bodies of rules can be conveyed by
noting that a collection of the rules concerning the imperial Hall of
Light, a ritual building used by the emperor, totaled twelve hundred
volumes. Another compilation, of rules bearing on the reception of
tributary envoys from Korea, ran to almost fifteen hundred volumes. 6 1
These various categories of laws were administered by a hierarchy of
administrative and judicial officials who staffed the central government
organs or were posted to the three principal levels of local administra-
tion. Sung local administration was heir to a long tradition. From late in
the Chou dynasty (1027-256) some kingdoms governed local areas
through a nested hierarchy of territorially defined units staffed by cen-
trally appointed officials. T h e officials in charge of the lowest levels
of these hierarchies generally had catholic responsibilities, including
general administration, judicial administration, and revenue collec-
tion. Although there were some changes over time in terminology and
character, generally the system used in imperial China was based on
local units called districts (hsien). Although in some eras (including the
Sung) the districts did have some subunits staffed by formal government
agents, the hsien remained the fundamental units of local control (see
Figure 2.2).
During the Sung a district was headed by a district magistrate who
had general responsibilities for district affairs. He was (usually)
assisted by a recorder concerned primarily with taxation and other
documentary matters, a sheriff concerned with law and order, and at
times an assistant magistrate. This rather simple distribution of offices
and duties was in practice made more complicated by a pattern of
sometimes-overlapping functions and, in the smaller districts, by a policy
of having a single official hold several of these offices concurrently.

60 Brian E. McKnight, "Patterns of Law and Patterns o£ Thought," Journal of the American
Oriental Society 106 (1982): 323-31.
61 McKnight, "From Statute to Precedent," p. 122.

63
 Law and order in Sung China

Figure 2.2. Sung local administrative structure: circuits (lu or tao) (between
fifteen in the late tenth century and twenty-three in the reign of Shen-stung,
1165-1085); prefectures (approximately ten to twenty per circuit); districts
(hsien) (from two to twenty per prefecture)

In trying to visualize the functioning of these offices one should bear in

mind that many local functions were actually carried out by unpaid local
agents working below the district level. The population of China had
long been divided into small subunits for purposes of control. The
administration therefore called on some members of these local popula-
tion groups to perform needed functions.
A number of such hsien would be grouped together to form the next
higher-level administrative unit, the prefecture. This middle, prefectural,
level of government had been a key level of administration in pre-Sung
times. During the Sung a pattern developed in which greater emphasis
was placed on the districts and on the larger units called circuits, with
the prefectures playing a somewhat smaller role.
During most dynasties the prefectures in turn were grouped together
to form the largest local units, called circuits in the Sung (when they
were only partly administrative in character) and provinces under the
later dynasties (when they were basically administrative).
This largest unit did not exist as an administrative entity under the
early dynasties. 62 Instead, in the Han period the prefectures were
grouped together as units for purposes of inspection and surveillance.
The officials charged with these inspection duties were in no sense
the line superiors of the prefects. Indeed, they were of much lower
bureaucratic rank, and did not have administrative (as opposed to
surveillance) duties. Such grouping of prefectures for purposes of over-
sight continued in use sporadically under ensuing dynasties, but only in

62 On the circuit system and its antecedents, see the excellent description in Winston Lo,
"Circuits and Circuit Intendants in the Territorial Administration of Sung China,"
Monumenta Serica 3 (1974-75)- 39—107.

64
 The historical context

the Sung did the officers in charge of such units begin to have some
administrative responsibilities.
The Sung units, headed by officials called circuit intendants, were
usually termed circuits (lu). In many cases the boundaries of these
Sung circuits correspond roughly to the boundaries of the provinces of
the later dynasties, and indeed, Sung circuits should be seen as the
progenitors of the provinces. However, although ancestral to the later
provinces, Sung circuits were not provinces. First, the duties of the
circuit intendants remained basically inspectorial, despite the accre-
tion of some administrative responsibilities. Second, despite the trend
in practice toward multifunctional circuit intendants, the Sung cir-
cuit intendants were in name, and largely in practice, functionally
specialized. Thus a given circuit might have a fiscal intendant respon-
sible for the efficient mobilizing of resources, a judicial intendant
charged with seeing that justice was fairly administered, a military
intendant overseeing certain military matters, and perhaps other sorts of
intendants with special duties. No intendant had line superiority over
his colleagues, and so no intendant corresponded exactly to the later
governors of the provinces. In general the number of circuits into which
the empire was divided grew over time. Finally, the circuit boundaries of
the intendants also did not necessarily coincide, so that the area overseen
by a military intendant, for example, might be different from the area
overseen by his colleague the judicial intendant.
The judicial hierarchy in local areas corresponded to this administra-
tive hierarchy. Although the heads of subunits of the district occasionally
might have some minor judicial responsibilities, the basic level of judicial
administration was the district. The staffs of these districts varied some-
what depending on size. As described, in the most populous Sung dis-
tricts there might be several civil service officials but in less-populated
places some of these posts would not be established, and single individ-
uals might serve concurrently in two or more of the posts that were
mandated.
Most legal cases probably entered the district court as the court of first
instance, although it was possible for cases to begin at the prefectural
level. The magistrate had the power to rule on and carry out penalties in
all cases in which the punishments did not exceed one hundred blows of
the heavy rod. This would cover many minor crimes and most of the
cases that we would classify as civil. Cases involving crimes calling for
penalties heavier than beating with the heavy rod had to be sent up to
the prefecture or higher levels for review, although the magistrate issued
a preliminary ruling and a recommendation for sentencing.
Sung prefectures had control over a number of districts. The least
65
 Law and order in Sung China

important prefectures might control only one or two districts, but the
average was closer to ten. Prefectural administration was headed by a
prefect, assisted by one or two controllers general (t'ung-p'an) who dealt
with general administrative matters and headed the judicial administra-
tion. Under these prefectural heads there were several staff organs.
These included organs primarily concerned with judicial administration,
which sometimes served as courts of first instance but did not specifically
deal with law enforcement. 63
The Sung state thus had a relatively complex structured judicial
system and a law-enforcement system that partly paralleled and partly
diverged from the pattern of this judicial system. Both rested on informal
and semiformal mechanisms among the people that settled disputes and
captured and sanctioned wrongdoers, and both were the product of
current problems and inherited practices.
The structure of law-enforcement agencies that evolved during the
Sung, with its mixture of civil and military agencies, can be understood
only in light of its historical background. The problem was set by the
complex, vital, creative, and rapidly developing society of Sung times.
The response of those in the government who established the actual
policies regarding law enforcement was conditioned by their views on
the nature of deviance, known technological and fiscal limitations,
and a shared historical understanding that suggested certain possible
responses. These, combined with an analysis of the historical mistakes of
the Sung predecessors, led to the creation of a complex interlocking set of
law-enforcement institutions.

63 A good brief description can be found in Kracke, Civil Service, pp. 45-53.

66
 3
Crimes and criminals

Crime as a function of law

During the early 1960s the conventional wisdom of criminology - the
positivist school which adopted the view that crimes were the objective
properties of certain types of acts - was challenged by a new school
of thought. This new approach, labeling theory, took the position that
"crimes" were created when certain individuals defined certain examples
of behavior as falling into categories of forbidden and punishable actions. 1
In one rather simple sense their approach was rather like that of Thomas
Szasz, who in his controversial book The Myth of Mental Illness main-
tained that the mentally ill are created by definition. 2 The labeling
theorists focused on how categories ("criminals") are socially con-
structed from the outside. People are ill or are criminals because their
society has defined certain behaviors as evidence of illness or criminality.
These misbehaviors are legislated into existence. These recent theorists
are unconsciously echoing here a theme reminiscent of the classical
Taoists. Men create deviance and crimes by defining propriety and
legality. In simple terms a crime is a violation of state rules that makes
the perpetrator liable for sanctions. "A crime is any act committed in
violation of a law that prohibits it and authorizes punishment for its
commission." 3 The more laws there are, the more crimes there will be,
and the fewer laws there are, the fewer crimes there will be. In Western
jurisprudence this attitude is expressed by the legal principle of nullum
crimen sine lege, that without a law against it no act can be considered
a crime. The Sung period itself provides a classic illustration of this
process of creating new crimes through legislation. The Sung History

1 For a good brief discussion of labeling theory, see Gary D. LaFree, Rape and Criminal
Justice: The Social Construction of Sexual Assault (Belmont, CA: Wadsworth, 1989), pp. i8ff.
2 Thomas Szasz, The Myth of Mental Illness (New York: Harper & Row, 1974).
3 Wilson and Herrnstein, Crime and Human Nature, p. 22.

67
 Law and order in Sung China

Table 3.1. Numbers of penal rules

Penalty

Penal Heavy Light Penal Death

Date Death Exile servitude rod rod registration with appeal

1029 17 34 106 58 76 63
1041- 48 55 211 326 88 144 53
1056- 108 105 272 399 126 189 91

remarks that in the period 1008 to 1016 there were 40 items of Sung law
that called for the punishment of penal registration (registration in
certain labor battalions of the army); by 1041 to 1047 t n i s had increased
to 70 items; and by 1174 to 1189 it had reached 570 items. Therefore,
"because the laws on penal registration were numerous, violators in-
creased day by day. Men being punished by penal registration were
everywhere" 4 (see Table 3.1).
English law in the eighteenth century showed a similar pattern of
development, with an enormous increase in the number of laws. 5 In the
English case we know that these new laws were laws to protect private
property and reflected the rising power of the capitalist class. Given the
rapid economic growth in China during the Sung period, the influence of
those seeking to defend property rights might have been a factor in the
increased number of laws, but since the laws themselves are no longer
extant, we cannot know for certain.
The labeling theorists were also saying something rather more subtle
than this question of legislating crimes. They were suggesting that
crimes are created by adjudication, when some person or persons with
the power to make a ruling declares that a specific instance of behavior
falls within the scope of previously enacted legislation. These advocates
of the labeling theory appeared in force in the 1960s. In the following
decade the problem of defining crime was further complicated by the
appearance of conflict theorists, who went beyond the position of the
labelers to point out that those doing the defining were the ruling elites
and that those with power made the rules and set down the definitions
as devices designed to defend their privileges. States create crimes by
enacting laws, laws that serve to protect the power of the ruling group. 6

4 T'o T'o, Sung shih (hereafter cited as SS) (Beijing: Chung-hua shu-chu, 1977), 201.5020.
5 E. P. Thompson, Douglas Hay, Peter Linebaugh, John G. Rule, and Gal Winslow,
Albion's Fatal Tree: Crime and Society in Eighteenth Century England (New York: Pantheon,
I976)-
6 LaFree, Rape and Criminal Justice, chap. 3.
68
 Crimes and criminals

To a historian the dispute among the positivists, labelers, and conflict

theorists seems to result primarily from their particular perspectives on
the problems of deviance. Certainly, some forms of "deviance" are
created by labeling, whether legislative or adjudicative, if we use that
term broadly to include all the rules that can lead to punishment by the
state. Driving fifty miles an hour on some American highways is not
punishable. Driving sixty miles an hour on those same roads is, because
of a defined speed limit. When deciding that A committed certain acts
against B and that those acts fall within the scope of a statute, judges or
juries are defining a particular act as being a crime. But the conflict
theorists are also right. It can be argued that almost all laws, to one
degree or another and in one way or another, serve to buttress the social
system dominated by the current ruling elite. And yet a great many laws
punish behaviors disapproved by large segments of society. Certain
objective actions - the deliberate killing of one person by another, the
forcible seizure of property viewed by the community as belonging to
another person, and violent sexual assault - seem to be behaviors ab-
horred by all known societies. The fact that in a specific instance we may
have difficulty deciding whether the act falls within our definition of rape
or robbery or murder is a commentary on the complexity of making
judgments in the real world, not an argument against the general agree-
ment that such acts are seen as crimes. In this book we shall focus
largely on these categories of universally abhorred crimes, because they
dominate our sources and also because such a focus enables us to make
comparisons with other cultures.

Who were the criminals?

Even with this narrower focus on major, and usually violent, crimes, we
would like to know what sorts of crimes were committed, how often, and
by whom. It is a truism of contemporary criminology that the typical
criminal is a young, unattached, adult male. 7 The same pattern seems to
have held true in Sung China. Although the sex and age distribution in
individual and idiosyncratic crimes - most homicides, for example -
appears to be relatively wide, when we look at comments on banditry -
the crime of most concern in Sung China - it is clear that the authorities
envisioned the criminals as relatively young adult males, usually un-
attached, and frequently deserters from the armed forces.8 Such criminals
7 Wilson and Herrnstein, Crime and Human Nature, p. 26.
8 See, for example, WHTK 168.1460, in which a twelfth-century official is quoted as
saying that in recent years a great many people had been sentenced to penal registration,
largely for having committed robbery, and that "in each case those involved are
deserters." See also SHY, hsing-fa 4.63a.

69
 Law and order in Sung China

ranged from those who probably were in their teens - "evil youths" in
the Chinese phrase - to men presumably in their twenties and thirties
who had escaped from the armies. 9
Soldiers are often accused of crimes, ranging from killing commoners
and claiming rewards on the grounds that the dead were bandits, to
fraternizing and cooperating with bandit gangs, to illegally taking public
property, to being themselves robbers. 10 Ts'ui Tun-li (doctorate of letters
degree 1160) noted that many of the soldiers stationed in the Huai River
area, which formed the boundary with the Sino-foreign state of Chin,
would regularly cross the border into Chin territory to steal horses. 11
And of course, soldiers committed at least their share of other crimes,
such as gambling, theft, and homicide.
Deserters were an even bigger problem. 12 The danger of deserters'
committing crimes was an old one, and rules concerning it occur in the
T'ang and the early Sung codes. There was a common assumption that
many such men would attempt to return to their native places. Thus one
of the ordinances regarding arrest recorded in the early Sung Code
states:
When prisoners, men being sent on military expeditions, or men being trans-
ferred desert and join up with bandits, the officials nearest their route should
send a report to the jurisdiction in which the families of the escapees live, as well
as to the prefectures and districts near where they deserted that are responsible
for pursuing and arresting them. The jurisdictions that receive these notifications
should forward them to the villages and mutual security units, ordering them
to increase their vigilance while trying the capture the men.13
During the Sung period, the rate of desertion could be extraordinarily
high. A report from 1083 s a ^ t n a t m seven months 3382 men had
deserted from a single prefecture!14 This startling figure vividly indicates
the limits on the state's ability to control the movements of its people.

9 On "evil youths," see Ch'in Kuan, Huai-hai chi, i 7 4 a - 5 b ; Han Yiian-chi, Nan-chien
chia-i kao, SKCSCP 1975, 10.15a. Ch'in Kuan also comments on the role of army
deserters in banditry. Note also the comment in the HCP 309.5a for a description of the
"young hoodlums" near the capital city.
10 For soldiers as robbers, see, for example, SHY, hsing-fa 4.15a, 5.1b, 7.1a; SHY, ping
11.6b.
11 Ts'ui Tun-li, Kung-chiao chi, SKCSCP 1972, 5.i7b-i8b. Cattle theft was another
common crime. See Ch'en Chun, Huang-ch'ao pien-nien kang-mu pei-yao (Taipei: Ch'eng-
wen ch'u-pan she, 1966), 40.4b, 48.11b.
12 Ch'in Kuan, Huai-hai chi 17.4a. Deserters are frequently mentioned as criminals. See,
for example, HCP 53.9a, 293-8a-b, 4 6 5 . u b - i 2 a ; SHY, ping i3.27b-28a, 28a-b;
TFSL 502.
13 Tou I, Sung hsing-t'ung (hereafter cited as SHT) (Beijing: Chung-hua shu-chu, 1984),
28.464.
14 HCP 338.10a.
70
 Crimes and criminals

Soldiers apparently could desert almost at will and, under the law, could
return within a set period of time without fear of penalty. Many criminals
were deserters from the military; most soldiers came from the working
classes of rural and urban China.
When discussing military deserters as criminals, it is important to
bear in mind, however, that a substantial number of men serving in the
military labor battalions called the provincial armies (hsiang-chiin), and
some of those in the fighting units called the imperial armies (chin-chun),
had been sentenced to register in the armies to serve as laborers after
being convicted for major crimes. Thus, when the sources say deserters,
they may in fact be referring to recidivist criminals, men sent to perform
forced labor who were soldiers only in an extended sense. An official
commented in 1204 that "of those sent to penal registration, many flee
while en route, and when they get to their place of registration, they
dread having to find sufficient food. Often the authorities deliberately
tolerate the escapees and do not recapture them." 15
As this material on escaping from registration in the armies suggests,
the Sung judicial administration was plagued by the problem of reci-
divists, men who were in many cases habitual criminals guilty of a
succession of crimes. The general response to recidivism was to increase
the penalties for such individuals and to reduce their chances of benefiting
from the various devices that the state created to enable the reintegration
of ex-criminals into society.16 The general pattern is reflected in a policy
sometimes in force, under which a criminal twice guilty of robbery was
to be sentenced to death, even if he were a follower and not a leader. 17
Whether escaped convicts from the provincial (labor battalion) armies
or runaway soldiers, when deserters cooperated with solders still in
service, the problems were doubly compounded. During the Northern
Sung a Buddhist monastery in Ssu Prefecture on the Grand Canal had
been given the task of burying the many corpses that were taken from
the canal. An investigator in the late eleventh century reported that the
problem was especially serious because
there are many evil deserters hidden in the post stations above and below the
river dikes, who cooperate with the post station troops to commit robbery. They
take the clothes and goods from their owners, kill them, and throw the corpses
15 WHTK 168.1462. See also SHY, hsing-fa 4.64a-b. On the punishments of officials who
allowed escapes, see TFSL 528. For other comments on deserters as criminals, see, for
example, SHY, ping 13.29b; SS 35.670; SHY, hsing-fa 4.56a.
16 See, for example, SHY,, hsing-fa 4.5a; HCP 143.27a; TFSL 502. Hsueh Ying-ch'i,
Sung-Yiian t'ung-chien (Taipei: Commercial Press, 1973), 88.10b repeats the policy
reported in SS 35.684, under which anyone guilty twice of robbery was to be executed,
even if he were a follower and not a leader.
17 SS 35.684.

71
 Law and order in Sung China

into the water. The bodies come to the surface only after several days, by which
time they are already several hundred lid [a lid was approximately one-third of a
mile] from the place where the robbery occurred. Therefore few robbers are
caught. 18

Ex-militiamen or constables also reportedly sometimes became

robbers. In 1086, in an attack on the policies of Shen-tsung's reign, Liu
Chih spoke of the militia-constabulary group called archers (kung-shou)
who, having been mustered out of service and lacking a means of
livelihood, became bandits. 19 Writing in the Southern Sung and de-
scribing a major bandit, Ts'ai K'an observed that he "in the beginning
was just an archer who stole arms." 20 Such men quite naturally formed
themselves into gangs and exerted some discipline over their members.
As ex-military men they recognized the importance of weapons. Often
deserters took weapons with them when they fled, and gangs of robbers
sometimes attacked towns, not only looting them of various valuables,
but also seizing weapons. In the last half of the twelfth century the
official Ts'ui Tun-li (doctorate of letters degree 1160) described a road in
Huai-yin District (in Huai-nan East Circuit) where travelers were ter-
rorized by a large gang of bandits, who "killed, burned, set free one
prisoner from the district jail, and also took official army bows, cross-
bows, implements, and staffs."21 Another report from about the same
time recounts an incident in which local officials, in order to provide an
escort to an arriving guest, took the local constables away from their
posts. Local bandits seized their opportunity, entered the district
town, freed the prisoners from the jail, stole military equipment, killed
some people, and fled. The responsible officials in this incident were
demoted. 22
This combination of rudimentary organizational skills and skills in
arms made these groups especially frightening. Those charged with
enforcing the law sometimes responded by punishing deserters with
great ferocity. We are told of military authorities in the northwest who
beheaded deserters, but only after having branded them and broken
their legs. Another report, from the early Southern Sung, says that when
prefectural authorities captured bandits who were military deserters,
they also executed their families. 23 Finally, in his Po-chai pien, Fang

18 SHY, ping 12.13a.

19 HCP 3 8 9 .6b-7b.
20 Ts'ai K'an, Ting-chai chi, 1.14b.
21 Ts'ui Tun-li, Kung-chiao chi, 5.17b.
22 SHY, ping 12.19b. The improper use of constables as escorts is reported elsewhere also.
See HCP 120.18a.
23 WHTK 167.1453.

72
 Crimes and criminals

Shao describes a sheriff who "in all cases when he seized military
deserters immediately executed them, so that in his district there were no
robbers." 24

Other criminal types

Such deserters were not, of course, the only criminals of concern. The
illicit salt dealers of the capital region around Lin-an in Liang-che
Circuit were characterized as vagabonds, who came from the state-licensed
commercial families called shopkeeper households (p'u-hu), from temples
and monasteries, from military encampments, or from families of either
commoners or literati. 25 The dangers of illicit dealers in salt was a
traditional concern. Sung leaders were certainly aware that the great
uprising in the late T'ang, which broke the central power of the state,
was supported by criminals involved in the salt trade.
Like many societies at many times, Sung China also had its share of
juvenile criminals. The problem of youthful offenders existed throughout
the dynasty. The Northern Sung official Ch'in Kuan (1049-uoo), in
describing the origins of criminals, refers to some as "evil youths in the
villages, who are ill tempered, who beat people to death and bury the
bodies, or who practice counterfeiting and who will not fit in again with
ordinary people." 26 An early Southern Sung official, Han Yiian-chi
(1118-after 1178), wrote of the "youthful banditti" then infesting the
roads to Yen Prefecture in Liang-che Circuit. 27 In the early 1160s such
youths are blamed for illicit dealings in tea and salt. 28 Finally, an official
writing in 1214 described a market at which "the number of residents
increases day by day and there are many merchant hostels. Juvenile
delinquents [o-hsiao] cause trouble, and there are many fights." He
petitioned, successfully, for the appointment of a sheriff at the market. 29
We do not know the social status of these young criminals. Many of
them may have been from the lower strata of Sung society. However, it
is clear that not all wrongdoers were from the working classes of Chinese
society. Indeed, Sung authors warn against the dangers represented by
the literati shih-ta-fu class. 30 Given the chance, members of this group
would use their local power to oppress others, indulging in person or

24 Fang Shao, Po-chai pien (Beijing: Chung-hua shu-chu, 1983), 2-3b~4a.

25 SHY, hsing-fa 2.127a.
26 Ch'in Kuan, Huai-hai chi, SPTK ed., i7.4a-5b.
27 Han Yiian-chi, Nan-chien chia-i kao, 10.15a.
28 Li Hsin-ch'iian, Chien-yen i-lai hsi-nienyao-lu (hereafter cited as HNYL) (Beijing: Chung-
hua shu-chu, 1956), 3407.
29 SHY, chih-kuan 48.85b.
30 Hsu Yuan-chieh, Mei-yeh chi, SKCSCP 1975, 4.6a.

73
 Law and order in Sung China

through their gangs of hangers-on in robbery and extortion. 31 A report

from the late twelfth century speaks of gangs who trafficked in illegal
goods being protected by powerful families that concealed them and
assisted them in their pillaging. 32 Another report from the opening years
of the thirteenth century describes "rich and powerful households who
promote the evil doing of gangs." 33 The collection of case reports called
Che-yii kuei-chien speaks of upper-class bullies, one of whom kept several
tens of hangers-on who would attack or kill to intimidate the local
people, so that "for years the people of the countryside feared them and
did not dare to report about the problem." 34 Such families also bribed
the local clerks to be their agents: "In causing injury to the people,
nothing is worse than rapacious clerks. . . . The literati families of today
nourish such people. The rich families use them as their dogs." 35 A
report from the opening years of the twelfth century even speaks of
powerful families who had made their own prison instruments and who
chained and tortured their poorer and weaker neighbors. 36 And in the
capital area it was reported that relatives of officials, and even mem-
bers of the imperial clan, were involved in illegal activities, including
robbery. 37 The thirteenth-century official who said that "Liu-tu Fort is
not a fort controlled by the public authorities but, rather, a jail for the
use of the powerful families; Yii-shan District [present-day Yii-shan in
Kiangsi] is not a district controlled by the authorities but, rather, a
flogging place for the powerful families" 38 was speaking hyperbolically
but certainly in earnest.
Finally, the authorities were well aware that a great many officials
committed crimes. The general impression given by the sources was that
the great majority of such crimes by officials involved in one way or
another some abuse of their powers. Perhaps most commonly they were
guilty of corruption in office. There is even one fascinating, but un-
substantiated, statistic on officials and crime. The collection of random
notes called the Po-chai pien says that in the years 1075 *°I 0 7 ^ some 2592
officials had been sentenced to penalties, from tattooed penal registration
to commutation, 39 this at a time when the total number of civil service
officials probably did not exceed several tens of thousands.

31 Fan Chun, Hsiang-ch'i chi, 15.149-150.

32 SHY, ping 13.25b.
33 SHY, hsing-fa 4.63a.
34 Cheng K'o, Che-yii kuei-chien, TSCC ed. (hereafter cited as CYKC), p. 66.
35 SHY, hsing-fa 4.6313-643.
36 SHY, hsing-fa 2.133b.
37 SHY, hsing-fa 2.147a (1133).
38 Ming-kung shu-p'an Ch'ing-ming chi (Beijing: Chung-hua shu-chu, 1987), p. 33.
39 Fang Shao, Po-chai pien, p. 58.

74
 Crimes and criminals

Religious groups
Traditional Chinese governments were notoriously hostile to the trans-
mitters of many popular magical, religious, or quasi-religious beliefs.
This category included individuals who were teaching and practicing
black magic sorcery or groups devoted to religious ideas traceable to
the major world religions but espoused and expressed in ways that
were felt to threaten the state and its established values. "People who
have adopted heretical beliefs" are commonly accused of becoming
criminals. 40
Private households were forbidden to possess prognosticatory texts,
astrological charts, or books on divination, for fear that such works
would lead them to question the legitimacy of the existing order. 41 All
those who made books on prophecy or portents, transmitted them, or
used them to delude others were to be severely punished. 42 The mere
ownership of such works, even if they were never used, carried with it a
penalty of two years of penal servitude. 43 Even works that resulted in
harmless prophecies, for example, predicting floods or droughts, were
forbidden, and their possessors were liable for one hundred blows of the
heavy rod. 44 A decree of the Later Chou (951-60), which was copied
into the early Sung Code, specifies that books of this sort were to be
seized and burned. 45
Another much feared and related practice was black magic sorcery,
especially the making or use of the magic poison called ku. In The
Vermillion Bird, Edward Schafer describes it as follows:

Ku was evil sorcery, erotic charm, malignant disease - and the artful or natural
causes of these things. It was the black magic of Nam-Viet. Usually it has been
described as a poison, concentrated from the venoms of insects and reptiles.. . .
In medieval times it was the common belief that the poison was prepared by

40 Ch'in Kuan, Huai-hai chi, i7.4a~5b.

41 Chang-sun wu-ch'i, T'ang-lii shu-i (hereafter cited as TLSI) (Taipei: Commercial Press,
1970), 2.82; SHT 9.155. See also a decree dated 1004. Anonymous, Sung ta-chao ling-chi
(hereafter cited as STCLC) (Taipei: Ting-wen shu-chu, 1972), p. 733. And see a case of
1016 that involved an official who was expelled from the civil service and sent into
penal exile (SHY, hsing-fa 4.8a).
42 According to the T'ang Code and the early Sung Code, such criminals were to be
strangled. By the time the TFSL was compiled in the early thirteenth century, the
penalty had been reduced by one degree, to exile at one thousand li.d TLSI 3.56; SHT
18.289; TFSL 251.
43 TLSI 3.56; SHT 18.289.
44 TLSI 3.56; SHT 18.289.
45 SHT 18.290. Such book burning was a regular Sung practice. See, for example, an
order dated 1121. SHY, hsing-fa 2.83a-b.

75
 Law and order in Sung China

putting poisonous creatures in a sealed vessel, where they ate each other until
their poisons were concentrated in a single survivor.46
In the T'ang and early Sung codes, people who practiced or taught
this magic were severely punished. Women sentenced to exile for other
crimes were allowed to remain in custody at home, but "those who make
or teach the making of ku poison, together with all those living as
coresidents in their households, are to be sent into exile, even if an
amnesty has been issued." 47
Other sorts of magic aimed at killing people were also heavily punished,
even if no one had actually been killed. "All those who hate someone
and for that reason make 'hate demons' [yen-mei] or amulets for cursing,
in order to commit murder, are to be punished as if guilty of premeditated
murder, with the penalty reduced by two degrees." 48 Despite the severe
penalties laid down in the codes, people continued to transmit the arts of
• • 4-Q

incantation.
Individuals who performed such acts of sorcery were severely punished;
organized groups whose members or leaders claimed supernatural
powers were even more worrisome. Their members, and especially their
leaders, might be punished with great severity. 50 Chinese dynasties in
traditional times were always chary of any organized religious groups
not under their control, recognizing that the combination of formal
organization and religious commitment could be especially dangerous to
the state. Such religious groups were seen as likely to resort to antisocial
or politically disruptive behavior. Such problems were often on the
minds of Sung leaders, because China in the Sung seems to have been
filled with some rather bizarre religious cults. Their practices could
directly contradict the ideas of the dominant social system. An official
wrote in 1115 about members of a sect whose members burned their
foreheads and backs and cut themselves, mutilations similar to those
practiced by some Taoist priests today in Taiwan. T h e decree re-
sponding to his report is thoroughly Confucian in tone: "Mutilation
is injurious to good customs. How much worse is it when barbarian
customs are practiced among Chinese?" 5 1 An even more bizarre cult
was found in the valleys of northern H u n a n . A report from the early
46 Edward H. Schafer, The Vermillion Bird (Berkeley and Los Angeles: University of
California Press, 1967), p. 102.
47 SHT 3.45.
48 TLSI 3.53; SHT 18.284.
49 STCLC 735.
50 See, for example, an order dated 1019 calling for the execution of cult leaders. HGP

51 SHY, hsing-fa 2.65a. See also a report of a cult in Yiin Prefecture in Chiang-nan West
Circuit, which mutilated women and children. SHY, hsing-fa 2.122a.

76
 Crimes and criminals

eleventh century describes a religious cult in this region that practiced

human sacrifice; it was probably the same religious cult described during
the Southern Sung in the same region. According to the Southern Sung
reports, this group, during each year that included an intercalary month,
kidnapped and murdered children, a practice they called "choosing the
living." 52 Cult leaders from such groups were very harshly punished,
often being executed. 53
Some of these groups can be traced to heterodox offshoots of Taoism
or Buddhism. 54 But even when such groups could be clearly associated
with these accepted religions, Chinese lawmakers often thought that
their members, who were forced to meet clandestinely, were guilty of
outrageous social and sexual behavior. A report from the late 920s, on
heretical Buddhist believers, comments that "sometimes Buddhist clergy
and laity are ignorant and thoughtless. Men and women live together
illicitly, forming themselves into groups, gathering at night and dis-
persing at dawn, speciously proclaiming and handing down a 'Buddhist
law society' [fa-hui], clandestinely being loose in their morals." 55
The phrase "gathering at night and dispersing at dawn," with its
overtones of illicit association between the sexes, recurs frequently in
Sung sources. The authorities were obviously very worried about this
sort of behavior. Indeed, an edict in 1035 offered a substantial reward,
thirty strings of cash, to anyone who was able to seize such sectaries or
who informed on them leading to their capture. (We can get some idea
of the value of this reward by noting that thirty strings was the estimated
cost to the state of supporting a postal worker for one year.) This report
concerned the western circuits but people accused of similar practices
could also be found in the east. 56
During the Sung dynasty, the authorities were even more concerned
about the Manichaeans. 57 Mani (215-73) drew on elements from a
number of traditions to found a dualistic religion, which asserted that
good and evil were engaged in an all-but-eternal struggle. Man, a crea-

52 HNYL 2636; SHY, hsing-fa 2.152a.

53 See, for example, STCLC 638, 736. For a further description of cults practicing human
sacrifice, see Kawahara Masahiro, "Sodai no sai-jin sai-ki ni tsuite" (On cultic human
sacrifice in the Sung dynasty), Hoseishi kenkyu 19 (1969): 1-18.
54 This would seem to have been the case with the groups discussed in a memorial of 1114
in which an official asks for stricter controls over Societies for Fasting and Incense
Burning. See SHY, hsing-fa 2.61b.
55 SHT 18.290.
56 HCP 117.19a; SHY, hsing-fa 1.49b, 2-78a-b, 81a, n i b .
57 For a discussion of the Manichaeans in the Sung, see Brian E. McKnight, "The
Rebellion of Fang La" (M. A. thesis, University of Chicago, 1964), chap. 2. For a more
general discussion of Manichaeans in China, see Samuel Lieu, The Religion of Light
(Hong Kong: University of Hong Kong Centre of Asian Studies, 1979).

77
 Law and order in Sung China

ture born of this struggle between light and darkness, contains the seeds
of light, which can be liberated if the proper religious practices are
followed.58 This new religion spread rapidly in the centuries after his
death, from the shores of the Mediterranean to the borders of China. 59
Manichaeism, which first entered China during the closing years of the
seventh century, became most widespread and open during the first half
of the ninth century, under the protection of the Uighur Turks who had
adopted Manichaeism as their state religion. When the Uighurs lost
power in the early 840s, the Chinese authorities moved quickly to
suppress Manichaeism. The surviving believers thus became part of an
underground cult which attempted with some success to disguise itself as
a Taoist sect. There is little information on the history of this religion
in China during the period from its suppression in the 840s until it
reappears in Chinese records from the eleventh century onward, though
its members were involved in one unsuccessful rebellion in the early
tenth century. In the eleventh century, however, there are reports about
the prevalence of the religion in southeast China, where local authorities
were accused of being indecisive in their suppression of the illegal sect.
During the Sung the Manichaeans, who seem to have been especially
numerous in eastern Chekiang, were often involved in trade, and, in
addition to offending Chinese sensibilities by failing to observe proper
decorum between the sexes, were stereotyped as people who would do
anything for a profit, without being concerned about its legality. They
thus affronted espoused elite values on almost every level. The state
authorities repeatedly tried to suppress the religion. An early Southern
Sung official, Chang Shou (d. 1138), recommended tactics that epitomize
the standard state response to such dangerous cults. The authorities, he
said, should first separate the leaders from their followers, severely
punish the leaders, and publicly and widely announce the punishment.
Second, prohibitions against nighttime gatherings for worship should
be publicly proclaimed and enforced.60 By cutting off the leadership
and forbidding the functions that would allow the continuing trans-
mission of the doctrines, the state could in time suppress the movement
without having to punish all its adherents. It seems likely that in some
areas the local authorities did attempt to suppress the religion, for
the Manichaeans were involved in a number of uprisings against the
58 On Manichaean theology, see F. G. Burkitt, The Religion of the Manichees (Cambridge:
Cambridge University Press, 1925).
59 Some authorities believe that Mani "translated" his ideas into two different idioms,
gnostic Christianity in the West and Iranian speculation in the East. See Geo
Widengren, "Mesopotamian Elements in Manichaeism," Uppsala universitets arsskrift, no.
3 (1946).
60 Chang Shou, P'i-ling chi, TSCG ed., 3.33.
78
 Crimes and criminals

government. In the Sung, Manichaeism was a classic example of an

illegal religion forced into functioning like a secret society and also into
covert or overt resistance to the state. 61

Non-Han criminals
Most of those involved in criminal activities were Han Chinese, but in
border regions non-Han minority peoples also participated in crimes.
Along the northern frontier this meant such things as cross-border
smuggling, horse theft, and the sale of illegal goods. In the south both
the non-Han minority people called the Man and the Li b minority
people of Hainan Island were accused of criminal behavior. Such peoples
might be punished differently than the Chinese were. The Man, for
example, were allowed to pay fines rather than suffer the regular Sung
penalties. 62
Dangerous incidents in the regions inhabited by the Man peoples were
endemic, occurring most frequently in the middle of the eleventh century
but never dying out entirely. 63 The great writer and statesman Ou-yang
Hsiu estimated in 1043 t n a t m t n e half-dozen or so prefectures that were
most affected, there were several thousand Man bandits in all, the gangs
ranging in size from about ten men to several hundred. 64 At the same
time the official Yii Ching wrote a number of memorials on the Man
problem. In one he reported that of the whole group, only a hundred or
so could be classified as leaders. The appropriate policy would be to
follow the traditional pattern, prosecuting only the leaders and pardoning
their followers. He referred to the Mencius: "I have heard that Mencius
said, 'A policy of benevolence can protect all within the Four Seas; a
policy without benevolence will not be capable of protecting your wife
and children.' " 6 5 He also agreed with Ou-yang Hsiu about the import-
ance of selecting good officials for the region. "If the court selects chief
officials and they select bandit-catching officials and patroling in-
spectors, who gather together courageous officers to seize the bandits,
and if these men are repeatedly granted rewards, then the bandits will
certainly be destroyed." 66
Inappropriate action by such officials could exacerbate the troubles:

61 For the later history of the Manichaeans in China, see Paul Pelliot and Edouard
Chavannes, "Un traite manicheen retrouve en Chine," Journal Asiatique, 10th series, vol.
18 (1911) and n t h series, vol. 1 (1913).
62 H C P 4 6 2 . i a - b .
63 See, for example, HCP 147.40-9^.
64 HCP 143.30a.
65 Yii Ching, Wu-ch'i chi, tsou-i shang 12a.
66 Ibid., tsou-i shang 12a-13a.

79
 Law and order in Sung China

I inquired and learned that in Ching-hu South Circuit, because the Man bandits
have come forth from their dwelling places, the judicial intendant, Chao Shih,
has ordered those people who lived near the mountains not to till their fields, to
burn their dwellings, and to enter the cities to live. Thus these people have lost
their patrimonies and so have joined with the bandits.67
In another essay Yii Ghing expanded on this theme:
I have heard that because of the bandits in Ching-hu South Circuit, the people
are forbidden to farm within twenty lid of the hills. I do not know from whom
this order came, but it forces people to become bandits. I feel that from the
middle rank down, people simply have no accumulated wealth. For their daily
expenses they must depend on farming. If they do not farm for a single year, they
will suffer hunger and cold. Now the bandits burn their homes and pillage their
goods, and the government forbids them to farm. They have nowhere to turn.
They do not want to be robbers, but what else can they do? . . . When the bandits
first arose, there were only a hundred of them. The authorities called in troops to
suppress them, and the troops killed the commoners near the mountains. Those
who survived this were forbidden to plant crops, so for a thousand lid people lost
their livelihoods. If they did not turn to banditry, what could they do? Therefore
in half a year there were three to four thousand bandits. This is a matter of
wanting to suppress bandits but instead forcing people to become bandits.68
In this region another key to the continuing ferocity of the resistance was
that
the officials and the military kill ordinary people in order to claim rewards.
Therefore the residents do not enter the cities but instead join with the bandits to
avoid being killed. The countryside has been deserted by the farmers, so that the
bandits have nowhere to pillage and must attack the towns. Fortunately they
have not taken any towns. If that happens, the evil consequences will be great,
for even if you wanted to pardon them you could not.
It would be necessary to summon people to surrender, to soothe their
feelings, and to indicate clearly that they would be forgiven by the
authorities. 69
The authorities in these areas were especially concerned about the
dangers of collusion between non-Han criminals and Han criminals,
especially army deserters. Acting together, such criminals could pose
dangerous security problems. In the early twelfth century an official
wrote about an uprising among the Li b people on Hainan Island,
where many of the worst Sung criminals were sent for registration. He
commented:
67 Ibid., tsou-i hsia 19b.
68 Ibid., tsou-i shang 16b.
69 Ibid., tsou-i shang 5a. In another memorial on the same subject he expanded on his
interpretation. Ibid., shang 17b-18a.

80
 Crimes and criminals

In the various prefectures on Hainan there are escaped registered criminals who
go into the Lib borderlands. Men who are registered in the prefectures of Hainan
are criminals whose crimes were marked by truly evil circumstances, or they are
men sentenced to death and spared. Recently, from both the East and West
circuits (of Kuang-nan), troops have advanced to attack the bandits, but before
anything happens the men fraternize and are made welcome. Many of these
criminals are men who bear large tattoos [i.e., are very serious criminals]. They
add to the bandits' authority, and the borders are disturbed.
The Court approved his suggestion that serious criminals who would
ordinarily have been sent to Hainan for registration would temporarily
be sent elsewhere.70

What were the crimes?

The kinds of crimes recognized or created by the state reflect the con-
cerns of both rulers and ruled and are embodied in complex literate
societies in codes of laws. Codes are normative. They spell out what the
codifiers say should and should not be done. Although they tell us
relatively little about actual social behavior, they do mirror to some
degree the attitudes and feelings of their creators. Chinese law codes, like
law codes elsewhere, were founded on the lessons of experience. They
reveal the forms of wrongdoing that had plagued the state authorities in
earlier times, to which lawmakers had responded by generalizing from
specific instances to create general rules.
Any attempt to understand the concerns of the people of Sung times
should begin by analyzing the earliest Sung criminal code, the Sung
hsing t'ung of 962. This code, compiled almost at the beginning of the
dynasty and based on earlier prototypes that had long been in use,
contains many of the penal rules used during the Sung, though by no
means all. Large numbers of penal rules that applied specifically to the
military were excluded from the code, and a host of other penal rules
were compiled in collections of edicts. 71
This early Sung Code is largely a repetition of the T'ang Code of 737
and, like that T'ang Code, contains twelve sections. Excluding the sec-
tions on general principles, arrest and flight, judgment and prosecutions,
and miscellaneous articles, there are eight substantive sections, those on
(1) the imperial guard and prohibitions, (2) administrative regulations,
(3) households and marriage, (4) public storehouses and granaries, (5)

70 SHY, hsing-fa 4.33b.

71 For examples of such penal laws for the military, see Tseng Kung-liang, Wu-ching
tsung-yao, SKCSCP ed. On the compilations of edicts, see McKnight, "From Statute to
Precedent," pp. 111-32.

81
 Law and order in Sung China

unauthorized levies, (6) violence and robbery, (7) assaults and accusa-
tions, and (8) frauds and counterfeits.72
These eight sections are composed of two large groups. Four sections
contain rules that are concerned with the acts of those in government
service (imperial guards and prohibitions, administrative regulations,
public storehouses and granaries, and unauthorized levies), and the
other four apply to the people at large as well as to government servants
(households and marriage, violence and robbery, assaults and accusa-
tions, and frauds and counterfeits). A proper study of violations of
the first four sections, most of whose rules are applicable primarily
to government servants, would require an in-depth analysis of Sung
government systems. We shall be concerned largely with the last four
sections.
The section on households and marriage contains twenty-five articles
in its three chapters. These pertain to crimes in the technical sense,
because the statutes give penalties for infractions but include acts which
in our system would not be classed as crimes. The rules concern such
questions as the proper disposal of the goods of foreign merchants who
die in China, how to treat people who become Buddhist or Taoist clergy
under improper circumstances, punishments for those who illegally
divided households in order to escape taxation, encroachments on prop-
erty, and illegal marriages.
Most of the major crimes are found in the remaining three sections.
The nineteen articles under violence and robbery cover such crimes as
sedition, rebellion, homicide, the manufacture of poisons, certain as-
saults and batteries (including sexual assaults), burglary, grave robbing,
heresy, robbery, destruction of property, theft, arson, and extortion.
The twenty-six articles under assaults and accusations include some
homicides, some assaults and batteries, some deprivations of property,
and false and improper accusations. Frauds and counterfeits has only ten
articles, covering the illegal manufacture of seals, the impersonation
of officials, improper manufactures, coercion of people into criminal
behavior, the illegal misuse of government horses, and improper re-
porting to the authorities. In short, these four sections contain many of
the rules on the major crimes. The penalties given in this code are the
traditional Five Punishments (wu-hsing) inherited from the T'ang,
beating with the light rod (ta) in five degrees from ten to fifty blows,
beating with the heavy rod (chang*) in five degrees from sixty to one
hundred blows, penal servitude (t'u), exile (liu), and death (ssu).

72 I am using the translations of these terms suggested by Wallace Johnson, The T'ang
Code: General Principles (Princeton, NJ: Princeton University Press, 1979).

82
 Crimes and criminals

Historiography
When we turn from the code to other sources to see what crimes Sung
men (almost exclusively Sung elite men who served as officials) dis-
cussed, there is a marked difference in emphasis. This is hardly sur-
prising. Codes are normative. In China they aimed at covering all
possible wrongdoings. Therefore, crimes that would have been considered
truly shocking and terrible may occupy no more space in the code than
do trifling misdemeanors, and crimes thought of as dangerous may be
covered in only a few lines of text.
Edicts were responses to specific immediate problems. Repeated
edicts on a topic reflect the authorities' failure to deal with a serious
situation. The numbers of examples we have on a specific topic thus
suggest the continuing concern of the state. When men feel it necessary
to describe in detail the methods that they have used to handle a
particular situation, presumably they feel that others might benefit from
their experience.
Because they were members of the ruling elite, the men who created
the records were less concerned with the frequency of crimes than
with whether the particular crime in question threatened their positions
and values. Like most groups of people they did not hesitate to identify
their own interests and understandings with the general social interest.
Sometimes, no doubt, their interests were in fact the same as those
of most groups in Chinese society. Most people, in China as elsewhere,
whether or not they are members of the elite, abhor homicide, robbery,
rape, and other serious felonies. When there were clear differences
between the interests of the Chinese elite and other groups in society, the
response of the elite was that the values of the others were simply wrong
and so needed to be changed. When they felt that perhaps their values,
the "correct" values, were not shared, they responded by trying to
educate people to propriety, through what they called "educational
transformation" (chiao-hua) and by punishing improper behavior. Be-
cause members of the ruling group viewed their relationship to ordinary
people as analogous to the relationship of parents to children, the trans-
formation they sought was reflected not simply in their attempts to
suppress what they saw as social crimes (like gambling) but also in their
attempts to promote "good" mores, which would make ordinary people
conform more closely to the elite ideal. These reformers stressed filial
obedience, the proper hierarchy of power distributed according to age
and sex, and, in general, a respect for superior authority, perhaps out of
a concern that the common people might not truly support these but-
tresses to social stability.
83
 Law and order in Sung China

In their writings about the need to remold the people and on law-
breaking, Sung writers concentrate most often on those crimes and
customs that seemed to threaten social order (and thereby their own
positions). As a result, scholars studying law and order in traditional
China face a serious historiographical problem. Some types of crimes,
like minor assaults, gambling, fraud, and extortion, are little discussed in
the sources, though they seem in fact to have been quite common. 73
Assaults against officials, which were punished with great severity, seem
to have been relatively rare, and our authors, who were almost all
officials, were little concerned with the brawling of commoners or their
other minor wrongdoings as long as they did not escalate into wide-
spread unrest. Because the Sung authors were overwhelmingly pre-
occupied with acts that seemed to them to threaten with special urgency
the control of the ruling elite, they usually described two general cat-
egories of crimes — corruption among government officials and clerks and
robbery-banditry - far more often than all other crimes combined.

Corruption
A thorough study of the problem of corruption among officials and clerks
would have to describe the spectrum of attitudes toward corruption, the
variety of corrupt practices, the social and other characteristics of those
found guilty, and their treatment by the state (with appropriate analyses
of the ideological implications of these matters). Before answering these
questions we would have to know more about the functioning of the
Sung civil, clerical, and military services. Here, however, we can touch
on only a few aspects of this important topic. In brief, corruption or
rapacity most commonly meant the taking of bribes from the people or
the misappropriation of public goods.
The authorities were extremely concerned about the taking of bribes,
and nowhere more so than when judicial officials were involved. A rule
contained in the thirteenth-century code called the Ching-yuan t'iao-fa
shih-lei states: "Supervisory officials who accept valuables in return for
perverting the legal process are liable for strangulation if the goods in
question have a value of twenty lengths [pi] of cloth (or twenty-five
lengths if the functionary is unsalaried)." For simple corruption not
involving perversion of justice, the penalties were not quite so severe.
The passage just quoted goes on to say: "If the appropriate sentence
calls for exile, or if it involved corruption but not perversion of justice,
for a value of fifty lengths of cloth, the sentence should be criminal

73 Wang Chih, Hsiieh-shan chi, 3.24.

84
 Crimes and criminals

registration in a base citadel command." 74 At approximately the time

when the Ch'ing-yuan t'iao-fa shih-lei was published, a length of cloth
would have been worth (very roughly) four strings of cash, so the
authorities were talking about corruption involving perhaps eighty to
one hundred strings of cash. 75
The seriousness of misappropriating funds is clearly indicated by a
decree from the early thirteenth century that modifies the earlier rules:
In all cases, those who are indicted for rapacity according to the statutes,
because they have used official goods that ought not to have been used, from now
on are to be tried for having committed a private infraction [i.e., an infraction
committed for the sake of personal advantage] if they have taken the goods for
their own private uses. If they have taken the goods for public purposes, they
should be tried for having committed an administrative misdemeanor [literally, a
"public crime"]. 76

Both bribery and misappropriation for private motives could be

punished very harshly, depending largely on the value of the materials in
question. The penalty could even be death. Beating plus exile, some-
times accompanied by enslavement of the families of the guilty, was a
common punishment. Less serious cases often led to expulsion from the
civil service, exclusion from holding certain sorts of posts in the future,
or confiscation of the property of those guilty. 77
The severity of these penalties was in fact even greater than appears
on the surface. Officials guilty of many sorts of crimes were able to
commute their sentences to various administrative sanctions or to pay-
ments of fines. In many such cases the penalties would be automatically
reduced because of the criminal's official status. However, for certain
crimes, corruption among them, such privileges were not allowed. 78

Banditry
Contemporaries who wrote on crime in Sung China clearly felt that
suppression of banditry {tao-tsei) was central to the problem of law and
74 TFSL 537. On base citadel commands, see Chapter 12. Although the text here and
elsewhere in the TFSL does not specify the kind of cloth involved, passages from the
T'ang Code (which were adopted in the early Sung Code) suggest that the measure
referred to chilan, a thick, loosely woven, raw silk fabric. Where the T'ang Code does
specify the type of cloth being referred to in determining sentences, this seems always to
be chuan. For these references, see Wallace Johnson, T'ang-lii shu-iyin-te (An index to the
T'ang-lii shu-i) (Taipei: Wen Hai, 1964), pp. 144-45.
75 I am basing these very approximate estimates on price figures kindly supplied to me by
Robert Hartwell.
76 WHTK 167.1456.
77 WHTK 167.1453; SS 199.4991.
78 The Sung system of privileges will be discussed in more detail in Chapter 14.
85
 Law and order in Sung China

order. The attitudes underlying the reaction of the elite to the problems
presented by bandits are expressed in numerous memorials on bandit
suppression. They provide us with our most vivid portrayal of the
bandits in action. The official P'eng Kuei-nien (i 142-1206) wrote of the
campaign against the bandits in Fukien:
On the twenty-sixth day of the eighth month the bandits came by way of Liang-
tzu Gully in An-fu. They crossed P'ing Riding79 and divined at the Dragon King
Shrine in Ta-an. When their divination did not succeed, they gathered a crowd
at the home of the Chou family in Tung-kang. On the twenty-ninth, Hsieh Yen-
hsiang ordered four soldiers to reconnoiter. They encountered the bandits at the
embankment of the Chou family and were captured. Two of them were killed,
but two escaped and returned to report. These men were from the command of
Ma Hsi, a military officer serving as the border control patroling inspector.
Hsieh Yen-hsiang was familiar with the place where the bandits were located.
Therefore he used Ma Hsi's troops to guard the village, and he himself led a
group to the east ridge. The bandits were lined up in the fields before the gate
of the Chou family. The fields were all muddy, with a path only a foot or so
wide. The bandits were camped by the fields. Our troops drew out their bows
and crossbows. One bandit was tall and bearded. He advanced to the fore, and
Hsieh Yen-hsiang shot him with an arrow, so that he fell forward into the mud.
Our troops cut off his head. . . . The [other] bandits lost heart and our troops
were inspired. By the late afternoon they had captured twelve bandits. The
bandits gradually retreated. . . . When I reached the area of the White Cloud
Temple, the bandits had recently withdrawn. I questioned the local people.
They all said that the bandits had remained in the ridge area for three days.
Forty or fifty had been wounded. The bandits themselves killed the wounded
who could not withdraw quickly and then left. At Small Mountain there is a
locally influential man named P'eng Tao. He carried out the order to capture
bandits. He searched the mountain ridges and found a number of bodies under
the trees. They all had suffered grievous wounds and died. This was the first
time that people realized that the tea bandits in P'ing Riding were not a trifling
matter.80
The officials were sometimes concerned not just with the bandits
themselves but perhaps even more with the possible increase of local
vigilante power among the local elites, who could use their power to
abuse ordinary people and even threaten the control of the state. The
official Li Kou (1009-59) wrote:
Bandits are not capable of causing a disaster to the world, but they can summon

79 Ridings (hsiangh) were one of the types of territorial subdivisions of districts in use
during the Sung.
80 P'eng Kuei-nien, Chih-t'ang chi, 11.135- These bandits were called tea bandits, apparently
because of their violations of government controls on the production and distribution of
tea, though in practice they also seem to have engaged in ordinary banditry.

86
 Crimes and criminals

forth such disasters. Those who kill in order to eat are hated by others. How
could such men cause a disaster to the world? When robbers are numerous, there
are many killings. The ordinary people who are the enemies of such bandits
are unable to take revenge. Therefore treacherous scoundrels arise among local
leaders. The bandits are the enemies of the people, but these treacherous
scoundrels use the excuse of exterminating the bandits to oppress the people. . . .
Those who murder the people are bandits of the people; those who suppress the
bandits are bandits against the state.81
The great frequency with which these men discuss banditry is not, of
course, an index of how common it was relative to other crimes. In any
society, only an indeterminate fraction of crimes committed are reported
to the authorities. 82 It also seems clear that the rate of reporting varies
with the types of crime and the social circumstances. Some crimes, like
rape, are apparently greatly underreported in the contemporary United
States, and we may be almost certain that they were even more under-
reported in traditional China.
If we consider why banditry loomed so large to the authors of our
sources, some obvious reasons stand out. First, banditry was often
obvious. Even if the locally responsible authorities wanted to conceal
such crimes, it would have been difficult for them to do so. Second,
banditry was a type of crime that the state found particularly threaten-
ing, because it involved violence and presumably some degree of organ-
ization and cooperation. The traditional Chinese state was always
particularly sensitive to the growth of any organized centers of resistance
to its own powers, especially if those centers had the potential for
growing into significant threats to the state's own pretensions. Therefore,
banditry, smuggling, the production and distribution of illicit goods,
and other forms of criminal activity that required the cooperation of a
number of individuals were of special concern to the state. Third, ban-
ditry may have seemed an "attractive" crime to the Sung bureaucracy
because it could be quantified. The numbers of bandits captured or not
captured could, at least in theory, be counted, and so the local officials'
success rate could be measured, thereby giving a (perhaps spurious)
objective appearance to their personnel evaluations. Finally, the bulk of
reports on banditry is swelled by the officials' tendency to describe both
foreign raiders and native criminals as bandits.
The problems presented by banditry were made more intractable by
the character of the official responses. When few in number, bandits
could, in the words of one twelfth-century official, "hide in the moun-

81 Li Kou, Chih-chiang Li hsien-sheng wen-chi, SPTK ed., 22.3a.

82 See Wilson and Herrnstein, Crime and Human Nature, pp. 30 flf.

87
 Law and order in Sung China

tains, coming out at night to make minor raids and scattering again by
the time the officials are informed." Worse yet, ordinary people, seeing
the official inaction, did not report what they knew. Or they might be
driven by official brutality into alliance with the bandits. In words that
sound all too familiar in this century of guerrilla wars, this official went
on to say that the soldiers complained that because the bandits had no
fixed locations, they were difficult to catch.
[But] the bandits do have to eat during the day and sleep at night. They have
goods they must sell. In coming and going during their plundering, the bandits
at least must cover some miles. How is it possible that no one sees them?
However, if people say they have seen the bandits, the officials will just think
that the informers are working with the bandits [and so will] seize them, bind
them, and beat them. Having gotten nothing from them, the authorities will
release them. The officials thus block up their eyes and ears, so that the plans of
the bandits are numerous and the plans of officials are wanting. After a year or
so of such punishment, the people, even if they were very close to the bandits,
would not be willing to say anything to their neighbors, much less the officials.
Therefore, the bandits are given advance notice of official troop movements,
whereas the official troops would not know it even if the bandits were ten feet
away.83

State response to banditry

The state was supposed to have a measured series of responses to
banditry. The first step was to attempt to split leaders from followers by
offering amnesty to those who surrendered. This was to be accompanied
by steps to alleviate the conditions that had led to banditry in the first
place. If these actions were ineffective, the state would begin to employ
force, gradually, increasing the level of force as necessary until the
bandits were destroyed. Unfortunately, in practice the local authorities
frequently waited for the arrival of major military units, and even these
often proved ineffective. The Southern Sung official Chao Shan-k'uo
vividly described the process and its flaws:
South of the lake and west of the Yangtze River several tens of poor people have
become bandits, taking up bows and knives, harassing the prefectures and
districts, and causing concern to the Court. The regional military commissioner
has, over an extended period of time, sent in generals with several thousands of
troops, [but] to no avail. Still, the Court's urgent orders have been delayed. Why
so? Is it not a case of "swatting at oxflieswill not kill lice" or "a crossbow built
for cave rhinoceroses cannot kill rats"? . . . The regular armies must be fed and
housed according to fixed arrangements. They must move in an ordered way.

8 3 L u Y u , Wei-nan wen-chi, 13.11b.

88
 Crimes and criminals

Halberds and spears are their weapons, and armor their clothing. Without
orders they will not advance. Without drums they will not line up. Yet you wish
them to move swiftly into dangerous mountainous terrain. Their opponents can
leap over crags and valleys, walking on thorns. . . . Their hideouts are difficult to
locate, and their patroling can never be comprehended. . . . Can the regular
system be used to respond to this? The mountains' loftiness and the forests'
thickness are like a screen. The climate is unhealthy. It is easy to catch diseases
from the miasma. The soldiers come and go. The officers and men arrive but do
not fit into this region, so many die. . . . What do the bandits have to fear? . . . In
this area of several hundred lid the powerful families have numerous followers.
Because this is bandit country, the powerful all have strongholds to protect
themselves from attack. They also safeguard their houses. The people and the
bandits understand the situation and exchange things with each other. Yet, on
arriving, the official troops seek food for their stock. They order local households
to provide labor. They even kill people's dogs and chickens. Sometimes, because
they are suspicious, they abuse their authority to the point of killing people.
Thus they terrify the people, whose anger is extreme. Because of this the bandits'
sources and growth remain a secret. No one is on the side of the government
troops. If the troops move even an autumn hair's worth, this will be known to
the bandits. . . . Moreover, the troops say "if we fight and win, this will not
necessarily be considered martial, and so our rewards will not necessarily be
good. If we fight and lose, then our punishments will certainly be heavy." 84

Terminology
The word banditry as I am using it here is a translation of the Chinese
tao-tsei. The terms tao means "robbery," "taking openly." To take openly
with the threat of force was called chiang-tao, "robbery by strength," but
the threat of force did not necessarily mean the force of arms. The
authorities thus can speak of robbery with force, but without weapons,
just as they can speak of thieves, that is, those who steal clandestinely, as
either armed or unarmed. 85 (Theft, to steal without confrontation, also
included burglary.) Tsei, "pillaging," which is used by itself to describe
foreign enemies, seems in the Sung usually to imply the use of arms.
Within the categories that resulted from the coming together of these
factors, the Sung authorities distinguished the seriousness of the crimes
by comparing the value of the goods taken, using measures that were
stated in terms of lengths of cloth but were convertible to coinage at set
rates.
A brief history of the crimes of theft and robbery from the T'ang
through the Sung may illuminate more general problems in under-
84 Chao Shan-k'uo, Ying-chai tsa-chu, i . i 8 a - i g a .
85 See, for example, a report from 1098 that speaks of trying certain criminals according to
the laws on "forceful robbers who do not carry weapons." HCP 5oo.5b-6a.

89
 Law and order in Sung China

standing crime and criminals in Sung China. The relevant section of the
T'ang Code of 737, and the early Sung Code, Sung hsing-t'ung, contains a
number of provisions concerned with thefts and robberies. These are
arranged in order of descending seriousness, from theft of major imperial
ritual objects, to the taking of religious objects including Buddhist
images, to the opening and robbing of graves, to ordinary robbery.
These T'ang rules, as reaffirmed in the Sung hsing-t'ung, form the founda-
tion of later Sung laws against robbery and related crimes.
The statute pertaining to robbery and theft is as follows:
In all cases of [unarmed persons] committing robbery by strength (chiang-tao), if
no booty at all was taken, the penalty for the robbers is two years of penal
servitude. If booty worth one bolt of cloth [ten Chinese feet] was taken, the
penalty is three years of penal servitude. If booty worth two bolts was taken,
increase the penalty by one degree [to exile at two thousand lid]. If the value of
the booty reaches ten bolts of cloth or someone is injured, the penalty is death by
strangulation. If there is a homicide in the course of the robbery, then the
penalty is death by decapitation. ("Robbery" means to take valuables with
threat or when armed, whether one is first forceful and thereafter takes the
valuables, or takes the valuables first and is thereafter forceful. This also applied
to those who add drugs to drink or food, and then abusively take goods . . .).
The statute goes on to explain that it is meant to cover robbers who
use threats or force as long as property was taken. It also covered
extortion under duress. The phrase about taking property and thereafter
being forceful was designed to cover cases in which someone, un-
beknownst to the victim, stole property but, upon being discovered, used
force or threats of force to resist arrest. The text as written is a little
misleading, because it does not state specifically that the penalties will
apply only if the robbers were unarmed. However, the sentences fol-
lowing the section just translated indicate that the penalties listed are for
robberies by unarmed criminals, saying that "if the robbers are armed,
even if they get no booty, they will be liable for the penalty of exile to
three thousand lid [approximately one thousand miles], and if they
cause injuries, they are to be beheaded." 86
Theft, that is, the clandestine taking of others' goods, was a less
serious crime than robbery, as measured by the sentences involved. As
with other crimes against property, the sentences were determined in
part by the amount of goods taken. According to the early Sung Code, a
thief who did not take any valuables was punished by fifty blows of the
light rod (which in the Sung period would have been converted to a
penalty of ten blows of the heavy rod on the buttocks). By contrast, a

86 SHT 19.300.

90
 Crimes and criminals

robber who got no booty was liable for a penalty one degree heavier,
sixty blows of the heavy rod, which in Sung practice would have been
converted to thirteen blows on the buttocks. The sentence for thieves
was increased by one degree when the booty was valued at one length of
cloth, which in the late tenth century would have been worth less than
one string of cash. For each added length of cloth this sentence was
increased by one degree, in theory adding ten blows of the heavy rod.
Then when the total reached one hundred blows, the sentence shifted to
levels of penal servitude. Thus, a thief whose booty was worth five
lengths of cloth (roughly four to five strings of cash in value) would be
liable for one year of penal servitude (which in practice meant forced
labor at some relatively nearby prefecture). For each additional five
lengths, this would be increased by one degree, reaching a penalty of
exile with added labor for a theft with a value of fifty lengths of cloth. 87
(The basic difference in T'ang law between penal servitude and exile
was the freedom of those sentenced to penal servitude to return home
after completing their sentences, and the distances to which convicts
were sent. As will be described in Chapter 12, during the Sung dynasty
these sentences were in practice converted to penal labor in certain
military units.)
Though less serious as measured by the penalties attached to it, theft
was endemic in Sung society. Yuan Ts'ai advised his kinsmen to guard
against robbers and thieves by arming themselves, patroling their prop-
erty, being cautious in displaying their wealth, and keeping it securely
locked up. 88 The point of his advice is reinforced by sayings attributed to
the criminals of the age: "We do not fear your iron walls, only your evil-
tempered dogs and robust men," and "if silver and gold are in a great
chest with iron fastenings, then bandits will not be able to get to them."
Burglars had their own sayings: "When entering a house at night there
are three things to be afraid of, old people, children, and nursing
dogs." 89
Cases of theft with special circumstances were occasionally reported in
detail, illuminating some facets of Sung crime and law enforcement:
At one time there was a Buddhist monk who passed through a village late in the
day. He sought a place to stay with a family, but the head of the household
would not allow it. The monk then asked to be allowed to sleep outside the gate
in the carriage stall and was permitted to do so. That night a thief entered the
house. From the top of the wall he helped a woman up, and they went away
87 SHT 19.302.
88 Patricia Ebrey, Family and Property in Sung China: Yuan Ts'ai's Precepts for Social Life
(Princeton, NJ: Princeton University Press, 1984), pp. 278-81.
89 T'ai-ping lao-jen, Hsien chung chin (Hsiieh-hai lei-pien ed.), p. 73.

91
 Law and order in Sung China

carrying a sack. The monk, who had not been able to sleep, saw this. He thought
to himself, "I was not admitted by the master of this house, although I pleaded
with him that I be allowed to stay the night. Now in the night the master of the
house has lost his wife and his valuables. Tomorrow he will surely seize me and
later turn me over to the district." The monk then fled. Not daring to go by the
regular routes, he walked through the wilds. Suddenly he stumbled into a well.
The body of the wife, who had already been killed, was in the well. The next day
the head of the household searched out the missing monk and found him with his
wife's body in the well. He seized the monk and sent him to the district to be
questioned underflogging.The monk falsely accused himself, saying that he had
seduced the wife and fled with her. Fearing that he would be captured, he had
killed the wife and thrown her body into the well but had lost his footing and
also fallen into the well. The booty had been abandoned beside the well, and he
had no idea who had taken it.
In this case a clever official on the prefectural level who was reviewing
the case was eventually able to learn the truth and in the end captured
the youth who had actually committed the robbery and murder. 90
The statutes just described indicate the general character of tradi-
tional Chinese laws against theft and robbery. Other statutes in the
T'ang Code of 737 and the early Sung Code, Sung hsing t'ung, spell out
special situations, for example, setting penalties for those who assault
others without initially having any intent to rob but who then do steal
goods, or those who having left goods behind when going away from a
place, later return, and must dispute the ownership of the valuables.
Later Sung additions to these rules generally take one of two forms.
Either they created special rules for limited areas or for specific sorts of
crimes, or they changed the amounts of booty needed to determine levels
of penalty.
The general trend of Sung changes in these laws was toward lessening
the severity of the laws inherited from the Five Dynasties, but not to the
degree of lenience characteristic of the T'ang period. The official Fan
Tsu-yii observed:
Under the Later Chou ruler T'ai-tsu [r. 951-54], a thief who stole booty worth
three lengths of cloth would be executed. When the Sung founder took over, he
made the laws on robbery and theft more lenient. A succession of benevolent
sages [i.e., the succeeding Sung rulers] progressively reduced this. Therefore
thieves were not executed. However, at present in the compiled edicts the
penalties for robbers are still three times as heavy as the penalties laid down
in the [T'ang] statutes. We inherited our system of penalties from the Five
Dynasties and have reduced the harshness of that era, and yet our "lenience" is
still like this.91

90 HCP 65-2ia-22a. 91 Fan Tsu-yii, Fan t'ai-shih chi, 22.1 ia.

92
 Crimes and criminals

The penalties were reduced primarily by increasing the amounts of

booty that thieves or robbers had to take in order to make them liable for
a particular punishment. Not all officials were satisfied with the system
of depending principally on the amounts of booty taken to determine the
penalties. There was concern that penalties so determined did not place
sufficient weight on the circumstances of the crime, whether or not, for
example, the criminals had been abusive to their victims. The late
Northern Sung statesman Tseng Pu (1035-1107) memorialized to
complain about the system then in use:

The circumstances of a robbery may be serious or not serious. The booty may be
large or small. At present we just use the amount of the booty to determine the
penalty. Thus, if a poor family is pillaged, even though the circumstances are
serious, because the amount of booty taken was small the robbers avoid major
penalties, whereas someone who pillages a rich family will be executed because
of the large amount of the booty, even though the circumstances are not serious.
Thus whether robbers live or die depends on whether the family robbed was rich
or poor.

When Tseng Pu became chief minister, his ideas were, for a time, put
into practice, but they were reversed after he fell from power. 92
It is noteworthy that robbery differed from most other crimes in that
no distinctions were drawn among those who participated. Ordinarily in
Chinese law the principal in a crime was punished the most heavily. All
those who were accomplices, which included those who harbored fugitives
as well as secondary actors in the major crime, were punished less
severely. The rule for robbery, however, was that all those who par-
ticipated were to be punished equally, and each was to be punished as if
he alone had stolen the whole amount taken. Only those who did not
participate but who did receive a share of the proceeds were to be
treated according to the rules for principals and accomplices. 93
If we look at what the officials recommended, in addition to the
constant litany about the need for appropriate rewards and punish-
ments, we find that they recognized the endemic nature of robbery. Wu
Ching, an official active in the later half of the twelfth century, who had
dealt with bandits with conspicuous success while serving as a sheriff
and then as a district magistrate, advised his contemporaries that

the world's having bandits is like a house having rats or a man's being ill.
Houses will always have rats; you keep them under control with cats. Men will
always have illnesses; you use medicine to help them get better. If you raise a cat

92 SS 1994978-79. 93 TLSI 3.76; SHT 20.11a.

93
 Law and order in Sung China

and it will not catch rats, then the rats will dance in the morning. If you take
medicines but not soon enough, then the disease will become more serious. 94

This simile of bandits as rats is common among Sung commentators.

Ch'in Kuan (1048-noo) suggests in an essay different responses to the
foreign enemy and internal bandits. The foreigners are tigers who must
be destroyed with traps and weapons. The bandits are like rats who
should be surrounded and smoked out. The bandits are easy to pacify at
the same time that they are difficult to get rid of entirely. The state
should respond by cutting off their sources of support, dispersing their
gatherings, and striking quickly at the beginning in order to break their
spirit. 95

Economic crimes
The great upsurge of the economy in the late T'ang, Five Dynasties, and
the Sung periods was accompanied by unprecedented concern by the
state authorities with economic crimes. In earlier times there had been
laws against counterfeiting, the illicit production of state monopoly
goods, and the like. However, during this era of rapid economic growth
the authorities seem to have been especially worried about enforcing
such laws. During the Sung, the government was concerned about
counterfeiting but even more so about illicit dealings in tea and salt.
The tea and salt monopolies produced a substantial amount of revenue
for the state. Because in this era the state tried to maintain high prices
for these restricted goods, men willing to risk the penalties continued
to manufacture and distribute them. In both the Northern and the
Southern Sung, organized groups of men engaged in this illicit traffic. At
times the authorities report that the criminals forced local people to buy
their products, but for the most part the officials recognized that these
illegal activities were provoked by the high prices of official goods. 96 The
penalties for breaking these laws could be very severe. Before 979, people
who bought a single pound {chin) of illegal tea were liable for one
hundred blows of the heavy rod. For twenty or more pounds - a figure
presumably set to distinguish dealers from users - the penalty was
execution. (An order of that year reduced the penalties somewhat.) 97

94 Wu Ching, Chu-chou chi, 2.2a.

95 Ch'in Kuan, Huai-hai chi, i7.2afT.
96 There is a huge amount of material on these kinds of crimes. A proper study of them
would require of book by itself, but see, for example, the kinds of reports found in SHY,
ping n . 2 6 a - b ; SHY, shih-huo 25.22a; HNYL 3407.
97 HCP 20.20b.

94
 Crimes and criminals

Arson
Arson as a crime was closely associated in the minds of Sung law-
enforcement officials with robbery and looting. The early Sung Code
even contains a specific rule designed to cover those who "deliberately
set fire to people's houses and buildings or to collections of goods in
order to commit robbery." When such people were to be punished under
the laws on robbery, the amount used to determine their sentences
included not only the value of what they had taken but also the value of
the losses suffered by the owner as a result of the fire.98 One magistrate
described in the collection of case solutions called the Che-yu kuei-chien
was determined to register all the juvenile delinquents in his jurisdiction
in mutual security groups, in order to prevent them from committing
arson with looting." As we might expect, arson seems to have been
especially a problem in urban areas with their dense populations and
crowded housing conditions. The urban soldiers used in fire fighting
were themselves notorious for looting burning buildings while sup-
posedly engaged in fighting the fires. The authorities wanted people to
start fighting fires as soon as they arrived on the scene, yet they hesitated
to let the soldiers begin before their superiors arrived, for fear of looting.
In the end the threat of uncontrolled fires forced them to authorize
immediate action, without waiting for the officers to arrive. 100
In one case a commoner from the capital city who had attempted to
commit arson was stopped by his wife. In a later incident, when he
ordered her to rob a neighbor and she refused, he threatened her with a
knife. She was so terrified that she reported him to the authorities. The
case had a very traditional Chinese ending. Under the statutes any wife
who accused her husband of a crime was liable for execution. The
prefectural authorities who were charged with settling the case were very
disturbed by what they saw as a gross miscarriage of justice, because the
wife, even with mitigation, would be liable for exile, but the husband,
the real criminal, would be released unpunished. They asked for an
imperial ruling. The emperor ordered that the husband be beaten, tat-
tooed, and sent to Kuang-nan for penal registration. The wife was
freed.101

98 SHT 19.304-5.
99 CYKC n o . There are a number of reports on the problem of looting following arson.
See SHY, hsing-fa 4.5a (1010); SHY, ping 3.2b (1020); SHY, ping 2.i7a-b.
100 SHY, ping 3-ia-b, 4b; HCP n o . i b .
101 HCP 106.15a.

95
 Law and order in Sung China

Crimes of violence — assaults and batteries

Minor crimes of violence against the person are said to have been
common. In a memorial about banditry in two prefectures in Chekiang
the Southern Sung official Wang Chih (1127-89) remarked that "never
a day goes by without there being cases of beating; never a year goes by
without there being cases of assault with a knife or saw." 102 However,
such crimes are little noted in the sources, compared with some other
crimes. Since for most such assaults and batteries the penalties were
relatively light, they could be assessed and inflicted by the district
magistrates without being reviewed at higher levels. Thus they would
tend to come less often to the attention of important officials, and so
records of them would be less likely to be preserved in the sources.
Although most assaults and batteries were relatively minor crimes, as
measured by the penalties inflicted, they were of greater concern when
they involved organized groups. In late imperial China the state was
plagued by organized feuding and violence. The origins of violence have
been traced back to the Sung dynasty, particularly the Southern Sung.
Such feuding seems to have originated at least partly because the state
had given local people in the pao-chia militia system training in weapons.
The state also persecuted local religious organizations, some of which
used weapons as part of their ritual paraphernalia. As a result of these
two developments, there were in the Sung countryside both large num-
bers of ordinary people who had had some training in arms and the
experience of being organized for combat, and the kinds of stimuli that
might cause them to band together to cause trouble. Such feuding by
groups could continue for extended periods, especially if it reflected
disputes among locally influential kin groups who used sizable gangs,
including pao-chia militia groups, to carry on their fights.103

Situational character of crimes

When assaults and batteries were committed by individuals or casual
groups, they were relatively minor matters. Although the penalties were
usually light, the category itself clearly displays some basic characteris-
tics of traditional Chinese views of crime and criminality. It is appro-
priate to call these traditional Chinese, and not merely Sung, views, as
the rules used by the Sung were adopted unchanged from the T'ang,

102 Wang Chih, Hsiieh-shan chi, 3.24.

103 See Sogabe Shizuo, "Sodai no ketsu-kan" (Concerning feuds in the Sung Dynasty), in
Ishida Hakushi Koki Kinen Jigyokai, ed., Ishida Hakushi shoju kinen Tqyoshi ronso
(Tokyo: Ishida Hakushi Koki Kinen Jigyokai, 1965), pp. 311 ff.

96
 Crimes and criminals

who in turn had inherited them from earlier dynasties. Similar rules
continued in use in later times.
The rules for punishing assaults and batteries reveal the Chinese
concern about the situational character of crime. That is, the serious-
ness of a crime was measured only in part by the degree of violence done
or the size of the booty taken. Equally important were the social statuses
of the perpetrator and victim, and their relationship. Generally, crimes
committed by status inferiors against superiors were punished more
heavily than were crimes against equals, and crimes of status superiors
against status inferiors were punished more lightly.
The most basic rule, found in both the T'ang Code and the Sung Code
of 962, states that the penalty for battery (by an ordinary person against
another ordinary and unrelated person) was forty blows of the light rod.
This penalty would be increased if injuries had resulted or if weapons
other than fists had been used. The statute states:

If there are wounds or if weapons other than the fists are used, the penalty is
sixty blows of the heavy rod. If the wounds reach to the loss of a square inch of
hair or more, the penalty is eighty blows of the heavy rod. If blood is drawn from
ear or eye or there are internal injuries with the spitting of blood, increase this by
two degrees [to one hundred blows of the heavy rod].... When those who
commit battery cause a broken tooth or the loss of an ear or nose or of some
vision in one eye, or if toes or fingers are lost, or if bones are broken, or if boiling
water or fire is used to injure someone, [the sentence is] one year of penal
servitude.104

It should be noted that with the exception of the last provision, which
calls for penal servitude, all the other cases, which involved penalties of
one hundred blows of the heavy rod or less, could in Sung times have
been adjudicated on the district level without being reviewed by higher
authorities.
As this passage indicates, in addition to differentiation according to
the degree of injury, the sentences also varied depending on the weapons
used. The statute cited also states that when weapons other than fists
were used, even if no injury resulted, the penalty was sixty blows of the
heavy rod, not forty blows of the light rod. If a military weapon was used
to attack someone, without any injury having been inflicted, the penalty
was one hundred blows of the heavy rod. Similarly, if boiling water or
fire was used to commit battery, the sentence was one year of penal
servitude. Naturally these two factors could be combined, so that if a

104 SHT 21.324*1:

97
 Law and order in Sung China

person actually wounded someone with a knife, the penalty would be

two years of penal servitude. 105
Equally important in sentencing, and indeed having logical precedence,
was the question of the parties' status and their relationship. Status
considerations, which were based on age, sex, and membership in cer-
tain social groups (officials and their relatives, relatives of the imperial
family, certain religious orders, ordinary people, unfree groups, and so
on) helped determine whether a crime had been committed and the
penalties to be inflicted. The relationship between the parties was even
more important. Certain actions, for example, the striking of one person
by another, might be an extremely serious crime, a minor crime, or no
crime at all, depending solely on the relationship between the parties. A
passage from the T'ang law, copied into the Sung Code (even though
some of the categories of persons named no longer existed in Sung times)
serves as an example of these principles:
When a pu-ch'ii [a type of unfree personal retainer] assaults or injures a com-
moner, his penalty is to be one degree heavier than that for an ordinary person.
For slaves in similar circumstances, increase it by another degree. If slaves fight
with commoners [literally, "good people"] and break a bone or injure an eye,
they are to be executed by strangulation.106
Thus a commoner injuring another commoner in these ways would
suffer one year of penal servitude, whereas a slave injuring a commoner
would be executed.
The reverse also held. This statute goes on to say that when an
ordinary commoner injured another person's pu-ch'u, the penalty was one
degree less than for having injured a commoner. Masters also were
legally entitled to beat their unfree dependents and would be punished
lightly even if the dependents died. Those masters who beat pu-ch'u to
death without having intended to kill them were liable only for one year
of penal servitude (the ordinary penalty for the crime of homicide in an

105 SHT 21.326; TLSI 3.81. In a puzzling passage recorded in the Sung History, the late
Northern Sung official Tseng Pu argued for an increase in the severity of penalties for
certain sorts of assaults, stating, "There is certainly a difference between using the
hands or feet or strike someone and using a sword, boiling liquid, or fire; yet all of
these are called 'injury'." He seems to be implying that people were being punished
more or less equally no matter what they used in their assaults, which would have
been flatly contrary to the rules in the statutes. However, he also was concerned
because some such cases were being sent up for central review and some were not,
perhaps fearing that the culprits who used weapons such as swords or fire, but whose
cases had been memorialized, would be punished less heavily than would those whose
cases were not sent up, even though they had only used their fists (see SS 199.4978—
79).
106 SHT 22.341.
98
 Crimes and criminals
107
affray was strangulation). If they had deliberately killed the pu-ch'u,
the penalty would be increased by one degree, to one and a half years of
penal servitude. If the death resulted from the pu-ch'u's having been
beaten as punishment for their wrongs or if the death was accidental, the
masters were not to be tried. 108
This should be contrasted with the treatment ofpu-ch'u or slaves who
accidentally killed their masters. They were strangled. They also were
strangled if they committed battery against their master's mourning-
degree relatives or his grandparents-in-law or parents-in-law. 109 The
same status differences protected the elite. Assaults and batteries
against imperial relatives or officials were punished more heavily than
were ordinary assaults and batteries if they were committed by status
inferiors.110
Relationship, especially the kinship relationships defined according to
the system of five degrees of mourning, also affected the character of a
crime. In the usual case, when a kinship relationship also involved
relations of inferiority and superiority, the principles just discussed
applied. Husbands who assaulted and injured their wives were liable for
penalties of twenty blows of the light rod, that is, penalties two degrees
less than those called for in cases of ordinary assault. By contrast, wives
who merely struck their husbands were liable for one year of penal
servitude, and if serious injury resulted, they were to be punished three
degrees more heavily than the penalty for having injured ordinary un-
related persons. 111 The state also mandated divorce for such crimes. The
closeness of the relationship, usually as measured by the system of
mourning gradations, also affected the weight of penalties, with a closer
relationship generally having a greater impact. The T'ang and Sung
codes contain many more rules in this same vein, spelling out the
degrees of liability associated with various statuses and relationships, but
the principles behind them are identical with those just described. 112
Finally, the circumstances were supposed to affect the penalty.
Chinese laws from the Sung very often say that whether the circum-
stances were "heavy" or "light" should affect the judgment. In part
what was to be considered heavy or light was set down in the Sung
laws themselves. Unfortunately, almost all the rules that specified such
circumstances are no longer extant.
107 SHT 21.328.
108 SHT 22.343.
109 SHT 22.343.
110 SHT 2i.334flT., 22.339, a n d passim.
111 SHT 22.346.
112 See, for example, the statute from SHT 22.351, which specifies penalties for such
crimes as assaulting an elder brother's wife.

99
 Law and order in Sung China

In addition to such formal rules, it appears that Sung magistrates

were supposed to take into account the particular facts of a case. One
official wrote in the late twelfth century to complain about the failure
of his colleagues to deal adequately with cases in which there were
mitigating circumstances that made the stipulated penalty excessively
harsh. He used three hypothetical examples: the case of a man who had
killed someone in a fight when prior to the fight the man killed had
invaded and injured the killer's fields or crops; a case in which the man
killed was a debtor who, prior to the fight, had vilified the man accused;
and a case in which the man killed had attacked the elders or parents of
the man accused. All such things could cause the case to be labeled light
circumstances.113

Major crimes of violence — homicide

Many of the same principles applied in the treatment of more serious
crimes of violence. In homicides as in assaults, attacks by social inferiors
against their superiors were more severely punished than were attacks
between equals in status, and those by superiors on inferiors were treated
more lightly, though a husband who killed his wife was treated as a case
of ordinary homicide.
Homicides were also differentiated according to the killer's intent.
T'ang and Sung law distinguished between homicide committed in an
affray (tou-sha) (in which there had been no intent to kill), deliberate
homicide (ku-sha) (which did not involve premeditation but in which the
killer at the time of his act recognized what he or she was doing),
premeditated homicide (mou-sha), homicide occurring during horseplay
(hsi-sha),114 and homicide through accident or mischance (wu-sha).
The most serious crime was premeditated homicide, which carried a
statutory penalty of beheading. The next most serious was deliberate
homicide, which also carried a penalty of beheading. Then, in order,
homicide during an affray earned a penalty of strangulation; homicide
in horseplay (hsi-sha) had a penalty of exile; and homicide through
accident or mischance carried a penalty of one and a half years of penal
servitude, though in this last situation people could usually commute the
penalty. 115
113 TFSL503.
114 For homicide in horseplay, see Benjamin E. Wallacher, "The Chinese Offence of
Homicide Through Horseplay," Han-hsiieh yen-chiu 1 (June 1983): 259-316. On the
system of commutation, see Chapter 14.
115 SHT 17.275; 23.360; 21.328. Of course, I am giving here the simplest cases; the actual
law was more complex. The penalty for homicide in an affray was ordinarily
strangulation but if a knife were used, the penalty was beheading. The laws for other
crimes are similarly complex.

100
 Crimes and criminals

As a more serious crime, homicide, though probably far less common

than simple assault and battery, is described quite often and in some
detail in Sung sources. The hidden potential for violence which ran
under the surface of Chinese life sometimes exploded for trivial reasons
and ended in death.
Seng Fa-tuan was outraged at the price that a fisherman demanded for his fish.
Seng ordered Seng Shou-kung to kill his estate's dog and then went to the
officials and falsely accused the fisherman of being a robber (who had killed the
dog) and fled the district. Runners were dispatched. They seized the fisherman
and his father. They bound the two men and sent them to the estate, where Seng
Shou-kung beat them to death. He also bribed the district records clerks to
gather the village officers to go to arrest the two younger brothers of the
fisherman and had them killed. He also wounded the mother of the fisherman
with a knife. Then he informed the sheriff that he had killed some bandits in
trying to capture them.

In this grim case the investigation by the sheriff and the magis-
trate eventually uncovered the actual facts, and the criminals were
executed. 116
Then as now, murders often followed in the wake of sexual abuses.
Adultery followed by killing is a recurrent theme:
A commoner of Hsiian Prefecture named Yeh Yuan strangled his wife and his
elder brother with whom he had lived, because the elder brother had previously
committed adultery with Yeh Yiian's wife. He also killed his elder brother's son
and forced his own father and the elder brother's wife to agree not to inform the
authorities.

In this case the neighborhood mutual security group found out about the
killings and informed the prefectural authorities. The case became
something of a cause celebre. The legal problems were difficult to dis-
entangle, because the reputed adulterer and the wife were dead. It was
therefore impossible to determine whether the elder brother had raped
the wife, which would certainly have justified some mitigation in the
sentence of the murderer-husband, or whether the adultery was con-
sensual. In the end the emperor himself intervened and ruled that the
man should be executed. It was no longer possible to determine whether
the wife had been raped. Therefore the case had to turn wholly on the
behavior of the defendant. If he had killed only the supposed adulterer,
there might have been grounds for mitigation, but he had also killed
the adulterer's son and coerced his own father. Therefore, even if the

116 HCP87.i7a-b.
101
 Law and order in Sung China

adultery could have been proved, it would have been insufficient to

mitigate the sentence. 117
Because homicides called for more serious penalties, they inevitably
entailed scrutiny by higher officials. They sometimes also raised serious
questions of formal legality and legal process. In a case from Shou
Prefecture in Huai-nan West Circuit, a man had killed his wife's parents
and his elder and younger brothers. The prefectural authorities felt that
it was a case falling under the category of "depravity" (pu-tao). As such
it required the joint adjudication and punishment of the man's wife
and his son, along with the murderer. The circuit judicial authorities
reversed this decision. Their reasons are thoroughly lawyerly. They
said that under Sung law if a man beat his wife's parents, the state
automatically mandated that the couple be divorced under the system
called righteous separation. If beating automatically created such a separa-
tion, then murder certainly would do so also. Therefore, because the wife
had, through the act of her husband, been divorced from him, she
was no longer his wife under the law and so could not be jointly
adjudicated. 118
Adultery could lay bare the kinds of tensions that divided some
households. In the late tenth century an adulterous widow, worrying
about people finding out, became ill. Fearful that her son would report
her, she accused the son of having put poison in her food. When the
incident was investigated, the widow bribed the investigators to arrest
her son. The son was held in jail for months. Eventually, the emperor
issued an order of disposition. However, the accused continued to protest
his innocence and demanded that the corrupt functionaries be punished.
Finally, many months later, the truth was uncovered. 119
With all their various circumstances, homicide cases often raised
difficult legal questions. Settling them was important not only to the
accused but also to the officials involved. Officials could be sanctioned if
they made serious errors in judicial judgments. A case that involved a
mixture of adultery and murder reflects this aspect of the legal system.
In the late eleventh century an adulterer plotted with her lover to
murder her husband. After the plotting had been completed, the husband
unexpectedly returned, and the lover killed him. Most of the legal
officials in the capital who dealt with the case felt that because the wife
had not herself participated in the actual killing, she should be treated as
an accomplice. This would have spared her execution. However, one

117 Hung Mai, Jung-chai sui-pi (Shanghai: Shanghai shu-tien, 1984), 5th collect., p. 147;
SHY, hsing-fa 5.10b; HCP 303.17a.
118 Chiang Hsiao-yii, Huang-ch'ao lei-yuan (Beijing: Chung-wen ch'u-pan she, 1981), 22.1a.
119 SHY, hsing-fa 5.2a-b.

102
 Crimes and criminals

official in the Ministry of Justice argued that because of the prior

plotting, the woman should be executed.
This same official, apparently at approximately the same time, also
contradicted his colleagues in a case involving another homicide. A
father had brought some grain to feed to his wife, who was ill. When
their son then ate the grain, the father beat him to death. The legal
authorities felt that because the son had in effect stolen the grain,
beating him to death fell under the category "miscellaneous crimes
calling for the death penalty," which meant that the criminal would be
eligible for, and spared by, an amnesty that had been issued between the
time the crime had been committed and the time of the trial verdict. The
hard-line official again disagreed. In the end the official who had dis-
agreed with his colleagues was fined, apparently for being inappropriately
harsh in his judgments, and the more lenient penalties were inflicted. 120

Crimes of violence - rape

Rape almost certainly was regularly underreported in Sung China. The
work of Vivian Ng on rape in Ch'ing China discusses both the practical
and the ideological reasons why the authorities made rape a difficult
crime to prosecute, and therefore a crime even less likely to be brought
to light in China than elsewhere.121 As Gary LaFree notes in his book
Rape and Criminal Justice, because rape cases rarely include eyewitnesses,
the processing of the cases "is often reduced to a direct confrontation
between accuser and accused." Such a
two party confrontation in which each party insists on a different reconstruction
of an event is likely to be highly ambiguous. Without eyewitnesses processing
decisions may depend less on an assessment of whether a rape has occurred than
on a perception of whether the victim and the assailant are the kind ofpeople who
could have been involved.122
In Chinese law the burden of proof that the victim resisted the advances
of her attacker rested on the woman. When the woman could not prove
her resistance, the crime would be classified as adultery by mutual
consent. Under these conditions it is hardly remarkable that rape cases
were underreported. Indeed, many cases seem to have come to the
attention of the authorities only because they were accompanied by other
serious batteries that could not be concealed.

120 WHTK 170.1476; SS 201.5010-11.

121 Vivian Ng, "Ideology and Sexuality: Rape Laws in Ch'ing China," Journal of Asian
Studies 46 (1987): 57-70.
122 LaFree, Rape and Criminal Justice, pp. 27-28 (italics in the original).

103
 Law and order in Sung China

In Sung law, rape was treated as adultery with force, and the crime is
listed at the end of the statute on adultery:
For committing adultery the penalty is one and a half years of penal servitude. If
the woman is married, increase the penalty to two years. For people of unfree
status who commit adultery with ordinary commoners, the penalties should be
increased by one degree. For adultery with an official slave or private slave, the
penalty is ninety blows of the heavy rod. For committing adultery with another
man's pu-ch'u\ wife or the wives of men belonging to two other unfree categories
[tsa-hu or kuan-hu], the penalty is one hundred blows of the heavy rod. For rape,
increase these penalties by one degree. If injury occurs, use the penalties for
battery and injury increased by one degree.123

Social crimes - gambling

The state was concerned about gambling because the authorities be-
lieved that it contributed to other sorts of crimes. Gambling often in-
volved drinking too much and exchanging insults, with fights breaking
out among the gamblers.124 Wang Yung speaks of "corrupt petty
men who have no sense of responsibility and who are unrestrainedly
evil." They may be gamblers; they may illegally butcher oxen or make
counterfeit money. "If they lose money through gambling and have no
means to pay their debts, then they will become thieves or, if they are
numerous, bandits and arsonists."125 His comment was echoed by an
official who commented in 1087 that "the rustic, lazy people of the
prefectures and districts gamble and, when they have lost, become
bandits."126
In one well-known and illustrative incident, three soldiers who were
serving as night police in the capital pawned their clothing in order to
gamble. When they lost their stake, they seized a passerby, killed him,
threw his body into the canal, and pawned his clothing to raise money
so that they could redeem their earlier pledge.127 The collection of
thirteenth-century case decisions entitled the Ch'ing-ming chi also pro-
vides us with a colorful description of a gambling house. A man of Ch'ii
Prefecture in Liang-che Circuit who was married to a "singing girl"
(prostitute) installed her on the first floor of a building that fronted
on the marketplace. There she presided over a tea house. Upstairs he set
up a regular gambling operation. Men who were originally attracted to

123 TLSI 4.38 ff; SHT 26.421 ff.

124 Wang Chih, Hsiieh-shan chi, 3.24.
125 Wang Yung, Sung-ch'aoyen-i i-mou lu, Pai-ch'uan hsiian-hai ed., 2.5b.
126 HCP 398.12a-13b.
127 SHY, ping 3.1b.

104
 Crimes and criminals

the tea house and its mistress stayed on to gamble, usually at dice
games. Swindlers and sharps set about fleecing the customers. Large
sums were gambled and lost. One man lost 250 strings, a true fortune
when you remember that the state figured that the total cost of suppor-
ting a member of the postal system was about 30 strings a year. Others
also lost large sums before the authorities cracked down and arrested
the proprietor. 128 Little wonder that the authorities tried, but without
success, to stamp out gambling operations.

Locations of crimes - piracy

To speak of "crime in Sung China" is somewhat misleading. Sung China
was a diverse empire, and crime varied depending on the regions con-
cerned. Urban crime (and law enforcement) are distinct and will be
covered in a separate chapter, but the patterns of crimes in other areas
were also sometimes quite distinctive. Piracy posed its own problems,
and as we noted, crimes in areas with substantial non-Han populations
also demanded some adjustments in law enforcement and legal patterns.
Piracy presented peculiar problems of law enforcement. A number of
Southern Sung officials wrote in detail about the methods of suppressing
piracy. Their suggestions indicate that policies like those used against
land-based bandits were used but modified for the problems of dealing
with an enemy who could strike quickly and with little warning. The use
of mutual security groups among the people,129 the elimination of the
underlying economic and political problems, government support for
local self-defense, proper training and equipping of government forces,
an emphasis on unified leadership and on the importance of cooperation
among state agencies - all these were elements in the regular pattern of
bandit suppression. Their importance was merely underscored by the
difficult problem of stopping sea-based criminal bands. 130 Even the
chain of command was also the same, from the local patroling inspector
to the prefecture, to the circuit, to the central authorities. 131 But in
addition to these general policies, the authorities also acceded to re-
quests by local officials that boats and ships be registered, as a control
device.132

128 Ming-kung shu-p'an ch'ing-ming chi, vol. 2, p. 530.

129 For the use of this device in areas threatened by piracy, see SHY, ping 2.43a, 46a.
130 For some of the sources, see, for example, essays in Ch'en Ch'un, Pei-ch'i ta-ch'iian chi,
45.3b ff; Cheng Hsing-i, Cheng-chung su tsou-i i-chi, SKCSCP ed., shang 17b; Ts'ai
Hsiang, Tuan-ming chi, 21.9a; Chen Te-hsiu, Hsi-shan hsien-sheng Chen wen-chung kung
wen-chi, SPTK ed., 8.5b, 17.20b; Wu Ch'ien, Hsu kuo-kung tsou-i, Pai-pu ts'ung-shu
chi-ch'eng ed., 4.1a.
131 Chen Te-hsiu, Hsi-shan hsien-sheng Chen wen-chung kung wen-chi, 8.5b-10b.
132 SHY, ping 2.46a, I2.i7a-b, 13.33b. See also Wu Ch'ien, Hsu kuo-kung tsou-i, 4.1a.

105
 Law and order in Sung China

Piracy also reveals another facet of Sung crime, its seasonality. In the
case of piracy along the south and southeast coast, the criminal season
corresponded to the period of fair winds that brought merchants from
southeast Asia to the ports of southern China. In other areas criminal
activity by gangs occurred during the agricultural off-season, suggesting
another aspect of peasant criminal culture. Presumably, the gang mem-
bers in such areas worked as farmers during the growing season, joining
together in the winter to rob. 133
Much of our information on piracy comes from the Southern Sung.
This is hardly surprising, considering the importance of water-borne
commerce during that period and the proportion of state revenues that
were derived from taxes on overseas trade. It is also not unlikely that
piracy actually increased as the numbers and values of water-borne
cargoes increased. Toward the end of the dynasty, the problem of piracy
became entangled with problems of national defense. The state had to
preserve its control over the sea lanes for fiscal reasons but understand-
ably had to limit its commitment of resources to the coasts so that it
could effectively combat the Mongol enemy.
The official Cheng Hsing-i (1126-99) summarized the special prob-
lems of suppressing pirates:
The methods useful against bandits on land are not useful against pirates. How
so? The sea is vast, who knows how many thousands and tens of thousands of li.A
Roving criminals take advantage of the winds to arrive unexpectedly, spy out the
unprepared, destroy the people's homes, and steal their food. The prefectures
bordering the sea are frequently raided. The Court dispatches troops to suppress
them, but these are men sent from the interior. How could they arrive in the
morning and go out in the evening? The troops man their positions, but the sea
raiders are like the winds and clouds. In the twinkling of an eye they have struck
and vanished. If you want to pursue them amid the vast waves, they have no
fixed place. How could they be effectively attacked?134
Like other sorts of bandits, pirates were frequently bought off by the
state. It even was common practice for the authorities to give honorary
ranks to major bandits when they surrendered. A report from 1135 notes
briefly:
The Fukien pirate Chu Tsung was awarded the rank of a Pao-i dignitary (rank
9A). His followers were given awards of official rank by degrees. During the first
month of this year Chu Tsung gathered a group of thirty ships on the sea, with
two hundred or so men, and entered the various districts of Kuang-tung to burn
and kill. After this the Court empowered the general of Fukien-Kuang-tung to

133 SHY, pine 13.37a.

134 Cheng Hsing-i, Cheng-chung su tsou-i i-chi shang. 17b.

106
 Crimes and criminals

devise methods for summoning him to surrender or for capturing him. At this
point Chu Tsung came in with his followers and surrendered.135
As this material suggests pirate gangs perhaps included somewhat
larger numbers of members on average than did other sorts of bandit
groups. Other sources confirm this possibility. 136
Most such pirate gangs probably operated along the seacoasts, but
piracy was an inland occupation also, especially on the larger lakes and
rivers. The authorities of Lu Prefecture in Huai-nan West Circuit asked
in 1175 that boats and families be registered there, to prevent piracy
under the guise of fishing.137
Penalties for piracy seem to have been determined by a separate set of
rules, the "precedents on piracy" (hai-tsei li).138 Men convicted of piracy
and sentenced like other robbers to penal registration were sometimes
registered in naval units or in riverside prefectures where their skills
would be useful. A report from the 1160s indicates that at this time
many were being sent to the "distant, evil prefectures" on the coast
of Kwangtung. The official who submitted the report was concerned
because, he said, at the seaside they often built boats and caused trouble.
He asked that they be sent to serve in naval units in the Liang-Huai
region of central China. His request was approved. 139 Another twelfth-
century document lists a set of prefectures to which certain pirates were
to be sent for registration. All of them bordered the Yangtze River. 140

How often were crimes committed?

True crime rates are notoriously difficult to establish. Even in con-
temporary states, reported crime rates are not completely trustworthy. 141
The crime situation in Sung China must remain problematic. Im-
pressionistic and indirect evidence, however, may allow us to comment
tentatively on changes in crime rates. Clearly there were periods during
which crimes, especially organized crimes like banditry and smuggling,
were more widespread. In some cases, the 1040s and the late 1120s, for
example, the rise of internal unrest and accompanying crime waves were

135 SHY, ping 13.18a.

136 Wang Chih-wang (d. 1170), reporting on seizing a pirate gang in T'ai Prefecture in
Liang-che Circuit, says that fifty-seven men were arrested. See Wang Chih-wang, Han
p'in chih, SKCSCP 1975, 7.24a-25a.
137 SHY, ping 2.46a.
138 SHY, hsing-fa 4.47a.
139 WHTK 168.1460.
140 Ibid.
141 See the discussion of crime rates and the problems associated with their interpretation
in Wilson and Herrnstein, Crime and Human Nature.

107
 Law and order in Sung China

associated with external military problems. More generally, I think, we

can argue that the long-term general increase during the Sung in the
number of local police officials and their supporting soldiers and con-
stables reflects a decrease in local security. Surely a state with only
limited resources would not invest them in police activities if it were not
necessary. More crime seems intuitively to be the most satisfying ex-
planation, though of course there are other possibilities, for example,
that the authorities felt it important that a higher degree of security be
maintained in the empire and thus raised the number of police personnel,
even though the amount of crime itself had not increased. No doubt
other possibilities can be imagined, but none seems as likely as the
simplest, that more crimes were being committed.
The general impression given by the sources is that the Northern Sung
period was somewhat less disturbed by crime than was the Southern
Sung, with the exception of the 1040s and the reign of Shen-tsung
(1067-85). In the first Northern Sung period of high crime rates, the
Sung state was involved in a very expensive foreign war, and in the
second the accumulated fiscal problems inherited from earlier reigns had
created an internal governmental crisis, to which the reforms of Wang
An-shih responded.
Because the sources are of uneven quantity and quality, it is possible
that some of the apparent rise in the rate of crime may reflect differences
in the amount of information that has come down to us. Thus we have
far more information on the reign of Shen-tsung than on any other
period in the Northern Sung and so should expect to find more reports of
criminal activity, and indeed of all kinds of activity. Still, the sheer
volume of complaints does suggest that the rate of crime rose in this
period.
The Southern Sung sources are much less full than those for the
earlier period. Given this disparity, the frequency with which Southern
Sung officials discuss crime and its suppression suggests that its incidence
had in fact increased, a suggestion most strongly supported by the larger
numbers of police officials appointed. Such an increase seems probable
also if we generalize from the findings of modern criminologists. It seems
clear from their work that the typical criminal in all societies that have
been subjected to study is not only young and male but also urban. 142 If
this trait — which seems to hold universally, judging from recent evidence
— was also true of traditional societies, then a rise in crime rates would
be predictable from the increasing urbanization of Sung times. In the
preceding chapter we described the demographic transformation from

142 See ibid., p. 26.

108
 Crimes and criminals

the T'ang to the Sung periods, marked by an enormous increase in the

number of small market towns and by the breakdown of internal controls
within the larger cities. During the Sung period itself the number of
larger cities grew, and therefore a much larger proportion of the popu-
lation lived in urbanized areas. This implies that if and when we are able
to learn the origins of an appropriate sample of Sung criminals, we will
find that Sung criminals came in disproportionate numbers from cities
and market towns. There are a variety of reasons that an increase in
urbanization may (but does not always) lead to an increase in crime.
One is simply the density of criminal opportunities:
An individual encounters opportunities to commit crimes at varying rates as he
goes about his daily business. Crime rates may go up or down because of changes
in the density of such opportunities or their accessibility to criminal penetration
or both, all without any change in the dispositions of the individual.143

Regional variations in crime rates

Some kinds of areas - borders, coasts, boundary regions, marshes, moun-
tains - seem to have been especially attractive to organized criminals.
However, even within those parts of rural China where the topography
was not unusual, there seem to have been major regional differences in
crime rates. Some areas were notorious as homes of unruly populations
and havens for criminals. Hsiu-ning District in Chiang-nan East Circuit
was, in the words of Wu Ching, "a violent district," 144 and Tu Fan
(doctorate of letters degree 1208) said that in the I-ch'eng and Ning-kuo
districts of Ching-hsi South Circuit, "pillaging and killing by bandits are
widespread." 145 Crime rates no doubt varied from place to place even in
relatively small regions. In 1157 Hsiieh Chi-hsiian describes serious
banditry problems in Hsing-kuo Prefecture in Chiang-nan West Circuit.
Two decades later Chu Hsi, writing about Tu-ch'ang District in Chiang-
nan East Circuit, an area only about one hundred miles from Hsing-kuo
Prefecture, derided a request for establishing another police fort. He said
that the area had been peaceful for the last century. There had indeed
been one gang of robbers, in all three men, who came along the river but
were too afraid to land on shore and who had promptly been captured
by the local sheriff.146 Certain districts were thus seen as basically law-
abiding.

143 Ibid., p. 410.

144 Wu Ching, Chu-chou chi, u . 8 a - i o a .
145 Tu Fan, Ching-hsien chi, SKCSCP 1971, 19b-20b.
146 Hsuch Chi-hsuan, Lang-yu chi, SKCSCP 1971, 20.9b-11 a; Chu Hsi, Hui-an hsien-sheng
Chu Wen-hung wen-chi, SPTK ed., 20. nb-14a.
109
 Law and order in Sung China

This regional variation is probably related to one important exception

to the characterization of the typical Sung criminal as an unattached,
young adult male. In certain regions of China there existed what we
might call a culture of peasant crime. This seems to have been especially
true in areas where the geography made escape from the authorities easy
and life itself hard - mountainous regions or swampy places, for example.
People in these regions exhibited behaviors that in the European tradition
are sometimes thought of as characteristic of the peasants of some parts
of southern Europe. In times of plenty most people would be law-abiding,
but they were always ready when opportunity offered, or when circum-
stances demanded, to turn to crime as a way of life.147 It is to this
penchant that Liu Chih (1030-97) was referring when he said that "the
people nourish the hearts of bandits," so that evil men could recruit from
among ordinary villagers as well as from among military deserters. 148
This problem was beyond the power of the state to settle definitively.
The authorities could try only to reduce it somewhat by indirect means,
for example, by outlawing the ownership of certain sorts of weapons. 149
Furthermore, in Sung China, as in many other times and places,
criminals tended to cluster along jurisdictional boundaries. This may
have been a geographical phenomenon. Such boundaries often ran along
watersheds, that is, in the highest and least accessible locations, which
also often were the most heavily forested and mountainous terrains in
the region.150 An official writing in 1117 characterizes a bandit strong-
hold in typical fashion as being located in "one of the most inaccessible
and mountainous places in Ho-tung Circuit." 151 In other parts of China,
jurisdictional boundaries ran through marshlands, also excellent hideouts
for gangs. 152 This pattern was also political, because it allowed the
criminals, if pressed by the law-enforcement apparatus of one jurisdic-
tion, to slip across the border into another jurisdiction. Then, as now,
cooperation among the law-enforcement officials of neighboring juris-
dictions was mandated in theory but often ignored in practice. 153 In the
early twelfth century an official remarked:

147 For some comments on this marginality of people in certain regions of China, see John
W. Haeger, "Between North and South: The Lake Rebellion in Hunan, 1130-1135,"
Journal of Asian Studies 28 (May 1969): 469-84.
148 HCP 365.3a-4b.
149 HCP 117.19a.
150 See, for example, the comments of an official writing in 1113 about problems in
Yung-hsing Chun Circuit: SHY, hsing-fa 4.34b-35a.
151 SHY, hsing-fa 4.36a-b.
152 SHY, hsing-fa 4.36b.
153 See Wu Ching, Chu-chou chi, n . 8 a - i o a , on the role of boundaries and borders. See
also, for example, SHY, ping 13.47b, on the problems of cooperation among
jurisdictions. In the American West, at the place where the states of Colorado,

110
 Crimes and criminals

In the area of the western mountains along the northern banks of the river, the
forests are heavy. There are many fugitives hiding out there. If there is any
weakening of control, then many of them will turn to banditry. Generally, when
this happens, because the boundaries are unclear, the prefectures mutually avoid
them, neglect to exercise control, and stay away from the distant places.
He recommended that in hard-to-control areas such as forests, moun-
tains, swamps, areas with lakes, places where horses were raised, and
places where there were abandoned buildings that might serve as head-
quarters for bandits, the circuit intendants should cooperate in deter-
mining the boundaries and should have maps drawn of especially difficult
areas. The Court approved his proposal. 154 However, we may suspect
that the impact of policies such as these would be minimal. The key
problem was probably that an official's career was harmed only by
crime committed within his jurisdiction. If he could succeed in driving
criminals over the border into his colleagues' territories, he might benefit
from his apparent success in preserving law and order. He thus would
have little to gain by cooperating with his colleagues in neighboring
jurisdictions, as their joint efforts might fail, and in the worst case the
criminals might end up back in his jurisdiction.
Criminals might cluster in regions difficult to reach. They also might
be found in disproportionately large numbers in places where the oppor-
tunities for crime were greatest, such as the large cities. Finally, in any
area they tended to seek out certain sorts of places as bases. The early
twelfth-century book of advice to local officials, entitled the Tso-i tzu-chen,
recommends that local officials have abandoned kilns blocked up, so that
criminals could not live in them, and a decree from the early eleventh
century instructs officials concerning certain unregistered temples, which
"have no abbot to manage them and are located in hidden, solitary, and
mountainous areas. Those that serve as the hideouts of bandits should
all be destroyed." The passage continues, "At this time in Fang Pre-
fecture [in Yung-hsing chun Circuit] there was a bandit gang that had
gathered at a Buddhist center in the mountains." 155

Number of criminals
Occasionally the sources give us figures on the number of criminals. 156
These figures, widely scattered and of differing degrees of reliability,
Wyoming, and Nebraska come together, nineteenth-century horse thieves and cattle
rustlers usually had three hideouts, one in each state, and, when pressed hard by the
authorities in one jurisdiction, would simply cross the line into another.
154 SHY, ping I2.i6b-iya.
155 SHY, ping u . 8 b .
156 The figures we have must be used with caution. It is important to bear in mind the

111
 Law and order in Sung China

cannot be of much help in tracing long-range changes in criminal be-

havior, but they still offer some glimpses of the difficulties faced by the
law-enforcement system. In 968, for example, in one sweep of the capital
city, the authorities arrested 367 juvenile delinquents and deserters who
had turned to robbery. 157 About a century later, an official wrote that in
Ching-tung Circuit some 80 forceful robbers had been arrested in some
fifteen months, this in a region that at this time had a population of
almost 7.5 million. 158 Without trying to push these figures too far, we
can probably say that the arrest rates were low relative to the popu-
lation. One thing this indicates is the ability of even small numbers of
troublemakers to disturb local tranquillity.
The numbers of men held in jails may also help in trying to form a
picture of the numbers of criminals in an era. We learn of the numbers
held in the capital because the emperor periodically examined cases
there and freed or reduced the sentences of most prisoners. In 995 it is
noted that when he had inspected the records of the prisoners held in the
various jails in Kaifeng, "those who received mercy numbered several
hundred." 159 This figure seems to have been not unusual for the capital
area. In 1005 when the emperor repeated this process, there were again
two hundred prisoners. 160 In 1078 the jail of the State Finance Com-
mission had more than seventy criminals, and the offices of the Kaifeng
Prefecture had more than one hundred. 161 This rather large number of
prisoners in the jail of the administration of the capital prefecture is
perhaps not surprising. Kaifeng was certainly the largest city in the
empire at the time, and the number of prisoners held in the numerous
jails there could be very large. In addition to the jails belonging to local

process by which the figures were generated and that the officials who created them
might have reason to alter the facts. An official writing at the turn of the twelfth
century commented that the Court had been receiving reports that suggested that the
numbers of robbers had decreased since certain changes in the laws had been enacted
but that in fact the reduction "really is the result of subordinates fearing to report
robberies for fear of being censured or having their actions reversed, not really because
robbers are fewer in number" (WHTK 167.1451).
157 HCP 12.11b.
158 HCP 310.12a. The men had been arrested since the last Hall of Light amnesty, which
seems to have been on the fifteenth day of the tenth month of the second year of the
Yiian-feng reign period (November 11, 1079) (see STCLC 53-54 and SS 15.23a). The
population figure is only very approximate. I have used the figures originally from
the Wang Ts'un, Yiian-feng chiu-yu chih (Taipei: Wen Hai, 1967), for 1080, as cited in
Hope Wright, Geographical Names in Sung China (Paris: Ecole pratique des hautes
etudes, 1956), supplemented in the case of Ch'i Prefecture by the figure from the Sung
shih which is dated 1102. The household total derived in this way was multiplied by a
factor of five to give a rough population total.
159 HCP 37.6b.
160 HCP 6 1 . n b .
161 HCP 288.9a.

112
 Crimes and criminals

Table 3.2. Numbers of bandits reported

Place Numbers Households

Shih Prefecture in K'uei-chou Circuit 400 bandits 19,802 (A.D. 1102)
Hua Prefecture in Ching-hsi North Circuit 40 bandits 13,542 (ca. 1000)
Wei Prefecture in Ho-pei West Circuit 70 bandits 10,482 (ca. 1000)
Yuan-ch'u District in Ho-tung Circuit 80 bandits (nofiguresavailable)
T'ai-p'ing District in Ho-tung Circuit 153 bandits (nofiguresavailable)
Wang-wu District in Ching-hsi North Circuit 100 bandits (nofiguresavailable)
Chi-lu Garrison Town in Yung-hsing Chun 60 bandits (nofiguresavailable)

Source: HCP 42.9a-lla.

administrative units, because a number of central government offices

and army units had their own jails, the total number of prisoners could
be huge. A complaint from 1091 indicates that in that year there
were almost one thousand prisoners in the capital and its surrounding
prefecture.162
At those times when the emperor was reviewing cases in the capital,
he also usually sent officials out to the circuits to review cases and extend
mercy. In 998, as a result, 3000 prisoners were freed.163 A few years later
the emperor himself examined the lists and freed 4606 men. 164 The large
numbers in these two cases, early in the dynasty when Kaifeng was
considerably smaller than it was to become later in the eleventh century,
suggests that the emperor examined the lists for the outer circuits as well
as those for the capital in these instances. This may also have been the
case in 1076. In that year, as a result of his review, 1726 men out of a
total of 2500 were freed or had their sentences reduced. 165
Another source which throws some light on the prevalence of crime is
the figures given by local officials, who were required by law to report on
the incidence of banditry in their jurisdictions. Other concerned officials
also could and did send in their own figures. In 997, for example,
officials commenting on bad omens provided the figures on bandits that
are given in Table 3.2.
Scattered individual reports suggest that gangs could range from a
handful of men to hundreds. In 1042 the "border bandits" in the capital

162 HCP 459.1a.

163 HCP 43.5b.
164 HCP 55.18b.
165 HCP 42.9a—1 ia. A comment from 1036 mentions a dozen bandits in Fu Prefecture
(HCP 63.13b), and from 1049 we have a mention of some twenty-two bandits in Tz'u
Prefecture (HCP 166.17a).

113
 Law and order in Sung China

prefecture itself were said to number some 690 men. 166 A gang of 100
men was reported for Kuei-yang Industrial Prefecture (a special type of
prefectural unit in which industry, such as mining, dominated the
economy) in Ching-hu South Circuit, and another gang of 400 men in
Chien-ch'ang Military Prefecture in Chiang-nan West Circuit. In the
following century, Wang Chih-wang (d. 1170) reported on the capture
by the authorities in Wen Prefecture of a gang of 96 pirates, and by the
authorities of T'ai Prefecture of a gang of 57 pirates (both prefectures in
Liang-che Circuit). 167 Clearly, gangs of such size were beyond the power
of the regular law-enforcement agencies. This information reinforces the
evidence just presented that even small groups of men were capable of
seriously disrupting local life.
Even the smaller gangs were a serious problem. The official Yii Ching
complained in the 1040s that although in Chieh Prefecture in Yung-
hsing Chun Circuit and in Ch'ih Prefecture in Chiang-nan East Circuit,
the bandits numbered no more than ten, they still were able to enter the
towns openly. In Teng Prefecture of Ching-hsi South Circuit, although
the bandits numbered not more than twenty, they had been able to elude
capture for several years. An official reported that in the strategically
vital circuit of Ching-tung, the larger bandit gangs numbered no more
than fifty to seventy men, whereas the smaller ones numbered only
twenty to thirty men. 168
In deciding how to respond to gangs, ten or more bandits seems to
have been a key number in the thinking of Sung officials. There is some
evidence that when there were gangs of this size or larger, the officials
could mobilize the military. 169 This response was presumably based on
the belief of higher authorities that the patroling inspectors and sheriffs
could not handle even the smaller gangs effectively. Local authorities
seemingly often simply waited for the arrival of major military forces,
but when faced with bandit incursions, even the military officers in some
areas "withdrew and would not go out of the gate." 170 Therefore, the
effects of bandit gangs were worse because the authorities sometimes
had great difficulty eliminating them, so that a gang might continue to
plague an area for years. One comment in the institutional encyclopedia
Wen-hsien t'ung-k'ao mentions gangs that had been "noted on the rolls ten
or more times." 171 The ability of such small groups to persist over time
166 HCP 12399.25b.
167 Wang Chih-wang, Han-p'in chi, 7.24a-25a.
168 HCP 141.16a.
169 See also a report of 1074 in which a decree commented on a rash of reports of gangs of
ten or more bandits (HCP 253.3a).
170 HCP 141.16a.
171 WHTK 167.1452.
114
 Crimes and criminals

and, as we stressed, to cause trouble all out of proportion to their

numbers, perhaps demonstrates the difficulties of law enforcement under
premodern conditions of transportation and communication, as well as
other factors such as local sympathies for the bandits and dissaffection
from the government, and the protection of bandit groups by locally
powerful figures.

Conclusion
Crime is a mirror of society. Many acts - homicide, robbery, arson,
fraud - are crimes in most societies. However, if we look at what crimes
were felt to be especially worrisome, we can see reflected the image of
those issues of greatest concern, especially to the ruling elite that creates
the legal system. The elite is concerned both with acts committed by its
members which undermine its position and with acts of outsiders which
threaten to weaken the hold of the rulers on the ruled. For the traditional
Chinese government elite, abuse of office, a threat from within, included
various criminal acts such as the abuse of prisoners and the disobedience
of orders, but it was corruption, the abuse of office for personal gain, that
troubled Sung officials the most. The other crime category of greatest
concern was banditry, organized robbery by groups of people outside the
elite and implying the use of force. Sung society suffered from many
other general types of criminal actions familiar from other cultures -
homicide, rape, theft, assault, economic crimes such as counterfeiting -
but banditry was particularly troublesome, because it implied violence
plus organization.
Bandits were also worrisome because of who they were, young un-
attached males, often with some military skills though probably little
formal education. Such men were capable not only of causing trouble by
themselves but more dangerously of rallying behind a leader with larger
and more dangerous aims. The same concern marked Sung government
attitudes toward religious groups. Officials were clearly aware of the
explosive potential of religious commitment combined with seditious
aims. A religion like Manichaeism — secretive, intensely solidary, and
espousing values sharply in conflict with some of the values of the
Chinese elite - was the very archetype of the threatening cult. The Sung
authorities attempted to repress Manichaeism in that part of southeast
China where it flourished and fulfilled their own prophecy by pushing
the Manichaeans into repeated revolt.
The very existence of widespread Manichaean influence in some
regions of China reflects another important truth about Sung China.
Looking backward we tend to overestimate how thoroughly settled
115
 Law and order in Sung China

China was and the degree of local control exercised by the authorities. In
practice, Chinese society, especially rural society, continued on its own
way without official interference as long as the people paid their taxes
and did not cause too much trouble. The pattern of de facto local
autonomy meant that the local common people feared the locally power-
ful families more than they did a distant formal government. Among
themselves these same commoners might come to view all outsiders as
fair game. In good times the people in some areas had no reason to
break the laws; in bad times the shift to banditry was an easy and
natural one.
The society in Sung China was also loosely knit. Certainly, for most
Chinese, kin groups and neighbors were of great importance. And yet
because of the increase in trade, urbanization, and improvements in
transportation, the Sung dynasty was an era of unprecedented personal
mobility. People moved around in large numbers. Soldiers, merchants,
itinerant artisans, Buddhist and Taoist clergy, officials, and even people
who were in essence tourists, traveled about the Sung domains to a
degree certainly unequaled in earlier times and possibly not exceeded in
later dynasties. Controlling this bubbling, adventurous population was a
knotty problem for a government that had only limited resources and
had some doubts about the agents it would have to use for such control.
This combination of a relatively open society and limited formal re-
pressive institutions helps explain why the state could be so worried
about gangs as small as ten men that it would sanction mobilizing the
military. It also explains why, despite such authorization, the local
authorities could fail year after year to root out criminal groups and
how such groups could occasionally grow to number hundreds or even
thousands of members. The government really had no choice. It was
forced to rely not just on its own formal mechanisms but also on the
people themselves and groups organized among the people that would
work for law and order.
Many aspects of Chinese life thus did not, in practice, fall under the
purview of the formal government. Still, despite this caveat the amount
of control exercised by the central authorities was truly impressive for a
premodern society. Certainly, crime was endemic, but the ability of the
Sung state to keep it under some control is striking. This ability partly
reflected the effectiveness with which people were socialized into values
supportive of good order, but it was also a tribute to the effectiveness of
the law-enforcement agencies and a penal system that seemed to succeed
in many cases in turning violators toward new noncriminal behaviors.

116
 4
Informal and
semiformal agencies of
law enforcement
In the abstract, the problem of law enforcement can be simply stated.
Those with a monopoly of legitimate force must receive accurate and
timely intelligence of wrongdoing and must be able to mobilize a suf-
ficiently large force of adequately armed, trained, and properly motivated
men and to move them to the affected area in a timely fashion, at a cost
acceptable to the state. The problem lies in the practice. For economic
and ideological reasons, the level of applicable force is usually inversely
related to its closeness to the triggering events and to information
about those events. Most crimes occur at some remove from the major
centers of state force, which are the most highly trained government
military units. Those most likely to have accurate information are
victims, witnesses, and neighbors. They could react most immediately to
the commission of the crime and inform higher authorities. But would
they? Even today, with the speed (and anonymity) of electronic com-
munication, a great many crimes are not reported or are not reported
immediately. If the authorities come to depend on self-defense by parts
of the society outside the formal government apparatus, how can they
ensure that these outside agencies will not become too powerful? For
premodern Chinese states this was always a dilemma to be lived with,
not eliminated.

Self-defense
In the past as in modern times, the first line of social defense was the
people themselves. In modern China as in the Sung period, walls go up
as wealth goes up. In much of the contemporary Chinese world the tops
of walls around houses are studded with glass fragments, and high-rise
China covers its outer windows with steel gratings, creating dreadful
firetraps but deterring burglars. Now, as in premodern China, given the
incomplete success of socialization, the problem is first how crimes can
117
 Law and order in Sung China

be prevented or at least discouraged and, second, how to respond effec-

tively to crimes in process or already committed. Obviously, the ideal
is to keep crimes from being committed. But crime was endemic in
traditional Chinese society. The burial pottery of Han times and later,
with its models of houses built like small fortresses, vividly demonstrates
the price to be paid for personal security in traditional times. That the
men of Sung also subscribed to the English adage that an ounce of
prevention is worth a pound of cure is made abundantly clear in the
writings of the Southern Sung official Yuan Ts'ai (fl. 1140-95):
Solid houses
In managing a family, you should see to it that the fence walls around your
house are high and sturdy, the hedges dense, and the window, walls, doors, and
gates secure. Repair any damage promptly. If there are drainage outlets for
water, put grates over them. Keep them new and strong and do not neglect
them. Even though clever thieves will be able to find ways to enter in a moment
by boring holes in fence walls, cutting through hedges, knocking holes in house
walls, and breaking open doors, you will at least not be inviting them in as you
would by allowing dilapidated fence walls, broken hedges, rotten house walls,
and worn-out doors. Moreover, through these maintenance efforts you will avoid
the problems caused by slaves running off, or unworthy sons sneaking out at
night. Although the authorities can deal with robberies, runaways, or trouble
caused by unworthy sons, the cost will be high.1
Estate tenants
Your residence may be off by itself in a quiet rustic spot in a mountain valley. If
so, be sure to set up estate tenant houses at strategic spots in a circle around the
house. Recruit families with several adult men to live in the houses. Then if there
is a fire or an intruder, the tenants can come to your aid when you call for help.

At this primary level, early knowledge was crucial. Yuan Ts'ai continues
his advice:
Raising alarms at night
The barking of dogs at night is not always a sign of the arrival of intruders. Yet
because there could be a thief making a test, you must not assume an innocent
explanation and fail to raise an alarm. The same holds true for strange noises
heard at night; do not simply attribute them to mice and fail to raise alarms.
Patrols
You ought to see that there is a passable road in a full circle around your house.
At night have ten or more patrols on this path. Careful planners live near the
city wall and away from empty lots. In addition they have a double set of walls
and send someone to patrol in between them. Within the house itself, you can
have the young men and the bondservants take turns making patrols.

1 For this and the following quotations see Ebrey, Family and Property in Sung China, pp.
278-81.

118
 Informal and semiformal law enforcement

Potential victims sometimes created more problems by tempting criminals

or by fostering animosity among their poorer neighbors. In Yuan Ts'ai's
opinion, crime could be reduced if people treated their neighbors
humanely and if they exercised a sensible restraint in displaying their
wealth.
Tempting thieves
Families with great accumulations of goods are the ones thieves love. And
because the people in these families often like to display their magnificence, they
set out their treasured articles; this makes the thieves salivate even more! Rich
and substantial families may store money and grain but should not collect
precious objects of gold, jewelry, or silk. That way, even if they are robbed, their
losses will not be large.
Oppression as a cause of robbery
Although robbers are the clearest examples of inferior people, they also have
rationales. Rich families who normally are not oppressive and moreover are able
to be charitable and to do good deeds will be spared during disturbances created
by fighting or fire. The robbers will be merciful to them and not burn down their
houses. Generally the families robbers derive satisfaction from burning, looting,
and debauching are those of people who have done many evils. Rich families
each ought to examine themselves in this regard.

However, even if people behave in a humane and prudent way, there

will still be crimes. Those most immediately affected should be armed,
trained, and ready to react.
Chasing thieves at night
If you hear a thief at night, you should immediately shout "There's a thief" and
proceed to catch him in a deliberate fashion. The thief will surely try to escape.
You should not try to hit him in the dark, for there is the danger that the thief's
panic will lead him to wound you with a knife, or by mistake you might hit
someone in your own family. If you take a candle to look at the thief, there will
still be time to strike him. When a thief is captured, naturally you should abide
by the law. Do not beat him excessively.
In case of armed robbery
Some robbers come in the middle of the night, with torches and unsheathed
knives; they force open the door and enter the house. It is essential that you be
prepared for such incidents. You should send people to keep a lookout at the
entrance to nearby roads. If anything out of the ordinary occurs, they can warn
you. Also, arrange for a side exit that the old, the young, and the women can use
for escape in case of emergency. You should also, as a routine matter, have the
young men in your family and the male servants maintain the weapons and
know what steps to take to defend against attack. If they can withstand an
attack, they should do so; if not, they should avoid confrontation. But by no
means let the robbers capture any family members. If someone is taken, he or
119
 Law and order in Sung China

she will be made a hostage, and the local mutual security group and the police
will not dare pursue the robbers.
Yuan Ts'ai's analysis suggests that he felt the level of security during
his era was not especially high; his advice suggests that the formal
apparatus for maintaining law and order could not be relied on to defend
property in a timely fashion. The first line, therefore, had to be self-
defense.
When feasible, people also needed to make preparations to resist
criminals. Occasionally the sources describe such self-defense. In i o n a
patroling inspector, the chief local military law-enforcement agent,
reported that a commoner in his jurisdiction, when bandits attacked,
had "rallied the people, who grasped clubs and fought." The patroling
inspector was eventually able to seize all of the bandits, and the com-
moner's family was rewarded by imperial decree. 2 Resistance was
mandatory, but it still was risky. Adequate force did not mean excessive
force. Yuan Ts'ai's advice against beating robbers excessively was very
much to the point. According to Sung law, killing a robber might be
prosecuted as homicide and could lead to a death sentence. 3
Cooperative self-defense was necessary, because the formal authorities
obviously would arrive too late to be of any immediate help in protecting
persons and property. Furthermore, in the absence of modern com-
munications, even if local people did try to report a crime, time would
pass before the police agencies could react. The investigative stage of law
enforcement and the pursuit of the criminals ordinarily could begin only
some time after the events in question. As Lu Yu (i 163-1210) remarked
about petty bandits, "They hide in the mountains and forests, and come
out at night to raid. By the time the officials learn about it, they are
already scattered." 4 Even when they did learn quickly, officials still had
to mobilize in order to respond, with all the delays that this entailed.
Although traditional Chinese empires had, by premodern standards,
exceptionally good systems of communication and transportation,
they were nonetheless premodern. The most important sorts of com-
munications could be forwarded very quickly during Sung times, reaching
a standard of 175 miles per day. This extraordinary speed was ordinarily
used only for sending out imperial grants of amnesty. Messages classified
under the general category of urgent, including most messages reporting
local unrest and crime, were carried only 140 miles a day (using runners
who traveled day and night between posts that were relatively close

2 SHY, ping n.6a-b.

3 Compare CYKC 54.
4 L u Y u , Wei-nan wen-chi, 13.11b.

120
 Informal and semiformal law enforcement
together), and the regular horse post moved at only 105 miles per day.
These rates are far better than those reached by the Romans, better even
than those of the runner system of the Incas, and were exceeded in
premodern times only by some of the postal systems of the Sino-barbarian
empires of east Asia. 5
Transmitting information about local problems took time, and then
more time would be spent in the formal process of mobilizing troops.
Mobilization required formal authorization from the regular chain of
command. Finally, even after mobilization, the troops would ordinarily
have to march to the area affected. In short, in most cases official help
could hardly be expected to arrive in time to help the immediate victims.

Mutual responsibility
The need for timely and effective action is a constant in law enforcement.
One traditional Chinese answer to this problem, which fit well with the
desire of the state to minimize expenses, was to make each man his
brother's keeper. In the words of Liu Yueh (1144-1216), "When one
family has a robber and cannot seize him themselves, then the group of
neighbors is to arrest him." 6 Again, the very process of law enforcement
both encouraged and depended on cooperation. The ideological roots of
this commitment to mutual responsibility go back to Chinese antiquity.
There was a deep-seated, though rarely articulated, conviction in
premodern China that everyone participated in the creation of good
order. This clearly implies the responsibility of all, but it also suggests on
another level the feeling that guilt was shared, that lawbreaking was in
some way and to some degree the responsibility of those close to the
lawbreaker, in terms of both physical distance and relationship. This
attitude is akin to the Western legal concept of due care. People have a
responsibility to exercise due care to prevent harm. This implies a duty
to supervise appropriately people with whom they are associated. The
Chinese belief in the analogous responsibility, which underlies some
of the current police practices on the China mainland, goes back to
classical times in China.
From a legal point of view this belief was embodied in the rules which
made bystanders responsible for intervening to halt lawbreaking and
punishing them if they failed to do so. In premodern China, witnesses
and neighbors (and victims if they were still able to help) were expected
to attempt to apprehend criminals committing major crimes or at least
5 See Peter Golas, "Courier Transport System," Harvard University, East Asian Research
Center, Papers on China 20 (1966): 1-22.
6 Liu Yueh, Yiin-chuang chi, 7.6a.

121
 Law and order in Sung China

to go to the aid of those attacked. (For minor crimes they might report
the matter to the authorities.) 7 Rules concerning this sort of responsi-
bility can be found in some of the earliest examples of Chinese written
law, from the preunification Ch'in state. 8 Such rules can also be found in
the T'ang Code and the earliest Sung Code (which reproduces the T'ang
Code almost unaltered):
When neighborhoods and administrative villages receive a report of a robbery or
homicide but fail to render assistance, [the responsible agents] are liable for one
hundred blows of the heavy rod. When they merely hear of such an incident, this
penalty should be reduced by one degree. If their strength is insufficient to
render assistance, they must immediately inform the nearby offices. If they do
not report it, they are to be tried according to the rules on not having rendered
assistance.9
The sources rarely mention anyone's actually being prosecuted under
this rule or the analogous ones that mandate the responsibility of neigh-
bors in cases of fire, but again Yuan Ts'ai suggests that people might
indeed be punished. In advising the rich to be charitable, he states:
There are gentlemen-officials [shih-ta-fu] who in ordinary times use their official
influence to maltreat and bully their neighbors. One day an enemy may come
and attack their family or burn their house, but the neighbors will warn each
other, "If we put out the fire, afterwards not only will we receive no reward, but
that man will bring charges against us, saying we stole his family's goods. Who
knows how long the case will be at court? If we do not go to his aid to put out the
fire, all we'll receive is one hundred blows."10
The code provisions called for a twofold responsibility: to resist crimes
and seize the criminals if possible and to inform the authorities in other
cases. Getting information, the key to effective action, was always a
problem for the authorities. Crimes were supposed to be reported, under
threat of sanctions, but then as now the authorities had reservations
about the willingness of local people to involve themselves in such
matters. Modern studies indicate that only a fraction of crimes com-
mitted are reported to the authorities. In part this reflects the age-old
unwillingness to get involved with the law. Such involvement meant
inconvenience and loss of time, for little apparent benefit. This general
distaste was strengthened in the Sung, as it would be in most societies,
7 SHT 28.451; see also TLSI 4.59.
8 See A. F. P. Hulsewe, Remnants of Ch'in Law (Leiden: Brill, 1985), p. 146.
9 SHT 28.451. This rule was taken over from the T'ang Code. For such systems in the
T'ang see Wallace Johnson, "Criminal Procedure at the County Level in T'ang
China," paper presented at the conference "The Transformation of Chinese Law,
T'ang Through Ming," Bellagio, Italy, August 10-14, l9^1-
10 Ebrey, Family and Property, p. 282.

122
 Informal and semiformal law enforcement

when neighbors were implicated or the criminals were dangerous. 11 In

n o o , commenting on the recent reduction of penalties for robbery, a
censor said that because of this, "ordinary people are afraid to report
robberies, because they know that the robbers will not be executed, and
the local neighborhood groups are unwilling to seize such men for fear of
their vengeance." 12
Even before the empire was founded in the third century B.C., Chinese
governments responded to the unwillingness of the local people to give
information by establishing among them groups whose members were
mutually responsible for reporting on and suppressing one another's
wrongdoings. 13 The origins of this practice are often traced to the legalist
reformer Shang Yang. The biography of Shang Yang in the Historical
Records reports that on his advice, the people of Ch'in were "organized
into groups of fives and tens mutually to control one another and to
share one another's punishments," and the Han Fei Tzu says that "Lord
Shang taught Duke Hsiao of Ch'in to organize groups of ten or five
families and established a system of denunciation of crimes and joint
responsibility for offenses."14 Shang Yang sought to prevent the growth
among the people of any organized centers of resistance to state power.
By mandating mutual responsibility groups, he helped pulverize society
by sowing the seeds of mutual distrust among neighbors and associates.
In ancient China as in the People's Republic today, the question of
whom you can trust with your inner feelings was an agonizing one. Such
formal devices for applying negative sanctions (and positive sanctions
in the form of rewards) were one creative (if, from a contemporary
perspective, distasteful) answer to the continuing problem of inadequate
information about crimes.
Such mutual surveillance organizations continued to play a role of
varying importance from that time on. Sources concerning the early
Sung suggest that such groups were not a major concern to the auth-
orities. This paucity of references may mean a lack of sources for this
era, but it also probably means a relatively low level of lawlessness in
many areas of the empire before the 1040s. In the second half of the
eleventh century, and especially after the introduction of the pao-chia
militia reform in the 1060s, such mutual security organizations pro-
liferated. During the Southern Sung most categories of people, including

11 Compare CYKC 66.

12 WHTK 167.1451; see also SHY, ping 12.3a.
13 See Hulsewe, Remnants of Ch'in Law, passim.
14 J. J. L. Duyvendak, The Book of Lord Shang (London: Arthur Probsthain, 1928), p. 14;
W. K. Liao, trans., The Complete Works of Han Fei Tzu (London: Arthur Probsthain,
i959)> vol. 1, p. 115.

123
 Law and order in Sung China

officials, clerks, students, miners, soldiers, and fishermen, as well as the

peasantry, were supposed to be enrolled. The frequency with which
Sung officials discussed such organizations indicates that they not only
became more numerous but also were seen as more important as devices
for social control.
By mandating the organization of such groups and making their
members mutually responsible for one another's behavior and for
reporting the presence of wrongdoers in their jurisdiction, the state could
hope to intensify its degree of control at little cost. In addition to
mandating groups for purposes of mutual responsibility, the state also
linked the culpability of members of certain involuntary associations,
most importantly families and official bureaus. Members were held
responsible for (and therefore had a motive to stop or report) the illegal
activities of fellow members. 15
When such systems failed in their purpose, the security of the state
could be endangered. Thus officials were quick to complain of ineffective
enforcement of mandated systems. In 1080 an official complained that
because the pao-wu militia system was not functioning properly, young
hoodlums remained concealed among the people. 16 And a decree in
1078, reflecting the deliberate use of such groups as a control device,
stated that "these miners . . . are riff-raff from all over. If they are not
organized into mutual security groups [shih-wu], with the members
subject to joint adjudication for the concealing of wrongdoing, then
the evildoers will stay hidden among them, and it will be difficult to
investigate." 17
This passage highlights one weakness of such systems. The groups
within which crime was most likely to occur were those comprising men
outside the usual stable structures of Chinese traditional agrarian society.
Yet such men, themselves strangers in their communities, were perhaps
least likely to have good information on wrongdoing by anyone outside
their immediate circle and almost by definition would have little incen-
tive to inform the authorities. The settled people, the peasantry that
formed the great majority of the population, might have a better idea of
local wrongdoing, but their close ties with relatives and neighbors, which
made possible their superior knowledge, at the same time might dampen
their willingness to inform, especially if local people were involved in the
criminal acts.

15 Brian E. McKnight, "T'ang Law and Later Law: The Roots of Continuity," Journal of
the American Oriental Society 112.3 (1991).
16 H C P 309.5a.
17 H C P 293.8a.

124
 Informal and semiformal law enforcement

Information systems
The system of mutual security units was thus a device for generating
information about crime and criminals. The desire for a regular flow of
information on local conditions may have been one reflection of the
general Sung government mania for compiling records. As a political
system the Sung is famous for its attempts, which had considerable
success, to control local functionaries. These attempts were based on an
elaborate system of records. Such records were compiled not only within
the formal government but also by the local units into which the popu-
lation was divided. These local records served to control strangers, who
were always considered suspicious persons, by locating all legitimate
people within groups, so as to isolate those who were not legitimately
residents of a particular area. Chu Hsi speaks of ordering the village
officers called superior guard leaders and their assistant leaders to
recompile the guard registers so as to ferret out concealed deserters or
roving criminals. The village officers were to refer such cases to the
chief civil law-enforcement official, the sheriff, who would see to their
apprehension. 18
Chinese investigators were not so naive that they were willing to
depend exclusively on such groups; they were well aware that local
people tended to protect their own. 19 They therefore made regular use of
spies and informers to gather information. The Sung casebooks have
numerous anecdotes about the skillful use of such agents. 20 The early
Sung statesman Wen Min-chung is said to have doubted the guilt of the
aforementioned Buddhist monk who had confessed under torture to
having murdered a woman. So he sent to the village in question a clerk
dressed not in his usual clerk's clothing but in the kind of clothing that
might have been worn by a bandit. The clerk ate at the local inn. When
the old woman working there heard that he was from the prefectural city
and did not recognize him as a clerk, she asked about the monk. The
clerk said that the monk had been beaten and killed the previous day.
The old woman asked what would happen if the real bandit were
caught. The clerk replied that since the prefecture had already made an
error, it would not dare to raise the issue if the real bandit were caught.
The old woman then said that the real killer was a youth of the village,
from such and such a neighborhood guard unit. The clerk asked her to

18 Chu Hsi, Chu Wen-hung wen-chi, 99.15a ff.

19 See SHY, ping 11.7b, for a report on how the failure of local village officer leaders to
report crimes contributed to an increase in banditry.
20 For examples, see Robert Van Gulik, trans., T'ang-yin pi-shih: Parallel Cases Under the
Pear Tree (Leiden: Brill, 1956). See also Li Kuang, Chuang-chien chi, 12.17b; CYKC 66.

125
 Law and order in Sung China

point out the youth's house, and when she had done so, he seized the
culprit. The monk was freed.21
Motivation was as critical to promoting a flow of accurate information
as it was to all other facets of law enforcement. Given the understand-
able reluctance of the local people to get involved, the authorities had
to provide both positive and negative sanctions to promote popular
cooperation. The spies and informers on whom the authorities depended
were paid through an elaborate system of rewards for information. And
the authorities paid well for what they needed. Reports from the 1030s
indicate that anyone who brought an accusation of homicide was given
fifty strings of cash (presumably after the criminals were convicted), and
those who reported the deaths of prisoners from abuse by their jailers
were to be given one hundred strings (each string being valued at one
thousand coins). 22 In this same period any person reporting on a group
of ten or more bandits who had killed some of their victims was to be
paid ten strings. 23 A decree from the late 1020s indicates that people
could be given other things besides money. Those in the army might be
granted increases in rank. Commoners might be spared taxes. Men
liable for corvee might be exempted from performance. And if the report
resulted in the capture of the whole gang, the informers might be given
commendations in addition to their other rewards. 24
Rewards were graded according to the sentences eventually given to
the criminals. According to the rules of the 1090s, people were rewarded
with ten strings of cash for robbers whose sentence was beating with the
heavy rod, twenty strings if the sentence were penal servitude, and thirty
strings if it were exile.25 The meaning of rewards of this size may be
made clearer by comparison. In the late eleventh century a postal
courier in Sung service is said to have cost the government approxi-
mately twenty strings per year in salary and benefits. 26 In the relatively
early part of the Northern Sung, one string of cash would buy from two
to three bushels (shihh) of rice, enough to feed a single person for the
better part of a year.
The state also wanted to establish procedures that would facilitate the
timely and effective use of this information-gathering system and its
attendant system of rewards. The central authorities were not sanguine
about the commitment of local officials to the effective execution of
orders. In their search for methods to control local functionaries, the
21 CYKC 16-17; see also CYKC 66.
22 HCP 114.19b; SHY, ping ii.i5~a-b.
23 HCP 107.4b.
24 SHY, ping, ii.nb—12a.
25 HCP 453.13a.
26 Golas, "Courier Tranport," p. 12.

126
 Informal and semiformal law enforcement
Sung central government depended on an elaborate network of admin-
istrative regulations that specified the requirements of government func-
tions. Such administrative laws were established to implement the
reward systems. Guidelines were provided for the officials, and penalties
were drawn up for those who hindered the effective working of the
reward system. Officials who sought to block rewards might suffer penal-
ties, albeit relatively moderate ones. One report from the last years of the
eleventh century says that the penalty was to be equivalent to one
hundred blows of the heavy rod, and another from the early twelfth
century says that they would be liable for a penalty equivalent to eighty
blows.27 As officials, the men so charged would have been able to
commute such penalties into relatively minor administrative sanctions,
though the entry would not look good on their dossiers. Eventually the
officials were ordered to certify and process reward claims according to
fixed time limits. In the early twelfth century an official complained that
no time limits had been established except for the thirty-day time limit
for processing rewards in cases involving robbery. He petitioned success-
fully to have these limits fixed, in a system analogous to that used in
setting time limits for capturing criminals. 28
The state's perennially weak fiscal position also influenced its policies
on rewards. By the standards of most other premodern states, the Sung
empire had a very effective taxation system, but there still were limits on
its ability to raise revenues. Much of what was raised had to be devoted
to national defense (which absorbed about half the annual income) or to
supporting the civil bureaucracy and the Court. Relatively little was
available for other uses. In many areas of government concern, including
the financing of rewards, the authorities responded by trying to shift
the burden. Although such rewards might be paid from official funds
established to reward those who informed about or helped capture
criminals 29 or might be drawn from various other official sources, 30 they
seem more often to have come from monies gathered by village officers31
or from property confiscated from the accused or his followers.32 In 1097

27 SHY, hsing-fa 2.86a; HCP 494.10a.

28 SHY, hsing-fa 2.85b-86a.
29 See, for example, the request of the judicial intendant of Yung-hsing chun Circuit that a
quota of four thousand strings of cash per year be set aside as reward money (SHY,
ping 12.9b).
30 HCP 374.6b-7a, 398.9a. See also SHY, ping 12.2b-3a, in which the report says that
Hopei East Circuit had only two thousand strings per year of official reward money.
This was insufficient, and it was decreed that it be supplemented.
31 Ts'ai K'an, Ting-chai chi, 5.1b; see also SHY, ping 13.20b-21 a.
32 The use of the criminal's property was a traditional practice. See, for example, the rule
in the early Sung Code, SHT 28.451. It continued to be standard Sung practice in later

127
 Law and order in Sung China

the emperor approved a suggestion by the Ministry of Personnel that

in cases in which the robber was truly indigent or had insufficient
resources, the needed goods could be taken from his followers or from
people who had been aware of the crime but had done nothing. 33
Rewards might also be taken from the illicit goods involved. In the 1080s
in an attempt to suppress the smuggling of Korean goods by the people
of Wen Prefecture in Liang-che Circuit, the authorities instituted a
policy of giving informers three-tenths of the goods recovered. 34
The never-ending search for information also affected other law-
enforcement policies. Because the criminals themselves knew more than
any others about their own activities, the state also tried to induce
criminals to inform on other criminals. Under one policy, enacted in
1087, when a person had been convicted of a robbery, a sign was to be
hung on his gate describing the crime. However, if such a person, having
been sentenced to penal servitude, was able to supply information lead-
ing to the arrest of one robber or two thieves also liable for penalties of
penal servitude or exile, his family could be spared the shame of such a
notice. A criminal sentenced to beating with the heavy rod also would be
spared the posting of the sign if he could accuse or enable the arrest of
one thief liable for penal servitude or exile. 35 (The authorities were, of
course, aware of the temptations of bringing false accusations. The
general rule was that anyone who brought a false accusation would be
liable for the penalty that would have been inflicted on the supposed
criminal.)
This information-generating function of local people was coupled with
the responsibility for restraining or seizing criminals when this was
possible. Again the state used a complex pattern of rewards as an
inducement for capturing or, in some circumstances, killing criminals.
Ordinary people, as well as militiamen, constables, and soldiers, might
be richly rewarded for their work. State authorities also recognized the
necessity of compensating those who were injured while trying to en-
force the law. A decree of 1023 stated that "from now on, the [constables
called] archers, [the village servicemen called] elders, [the elders' aides
called] stalwart men, and commoners of the various circuits, who in the
course of capturing bandits suffer medium or serious injuries, should be

times; for the rule, see TFSL 186. See, for examples, HCP 132.1a, and SHY ping
11.15a—b. One problem of using the bandits' booty for rewards was that it prevented
the robbery victims from being recompensed. See an order of 1012 that official funds be
used, so that the booty could be returned to its rightful owners (SHY, ping 11.6b).
33 HCP 493.25b; see also HCP 5oo.5b-6a.
34 HCP 302.6a.
35 HCP 398.12a-13b.

128
 Informal and semiformal law enforcement
given two thousand coins. Those who suffer light injuries should be
given one thousand coins." 36
As noted, even the criminals themselves could use the reward sys-
tem to reduce or eliminate their penalties. These rules for rewarding
criminals were in regular use. 37 Under the rules in force in 1077 for
Ho-pei and Ching-tung circuits, even if robbers themselves had killed
people, they could benefit. A bandit leader who captured or killed two
bandits, or a regular bandit gang member who captured or killed a
single bandit, might be rewarded as well as forgiven crimes if he sur-
rendered and confessed. (This policy was soon made somewhat less
lenient.) 38

Weaponry
These rules on seizing lawbreakers assumed that the local people had the
means to respond quickly and effectively to immediate threats, without
having to appeal to some distant center. Obviously their ability to
respond would vary directly with their degree of military training and
the weapons they possessed and inversely with the dangerousness of the
criminal situation. Unarmed witnesses might have no trouble subduing
an unarmed petty thief; even a single armed bandit might be beyond the
power of witnesses to control unless they were themselves armed.
Yuan Ts'ai indicates that the wealthy kept weapons in their homes
and might encourage both male family members and other male depen-
dents to learn how to use them. Other sources also suggest that weapons
and a knowledge of their use were quite widespread among the people. 39
However, any diffusion of power through the possession of weapons by
those outside the government ran counter to the feeling among Sung
officials that the presence in local areas of armed groups not under direct
official control might dangerously weaken the center. The idea of the
people in arms was always a specter to Sung authorities. The traditional
response was to try to walk a thin line between allowing local people so

36 HCP 100.8a.
37 See, for example, an incident recounted in the Che-yu kuei-chien in which an initial death
sentence for a robber who had killed his partners and stolen their booty was reversed on
the grounds that a robber who killed his fellows was to be spared. Admittedly, this was
an arguable point in this case (CYKC 52).
38 HCP 282.2b, 282.9b. See a story in CYKC that says that a robber had killed his
associates and taken the stolen goods, after which he was captured. Some legal officials
wanted to spare the man because he had killed his partners, but another official pointed
out that only those who surrendered and confessed could be spared (CYKC 52).
39 Brian E. McKnight, "The Rebellion of Fang La" (master's thesis, University of
Chicago, 1964), p. 58.

129
 Law and order in Sung China

little power that they could not protect themselves from lawbreakers and
giving them so much that they could ignore higher authorities.
An armed populace might be able to suppress criminals, but it might
also go into business for itself. The problem of ideology appears again.
Confucius had said that the people should be led with rituals and virtue.
In the real world, the coterie of people working for the interests of
the emperor could promote order by using rituals and the example of
virtuous behavior only to a limited degree. The emperor might want to
have as subordinates men who would act in his interest. Given the
numbers of those needed in the bureaucracy, it was clearly impossible
for the emperor to assess their qualifications himself. Furthermore he
had to compromise with the interests of those from among whom officials
were to be chosen. The Sung emperors dealt with this problem by
selecting one large group of subordinates in ways that clearly demon-
strated their power. These men entered the civil service as a result of the
emperor's grace, not as a result of their own objectively verified qualities.
To balance this group the emperor then brought into the civil service,
and indeed gave the greatest power to, a second group of men who
entered through the civil service examination system. These examination
graduates understandably felt that their selection rested at least in part
on their own abilities.
The civil service examination system served to certify the candidates'
knowledge of the Confucian Classics. The rulers might at least hope that
such knowledge in most cases could serve as a surrogate measure of
socialization into appropriate attitudes. If officials acted in accord with
the values enshrined in the Classics, they could at least be trusted to
behave in predictable ways which would not often threaten basic state
interests.
The loyalty of the more well-to-do among his subjects - though per-
haps less certain than that of members of the civil service - might seem
relatively reliable, since they depended on the state to defend their
privileges and wealth and had been exposed to one degree or another to
Confucian education. The danger was that if members of the wealthy
classes proved disloyal, they would be fearful enemies, because they
possessed the organizational and educational skills needed to threaten
the state. A twelfth-century official, analyzing the difficulty the state had
in controlling the ordinary inhabitants of two prefectures in Liang-che
Circuit through the standard mixture of rewards and punishments,
commented that he had heard a local pointing out that

the men of these two prefectures "do not fear the officials of the emperor but fear
the powerful people of the villages." Therefore they cannot be controlled by the
130
 Informal and semiformal law enforcement

officials but obey the orders of the powerful families. Generally, the power of the
powerful families has three bases. They surpass others in knowledge, in courage,
and in worldly goods. Having these three, they can overawe the commoners, who
cannot but bow their heads, submit in their hearts, act obsequiously, and follow
the lead of the powerful.40

The official's solution was to win over the local elites by giving them
rewards and honors and then to use them in controlling disorder. The
problem with this approach had been described a century or so earlier
by another official, who wrote of the dangers of having "traitorous
scoundrels," that is, local elite leaders, use the excuse of exterminating
bandits when committing outrages. 41 However, since the authorities
needed the help of the wealthy in controlling the peasantry and as
unofficial law-enforcement leaders, they overlooked violations of the
rules on owning weapons.
In good times, ordinary peasants also had a major stake in maintain-
ing the local order. They might at most times feel for the ruler, and the
elite which represented the ruler, a sense of awe and respect. However,
they were often living on the edge of bitter poverty. When times were
bad, for whatever reason, the officials of Sung times felt that the loyalty
of ordinary people was very questionable. At such times, because they
had neither wealth to defend nor values inculcated by Confucian educa-
tion, they were viewed with mistrust. The state probably could not have
disarmed the elite families even if it had wanted to do so. Peasants were
another story: They were less able to afford arms and training, and
so despite their numbers, they were more easily disarmed. The ques-
tion was, To what degree could the state presume upon the loyalty of
these people whose indoctrination was problematical? The overriding
aim of the authorities was to maintain their power. In the event, they
understandably chose the safer course. It might be neither possible
nor desirable to keep the rural rich from possessing arms; certainly,
the authorities would seek to limit the ordinary people's ownership of
weapons. Better a disarmed population, even at the cost of a higher level
of endemic crime, than an armed population of doubtful commitments.
This decision was reflected in laws that made it illegal for people to
possess arms, though there were occasional exceptions. 42 The early Sung
Code statutes provided that

40 Wang Chih, Hsueh-shan chi, 3.25.

41 Li Kou, Chih-chiang Li hsien-sheng wen-chi, 22.3a.
42 HCP 81.5a. Indeed, the local authorities at one point even attempted to punish the
state-selected archers in Szechuan for having armed themselves (HCP 86.17b; SHY,
chih-kuan 48.62a).

131
 Law and order in Sung China

all those who privately possess forbidden military materiel are liable for one and
a half years of penal servitude. [Note: This refers to weapons other than bows,
arrows, knives, spears, and short spears.] For a single crossbow the penalty is
increased by two degrees. For one suit of armor and three crossbows, the penalty
is exile to two thousand lid. For three suits of armor and five crossbows, it is
strangulation. For privately manufacturing, the above penalty is increased by
one degree. 43
This rule was specifically reaffirmed later in the dynasty. 44 Military
knowledge and skills among the people were also greatly feared. Books
on military topics were proscribed; men with military skills were
regarded as even more dangerous. According to the tenth-century Sung
Code, owners of forbidden military texts were liable for two years of
penal servitude. 45 By the time of the early thirteenth-century code, the
penalty had been increased to exile at three thousand li.d46 Actual
military skills were much more frightening. A decree in 1041 ordered the
beheading of the leaders of groups that gathered to study the martial
arts. Their followers were to be registered on army rolls as laborers at
the maximum-security location on Sramana Island, in the sea off the
north coast of the Shantung peninsula. Anyone who seized them was to
be given a reward of three thousand coins per person. 47 The fear of
martial arts was greatly increased by their close association in traditional
China with magical incantation and occult practices, which might lead
to organized and disciplined resistance to the state. 48

Semiformal agencies
Informal agents might be adequately informed and close to the scene,
but they were likely to be inadequately armed, trained, and led, and
of sometimes questionable motivation. One possible answer to this
dilemma of the state was to have the local authorities select or sanction
the selection of subsets of the people as a whole and to give these subsets
some means of enforcing order. Such groups would have the advantage
of involving fewer people, and so they would therefore be easier to
control. Because they were presumptively easier to control, the auth-

43 SHT 16.264.
44 HCP 117.19a.
45 SHT 9.155; TLSI 2.82.
46 TFSL 251.
47 HCP 134.9a-b; SHY, ping 11.18a. The severity of these penalties is perhaps related to
the particularly delicate internal security situation in the early 1040s when the
Sung state, engaged in a war on the northwestern border, suffered a rash of internal
disturbances.
48 See STCLC 735.

132
 Informal and semiformal law enforcement
orities were not quite so concerned about these men being armed and
trained. These groups in turn could perform functions which otherwise
might have fallen to regular police or troops and, in practice, were called
on to perform other necessary local duties. By creating such agents, the
state was able to shift a large part of the cost of local control onto these
men and their families.49 The state could also mandate the use of criteria
that would tend to select as members of the subsets men who by their
life circumstances might be better indoctrinated and more committed
to social and political stability, choosing, for example, only men who
owned land and requiring that those in leadership positions be the more
wealthy. By moving from the informal to the semiformal, the state
gained in security without sacrificing much in coverage.
Allowing such unpaid groups to bear some local responsibilities would
have the advantage of widely diffusing power without burdening the
state with unacceptably high costs. In premodern China there were two
types of such groups, local militia organizations, which aimed specifi-
cally at self-defense and crime prevention, and organizations of village
officers. Village officers were fewer in numbers than militiamen and had
broader functions. These village officers performed most important local
government tasks, including not only law enforcement but also tax
collection, population registration, and miscellaneous services.
During the Sung the terminology and organizational forms used for
these village officers varied considerably over time and from area to area,
despite the similarity of their functions. 50 At the beginning of the Sung,
the system in most widespread use closely resembled that in use under
the T'ang. Unpaid functionaries were conscripted from certain types and
grades of households to serve for limited periods of time in specific offices
with specific duties. The key unit of control was the administrative
township (lid), nominally composed of one hundred families.
Of the five major types of village officers selected in the early Sung,
some were concerned most especially with tax collection, some with law
enforcement, and some with both duties. The key figures in the T'ang
were the township leaders (li-cheng). During most of the Sung, their
function in most places dealt with general administration and taxation,
though a twelfth-century book of advice to local officials indicates that in

49 Liao Hsing-chih, Sheng-chai chiy SKCSCP ed., 5.18a.

50 Authors use terminology that varies rather widely. For example, the handbook of
advice of magistrates called the Chou-hsien t'i-kang states that "the responsibility for
pursuit and arrest often falls on the township leaders [li-cheng]." See Gh'en Hsiang,
Chou-hsien t'i-kang, TSCG ed. (hereafter cited as CHTK) 2.19. This term li-cheng is
derived from the T'ang and was used early in the Sung but was in disuse in most places
at the time when the Chou-hsien t'i-kang was written.
133
 Law and order in Sung China

some parts of China they were heavily involved in law-enforcement

activities:
With regard to pursuit and arrest, the district officials often assign the respon-
sibility to the township leaders. If a township leader exceeds the first time limit
for arresting a fugitive, do not be hasty in beating him. Now suppose this is an
urgent matter of pursuing someone and because a township leader has broken
the first time limit offivedays, you have hastily beaten him. When again he does
not make the arrest during the second time limit of five days, if you do not beat
him a second time, he will not pay attention to your orders, and so the previous
beating will have been meaningless. As to beating him a second time, the
wounds from thefirstfloggingwill not have healed. To beat him again would not
simply be contrary to law but would also cause excessive injury.51
Township leaders thus might be involved in law enforcement at some
times and places, but the more typical pattern was to make law enforce-
ment the particular responsibility of men called elders (ch'i-chang),
assisted by unofficial constables called stalwart men (chuang-ting). The
office of elder first appeared during the Later Chou (951-60) and con-
tinued in the Sung. 52 The elders were responsible for suppressing
bandits, settling minor disputes, fighting fires, caring for roads and
bridges, keeping logs of local affairs, managing message boards, seeing to
the treatment of men who fell ill on the roads, caring for signal beacons,
keeping weeds cut back at dangerous locations so that bandits could not
lurk there and, for the same reason, blocking up empty unused kilns, and
encouraging agriculture and handicrafts. They took part in revising the
five-grade population registers on which, after 1022, rural households
were ranked according to wealth, and in creating local maps, which
included information on topography, boundaries, hamlets, temples,
monasteries, shrines, pavilions, wineshops, river crossings, postal sta-
tions, remains of ancient buildings, guest houses, and the locations of
police forces.53 In short, the elders had very broad responsibilities, but
most of these were related to maintaining local security.
Early in the dynasty these elders were chosen from first- or second-
51 CHTK 2.19. This use of township leaders may reflect the continuing influence in some
areas of an order from the early 1070s that abolished the post of elder and transferred
the duties of that post to the township leaders and their assistants (see HCP 257.8a).
This is also reported by Ch'en Fu-liang, 21.3a; Chih-chai wen-chi, SPTK ed. and in Chen
Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, Sung-Yuan ti-fang chih ts'ung-shu ed. (hereafter cited
asSYTFCTS), 17.8a.
52 Sudo Yoshiyuki, To-So shakai keizaishi kenkyu (Studies on the economic history of T'ang
and Sung China) (Tokyo: Daigaku shuppan kai, 1965), p. 570: Sogabe Shizuo, Sodai
zaisei shi (Sung financial history) (Tokyo: Dai Nippon insatsu kobinshiki kaisha, 1941),
p. 135; on Wu-tai hui-yao, chap. 25.
53 See Sogabe, Sodai zaisei shi, pp. 136-37. See also Sudo, To-So keizai shi kenkyu, p. 588;
SHY, shih-huo 14.35b.

134
 Informal and semiformal law enforcement

grade households (in the nine-grade system of household registration

used during the opening decades of the Northern Sung) or from house-
holds classed as "official households." First- and second-grade house-
holds would have been among the wealthiest local families. Official
households were also important parts of the local elite, because they
were households nominally headed by an official or ex-official. Only if
people in these categories were insufficient in numbers were men drafted
from third-grade households. They reportedly served one-year terms. 54
The assistants to the elders, called stalwart men, were chosen from
fourth- or fifth-grade households, households of moderate means. 55
According to Ch'en Fu-liang (i 137-1203), during the late eleventh
century these men were selected from households with two or more adult
males and served in rotation for terms of half a year.56 These stalwart
men helped the elders by running messages to and from the district,
acting as their constables, escorting people to the district when neces-
sary, subpoenaing witnesses in cases of fights, and otherwise supporting
the elders in their law-enforcement duties. The stalwart men were thus
the actual agents responsible for carrying out the orders sent from the
districts to the elders. Serving in this way, the stalwart men were also
used by elders as guards to escort participants in legal matters to the
district town. These duties were managed through a regular system of
documentary control, including a log of documents and activities. When,
for example, an elder assigned a stalwart man to escort someone to the
district seat, the date was marked on the escort document, which was
then sealed and clamped shut to prevent abuses. 57 Village officers were
also charged with escorting to the district town some men involved with
the law.58
Generally, these men were not well trained or armed, though there are
examples of groups armed with swords and bows who were to be actively
concerned with seizing bandits. 59 For the most part, however, in most
times and places, although such men might have been capable of dealing
with a few troublemakers, for more serious problems they had to ask for
outside help. Because it was the rule that if the criminals were too
strong, the elders were immediately to inform the authorities, it appears

54 Ch'en Fu-liang, Chih-chai wen-chi, 21.2a; Li Yuan-pi, Tso-i tzu-chen, SPTK ed. (hereafter
cited as TITC) 7.35; Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, 17.8a.
55 Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, 17.8a. For more about such men, see Sudo,
T0-S6 shakai keizaishi kenkyu, p. 636.
56 Ch'en Fu-liang, Chih-chai wen-chi, 21.2a.
57 TITC, 7.35a.
58 Sudo, To-So shakai keizaishi kenkyu, p. 637.
59 SHY, ping 2.43a.

135
 Law and order in Sung China

that the primary role of such men was often simply to report to higher
police authorities, who would arrest the suspects. 60
This early Sung system continued with only minor changes until the
reign of Shen-tsung (r. 1068-85), w n e n it was profoundly affected by the
move to pay village officers and even more by the impact of the pao-chia
militia reforms that divided the populace into a nested hierarchy of small
groups. Nevertheless, despite striking changes in some of the details of
the system, its functions still had to be performed, and analogues of these
early Sung officers continued to exist.

Militia organizations
Sung law and order also depended on militia organizations, some of
which were closely tied to law-enforcement agencies and some of which
were largely independent. Such organizations involved far more men
than the village officer organizations. 61 They thus provided the state
with a compromise solution to the problem of maintaining order. Mem-
bers were usually armed (though not necessarily heavily armed), had
some training, worked under leaders with a modicum of experience, were
close to the people, and cost little. At their most martial, they blended
with the Sung regular armies; at their least martial, they were in effect
small-group mutual security organizations whose responsibilities were
very like those of the village officers, providing information to the auth-
orities while not necessarily being effective at law enforcement. In
describing one such organization the Southern Sung official Liao Hsing-
chih (1137-89) commented:
The duty of the sheriff is to control evildoers.... In doing this, nothing has
priority over using the pao-wu [guard] units. . .. Members of the same guard unit
support one another. Because they live together, it is easier for them to inves-
tigate. Also, the leaders of the superior guards are chosen from among trusted,
powerful men. The leaders of the large and small units are also vigorous men
from the multitude. The sheriffs and patroling inspectors are responsible for
organizing them, for seizing the guilty and establishing rewards, and issuing
rules. When there are wanderers, the small-unit leaders inform the large-unit
leaders, and then the leaders of the superior guards tell the officials.62
60 C Y K C 19; Brian E. McKnight, Village and Bureaucracy in Southern Sung China (Chicago:
University of Chicago Press, 1971), p. 44. Huang Kan, Mien chai chi, SKCSCP 1971,
25.27a, remarks that the village officers' jobs were simple in practice. If the troubles
were serious, they just called on the sheriff. Some documents make it clear that
sometimes the local people dealt directly with the constables called archers and the
clerks without the intervention of the sheriff, but it is not entirely clear how this worked
(see HCP 97.11b).
61 HCP 307.18a.
62 Liao Hsing-chih, Sheng-chai chi, 5.15b-17a.

136
 Informal and semiformal law enforcement

During the first century of the Sung, there were more than twenty
different militias scattered throughout China, although the majority were
along the northern and western frontiers, regions of China with large
unassimilated groups of minority peoples. The militiamen were usually
farmers, who served in the off-season. For example, during the second
decade of the eleventh century, in some of the northern border prefec-
tures men were chosen from among the adult males to serve in units that
guarded the river. Along some nine hundred lid (about three hundred
miles) there were 26 forts and 126 post stations. Men patroled in rota-
tion and in the agricultural off-season received rations while serving
under a patroling inspector. 63
Some of these militias were established by the local people without
official encouragement. Others were established with various sorts of
state help, including remuneration and the use of state population
documents to provide lists of eligible men. Such organizations of local
people also were customary in parts of south China, such as Kuang-nan,
where traditional organizations of adult men existed to preserve law and
order. 64 Such militias played critical roles in periods of unrest such as
the 1040s and 1120s.

Pao-chia militia
T h e most famous Sung militia organization was the pao-chia system
m a n d a t e d as a part of the new laws of W a n g An-shih during the reign of
Shen-tsung (r. 1067-85). T h e basic motive for creating the pao-chia was
to reduce state expenditures. Even before the reign of Shen-tsung, the
central government was unable to meet current expenses from current
income. In 1064, under the reign of Shen-tsung's predecessor Ying-
tsung, the statesman H a n C h ' i wrote a long memorial advocating greater
reliance on a militia system. Citing precedents back to antiquity and
more recently from the T ' a n g fu-ping system, as well as the Sung's use of
archers during the Hsi-hsia wars of the 1040s, he suggested that a militia
could increase the n u m b e r of potential soldiers while decreasing the
state's costs. 6 5
T h e most immediate impetus for founding the pao-chia^ however, was a
report submitted in 1069 an official who had been ordered to survey
conditions in the capital region. T h e report suggested that in order to
strengthen security the villagers be organized into groups of ten. H e
commented that in recent years there had been an upsurge of banditry.

63 HCP 8o.8a-b.
64 HCP 254.10a, 246.1 i a - b .
65 HCP 203.4b; see also HCP 203.5a for Ssu-ma Kuang on this issue.

137
 Law and order in Sung China

Although the villages had certain servicemen such as the elders, stalwart
men, and other village officers, their forces were too weak to deal with
the new threat. A mutual security system called pao-chia had existed in
the past, and he wished to revive it.
In the following year an edict ordered that the villagers in this area be
organized into groups. Ten families were to form a guard {pao), led by
an adult male from the resident household serving as the guard head.
Five guards (fifty households) would form a large guard, headed by a
large guard head. Ten such large guards (five hundred households)
would form a superior guard, with a leader and an assistant leader. Each
family with two or more adult males was to have one serve as a guard
member. These men were not to be armed with certain weapons used by
the regular military, though they could have bows and arrows. Under
the pao-chia system, a substantial number of people who were supposedly
concerned with maintaining law and order were scattered throughout
the rural areas of China. In the early Southern Sung a prefect remarked
that in his prefecture in each riding there were about twenty guard
leaders and their assistants and some two hundred large and small guard
heads. 66 These organizations were set up simply to serve as security
units. From each large guard five men might be selected to serve in
rotation as patroling police. In some areas, they were expected to estab-
lish a headquarters, beat the drum to announce the hours, and patrol at
night. 67 When bandits appeared, the drum was to be beaten, and the
men in the units were to assemble and give chase. If the bandits fled to
another guard, the initially involved guard members were to cooperate
in catching them. 68
The authorities again recognized the importance of creating positive
incentives for effective performance. Pao-chia members who captured
criminals were to be rewarded. Criminals guilty of crimes calling for
penal servitude or worse were worth three strings of cash. Those guilty of
crimes calling for beating with the heavy rod were worth one string, the
rewards to be paid from the property of the criminals. The pao-chia
members or their families were also compensated for injuries or deaths in
the line of duty. For any injury a member was to be paid five strings of
cash; for serious injuries, ten strings; for a disabling injury, twenty
strings; and the families of pao chia personnel who were killed in the line
of duty were given fifty strings of cash and forgiven three years of taxes.69

66 SHY, shih-huo 14.25b.

67 HCP 218.6a, 343.3b. There were areas where they did not patrol at night (see HCP
246.6a).
68 HCP 246.6a; see also SHY, ping 1.27b.
69 SHY, ping 2.8a.
138
 Informal and semiformal law enforcement

The guard also served as a device for securing internal information. If

the people in a guard knew of fellow members who were guilty of
robbery, theft, homicide, arson, rape, kidnapping, heresy, or the making
of black magic (ku) poison but did not report them, then the guard
members would be punished according to already-existing laws on small
groups. 70 Even if the guard members claimed a lack of knowledge of
such culprits, if three or more criminals were involved and they had
remained in the guard for three or more days, the guard members were
still to be punished for having been negligent in their investigation. The
guard also served as a device for generating the kind of information
necessary to control the people. If in a guard a family died out or fled,
the guard was responsible for reporting this to the proper authorities. 71
Thus it is important to recognize that the original intent of the man who
suggested the pao-chia system was not to create a true militia organiza-
tion but rather an internal security system. A guard was to serve as a
source of information and a system for policing its own area, not a type
of military unit. But in any society, changes in the distribution of control
over armed forces cannot be separated from politics. Emperor Shen-
tsung, who had inherited an empire with extremely serious fiscal prob-
lems, was seeking to reduce government expenditures. At the same time
he also sought ways to reincorporate territories under foreign control
that Sung leaders thought of as rightfully belonging to the Sung. Much
of the politics of his reign must be understood as revolving around
attempts to arrive at policies that would relieve fiscal pressures and
restore Sung military might. The pao-chia was suggested at a time when
Emperor Shen-tsung was deeply concerned about foreign military prob-
lems. His chief adviser, the great reformist statesman Wang An-shih,
though less bellicose than his imperial patron, was greatly concerned
about the extraordinary strain that the military establishment placed on
the state's revenues. The proposed pao-chia system appealed initially to
the emperor and then to Wang An-shih as a way out of their dilemma.
From the beginning they seem to have envisioned the transformation of
this internal security device into a militiary system, a system that could
maintain or even increase their available military forces while sharply
cutting expenditures. Both men had been thinking about the possibilities
of a militia system even before the pao-chia was proposed and so were
enthusiastic about the possibilities. By simply arming the men of the
guard units and giving them military training, the state could create, at
relatively little expense, a vast reservoir of potential soldiers. In 1071 the

70 HCP 2664b. 71 HCP2i8.6afT.

139
 Law and order in Sung China
pao-chia members in the capital area began to receive military training
in mounted and foot archery. Later this same year this system was
extended to certain other circuits along the northern frontiers. The
trained personnel were expected to "volunteer" for ten-day periods of
service under their local patroling inspector. During this time they were
to receive state rations and pay.
This training was vital if such men were to be successful against
foreign soldiers or even against criminals, since many criminals were
military deserters. The men chiefly responsible for training the local
military groups were the patroling inspectors, 72 though they were at
times aided or replaced by sheriffs,73 who might also take part in actual
military actions. 74 In 1073, when the Court wished to assist certain
prefectures in establishing the pao-chia militia system, it commissioned
two former sheriffs to help the resident officials "in arranging the pao-chia
system in accordance with the rules used in the Kaifeng Prefecture
region." 75 Sheriffs might also be sent to inspect the working of the pao-
chia system in remote border areas. 76
The advent of the pao-chia militia system affected various facets of the
law-enforcement system and changed the character of the troops serving
under the key local military officials involved in law enforcement, the
patroling inspectors. In some areas at least, pao-chia militiamen were sent
to serve under patroling inspectors, taking the places of the regular
troops, who returned to their camps, 77 though the poorer pao-chia mem-
bers were not supposed to be rotated for service or to be forced to
undergo teaching and inspection. 78
Here again the state's desire to lower its expenses is clear. If pro-
fessional troops could be freed for regular military duties and replaced
by relatively low-cost pao-chia personnel, the military position of the Sung
could be strengthened at little expense. The patroling inspectors were
apparently allowed to retain a small complement of regular troops,
perhaps because the state recognized their need for a personal guard and
also because it recognized that the pao-chia members were simply not as
well trained as the regular troops. One report from Ho-pei West Circuit

72 HCP 84.12b (1015), 237.12b, 26.14b.

73 See HCP 274.12b for the instructions of tribesmen during the slack agricultural season
by the patroling inspectors and the sheriffs. See also SHY chih-kuan 48.67a ( n 16),
48.7ob~7ia (1141); HCP 309.12b (1080), 330.7b (1082).
74 HCP 284.8a.
75 HCP 235.13b.
76 SHY, ping 2.36b.
77 HCP 260.4a, 278.6a-b, 402.14b-16a.
78 HCP 281.14a.

140
 Informal and semiformal law enforcement

in 1076 indicates that the patroling inspectors there could keep twenty
infantry and ten cavalry. 79
Pao-chia militiamen were also sometimes sent to take the places of
some of the sheriffs and archers. In the late 1070s the number of archers
in many areas was ordered reduced, with a small number, about 15 to
30, being left to serve as the sheriffs' personal attendants in the district.
This was a reduction from the 70 to 140 used previously. 80 In 1076 in
Ho-pei and Ho-tung circuits the pao-chia personnel were rotated for
fifteen-day tours to take the places of the released men. 81 Again, this
seems to have been an economy measure. Still, in many areas archers
and pao-chia coexisted.82 By a decree of 1085 in the five northern border
circuits the sheriffs were to retain approximately one-third of their pre-
vious quotas of archers. The places of those dismissed were to be rilled
by pao-chia personnel rotated for service. 83 However, an order of the same
year also called for the reestablishment of archers and troops according
to the pre-pao-chia system, in the circuits around the capital. 84 Perhaps
the overriding importance of security in the capital area convinced the
officials that the money saved was not sufficient to justify using a group
with lesser military skills.
The pao-chia heightened in the minds of some the old specter of the
people in arms. The Sung authorities had grave doubts about how far
the people, including the locally influential elites, identified with the
interest of the state. In Kuang-nan, pao-chia personnel were not allowed
to have swords or spears. During training or when chasing bandits, they
would receive weapons from the officials.85 It was necessary for such
men to be adequately armed, but armed men were always seen as a two-
edged sword. Ssu-ma Kuang spoke of "bandits among the pao-chia who
take advantage of the pao-chia and the horse-raising system in order to
pillage." 86 Presumably Ssu-ma Kuang was worried by the kinds of
organizational connections formed within the militia and among the
peasant families who raised horses for the state. He was especially
concerned about the large numbers of men involved:

79 HGP 277.15b.
80 HCP 279-i4a-b.
81 HCP 274.12b, 279-i4a-b, 329.13b, 267.5a.
82 HCP 327.10a.
83 SHY, ping 2.12a.
84 HCP 356.8b. From this time on the militia functions of the pao-chia became less
important than their functions as local servicemen; see also HCP 364.2a. There is
detailed information on the pao-chia in Hopei Circuit in 1083; HCP 337.6a-ga. The
funds that had been used to pay the soldiers were diverted to other uses; HCP 276.9a.
85 HCP 327.11b (1082).
86 HCP 355.3a. Under the horse-raising system the state provided horses to commoner
families, who became responsible for raising them.

141
 Law and order in Sung China

I have noted that although it is said that the soldiers came out from the people in
ancient times, in antiquity out of eight hundred families there were only three
men who served as armored knights and seventy-two as foot soldiers. . . . Now
the villagers are registered, and one man out of two is inducted into the pao-chia.
Each man is trained with weapons, making half of the adult men of peasant
families into soldiers. . . . If these men are for catching bandits and guarding
villages, why do we need so many? 87

It was not simply that the pao-chia rank and file might get out of hand
but that they might be used by their betters. When discussing the causes
of disorders, the official Feng Ching (1021-94) argued that the pao-chia
would be dominated by locally powerful families. This danger, however,
had to be weighed against the dangers of not having the people organ-
ized into mutual security groups. W a n g An-shih disagreed with Feng
Ching, saying that it was when the people were not organized that
problems arose and that when the people were organized, such problems
did not arise. 8 8 T h e problem, however, had not been solved. Chu Hsi, in
describing Sung local feuding, cited the example of two locally powerful
individuals using pao-chia militia units to fight each other. 8 9 This situa-
tion exemplifies the dilemma faced by officials who were troubled by the
arming of the people. How far could the people be trusted? T h e late
Northern Sung official Liu Chih spoke in 1086 about the problem posed
by archers who, being mustered out of service and finding no employ-
ment, had become bandits. 9 0
T h e same concern is reflected in Sung policies regarding the theft of
weapons and regarding discipline within the militia. T h e authorities felt
compelled to institute very severe penalties for such abuses. T h e concern
about internal discipline is reflected in the rule that pao-chia soldiers who
scuffled with their superiors were to be punished as if for fighting (tou*),
with the penalty increased by two degrees. 9 1 Stealing weapons was
even more severely punished. T h e penalty for forcibly stealing pao-chia
weapons was death. However, if the circumstances were not serious, the
case could be memorialized. For the theft of twenty arrows, the penalty
was one year of penal servitude; for a bow, two years; and for a cross-
bow, exile to three hundred li.d (Those subjected to penal servitude
were registered at five hundred li.d Exile meant registration for penal

87 HCP 355.7a-8a.
88 HCP 230.4a.
89 Sogabe Shizuo traces the rise of Sung feuds in part to the pao-chia reforms. See Sogabe
Shizue, "Sodai no ketsu-kan" (Concerning feuds in the Sung dynasty), in Ishida
Hakushi koki kinen jigyokai, ed., Ishida Hakushi shoju kinen Toyoshi ronso (Tokyo: Ishida
Hakushi koki kinen jigyokai, 1965), pp. 311-18.
90 HCP 389.6b.
91 H C P 2 3 7 . i 2 a - b .

142
 Informal and semiformal law enforcement

servitude at one thousand li.d)92 The role of the pao-chia as a militia

organization declined rapidly after the 1070s (though it was occasionally
revived during times of crisis like the late 1120s and the 1130s). 93 In
some areas pao-chia personnel became more purely village service officers
and neglected their role as local security forces. The leaders of the
pao-chia units became responsible for various local services, especially
tax collection. However, they might also be in charge of investigating
banditry, a law-enforcement function they sometimes shared with the
elders of their area. 94
Where the pao-chia no longer performed military functions, their places
were taken by other organizations, with structures and functions rather
like the older pao-chia's, but with different names. The pao-wu militia,
which though overshadowed by the pao-chia, had existed during the
Northern Sung, again became important during the Southern Sung. 95 In
other parts of the empire, organizations with similar functions but dif-
ferent names were established. A document from 1127, for example,
indicates that in both rural and urban areas there were organizations
called patrol societies (hsiin-she), small organized groups of men who
were excused from serving in the pao-chia itself, who acted as mutual
guarantors of one another's behavior and investigated banditry. 96
Documents from another time of troubles, the 1160s, show the estab-
lishing along the rivers and coasts of institutions, under a variety of
names, which functioned like the old pao-chia. The authorities wanted to
enroll all the coastal folk, as they were the ones who provided the pirates
with their manpower. If a man refused to join such a mutual security
group, he was to be taken to the officials, who would punish him. 97
Such variant forms of a general type of organization continued to exist
until the end of the dynasty, since they fulfilled a real social need. Huang
Kan describes a typical late Southern Sung system from Han-yang in
Ching-hu North Circuit under which five families formed a small group
(chia), five small groups a large group (ta chia), and four large groups a
company (t'uan). Companies were joined together to form administrative
villages (lid). Villages were grouped in turn into ridings (hsiang), with a

92 HCP 322.3b (1082).

93 Pao-chia acting as militia are noted frequently in annalistic sources on the early
Southern Sung. See, for example, HNYL 204, 273, 494, 1101, 1621.
94 SHY, shih-huo 14.35b (1159); WHTK 13.138, 139.
95 See CHTK 2.23 for advice on creating pao-wu registers and on structuring such a
militia. The author comments that the system was useful in case of bandits, if labor
were needed, or in case of natural disasters.
96 SHY, ping 2.50a fT. See HNYL 2766 for a report on the pao-wu's being responsible for
reporting illegal wine sales and production.
97 SHY, ping 2.43a.

143
 Law and order in Sung China

border officer (yu-kuan) charged with investigating wrongdoing and

arresting those involved. 98 Thus in this region, for every few hundred
families, there was a functionary concerned with law enforcement.
Whether or not such organizations functioned effectively depended on
the energy and initiative of the local officials. They often had to revive
village officer organizations that had fallen into disuse. Yuan Fu (fl.
1237-40) reported on his own service in Ch'u Prefecture in Liang-che
Circuit:

I have heard it said that the pao-chia system is that whereby to unite the villages
and bring the people closer to one another. It embodies the ancient intent of the
neighborhood groupings and the kin associations (pi-lu tsu-tang) by promoting
mutual aid. In recent years we have had the name of pao-chia but have been
without the organizational reality. Although men live as neighbors, their hearts
are scattered. . . . I have examined the situation in this prefecture. . . . Recently I
took a careful look at the problem of fire control. I gathered together officials in
the office and discussed it. Someone asked, "If we want to bring together the
neighborhoods we already have the law, but how can this be done without
people?" Subsequently I examined the riding (hsiang) records and selected a
group, asking them to serve as riding officials and making them leaders. For each
fort district there were three or four, or five or six men. . . . I examined what had
been done according to the Offices of Chou, which described the practice of
holding a monthly meeting, and gathering the people to read the law. Sometimes
they read about the virtuous practices, or the practices of the Way, sometimes
about being filial and so on. Now I wanted to imitate this goal, and so I set up a
register, called the riding record. Within the villages, when admirable things
were done that really were worth reporting, the neighborhood groups were to
report them, and the riding officials were to certify them and report them to the
prefecture, which would investigate the facts and enter them in the riding record.
And for all wrongdoing, like gambling, fighting, killing or injuring, or banditry,
the officials were to be at all times concerned with what went on in their
jurisdictions, warning men to be peaceful and not to break the law. This also was
to be entered in the riding record. At the end of each month we would examine
what had been written in order to assess the leadership of the riding officials.
Those who had good things entered and no bad things were the top class. Those
with no good things but also no bad things were the second class. These two
degrees were praiseworthy. This prefecture has separately set aside some official
funds . . . in what is called the riding officials' reward money, so that the people
will feel pleased and will mutually encourage one another and customs will
gradually be improved."

98 Huang Kan, Mien-chai chi, 24.11b. There were also a variety of other militia
organizations at this time, perhaps the most widespread being the organization called
pao-wu. See Chang Shou, P'i-ling chi, 3.35; Hsiieh Ghi-hsiian, Lang-yu chi, 15.6b-7b.
99 Yuan Fu, Meng-chai chi, 3.32 fT.

144
 Informal and semiformal law enforcement

As Yuan Fu's report implies, the people on the spot were the ones who
had the greatest knowledge of both good and bad behavior and could
respond to it most quickly, but they were also - from the point of view of
armaments and training - far weaker than full-time police or soldiers.
The paradigm for the inverse relationship between the intensity of cover-
age and the level of power is reflected in relative numbers. A report from
the late 1070s shows that there were many hundreds of village officers
concerned with law and order scattered throughout each district, but
only a few professional police officials (the sheriffs and the patroling
inspectors) with their few hundreds of subordinate soldiers and archers
concentrated in the district towns and a few rural forts, 100 and even
these men might not really be doing police work. Thus militias and
many village officers, both the pao-chia and others, had to be principally
devoted to maintaining order. As Chang Shou wrote in 1138:
From the time of the dynastic ancestors in the systems of capturing bandits,
below we had the pao-wu [militia] and above the sheriffs and patroling inspec-
tors. When someone committed a robbery, the responsibility fell on the pao-wu
militia. If the robber was not caught, the blame fell on the sheriffs and patroling
inspectors.101 *

Conclusion
The informal and semiformal levels clearly reflect some of the dilemmas
of law enforcement. As in many bureaucratic structures, those with a
knowledge of the problems often lacked the power to deal with them, and
those with the power lacked the knowledge. The people with the most
timely information on crimes - witnesses, bystanders, and neighbors -
who also were in a position to respond immediately almost always were
not themselves members of the formal law-enforcement agencies. The
arms they possessed and their skill in using them were usually inferior to
the arms and skills of the regular formal agents. Despite these weak-
nesses, traditional Chinese states nonetheless had to rely on informal
and semiformal agents. Lacking rapid systems of communication and
transportation, the formal authorities were almost never in a position to
respond sufficiently quickly to be of help with an immediate problem.
They almost always had to follow a cold trail.
Even if they had been able to solve the problems of communication
and transportation, the authorities would still have faced fiscal and
political problems that would have ruled out creating a body of armed
state functionaries large enough to maintain order everywhere. The state

100 HCP 307.2a. 101 Chang Shou, P'i-ling chi, 3.37.

145
 Law and order in Sung China

simply could not pay for such a large force and in any case would no
doubt have had grave reservations about creating an armed force of the
required size. Traditional Chinese governments were also unwilling to
see the people themselves sufficiently well armed and trained for full self-
defense. For good reason, Chinese regimes were deeply suspicious of
local centers of power, especially the people in arms.
Chinese dynasties were forced to try to walk a narrow line. They had
to permit - indeed, they were probably not in a position to forbid
completely - the arming of the people to a certain level. A completely
disarmed populace would be easy prey for the sort of men who might
fatten themselves on the helpless until they became a danger to the state.
However, the central authorities also could not allow the people to
become so well armed and trained that they were no longer under some
state control. Chinese states early devised a compromise solution to this
conundrum. They selected (or certified the selection after the fact) sub-
sets of the local population that were allowed, encouraged, and even
mandated to learn the martial arts and bear arms. These groups would
then be used by the state as its semiformal agents to maintain order.
These agents, village officers and militiamen, were important to the
state's overall law-enforcement strategy. Of course, such groups them-
selves, especially the more heavily armed and numerous militiamen,
could pose a threat to state control. Most of the dynasties tolerated the
militias in times of unrest but distrusted them otherwise. The Sung is
unusual in its attempt, during a time of peace, to create an empirewide
militia system, the pao-chia, which was to be armed and trained by the
state. This Sung pao-chia system soon evolved so that at most times in
most regions its members served as village officers rather than as a
militia. Still, the authorities' willingness to try to create this system
demonstrates their underlying confidence in the stability of the Sung
state. However, even when they were working well and as intended, such
informal and semiformal systems usually only supplemented and did not
substitute for the state's formal power. When they became so powerful
that for extended periods they became the dispensers of law and order,
the state itself was endangered from within. This did not happen during
the Sung.

146
 5
Formal civil agencies of
law enforcement

Role of the sheriff

Although they were clearly important to the general processes of keeping
order, the village officers and militia personnel were hardly powerful
enough to handle serious disorders by themselves, nor were they allowed
to fetter people or hold them in detention. 1 The superiority of their
sources of information and the advantage of being close to the scenes of
the crimes were offset by their inadequate arms, training, and leader-
ship. Therefore, above them the formal state apparatus took over the
problem, using both civil officers and the military. A wide variety of
officials and officers had some responsibilities for law and order during
the Sung, the two most important field-level offices being the (civil)
district sheriff (hsien-wei) supported by a group of several dozens of
civilian constables called archers (kung-shou) and the (military) patroling
inspector (hsiin-chien) with perhaps one hundred regular soldiers under
his command; on the supervisory level most tasks were eventually the
concern of the circuit intendants. Though such formal agencies could not
be spread as widely as village officers, the authorities did station them
at critical points where their very presence might deter lawbreaking.
Sheriffs were usually stationed at the towns that were the district seats
and sometimes at other towns. The patroling inspectors were mostly
commonly stationed in forts located at dangerous and strategic places in
the rural areas. Both sorts of officers on occasion were headquartered at
other population centers such as fords, bridges, and markets. Finally,
small complements of soldiers with law-enforcement duties were stationed
at the postal stations that dotted the transportation routes.
Some idea of the ways in which local civil officials might respond to
reports of lawbreaking can be found in the advice given in a twelfth-
century handbook written to help local officials:
1 HCP 477.ib-2a. TFSL 537 says that if powerful families made prison instruments and
detained people, they could receive a penalty of two years of penal servitude.

147
 Law and order in Sung China

If you have cases that involved such unexpected things as fights, killings or
injuries, floods or fires, or banditry, they should be handled one by one and not
delayed. As noted previously, you should install a gong outside the district gate.
Without regard to the time of day, when the gong is struck, you should have the
person questioned and act on his complaint. If the matter is urgent and help is
needed, provide that help. If criminals can be captured, capture them. If there is
need for an inquest, go to conduct it so that the affairs will not be unsuccessfully
handled because you have tried to deal with them following some sort of rigid
order. This is the second key thing about administration. 2

The principal officials concerned with law enforcement were the dis-
trict sheriffs with their complement of archers armed with bows and
swords. When the founder of the Sung began to reemphasize the role of
the regular civil system early in his reign, he chose to revive this earlier
office. In 962, on the advice of the head of the Bureau of Military Affairs,
Chao P'u, he ordered the establishment of a district sheriff (hsien-wei)
in each district. 3 These officials received a salary equal to that given
their colleagues, the recorders (chu-pu), but stood below them in the
protocol order. Assisted by a body of clerks (chieh-chi) and constables
called archers (kung-shou), the sheriffs were to maintain order in the
district, including its rural areas. 4 The garrison troop commanders,
who had been the key law-enforcement officials during the preced-
ing Five Dynasties period, retained control only over disorders in the
garrison's environs. 5 The sheriffs were usually stationed in the district
towns. Their offices, although ordinarily located in the towns and often

2 Hu T'ai-ch'u, Chou-lien hsu-lun, TSCC ed., pp. 7 - 8 .

3 SHY, ping 11.1a. The post of district sheriff first appeared under the Ch'in (see Pan Ku,
Han shu, 19 shang). Note that although the general Sung practice was to appoint one
sheriff per district, some districts had more than one sheriff, most of the cases being from
the Southern Sung. SHY, chih-kuan 48.66b (1114), 48.79b (1181), 48.81b (1192),
48.8ib-82a (1201), 48.85b (1214). Because all of these instances are from the twelfth
and thirteenth centuries, it seems that this practice was a late development, although it
also is possible that it occasionally occurred earlier and that the references to it have
been lost. There are also cases of districts without sheriffs; see HCP 243.10a.
Wu Ching explained the character of this position by referring to the Shuo-wen, where
wei is defined as weih (terrifying). The word, he said, is composed of ts'un, which
expresses the need to assess so as to show respect for superiors, shih,c which refers to the
need to make instructional orders clear so as to guide inferiors, and shih,d which refers to
defining the office so that it could be established; see Wu Ching, Chu-chou chi, 11.8a.
4 Initially clerks and archers were transferred to their control from the staffs of the
garrison troop commanders; see SHY, chih-kuan 48.60a-b. The sheriffs general
responsibility for maintaining the local peace is suggested by his responsibility for such
related tasks as trying to prevent Chinese from causing trouble among the minorities; see
an order dated 1090 to the effect that sheriffs were to prevent skilled Chinese from
having contact with the Man tribes (HCP 442.6b).
5 HCP 3.14a. A few years later the tie between the military commissioners and the
garrison troop commanders was broken when the regional commandants were
forbidden to appoint their followers to these posts. Such posts were to be filled only by

148
 Formal civil agencies of law enforcement
within the headquarters itself, were frequently physically separated from
the offices of the district magistrate. This independence may have con-
tributed to the tendency of some sheriffs to act in a high-handed manner
toward prisoners. 6
This problem reflects the inverse relationship that existed in tradi-
tional Chinese government between the level of socialization into civilian
values and the possession of the authority to use physical force. The
closer the formal governmental agents were to the criminal incident, the
larger their numbers and the lower their level of civil socialization were
likely to be. The sheriff's position was a compromise one. By origin the
position of sheriff (wei) was a military office, a fact of which Sung
officials were well aware. 7 The ambiguity of Sung law enforcement, with
its mixture of civil and military agencies, is emphasized by this use of an
originally military office as a civil office. This ambiguity became even
more pronounced in the late Northern and the Southern Sung, when
some of the men chosen for this post had military connections, but the
post itself was part of the civil hierarchy.
Throughout the period, these sheriffs were the main civil officials in
law enforcement, though their civil superiors, the magistrates, prefects,
administrators for public order (ssu-li ts'an-chun),8 and some circuit
officials also were involved in supervision and, at times, in actually enforc-
ing order. 9 The primary role of the sheriff is outlined in an order issued
men from the prefectural staffs (SHY chih-kuan 48.92a). We do have reports, however,
that the commanders continued to seize "village bandits" and to hear complaints (HCP
61.2b) and, at least as late as 1013, to try cases and punish offenders (HCP 80.9b). See
also SHY, chih-kuan 48.92a.
6 See the maps in Ch'en Kung-liang, Yen-chou t'u-ching, TSCC ed., pp. 12, 14, 15, 16, 17,
136, 198-99. As these materials indicate, the sheriffs' offices were sometimes outside the
headquarters area. For instance, in Yen Prefecture itself, "the sheriffs office was outside
the Shen-li Gate" (Ch'en Kung-liang, Yen-chou t'u-ching, p. 5). This places it just outside
the town wall. We also have examples of sheriffs' offices located far from the district
towns (SHY, chih-kuan 48.69b, 756b, 80a). Hsiieh Chi-hsiian, Ken chai hsien-sheng Hsu'eh
Ch'ang-chou lang-yu chi (Yung-chia ts'ung-shu ed.), 20.6b, 26.ia-b, gives a remarkably
full picture of the process of petitioning for the establishment of such an office, since the
two documents just cited refer to the same example.
7 See, for example, the comments of Ma T'ing-luan, Pi-wu wan-fang chi, SKCSCP 1975,
17.8a.
8 Compare WHTK 166.1444.
9 See, for example, an order of 1088 that controllers-general in Ho-pei Circuit were to be
allowed to select a dozen imperial soldiers to go out to arrest bandits as necessary, and in
1098 a similar order allowed the controllers-general to send out thirty armed imperial
soldiers to capture bandits (HCP 417.b, 497.19a). One official even suggested that
bandits could not be suppressed because the system of rewards and penalties focused
too heavily on the sheriffs and patroling inspectors and did not cover higher local
officials like district magistrates. He maintained that because the district magistrates had
broader powers, they could plan campaigns and try to win back criminals through
education. However, his plea itself demonstrates the predominant roles of the sheriffs
and patroling inspectors (see Su Sung, Su-wei wen-Kung chi, SKCSCP ed., 19.ia-b).

149
 Law and order in Sung China

in 962, only two years after the dynasty was founded: "In all cases of
banditry or murderous robbery, the district authorities are responsible
for immediately trying to capture the criminals. If the sheriff has already
gone out in pursuit of the criminals, then the responsibility for capturing
them falls on the magistrate." 10
From the revival of the position of sheriff in 962 until the end of the
dynasty in the thirteenth century, sheriffs overshadowed the other civil
service officials serving in local posts in the police aspects of maintaining
law and order. The initial edict of 962 decreed: "If there are bandits, the
sheriff in person is expected to lead his followers to capture them and to
transport them to the prefecture." 11 The sheriffs sometimes left routine
law-enforcement tasks to their subordinates, but the records repeatedly
take note of sheriffs who themselves fought against outlaws, like Shang
Chih-chung, the sheriff of An-hsi District in Ho-pei West Circuit, who in
1016 "fought bravely against bandits. He was himself wounded by an
arrow. When the bandits fled, Shang Chih-chung jumped onto his horse,
grabbed his bow and arrow, and pursued them for seventy lid to the
border to his district. He cut off two heads and returned." 12
If the bandits were in gangs too large for the sheriff to capture,
the state faced the possibility of serious civil disorder and not merely
robbery. The sheriff in such cases was "immediately to inform the
prefectural bandit-capturing servitors [tsu-tsei shih-ch'en]." (Servitors were
low-level officials in the military hierarchy.) These bandit-seizing servi-
tors were charged with suppressing bandits and would be responsible for
mobilizing the military for this task. 13 Though the specific position of
bandit-seizing servitor was used only sporadically during the Sung, the
principle that when the civilian authorities were unable to deal with
disorders they were immediately to call on the military for help remained
true. A decree of the 1080s further illuminates this chain of respon-
sibilities, saying that prefectures or districts that had ten or more robbers
and were unable to handle them could mobilize troops, beginning with
the troops within their own jurisdiction. If these were insufficient, they
then could call for outside help. 14
Thus, when major bandit gangs appeared, the sheriff was supposed to
solicit help from his military colleagues. After the increase in the number
of (military) patroling inspectors during the eleventh century, the sheriffs
were to inform the nearest patroling inspector, as well as the prefectural

10 SHY ping u.ib-2b.

11 SHY, chih-kuan 48.60a.
12 SHY, ping 11.7b.
13 SHY, chih-kuan 48.60a-b.
14 HCP 348.10a.
150
 Formal civil agencies of law enforcement
authorities, and ask for the dispatch of military personnel to aid in
capturing the criminals.15 The prefectural authorities and the patroling
inspectors were expected to send help to the sheriffs or, if their forces
were insufficient, to request assistance from the regular military. If they
did not aid the sheriffs, they might be punished.16

Patroling
Patroling was one of the key tactics used by the Sung to maintain law
and order in the empire. In effect, by moving police forces around the
territory of a jurisdiction, the authorities multiplied the potential for
deterring criminals while not having to increase the number of law-
enforcement officials. Conversely, the lack of such patroling might lead
to an increase in banditry. In 1072, archers were ordered not to patrol in
the areas along the northern border and were returned to military
camps. "As a result, the number of men on preventive patrol certainly
was reduced."17 Banditry soared, necessitating the ending of this pro-
hibition two years later.18
Patroling itself entailed some costs, and the potential for abuse of
the helpless population was always present when armed men entered the
countryside, but the returns were generally judged to outweigh the
possible problems. Patrol systems had existed from the beginning of
the dynasty and included, where appropriate, patrols at night and
patrols by boat.19 These patrols might be conducted by the village
officers,20 the militia, or the regular military. The military patroled
from early in the dynasty. In 1019 an official reported on a settlement
associated with the small rural market in Ping Prefecture of Ho-tung
Circuit, which had a military camp but only four provincial army soldiers
(hsiang-jen) assigned to patrol. He requested that the military intendant
select an official with five assigned soldiers to patrol every night.21
In areas that were patroled by units under regular military officers,
such as superior joint patroling inspectors (tu-t'ung hsiin-chien) or the
15 SHY, ping 11.2a. See also SHY, chih-kuan 48.6oa-b, 48.66a, 48.71b. I believe that
there were two types of reports involved. One was a regular report to the prefecture
noting any bandits active in the district, and the other was a specific report requesting
assistance from higher authorities.
16 HCP 348.10a, 386.3b. According to an order of 1084, the key number for bandits was
ten men. If the local authorities could not handle bandit gangs of ten or more, they
might call for troops (HCP 348.10a).
17 HCP 260.4a.
18 HCP 2354b-7b, 242.4a, 257.4b.
19 HCP44.i5a-b, 193.12b.
20 HCP 7.13b.
21 SHY, ping 3.2a-b.

151
 Law and order in Sung China

servitors acting as inspectors general for seizing bandits (chu-po tsu-tsei

shih-ch'en), the patrols were forbidden to stop at the district or garrison
towns for more than a few days. The district was to report the patrols'
arrival to the prefecture. When the district reported bandits, this was to
be marked on the log of the patroling inspectors general (chu-po) and
crossed off only when the bandits had been captured. 22 Although there
were such patrols by regular military units from regular posts, the
primary military figures in patroling, as in other aspects of law enforce-
ment, were the patroling inspectors, who regularly patroled from the
early eleventh century. 23
Followers of the sheriffs were also occasionally used for periodic
patrols from early in the dynasty, 24 but it was not until the 1070s that
sheriffs also took part regularly in such patrols. 25 Under policies that
began at this time, the sheriffs were expected to pay monthly visits to
each of the superior guards in their district, which would have taken
them regularly into the rural hamlets as well as the markets. 26 This
pattern of patrols was modified for a few years in the early 1080s, when
the sheriffs were ordered to deal only with problems in the markets and
district towns. In theory, this would have reduced their travels on tour,
but in practice, according to one observer, since the sheriffs were respon-
sible for the various small rural markets as well as for the tax collection
stations, they still had to leave the district towns hundreds of times each
year. 27 In rugged or large districts the patrol system greatly burdened
police officials. The authorities responded by increasing the numbers of
officials,28 or more systematically, by permitting the patroling inspectors
and sheriffs to divide up their jurisdictions - each might cover a part of
the required total area, or they might take turns touring. 29
Between 1111 and 1117a system was established under which the
sheriffs and patroling inspectors kept registers of their tours of rural
areas. Judging by the later system these registers probably were carried

22 SHY, chih-kuan 48.126b-127a.

23 SHY, chih-kuan 48.24a, 24b. See also an order of 1045 t n a t t n e P a o Prefecture,
Kuang-hsin Prefecture, and An-su Prefecture police commissioners and directors in
chief were every month to rotate the duty of patroling the border (HCP 155.12a).
24 See, for example, the tour orders for the capital area in 1002 (HCP 52.1 i a - b ) .
25 HCP 242.4a, 248.2b. Tseng Pu suggested that sheriffs and patroling inspectors in
heavy-penalty places tour constantly, sending reports on the places visited to the
prefecture at ten-day intervals. The prefectural officials would compare these reports
quarterly with the road movement registers to see whether these officials were really
complying with the policy. Tseng Kung, Yuanfeng lei kao (SPTK ed.), 32.6b.
26 SHY, chih-kuan 48.67a ( n 15), 48.68b (1127), 48.136b (1212).
27 HCP 349.2a.
28 SHY, chih-kuan 48.85a-b.
29 SHY, chih-kuan 48.7oa-b; TFSL 93.

152
 Formal civil agencies of law enforcement
into the rural areas by the officials on tour so that the local village
officers could countersign the statement that a visit had indeed been
made to a specific locality on a specific day. 30 In 1120 this system was
refined. At the request of the intendant of salt, perfume, tea, and alum
for the Ching-chi and Ching-hsi circuits, whitewashed notice-walls were
ordered established in each hamlet according to the pattern used by the
inspectors of postal relays. Every month the local village officer was to
write on the face of this wall the month and day of the visit by the sheriff
or patroling inspector. The officials in question then had to sign this
wall.31
A further refinement was added in 1133 when the local superior guard
leaders were ordered to be given sealed registers to keep as a way of
recording the visits. When the sheriff or patroling inspector reached a
particular superior guard, he would sign both the whitewashed wall and
the sealed register, writing not only his name but also the date and his
office. It appears that the police official also carried with him the register
first mandated in the Cheng-ho period (1111-17) and that this too was
validated at this time. This personal register (pen-shen li) was sent up to
the prefecture when the police official returned from his monthly rounds.
The prefecture official stamped and returned it on the same day. The
responsible circuit intendants were expected to inspect this system
quarterly. 32 A report from 1214 indicates that having the sheriff visit
each superior guard in his jurisdiction once each month continued to be
the customary practice. 33
Police officials who abused the system by not going on tour or by
dispatching others to tour and sign in their places were to be punished.
Initially, such offenders could incur a penalty of one hundred blows of
the heavy rod, which was later increased to one year of penal servitude. 34
Officials also abused the system in other ways. In south China, with its
many rivers and canals, police officials sometimes treated the patrols like
excursions, taking their relatives along when patroling by boat. But
under these circumstances when they did meet bandits, they were un-
willing to fight.35 Although the rule seems to have remained that the
sheriff was to patrol, some writers suggest that in practice it was the
archers who did the actual work. Chu Hsi, writing around 1180, says
that in Tu-ch'ang District in Chiang-nan Circuit, the actual patroling

30 SHY, chih-kuan 48.68a. See also SHY, chih-kuan 48.85a-b (1214).

31 SHY, cjiih-kuan 48.67b.
32 SHY, chih-kuan 48.70a-b.
33 SHY, chih-kuan 4 8.8 5 a-b.
34 SHY, chih-kuan 48.67b, 48.68a; TFSL 92.
35 SHY, chih-kuan 48.126b (1026).

153
 Law and order in Sung China

was carried out by the archers alone. Moreover, the fort soldiers simply
guarded the town and did not go into the countryside. 36
Properly done, such patrols could be a law-enforcement aid, but this
presumes that the men used were appropriately socialized and would
not abuse their powers. In practice, however, problems arose almost
immediately. According to the edict of 962 which revived the post of
sheriff, sheriffs were not supposed to go into the countryside except to
pursue bandits. 37 In 966 an order instructed the district sheriffs to forbid
their subordinate clerks (chieh-chi) to go into the countryside, where,
using the excuse that they were dunning people for taxes or were on
regular patrol, they annoyed the people. 38 The abuses by touring sheriffs
even inspired satirical verse. 39 The local villagers and village officers
suffered because they had to provide for the touring parties, including
not only the police officials themselves but also large bodies of armed
followers. The rules against abusing the system thus were of doubtful
efficacy.40
Despite earlier proposals, until 1121 there was no clear rule for sheriffs
(as there was for patroling inspectors) concerning irregular trips into
rural areas. The general regulation governing civil service officials, that
they were not to enter the countryside except on official business, was
applied only loosely to the sheriffs. As a result they used the excuse of
chasing bandits and caused trouble to the villagers. In 1121 the Court
approved an official's request that the word sheriff be inserted in the rule
that made patroling inspectors liable for a beating with the light rod if
they improperly entered the countryside. 41 Despite these rules, the prob-
lems continued, with the troops, particularly those using the excuse of
patroling the countryside, still extorting money from the people. 42
Although the problems caused by patrols were never completely
solved, such patroling continued to be one of the key crime prevention
policies used by the authorities. Although the sheriffs and patroling

36 Chu Hsi Chu wen-hung wen-chi, 20.11b-14a.

37 HCP 3.13b-14a, SHY, chih-kuan 48.60b. See also McKnight, Village and Bureaucracy, p.
44; HCP 5.4a (964); SHY, chih-kuan 48.60b. In the twelfth century, patroling
inspectors and sheriffs who used chasing criminals as an excuse for going into the
countryside were supposed to be beaten with the light rod; SHY, chih-kuan 48.68a
(1121); TFSL 93. See also HCP 7.13b (966), in which a similar complaint is recorded
about the clerks and garrison troop commanders.
38 HCP 7.13b.
39 Yii Sou, Sung-jen hsiao-shuo lei-pien (Beijing: Chung-kuo shu-chu, 1985), hsiao-t'an lei
3-5a.
40 SHY, chih-kuan 48.68a; SHY, shih-huo i442b~43a; Liao Hsing-chih, Sheng chai chi
5.i7a-b.
41 SHY, chih-kuan 48.68a.
42 SHY, chih-kuan 48.71a (1135), 71b (1140).
154
 Formal civil agencies of law enforcement
inspectors were the chief officials concerned with it, other agents could
be involved. The 1259 gazetteer of Ming Prefecture in Liang-che Circuit
mentions that during his tenure as administrator, Fan Ch'eng-ta had
"established ten patrol huts, which enrolled four hundred stalwarts, with
sixty boats. The ten huts each used groups of three men for night
patrols." 43
The idea that crime could be deterred by having law-enforcement
personnel close at hand was also behind some of the other functions of
the sheriffs and their military counterparts. The sheriffs were respon-
sible for convoying men and materials. Thus when "barbarian" envoys
entered Sung territory bearing tribute, the prefectures and districts
through which they passed were supposed to order the patroling inspec-
tors and sheriffs to escort them. 44 Presumably the sheriff would escort
such visitors from one border of his district to the other and then turn
them over to the sheriff of the neighboring jurisdiction. In 1144 an
official complained about the plundering of convoys: "When a con-
voy reaches the boundary [of the jurisdiction of] a sheriff or patroling
inspector, these officials are immediately to inform the circuit offices of
the next jurisdiction, to escort the convoy to the far border, and there to
turn its security over to their counterparts." The appropriate offices were
ordered to devise measures to implement this request. 45 (At times the
convoying was less to protect the men being escorted than to defend
the local populace from being despoiled by the retinues accompanying
the convoys, who would demand hospitality and gifts from the local
people.) 46 Patroling inspectors and sheriffs were also supposed to meet
and travel with prefectural and circuit officials en route to office.47

Inquests
Information gathering also required direct investigation by the police
authorities. When deaths in unusual circumstances were reported, the
sheriffs (or, in certain cases, other local officials) were expected to go
immediately to investigate the situation. If they judged the death to have
been accidental or the result of disease, this investigation was deemed
sufficient. If, however, the death was suspicious or involved some breach
of law (including, of course, homicide) then the case had to be reinves-

43 Mei Ying-fa, K'ai-ch'ing ssu-ming chih (SYTFCTS ed.)3 5413-5414.

44 HCP 121.16a (1038).
45 SHY, chih-kuan 48.72a.
46 SHY, chih-kuan 48.72a-b (1144).
47 SHY, chih-kuan 48.74b~75a (1167).

155
 Law and order in Sung China

| | *'

Figure 5.1. Chart to be used in preparing inquest reports, ventral view. Wounds
found on a corpse were entered in a chart such as this, summarizing the more
exact indications of location given on a detailed drawing of the body surface.
From the Hsiyuan lu hsiang i (preface dated 1854).

tigated by another official, ordinarily a man sent from the prefecture, 48

with the expenses falling on the local villagers and sometimes especially
on the guard leaders. 49 These inquests produced a series of formal
documents, prepared by the authorities and signed by witnesses, which
would be used in any subsequent court proceedings. These documents
eventually included outline sketches of the victims on which the locations

48 Brian E. McKnight, The Washing Away of Wrongs: Forensic Medicine in Thirteenth Century
China (Ann Arbor: University of Michigan Center for Chinese Studies, 1981), esp. pp.
57
49 Sun Ying-shih, Ch'in-ch'uan chih (SYTFCTS ed.), 2711.

156
 Formal civil agencies of law enforcement
and characters of wounds could be marked 50 (see Figure 5.1). When the
identity of the killer was known or suspected, the sheriff would normally
be responsible for apprehending him.

Investigation
During the Sung, as during other dynasties, the district magistrate was
the usual judge of first instance, 51 but in the Sung he was not the
detective. Instead, the sheriffs seem to have had charge not only of
apprehending suspects but also of the initial investigations. (Some inves-
tigations were handled by jail personnel.) Such investigations included
the collection of physical evidence and the interrogation of witnesses. Of
course, higher authorities also were occasionally involved in gathering
evidence. During the trial, the district magistrates and prefectural
legal officials also sought evidence when they reviewed the more serious
criminal cases. Here, as in all aspects of Sung government, there was a
deliberate policy of diffusing functions among offices.52
Like other aspects of Sung governmental operations, such investiga-
tions were governed by sets of formal rules. A typical rule, dating from
1114, established time limits for gathering men for interrogation, allow-
ing five days in cases in which the crime called for the death penalty,
three days in which it called for exile, and one day when beating with
the heavy rod was the sentence. 53
During these investigations, beatings were often used to gain informa-
tion. The number of blows allowable during judicial torture was limited
by law to twenty (by an order of 1023).54 The authorities also tried to
limit abuses of prisoners by punishing functionaries guilty of excessive or
improper use of torture and by specifying that senior officials were
supposed to be present when the torture was applied, though in fact this
was often left in the hands of the clerks. 55 Thus, the jail functionaries
were supposed to report to the chief local officials before they flogged
prisoners. If they did not report, they might be punished. In 1018 the
central judicial offices recommended a new system under which

50 See Cheng Hsing-i, Cheng chung-su tsou-i i-chi, shang 15b.

51 For a picture of trial on the local level, see Miyazaki, "Administration of Justice," pp.
56-76. See also Robert Van Gulik, T'ang Yin Pi Shih (Leiden: Brill, 1956), chap. 3 of
the Introduction.
52 On the role of prefectural judicial offices, see Shih Hao, Mou-fang chen-yin man-lu,
SKGSCP 1973, i7.6b-8a.
53 WHTK 167.1452.
54 HCP ioi.8b-ga.
55 Chang Nieh, Tzu wei chi, SKCSCP 1975, 25.15a.

157
 Law and order in Sung China

when robbers have been seized, if the functionaries managing the jails, without
having reported to the chief officials, flog the prisoners but do not cause severe
injuries and do get the facts, they should be tried only for having violated
regulations through mistake (which carried a penalty of one hundred blows of
the heavy rod). If they exceed the measure and yet do not get the facts, for
extortionate private motives beating the prisoner and being avaricious, then try
them for violating regulations for private motives (which carried a penalty of two
years of penal servitude). 56
Those who caused the deaths of innocent people through judicial
torture could be tried on a charge of deliberate homicide, 57 though the
official involved could protest such charges, which might lead to a
reconsideration. 58
Despite such threatened sanctions, the local functionaries did, how-
ever, use judicial torture far too freely. In one case during 1007, after a
night robbery the neighborhood unit leaders reported on two men who
supposedly had not been at home on the evening in question. "The
district sheriff arrested them and interrogated them. The district clerk,
Wang Ssu, tortured them without restraint, and as a result they died."
(Shortly thereafter the prefecture seized the actual robbers.) 59 Two years
later a sheriff who seized three commoners thinking them bandits was
himself present when the clerks applied more than one hundred blows as
judicial torture and used additional illegal punishments, thereby break-
ing the men's ankles. Yet the magistrate did not know how these men
had been injured. The sheriff's office made up a false statement in-
dicating that the prisoners' ankles had been broken during beatings by
the father and elder brother of one of the suspects. 60
In some cases the sheriffs and others involved had prisoners tortured
to death 61 or used gruesome (and illegal) instruments of torture. 62 We
cannot now identify all these devices exactly, but even their names are
all too suggestive. In 1091 we are told of a district magistrate who built
instruments of torture including a "wooden steam cake" (mu hsiang-ping),
a "wooden donkey" (mu-lu), a "wooden press" (mu-hsieh), a "wooden
rack" (mu-chia-tzu), a "stone box" (shih-chia), an "iron wrapper" {t'-ieh-
kuo), and "long fetters" (chang-chia). He also secretly increased the size of
the heavy rod, with which he frequently beat prisoners. Many died.
56
57 HCP ioi.8b-ga.
58 See HCP io6.2b-3a.
59 HCP 67.14b.
60 HCP 72.5b-6a.
61 HCP 67.14b. In this case, the men killed were innocent, and the true story came out
when the prefectural authorities captured the real culprits.
62 In theory the instruments of judicial torture were carefully regulated; HCP 65.5b;
Wang Ch'eng, Tung tu shih lu'eh (Taipei: Wen hai, 1967), 4.4b.

158
 Formal civil agencies of law enforcement
Beyond all this he himself stole government funds. His case became
something of a cause celebre, with officials complaining that even distant
banishment was insufficient punishment. 63
In other cases the instruments are said to have been illegally used by
the jail clerks, though it is unlikely that this could have been done
without the knowledge of the sheriff. There are complaints that such
functionaries beat the hands and feet of suspects with wooden staffs;
used throat presses made of wood and cloth; made suspects kneel while
men jumped on a wooden staff laid across their legs; or applied the
"brain hoop" by wrapping the suspect's skull in cloths and then squeez-
ing it by inserting wooden wedges. 64 Another group of clerks was said to
have wrapped a suspect's head in a belt of wet leather. As it dried it
contracted, and when the prisoner could no longer bear the agony he
signed a false confession.65
The central authorities did try to stop such abuses. For example, after
the night robbery incident in 1007 they issued an order stating that "the
rules on judicial torture have already been determined in detail. Illegally
and without authorization to go beyond the rules is cruel. Officials in the
various circuits are to see to the destruction of illegal instruments of
torture." 66
Sometimes the authorities tried to reduce the incidence of abuses by
heavily punishing the functionaries responsible. In 1028 a sheriff who
had had a man beaten to death was himself beaten, tattooed, and sent to
Kuang-nan to serve as a laborer registered in the provincial armies. 67
The state itself, however, certainly contributed to conditions that encour-
aged such acts, since sometimes it punished officials in only the most
trifling ways for truly gruesome abuses. In the early 1080s a circuit
judicial official who burned a man's face was merely fined and was
subjected to a minor setback in his career - the authorities ruled that he
had to serve one extra year in his post before he could seek promotion.
After this it is said that he burned the faces of female slaves three times
in one year, but even when an official impeached him, he was simply
sanctioned by having five more years added to the seniority require-
ments needed for promotion. 68
The temptation to use excessive force stemmed in part from the key
role that information gained in this way played in later trials. District
magistrates frequently conducted trials largely on the basis of reports
63 HCP 45g.6a-b. See also HCP i6a-i8a, 460.7b, 9a-b, 9bff, 17a.
64 SS 200.4996.
65 HCP 67.5b.
66 HCP 67.5b.
67 HCP.106.10b.
68 HCP 330.7b.

159
 Law and order in Sung China

drawn up by clerks using information secured by the sheriff and his staff.
Although their primary functions were to capture criminals and gather
information, occasionally the sheriffs might be called on to reexamine
cases that had been appealed or to reinvestigate cases that had already
been decided, though concern about possible conflicts of interest
prompted the creation of a rule that only those men not involved in a
case could be so used. 69 There even were some limited circumstances in
which sheriffs not only functioned as the investigators but also were
considered competent to judge minor cases and have punishments
inflicted.70 In general, however, their power to adjudicate was severely
limited, and officials who did conduct their own trials might be subject
to quite severe sanctions. 71 According to an order of 986, "trials may not
be conducted in district sheriffs' offices."72 The meaning of this order is
made clear by a remark in Wang Yung's Sung-ch'aoyen-i i-mou lu:
The sheriff's duty is the policing of robbers. When there were fights in the rural
areas, the people, fearful of becoming involved with the [higher] prefectural and
district authorities, often turned the culprits over to the sheriff's office. For this
reason the sheriffs conducted trials in which men were subjected to the bitterness
of judicial torture. Therefore in the Hsien-p'ing period [998-1003] there was a
decree that forbade these trials and returned to the district magistrates the
responsibility for all legal suits.73

Communications
Traditional Chinese governments were acutely aware of the vital role of
information transmission. 74 If the total system was to function properly,
accurate messages had to be sent by the appropriate parties and received
and understood by the appropriate parties at the proper times in the
proper places. The pervasive traditional Chinese concern with this need
was based both on a general underlying ideological orientation concern-
69 HCP 72.4a (1009). This incident provoked an order that officials involved in the initial
stages of a case could not be called on to reinvestigate it.
70 SHY, chih-kuan 48.80a, records an order dated 1184 in which a sheriff's office was
established at a distance from the district town. The sheriff was given the power to try
and punish offenders when the sentence was seventy or fewer blows of the heavy rod.
According to Miyazaki Ichisada, the general Sung rule was that district magistrates
could administer punishments of the heavy rod or less. See Miyazaki Ichisada, S 6 -
Gen, p. 137.
71 See HCP 477.1b-2a, which says that such men were liable for one year of penal
servitude.
72 SHY, chih-kuan 48.61a; HCP 43.10b.
73 Wang Yung, Sung-ch'ao yen-i i-mou lu, 1.8. For chih yu as "conduct a trial," see HCP
106.2b; 180.9b.
74 For some further commentary and illustrations of this awareness, see McKnight,
"Patterns of Law and Patterns of Thought," pp. 323—31.

160
 Formal civil agencies of law enforcement

ing the role of communication in the creation of order and on the

practical character of bureaucratic politics.
In the realm of law enforcement the need for efficient communications
was even more immediate. T h e first part of the c o n u n d r u m of law
enforcement is the need to learn about violations and to alert law-
enforcement agents in a timely fashion. This problem was very much on
the minds of effective local officials. W u Ching, whose effectiveness
against bandits during his first appointment as a sheriff led to his
promotion and transfer to the post of district magistrate, built d r u m
towers in his area to be used to alert the constables and soldiers to the
incursions of bandits. As a result, in a time of natural disaster when
bandits were r a m p a n t in the surrounding areas, " t o the end not one m a n
dared cross my b o r d e r s . " 7 5
In regard to transferring information within the bureaucracy, a pro-
perly working communication system of the type created by traditional
Chinese empires functioned as an important tool of power, providing
superiors with the information necessary to control their subordinates
a n d limiting the subordinates' access to information that would increase
their control beyond that needed for the tasks of their office. In prac-
tical terms, as exemplified by the Sung bureaucracy, this meant the
voluminous production of reports, statistics, m e m o r a n d a , and sugges-
tions generated at lower levels and sent u p through the hierarchy, with a
smaller return flow of orders and guidelines.
Within the law-enforcement agencies, local officials were required to
submit regularly a series of reports on conditions in the areas under their
jurisdiction. Early in the dynasty the rule was that if the bandit-seizing
officials delayed in informing the prefecture, they were to be considered
as having violated the system for private motives (wei-chih), which
carried a penalty equivalent to two years of penal servitude. In 1018 as a
result of an official complaint that this penalty was too harsh, the rule
was changed so that such men were considered to have violated systems
only by mistake, a lesser offense that carried a penalty equivalent to one
h u n d r e d blows of the heavy rod. 7 6 A decree of 1043 clarified the pro-
cedures. Those officials in charge of capturing bandits, on the day that
the attacks occurred, were to report to the prefecture the n u m b e r of
persons plundered. T h e prefecture in turn was to report to its superior
jurisdiction on the day that it received such a report. 7 7 After 1043, t n e
officials were also supposed to report to the prefecture the capture of

75 Wu Ching, Chu-chou chi 8.3b.

76 HCP 91.8a—9a. See also SHY, hsing-fa 1.17b, which indicates that this rule continued
in force at least until the late eleventh century.
77 HCP 144.6b.

161
 Law and order in Sung China
violent bandits on the day they were seized. The prefecture in turn was
supposed to notify the higher authorities on the day that it received the
district report. Failure to do so was a punishable offense,78 but predict-
ably, the officials locally responsible for capturing bandits continued
to conceal crimes because they were afraid that they would suffer
sanctions. 79
The prefects also had to submit semiannual reports on the numbers
of robbers in their jurisdictions. Under a system started in 1090, the
Ministry of Justice was to use these reports to compare the numbers
arrested and at large during the preceding five years and to forward
these reports to the Court. 80 These reports apparently were sent by way
of the circuit intendants, who consolidated them, but also used them as
evidence for distributing rewards in money. 81 If the time limits for such
reporting were violated, the penalty was the equivalent of one hundred
blows of the heavy rod. 82
Officials could be sanctioned for having failed to capture wanted men.
And if their failure led to the spread of banditry, they might be severely
punished. 83 From 1093 onward, prefects leaving their posts after having
served one year or more, with some limited exceptions, had their records
scrutinized for bandit capturing. If overall they had captured fewer than
half of those at large, a report about this was memorialized, presumably
so that they could be penalized. 84 For some years during the Yiian-yu
period (1086-93) officials could reduce their penalties by balancing the
number of men caught against those not caught. However, in 1099
this Yiian-yu proportional reduction law was abolished, and the policy
reverted to the pattern followed in the Yiian-feng period (1078-85). 85
This policy on reporting bandits may have been strictly enforced
during the Northern Sung, but by the 1130s it had fallen into disuse as a
result of the military troubles of the 1120s. In 1138 Chang Shou sug-
gested reviving a system of district records, which would record the
names of the families who had been robbed and the dates of the in-
cidents. Arrests would be noted in red ink on the bottom of the docu-
ment. The record could be checked quarterly by the controller-general.
At year's end, the circuit officials would check them to find out how
78 HCP 144.6b.
79 SHY, ping i 2 . i g a - b (1126).
80 HCP 437.21b.
81 See HCP 300.6b-7a, 447.9a-b.
82 HCP 460.1a.
83 HCP 465.1a.
84 HCP 481.1b. In the Yiian-yu period (1086-93) a policy was initiated under which only
those bandits were counted whose activities had been initially reported after the last
preceding amnesty.
85 HCP 310.12a, 437.21b, 507.16b.

162
 Formal civil agencies of law enforcement

many robbers there had been, how many had been captured, and how
many were still at large, giving rewards and punishments in accordance
with the number and the speed with which the criminals had been
captured. 86
These policies were in part designed to deal with one of the general
problems of law-enforcement systems. The interests of central authorities
and local police officials may not be the same. In traditional China the
center needed accurate information about local unrest, but the local
officials were tempted to conceal or downplay such problems, because
the appearance of bandits reflected poorly on their administration and so
might adversely affect their careers. 87 If there were no reports of bandits
during an official's term in office, then he would receive a commenda-
tion; if there were bandits who were captured within the allotted time
limits, he would not be reprimanded; and if he failed to capture them
in time, then he would suffer progressively more serious bureaucratic
sanctions as time passed. 88 Because officials were tempted to ignore or
downplay reports of unrest, the state instituted harsh penalties for failure
to report. Even in ordinary areas, not reporting might carry a penalty
equivalent to two years of penal servitude. 89 After n o o , in those vital
areas of the empire where especially heavy penalties were inflicted,
sheriffs or patroling inspectors who failed to report gangs of ten or more
bandits were to be excluded from the group of those whose sentences
could be reduced through amnesties. 90
Reports were, of course, part of a two-way communication system.
Under Sung law, according to a report of 1091, the fastest communica-
tion method was used for amnesty documents, the imperial decrees sent
from the capital that freed or reduced the sentences of many criminals. 91
These were carried by relays of runners at the rate of about 170 miles
(500 lid) per day. Reports on border affairs or extraordinary banditry
were carried by runners at about 133 miles (400 lid) per day. Ordinary
banditry reports were sent by horse post at about 100 miles (300 lid) per
day. If the carriers failed to meet their deadlines, they were to be beaten
fifty blows of the light rod for being less than an hour late, eighty blows
of the heavy rod for one hour up to one day, and one hundred blows for

86 Chang Shou, P'i-ling chi, 3.37-38.

87 HCP 129.3b, 129.3b (1040).
88 SHY, chih-kuan 59.1b (962).
89 SHY, chih-kuan 48.66a. This ruling differed somewhat from the rules found in the Sung
hsing t'ung; see SHT 28.448 ff. See also TLSI 28.57, 61; Niida Noboru, Torei shui
(Collection of T'ang ordinances) (Tokyo: University of Tokyo Press, 1964), p. 729.
90 SHY, chih-kuan 48.66a; HCP 494.29a. For the heavy-penalty system, see Chapter 14.
91 Amnesties will be discussed in more detail in Chapter 14.

163
 Law and order in Sung China
one day, with an increase of one degree for each additional two days'
delay, up to a maximum of three years of penal servitude. 92

Bureaucratic functions
The state's ability to respond effectively to disorder was limited not only
by considerations of power but also by fiscal constraints. With the
sheriff, as with other offices, the state had to balance its desire that the
sheriff (and, similarly, other law-enforcement personnel) perform as well
as possible his principal role as a police official, against its desire to
reduce the cost of local administration by burdening him with other
tasks.
At some times and places the sheriffs were assigned a variety of
administrative, fiscal, or supervisory tasks, which complicated and inter-
fered with their law-enforcement duties. At least during part of this
period, the sheriff- like the recorder, the assistant magistrate, and the
magistrate - was expected to go daily to the Senior Officials' Court
(Chang-kuan t'ing) of the district headquarters to discuss affairs and
endorse documents. 93 Sometimes in the less-populous districts, sheriffs
were appointed concurrently as recorders (who managed the flow of
documents, especially fiscal documents, in the district). These added
responsibilities greatly broadened the sheriffs' duties, though their total
workload may have been modest because of the small populations of
such districts. 94 According to a rule of 970, districts with one thousand
or more households were to have a magistrate, a recorder, and a sheriff;
those with two hundred to one thousand households were to have a
magistrate and a concurrent recorder /sheriff; and those with fewer than
two hundred households were to have only one man serving concurrently
as recorder and sheriff.95 In the early years of the Sung this system was
applied only to districts of two hundred or fewer households, 96 but after
978, at least in Kuang-nan, it was used in districts with five hundred or
fewer households. 97 Apparently over time there was a tendency in some
areas for the recorders and sheriffs to share some types of duties, even
when there were separate offices in the district. In 1070, as a result of a
complaint about this situation, the Court reaffirmed the distinction in

92 HCP473.ua (1091).
93 SHY, chih-kuan 48.53b; HCP 499.9a.
94 HCP 118.4a. From 1001, in Shensi and Szechuan, all districts with fewer than five
thousand households had registrars concurrently serving as sheriffs; see HCP 49.14a.
95 HCP 11.6b.
96 HCP 11.6b.
97 HCP 19.na.

164
 Formal civil agencies of law enforcement
tasks: The recorder was in charge of written records, and the sheriff had
sole control over law and order. 98
Early in the dynasty (in 987) the sheriffs also were sometimes assigned
the job of watching tax collection depots, but only three years later, as
a result of complaints about the sheriffs' and archers' abusing their
powers, this order was rescinded, and the village officers took over the
task." From this time on, sheriffs were rarely associated with tax collec-
tion, though they occasionally were called on to manage tax depots or
monopoly stations in outlying areas 100 or to dun recalcitrant households
for tax debts. 101 Biographical reports from the thirteenth century also
indicate that sheriffs might manage local welfare granaries, 102 and the
1208 gazetteer of Chen-chiang Prefecture in Liang-che Circuit mentions
a sheriff who was responsible for supervising the management of two
government estates.
The sheriffs' miscellaneous duties might even include managing local
temples. An order of 972 states that the magistrate or sheriff of the
district in which a temple was located was to serve concurrently as the
temple manager of temples such as those dedicated to the Five Sacred
Mountains, the Four Rivers, or the Eastern Sea. 104 In the same general
line of ritual duties, the recorders and sheriffs were ordered to manage
the rites of gratitude performed in the schools. 105
In addition to these recurrent duties, the sheriffs might also perform
extraordinary tasks. At certain times individual sheriffs submitted their
advice on a wide variety of topics. Sometimes the advice seems to have
been requested and dealt with the maintenance of order 106 or the pre-
vention of disorder (in times of natural disasters), 107 but sheriffs also

98 SHY, chih-kuan 48.63b-64a; HCP 214.24a.

99 SHY, shih-huo 70.4a.
100 For instance, in 1208 there was an order for a sheriff to oversee a silver tax-collecting
station and wine tax revenues; SHY chih-kuan 48.83b. And from 1165 there was a
similar order for a sheriff to oversee a tax depot in Szechuan; SHY, shih-huo 18.3b.
See also Hsiieh Chi-hsiian, Lang-yu chi, 20.6b, 26.ia-b.
101 Chu Hsi, Chu tzu ta ch'u'an (Ssu-pu pei-yao ed.) (hereafter cited as SPPY), wen 18.17b.
102 Yuan Fu, Meng-chai chi, 17.251-52.
103 Ch'en Ch'i-ch'ing, Chia-ting Ch'ih-ch'eng chih (SYTFCTS ed.), 2842.
104 HCP 22306.1 ia. The Five Sacred Mountains are T'ai-shan, Hua-shan, Heng-shan,
Hang-shan, and Sung-shan. The Four Rivers are the Yangtse, the Yellow, the Huai,
and the Chi.
105 This is an order of 1141, but it enjoins these officials to have general management
"according to old precedents" and so must presume an earlier origin of the system. It
is not at all clear what these ceremonies were; see SHY, chih-kuan 48.72a.
106 HCP 283.14a says (in 1077) that "the emperor took note of Huang K'o-chiin, the
sheriff of Nan-feng District in Chien-ch'ang Prefecture. From the time when Liao En
became a bandit, among the multitude of plans submitted for suppressing them, none
was as good as that of [Huang K'o-chiin]."
107 See an order of 1074 that the patroling inspectors and sheriffs in the areas of Huai-nan

165
 Law and order in Sung China

memorialized on topics only peripherally within their purview, such

as foreign border problems or monopoly systems, and sometimes carried
their zeal to the point of potentially dangerous effrontery. In 993 a
sheriff memorialized asking that the number of imperial concubines be
reduced. 108 They might also be specially ordered to go out for various
commissions, to inspect such facilities as the ever-normal granaries, or
the salt, tea, and alum monopoly systems. 109
Finally, the sheriffs might be called on to act as general district
administrators when the other district officials were out of town on
business. As an official pointed out in 1109, this could cause grave
problems if, while a sheriff was filling this role, there was a major
disorder in the district. As a result of his recommendations, it was
ruled that at least one of the other district officials (the magistrate, the
assistant magistrate, or the recorder) had to remain in the town. They
all could not leave at one time. If under extraordinary circumstances
there were no such men in these posts, then the prefecture should
dispatch an official to act on a temporary basis. 110
Most of these data preserve the dry bones of the bureaucratically
operated system. Occasionally the reports allow the personalities of these
men to appear for a moment:
The doctorate-in-letters graduate from Kaifeng, Sang Shih, who was appointed
sheriff of Wei-nan District, was a courageous man from Hsiung-chiu. He was a
man of great strength and skilled with the sword and iron rod. . . . In the
neighborhood ofJu Prefecture the various districts had many bandits. Shih asked
the help of the elders and went about investigating the evil behavior. Then he
called together the juvenile delinquents of the village to warn them that he would
not endure banditry. There was a father of Ch'ing hamlet. His son had died, and
at night, before the son had been dressed for burial, a thief stole the burial
clothes. The father was afraid to report it to the authorities. Shih suspected a
juvenile surnamed Wang. At night he went to the Wang house and secured the
stolen clothes, without the Wang boy's finding out. The next day Shih met him.
"You promised me you would not steal. Are you not the one who stole these
clothes?" The boy turned pale. Shih threw him to the ground and bound him.
East that had suffered be allowed to send in memorials without regard to the regular
system. When there had been a good harvest, then the older rules would come back
into use; HCP 255.2b.
108 HCP 338.7b, 67.5b, 34.6b. In this particular case the man seems to have provoked the
emperor's admiration rather than his wrath. In Chu Pien, Ch'u-wei chiu wen (Chih pu
tsu chai ts'ung-shu ed.), 7.5b, it says that "T'ai-tsung spoke to the chief minister
about it, observing, 'A petty official who dares to discuss the affairs of the inner
palace, this is to be admired.'"
109 HCP 248.20b (1073), 216.13b-14a. In 1135 the newly added military sheriffs were
ordered to tour and investigate illegal salt and tea dealing; SHY, chih-kuan 48.69b.
110 SHY, chih-kuan 48.66a-b. At times the sheriffs themselves were sent out to other
districts to manage various affairs; HCP 104.13b (1026).

166
 Formal civil agencies of law enforcement

He turned over the names of the criminals to the local authorities. All confessed
their crimes. 111

Remuneration
According to Winston Lo, the differentials in monetary salaries of Sung
bureaucrats were relatively small, with the top grades receiving only
about ten times the amounts given to those lowest in the hierarchy. 112
Still, the officials on the bottom of the pyramid, like the sheriff, were not
well paid in money. The cash income of a sheriff seems to have been only
slightly more than three times the income paid to one of his constables.
Of course, as an official the sheriff also received substantial income in
kind and in perquisites. Nonetheless, sheriffs can hardly be considered
generously remunerated.
Moreover, although the sheriffs' functions were clearly important to
the proper functioning of local government, perhaps because of their
lowly bureaucratic position the sheriffs often were forced to work in
inadequate facilities. The sources often note the decrepit conditions of
jails and offices.113 The sheriff of Sui District in Yen Prefecture in Liang-
che Circuit was at one point housed in an abandoned mail station. 114
And Mao P'ang commented about a prefecture in Ghing-hsi South
Circuit:
Officials in the nine grades view the post of sheriff as the lowest. Within the nine
territories, this Ying is the most rustic. One hundred paces north of the pre-
fectural wall there is the sheriff's office. This old building of ten bays [chien,h a
traditional unit of architectural measurement equivalent to the distance between
two upright pillars] was rotting and dark, bent and broken. When it rained,
water came in, so that the clothes of those inside the building became wet.115
Even when there were adequate facilities, they seem almost always to
have been created on the initiative of especially concerned and energetic
sheriffs, who sometimes used their own funds to pay for improve-
ments. 116 Particular structures are often traced to particular incumbents,
who sometimes gave the facilities names that belied the grim overall
111 HCP 105.15b-16a. In another incident, Shih personally attacked and killed a number
of bandits, tied up the rest of the gang, and later went to another area suffering from
bandits and again killed or captured a whole gang. For more on this heroic lawman,
see CYKC 99.
112 Winston Lo, An Introduction to the Civil Service of Sung China (Honolulu: University of
Hawaii Press, 1987), pp. 160-61.
113 See, for example, Li Liu-chien, Tan-chai chi SKCSCP 1971, 15.16b-17a.
114 Cheng Yao, Ching-ting Yen-chou hsii-chih, TSCC ed., 7.78.
115 Mao P'ang, Tung-t'ang chi, SKGSCP ed., 9.23a. For a similar complaint, see Shih
Chieh, Tsu-lai chi, SKCSCP ed., 19.17b.
116 See, for example, Yuan Fu, Meng-chai chi, 18.258.

167
 Law and order in Sung China

purpose of the office. The gazetteer of K'un-shan District in Liang-che

Circuit, dated 1251, stated:
The sheriff's office was located fifty paces south of the district headquarters. It
once had had a Shadowy Hall, a Pavilion of the Laughing Moon, and a Separate
Kingfishers Studio, all of which had been built by the Sheriff Wang Chi-yii. .. .
In 1245-46 the sheriff, Ch'en Hsiao-sun, added a library east of the office.117

Rewards and punishments

Like informal and semiformal agents, formal police agents could benefit
from the system of rewards for capturing criminals. Like most aspects of
Sung government these rules were periodically collected for more con-
venient reference by the authorities. These reward rules, which were
governed by a type of law in the Sung called ko, might refer to specific
areas or specific groups. In 1086 the Ministry of Finance compiled the
Reward Laws for the Capture of Bandits in Cheng and Hua Prefectures.118 In the
same year the Bureau of Military Affairs petitioned successfully to have
a recompiled the rules on rewards for capturing bandits applicable to
Shensi and to Ho-tung Circuit. (The bureau was dissatisfied with the
existing compilation because it had been created during the reform
period when the partisans of Wang An-shih were in power.) 119
Such collections could be quite sizable. In 1179 an official submitted
four separate compilations, the Reward Law for the Prefectures of the Various
Circuits (Chu-lu chou-chiin shang-fa) in 139 chapters, the Circuit
Intendants} Commendation and Reward Law (Chien-ssu chou-shang fa) in 47
chapters, the Generally Applicable Reward Law (T'ung-yung shang fa) in 13
chapters, and the Traditional Rewards for the Northwestern Prefectures (Hsi-
pei chou-chiin chiu-shang) in 1 chapter. These compilations were issued
as a unit entitled Individual Prefecture Individual Circuit Commendation and
Reward Laws of the Ch3un-hsi Period (Gh'un-hsi i-chou i-lu chou-shang fa).
The compilation was needed because officials had been faced with a
welter of laws, some going back a century. Selecting the pertinent law
was not always easy. Furthermore, many of the cases and recorded
precedents which might have been used when deciding on rewards had
been lost during the military troubles of the 1120s and 1130s.120 The
sources contain many scattered examples of rewards (see Table 5.1).
117 Sun Ying-shih, Ch'in-ch'uan chih, p. 3809.
118 SHY, hsing-fa 1.13a; HCP 373.17b.
119 HCP 368.22b-23a. See also HCP 307.5a.
120 SHY, hsing-fa 1.52a. This was clearly not the first time that the Sung had compiled
such rules. On earlier compilations, see, for example, from 1133 a compilation of
Ministry of Justice precedents on rewards for the capture of bandits; SHY, hsing-fa
1.36a.

168
 Formal civil agencies of law enforcement

Table 5.1. Sample reward information

Amount in
Date Type of crime strings Source

1034 Prisoners killed 100 HCP 114.16b

1034 Homicide 50 HCP 114.19b
1035 Homicidal bandits
10 or more 100 HCP 117.4b
fewer than 10 Proportionate
1041 Teaching martial arts 3 per man HCP 134.9a-b
1047 Heresy 500 SHY, hsing-fa 2.27b-28a
1070 Corruption
(penalty) heavy rod 50 HCP 217.10b
penal servitude 100 214.27a
exile 200
exile to Sramana 300

Because the dates of the information given in this table are relatively
close together, we can probably assume that the differences in amounts
represent differences in the authorities' perception of the seriousness of
the problems rather than changes in the value of money. The amounts
are really quite large, up to ten times the annual pay of Northern Sung
first-grade constables, who only earned about thirty strings per year.
The rewards might come from the offenders' conscripted goods or from
official funds. Sometimes the informers were rewarded with a proportion
of the illicit goods recovered. 121

Extortion and bribery

In the deceptively simple paradigm of law enforcement, one phrase says
that the state must use men who are properly motivated. When men
given power over others have not absorbed those values supportive of the
interests of the state and therefore approach their tasks with "illicit"
motives, the results are sure to be troublesome. The paradigm of law
enforcement that calls for getting the proper number of men to the
proper place at the proper time also assumes that the goals of those who
lead these men will be the same as the goals of the central authorities. In
practice, low-level officials usually have goals of their own. Among other
things they may be tempted to use men supposedly intended for one
task, like law enforcement, for their own general purposes. The problem,

121 See HCP 302.6a.

169
 Law and order in Sung China

which can exist in any bureaucratic system, appears occasionally in the

records dealing with Sung police agencies. The sheriff could legitimately
use a number of archers as personal attendants, but there also are
reports that the village servicemen called stalwart men, who supposedly
had only constabulary duties, were sometimes asked by local officials to
serve as personal guards. 122
The demands and dangers of the job, the opportunities for abuse, the
inadequacies of the reward system, and the use of personnel from groups
of doubtful motivation all combined to create a situation in which abuses
were endemic. For police officials the temptation to abuse their power,
extort evidence from the accused, or money from all was a serious one.
The difficulty might have been reduced if the state could have attracted
able, well-indoctrinated men to these posts; since the posts were at the
very bottom of the civil service ladder and were poorly remunerated, the
state had to try to fill them from a pool of men who often had dubious
credentials.
The problem of finding effective local officials was exacerbated when
the central authorities left men in office for very short terms before
transferring them elsewhere. Figures for magistrates from the Southern
Sung show that these officials had terms averaging between one and a
half and two and a half years. 123 How could men serving such short
terms have had sufficient time to become familiar with their jurisdic-
tions? Because of their low position, relatively few sheriffs have bio-
graphies, too few for a systematic survey of the men from any given
district. We do, however, have some lists of names. The list from
Ch'ang-shou District in Sua Prefecture suggests that in this area, during
the twelfth century at least, the sheriffs, like Northern Sung prefects or
Southern Sung magistrates, served for very short terms. In the seventy
years between 1126 and 1196, thirty-three men were appointed sheriff in
this district, which meant average terms of less than two years. 124 As
with a number of other local offices in the Sung, these short tenures
must have had a negative impact on the effectiveness with which the
incumbent officials could perform their duties.

122 HCP 120.18a (1037).

123 See figures for the four districts in Chien-k'ang Prefecture, where the incumbents had
terms averaging two and a half years, and the figures for the districts in Yen
Prefecture where the average terms ranged from less than one and a half years to
about two years. Chou Ying-ho, Ching-ting Chien-k'ang chih, SYTFCTS ed., pp. 1141-
55; and Cheng Yao, Ching-ting Yen-chou hsii-chih, pp. 7042-43, 7050-51, 7054.
124 Sun Ying-shih, Ch'in-ch'uan chih, pp. 2690-92. For Northern Sung prefects, see
McKnight, "Administrators of Hangchow," p. 208.

170
 Formal civil agencies of law enforcement

Archers as subordinates of the sheriff

It was necessary to move sufficient numbers of appropriately trained and
armed and properly led and motivated personnel to the requisite places
in a timely fashion, but the state faced both an economic and a socio-
political dilemma. During the Sung district populations ranged from a
few thousand people to fifty thousand people or more. A large proportion
of these populations lived scattered in rural areas. To maintain order in
such an area how many men would be needed and how should they be
trained and armed? The state had only a limited fiscal base. What would
it cost to provide the requisite number of men? And even if this fiscal
problem could be solved, there was a potentially more worrisome social
and political problem. Who was to guard the guardians? Armed men of
any sort were a constant source of danger and unrest. They might
themselves cause troubles, or even more dangerously, they might fall
under the control of locally powerful figures, men who had the knowl-
edge and organizational skills to be serious threats to the state. There
were thus strict limits on the state's capacity to control its scattered
populations. Approximately ioo million people lived in China in
noo. State officials, civil and military, numbered a few tens of
thousands, perhaps half of whom were serving in the capital city. Of
those in local areas, only a fraction were primarily concerned with law
enforcement.
To exercise even a fiscally limited degree of control the state had to
provide this handful of officers with forces of full-time constables and
other law-enforcing subordinates. To understand the workings of Sung
law enforcement, we must know who these men were, how they were
recruited, and what functions they performed. The problem of how
many such men should be used and how they were to be armed and
trained was a perennial one. Both the sheriff and the patroling inspector
relied on these armed subordinates in carrying out their law-enforcement
duties and used both direct personal subordinates and militia forces only
partly dependent on them.

Archers: numbers and costs

The sheriffs' most important support was a local constabulary of archers
(kung-shou). These archers were armed, as their name suggests, with
bows, but also carried staffs and other simple weapons. An order of 1016
indicates that in some parts of China the state was to lend each archer a
crossbow. The archers were to supply their own swords. The sheriff was

171
 Law and order in Sung China

to train them in arms, though they probably received less training than
did the members of the regular armies. 125 The archers were to live at or
near the offices of the sheriffs, who were ordered in 1005 to provide them
with barracks. 126
The decree of 962 which reestablished the sheriffs also transferred to
the sheriffs' control the archers who had previously been subordinate
to the garrison troop commanders. According to the rules then promul-
gated, each district was to have a quota of archers that would vary
according to the number of households. (We may roughly estimate that
each such household numbered about 5 persons.) A district with 10,000
or more households was to have 50 archers, or approximately 1 per
thousand people. A district with 7000 to 10,000 households would have
40 archers, and so on down to districts with fewer than 1000 households,
which would have 10 archers. 127 In 968, on the grounds that condi-
tions had become more peaceful, the quotas were lowered: A 10,000-
household district now had 30 archers, and so on down to a district with
fewer than 1000 households which retained its old quota of 10 men. 128
Thus, if these quotas had been properly filled, there would have been
approximately 1 constable for every 500 to 1700 people. These archers
were probably slightly outnumbered by the soldiers who served under
the patroling inspectors. Thus in a district there would be 1 law officer,
constable, or soldier for every 250 to 850 people. This may be compared
with the situation in the rural United States, in which the number of
police, deputies, and other law officers is slightly less than 1 for every
450 people 129 (see Table 5.2).
Archer quotas were increased in the 1040s when the Sung was fighting
against the Hsi-hsia state to the northwest, possibly in response to a rise
in general lawlessness within the empire, as well as to the immediate
external threat. 130 In 1042 the controllers-general were ordered to select
villagers with military talents, in order to triple the number of archers
in certain circuits. 131 In some cases such men were apparently to be
125 HCP 86.17b.
126 HCP 52.17a.
127 SHY, chih-kuan 48.60a; HCP 3.13b-14a.
128 HCP 9.11b; SHY, chih-kuan 48.6ob-6ia. Note that the higher authorities did not
expect all the sheriffs to be pleased with this order. The decree specifies that sheriffs
who resisted the implementation of the order were to be punished.
129 To arrive at a roughly approximate figure for comparative purposes, I used the
number of police officials in the state of Mississippi (according to the Statistical Abstract
of the United States 1988) and divided this figure into the population of the state. The
figures are not all that different for Hawaii, a state in which the majority of the
population lives in Honolulu.
130 HCP 125.14b, 126.12b, 127.15b, 131.20b, i33.ia-b.
131 HCP 132.21b orders this for Ching-tung West Circuit; see also HCP 132.21b, 133.1 ia,
12899.14b, etc.

172
 Formal civil agencies of law enforcement

T a b l e 5.2. Numbers of archers in selected districts

Number of Number of
Date Place archers households Source

1079 Kaifeng Prefecture

Hsiang-fu District 60 HCP311.5a-b
1117 Ta-ming Prefecture
Yuan-ch'eng District 105 SHY, chih-kuan
48.67b
Ca. 1120 Su Prefecture
Ch'ang-shou District 500 Sun Ying-shih,
Ch 'in-ch 'uan chih
2709
1204 Shao-hsing Prefecture
K'uai-chi District 95 35,406 K'uai-chi Chih
4.5b, 5.2a
Shan-yin District 95 36,652
Sheng District 98 39,792
Chu-chi District 113 42,424
Hsiao-shan District 70 29,063
Yii-yao District 100 30,833
Shang-yii District 70 30,303
Hsin-ch'ang District 67 28,820
1228 Ming Prefecture
Ting-hai District 65 Lo Chun,
Pao-ch 'ing
Ssu-ming chih
5320
Yin District 85 5247
Feng-hua District 84 5274
Tz'u-ch'i District 80 5298
Ch'ang-kuo District 73 5336
Hsiang-shan District 80 5354
1265 Lin-an Prefecture
Ch'ien-t'ang District 53 87,715 Ch'ien Yiieh-yo,
Hsien-ch }un Lin-an
chih 57.6a-b
Jen-ho District 100 98,615 58.2b ff
Yii-hang District 62 26,581
Yii-ch'ien District 90 25,965
Fu-yang District 90 25,965
Ch'ang-hua District 60 13,687

173
 Law and order in Sung China

tattooed on the hand for identification purposes;132 in others they might

be tattooed on the face. The Court in 1042 approved a request that in
Ho-tung Circuit the archers to be inducted into the righteous warrior (i-
yung) units not be tattooed but that those in Shensi be tattooed on the
face for use in the victorious protector (pao-chieh) commands. 133 (These
tattoos indicated the name and location of a soldier's unit.)
These newly recruited men, as distinguished from the archers pre-
viously employed, should be thought of as militia soldiers rather than as
simple law-enforcement constables.134 They could also serve as regular
soldiers, though supposedly only if they volunteered for such service.135
According to an edict of 1040, the newly organized archers' units on the
northern frontiers were to be organized in units of twenty-five men called
companies (t'uan). Four companies were to form a group (tu), with five
groups comprising a command of five hundred men (the key unit in the
regular Sung military organization), headed by a command officer (chih-
hui shih). The men so inducted were to serve from age twenty until age
sixty. They were to receive one month of military training during the
agricultural off-season and were to be given rations during training as
well as when they were mobilized for other reasons. 136
It was in connection with this mid-eleventh-century expansion of
forces that the state's dilemmas were described most eloquently by one
of the truly brilliant scholar-officials of that era, Chang Fang-p'ing. At
that time he was serving as the controller-general of a relatively back-
ward prefecture in Chekiang. In 1041 he submitted a lengthy and de-
tailed memorial concerning the newly adopted policy of expanding the
number of archers. He began by detailing the underlying political and
fiscal problems presented by the new measure. Under it, fifty archers
would be drafted for every one thousand households. For prefectures
with more than ten thousand households, the quotas were lowered some-
what. Still, a prefecture with fifty thousand households would have five
thousand archers. Recruiting such a force would be a great burden
on the people, and the potential threat to security posed by the very
existence of such a large body of armed men was also extremely worri-
some. The geography of some prefectures was such that control was
intrinsically difficult. In other cases the force would be active in areas
where there had been famines, thereby presenting opportunities for

132 HCP 132.19b, 12400.15b, 134.20b; see also 12399.7b.

133 HCP 12399.7b.
134 HGP 133.2b, 133.5b.
135 HCP 133.2b, 12399.7b, 15a, 12400.15b.
136 HCP i28.6b-7a, i 5 a - b .
174
 Formal civil agencies of law enforcement

pillage. And there was always the danger of these men's falling under the
sway of troublemakers and becoming mutinous.
Moreover, under the new rules, during the agricultural off-season the
archers were supposed to present themselves for one month of training in
the use of weapons. During these periods they would be given rations by
the state, 2 bushels (shihb) and 4 pecks (toub) of rice per man. This
amounted to approximately 130 dry U.S. pints, that is, over 4 pints per
day. This was a substantial ration. Subsistence was considered to be the
equivalent of about one-half U.S. dry pint per day, and the regular
ration of men in penal registration was about 2.5 U.S. dry pints per day.
For Chang Fang-p'ing's prefecture - which for purposes of recruiting
archers was considered to have a population of about twenty thousand
households - this item alone would have meant an expenditure of almost
5000 bushels (shihb) of grain a year. For the empire as a whole, the
burden would have run to about 1 million pecks (touh) per year. The cost
of the archers was thus seen as excessive, especially in relation to their
duties.
The fiscal burden connected with such a force included not simply the
costs of their rations and other remuneration but also forgone state
income, since the families of the archers were forgiven certain fiscal
obligations. If the archers were called from relatively well-to-do families,
that would place the weight of the area's fiscal obligations onto the
shoulders of the poor. If the men were called from the poor, their
families could hardly afford the loss of a productive member of the
household, and in addition such men were much more likely to be
troublesome.
Adding to this excessive cost, the new plan raised the specter of the
men themselves stirring up local unrest. The edict called for appointing
an officer outside the regular civil and military services to lead each
group of five hundred archers. Chang Fang-p'ing considered such
appointments reasonable in the border regions, where external security
was a very real problem and where there was a tradition of militia
leadership. But, he said, in the inner regions of the empire, although the
local leaders were unfamiliar with military affairs, they were rich and
oppressive to their poorer neighbors. Therefore,
they unite as powerful villains and gobble up the poor and weak, doing injury to
local areas. How much worse it will be if they are openly allowed to have
management [over the archers]. Indeed, those who are called to be leaders of
groups of five hundred men will, of necessity, be important and cunning men
from the villages.137

137 HCP i3i.iob-i4a.

175
 Law and order in Sung China

Not all of Chang Fang-p'ing's suggestions were adopted, but it is clear

that the number of archers inducted was never as large as it would have
been if the law had been strictly enforced. After the emergency was over,
the number again declined somewhat, though it does not appear to have
fallen to the low level which had existed prior to the war against the Hsi-
hsia. An official in 1070 complained that the number of Archers had
been increased in the 1040s and had not been thereafter reduced, that
the archers served for indefinite terms, and that in return for training for
one month a year with the sheriff, they received complete exemption
from taxes and other services. The authorities thereupon ordered an
immediate reduction of the archer quotas in some southern circuits. 138
This reaction highlights a Sung dilemma. In times of unrest the
authorities wanted to increase the number of law-enforcement agents,
but such periods of crisis strained the state's finances, and so there was
also a contrary desire to lower the number of government-supported
functionaries. The number of police agents might be raised and shortly
thereafter suffer from pressures to reduce costs. In the case of the archers
this problem recurred during the reign of Shen-tsung (r. 1067-85).
Although during the troubles of the early 1070s more men were again
recruited, 139 the authorities were seeking some way of reducing their
cost. For a brief period the state attempted to cut the number of archers
and substitute pao-chia militia personnel. 140 In 1075, at least in Ho-pei
and Ho-tung circuits, the number of archers was to be reduced, with
fifteen to twenty being retained to serve as the sheriff's aides. The rest
of the sheriff's constables were to be drawn from the pao-chia militia
quotas. 141 Again, for a brief time in the early 1080s, when the sheriffs
were not concerned with the rural areas, the quota of archers was cut to
fifteen men. Some regions were exempt from this order, but in others the
number of archers seems to have been reduced even further. 142
This policy was soon abrogated, at least in the regions surrounding
the capital. By an order in 1085, m t n e capital and the three circuits
surrounding the capital (Ching-tung, Ching-hsi, and Ching-chi), the
sheriffs' archers and the patroling inspectors' soldiers were to be re-
established in accordance with the system in use before the advent of the

138 HCP 216.14b-15a.

139 HCP 4.23b (1077). A memorial of 1076 suggests that this order was generally
enforced, because it cites quotas of 140 archers in large districts, 100 in medium
districts, and 70 to 80 in small districts; HGP 27g.i4a-b.
140 HCP 358.4a ff.
141 HCP 267.5a; see also HCP 33i.i3b-i4a; SHY, chih-kuan 48.65a-b.
142 SHY, ping 12.18a. Sixteen districts in Ho-pei West in 1082 averaged ninety-four
archers; HCP 329.13b.

176
 Formal civil agencies of law enforcement

pao-chia militia system. 1 4 3 T h u s , just a few years after the fall of W a n g

An-shih, the archers were fully reestablished, though on occasion pao-
chia personnel were inducted as paid archers in place of those judged
incapable of doing the j o b . 1 4 4 Figures from the late Northern Sung
suggest that generally the quotas remained relatively close to those of the
1070s. 145
During the troubles of the 1120s the number of archers was again
temporarily increased, but this trend was soon reversed 1 4 6 so that quotas
of approximately the same size as those common in the Northern Sung
seem to have continued in use under the rest of the Southern Sung. In
this period the archers were paid from state funds, including money
collected especially to finance labor services. 147 In 1160, three of the four
districts of Ming Prefecture had quotas of 80 archers. T h e magistrate of
Hsiang-shan District, which had a quota of only 45 men, petitioned
successfully to have his jurisdiction's quota raised to 80, 1 4 8 and Chu Hsi,
writing about 1180, comments that in his day a large district sheriff's
office had 100 or so archers, whereas a small one would have several
tens. 1 4 9 O n e judicial intendant, writing toward the end of the twelfth
century, reported that according to his sources there were 2650 archers
in Ching-hu South Circuit, who were responsible for some nine prefec-
tures comprising thirty-eight districts, for an average of about 70 archers
per district. 1 5 0
Quotas by district population were changed several more times during
the dynasty, but in practice, in the Northern as well as the Southern
Sung, the number of archers seems to have been determined only in
part by population. It also depended on the amount of lawlessness in
the district. 1 5 1 Quotas could also be temporarily increased in times of

143 HGP 356.8b.

144 HCP 364.25a.
145 See, for example, Yiian-ch'eng District which in 1117 had 105 archers; SHY chih-kuan
48.67b. Note also that Hsiang-fu District in Kaifeng had 2 sheriffs with a total of 60
archers; HCP 311.5a—b. Some districts had much larger quotas; see HCP 12507.11a.
See also the quota of 500 men in Ch'ang-shou District in Ch'ang Prefecture in 1120s;
Sun Ying-shih, Ch'in-ch'uan chih, p. 2660.
146 SS 24.9a, 27.16a; SHY, chih-kuan 48.33b-34a. A memorial by Wang Yang (doctorate
of letters degree 1124) that criticizes the recent increase in the number of archers and
the addition of a second sheriff perhaps dates from this period. See Wang Yang,
Tung-mou chi, 9.19a-20a.
147 HNYL 15.325; Liao Hsing-chih, Sheng-chai chi, 5.23b; Pien Shih, Yu-feng hsu-chih,
SYTFCTS e d , 7b.
148 SHY, chih-kuan 48.74a.
149 Chu Hsi, Chu wen-kung wen-chi, 20.11b. Hsiieh Chi-hsiian mentions that Ta-p'ing
District had an eighty-man quota. Hsiieh Chi-hsiian, Lang-yu chi, 20.10a.
150 Ts'ai K'an, Ting-chai chi, 2.12a.
151 HCP 113.11b. For instance, in 1160 the district magistrate of Hsiang-shan District in
Ming Prefecture submitted a memorial in which he stated: "This district controls
177
 Law and order in Sung China

troubles. In 1032, for example, the quota in Mi Prefecture was raised by

twenty men because there had been a bad harvest, and the prefect feared
an increase in robbery. And in the 1120s, during the Chin invasion, the
quotas were increased everywhere. 152
Because the density of the population was greater in the Southern
Sung, the ratio of archers to the general population actually declined
from the ratio of the Northern Sung. This Southern Sung problem
worsened when the quota of archers was left unfilled. Liao Hsing-chih
(doctorate of letters degree 1184) complained that one district with a
quota of 120 men had in fact only 59 in service. 153 The drop is especially
significant in light of the increase in the number of military personnel
concerned with law enforcement in the rural areas. The number of
sheriffs did rise over time, especially during the Southern Sung, so that
many districts had two, and some more than two, sheriffs. However, the
number of patroling inspectors rose even more rapidly. Furthermore,
these patroling inspectors had complements of soldiers that were gener-
ally larger than the sheriffs' complements of archers.

Qualifications and character

The policies on selecting and employing archers provide another
example of the dilemmas faced by the Sung authorities. The early
Northern Sung government felt that it could not afford to pay the
archers from its own revenues and so had to levy the position as a form
of labor service. The burden had to be laid on prosperous households
that would not be ruined by the cost. Furthermore, moderately well-to-
do families were seen as more socially reliable than poorer families were.
However, the use of drafted men ran counter to the general trend of
Sung administration. More and more the government relied on long-
term professional personnel.
This trend toward using professional personnel itself was contrary to
some of the postulates of traditional administration, and conservative

broad and distant sea areas that abut the Wen and T'ai prefectures. In these regions
there are constantly groups of bandit boats. I have humbly noted that the sheriffs'
offices of the other five districts of this prefecture have eighty-odd archers, whereas my
district has only forty-five. I ask that in accordance with the practice of these other
districts, our quota be raised to eighty." The request was approved; SHY chih-kuan
48.74a. For further information on quotas, see HCP 131.10b-14a, 132.21b, i33.ia-b,
22399.14b, 216.14b-15a, 329.13b, 331.13b-14a; SHY, chih-kuan 48.63a, 65a-b, 67b,
70b, 86b; Gh'ien Yiieh-yu, Hsien-ch'un Lin-an chih (Gh'ien-t'ang Wang-shih chen-ch'i-
t'ang blockprint ed., 1830), 57.6a-b, 58.2b-6b; Shih Su, Chia-t'ai K'uai-chi Chih
(Chia-ch'ing blockprint ed.), 4.5b, 5.2a.
152 SHY, chih-kuan 48.33b-34a; HCP 113.1b; SS 24.446, 450, 451.
153 Liao Hsing-chih, Sheng-chai chi, 5.23b.

178
 Formal civil agencies of law enforcement

Sung officials repeatedly decried the growing tendency to substitute

other systems for those that used drafted men. These disagreements
were reflected in occasional shifts in government policy, which did
not, however, permanently affect the trend toward paying lower-level
functionaries.
At the beginning of the dynasty, at least in many areas, the post
of archer was in theory filled by drafting men from local third-grade
(moderately well-to-do) households for indefinite terms. 154 However,
under most of these drafting systems individuals were supposedly to be
called to serve for only limited terms. In IOIO a system was introduced
in some parts of the empire under which archers were drafted to serve
for three-year terms, and in 1033 the government approved a suggestion
of the important statesman Fan Chung-yen that archers be given seven-
year renewable terms. 155
This seven-year system was enforced in some places, but a memorial
by Ts'ai Hsiang, probably dating from the 1050s or 1060s, indicates
that in practice the situation varied geographically. Ts'ai Hsiang
wrote that by law, in the less-developed circuits of the south and west,
archers were to serve three-year terms but that elsewhere the terms were
indeterminate. While serving in Fukien, Ts'ai Hsiang had found the
local practice was to have the men serve for seven years and, despite the
problems associated with this practice, recommended its general use on
the grounds that it attracted experienced men willing to work for the
long term. 156
The desire for expertise no doubt contributed to the adoption of
policies such as the use of seven-year terms that would allow longer
service. Even under the three-year system that had been adopted in
1010, although the archers were required to serve for only three years,
they could volunteer for further service at the end of their term. 157
The extension of the length of actual service was accompanied by
policies allowing those drafted to hire substitutes, in the name of con-
venience for the drafted households and professionalism for the con-
stabulary. In practice, despite rules that called for drafting archers
for limited terms, from early in the dynasty if not from its beginning,
many of these archers seem to have been in for the long term. Hu Su
(996-1067) responded to Fan Chung-yen's seven-year term system:

154 SHY, chih-kuan 48.66b; HCP 73.16b. This report specifically refers to Szechuan, but I
suspect that this was a more general practice. This policy was occasionally suggested
in later periods; see HCP 285-5b-6a (1077).
155 HCP 113.15a.
156 Ts'ai Hsiang, Tuan-ming chi, 26.7a-b.
157 HCP 73.16b; SHY, chih-kuan 48.61b.
179
 Law and order in Sung China

Under the old system the districts originally established archers to guard against
bandits, selecting men who were skilled with weapons to fill the jobs without,
initially, having any regularly set term limits. Generally they wanted men to
serve for long periods so that they would become practiced and skilled with
weapons and become familiar with the local hills, waterways, and roads, and
their trouble spots. 158

D u r i n g the early part of this period such men were the paid replace-
ments of those purportedly serving as draftees, though there were orders
in 1020 and 1026 forbidding this practice (in Szechuan). 1 5 9 As a regular
practice it seems that hiring substitutes probably continued until it was
replaced by a system of state-salaried archers during the labor service
reforms of W a n g An-shih in the 1070s.
Occasionally there were specific orders reviving the drafting of men for
service in person, such as in 1040 when archers in Shensi were to be
drafted from resident village households with three or more males. 1 6 0
T h i s a t t e m p t to revive the use of draftees was connected with the great
increase in the n u m b e r of archers during the Hsi-hsia troubles. Again,
during the antireform period that began in 1085, some officials argued in
favor of returning to the practices of the early Sung, that is, drafted
service fulfilled in person.
Conservative officials argued for this return to past practices. In a
diatribe against hiring personnel, Liu Chih mentioned in passing that
these hired substitutes were frequently "guest households" (k'o-hu) - that
is, households that in n a m e at least were migrants who owned no
land. 1 6 1 H e called them "good-for-nothings." M a n y of those drafted, in
his opinion, could serve in person but were prevented from doing so by
overzealous officials who forced them to hire substitutes on the grounds
that the substitutes had experience and also were worthy of pity because
they were poor and needed the work. 1 6 2
Even when policies calling for drafted service in person were adopted,

158 Hu Su, Wen-kung chi, TSCC ed., 7.86-87.

159 HCP 96.3a, 104.14b. SHY, chih-kuan 48.62b. The order of 1020 is for Liang-ch'uan
and that of 1026 for Szechuan. These substitutes were said to be "lazy farmers."
160 HCP 126.12b.
161 The meaning of this term in Sung sources is hotly debated, in large part perhaps
because it was used with different meanings by different Sung authors. However, the
meaning given here is perhaps the best general one. See Joseph Macdermot,
"Charting Blank Spaces in Disputed Regions: The Problem of Sung Land Tenure,"
Journal of Asian Studies 44 (N984): 13-41.
162 HCP 389.6b-7b. There was also a complaint from the early twelfth century that many
of the archers v/ere old, ill, or too young or were skilled artisans; SHY, ping 12.16a.
The term k'o-hu has been the subject of much debate because in the Sung it was used
in several different ways, including immigrant household, tenant, and the like. Here,
from the context, "immigrant household" makes the most sense.

180
 Formal civil agencies of law enforcement
they were not long in force. By the late Northern Sung, important and
powerful officials had come to feel that the hiring system had proved its
worth. Su Shih, among others, argued strongly in favor of permitting
hiring. 163 A decree in 1086 reversed the previous government stand,
stating that "all those [drafted as archers] who do not wish to serve in
person may hire substitutes. Their pay is not to exceed the amounts
paid under the previous state's hiring system." 164 In these years it
appears that the very well-to-do households classified as first or second
grade were made responsible for this service.165 Thus, despite conserva-
tive criticism, hiring archers remained the general practice from this
time on. 166 At times the state used its own funds, but sometimes the
"drafted" households paid the costs (and were in return spared some
state exactions).167
Whether draftees or hired personnel, the men serving as archers were
chosen from the local population. The desire to use local people
stemmed from the authorities' belief that their knowledge of local con-
ditions and geography made them more effective policemen.168 Some
officials even considered them more trustworthy than fort soldiers,169
and in arguing (unsuccessfully) for a return to the policy of forcing
drafted households to serve in person, Liu Chih noted that such drafted
archers were familiar with local conditions, were concerned about them,
and had no incentive to flee. They would, moreover, involve their rela-
tives by blood and marriage, and even their neighbors, in the process
of apprehending criminals, for fear of the penalties that would follow
failure. "As soon as the time limit (for capturing the criminals) has been
established, everyone energetically uses his eyes and ears, going forth to
search out the bandits." 170

Training
The sheriff himself generally oversaw the training of his men. During the
second half of the eleventh century, when the number of patroling
inspectors in local areas increased, they too took an active part in
163 HCP 394.1 i a - b .
164 HCP 389.6b-7b.
165 HCP 39713a. There were conservative officials who did argue strongly for a return to
a system of drafted personal service, but their arguments were rejected by the
government; HCP 285-5b-6a.
166 HCP 285.5b-6a.
167 HCP 131.10b-14a, 331.13b-14a, 389.6b; SHY, chih-kuan 48.65a-b. A first-grade
archer could earn thirty strings of cash per year.
168 HCP 188.133-1058. See also Su Ch'e, Luan-ch'eng chi, 35.2b.
169 Tu Fan, Ch'ing-hsien chi, 8.19b-20b.
170 HCP 389.6b-7b.

181
 Law and order in Sung China

Figure 5.2. Archery practice. From Tsang Mao-hsun, Yuan ch'u hsu'an.

182
 Formal civil agencies of law enforcement

training their soldiers. 171 Such officials also sometimes helped provide
the physical facilities for training exercises. During the Northern Sung,
Tseng Kung describes a sheriff who, three months after he took office,
"enlarged the four walls of his compound to make an archery pavilion,
and when it was completed, trained his subordinate lawmen there."
Such training fields were probably a common, though not a universal,
adjunct of the sheriff's offices172 (see Figure 5.2).
The sheriffs continued to play an important role in training during the
Southern Sung. Yuan Hsieh (1144-1224) wrote:
When I was first appointed to serve as sheriff in the district, someone com-
mented that there were many bandits blocking the Yangtze. I was deeply
disturbed by this, and when I arrived and checked the rolls of the archers, I
found that there were many vacancies. Their military training had also been
neglected for a long time. Then I hired some more men, oversaw the repair of the
archery pavilion, and reviewed the practice system.173
Such constables certainly did not have the same level of training as
did the more elite units of the Sung professional military; nonetheless
they should have been more familiar with the use of weapons than were
most of the adult male members of the general population. A memorial
from the eleventh century indicates that the men were trained in the use
of "bows and arrows, swords, wooden spears, and clubs." Groups of
archers were trained in rotation, during the agricultural off-season,
and were given rations during these training periods. 174 Here, as else-
where, the rulings indicate only what was mandated. As suggested in the
passage from Yuan Hsieh just quoted, what was practiced depended on
the character and energy of the officials involved.

Functions
The main mandated task of the archers was to patrol the district,
sometimes without the sheriff, with the intention of spreading the deter-
rent effect of their presence throughout the jurisdiction. 175 For a brief

171 In 1086 an official asked that the responsibility for training archers be assigned to the
subordinate officials in the district, but earlier orders make it clear that such officials
would ordinarily be sheriffs (see HCP 73.16b, 86.17b, 364.25a). See also Chu Hsi, Chu
wen-kung wen-chi, 100.19.a-b; SHY, chih-kuan 48.61b.
172 See Sun Ying-shih, Ch'in ch'uan chih, pp. 2660, 2775.
173 Yuan Hsieh, Chieh-chai chi, TSCC ed., 9.138.
174 HCP i28.6b-7a, I 3 i . i o b - i 4 a .
175 SHY, chih-kuan 48.61b. The fact that archers sometimes patroled without the sheriff
is revealed in complaints that the sheriffs were claiming credit for capturing bandits
who actually had been caught by archers acting alone; SHY, chih-kuan 48.78a,
84b-85a. There is also one report, dated 1033, that says that patroling inspectors

183
 Law and order in Sung China

period in the 1070s, perhaps in an attempt to reduce the abuse of local

people by the constables, archers were forbidden to patrol, but with the
exception of this interruption, patroling was usually one of their prin-
cipal tasks. 176 The archers might also serve in rotation to patrol the
markets, as personal attendants for the sheriff or as jail functionaries. 177
The functions which took the archers out into the countryside reveal
one of the ongoing tensions of the traditional Chinese political process.
The state had to be able to project its power into rural areas. But its
agents all too often abused their power. Guarding the guardians was one
of the key conundrums of the Chinese past. In his book of advice to local
officials, the official Chen Te-hsiu (1178-1235) put it succinctly:
The little people of the rural areas fear the clerks like tigers. To allow clerks to go
into the villages is like permitting tigers to escape from their cages. Archers and
local soldiers must be strictly forbidden to go into the villages. Except when they
are engaged in arresting bandits, they ought not to be sent out.178
At times these men might go beyond abusing villagers while serving as
government agents. Sometimes they seized ordinary people, bound them,
flogged them, and seized their goods by claiming that the people were
bandits. 179 Sometimes they divided booty seized from criminals among
themselves and let the criminals escape. 180 Sometimes they sheltered
criminals in return for bribes. 181 And sometimes they themselves became
bandits. Speaking of a major bandit active in southern China in the late
twelfth century, the official Ts'ai K'an remarked that this man "in the
beginning was just an archer who stole arms, yet he has raised a crowd
of several thousand men, pillaged the prefectures and districts, and killed
officials. His position is really fearsome." 182
The archer gone bad was a recognized figure in fiction as well as real
life. In his book of miscellany, Sui-shou tsa-lu, Wang Kung recounted a
story in which a ghost came to seek justice for a drowned woman. The
ghost reported that two archers had chased her and caught up with her
near a small river. They demanded that she strip, and when she refused,
they pushed her into the river. In the end, with the intervention of the

sometimes used archers to capture bandits; even though the practice was forbidden
(see SHY, ping 11.14b).
176 See HCP 242.4a, 257.4b. For the continuation of such patroling during the Southern
Sung, see, for example, SHY, shih-huo I4.42b~43a. Kung-ping in this passage is a
synonym for kung-shou.
177 Sun Ying-shih, Ch'in-ch'uan chih 2709; SHY, chih-kuan, 48 6ib-62a.
178 Chen Te-hsiu, Hsi-shan cheng-hsun, TSCG ed., p. 3.
179 SHY, ping 11.15a.
180 SHY, ping 11.5b.
181 SHY, ping n . n b .
182 Ts'ai K'an, Ting-chai chi, 1.14b.

184
Figure 5.3. Weapons.
From Chi Yiin,
Chin-ting Ssu-k'u
ch'uan-shu.
 Law and order in Sung China

gods, the guilty men were brought to justice. 183 Then, as now, it seems
that at times some of those engaged in police work and those engaged in
crime were all too similar. Indeed, the constable gone bad was a staple
figure in traditional China. Civil service officials expected that their
underlings, given the chance, would abuse ordinary people.

Arms
If the archers were to be effective, they had to be at least as habituated
to the use of arms and as well armed as were the criminals whom they
were charged with capturing. To achieve this end, the authorities pro-
vided arms - bows, arrows, swords, spears, crossbows, and, in some
cases, armor made of layers of paper glued together (see Figure 5.3). 184
These weapons were given to the archers who were supposed to hand
them over to their successors when they retired from their posts. When
not in use, the weapons might be stored at the sheriff's offices, where a
system of registers was used to keep track of the weapons given out. 185
The system of having the state issue arms became policy early in the
eleventh century after an official complained that the archers in some
regions were being punished by their superiors for possessing swords and
bows of their own. 186 Liao Hsing-chih (doctorate of letters 1184) de-
scribed the situation when he arrived as a newly appointed sheriff:
When I arrived to take up office, I inspected the sheriff's equipment and arms.
We had, counting new and old, only fifteen flags, two bronze gongs, and three
drums (large and small). According to an affidavit, these all were pieces of
equipment the sheriff himself had collected. As to military equipment, armor,
and staves, there were none.187
After 1002 the archers were supposed to be quartered in barracks
which probably adjoined the offices and living quarters of the sheriff
himself or were in the side halls in which the sheriff's clerks were
housed. Yuan Hsieh outlined the real situation in his district. Puzzled by
the fact that the archers arrived at military practice at different times, he
looked for the cause and found that they lived scattered about. Some
indeed did live beside the sheriff's office, but others lived various dis-

183 Wang Kung, Sui-shou tsa-lu, Hsiieh-hai lei-pien, p. 7.

184 On Chinese armor of paper, see Needham, Science and Civilization in China, vol. 5, pt. 1,
pp. 114-16 and vol. 2, pt. 8, p. iii.
185 SHY, chih-kuan 48.61b; HCP 73.16b, 86.17b (1016), 127.2a (1040), I3i.iob-i4a; Hu
Su, Wen kung chi, pp. 85-86; Liao Hsing-chih, Sheng-chai chi, 5.25a-6.
186 HCP 86.17b (1016); SHY, chih-kuan 48.62a.
187 Liao Hsing-chih, Sheng-chai chi, 5.25a.

186
 Formal civil agencies of law enforcement

tances up to several miles away. Yuan Hsieh, however, was no ordinary

official, and he aggressively set about improving the situation.
I asked the intendant of ever-normal granaries, Mr. Lo, for seventeen hundred
strings of unspent hiring money collected in recent years. With this and using the
official fields, I could make a foundation. The intendant was delighted to help.
Our sheriff's fields were scattered. I exchanged them for fields belonging to
private owners, getting a width of thirty mou [a mou was approximately one-sixth
of an acre] that bordered our practice field. It was high, dry, level land. I divined
about its good fortune, assembled workmen, prepared materials, and checked
everything carefully. At this time the censor, Mr. Wu, strongly argued for the
value of archers and suggested that camps be built so as to redress this problem
of their living scattered about. The emperor approved. The governor, Mr. Hou,
received the order to carry it out carefully. . . . The work began in the second
month of spring in the year 1187 and was completed by the winter of this year.
In all there are 176 bays [chienh] of rooms. There is a room for the practice of
Taoist-style breathing and calisthentics and a room for the night guard. The
watchman has a pavilion. Generally, when assigning room space each man gets
one space. Meritorious men get one and a half or two, and those with particular
labors three. Those who do not yet have a room share with another. In this way,
those who had been scattered about are [now] gathered together.188
Like the clerks, groups of archers serving by half-month turns pro-
vided personal services for the sheriffs. As with other groups of sub-
ordinate personnel, local officials sometimes abused this system and used
the archers in excessive numbers as miscellaneous workers. 189

Grading and remuneration

The state also had to establish policies that would keep men in service as
archers and would encourage them to perform effectively. At a minimum
the authorities had to provide the archers with a livelihood. Because the
archers differed in the length of their service and in their skills, they were
divided into grades and were remunerated accordingly. 190 The informa-
tion on their rates of pay is scattered and contradictory, but it appears
that during the late Northern Sung the average rate of pay was approxi-
mately thirty strings of cash per year for a first-grade archer. 191 During

188 Yuan Hsieh, Chieh-chai chi, 9.138.

189 HCP i 3 i . i o b - i 4 a .
190 SHY, chih-kuan 48.i27b-i28a.
191 Some of the confusion perhaps stems from differences in the names and types of
remunerations involved. For example, Ch'en Ch'i-ch'ing, writing in 1220, describes a
sheriff's office as receiving thirty strings per month for rations, with a supposed quota
of fifty-two soldiers (though in fact most were missing), whereas a report from 1127
says that each archer was to receive three strings per month of ration money (SHY,

187
 Law and order in Sung China
the brief period in the early 1080s when the sheriffs were not responsible
for rural areas, the archers' salaries were supposed to have been reduced.
Officials argued that the new policy had spared the archers the cost of
traveling. But even in this period they are still said to have been paid
thirty strings per year. 192 This accords with reports stating that upper-
grade households were paying substitutes at the rate of thirty to forty
strings. 193 This pay scale increased very slowly. A report dated 1219
from Hsi-an District in Liang-che Circuit indicates that the average
archer under the newly established western sheriff there was to receive
fifty-three strings of cash per year. 194 Archers might also receive special
rewards if they demonstrated military skills during inspections by
the intendants. 195 Furthermore, the households to which the archers
belonged were spared various miscellaneous government exactions. 196
Figures from Robert Hartwell suggest that the buying power of those
strings of cash, at least in terms of rice costs, had declined greatly, so
that despite the increase in salary, the archer of the early thirteenth
century was less well off than his predecessors of the late Northern Sung
had been. 197
The archer system, like the militia system devised during the Sung,
was intended to reduce the government's costs as well as to provide on-
the-spot protection against disorder. Nonetheless, it was itself open to
criticism, since those inducted were given rations during training and
their families were freed from some taxes and from labor services. 198 The
brief attempt during the reign of Shen-tsung to substitute pao-chia per-
sonnel for the archers is probably best viewed as an attempt to cut costs,
but the archers proved to be irreplaceable, and in the late Northern and
the Southern Sung the state responded by using regular state funds to
pay the archers' salaries. These funds were usually taken from the
special labor service moneys collected by local administrative units. 199
These payments were necessary simply to secure minimal services,
chih-kuan 48.33b—34a). It seems likely that either the report of Ch'en Ch'i-ch'ing is
garbled or the men at that time received pay other than so-called rations. Another
confusing report, from Ch'ang-shou District in Liang-che Circuit, states that the men
received about nine strings per year (see Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih,
2704).
192 HCP 331.13b-14a; SHY, chih-kuan 48.65a-b says that first-grade archers received
thirty strings per year in 1082.
193 HCP 424.10a.
194 Thirty archers were assigned to this new office, with an annual salary total of 1602
strings, an average of 53 strings per man per year (SHY, chih-kuan 48.86b-87a).
195 TFSL 82.
196 HCP 131.10b-14a.
197 Personal communication from Robert Hartwell, April 1989.
198 HCP I3i.iob-i4a, 2i6.i4b-i5a.
199 See, for example, Pien Shih, Yiifeng hsii-chih, p. 7b; HNYL 15.325.

188
 Formal civil agencies of law enforcement
and even they sometimes were not given. At times, when the archers
were not paid on time, they preyed on the people in order to survive. 200
The state also faced the problem of motivating its agents to act effec-
tively and again fell back on a system of rewards and punishments.
Archers who captured criminals were rewarded. However, like their
official superiors and clerical colleagues involved with capturing crim-
inals, they were liable for punishments, including beatings, if they failed
to capture the criminals within the set three time limits. 201
The other side of motivation was the system of control. It was hoped
that the archers would do a satisfactory job, but some way of supervising
them was clearly necessary. In his early twelfth-century book of advice
to local officials, Li Yuan-pi recommended that the district magistrate
have a list of the archers' names. When they were dispatched for duties,
they should be sent in the order of the list. This list should also note
which men were capable of being used in serious situations. 202

Conclusion
Among Chinese dynasties the Sung is especially noteworthy for the
prestige, honor, and power given to civil bureaucrats by the ruling
house. Rulers of some other dynasties gave more honor to the military,
and others sometimes treated their civil bureaucrats harshly. In preced-
ing Chinese and later foreign-dominated dynasties, prestige might derive
as much from birth as from achievement. But in the Sung the civil
bureaucrats - the important ones being mostly graduates of the civil
service examinations - were treated with consideration and respect by
the emperors and had a dominant hand in shaping policy. I do not mean
that there were not some activist emperors who personally participated
in making major policy, but clearly the careers and lives of Sung bureau-
crats were safer, more honorable, and more powerful than were those of
bureaucrats in many other eras.
This situation was in part due to deliberate state policy. In seeking to
reduce the role of the military in politics, the Sung founder and his
successors deliberately fostered the prestige and power of civil officials.
Their treatment of the problem of maintaining law and order makes this
abundantly clear. Law-and-order functions in the dynasties before the
Sung had often been mainly the responsibility of the military. In the
Sung, however, although the authorities continued the tradition of using
the military to enforce law and order in the cities, from the early years of
200 CHTK 2.26.
201 HCP 97.11b.
202 TITC 2.9a.

189
 Law and order in Sung China

the dynasty the emphasis in all areas outside the cities was on using
civilian agents in law enforcement. The principal civil service figure was
the district sheriff with his civilian constables. The office of sheriff, as
revived by the Sung founder, resembled in its functions the office of
sheriff during the dynasties before the Sui, and it was usually staffed by
civilian officials. These men and their constables were to maintain order
by patroling regularly in their jurisdictions and by actively pursuing
known criminals. The sheriffs were also involved in some facets of
the investigation process, including inquests. Because they were civil
officials, they were seen as competent to perform a variety of minor
bureaucratic tasks, and indeed the state regularly assigned to the sheriffs
some bureaucratic functions or appointed one official to serve con-
currently as sheriff and recorder.
During the dynasty the growing role of the military in low-level law
enforcement in rural areas presumably reflected both changes in the
problem of maintaining law and order and the increasing confidence
among the higher civil officials that the military was firmly under con-
trol. In terms of distributing functions, these state policies regarding law
enforcement may seem well designed to achieve their goals. However,
because the position of sheriff was very low in the hierarchy of civil
officials and because the post combined this low position and prestige
with duties that could be unpleasant and were often dangerous, the
higher authorities often found it difficult to get enough good men to
fill the vacant posts. This difficulty may have contributed, too, to the
trend toward a more even balance between the law-enforcement role of
the civilian sheriff and that of his military counterpart, the patroling
inspector.

190
 6
The role of the military
in law enforcement

Introduction - wen versus wu

"The military and punishment are a single Way." 1 With this simple
declaration a Sung official put the Chinese assessment of the role of the
military in law enforcement in cameo. Since one facet of law enforcement
is bringing to bear force sufficient to deter or suppress lawbreaking, the
military clearly must play an irreplaceable role in dealing with disorders
too large in scale to be dealt with by the regular police system. Although
many modern states are almost always able to handle disorder without
calling on the military, the use of military force always remains as a last
resort. What was characteristic in the use of the military by many
traditional Chinese dynasties, of which the Sung is a good example, was
the integration of the military into the ordinary institutions of law
enforcement. The military and punishments were seen as a single way in
that they both embodied the sovereign's forceful responses to those who
would disturb the social order. The distinction between police and the
military (which the Chinese did recognize) was rooted in differences in
the scope of the threat, not in differences in the type of threat.
The role of the military in law enforcement was especially delicate in
the Sung. Early Sung analysts, who sought ways of reducing this power,
traced the social and political problems of the late T'ang and the Five
Dynasties to the delegation of excessive power to the military. This
reaction against excessive militarism was part of an overall pattern. In
the second century B.C. the great Han historian Ssu-ma Ch'ien had
searched for patterns in the Chinese past. He found there the continuing
interaction of the unchanging, the cyclical, and the evolutionary. Some
changes, he said, are caused by the leaders of each new dynasty as they
seek to learn from the past and avoid its mistakes. Such reforms might

1 Liu Cheng, Huang Sung chung-hsing liang-ch'ao sheng-cheng (Taipei: Yee Wen, 1967), 10.12a.

191
 Law and order in Sung China
be cyclical, as each succeeding dynasty reacts against the excesses of its
predecessor, only to push its own choices too far. Where there were pairs
of complementary opposite tendencies, there would be an alternation of
emphasis from era to era. When one tendency - military interference in
politics, for example - had been pushed too far, a natural and strong
impetus toward stressing the civilian side of government would arise.
In traditional Chinese thinking, the world of politics and government
was marked by tensions between the civil and the military, the wen and
the wu. Members of the elite in traditional China distinguished between
the nature and function of these two facets of the political world. To
understand how the Sung Chinese felt about this distinction, we must
recognize the associations surrounding the key terms. The philosopher
J. L. Austin remarked that "a word never - well hardly ever - shakes off
from its etymology and formation. In spite of all changes in and exten-
sions of and additions to its meaning, and indeed rather pervading and
governing these, there will persist the old idea." 2 The truth of this
observation is especially striking to anyone who reads classical Chinese,
in which words have many layers of meanings. Newer usage may overlie
the meanings of the past, but these earlier layers almost always color the
reaction to the newer meanings.
The character wen in its original form was perhaps a representation of
lines or cracks that came to mean ornamentation. It was used to describe
cultivated, civilized things, as epitomized in writing, and was set off in
Chinese minds against the character wu, which represents a footman
holding a lance and stands for all that is martial and military. These
concepts must not be understood as contraries. Rather, we would do
better to envision the totality of felt reality for the Chinese official elite of
the Sung (if not always of other dynasties) as a bounded solid figure,
with wen at the core and wu at the periphery. No rigid internal boundary
separated the spheres of the two concepts. At the center of the solid, wen
was pure but contained the possibility of wu, and at the surface wu was
pure but retained the possibility of wen. Between these contrasting points
there was no fixed borderline.
Among the leaders of the Sung, who embodied this tradition that
contained both wen and wu? During the Sung, men representing wen, the
civil side of this complementary pair, dominated government and society.
Those representing wu were both topologically and psychologically at the
periphery. In dealing with hsieh, or "deviance," the use of civilizing

2 J. L. Austin, "A Plea for Excuses," in J. L. Austin, Philosophical Papers (Oxford:

Clarendon Press, 1961), p. 149.

192
 The role of the military
change, wen, was the preferred response. This commitment to using
wen coexisted with the recognition that deviance could reach a point
at which resolution was beyond the short-run power of peaceful and
civilized transformation. At that point it became necessary to invoke
force.
Because there was no radical break between these two concepts, it was
certainly possible for carriers of the wen tradition to employ force in its
defense, and men, basically civil in their orientation, could and did
function as military commanders when needed. Only when the use
of force by nonmilitary or paramilitary groups was insufficient did
it become necessary to involve the specialists in wu, the professional
military. Thus the line between persuasion and force was not identical
with the line between civil and military. The military system depended
almost exclusively on the use of force (or the threat of force). The civil
authorities combined persuasion and education with the measures of
intimidation, including the use of force, embodied in the police and
judicial systems.
The Chinese viewed (and continue to view) deviance as a continuum
measured by the degree of the threat to the existing order of society. At
the low end of the continuum, because the deviance did not constitute an
immediate threat to the social order, the proper response was persuasive,
educational, and largely the responsibility of social associates. As the
degree of threat increased, the state gradually began to apply force
through the police function of the civilian government. Finally, when the
threat was too immediate and too extreme for the civil authorities, it
became the responsibility of the military.
During the Sung the harmonious working out of this tension between
the civil and the military intersected with the problem of self-cultivation.
The contrast between the wen and the wu, when embodied in individuals,
was the contrast between the exemplary gentleman (chun-tzu) and the
hero. Different thinkers might have somewhat different visions of the
Confucian gentleman; they generally agreed that the gentleman embodied
balance and measure and cultivated sobriety. Even when officials who
were basically committed to the chun-tzu ideal took up arms, as they did
on occasion and sometimes with marked success, their attitude toward
the use of force and the characteristics they admired in people were
radically different from those of the heroic warrior. During the Sung,
even in the aesthetic-cultural wing of Confucians led by the Su brothers,
there was a lack of extravagance and a concern for propriety. 3

3 For this group of Confucians see Bol, "Culture and the Way."

193
 Law and order in Sung China

The hero
In contrast, the traditional Chinese also had a vision of the hero. Even in
his Confucian incarnation the hero was a person who lived beyond the
ordinary strictures and everyday rules of his contemporaries. As pre-
sented in Chinese literature and as imitated in life, heroes might be great
warriors or generals, wandering knights-errant (yu-hsia), or even ban-
dits, but they all shared a larger-than-life quality. Among knights-errant
this heroic quality was expressed by an altruistic and total commitment
to seeking justice as they understood it, even beyond the boundaries of
normal morality, in defense of the weak and oppressed. The knights-
errant, in the common Chinese phrase, "seeing an injustice on the road,
would pull out their swords to help." Their contempt for ordinary
conventions was often reflected in their haughty manner, absence of
courtesy, and dissolute style of life, but this was coupled with a total
commitment to their friends and acknowledged superiors, great per-
sonal courage, and a deep belief in truthfulness and mutual good faith.
Generous to a fault, contemptuous of wealth, they sought honor and
fame rather than the security and comforts valued by ordinary men.4
The deep sense of honor, the commitment to justice, and the loyalty to
friends in these heroic figures were linked in Chinese minds with a
penchant for disorder. The Sung founder, a hero in numerous stories,
exemplifies this tradition in fiction. In "The Sung Founder Escorts
Ching-niang One Thousand Li," the Sung founder is called a "chivalrous
man, who would not hesitate to draw his sword, and offer help to right a
wrong," but the next sentence calls him "a master of busy bodies and a
leader of troublemakers."5 With minor variations, this same set of values
and characteristics would fit the bandit-heroes of Chinese fiction or the
great military figures of the Chinese past.
These traits were not mere stigmata of stock characters in popular
entertainments. Men could and did seek to model themselves after this
heroic image. The Sung dynasty produced perhaps the greatest of all
these military heroes - a man who set out to live a life that embodied
these great virtues - General Yueh Fei (i 103-41). His career, from the
time when he first took up arms at the age of twenty until his death two
decades later, is the stuff of which myths are made. The picture of him
that has come down to us presents a pattern characteristic not just of
Chinese heroes but also of many heroes in many cultures:
4 For a more thorough description of these men and their values, see James J. Y. Liu, The
Chinese Knight-Errant (Chicago: University of Chicago Press, 1967).
5 For information on this story, see Y. W. Ma and Joseph S. M. Lau, Traditional Chinese
Stories (New York: Columbia University Press, 1978), p. 61.

194
 The role of the military

The forceful youth of almost unknown family background, as proud as he is

naive, who overnight gains entrance into his chosen vocation; the loyal attach-
ment to those who guide him in his activities; his somewhat ostentatious display
of all the shining virtues the tradition demands; his always successful battles
against the enemy which is devastating his country; his sudden reversal of
fortunes at the peak of his success; and his end at the hands of villains of blackest
hue.6
A poem by Yueh Fei, reflecting his feelings about the foreign invaders
then ravaging north China, epitomizes his commitments and desires and
contrasts starkly with the measured pattern of nonheroic figures:
My hair bristles in my helmet
I lean against the railing, the pattering rain has ceased.
I raise my eyes, and toward the sky I utter a long-drawn shout.
My breast is filled with violence.
At the age of thirty fame and merits are but earth and dust,
Eight thousand miles of land are like the moon covered with clouds.
Do not tarry! The hair of youth grows white.
Oh, vain sorrows.
The shame of the year Ching-K'ang [i 126] not yet wiped away,
When will the hate of the subject come to an end?
Oh, let us drive endless chariots through the Ho-lan Pass.
My fierce ambition is to feed upon thefleshof the Huns,
And, laughing, I thirst for the blood of the Barbarians.
Oh, let everything begin afresh.
Let all the rivers and mountains be recovered,
Before we pay our respect once more to the Emperor.7
The self-cultivation of the hero thus differed strikingly from the self-
cultivation of the Confucian gentleman.

Reaction to T'ang militarism

In the eyes of early Sung leaders, the T'ang dynasty (618-907) stood as
the exemplar of both the glories to be gained by military excellence and
the price to be paid for military excesses. During the T'ang, military
ideals permeated the ruling class, affecting their behavior and their
understanding of and reactions to the political and social questions of the
day. During the early T'ang this pattern had resulted in a vibrant and
6 Hellmut Wilhelm, "From Myth to Myth: The Case of Yueh Fei's Biography," in Arthur
Wright and Denis Twitchett, eds., Confucian Personalities (Stanford, GA: Stanford
University Press, 1962), p. 146. For his career in more detail, see Edward Kaplan,
"Yueh Fei and the Founding of the Southern Sung" (Ph.D. diss., University of Iowa,
1976).
7 Wilhelm, "From Myth to Myth," p. 156.
195
 Law and order in Sung China

creative culture. But after the rebellion of An Lu-shan in the middle of

the eighth century, the pendulum swung too far in the direction of
transferring dominant military power into the hands of local leaders.
This pattern became even more blatant after the rebellion of Huang
Ch'ao in the late ninth century, and was characteristic of the Five
Dynasties period that followed that end of the T'ang in 907.
The leaders of the Sung wanted to swing the pendulum toward an
appreciation of the civil. Accepting the idea that the principal problem of
the recent past had been an excessive emphasis on the military, the Sung
founder, himself a usurping general, continued the reforms that had been
initiated by his predecessors. Following the pattern of the Later Chou
(951-60), he selected men to form the imperial armies (chin-chun) by
choosing the best soldiers from all his units. The remaining soldiers were
organized into the provincial armies (hsiang-chun). The imperial armies
functioned as the key fighting forces. Although some units of the
provincial armies did occasionally take part in the fighting, they were
primarily a reserve of state laborers.
Despite the central authorities' concern about the dangers of military
power, at the beginning of the Sung the military still played a significant
role in law enforcement. Military men serving as garrison commanders,
holdovers from the military-dominated system of the Five Dynasties,
initially contiriued to have a significant day-by-day role in maintaining
order, but the general trend was toward reducing their importance.
Edicts issued under the Sung founder deprived the garrison commanders
of much of their early power. Their loss of power was not, at first,
matched by an increase in the power of other military figures. By the
middle of the eleventh century, however, the curbing of the military on a
national level had the unintended side effect of making it politically
feasible to increase again the role of the military in low-level law enforce-
ment, if and when this participation were desirable.
From the 1040s onward, there seems to have been a deterioration of
local order in rural areas, accompanied by the greater use as police
officials of the military officers called patroling inspectors (hsun-chien).
Thus, the role played by the military grew in importance over time until
by the late eleventh century, the Sung had created a dual civil-military
police system in most regions. During the Southern Sung the military
came to dominate rural law enforcement, at least in the number of
officials.
The military played several roles in maintaining order in Sung China.
It enforced a special military law on its own members 8 and might
8 See Brian E. McKnight, "The Military and the Law in Sung China," Second International
Sinological Conference, Transactions (Taipei: Academia Sinica, 1990), pp. 855-74.

196
 The role of the military

sometimes be called upon to enforce this military law against the general
population. The military patroling inspectors became increasingly
important in regular day-to-day law enforcement. Finally, the military
played a major role in enforcing the regular civilian laws by suppressing
large-scale lawbreaking by the civilian population. In law enforcement,
the military represents a means that is significant not because it is
necessarily close to the scene of crime in either space or time but because
it can bring to bear more than sufficient force. The police agents of the
civil authorities - not to mention informal and semiformal agents - were
often inferior in numbers, equipment, and training to the criminal
gangs that they faced. Indeed, a Southern Sung official, commenting on
the suppression of bandits in Ching-hu North Circuit, pointed out that
the victims "summoned the adult males of the powerful families from the
mountain areas, but they still could not handle the bandits. An effective
response came only with the arrival of a large army from E Prefecture.
Thereafter the bandits were destroyed." 9
At the same time and by the same token, the commitment of the
military to the basically civil values of the dominant tradition was
suspect. Many officers as well as almost all soldiers were drawn from the
lower levels of Chinese society. The civilian authorities, themselves
drawn overwhelmingly from the upper levels of Chinese society, hesi-
tated to summon the military unless pressed to do so, since they feared,
justifiably, that the military might be more given to persecuting the
hapless peasantry than to fighting the bandits. Still, the military was
superior in weaponry and technical skills. The authorities wanted to use
the strengths of the military to advantage without paying an excessive
price in terms of military abuses. In the end, therefore, they adopted
policies that allowed the regular use of relatively small numbers of
soldiers serving under the patroling inspectors, most of whom during the
greater part of the dynasty were concerned almost exclusively with law
enforcement. (Earlier in the dynasty patroling inspectors had been
commanders of major military forces.) Larger units of the regular army
were not called in unless the local situation threatened to get out of
hand. The general rule, restated in a decree of 1084, was that only when
the prefecture and districts could not themselves handle the situation
could they call on other regular army units. 10 In the words of Chang
Shou (d. 1138):
When bandits arise, the responsibility falls on the sheriffs and patroling inspec-
tors. If they cannot handle them, then the responsibility shifts to the prefects and
their assistants. If the prefects and their assistants cannot handle the problem,

9 Ts'ai K'an, Ting-chai chi, 2.13a. 10 HCP 348.10a.

197
 Law and order in Sung China

the responsibility falls on the circuit intendants and the circuit military leaders.
If they cannot handle it, then the generals are mobilized.11
At the lowest level the civilian character of the sheriff was balanced
against the military character of the patroling inspector. On the inter-
mediate level of the prefecture this tension between wen and wu was
embodied in a single individual, the prefect, who was most often con-
currently the commander of local forces. And at the highest regional
level the civilian orientation of the circuit intendants was again balanced
against the more military orientation of the circuit military leaders.

Patroling inspectors
Although the sheriff was the initially responsible official charged with
suppressing criminals, he was not alone in carrying out his law-
enforcement duties. The garrison troop commanders themselves retained
at least some of their former powers during the early decades of the
dynasty. Furthermore, several sorts of military officers might take part
in bandit suppression, especially when gangs were large or active in
strategic areas. The most important officers involved on an everyday
basis were the patroling inspectors.
The term for the office of patroling inspector had long been in use as a
verb, meaning "to inspect" or "to investigate," but as the name of an
office, it appears to have been first used during the chaotic period at the
end of the T'ang. Certainly it was in use during the opening years of the
Five Dynasties. Such officers at this time seem mostly to have been men
dispatched by the central authorities to deal with foreign incursions, to
pacify disturbed localities, or to act as urban police functionaries in the
capital. During the Five Dynasties the role of these men continued to
evolve. They were appointed to local areas, to prefectures, and even to
districts, where they had charge of units of regular army troops. They
began to deal less with outside incursions and the pacifying of disturbed
areas and more with regular law enforcement, particularly including
the enforcement of state monopoly policies. The patroling inspectors'
activities may have contributed to the gradual weakening of the military
commissioners' powers and the strengthening of the central government,
but they also undercut the power of the regular local officials. Their
flexibility was due in part to the fact that the office was not a regular
post in the bureaucracy but, rather, a kind of commission. 12

11 Chang Shou, P'i-ling chi, 1.14.

12 For the early development of the office, see Habu Kenichi, "Godai no jun-ken shi ni
tsuite" (On the patroling inspectors during the Five Dynasties), Toyogaku 29 (1965):

198
 The role of the military
At the beginning of the Sung dynasty the officers bearing the title hsiin-
chien seem to have been major military commanders. 13 The hsiin-chien
frequently participated in the campaigns that led to the unification of
China under the Sung. 14 These relatively important military officers
were often titled superior patroling inspectors (tu hsiin-chien), and at least
early in the dynasty they often held important local offices concurrently,
sometimes including the post of administrators of one of the prefectures
under their military jurisdiction. 15 A superior patroling inspector might
have ten thousand men under his command. 16 During the early decades
of the dynasty, the lowest-ranking patroling inspectors were responsible
for several districts, whereas the highest-ranking ones might have charge
of as many as ten prefectures. 17
Over time the size of the jurisdictions of the lower-level patroling
inspectors shrank while their numbers increased. By the late eleventh
century the lower-level patroling inspectors, with perhaps one hundred
regular army soldiers under their command, seem to have become
merely military and rural counterparts of the sheriffs. Appointing patrol-
ing inspectors to small jurisdictions continued to be characteristic of the
system during the Southern Sung. Documents from the early thirteenth
century even note a superior patroling inspector with responsibility
for an area containing only three hundred households (his post was
abolished), and from this same era the sources refer to patroling inspec-
tors subordinate to individual districts. 18
Despite occasional and temporary reversals, this trend toward increas-
ing the number of patroling inspectors while reducing the size of their
jurisdictions continued from the late tenth to at least the late eleventh

45-55. The term itself could be translated in various ways. I am following the lead of
the Japanese scholars writing on the office in assuming that the word hsiin should be
read to mean "patroling." See also Wu Chien-fan, "Sung-tai hsiin-wei liang-ssu chih"
(The system of the two offices of sheriff and patroling inspector in the Sung Dynasty),
in Fa-chih shih yen-chiu shih, comp., Chung-kuo ching-ch'a chih-tu chien-lun (Beijing:
Ch'un-tsung ch'u-pan she, 1985), pp. 174-75. Patroling inspectors continued to be
important under later dynasties. Under the Ming they were closely identified with
economic controls, but because of their corrupt activities they were eventually
displaced. See Kawakatsu Mamoru, "Myodai jin shi no shi saku to jun ken shi seido"
(Market towns, watergates, and the system of patroling inspectors in the Ming
Dynasty), Tqyogaku 74 (1988): 101-15.
13 For a general description of the Northern Sung hsiin-chien see Habu Kenichi, "H0-S6
no jun ken to hoko ho" (Patroling inspectors of the Northern Sung and the pao-chia
system), Shien 92 (1962): 165-75.
14 See, for example, HCP 1.12a, 4.14b, 5.4b.
15 HCP 7.11a, 10.10b, 12307.ioa-b.
16 HCP 151.10b.
17 For more on these high offices, see Sun Feng-chi, Chih-kuan fen-chi, SPTK ed., 3543a-
48b.
18 SHY, chih-kuan 48.135b.

199
 Law and order in Sung China

century. Although early in the dynasty there had been calls for reducing
the numbers of these men as one part of more general calls for cut-
ting the total number of local officials,19 their numbers were gradually
increased through the piecemeal creation of new posts and the addition
of new personnel to old posts. 20 Then in 1058, a key order greatly
expanded the number of patroling inspectors. In that year it was stipu-
lated that each prefecture should have from three to five regular patrol-
ing inspectors, and each group of from three to five prefectures or up to
eight or nine prefectures should have a (military) director in chief
concurrently serving as a patroling inspector (tu t'ung hsiin-chien) or an
inspecting general for seizing bandits (chu-po tsu-tsei).21 (The office of
military director in chief was normally a concurrently held title that
gave territorially defined civil officials control over the military forces in
their jurisdictions. 22 )
The Sung authorities used a variety of policies to try to resolve the
jurisdictional issues created by the simultaneous presence of both civil
and military police officials in many regions in the late Northern and the
Southern Sung. By this time the difference between the bureaucratic
rank of the sheriffs and that of the low-level patroling inspectors had
become negligible. In a report of 1083 t n e t w o offices are referred to
together, using the abbreviation hsiin-wei, an expression that thereafter
becomes more and more common in the sources. 23 Because they both
were offices acting on the same level and concerned with the same task,
the possibility of conflict was certainly present. However, in some dis-
tricts this was not a problem because there were no patroling inspectors;
in others the jurisdictional boundaries were drawn geographically, with
the sheriff handling one part of the territory and the patroling inspector
another. In the thirteenth century the divisions occasionally were func-
tional, with the patroling inspectors handling smuggling, illegal salt
dealing, and so on, and the sheriffs handling ordinary banditry. 24
Most commonly the sheriffs controlled the towns, and although both
sorts of officers might work in rural areas the patroling inspectors seem
to have been concerned especially with large-scale lawbreaking such as
banditry and with security in the immediate environs of their head-
quarters. Thus, in most parts of the empire in districts where both
officials were present, the sheriffs were more closely tied to maintaining

19 HCP 43.16b.
20 See, for example, HGP 74.6a.
21 HCP 183.13a; SHY, chih-kuan 4 8.i28a-b.
22 See Hucker, Dictionary no. 4687.
23 HCP 356.8b.
24 HCP 129.5a, 139.7a, 141.13b.

200
 The role of the military

order in towns and markets, and the patroling inspectors were especially
concerned with rural crime. This was, of course, only a general pattern.
There were exceptions, since patroling inspectors were stationed with
some regularity in market towns, and from the Southern Sung onward
there are reports that both sheriffs and patroling inspectors were sta-
tioned in the towns. 25
For this system to work effectively, both offices had to cooperate to
deal with serious problems. Here as elsewhere in the Sung government,
there was concern about two quite opposite problems. First, the offices
simply might not cooperate, with potentially disastrous results. 26 Second,
they might cooperate all too well. The Sung authorities were always
haunted by the dangers of collusion between officials. Thus, the sheriff
and the patroling inspector were to cooperate in capturing bandits, but
not to deal with each other for any other reasons. According to the
bandit-capturing ordinances, "except when it is matter of cooperating to
capture bandits, the patroling inspectors are not allowed to exchange
documents with the sheriffs' offices."27
Although the general line of evolution is clear, there were occasional
experiments with different systems in some areas of the empire. In the
early 1080s, for example, an attempt was made to confine the sheriff's
jurisdiction to the district town and the district's markets, with the
patroling inspector having sole control over rural areas, at least in the
region surrounding the capital, including Ho-pei Circuit directly north of
Kaifeng.28 However, a few years later the new system was modified. An
official pointed out that along the borders, where the sheriffs were chosen
from military men, the situation differed from that in the interior of
the empire. He asked that in such areas the older system under which
the two offices shared responsibility in rural areas be revived, though the
sheriffs were still not to participate in training troops. 29 By the late 1080s
the policy of confining the sheriffs to the towns and markets seems to
have been abandoned.

Duties of patroling inspectors

Patroling inspectors were to be principally concerned with the actual
seizure of wrongdoers, especially bandits or military deserters. 30 They

25 See, for example, HCP 118.3a (1036); Hsueh Chi-hsuan, Lang-yu chi, 2 0 . 9 b - n a .
26 HCP 52.15a (1000).
27 TFSL 237.
28 HCP 311.5a.
29 HCP 324.13b; SHY, chih-kuan 4 8 . 6 ^ - ^ .
30 HCP 12307.10b, 23.17b.
201
 Law and order in Sung China

thus did not under ordinary circumstances share the evidence-gathering,

inquest-holding, and general detection functions of the sheriffs. A decree
pertaining to the patroling inspector in 1006 reveals their general
character:
Two servitors are to be selected to serve as patroling inspectors at Chang-ch'eng
k'ou. Each is to be given one hundred soldiers. They are to divide their jurisdic-
tion into two circuits and patrol [them]. This was done because the people of the
border often concealed forbidden goods or stole border-area horses.31
The inspectors thus were to be concerned with the kinds of organized
crimes represented by the smuggling of illicit goods or, near the border,
the stealing of horses.32 Also, their principal focus was on groups of
criminals rather than on individuals. One report, from 1076, shows that
the inspectors may have been called out mainly to deal with gangs,
stating that when ten or more bandits were involved, the soldiers were to
be dispatched. 33 This emphasis is consonant with the state's general
position, that the amount of force to be used should be adjusted carefully
to the dangerousness of the lawbreakers. The authorities wanted to limit
the amount of discretion granted to those who controlled major amounts
of force.

Training of troops
T h e training of troops to support the law-enforcement process raised
again the dilemma of creating men skilled in the use of violence whose
ethical commitments were in doubt. Such men were both necessary a n d
dangerous. T h e authorities treated the commitment of troops to accept-
able values as an accomplished fact, and they used military officers to
train them in violence. In connection with their duty to seize bandits and
other lawbreakers, the patroling inspectors also had to visit military and
paramilitary units to train or oversee the training of troops, 3 4 including
those in the pao-chia militia (though some patroling inspectors were said
to be incompetent to do so). 3 5 This general duty to supervise the training
of troops was the inspectors' from early in the dynasty. A decree of 1015
ordered that the "prefectural administrators who are concurrently
patroling inspectors in border areas are to go monthly to the various
army encampments to oversee training." 3 6

31 HCP 64.3a.
32 HCP 64.3a.
33 HCP 273.1 ia.
34 HCP 84.12*3-1015.
35 SHY, ping 2.24a.
36 HCP 84.12b describes a border-patroling inspector who was promoted to be an

202
 The role of the military
Patroling inspectors were first and foremost law-enforcement officers.
Indeed, they seem to have been more exclusively concerned with the police
aspects of maintaining law and order than were their civil colleagues, the
sheriffs. Still, like others on the lower levels of the bureaucratic system,
the patroling inspectors could sometimes be multifunctional. This policy
of using them for nonpolice duties both reduced government expendi-
tures and diffused the authority of the patroling inspectors, by giving
them duties in several areas. In addition to seizing wrongdoers, they
might also under unusual circumstances be used as investigating agents.
When expedient, patroling inspectors might be sent to investigate re-
ports of corruption by local administrators, 37 to report on the extent of
natural disasters, 38 to assess possible problems among refugees,39 or to
study other local problems. 40

Interference in trial functions

The problem of the delicate balance between giving such men so little
power that they could not perform their legitimate functions effectively
and giving them so much power that they could be abusive was reflected
in their penchant for intervening in areas beyond their purview. At times
during the early years of the dynasty, these patroling inspectors took the
law into their own hands, interfering in trial processes and punishing
criminals when not authorized to do so. In 965 the patroling inspectors
of Szechuan and Shensi were ordered to stop interfering in legal cases in
the prefectures and districts, 41 but such interference clearly continued.
Orders dated 975 and 982 expressly forbid them from holding and
interrogating suspects. 42 At about the same time patroling inspectors
and supervisors of militia who had seized salt smugglers were forbidden
to punish them summarily. Rather, they were to send the suspects to
their superior jurisdiction for trial. 43 As late as 1014 we have an order

audience-attending usher, a position ranked as a servitor major, as a reward for having

captured twenty-two bandits.
37 WHTK 167.1448.
38 HCP 155.2b (1074).
39 HCP 92.14a.
40 In one case, a police commissioner was sent to learn why certain gold mining
operations were not profitable. SS 201.5018.
41 HCP 6.15b. See also SHY, chih-kuan 48.122b (967).
42 HCP 12307.10b, 23.17b. Interference of military officials in the trial process was not
confined to patroling inspectors. A report from 972 accuses the headquarters of the
inspectors general of the armies (Ma-pu yuan) - men supposedly charged only with
maintaining discipline in the armies - of conducting trials in some prefectures. This was
strictly forbidden. See HCP 12306.11b.
43 HCP 8.3b.

203
 Law and order in Sung China

concerning the patroling inspectors in the capital regions, forbidding

them from passing sentence on and punishing certain offenders in the
city.44
This tendency to act in a high-handed way, however, was difficult to
suppress. As late as 1032 a superior patroling inspector reported that he
had, on his own authority, executed six recidivist bandits by slicing.
His report prompted an order that in future such cases were to be
memorialized first.45 Such complaints do not appear in the record from
later reigns, but because our sources are fragmentary, it would be
difficult to argue from this absence of evidence that the problem had
disappeared. And like the other agents arresting suspects, the patroling
inspectors are said to have abused their powers on occasion by seizing
commoners in the process of seeking criminals. 46 The extent of their
power over ordinary people is reflected in an incident recounted by the
official Hsu Ching-heng:
I have heard that a commoner named Chu Chin from the district of Hsiang-fu in
the capital city was seized and dragged away forcefully by four soldiers from the
Capital West patroling inspector's office led by Lui Hsi. They ordered him to
join the army. Chu Chin was unwilling, and the four men indiscriminately beat
him. Subsequently the patroling inspector, Chang Shih-ying, ordered that he be
bound and questioned under flogging. When he had been beaten a hundred
blows, he falsely confessed to having bought cloth from a bandit. He was then
tattooed and registered in the army.
Fortunately for Chu Chin, his wife had the courage to complain to the
authorities of Hsiang-fu, who investigated and uncovered the whole
story.47

Numbers of patroling inspectors

During the period in which lower-level patroling inspectors were increas-
ing in number and importance, the higher levels of the patroling inspec-
tor system seem to have become less important. Their duties were
assumed by units dispatched from the regular army. The expanding role
of the lower-level patroling inspectors reflects both a renewed import-
ance of the military to law enforcement and a general increase in the
number of law-enforcement personnel. Early in the Sung the most com-
mon pattern was to appoint one sheriff for each district, often apparently
without his having a military colleague to share the burden of law
44 HCP 83.5b.
45 HCP 111.19b.
46 SHY, chih-kuan 48.126a (1020).
47 Hsu Ching-heng, Heng-t'ang chi, SKCSCP 1975, 10.8b.

204
 The role of the military

enforcement. By the late Northern Sung, however, most districts had a

military patroling inspector who acted in concert with the sheriff. This
general trend toward increasing the number of functionaries in local
areas who dealt with law enforcement seems to have accelerated during
the Southern Sung. Occasional specific reports make this clear, as in the
case of Ch'ang Prefecture in 1133, which for its four districts had twelve
men serving as sheriffs or patroling inspectors. 48 About a century later
Tu Fan (doctorate of letters degree 1208) described I-ch'eng District in
Ching-hu South Circuit and Ning-kuo District in Chiang-nan Circuit as
having three forts with attached patroling inspectors. (He suggested that
two of the forts and their officials be replaced by two new sheriffs'
posts.) 49 This situation was common in the Southern Sung. The 1228
gazetteer of Ming Prefecture, which controlled six districts mentions nine
currently active patroling inspectors in the prefecture. 50 Thus, during
the Sung there was both a militarization and an intensification of law-
enforcement patterns.

Location and character

Although there were complaints of patroling inspectors being housed in
inadequate facilities, including temples, most of them were supposed to
have been stationed at forts (chat) located at strategic points in the rural
areas. 51 These forts seem to have been small walled enclosures, inhabited
by soldiers and some local people. Forts were also established at the
market towns (chen), which had originated as local garrisons in the
T'ang and Five Dynasties period but by the Sung had become primarily
market centers. 52
Gazetteers sometimes give us some idea of the number of such forts in
an area and the size of their garrisons. In 1268 in Ch'ang-shou District

48 SHY, chih-kuan 48.70a.

49 Tu Fan, Ch'ing-hsien chi, 8.19b-20b.
50 Lo Chun, Pao-chHng Ssu-ming chih, SYTFGTS ed., 5111.
51 Chu Hsi, Chu wen-kung wen-chi, 20.11b; SHY, chih-kuan 48, 133a, 134b, 135b. For the
complaints, see, for example, Fu Ch'a, Chung-su chi, SKCSCP ed., 6.74-75.
Unfortunately I have not found descriptions of the construction and cost of building
these forts. But an interesting recent article does describe their economic importance as
well as their security functions. See Huang Kuan-chung, "Nan-Sung shih-tai Yung-
chou te Heng-shan chai" (Heng-shan fort in Yung Prefecture during the Southern
Sung), Han-hsuehyen-chiu 3 (1985): 507-34.
52 As befitted a unit that was primarily of economic importance, such market towns (and
their forts) were sometimes headed by low-level officials called service agents, who were
responsible not only for the collection of taxes, their normal duty, but also for patrolling
their area to control bandits. One source even indicates that such men could administer
punishments of beatings with the heavy rod or less. WHTK 63.574. See also SHY,
chih-kuan 48.92b.

205
 Law and order in Sung China

in Liang-che Circuit, there were four forts. O n e , located outside the

district town on the river bank by the Watergate, housed 150 soldiers. A
second, about thirty miles south of the subprefectural town at a strategic
mountain, held 175 men, and the third, about twenty miles south of the
district town on a waterway, had 195 men. T h e fourth, about sixteen
miles north of the town on the Great Lake, had 185 men. 5 3 From
these forts the patroling inspectors, like their civilian counterparts the
sheriffs, conducted regular patrols of their jurisdictions. Because they
were intended to be rural officers, they were not supposed to visit the
district or prefectural towns except on business. And when they did visit
the towns, they were supposed to return to their own posts within two
days. 5 4
T h e higher-level patroling inspectors, who had jurisdiction over a
n u m b e r of districts, were also expected to visit the areas under their
control. O n e official in 1029 a sked that they visit each district but not be
allowed to stay in any one for more than a few days. T h e districts visited
were to report to their superior prefecture on the progress of the tour. 5 5
These higher-level patroling inspectors were also required to patrol the
northern borders. 5 6
Sometimes patroling inspectors' titles indicated their roles. O n e com-
mon title means literally the "patroling inspectors who control the
borders." 5 7 They were also often helped maintain security along the
waterways which were the principal highways for the transport of goods
and people in many regions of China. A document from the late 1020s
gives some idea of their role and distribution, suggesting that a patroling
inspector be established every two courier posts along the Pien Canal
from the capital at Kaifeng to Ssu Prefecture, a reflection of the vital
importance of this waterway as the supplier of resources from the
southeast. 5 8
In some areas, because of particular local problems, patroling inspec-
tors were not stationed in the countryside. In 1019, the seaside patroling
inspector in K u a n g Prefecture in K u a n g - n a n East Circuit was shifted to
a spot near the office of the commissioner of overseas trade (shih-po ssu)
to serve as a guard for the traveling merchants who congregated there, 5 9
and as we noted, the patroling inspectors were sometimes stationed at

53 Shih Neng-chih, Hsien-ch'un p'i-ling chih, SYTFCTS ed., 3560.

54 If they disobeyed this rule, they were liable for one hundred blows of the heavy rod; see
TFSL 92-93.
55 SHY, chih-kuan 48.126b-127a.
56 SHY, chih-kuan 48.127b (1043).
57 HCP 243.10a.
58 SHY, chih-kuan 48.126b.
59 HCP 94.6b.

206
 The role of the military

garrison towns {chert) where they appear to have had general administra-
tive responsibilities, including management of tax collection stations. 60
Finally, the 1268 gazetteer of Ch'ang Prefecture in Liang-che Circuit
mentions a border fort located on the river "outside the western water-
gate" of the prefectural town, which housed a patroling inspector and his
men. 61
Some writers give the impression that patroling inspectors were shifted
to the towns because officials ignored government policy. In 1182 the
official Yang Wan-li wrote:
The prefectures of the empire are supposed to have patroling inspectors in outer
forts, to control wild forests or to guard strategic places. This was to be a regular
preparation for catching bandits. Nowadays it is not really so. In name there are
outer forts, but in fact the officers and soldiers live in the towns.. . . Formerly I
served as the judicial intendant in Kwangtung. At one point, because bandits
had passed by the outer forts in Hui Prefecture, I investigated. The patroling
inspectors' headquarters had become a field of rubbish. I inspected the soldiers'
quarters. They were in ruins. I asked about the officers and men and learned
that they all lived in the town. If the bandits live in the forests and mountains,
and the officers and men live in the walled towns, the bandits will have nothing
to fear.62
The number and variety of these positions in local areas are revealed
in a report in the gazetteer of Ming Prefecture in Liang-che Circuit from
the year 1228, which lists border or valley superior patroling inspectors,
tribal or Chinese superior patroling inspectors, border patroling inspec-
tors for a group of prefectures or a group of districts, and border
patroling inspectors for single prefectures or districts. There were also
knife-fish boat assault patroling inspectors. In all, in this prefecture,
there were nine patroling inspectors, having under their command a
total of about 900 men. 63

Appointment process
We have far less information on the backgrounds of these patroling
inspectors than we have for their civil service counterparts, the sheriffs.
We do know that after 1007 the servitors (shih-ch'en) appointed as patrol-
ing inspectors or as supervisors of militia (chien-ya) were expected to go
to the Bureau of Military Affairs to be examined. 64 They might also be
60 HCP 401.1a.
61 Shih Neng-chih, Hsien-ch'un p'i-ling chih, 3560.
62 Yang Wan-li, Ch'eng-chai chi, SPTK ed., 6g.8a-9a.
63 L o Chun, Pao-ch'ing Ssu-ming chih, 5111. See also W a n g I-chih, Chih-yuan ts'o-yao, H s u
Chin-hua ts'ung-shu ed., 59.
64 HCP 66.20a.

207
 Law and order in Sung China

subjected to physical examinations while on duty, perhaps because the

higher authorities were concerned that some might be elderly or other-
wise not physically capable of doing their jobs. 65 The Bureau of Military
Affairs also participated in selecting some of the patroling inspectors,
many of whom were drawn from among the ranks of the servitors minor
(hsiao shih-ch'en), that is, low-ranking functionaries with credentials in the
military hierarchy. 66
In name, the servitors were military officials, but recent research by
Winston Lo has demonstrated that in fact many of these men had no
connection with the military. 67 They were, for the most part, men
without military connections who were entering the government through
the use of the "protection" (yin) privilege (a Sung civil service practice
that permitted officials of certain ranks to secure official rank for their
assignees). Many of these men served in such lowly positions as service
agents in charge of tax collection depots, though in the very early Sung
they were also appointed to a variety of local posts. 68
However, servitors' backgrounds were indicated by a series of appella-
tions that gave some indication of their qualifications. Those used as
patroling inspectors are usually referred to as "militarily talented," and
so they presumably were from that part of the servitor group whose
members had some military connection.
There was some concern that such men possess at least minimal
literacy, and so the group of servitors, from which many patroling
inspectors were drawn, was subject, from at least the middle 1040s, to a
test of literacy, requiring the reading of the statutes and the writing of a
family register.69 Again the authorities, ever sensitive to the predicament
of the late T'ang when military men became dominant in local areas,
tried to ensure that at least the rudiments of wen would characterize men
chosen primarily for their commitment to wu. Of course, mere literacy
did not necessarily mean that these men had been much affected by the
civil culture of Sung times, but it did at least enhance the possibility that
they might be touched by civil values.
Ideally, the authorities looked to men with military officer's cre-
dentials for candidates. In 1014 when four patroling inspectors were
appointed for the capital area, they were chosen from among the attend-

65 HCP 305.13a.
66 SHY, chih-kuan 48.67b.
67 Winston Lo, "A New Perspective on the Sung Civil Service," Journal of Asian History 17
(1982): 121-35.
68 On the Northern Sung shih-ch'en, see Koiwai Hiromitsu, "H0-S6 no shi shin ni tsuite,"
(On the shih-ch'en in the Northern Sung), Shukan toyogaku, October 1982, pp. 34-54.
69 HCP 156.7b.

208
 The role of the military

ant servitors (ch'i-pan shih-ch'en).70 In 1034 the Bureau of Lesser Military

Assignments (san pan yuan) reported that " u n d e r the old system, the
servitors minor who had served as service agents [chien-tang] for one term
of office [jen] [probably for three years] could be appointed as patroling
inspectors or supervisors of militia [chien-ya]." This agency asked that a
second term as a service agent in nearby territories be added to the
requirements. 7 1 Thirty years later it was the rule that servitors serving as
service agents could be chosen as patroling inspectors if they had served
in posts "close to the people," had bandit-capturing abilities, and had
records that included only minor misdemeanors. 7 2
T h e state did try to find suitable candidates, specifying various criteria
to indicate desirable qualifications and undesirable groups. Reports list
the types of positions and bureaucratic levels required of candidates for
different sorts of patroling inspectors. 7 3
As with many civil posts, one common method of identifying promis-
ing men was to ask the local circuit intendants to recommend candidates
for such positions. An order of 1043 called on the fiscal and judicial
intendants and the prefectural administrators to recommend militarily
talented infantry and cavalry directors in chief (ma-pu tu chien) or servi-
tors minor (low-level military-hierarchy officials) serving as taxing depot
chiefs. 74 It was also in the power of these intendants to dispatch men to
serve provisionally in vacant local posts, including that of patroling
inspector. In times of troubles like the 1040s, intendants were frequently
granted permission to appoint patroling inspectors (and sheriffs) without
prior consultation with the Court. 7 5
Although the state wanted the best, it often took what it could get,

70 HCP 83.7a.
71 HCP 114.17a.
72 SHY, chih-kuan 48.129a.
73 See SHY, chih-kuan 48.129b-130a, which says that the state uses men with the rank of
Court audience-attending usher (ko-men chih-hou, R8B) to serve as superior patroling
inspectors, commissioned patroling inspectors, or (regular) patroling inspectors. Court
audience-attending ushers ranked as servitors major. These officers had charge of from
two to ten prefectures, as well as border forts or forts in strategically important spots.
They took their titles from the names of the places where they were located. There were
also jointly serving superior patroling inspectors. If they ranked from functionaries at
the disposal of the emperor or below, they were not called commissioners. Near the
borders there were also those called patroling inspector commissioners who had control
over from one to nine prefectures or at times were distributed according to routes and
not by territories. There were also joint patroling inspector commissioners. When they
were palace attendants, they were not called commissioners. Areas with from two to
seven districts or, in Kaifeng, each district, could have a patroling inspector. Groups of
two or three districts could have a Chu-po patroling inspectors imperial troop-
patroling inspector.
74 SHY, chih-kuan 48.i27a-b.
75 See, for example, HCP I43.i9a-22a.
209
 Law and order in Sung China

including men with dubious qualifications. As with the post of sheriff,

the Court also preferred not to use men who had a prior history of legal
violations. In 1043, f°r example, the Court approved a suggestion that
circuit intendants not be allowed to send to local posts, including that of
patroling inspector, men who had previously been demoted to serve as
service agents. 76 As Ts'ai K'an (d. after 1195) remarked, "Patroling
inspectors are servictors who have been weeded out of the army." 77
An order of 1211, in regard to having local administrators weed out
old and infirm patroling inspectors, also says in closing that in the
future, men qualified through imperial grace, men without military
abilities, and men who entered state service as clerks or miscellaneous
officials were not to be appointed, thereby indicating that such men had
already been serving as patroling inspectors. 78 Nor was this problem
confined to the Southern Sung. One official, writing in 1006, recom-
mended the use of "militarily inclined vigorous and intelligent men," not
those from clerical backgrounds or with civil credentials. 79 A few years
later a decree ordered the Bureau of Lesser Military Assignments to
select servitors to be prefectural superior patroling inspectors, because
"prior to this time [the state] has often used men [for these posts] who
entered office through protection [yin] and who had no recorded
experience in administration." 80
Later in the eleventh century the Court approved the use of graduates
of the military examinations, students in the military college, and men
who had been awarded official qualifications for having submitted mili-
tary plans. 81 About a century later there was apparently a policy that
military graduates be appointed patroling inspectors in border areas. 82
The general trend seems to have been to raise the level of the can-
didates' formal qualifications, though there were exceptions. Occasion-
ally men might even jump from the ranks of noncommissioned officers to
the post of patroling inspector. 83 And in the middle of the eleventh
century the post of patroling inspector was also used as a title for tribal
leaders working for the Chinese. 84
The Sung authorities were fully aware of the potential dangers created

76 HCP 142.1a.
77 Ts'ai K'an, Ting-chai chi, 1.21a.
78 SHY, chih-kuan 48.135a.
79 HCP 63.4b.
80 HCP 77.7b.
81 HCP 328.14b.
82 HNYL, 181.3005.
83 See the example of the Kuang Prefecture noncommissioned officer of the prison citadel
command, who was appointed because of his merit in capturing bandits; HCP 302.5a.
84 HCP 152.1 ia, 156.2a.

210
 The role of the military

by using such men as police officials. Military men had been trained to
accept a set of values that stressed the use of violence to achieve goals
and at times had only the most tenuous commitment to the values
embodied in the civil bureaucracy. We cannot understand the function-
ing of the Sung government without recognizing that a concern about
controlling the military shaped many government policies. The feeling
that military men might be dangerous to civilian control was certainly
one factor in promoting the conscious and regular policy of using civil
officials concurrently in military posts. As early as 971 two men who had
served as district magistrates were appointed as patroling inspectors in
the capital area. 85 And later in that same year, in a development related
to the government's policy of reducing the role of the military, the
prefectural cavalry and infantry offices were transformed into public
order offices (ssu-k'ouyiian), which were to be staffed by new graduates of
the doctorate of letters examination. 86 There are numerous examples
of prefects' serving concurrently as patroling inspectors or superior
patroling inspectors, particularly in sensitive regions such as the Ho-
tung border north of the capital. 87 The problem of staffing this post
highlights the tension in the paradigm of law enforcement. It is neces-
sary to get to the appropriate spot those capable of exerting sufficient
force, but at the same time such men must be properly motivated, not
just to do an efficient job of destroying the enemy, but also to observe the
civilian values on which the government rested.
It was probably for this reason that promotions from the post of sheriff
were common. As a civil official, the sheriff presumably had been more
thoroughly socialized into elite civil values. If such men also had proven
military abilities, then they would be ideal for the role of patroling
inspector. In 1041 a sheriff who had captured a military bandit and been
wounded in the process was promoted first to be a patroling inspector
and then, when officials objected that this was too trifling a promotion
for his merit, to the post of superior patroling inspector (tu hsiin chien)88
In 1079 another man was promoted from sheriff to patroling inspector
because of his merit in capturing bandits. 89 In Kaifeng, where the post
was of considerable importance, we find examples of men appointed
after having served as district magistrates. 90 As with other Sung offices
there seems to have been a set of policies that ranked candidates from

85 HCP 12306.17b.
86 HCP 12306.18b.
87 HCP 82.i8a-b, 4O2.i4b-i6a (1087).
88 HCP i32.28a-b.
89 HCP 299.19b-20a.
90 HCP 12306.17b.

211
 Law and order in Sung China

various groups when filling positions. In an incident in 1187 t n e auth-

orities approved the request of an official that all patroling inspectors
serving along the coasts be either graduates of the military examinations
or men who had entered the service as a result of military merit. They
were to be no older than fifty years of age and to have recognized
military capabilities. The Ministry of Personnel was to select the can-
didates, filling the vacancies first with men who had already had experi-
ence in sea warfare, second with graduates of the military examinations,
and after that with men qualified under the then-current rules. 91

Terms of office
The patroling inspectors themselves served for limited terms. The
general practice, as exemplified by the patroling inspectors, supervisors
of militia (chien-ya), inspectors general for seizing bandits (chu-po tsu-tsei),
and service agent servitors (chien-tang shih-ch'en) of Szechuan, Shensi, and
Kwangtung, was to appoint officers for two-year tours of duty, though
they might be reappointed for a second or even a third term to the same
post. 92 In particularly unhealthy areas they might serve shorter terms.
Until 1090 in part of Yung Prefecture in Kuang-nan West Circuit, which
was an especially unhealthy area, they were rotated every six months. In
that year an official complained about the high death rate and success-
fully petitioned that they be sent to different locations each quarter. 93

Motivation
The authorities also sought to create a system of incentives to maintain
morale. On the most basic level they had to provide some regular system
of support for their functionaries. Beyond that, in typical traditional
Chinese fashion, they created a system of rewards for successful perform-
ance but then coupled this with a set of penalties for failures. The
problem of attracting competent personnel was made more difficult by
the low rate of pay. One report from the middle of the eleventh century
says that the men were paid five strings per month, as opposed to the
seven strings per month that had been paid to the recorders and sheriffs
in the smallest districts more than two decades earlier. 94 Patroling

91 SHY, chih-kuan 48.133a. Men from "outside the stream" - that is, men who were not
considered to be part of the regular structured civil service - were not to be appointed.
92 See HCP 94.1 ia for the Northern Sung, and SHY, chih-kuan 48.133b, for the Southern
Sung. For added tours of duty, see SHY, chih-kuan 48.133b.
93 HCP 408.12a.
94 HCP 102.8b, 175.6a.

212
 The role of the military

inspectors probably received some income in kind, but still they were not
paid well. The archers, mere constables who worked for the sheriffs,
were paid at half the rate of patroling inspectors, earning two and a half
strings per month. Thus it is hardly surprising that as with many of the
other lesser offices in the Sung, the authorities sometimes had difficulty
filling some of these bandit-controlling military posts. An official in 1185
complained about the long-unfilled posts, not only of patroling inspec-
tors, but also of inspectors general for seizing bandits {chu-po tsu-tsei).95
The state faced a serious problem in trying to motivate officials to act
effectively. The problem of incentives was especially delicate where law-
enforcement personnel were involved. Such agents controlled by superior
force. If they did not use it well, the state would be at risk. And if
they overused it, the state also would be at risk. The first possibility
is described laconically in a report of 1043 which states that "many
patroling inspectors are military officers who are unwilling to capture
bandits." 96 The root of the problem, as analyzed by a number of mid-
century critics, was a failure in the system of rewards and sanctions. As
Yii Ching pointed out, failure to pursue bandits was punished only with
a fine, whereas actually pursuing them might get the official killed.97 The
state responded to this by advancing the patroling inspectors' rank when
they did capture bandits. 98 The government also regularly gave awards
to the families of police officials who were killed in the line of duty. 99
During both the Northern and the Southern Sung, the state tried to tie
rewards directly to the successful performance of duties. An order of
1024 stated that patroling inspectors and bandit-catching servitors minor
who captured a whole gang of bandits or other evildoers should be
granted commendations; 100 a complaint of 1185 pertaining to the salt-
producing areas in Liang-che Circuit asked that rewards and advance-
ment be tied to performance, by examining the salt quotas to determine
whether or not they had been filled.101
The state also established negative sanctions, punishing unsuccessful
patroling inspectors in a variety of ways. Like other police agents, they
might be fined for not capturing wanted men. Moreover, fines were not
the only penalties to which they were subject. Failure could also affect

95 SHY, chih-kuan 48.132b-133a. This document refers to conditions on the Szechuan

border.
96 HCP 141.14b—15a, 142.23a—b.
97 Yii Ghing, Wu-ch'i chi, shang. 8b-9a.
98 See, for example, HCP 166.17a.
99 Compare examples from 1044 a n d IQ 6i of appointing their sons to official rank; HCP
146.20b, 192.1b, 293.7b.
100 SHY, ping u . i o a - u a .
101 SHY, chih-kuan 48.132b.

213
 Law and order in Sung China

their future careers. An order of 1024 s a ^ that "if they do not apply
themselves and, as a result, bandits injure the people or if they make few
arrests relative to those of other jurisdictions, they are to be severely
sanctioned in accordance with the current regulations." 102 An order of
1033, for example, indicates that patroling inspectors who had three or
more uncaptured gangs in their jurisdiction could be demoted to the post
of service agent. 103 This was no empty threat. In a case from 1059 a
patroling inspector was disenrolled and sentenced to the penalty called
registered control, in which the convict performed penal labor for a
fixed period while being registered in and controlled by a unit of the
provincial armies, for having failed to attack a salt smuggler.104
Failure to report bandits could also be punished. Both sheriffs and
patroling inspectors might be subject to a penalty equivalent to two
years of penal servitude for such a failure. Under some circumstances
this penalty might not be reduced through grants of imperial grace. 105 If
bandits entered the towns, the patroling inspectors, along with the other
local officials, were subject to penalties. 106
Failure to perform adequately was usually blamed on incompetence.
There are repeated orders throughout the dynasty for inspections of
patroling inspectors (and sheriffs) and the replacement of the unfit, even
if they had not yet committed any infractions of the rules or failed in
their duties. 107 In 1149, when the emperor was concerned about the
failure of the authorities in Fukien to suppress piracy, he traced the dif-
ficulty to incompetent or incapable police officials and ordered the cir-
cuit intendants to investigate and replace the unfit.108
Sometimes the officials failed to perform, not because of incompetence,
but because of corruption. The temptation to extort money must have
been severe; those who were caught were usually treated very harshly. In
1006 a patroling inspector who had accepted a bribe from the father of
an arrested bandit was beheaded. The punishment for his crime would
normally have been strangulation; it was made more severe by special
imperial order. 109
Inspectors might also have the statutory penalties reduced. In 1069
a patroling inspector, who had been sentenced to death for having

102 SHY, ping n . i o a - u a .

103 HCP 113.12b.
104 HGP 190.17a.
105 HCP 494.29a, 129.3b.
106 HCP 12400.24a (1042). See also HCP 113.12b.
107 HCP 27i.i4a-io8o, 365.^-1086.
108 SHY, chih-kuan 48.73a. See also a similar complaint in 1135; HNYL 84.1383, SHY,
chih-kuan 48.71a.
109 HCP 64.1a.

214
 The role of the military
accepted bribes, was spared execution and sent for registration to
Sramana Island, the Northern Sung maximum-security location off the
north coast of the Shantung peninsula. This particular incident was well
known, because the patroling inspector in question had originally been
sentenced to be beaten, tattooed, and then sent for registration. An
official complained that it was shameful for a member of the government
to be beaten and tattooed. As a result of this complaint, "from this time
on active officials were to be spared the beating and tattooing," 110
though in fact officials were occasionally still punished in this way. 111

Later career paths

The later career paths of these patroling inspectors frequently led to
other law-enforcement positions, though they might first have served a
number of two-year terms in a patroling inspector slot. 112 An order of
1020 indicated that the supervisors of militia {chien-yd) or patroling
inspectors from distant border areas who ranked as palace auxiliaries
(tien-chih) or lower (that is, in the bottom rank of the nine-grade civil
service) and who had law-enforcement experience might be rotated for a
term of duty as inspectors general for the seizing of bandits (chu-po tsu-
tsei) in the inner parts of the empire. After finishing their tours they were
to be returned to their "distant" territories. 113 In rare cases patroling
inspectors could hope for appointments to the capital where they might
have opportunities for further advancement. A report from 1049 records
the case of a patroling inspector who was promoted to the rank of court
audience attendant usher (ko-men chih-hou) because he had captured
twenty-two bandits. 114 This position, although it was not a high one, was
greatly coveted, in part because its incumbents took part in ceremonies
at Court audiences and thus had good access to high Court officials.115

Other military
As a matter of general policy, the Sung government did not confine the
performance of any major duty to a single office, though there might be
one official especially concerned with it. The role of the military in
law enforcement reflects this penchant for balancing offices against one

110 SS 201.5018.
111 See, for example, SS 200.4998 (1075).
112 SHY, chih-kuan 48.133b.
113 HCP 95.17a.
114 HCP 122.17a.
115 Lo, An Introduction, p. 67.

215
 Law and order in Sung China

another. Patroling inspectors were clearly the most important military

officials charged with the daily task of law enforcement, but many other
military officials were also involved at various times. Prefectural direc-
tors in chief (tu-chien), a post sometimes filled by the men in the lower
ranks of the military hierarchy who were called servitors major and
minor, had charge of the training and dispatching of their town's con-
tingent of military garrison soldiers. These officers could serve con-
currently as patroling inspectors but did not always do so. Those with
qualifications too low for appointment as patroling inspectors often
served concurrently as service agents. As with the post of patroling
inspector, civil officials were sometimes used for this post of prefectural
director in chief, at least during the early years of the dynasty. 116
Other sorts of military men were also used in law enforcement. In
1027 t n e central government established an envoy for capturing bandits
for every two post stations along the imperial canal, with a support force
of seventy soldiers, 117 and in 1052 a supervisor of bandit capturing (t'i-
chu tsu-tsei) was established for the two prefectures of Kuang and Hui
in Kuang-nan East Circuit. 118 Lower-level officials in the military
hierarchy were also used in the early years of the dynasty, 119 and under
a decree of 1012, in Kuang Prefecture, the military administrator (ch'ien-
hsia) was concurrently the overseer of security and bandit control within
the city. 120
Finally, the desire to avoid an overconcentration of responsibility
prompted the government to assign to civil administrators in many areas
the concurrent responsibility for mobilizing troops to guard against
bandits. Indeed, we cannot understand how the Sung government func-
tioned without remembering that the central authorities, while under-
standing the importance of distinguishing between military and civil,
blurred this distinction. They used civil officials in military roles in order
to control the military system. 121
The problems that faced the Chinese a millennium ago in their
attempts to use the regular military against bandits have a haunting
familiarity in our age of guerrilla warfare. The Southern Sung official
Chao Shan-k'uo, criticizing the use of regular troops against bandits in
central China, listed their weaknesses. Regular troops, he said, would
work only according to regular procedures. They were incapable of
116 Wang I-chih, Chih-kuan fen-chih,
117 SHY, chih-kuan 48.126b.
118 HCP 172.8b.
119 HCP 52.1 ib 12a mentions a servitor in 1002 who was assigned a group of soldiers and
the duty of patroling in the capital.
120 HCP 79.7a.
121 HCP 144.2b.
216
 The role of the military

striking swiftly and stealthily, whereas the bandits, masters of their

terrain, could come and go quickly. The soldiers in central China were
demoralized by the unhealthy conditions and by the knowledge that
even if they succeeded, they would be rewarded less than would those
who fought in supposedly more difficult places such as the northern and
western frontiers. And if they failed, they would be heavily penalized.
Because the locally powerful people fortified their homes, the bandits left
them alone and paid for what they needed. The soldiers, by contrast,
abused their power, confiscating and seizing what they wanted. As a
result the local people were sympathetic to the bandits; no one sided
with the government. Thus the bandits knew all the movements of the
government troops. Chao summed up his assessment: You cannot swat
at ox flies and expect to kill lice. To this classic guerrilla situation he
proposed a classic answer: Arm the local people, and build small scat-
tered forts so that they could protect themselves. Offer ample rewards
and military commissions to local fighters.122
Most of our reports on the workings of this military side of the law-
enforcement process are based on dry official documents. Occasionally
we have an opportunity to see the men themselves in action. In the
Northern Sung a man named Sang Shih was appointed the patroling
inspector in Yung-an District, which was at that time plagued by a gang
of twenty-three bandits that had arisen in the wake of a plague of locusts
and a drought.
Shih was summoned to the capital by the Bureau of Military Affairs and given
responsibility for the bandits. He said, "The bandits are afraid of me and will
disperse. If they disperse, they will be difficult to seize. We should first announce
my coming in order to frighten them." When he arrived at the post, he closed the
palisade around the fort. He warned the officers that no one was to be allowed to
leave the fort without authorization. Thus they remained for several days. His
subordinates did not understand what was going on. Several asked to be allowed
to go out and investigate, but this was not permitted. At night, with several
soldiers dressed like bandits, Shih visited the bandits' former haunts. They
entered people's homes and found that both old and young had left. There was
only one old woman left behind to prepare food, as if there were to be a
gathering of bandits. Shih went back and again shut the gate for three days.
Again he went out, personally carrying utensils to the cook and gave the old
woman a gift. The woman thought he really was a bandit. Then he spoke with
her and mentioned the bandits. She said, "When they heard that Master Shih
had arrived, they all went away. Recently they have heard that he has shut up
the fort and will not come out, so they know he is not to be feared. Now all are
returning, so-and-so to such-and-such a place, and so-and-so to such-and-such

122 Chao Shan-k'uo, Ying-chai tsa-chu, i.i8a-2ob.

217
 Law and order in Sung China

another place." Shih again in three days went out with a handsome present for
her. Then he spoke truly saying, "I am Master Shih. Investigate the truth for
me. Do not let my secret out. In three days I will come again." Three days later
he came again. The old woman had all the facts of the bandits' locations to tell
him. The next day Shih divided his troops and seized all the bandits, including
the most violent. Shih himself mounted his horse and captured them.123

Soldiers subordinate to the patroling inspectors

The military side of law enforcement also needed sizable numbers of
men to enforce the law. As the key local officers, the patroling inspectors
were assisted by a body of cavalry and infantry soldiers (or, where
appropriate, sailors). Not enough work has been done to allow us to
describe exactly the backgrounds of Sung soldiers, though it is clear that
some of them began as criminals. We can surmise that most of them
were probably drawn from the same social strata from which the civilian
archers came. The soldiers serving under patroling inspectors were
often men from the regular imperial armies (chin-ping),124 though some
might be "local soldiers" (t'u-ping) or a mixture of both sorts of troops.
Patroling inspectors might also use non-Chinese tribesmen, in accord-
ance with tribal law, under appropriate circumstances. 125 They were, on
the other hand, specifically forbidden by an order dated 1019 to use
criminals registered in the army as agents to try to catch bandits, an
order that indicates that criminals actually were being used in this
way. 126
The regular army troops sent to serve under patroling inspectors had
received regular military training, which was presumed to be adequate
in most regions. However, parts of China south of the Yangtze River
were crisscrossed by canals and rivers and heavily dependent on water
transportation, with a concomitant problem of piracy. For this reason,
when appropriate, the soldiers serving under a patroling inspector might
be given naval training. Under an order of 1015 the soldiers in Chiang-
nan and Liang-che who were serving under inspectors general (chu-po) or
patroling inspectors were given training in fighting from boats, without
which they could not succeed in capturing the local bandits. 127 In a
report from the early 1050s, Ts'ai Hsiang stated that in the past the
coastal cities in Fukien had had garrison forts to guard the inlets that led

123 CYKC 49 (in Beijing 1981 ed.).

124 See, for example, HCP 83.7a.
125 HCP 280.12a 1077.
126 SHY, chih-kuan 48.126a (1019).
127 HCP 84.17b.

218
 The role of the military

to them but that their troops had gradually been cut back during the
time of peace. Those troops currently stationed at such places did not
have naval training, and so when they encountered salt or tea smugglers
they were often defeated. He asked that the troops be given appropriate
training and arms and that the coastal patroling inspectors go to sea
during the trading season in order to protect merchant vessels.128 In
other cases the patroling inspector might be assisted by sailors rather
than soldiers, as was the case in 1061 when a patroling inspector was
established at a district in Kuang Prefecture in Kuang-nan East Circuit,
supported by two hundred sailors. 129

Local soldiers
When "local soldiers," that is, soldiers drawn from the local residents,
first were used, it was assumed that they would be particularly effective
in suppressing bandits, their familiarity with the area more than off-
setting any inadequacies in their military training. Regular soldiers,
although perhaps technically better trained, were at a disadvantage,
because they were usually serving in unfamiliar areas. To some extent
this was true. In 1080 an official from the southeast complained that
many of the regular army troops being rotated in for service were from
the northwest and so were unfamiliar with the southeast.
Indeed, you dispatch such men for a year. They develop some slight familiarity
with the roads, hills, waters, and people, and then they are compelled to rotate
out. If they go to the seaside, they are not familiar with it. For these reasons they
are unable to defeat their opponents when they meet them. I ask in the selection
of local soldiers you order that one-half [of a unit's complement] be veterans
from the provincial and imperial armies who are willing to fill out the number so
that the new and the old can instruct each other.130
His suggestion may not have been immediately accepted. However, by
the late 1080s, at the request of the Bureau of Military Affairs, it was
specified that one-half the soldiers under a patroling inspector be men
dispatched from the regular imperial army. 131 The bureau was con-
cerned not about the incapacity of the local soldiers but about their
tendency to become involved in the local community and thereafter to
collude in breaking the law. 132 This half-and-half order seems to have
been enforced. We have a slightly later approved request that when this
128 Ts'ai Hsiang, Tuan-ming chih, 21.10b-na.
129 HCP 12428.12a.
130 Lo Chun, Pao-ch'ing ssu-ming chih, p. 5155.
131 HCP 395.20b.
132 HCP 395.20b.

219
 Law and order in Sung China

new system required a reduction in the number of local soldiers, this cut
was to be taken wholly from those who were in the infantry, with no cuts
in the number of mounted troops. 133
There were problems associated with both local and regular army
soldiers. Soldiers were trained to be violent and not necessarily to behave
ethically. At times, they actively cooperated with robbers or became
robbers themselves.134 Or else they might simply be disinterested in their
work. Many of the imperial soldiers sent for such duty are said to have
been looking forward to the day when they would be rotated out. 135
Local soldiers, by contrast, were sometimes said to be incapable. One
official reported in 1083 that on a tour of local areas, he had found many
local soldiers to be old and weak. Not even 30 to 40 percent of them were
capable of fighting bandits. Many in fact were not even local men. 136 In
any event, critics still charged that neither local soldiers (t'u-ping) nor
imperial troops (chin-ping) were as effective as archers, because the
archers, as locally recruited village officers, were themselves villagers. As
a result the sheriffs' police work was said to be more effective than that
of the patroling inspectors.137 Even this bad situation became worse.
Writing about a century later, Chu Hsi remarked that in Tu-ch'ang
District (in Chiang-nan East Circuit) the fort troops merely protected
the town and did not go out to patrol the countryside.138

Numbers
There are various figures for the numbers of such troops. In 1006, when
two new patroling inspectors were established in a border area, they
were assigned 100 soldiers each. 139 This figure probably is fairly rep-
resentative. It appears again in a report from 1084,140 and in 1099 when
new supervisors of militia (chien-ya) and patroling inspectors were being
established at Ning-chai City, they were each given 100 troops.141 In the

133 HCP 4O2.5b-6a.

134 Tu Fan, Ch'ing-hsien chi, 8.19b-20b.
135 Lo Chun, Pao-ch'ing ssu-ming chih, p. 5515.
136 HCP 340.15b.
137 Su Ch'e, Luan-ch'eng chi, 35.2b.
138 Chu Hsi, Chu wen-kung wen-chi, 20.11b-14a.
139 HCP 64.3a.
140 HCP 343.3b. During the Hsi-ning period, an official assessing local troop needs in
Huai-nan and Liang-che reported on eighteen police commissioner jurisdictions, with
a quota of 1613 men, that is, an average of about 90 men. Lo Chiin, Pao-ch'ing ssu-ming
chih, p. 5155.
141 HCP 516.9b.

220
 The role of the military

Table 6.1. Numbers of soldiers in selected districts

Date Place Numbers Source

1196 Su Prefecture 705 (in four forts) Sun Ying-shih, Ch'in-ch'uan chih
Ch'ang-shou 2660
1221 T'ai Prefecture 40 Ch'en ch'i-ch'ing, Chia-ting ch'ih-
Hsien-chu 484 (in four forts) ch'engchik 7206 ff.
Ning-hai 124
Huang-yen
1228 Ming Prefecture 200 (in one fort) Lo C h u n , Pao-ch}ing Ssu-ming chih
Cheng 5297
Chu-t'ai 118 (in two forts) 5274
Tz'u-ch'i 73 (originally 90) 5298
Ting-hai 320 (in two forts) 5320
Ch'ang-kuo 643 (in two forts) 5336

following century Chu Hsi reported on a patroling inspector who had

106 local troops. 1 4 2
T h e size of this complement may have been determined largely by the
level of disorder and the strategic importance of an area; it also may
have been influenced by a desire to maintain a balance between the
number of soldiers engaged in law enforcement and the number of
civilian constables. According to one report, the number of local soldiers
in one circuit was about the same as the number of archers there. 1 4 3 T h e
regular complement could be increased if the patroling inspector in
question reported a large number of bandits and requested more men
and if his report were verified by the appropriate pacification com-
missioner (an-fu shih)144 (see Table 6.i).
In addition to their service under the patroling inspectors, local
soldiers also helped in law enforcement by their presence in small num-
bers at the innumerable post stations that dotted the Chinese landscape.
A report from the beginning of the thirteenth century suggests that such
post stations were staffed by from three to ten soldiers each. 1 4 5

142 Chu Hsi, Chu wen-kung wen-chi, 19.42a. See also, from 1196, a report that Ch'ang-shou
District had three patroling inspector forts, each with 144 soldiers and a report that
the Ling-san patroling inspector had a quota of 100 men in 1084. Sun Ying-shih,
Ch'in-ch'uan chih, p. 2660; HCP 343.3b.
143 Ts'ai K'an, Ting-chai chi, 2.12a.
144 HCP 398.13a. See the example from 1148 in Fukien of a patroling inspector with four
hundred troops; HNYL 2567.
145 Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, pp. 7205, 7209.

221
 Law and order in Sung China

Selection
The authorities needed men who were both adequately skilled in arms
and physically capable. Therefore they established guidelines for the
selection of soldiers and mandated continued training. When these
soldiers were men dispatched from regular army camps, they were to be
selected by senior prefectural officials. The patroling inspector in turn
was responsible for informing his superiors in the military hierarchy of
the names of the men under his command. 146 Such soldiers were trained
under their patroling inspectors and were usually given training facilities
rather like those available to the regular army. 147 The training results
were examined periodically. In some regions, the patroling inspectors
were supposed to inspect their troops every few days, 148 and the fiscal
and judicial intendants of the circuit were to inspect them twice a year,
in the spring and fall.149

Motivation
Maintaining morale required positive incentives such as pay and
rewards, negative incentives, and the provision of adequate working
conditions. The Sung continued an ancient Chinese practice by reward-
ing soldiers for capturing or killing wanted men. This practice highlights
the more general problem of motivating men not just to act but to act
properly. The Sung paid for severed heads, but who was to know
whether the head came from a bandit or an ordinary peaceful person? In
the mid-eleventh century the official Yii Ching remarked:
In Hunan, money and silk were paid out for the heads of the barbarian bandits
who belonged to the Man tribes. Soldiers greatly profited from the liberal
rewards, stopping people on the roads, treating them as bandits, and killing
them. So people were issued identity passes to prevent their being injured by
government troops. From this time on, the bandits killed those who carried the
identity passes, and the troops killed those who did not.150
This was a recurrent problem during the Sung (and in other dynasties).
The great Southern Sung poet (and official) Lu Yu reported that "fight-
ing and taking heads has been practiced since the Three Dynasties [of
Hsia, Shang, and Chou], so we cannot say that the practice is not an

146 HCP 398.13a.

147 HCP 380.3a; Shih Neng-chih, Hsien-ch'un p'i-ling chih, p. 3560, notes a training field in
1268 that was 14 changh (about 140 feet) wide by 80 changh (800 feet) long.
148 HCP 516.18b.
149 HCP 348.18b.
150 Yii Ching, Wu-ch'i chi, tsou-i shang. i6b-i7a.

222
 The role of the military

ancient one. However, in recent times those seeking bandits kill ordinary
people and use this as a basis for claiming merit." 151
Motivation was also intimately related to troop morale, and that in
turn was affected by the conditions of service. Security demanded that
some units be stationed in unhealthy or dangerous regions. The auth-
orities tried to reduce the demoralizing effects of such postings by limit-
ing the length of tours of duty. All soldiers under patroling inspectors
served for limited terms. In the most difficult and dangerous (and
unhealthy) places they, like their officers, might serve for only six
months before being relieved.152 In other, somewhat less unhealthy but
still difficult and unpleasant areas, these men might serve one-year
terms. Patroling inspector soldiers in Yung, Kuei, I, Jung, Ch'in, and
Lien prefectures in Kuang-nan West all seem to have served one-year
tours of duty, 153 as did the soldiers serving in patroling inspector posts
concerned with controlling salt production in Hang, Hsiu, Wen, T'ai,
and Ming prefectures (all in Liang-che Circuit) during the Huang-yu
period (1049-53). By contrast, in most areas they seem to have served
two-year tours. 154

Remuneration
Motivation was also affected by the physical facilities and remuneration
provided. Until 1004 in Szechuan and Shensi these troops, despite their
vital role in maintaining local security, were without adequate quarters.
In that year it was ordered that quarters be built. 155 The problem,
however, continued. Liu Chih complained in 1086 that he had learned
of soldiers' being quartered in temples because of a lack of suitable
barracks. 156
Sung gazetteers provide some information on the forts that housed
these soldiers. One fort built in Chen-chiang Prefecture late in the
Southern Sung had fifty bays (chien) of buildings and 150 men (a chien,
the space between two pillars, was the standard unit for measuring the
size of buildings). 157 A fort in Ming Prefecture, first occupied in 1259
and dubbed "Perfect Peace," had three bays (chien) of gates, three of
offices, three of halls, three of central halls, twelve of living quarters, and
four of kitchens. The fort had been built under the management of the
151 Lu Yu, Wei-nan wen-chi, 13.13a.
152 HCP 480.12a.
153 HCP 380.3a, 393.6a-b.
154 HCP 393.6a-b.
155 HCP 57.4a.
156 HCP 365.33-4^
157 Lu Hsien, Chia-ting Chen-chiang chih, SYTFCTS ed., pp. 2891-92.

223
 Law and order in Sung China

heads of two Righteous Service village officer groups, using 3000 strings
of prefectural funds. T h e authorities also spent 250 strings to build beds
a n d other needed items. 1 5 8
At times the soldiers did not receive a d e q u a t e rations and other re-
muneration. Although, as we noted, the soldiers in Szechuan and Shensi
were supposedly given quarters by an order in 1004, not until 1018 did
they have a regular system of rations. In that year it was ordered that
they henceforth be paid 200 coins per m o n t h as "food m o n e y . " 1 5 9 In
addition to this money income, soldiers also received some income in
kind. In a report written sometime after 1178, C h u Hsi criticized a
request to build a new fort with one h u n d r e d troops. H e said that the
a n n u a l cost would be 1800 bushels (shihh) of rice, 500 strings of cash, 500
lengths (p'i) of silk, and 1500 ounces of silk floss; that is, for each soldier
the cost would be 18 bushels of grain, 5 lengths of silk, 15 ounces of floss,
a n d 5 strings of cash. 1 6 0 T h e i r cash income at this time, about 400 coins
per m o n t h , was a b o u t double the pay scale of 1018, but was much less
t h a n that of their civilian counterparts, the archers, who were receiving
2.5 strings per m o n t h . Each soldier would have received in rice rations of
slightly more t h a n 2.5 pints per day, the same ration as men serving
in the provincial armies received. This would have been more than
a d e q u a t e for a single m a n , since the Sung estimated a full ration at 2
pints were day per m a n . H a d these men been remunerated at these
rates, they would have had a comfortable, if not overlarge, income.
Unfortunately, at various times there were complaints about deficiencies
in rations and wages, making it difficult to keep u p with the quotas of
men 1 6 1 a n d tempting those who did serve to practice extortion. 1 6 2
Effective performance of duty was rewarded; failure was punished.
Since 1021 the rule h a d been that those who did not succeed in captur-
ing criminals before the third time limit were to be punished by sixty
blows of the heavy rod. 1 6 3 In addition to more general rules on rewards,
in 1086 the government also instituted a series of rewards for capturing
wanted men, to complement the already existing system for punish-
ing those who failed to seize those wanted. Soldiers were to be paid
two h u n d r e d coins for each deserter a n d five h u n d r e d coins for each
bandit. 1 6 4

158 Mei Ying-fa, Pao-ch'ing Ssu-ming hsii chih, pp. 5473-74.

159 HCP 91.2a.
160 Chu Hsi, Chu wen-kung wen-chi, 20.11 b— 14a.
161 Tu Fan, Ch'ing-hsien chi, 8.19b-20b.
162 Ch'en Ch'i-ch'ing. Yiin-chuang chi, SKGSCP ed., 4-3b-5a.
163 HCP 97.11b.
164 HCP 92.7a; SHY, ping 3.2a, says that this order pertained to the new city at the
capital.

224
 The role of the military

Given their poor pay and other disadvantages of this service, why
would men serve as local soldiers? One answer for some of the soldiers
was that they could sometimes transfer into the regular army. An order
of 1090 states that local soldiers who volunteered for such service - if
found on inspection by the circuit authorities to be capable and under
forty years of age - could change their registration and enroll in the
imperial army. 165

Conclusion

In The Origins of the Family, Private Property, and the State, Friedrich Engels
r e m a r k s that t h e distinguishing m a r k of the state w a s first its organiza-
tion on a territorial basis a n d second

the institution of a public force, which is no longer immediately identical with a

people's own organization of themselves as an armed power. This special public
force is needed because a self-acting armed organization of the people has
become impossible since their cleavage into classes. . . . This public force exists
in every state; it consists not merely of armed men, but also of material
appendages, prisons and repressive institutions of all kinds. 166

W e m a y disagree with Engels's analysis of the causes a n d n a t u r e of

the growth of institutions of force. W e c a n certainly accept the fact of
their existence a n d their most typical role as defenders of the existing
social order a n d system, which at the s a m e time m e a n s as defenders of
the existing political a r r a n g e m e n t s a n d the prerogatives of the ruling
elite. Chinese states, including t h e S u n g dynasty, were n o exception. T h e
S u n g military w a s very large a n d usually w a s effective enough to defend
the state against enemies from within a n d without.
Like other Chinese dynasties, the Sung also used its military in a
variety of ways to enforce t h e law. L a r g e regular military units were used
against major outbreaks of disorder. T h e military also enforced its
own law, most i m p o r t a n t l y against its o w n m e m b e r s . Finally, smaller
d e t a c h m e n t s of troops were active from t h e beginning of the dynasty in
local law enforcement. Early in the dynasty, however, except in u r b a n
areas where t h e military remained clearly d o m i n a n t , law enforcement
was primarily a function of civilian agencies. I n time the military
b e c a m e increasingly significant in t h e countryside, so that by the e n d of
the d y n a s t y t h e sheriff a n d his supporting archers were o u t n u m b e r e d by
the patroling inspectors a n d their troops.

165 HCP 449.8b.

166 Friedrich Engels, The Origins of the Family, Private Property, and the State (New York:
International Publishers, 1972), pp. 279—80.
225
 Law and order in Sung China
Soldiers and constables differed in many ways. All constables were
local men; many soldiers were strangers. Constables served for long
periods at the same place; soldiers were rotated into and out of local
posts. Soldiers were part of the military hierarchy; constables served
under civil service superiors. The two groups also, however, had much in
common. Both were concerned with using force to maintain order. Both
ranked low on the scale of social prestige. Both were largely illiterate and
uneducated. And both were known for their tendency to abuse the local
people.
The state authorities were faced with a dilemma. They could not
avoid using such men, but they also could not hope to create a group
that would be sufficiently skilled in the use of force and also be thoroughly
imbued with the nonmilitary values of the civil government. What was
desirable - a sufficiently skilled force of thoroughly trustworthy agents
employed at a tolerable cost - was not possible. What was possible was
sometimes dangerous. A force of men at least minimally skilled in the
martial arts might be assembled at a bearable cost, but it might not be
thoroughly indoctrinated with civil values. Some devices, therefore, had
to be created to keep these men from becoming a danger. Some abuse of
the local people was probably inevitable, but the armed men had to be
kept from becoming a danger to the political system. The response of
the state authorities to this problem reflects the specific application to
this situation of the more general principles used elsewhere in the
government - balancing group against group, and preventing strong
internal bonds from forming between superiors and inferiors.
Among the military involved in law enforcement, the regular rotation
of both men and officers lessened the danger of the growth of strong ties
between soldiers and, more importantly, between officers and men. The
rotation of civil officials, with the sheriffs having relatively brief terms,
also precluded them from establishing excessively strong bonds with
local people or with their subordinates. And finally, balancing the civil
against the military, the sheriff's constables against the soldiers of
the patroling inspectors - combined with the policies that forbade the
two offices from having contact outside the immediate needs of a law-
enforcement problem - reduced the dangers of collusive misbehavior.
The Sung authorities could not, of course, eliminate all the potential
problems created by having such forces of armed men in local areas, but
the dynasty's record of internal peace suggests that the policies they did
use succeeded to a remarkable degree.
Despite the growth in the role of the military, the Sung government
was able to avoid the fate of the late T'ang and the Five Dynasties.
By adjusting the ever-changing balance between the wen and the wu,
226
 The role of the military
between the role of the gentleman and the role of the hero in politics, the
Sung authorities created a set of policies and practices that maintained
civilian dominance while building a military and police apparatus suf-
ficient to maintain a degree of internal order unmatched by any other
long-lasting dynasty. The Sung were never seriously threatened by
internal rebellions, and externally they were defeated only by the most
powerful military machines of their age.
The key to the control system was the intermixture and interlayer-
ing of civilian and military elements. The chief local military law-
enforcement officials, the patroling inspector, was responsible to the
prefectural military authorities, but in many prefectures the civilian
prefect was concurrently a key figure in this set of military commanders,
indeed was often himself the chief figure. The Sung authorities were thus
able to devise a mixture of civil and military that allowed the military to
be active in and important to law enforcement without deriving undue
power from this responsibility. And the Sung civilian authorities were
able to use the military without themselves being used by it.

227
 7
Supervision of
law enforcement —
the role of the intendants

General problem of control

Governance is a control problem. We generally think of law enforcement
as controlling those who are outside the ruling group, but actually
control has two faces: control of outsiders and of insiders. Dual control
exists in all parts of the political apparatus; nowhere is it more important
than in those segments of the government that have access to armed
force. Law-enforcement and military systems present the ruling group
with a serious dilemma. It is necessary to develop effective users of force,
to counteract those who want to do violence to the existing order, but
this need must be balanced against the necessity of keeping the users of
legitimate force from becoming too powerful.
The Sung were unsurpassed among the major Chinese dynasties in
dealing with this problem. Their success in creating a sufficient but not
excessive system of the institutions of force is reflected both internally
and externally. It is often said that the Sung were militarily weak.
Certainly this was true in the sense that the Sung never succeeded in
retaking the area inside the Great Wall held by the Liao dynasty. Then,
in the 1120s the Northern Sung was disastrously defeated by the newly
powerful Chin state and survived only as the truncated Southern Sung
state. Finally, in the late thirteenth century the Sung were overthrown
by the Mongol founders of the Yuan dynasty. However, strength and
weakness are measures not just of the internal power of a state but also
of the power of its enemies. The Liao, Chin, and Yuan armies were
among the most powerful military organizations of their times, and so
the survival of the Sung was itself a demonstration of impressive military
power.
The success of the Sung internally is even more striking. The Sung
generals almost never were in a position to mold policy, and when,
228
 Supervision - the intendants

as with the great general Yueh Fei in the early Southern Sung, they
appeared to be becoming too powerful, they were quickly brought to
heel. Furthermore, of all the six long-lasting Chinese dynasties, only
the Sung was never seriously threatened by internal unrest. The most
serious rebellion against the Sung was dealt with in a matter of months. l
Banditry was endemic but never endangered the state. This meant,
consistent with our paradigm, that the authorities were able to bring
sufficient, properly led forces to bear on resisting elements. Proper
leadership was crucial, both in the field and also at higher levels where
there had to be men who had the skills to plan and supervise action.
The Sung succeeded in keeping excessive power out of the hands of
the professional users of force. They had inherited a group of traditional
practices for guarding the guardians, and they used these practices in
creative ways, with some special Sung variants. One was the dilution of
military authority through the intermixture of civil agents in the chain of
command. In the Sung this control device appeared in several guises.
Civil officials might have concurrent control over military forces, and
they also were appointed to perform military functions in the military
chain of command. Moreover, for some functions the Sung established
an interlayering of civil and military levels of action.
This last practice was important to the supervision of law-enforcement
systems. At the bottom of the hierarchy of groups concerned with law
and order were the informal and nonmilitary village officers. Above
them, in the sense of being able to muster greater force, were the quasi-
military militias. On the formal level the lowest agent was the civilian
sheriff. Above him, in terms of control over force, was the military
patroling inspector. Above the patroling inspector were the civil prefects,
who usually concurrently held posts in the military hierarchy that gave
them control over the local military forces. Also at the prefectural level
were a cadre of regular military officers. Finally, at the top of the
local hierarchy were the supervisory officials who were members of the
civil service. Such officials played a critical role in overseeing the per-
formance of officials in their jurisdictions and in transmitting infor-
mation between the center and the local authorities. These were the
men called intendants. They had authority over various aspects of the
Sung law-enforcement and judicial systems. The evolution of the posi-
tion of intendant, therefore, highlights the character of Sung political
development.

1 McKnight, "The Rebellion of Fang La."

229
 Law and order in Sung China

Origins of the system of intendants

The Sung founder, Chao K'uang-yin, was certainly one of the most
perceptive of dynastic founders. He recognized the problems that had
beset the late T'ang and, in seeking to reunify China, made a succession
of wise choices. One of his early decisions was to concentrate first on
recapturing south China before attempting to confront the Sino-foreign
states on his northern and western frontiers. The military problems of
reconquest in the rich southern region proved relatively minor. Chao
had decided to move first in this direction so that he could gain con-
trol over the region's resources. Therefore he needed some mechanism
whereby to take possession of southern wealth quickly and effectively.
The device he used, the intendancy, was typical of his approach to
governing. Chao borrowed an office and a function with ancient roots in
the Chinese tradition and shaped them to his needs.
Even before the unification of the empire in the third century B.C., the
rulers of some of China's states were dividing newly acquired territory
into units under direct central control (the chun-hsien system) rather than
granting them to their supporters. These attempts by Warring States
rulers to concentrate power at the center, and the countervailing resist-
ance of local political figures seeking to retain their power, are early
examples of a tension that remained constant in imperial China, between
the desire of rulers to centralize power and the interest of local notables
to decentralize it. Under the chun-hsien system of dividing China into
units of administration staffed at least in part from the center, the
territory of the state was divided into commanderies {chun) that were
further divided into districts (hsien). After his victory, the first unifier of
the empire, Ch'in Shih-huang-ti (r. 246-210), extended this practice to
the whole state. Despite some minor changes in terminology in later
times - the term commandery {chun) being replaced in most dynasties by
the term prefecture (chou or fu) - this system has, in its basic outlines,
remained in use ever since. The only major innovation introduced by
later rulers was an intermediate level of administration between the level
of the prefecture and the central authorities. 2
The successors to the Ch'in, the rulers of the Han dynasty (203 B . C -
A.D. 220), introduced this new element when they appointed inspectors
to tour the empire. For this surveillance function they divided the empire
into thirteen circuits, each made up of commanderies. These inspectors
were members of the Censorate, charged with investigating administra-
2 For the material here on the intendants I am deeply indebted to Professor Winston Lo.
Indeed, the background material on the position of the intendants is almost entirely from
his article "Circuits and Circuit Intendants."

230
 Supervision - the intendants
tive abuses, but they themselves had no power to discipline the governors
of the commandaries and, indeed, ranked far below these governors in
the bureaucratic hierarchy. In the Later Han, as the powers of the
central government waned and instability in local regions put a premium
on the state's ability to react quickly to problems, there was a tendency
to allow those units originally charged only with surveillance to exercise
some control over the mobilization of resources and general administra-
tion. Later dynasties borrowed the Han surveillance model, hoping to
recreate the glories and powers of perhaps the most admired dynasty in
Chinese history. Under these later states, however, as under the Han,
this supervisory level of administration was at times irregular in status,
staffed by individuals appointed on commission, and in general had no
administrative functions.

Sung intendants
The creation of a permanent set of intermediate functionaries and
the beginning of their use of administrative powers came only with
the Sung. The Sung founder, needing to exercise surveillance over the
southern regions during the course of reconquest, and also to mobilize
the resources of the south for use in his further campaigns, used the
earlier model of the intendancy of a circuit. The first category of Sung
circuit intendant to be created, and the model for the others, was the
fiscal intendant {chuan-yiin shih), created by the founder for his southern
campaigns. The title itself had been in use under the T'ang for the
officials whose original duty was to oversee the shipment on the Grand
Canal of tax grains for the T'ang capital. These functionaries expanded
their powers during the late eighth century until they became the key
fiscal officials of the state. Eventually the title was used also for officials
in charge of provisioning the army. The Sung built on this foundation.
As the Sung armies moved into new territories, the fiscal intendant
charged with provisioning the armies would take control of all local tax
registers and other sources of fiscal information.
After the effectiveness of these officials had been proved by experience
in the newly acquired peripheral areas, they were appointed as regular
officials in the core regions of the empire. By the end of the tenth century
the Sung empire was divided into some fifteen circuits, their boundaries
coincidental in many cases with the boundaries of the states that had
formed the Five Dynasties or with the boundaries of natural regions. The
number of these circuits increased somewhat over time, reaching twenty-
three by the reign of Shen-tsung (r. 1067-85). During the Southern
Sung, after the dynasty had been driven out of the north by the Chin
231
 Law and order in Sung China

armies, there were seventeen circuits. There were occasional changes in

the size and boundaries of some of the Sung circuits, though others were
relatively stable, and some prefigured the provinces of later dynasties.
Initially created to mobilize resources, these fiscal intendants proved
useful also as agents for exercising general surveillance over local govern-
ment officials. The fiscal intendants did at times perform nonfiscal func-
tions, but their basic task, the mobilization of resources, remained clear
and specific. Rather than grant them exclusive control over new and
differing functions, the Sung authorities chose instead to create other
sorts of functionally specialized intendants to share surveillance over
other sorts of state concerns. This choice reflects the early Sung rulers'
grasp of the problems of power. By appointing several types of intend-
ants with overlapping (though not necessarily identical) territorial
jurisdictions but with differing functions, the central authorities created
a system with built-in checks and balances. Furthermore, it was quite
common for the different sorts of intendants active in an area to have
their headquarters in different cities within the circuit. At no time was
any one of them placed in a superior position to the others. Intendants
were colleagues, albeit sometimes jealous and competitive colleagues.
Here, as in many aspects of the law-enforcement system, we can see one
of the distinctive traits of the Sung pattern of governance: a willingness
to sacrifice a degree of administrative efficiency and control in order to
prevent the growth of excessive power in local hands.

Military and fiscal intendants

At various times and places during the Sung, different kinds of intend-
ants were appointed. In addition to the fiscal intendant, two types stand
out, the military intendant and the judicial intendant. In the core areas
of the empire, the fiscal intendants remained the key circuit officials
during the Sung. Although their functions were shared by others, they
were at least first among equals. In the border regions, however, the
fiscal intendants were overshadowed by the military intendants. 3 These
officials had both military administrative and military command func-
tions. Beginning with the reign of the founder, military intendants were
appointed sporadically to circuits along the northern and western
frontiers of the empire. By the beginning of the eleventh century they
had been appointed to circuits along the whole northern frontier and in
Szechuan. A number of new intendancies were created during the reign

3 The material on military intendants is from Michael McGrath, "Military and Regional
Administration in Northern Sung China" (Ph.D. diss., Princeton University, 1982).
232
 Supervision - the intendants
of Chen-tsung (r. 997-1022), but it was under Jen-tsung (r. 1022-63)
that they became a regular part of Sung administration. Finally, late in
Jen-tsung's reign, military intendants were appointed in the interior
circuits of the empire, where there were few imperial army troops and
where they were overshadowed by their colleagues, the fiscal intendants.
Like the fiscal intendants, the military intendants could at times be
called on to exercise surveillance over areas outside their military con-
cerns. This pragmatic use of intendants in differing roles fitted well
with the general Sung distaste for allowing any single type of official
to monopolize an important function. Furthermore, because they were
military intendants, it is hardly surprising that they at times were
involved in bandit suppression and other facets of the law-enforcement
process.
During the Northern Sung, however, the most important intendants
for law enforcement were, without question, the fiscal intendants and the
judicial intendants (t'i tien hsing-yii shih). They shared all of the law-
enforcement-related functions. Indeed, the sources very often pair the
two titles when describing the intendants' actions. If there were any
differences, it was perhaps that the judicial intendants handled more of
the routine aspects of judicial supervision.

Judicial intendants
The origins of the judicial intendants can be traced to the appointment
by the first Sung emperors of special judicial envoys to tour the areas
under Sung control. The sending of such envoys was an ancient practice,
which began at the latest in the Han dynasty and had been used also by
the T'ang and the Five Dynasties. 4 When in the Sung it became routine
to appoint such men and their tours were confined to the boundaries of
particular circuits, the office of judicial intendant was born. During the
first century of the Northern Sung, the office of judicial intendant existed
in most years but also was formally abolished on several occasions. By
the late Northern Sung, the office had become a permanent institution
and remained such during the Southern Sung period. 5 The sources give
the impression that for the most part civil officials were appointed to this
post, though sometimes officials from the military hierarchy were used. 6

4 For the Five Dynasties policy, which was the immediate predecessor of that of the Sung,
see Ou-yang Hsiu, Hsin Wu-tai Shih (Beijing: Chung-hua shu-chu, 1974), 30.327.
5 See HCP 34.9a, 106.ib, 109.6a, 113.13b-14a; WHTK 166.1445; $$ 5-93> 9-9a> 9 J 2 a ;
Sun Feng-chi, Chih-kuanfen-chi, 47.11b-13a.
6 HCP 120.ib (1037). See also Ss 34.649, in which in a comment dated 1136 it says that
"the military official judicial intendants were reestablished."

233
 Law and order in Sung China

As their title indicates, the judicial intendants were concerned with

judicial process. Their basic charge was to see that the judicial and law-
enforcement processes were fair and effective. This charge included
supervision, that is, the critical evaluation and direction of those officials
doing judicial or police work in their circuits; surveillance, in partic-
ular the systematic observation of and reporting to superiors about the
actions of law-enforcement and judicial functionaries; investigation of
certain sorts of legally important matters; review of decisions made by
judicial functionaries lower in the hierarchy; recommendations concern-
ing subordinates, including recommendations that they be appointed to
or removed from certain posts; verification and certification of the claims
of subordinates regarding legal and police actions and claims that might
lead to the granting of rewards; adjudicating and settling certain kinds of
legal cases; appointing certain police officials under some conditions;
and, especially during the Southern Sung, coordinating large-scale
bandit-suppression campaigns.
The judicial intendants' sphere of activities was wide but limited. The
general Sung policy of not allowing the unification of functions was
reflected in this office, as in others. The official Hsu Ching-heng, in
attacking a judicial intendant's request for general control over admin-
istration in a section of the empire put it neatly:
Judicial intendants travel around their jurisdictions. They may accept com-
munications and cases for judgment, or they may examine punishments and
jails. [But] matters not within the purview of their office, being matters of law,
are all sent to the appropriate overseeing intendant's office for action. The
judicial intendants do not dare assume sole control.7

Review of legal cases

The judicial intendants' functions thus were very similar to those of the
fiscal intendants when the latter officials dealt with judicial matters. As
the general description in the Sung History indicates, earliest in terms of
time, and always an area of concern for the judicial intendants specifi-
cally, was the surveillance of the legal process in the prefectures and
districts. This involved supervising the trial process itself, in part by
reviewing reports sent from subordinate units to the circuit head-
quarters. Under Sung law the magistrate of a district had the authority
to make final judgments in cases in which the penalty did not exceed one
hundred blows of the heavy rod. For cases calling for heavier penalties,
from penal servitude and up, such officials could make provisional

7 Hsu Ching-heng, Heng-Vang chi, SKCSCP 1975, 9.3a.

234
 Supervision - the intendants

decisions, but these had to be reviewed at higher levels. Prefectural

administrators had the authority to make final judgments on cases of
penal servitude or less, but they also had to submit more serious cases,
those calling for exile or death, to the circuit authorities for review. 8 An
annalistic history of the Northern Sung reports a case from 1026 in
which
a commoner from Shang-tang District [in Ho-tung] whose surname was Wang
was slandered with having secretly killed his stepmother. The case had already
been written up, and the clerks all thought that the evidence was adequate and
not in doubt. Only Wang Shu [the administrator of An Prefecture] said that
there were some doubtful points. When the case reached the judicial Intendant,
Tu Yen, he rechecked it and discovered the real killer. Because of this, Wang
Shu wrote the "Record on Judging Cases" to forewarn judicial officials.9
Judicial intendants thus automatically participated in reviewing
the capital cases submitted to them. Under a rule found in the early
thirteenth-century code ChHng-yuan tyiao-fa shih-lei, they were to write out
reports on these cases, giving the key facts and the sentence and indicat-
ing whether or not there had been delays in the case and whether or not
it had been remanded for reconsideration and the date of such remand-
ing. These reports were sent quarterly to the Secretariat and the Ministry
of Justice. 10 This order was in fact obeyed at least sometimes. One
source tells of a prefect who lost his position when someone at Court
pointed out that his judicial intendant had had to remand forty cases in
six months. 11
In addition to these automatic reviews at higher levels, the accused
parties or their families could also secure reviews or retrials by raising
accusations of judicial impropriety. Such initial retrials might be carried
out at the prefectural level if personnel were available who were not
debarred from participating because they had been involved in the initial
investigation. If such people were not available, then the fiscal or judicial
intendant of the area was responsible for selecting someone to conduct
the retrial. 12 He also might be asked to investigate cases involving
charges against officials in his jurisdictions. 13 Finally, the intendant
could be ordered by the Court to investigate reported problems. In the
early 1180s there had apparently been complaints that local adminis-

8 SS 199.4972; SHY, hsing-fa 3.13a.

9 HCP 104.6b-7a. It is perhaps worth noting that as a result of this case Tu Yen was
appointed to an important position in the Ministry of Justice; see HCP 106.2b.
10 TFSL501.
11 HNYL1968.
12 SHY, hsing-fa 3.55b; HCP 72.4a.
13 WHTK 166.1445 (1014).
235
 Law and order in Sung China

trators were using village officers for inappropriate tasks. The Court
ordered the judicial intendants to investigate. In the same document
these judicial intendants were ordered to look into cases in which local
administrators had confiscated the property of convicted criminals, to
make sure that such confiscation was in fact justified under the law. 14
When people appealed cases to the central authorities, as they had the
right to do, they still ordinarily had to pass through the circuit offices
first. A decree of 1005 ordered that men who came to the capital to ask
for a review of legal cases had to write descriptions of the ways in which
they claimed that the cases had been mishandled on the district and
prefectural levels and the date on which the matter had been reviewed at
the offices of the circuit intendants. 15 The general practice seems to have
followed a pattern described in a thirteenth-century order, which says
that if the appeals were frivolous and aimed simply at evading punish-
ment, the intendants should, after having conducted a thorough review,
send down an order for enforcement. If, however, there was evidence of
wrongdoing, delay, or incompleteness, they should write a memorial
asking for a central judgment, which would then be enforced. 16 This role
is reflected in a memorial from the judicial intendant of Ho-tung Circuit,
who reported in 1012 that
all the capital cases in the various circuits are, at the end of ten days, reported to
the offices of the fiscal or judicial intendants. If the fortnight period is just
starting and the office of the intendant involved is very far away from the
reporting prefecture, the prisoner may then be held for half a month so that [by
the time the intendant becomes involved] there may be uncertain circumstances
that can no longer be elucidated. From now on it is our hope that the prefectures
will report such cases to the intendant's offices on the very day [that their
decision is made].17
The judicial intendants who reviewed cases were expected to write
reports dealing with the facts and the law involved and report these to
the relevant authorities. Some idea of this process is conveyed in the
report on a case from the late 1070s. An acting judicial intendant
reported on a case in which a son had been killed, but the parents had
been persuaded to accept payments to settle the case out of court. For
having settled out of court both parents had been sentenced by a lower
court. The lower court cited a statute covering cases of mourning-degree
relatives 18 who reached private settlements after their relatives had been
14 SHY, hsing-fa 2.121a.
15 SHY, hsing-fa 3.13a.
16 SHY, hsing-fa 5.48a-b.
17 SHY, hsing-fa 6.53a.
18 The Chinese differentiated relatives according to degrees of mourning. These degrees

236
 Supervision - the intendants

killed. Under this statute those guilty were liable for a sentence of two
and a half years of penal servitude. The intendant in this instance goes
on to cite a statute defining the meaning of the term son for legal
purposes and a statute stating that those who deliberately kill their
junior relatives could be punished with strangulation. He also used a
statute stating that parents or grandparents who beat to death dis-
obedient children were liable for one and one-half years of penal servi-
tude. If the elders in these cases had intended that the disobedient ones
die, the penalty would be increased by one degree, to two years of penal
servitude. The intendant thus suggested that it was unjust to punish
parents guilty only of having reached a private settlement after their
children had been killed by someone else more severely than parents
who did the killing themselves and so asked that the central authorities
promulgate instructions to guide local magistrates in balancing conflict-
ing rules that called for different levels of penalties. 19

On-the-spot investigations
When the regularly submitted prefectural reports indicated that there
were problem cases in a judicial intendant's jurisdiction, he was to go, or
to send authorized subordinates, to investigate on the scene. Intendants
had this responsibility from the late tenth century onward. A decree of
991 states that "where there are doubtful cases that have not yet been
decided, the judicial intendants are to go to examine them." 20 Five years
later the power of the intendants to send investigators is made clear in a
description of the origins of the relevant decree:
The emperor had heard that the various prefectures, when trying to pass
judgments on capital cases about which the circumstances were in doubt, were
afraid to send such cases because they feared that the cases would be reversed.
Subsequently he [the emperor] issued a decree that in doubtful capital cases the
fiscal intendants should select subordinates who were knowledgeable and experi-
enced in the law to pass judgment on such cases. If there were cases that truly
needed to be memorialized to the center, then that should be done.21

determined what mourning rituals were to be performed on the death of those relatives
that fell within the mourning-degree circle. For examples of mourning charts, see P. L.
F. Philastre, Le Code annamite (Paris: E. Leroux, 1909). This is a translation of the code
of the Nguyen dynasty in Vietnam, but it is at almost all points a copy of the code of
the Chinese Ch'ing dynasty.
19 HCP 299.i2a-b.
20 WHTK 166.1445.
21 SS 199.4972. See also the detailed decree dated 1003 and recorded in SHY, hsing-fa
3.11b-12a, and the description of an example of this process in Huang Huai, Li-tai
ming-ch'en tsou-i (Taipei: Tai-wan hsiieh-sheng shu-chu 1964), 218.17b.

237
 Law and order in Sung China
The judicial intendants also participated in inquest procedures. They
were to go - or probably, in the ordinary case, to send a subordinate - to
conduct inquests on people who had died as a result of judicial torture
during trial. This was certainly true in the Southern Sung, and a similar
policy might have existed earlier. An order of 1159 also made judicial
intendants responsible for investigating certain very difficult sorts of
homicides, those in which there were no witnesses or in which the corpse
had not been subjected to autopsy. Such cases were to be memorialized
for a ruling from the central authorities. This order incidentally reveals
the intendants' power to dispatch subordinates from their offices to
conduct investigations. 22
This investigatory and judgmental responsibility was not confined to
capital cases. People also could and did pursue legal settlements of other
issues. We learn from an order of 1038 that people could come to the
circuit intendants' offices to resolve disputes over property. In such cases
the intendants were supposed to dispatch officials to investigate. How-
ever, such cases were supposed to be accepted only after the beginning of
the tenth month, when the harvest was presumably completed, so that
agricultural work would not be hindered. 23

Periodic visitations
Intendants, and at times their designated agents, also were supposed to
pay regular visits to the territorial units in their jurisdictions, accom-
panied by their retinues of soldiers. (An order dated 1037 says that if he
were a Court-rank official, a judicial intendant was to have a guard of
thirty soldiers; if a military official, twenty-five. 24) When this guard went
with the intendant on tour, the local authorities were responsible for the
costs. The central authorities were understandably concerned that local
officials might attempt to influence the intendants' reports by giving
them especially lavish treatment. Indeed, the order of 1007 that re-
established the office of judicial intendant specifies that "the prefectural
officials are not allowed to go out to greet the arriving intendants or to
assemble with them." Records of the tours were kept on logs written on
imperially stamped paper, and they were reviewed at the capital. 25
When on such tours the intendants were expected to examine out-
standing cases, review judicial records, and even interview prisoners in
order to discover cases of judicial abuse (or judicial excellence). During

22 SS 201.5024.
23 SHY, hsing-fa 3.45a.
24 HCP 120.1b.
25 WHTK 166.1445.

238
 Supervision - the intendants

their visits, the intendants were also expected to expedite cases, by

rendering decisions on cases that seemed clear-cut. In the words of the
twelfth-century official Li Kuang:
On the day when the imperial order arrives, the judicial intendant should
circulate in person through the subordinate prefectures, himself going to the jails
and taking the names of the current prisoners. If among them there are some
about which the major points are clear, with only minor points not resolved, he
should weigh the facts and decide the case, not allowing any further delay.
During these same visits the intendants or their authorized agents
were to check on the physical treatment of the prisoners, including the
enforcement of rules on the cleaning of jail cells, the washing of fetters,
and the size and condition of the prison instruments. 26
Probably in many cases the intendants sent deputies to make these
inspections. This is suggested by numerous comments in the sources
about choosing appropriate subordinates for these tasks. Moreover, the
inspections were supposed to be made annually, but a decree of 1090
specifies that the judicial intendant on his tour was to visit every pre-
fecture and district once every two years. Presumably, deputies handled
the other visits.27

Welfare of prisoners
This system of visitations was only one of the intendants' more general
responsibilities for the welfare of prisoners. Intendants reviewed the
prisoners' records, which were to be sent to them every ten days,
checked for doubtful cases, and impeached officials if there had been
unjustified delays (or wrong decisions) in the judicial process. 28 Accord-
ing to the Sung History.
At the beginning of the Ch'un-hua period [990-94] the office of judicial intend-
ant was first established in the various circuits. These men had control over the
prefectural reports of prisoner records which were sent in every ten days. When
there were problematic cases that had not been settled, they were quickly to send
an observer to look into the situation. If the prefectures and districts delayed and
did not reach a judgment, or if on examination the information was not accurate,
the chief officials were to be impeached.29

26 Li Kuang, Chuang-chien chi, 12.21-b.

27 HCP 442.8a.
28 WHTK 166.1445 (991)- For continuation of these functions in later times, see, for
example, HCP 66.7a (1007); SHY, hsing-fa 6.55a-b (1047), 5 5 ^ - 5 ^ (1050); WHTK
167.1452 (1119).
29 SS 199.4971.

239
 Law and order in Sung China

The intendants were also responsible for conducting inquests on those

who died in jail. In 1047 t n e fiscal intendant of Ho-tung Circuit reported:
In recent years many of the men accused in criminal cases are incarcerated for
months. They do not have [enough] food and drink, and they are subjected to
judicial torture and die. These deaths are concealed by higher and lower func-
tionaries, so that when it comes time for the inquest, the deaths are listed as
deaths from illness. I ask that you order the fiscal and judicial intendants, when
they go on their tours of the prefectures and districts, first to go into the jails and
interrogate the prisoners. When prisoners die, inquest forms should immediately
be sent up to the two intendancies' offices to be reviewed. If the circumstances
seem unclear or if it is indicated that there are wounds from judicial torture, then
an official should be dispatched to conduct a detailed inquiry and to secure
justice.30
From at least the twelfth century, the intendants were also supposed
to keep records of the number of prisoners in the jails in their circuits.
An official active during the closing years of the Northern Sung sug-
gested that the judicial intendants be responsible for annually reviewing
the number of prisoners who had died. Mu-jung Yen-feng said:
I have noted that the emperors, out of pity for prisoners, have established the
law on the investigation of prisoners who have died of illness. . .. We ask that
you order the various circuits' judicial intendants at the end of each year to
assemble the records on the numbers of prisoners who have died of illnesses.
On the basis of these reports, local officials might be either punished or
rewarded. These figures would be reported to the Court and would
result in sanctions against officials in units where the death rate was
excessive and commendations for those where the death rate was very
low.31 We know from other records that such systems were in fact used.

Rewarding effective jurisdictions

The other side of this coin was the intendants' involvement in systems
for rewarding jurisdictions that handled judicial matters especially effec-
tively. Intendants participated in key ways in the investigation, verifica-
tion, and certification of law-enforcement functionaries' applications for
rewards for the successful performance of their duties. The intendants
were responsible for verifying the accuracy of the claims and for guaran-
teeing the performance of law-enforcement agents. 32 They might also

30 SHY, hsing-fa 6.55a-b; HGP 73.2a (1010).

31 Mu-jung Yen-feng, Li-wen Vang cki, Ch'ang-chou hsien sheng ts'ung-shu, 10.2b.
32 SHY, ping n . i o a (1024), 13b (1032), 26a (1059).

240
 Supervision - the intendants

sometimes recommend new systems of rewards. 33 The best-known policy

for rewarding such units was the "empty jails" system, a policy, des-
cribed more fully in Chapter 12, according to which units that had no
prisoners in jail, and therefore no caseloads, were rewarded. An imperial
order establishing this policy was issued in the late tenth century and
continued in force thereafter. Such reports of empty jails in local areas
were forwarded to the Court by the judicial intendants after they verified
the accuracy of the claims. 34

Recommendations on personnel and policy

Their position in the hierarchy that dealt with the judicial process made
the intendants key officials in submitting recommendations to higher
authorities on all aspects of the judicial system. Such recommendations
might pertain to judicial procedures, including procedures for investiga-
tion, the nature of case records, and ways of bringing accusations or
of reporting cases. 35 The intendants' recommendations on personnel
were part of general Sung policy, since candidates for promotion from
the executive to the administrative class in the bureaucracy (the most
important single step in promoting an official) required recommenda-
tions from their superiors, including circuit intendants. 36 The executive
class, comprising the ninth and lower eighth ranks of the nine grades of
civil service hierarchy, contained the majority of Sung officials. These
functionaries mainly administered policies created by others. Promotion
to the elite group of policymaking administrative officials was the avidly
sought goal of most Sung bureaucrats. This step could not be made on
the basis of seniority alone, nor simply on the basis of success in the
more routine measures of achievement. Having a set number of recom-
mendations from superiors was necessary.
In addition to these special recommendations, the judicial intend-
ants also made regular reviews of the officials under their jurisdiction.
According to the rule found in the early thirteenth-century Ch'ing-yuan
tyiao-fa shih-lei, they were to report twice each year, in the fourth and
tenth months, on the conduct of the senior officials in their circuit. Given
the timing of these reports, they probably were to be based on the twice-
yearly investigation of judicial affairs in each prefecture and district. 37

33 HCP 129.5a, 2i6.i4a-b, 302.6a, 328.2a; SHY, ping 11.19a.

34 HCP 72.16a; WHTK 166.1445.
35 SHY, hsing-fa 3.11b (998), 4.7a (1013), 6.59a (1114); SHY, ping 11.15a (1034); HCP
302.6a (1080).
36 On this process, see Lo, An Introduction, pp. 165-70.
37 TFSL 226.

241
 Law and order in Sung China

In regard to law enforcement, the intendants frequently made

recommendations about the establishment and locations of local
law-enforcement posts and about jurisdictional boundaries between
law-enforcement units. They sometimes recommended persons for
bandit-suppression campaigns and made more general suggestions about
personnel policies. Their knowledge of the men working as functionaries
in a large region made the intendants uniquely qualified to render
judgments about appointments. They might also be charged with super-
vising the provision of appropriate quarters for these law-enforcement
personnel. 38
Intendants also investigated and made suggestions about militia per-
sonnel and constables. The official Mu-jung Yen-Feng asked that the
emperor "especially order the judicial intendants to assign officials the
responsibility for examining, selecting, and retaining strong stalwart men
and dismissing the old and weak." 39 Perhaps most importantly, the
judicial and fiscal intendants regularly assessed the capabilities of law-
enforcement officials like the local sheriffs and patroling inspectors, and
recommended the replacement of those unfit for duty. 40 This practice is
indicated in a report from 1079:
The emperor noted that because there had been natural disasters this year in the
circuits of Huai-nan, Chiang-che, and Ching-hu, people would find it difficult to
feed themselves come spring, and it would be easy for bandits to arise. An order
was to be sent down that the various circuit fiscal and judicial intendants and the
military administrator offices (ch'ien-hsia ssu) examine the physical conditions of
those serving as patroling inspectors and sheriffs and memorialize the names of
those found not competent.41

Supervision of penal practices

T h e intendants' duties also extended to penal policies and practices. T h e
intendants were to see that local authorities did not abuse their power to
punish, for example, by increasing beyond the legal limits the number of
blows inflicted. 42 They oversaw, as well, the proper and careful guarding
of convicts in transit. 4 3 Early in the eleventh century, one intendant
recommended a better system of supervising the transfer of convicts, so

38 Fan Ch'un-jen, Fan chung-hsiian chi, SKCSCP ed. 6.74.

39 Mu-jung Yen-feng, Li-wen fang chi, 10.1b.
40 SHY, ping 11.15a (1034), 11.23b-24a (1050); SHY, chih-kuan 48.64b; HCP 116.12b
(1035), 216.14b (1070), 271.14a (1074), 275.4b (1076), 328.2b (1082), 334.19a (1083),
338.10a (1083), 34<>-5b (1083), 389.6b (1086).
41 HCP 271.14a.
42 HCP 82.16a.
43 SHY, hsing-fa 4.i2a-b.

242
 Supervision - the intendants

that people could no longer abuse the prisoners during transport. 44 This
sort of concern is illustrated by a recommendation submitted in 1028: An
intendant asked that when men were being taken to their place of penal
registration, a record be made of the money and goods that they were
taking along in order to defray their expenses en route. The escort-
ing guards would be responsible for these registers and would deduct
expenses as they were made from the register, to prevent theft of the
convicts' goods. 45
Intendants might also be made responsible for assigning to specific
prefectures men sent for penal registration. 46 This duty was related to
their responsibility for seeing that no given prefecture was forced to
accept a dangerously large number of serious criminals. An order of 1044
regarding the system of penal registration stipulated that the fiscal and
judicial intendants were regularly to assess the working of the system,
and if in a given prefecture "there are too many men, a report should be
sent, on the basis of which men could be transferred to jurisdictions with
few registered criminals." 4 7
In the mid-eleventh century the intendants in certain southern and
central circuits were given the task of examining the men being held in
various sorts of penal registration in their areas. If they found potentially
dangerous convicts gathered together in certain prefectures, they were to
transfer such men to other areas, in order to reduce the threat of
collusive misbehavior. 48 Their access to and familiarity with the records
of convicts in their areas made them the logical persons to report on
convicts who were to be transferred. 49
The intendants were in an ideal position to recommend penal policies
to the central government. They made suggestions about systems of
communication and information 50 and about the treatment of prisoners. 51
It was an intendant who, in 1058, submitted a highly critical report
about the treatment of prisoners at the Sung maximum-security facility
on Sramana Island off the north coast of Shantung. 5 2 They also might
advise using specific penalties in some areas, as when an intendant asked

44 SHY, hsing-fa 4.7a.

45 SHY, hsing-fa 4.15a.
46 SHY, hsing-fa 4.4b. Intendants could also ask that men be registered in certain specific
prefectures in order to meet labor needs there. See HCP 256.2b.
47 SHY, hsing-fa 4.21a.
48 SHY, hsing-fa 4.21a. See also SHY, hsing-fa 4.31b-32a.
49 SHY, hsing-fa 4.25b.
50 SHY, hsing-fa 3.11b.
51 In 1007 it was an intendant who recommended standards for certain of the fetters worn
by prisoners; see HCP 67.18b; SHY, hsing-fa 6.77b; WHTK 166.1445; Chiang Hsiao-
yu, Huang-ch'ao lei-yuan 20.7b.
52 SHY, hsing-fa 4.23b-24a.

243
 Law and order in Sung China

that the families of evil habitual criminals in his circuit be confiscated by

the authorities or when an intendant asked that those along the border
guilty of kidnapping people for later sale outside the borders be executed.
(His request was not approved.) 53 Or they might ask for lenient treat-
ment of certain groups. An intendant petitioned successfully that in
prefectures that had suffered natural disasters, people driven by want to
commit robbery be spared the punishment due under the regular laws. If
these people "had shared their spoils with their families, even if they
were liable for heavy punishments, their cases should be examined
individually. If they are not basically evil and the circumstances of the
crime are not serious, they should be released after being beaten." 54
Intendants were also involved in the transferring or freeing of those
benefiting from imperial amnesties. Amnesties, which will be discussed
in more detail in Chapter 14, were issued frequently during the Sung.
On receipt of a letter of amnesty the circuit authorities informed their
subordinate subunits. These subunits in turn prepared statements con-
cerning the behavior and crimes of those under their control. Those with
good records could benefit from the amnesties; those with bad records
could not. The assessment of who was to benefit was a duty shared by
the intendants and the officials of the prefectures and districts. 55 The
intendants also could ask the central authorities to extend grants of grace
to areas under their supervision, especially in times of natural or other
disaster, 56 and sometimes they were called on to investigate reports that
some persons who ought to have benefited from amnesties were being
improperly detained. 57 A decree of 1070 said:
With regard to men held in [the kind of penal registration called] registered
control, the judicial intendants are ordered to select officials for the various
prefectures who will, together with the currently incumbent prefectural officials,
investigate the original crimes involved and make judgments as to the law and
the facts. They are to determine whether there really are men being held illegally
in registered control by the prefectures. Such men are to be released as soon as is

Supervisory and administrative responsibilities

The other side of this duty was the intendants' responsibility for report-
ing failures to capture criminals. At times they were required to submit
53 SHY, hsing-fa 4.20b; HCP 91.7b.
54 SHY, hsing-fa 4.19b. See also SHY, hsing-fa 4.5b, 19a.
55 SHY, hsing-fa 4.23a-b (1051), 24b (1063), 27b (1080); HCP 12428.14a (1061); SS
24.450 (1127). See also Huang Huai, Li-tai ming-ch'en tsou-i, 217.9a.
56 SHY, hsing-fa 4.5b (1012), 19a (1034), 19b (1034).
57 SHY, hsing-fa 4.24b. See also HCP 217.6b.
58 SHY, hsing-fa 4.25b. See also 4.45a-b.

244
 Supervision - the intendants

biweekly reports on the number of bandits captured, but usually the

reports were due much less often.59 In 1066 the judicial intendants were
ordered to review reports on the number of bandits caught and not
caught, submitted when local officials had completed their terms in
office, and, on the basis of these reports, to recommend either rewards or
punishments. 60 It is not clear how well this policy was enforced, since in
1093 a judicial intendant recommended that when prefects who had
been in office for a year or more were leaving their posts, their record for
capturing bandits should be examined and reported to the central auth-
orities.61 Finally, intendants were to oversee attempts by officials in
their jurisdiction to capture bandits, to be certain that "if there were
deficiencies in the pursuit and anyone dares to conceal them, those guilty
would be punished for the crime of having violated regulations [which
called for a penalty equivalent to two years of penal servitude]." 62
Officials naturally wanted to avoid becoming involved in this system and
were tempted to conceal evidence of banditry in their areas. 63
As these examples should make clear, the intendants were key figures
in transmitting reports on particular cases 64 and generalized statistical
reports on the workings of the judicial and police systems passing from
the circuits to the agencies of the central government, and transmitting
via the circuits of some sorts of orders and instructions from the central
government for enforcement on the local level. Intendants also provided
the central authorities with a flow of information on affairs in local areas.
They investigated and reported on such matters as cases brought against
local officials, the numbers of prisoners currently incarcerated (including
names, crimes, dates of detention and so on), the numbers and locations
of vacant law-enforcement posts, the numbers and types of penalties

59 SHY, ping 11.21b, recommended frequent reports, but in the 1090s the Department of
Ministries recommended that reports be submitted twice a year; see HGP 460.1a.
60 SHY, ping 11.27a.
61 SHY ping 12.13a. Such reporting systems may have fallen temporarily into disuse
during the troubles that accompanied the fall of the Northern Sung. Between n 26 and
1131 the official Chang Shou asked that the districts be ordered to create registers
recording the names of the victims of robbery and the date (Chang Shou, P'i-ling chi,
3.6a-b).
62 HCP 125.3b.
63 SHY, ping 11.25b; HCP 176.8a. Of course, the intendants themselves were also
subjected to periodic evaluations of their performance. Lo, An Introduction, p. 243,
translates the standard form for a merit evaluation of circuit intendants. As might be
expected, it includes such items as whether the intendants had committed any
improprieties in handling lawsuits, and whether they had impeached or failed to
impeach subordinates guilty of crimes of bribery serious enough to warrant
punishments of exile.
64 See, for example, an order of 1024 that the intendants inform the prefectures of a decree
on the local investigation of bandits (SHY, ping 11.10a).

245
 Law and order in Sung China

inflicted in the area of their jurisdiction, problems with heretical reli-

gions, and other aspects of the general security situation.65 They eval-
uated the persons in their jurisdictions, not just the officials, but also the
troops who were, after all, the basic force for suppressing large-scale
disorder.66 As we noted, they sent in regular reports on banditry, some-
times twice a year and sometimes more often, giving the numbers caught
and not caught. 67
When suggesting a general policy for keeping track of local crimes, the
official Chang Shou gave the intendants an important supervisory role:
I now ask that. . . each district establish a register on which will be written the
name of the family robbed and the date. When the criminals are seized, a red
mark should be made after the entry. The prefectural controller general [t'ung-
p'an] should examine these registers every quarter. At year's end the judicial
intendants should assign officials to collect these registers and check to see how
many bandits have been captured and how many have not been captured. In
accord with the success in capturing [them] and the time spent in so doing,
reward or punish the local officials.68
Intendants could also be asked to act as judges. Sometimes they were
even given the power to condemn men to death, to execute them, and to
report thereafter to the central authorities. 69 More generally they had
the power to settle certain lesser cases without necessarily consulting
the central authorities. In the late eleventh century when there was a
major restructuring of the government, this power was reaffirmed by an
imperial order that "local cases that do not require a memorial are to be
settled by the judicial intendants." 70
During the Sung, intendants eventually began to make certain admin-
istrative decisions. In times of crisis they might be given the power to
select and appoint local law-enforcement agents without waiting for the
approval of the central authorities. 71 In response to an official who
submitted a report concerning problems in mining regions, the Court
ordered the fiscal intendants
to assess on a regular basis the patroling inspectors, sheriffs, and bandit-capturing
servitors. If among them there are some who are incapable by reason of illness,

65 HCP 133.2b (1040); SHY, ping 11.21b (1046), 27a (1066); SHY, hsing-fa 4.19a (1034)
6.55b-56a (1050), 57b (1094).
66 HCP 338.10a.
67 HCP 460.1a (1091), 443.1a (1090).
68 Chang Shou, P'i-ling chi, TSCC ed., 3.37-38.
69 SHY, hsing-fa 4.21a; HCP 1143.10a.
70 WHTK 167.1450.
71 HCP 143.19a-22a (1043). This also seems to have been the case in 1076; see HCP
275.4b.

246
 Supervision - the intendants

stupidity, and so on, the fiscal intendants are to select men from their jurisdiction
to replace the incompetent. Then the intendants should submit a report.72
More significantly, especially during the Southern Sung, the intend-
ants moved from monitoring the bandit-capturing activities of other
officials to personally coordinating the major antibanditry campaigns. 73
The first step along this path was perhaps the order of 1043 gi y m g t n e
judicial intendants sole control over the bandit-control activities of the
patroling inspectors, and there are other examples from the Northern
Sung. In 1062, for example, judicial intendants were given sole charge of
controlling bandits in certain prefectures and, concurrently, of super-
vising the salt monopoly. Before this, "in Chiang-nan West and Fukien
circuits the salt-smuggling gangs included thousands of men who have
gone about openly pillaging, killing, and injuring, and the officials have
been unable to stop them." 74 And in 1080, because of an increase in
banditry in the area around the capital, the judicial intendant there was
ordered to go in person to supervise their capture. In a decree issued a
few months earlier, the intendants had been given the authority to
dispatch fifty cavalry or infantry soldiers to capture robbers. 75 In the
Southern Sung this became a general pattern. 76 A judicial intendant
commented in 1180 that "recently commands have been sent down that
from now on, in the various prefectures and districts, if there are places
with many bandits, the judicial intendant of that region is to be per-
sonally responsible for mobilizing resources and capturing them." This
official noted that the judicial intendants did not have control over the
regular soldiers in their jurisdiction, but only over the local soldiers and
the archers. He pointed out that the intendant might have to mount a
sizable campaign at some distance from the location of these local
soldiers and asked that a force of some 670 regular soldiers be placed
under the control of the intendant for use in such campaigns. His request
was approved. 77
The importance of such a coordinating official was underscored by the
late Southern Sung official Hsu Ying-lung. In a passage that could
be duplicated in modern books on management, he emphasized that
the essential task of control is the proper selection of subordinates.
A superior should choose capable men and rely on their judgments,

72 SHY, ping n.23b-24a.

73 HCP 129.3b (1040).
74 SHY ping 11.26a.
75 SHY, ping i2.8a-b; HCP 305.13a.
76 HCP 143.27a.
77 SHY, ping i3.33a-b.

247
 Law and order in Sung China

thereby creating a unified chain of command to preclude infighting and

competition/ 0
We can get an idea of how this system worked from a long memorial
submitted by the Southern Sung official Ts'ai K'an (doctorate of letters
1166), who reported on his early tenure as a judicial intendant:
Last year, on the twenty-sixth day of the eleventh month, I reached the border of
I-chang District in Ch'en Prefecture [in Ching-hu South Circuit] to accept the
handing over of the affairs of this office. Since my arrival I have examined the
whole circuit's criminal cases, its judicial abuses, incarcerations, and delays for
those who could be currently sentenced. . . . Circuit intendants are also respon-
sible for capturing bandits. Recently there was an order that from now on if
bandits secretly arise . . . the judicial intendant is himself to devise ways to catch
them, so that they will not increase. With this order the responsibility of the
judicial intendant becomes even heavier. But to carry it out we have only the
local soldiers and the archers, and they are inadequate to destroy bandits. If you
assign heavy responsibilities but provide inadequate forces, how can there not be
trouble?... I have checked and found that in this circuit there are 2650 archers
and 2429 local soldiers. This is a fair number, but they are scattered among nine
prefectures and thirty-eight districts. This region has rugged terrain. If there is
an emergency, they cannot be fully mobilized. . . . When we had trouble with Li
Chin and Ch'en Tung we could dispatch only the archers and local soldiers from
the nearest few districts, who were not very numerous and in any case were
a rabble and not of much use. Therefore we called up the locally powerful
groups, but they also were unable to manage the bandits, and so we had to wait
on the arrival of a large military force from E Prefecture. Then the bandits
were destroyed. . . . I have noted that the judicial intendant of Kuang-nan West
Circuit, Hsu Hsu, has requested the transfer of regular circuit troops, some 500
men, to be put under the command of the judicial intendant's office and the
other intendants' offices. . . . The Court approved this. . . . The Kuang-nan East
judicial intendant. . . followed a precedent allowing him to call for the prefec-
tures to dispatch 50 imperial soldiers per prefecture, or for fourteen prefectures a
total of 700 men, to be used in arresting the bandits. . . . In this circuit we have
no such precedent. . . . I do not dare to ask that a special force be put under my
command, but I do hope that. .. you will allow me to order from among the
prefectures a force of 50 men to be sent to work with the archers and local
soldiers in making arrests.79

Conclusion
The patterns of responsibility and function we have described, estab-
lished during the Northern Sung, continued during the Southern Sung,

78 Hsu Ying-lung, Tung-chien chi, 8.16a-19b.

79 Ts'ai Kan, Ting-chai chi, 2.12a-14b.

248
 Supervision - the intendants
but the distribution of these duties among the various intendants dif-
fered. During the Northern Sung, the fiscal intendants played a major
role in the law-enforcement and judicial process; during the Southern
Sung, the fiscal intendants were much less important to judicial matters
than the judicial intendants were. In Southern Sung material on legal
matters, the fiscal intendants and other intendants such as military
intendants are mentioned only rarely. Rather, during the Southern Sung
it was the judicial intendants who submitted reports on bandits caught
and not caught,80 who reviewed and reduced sentences,81 who handled
the implementation of amnesties,82 who investigated rapacious clerks,83
who checked on the local handling of cases and impeached officials who
erred in their judgments,84 and who reviewed cases and reported on
delays.85 The judicial intendants or their appointed agents visited the
areas under their jurisdiction and checked on current cases, expedited
settlements, and freed prisoners or reduced their sentences. Eventually
the judicial intendants performed this last duty not only during the
summer heat but also during the dead of winter.86
The Southern Sung judicial intendants also participated in the inquest
process that was used in cases of homicide. Indeed, it was a Southern
Sung judicial intendant who suggested that all authorities use a standard
form showing the outline of a body so that local officials could mark the
location and character of wounds.87 They investigated the problem of
escapees, enforced the prohibitions against heretical religious sectaries,
oversaw the capture of bandits by the prefectures (setting their rewards
and the time limits for capturing the fugitives), and kept watch over the
quotas of men penally registered in their jurisdictions.88 Finally, the
Southern Sung judicial intendants were charged with mobilizing local
resources and capturing bandits.89 During the Northern Sung, the
judicial intendants had also performed these functions, but during that
earlier era they shared many of these tasks with fiscal intendants.
During the Sung the office of circuit intendant became increasingly
important to local administration. Intendants came to exercise certain
80 Chang Shou, P'i-ling chi, 3.6a-b.
81 SS 24.450.
82 SHY, hsing-fa 4 4 5 a - b (1131), 46b (1147), 51a (1165), 67a (1218); WHTK 167.1454
(1156).
83 SHY, hsing 4.63b (1203).
84 SHY, hsing-fa 6.78b (1142), 2.121a (1182).
85 HNYL 2568 (1148); SS 200.4994 (1174).
86 SHY, hsing-fa 5.39a (1163), 42b (1174), 41b (1189), 45a (1191), 4 6 ^ (1206), 47a
(1212), 47b (1213).
87 SS 200.4994 (1174).
88 SHY, hsing-fa 4.53a (1174), 63a (1203); TFSL 502.
89 TFSL 500; SHY, ping 13.33a (1180).
249
 Law and order in Sung China

administrative powers. At the same time, the Sung government, ever

concerned about losing power at the center, reduced the general role of
the fiscal intendants, in part by emphasizing the roles of other intend-
ants, such as the judicial intendants. The tension between the politics of
centralization and decentralization is reflected in the growing import-
ance of positions whose functions represented the interests of the center
but whose incumbents had always to be kept from becoming too powerful.
However, with the intendants as with other local officials, some
policies actually made them less effective. The local gazetteers show that
in the Sung, intendants, like prefects and other local officials, were in
office for very brief tenures. Figures in the K'uai-chi hsu-chih indicate that
in the late Northern Sung, from i ioo to 1123, 32 men were appointed to
the post of judicial intendant in Liang-che Circuit, and in the 135 years
from 1129 to 1264, 105 men were appointed to this post. Judicial intend-
ants thus held their appointments for 1 year on average. Figures from
the San-shan chih suggest that in the area covered by that gazetteer,
judicial intendants during the Northern Sung served an average of
slightly less than 18 months, and in the Southern Sung slightly more
than 18 months. 90 We might have a better picture if we assumed that the
bulk of the routine work continued to be done by subordinate officials
and clerks and that the judicial intendants were truly useful only when
there was a fortuitous combination of a competent and energetic official
and a crisis that caused the state to leave him in his post long enough to
be effective.

90 Liang K'o-chia, San-shan chih, SYTFCTS ed., p. 7843.

250
 8
Personnel selection

Introduction — general personnel patterns

To deal with disorders, the authorities must get enough adequately
armed and trained personnel to the site in time, but such men will be
useful only if they are properly led. Proper leadership implies experi-
enced leaders with sufficient training in practical skills, as well as appro-
priate aims and motivations. The worst possible situation would be to
have field-level leaders with practical expertise who held goals hostile to
the general aims of the state. The tension between service to state and
service to self, family, or some other limited group was nowhere in
sharper focus than on this low level of the bureaucracy. Many men on
this level could never be held to high standards of behavior by the lure of
high office. Their chances of advancement were slight, especially if they
entered the civil service through one of the less preferred methods.
Therefore, if they were to be kept from abusing their positions to enrich
themselves and their families, they had to be controlled by superiors and
adequately remunerated. But what was desirable might not be feasible.
Given the limited fiscal base of the traditional Chinese state and other
competing demands on its resources, its ability to remunerate its func-
tionaries was also limited. Therefore selecting men who, by their charac-
ters, would be inclined to behave properly was all the more important.
To select a few such men would have been no problem. With an empire
the size of the Sung, in which low-level bureaucrats numbered in the
tens of thousands, it was a formidable task. Recent research suggests
that the total bureaucracy in the Sung may have numbered about
forty thousand men, of whom the vast majority were in the lower-level
offices.1 Tens of thousands of officials had to be selected. If the Court did

1 See Li Hung-ch'i, "Sung-tai kuan-yiian shu te t'ung-chi" (Statistics on the number of

Sung dynasty officials), Shih-huoyueh-k'an 14 (September 1984): 227-39.

251
 Law and order in Sung China
not want to delegate the power of selection to people away from the
capital, if it wanted to maintain central control over appointments, it
had to devise bureaucratic mechanisms for selecting skilled and motiv-
ated men for specific posts.

Sung system of offices

The desiderata were easy to identify. Sung statesmen recognized the
critical role of personnel selection in maintaining order. The rise of
banditry was repeatedly blamed on the use of incompetent officials.2
Although the authorities recognized that circumstances beyond the con-
trol of local officials, such as widespread natural disasters, might contri-
bute to disorder, they also felt that such underlying causes usually
became truly threatening only when their effects had been exacerbated
by the actions of ineffective or venal officials. The answer, to choose good
officials, was simple in theory. However, in practice, getting the right
officials to the right place at the right time was not always easy. Were
the officials chosen equipped by background and training to perform
their functions well? Did they have the requisite combination of
bureaucratic skills, commitment to the proper values, and martial abil-
ities? The history of the position of sheriff under the Sung reveals the
difficulty that the government had in finding men who combined these
somewhat disparate talents, so that people were usually selected on the
basis of only one criterion or by the grim reality of the available pool of
candidates.
The Sung civil service, in which the sheriff played only a minor role,
was heir to a bureaucratic tradition already more than a millennium old.
The Chinese had early learned the enormous power of systematically
organized behavior. This was reflected in a carefully structured formal
political apparatus, composed of three segments: a civil government
branch divided in the Sung into nine grades, a military service, and a
clerical service that was despised and feared by many members of the
regular civil service, feared perhaps because of its functional indispens-
ability and its huge size and despised because its members were viewed
as venal and unethical.
The founders of the Sung had inherited from their predecessors a
system of offices that had been functional in the T'ang but that over time
had become titular. These offices were divided into nine grades (p'in)
and each grade was subdivided into from two to four subgrades (chieh*).
In the early Sung (before 1080) these titular offices were used primarily

2 See, for example, Wang Shih-p'eng, Mei-ch'i chi, 2.10b- n b .

252
 Personnel selection

to determine salary and formal rank. They were periodically listed in

rank order on protocol registers that provided in themselves a much
more exact indication of formal standing. These titular offices were
divided into two major groups, executory officials (hsiian-jen) and admin-
istrative officials (literally, "capital and court officials"; ching-ch'ao kuan).
All the higher offices were staffed by the court officials. The capital
officials included the middle group. The executory officials included local
administrators, from the prefectural staff supervisors down to the district
sheriffs, the recorders, the prefectural inspectors of education, and the
teaching assistants who served in the state-supported local school sys-
tem. Of the whole hierarchy of Sung titular civil officials (the protocol
list of 1038 names nearly two hundred), the sheriffs outranked only the
lowly educational inspectors and the teaching assistants. The rank of
their military counterparts, the patroling inspectors, also was low. The
sheriffs, being near the bottom of the hierarchy, were supposed to be
drawn from among executory class officials, the men who formed the
lower echelons of the civil service. 3 This group continued to be a major if
not the major source of candidates for sheriff throughout the dynasty.

Profile of the Sung civil service

During the Sung, the civil branch of the government was especially
powerful. This bureaucracy of nine grades should be visualized as a
sharply tapering pyramid, with most of the officials being in the lowest
grades, lower eight and nine. Although the hierarchy of officials came to
be based on the titular office held by an official, rather than the grade he
held, for general purposes we can think of this grouping of officials in
grades 8b and 9 as comprising the executory officials, with officials
ranked at grade 8a and higher comprising a much smaller group of men
occupying the policymaking levels of the government. The sharpest
divide within the bureaucracy thus separated the capital and court
officials - the administrative officials who made public policies - from
the executory officials who carried out the policies. Almost all the men
concerned with law enforcement came from the executory group.
Men in the lowest grades far outnumbered their superiors, but this did
not mean that the authorities found it easy to fill these low-level posi-
tions. First, because there were simply more positions on the bottom level
than in the higher levels, large numbers of men were needed. Second,
men nominated for positions could refuse to accept them, and often did
so if the positions were unpromising. Men competed avidly for entrance

3 SHY, chih-kuan 48.65b-66a; HCP 385.3a.

253
 Law and order in Sung China

to the civil service, because this carried with it great prestige and the
possibilities of power, but might decline to serve in posts in poor loca-
tions or with little chance of advancement.

Selection methods
The Sung government used a variety of methods to select officials. Many
of the methods used had been inherited from earlier dynasties, though
some were carried further and given new importance under the Sung.
The most famous selection device, and in many ways the most influen-
tial, was the traditional examination system. These examinations tested
men on their knowledge of the Confucian Classics and on their ability to
compose certain literary forms. The use of the examinations was based
on the assumption that a knowledge of the Classics could measure the
candidate's acquisition of appropriate values and of political acumen
and that an ability to compose literature could indicate intelligence.
Given the determination of top government leaders to reduce the
power of the military and to emphasize civil values, and the absence of
the aristocratic elite that had dominated T'ang politics, the examination
system appealed to the emperor and his advisers. The Sung brought into
the government an unprecedented number of examination graduates.
The degree holders, although they were slightly outnumbered in the
civil service as a whole by men who entered the civil service in other
ways, dominated the policy-determining levels of Sung political life
to an astonishing degree. An example is the position of prefect of Hang
Prefecture, an important prefecture on China's southeast coast. Exclud-
ing those men appointed early in the dynasty, most of whom entered the
civil service either during the Five Dynasties or in the opening years of
the Sung, of the men on whom we have information, some 92 percent of
the Northern Sung prefects were examination graduates. All seem to
have been capital or court officials.4
The central authorities could assume that the graduates subscribed to
the proper values and were qualified to perform the sheriff's bureau-
cratic duties, though they were open to criticism for not having military
skills.5 A large number of sheriffs were selected because they had grad-
uated from this civil service examination system. Some of these men
were ordinary graduates. From at least the 980s, it seems to have been
customary to nominate the lowest grades of graduates (both the doctor-
ate in letters and other fields) to serve as recorders or sheriffs.6
4 McKnight, "Administrators of Hangchow."
5 HCP I 4 i . i 4 b - i 5 a .
6 HCP 21.3b (980), 24-ga-b (983), 29.8a (988), 46.13b (1000), 51.12b-13a (1002), 59.13b

254
 Personnel selection

During the Sung the regular examinations, given triennially, grad-

uated an average of 140 men per year.7 Examination graduates were
grouped in classes depending on their graduation level. Men from the
lowest levels of graduates were the more likely candidates for the post of
sheriff, with those in the higher brackets being offered more desirable
posts. There were in the Northern Sung about twelve hundred districts,
which suggests about an equal number of positions for district sheriffs.
In the Southern Sung there was probably almost an equal number of
sheriffs' posts, since there were then fewer districts, but many had two
and some more than two sheriffs. Given the number of potential posi-
tions and estimating that on average a sheriff did not serve more than one
term, the state would have needed approximately 300 men per year to
fill the vacant positions. Graduates of the regular examinations could
have provided at best only a fraction of the men needed.
Examination graduates probably continued to be an important but
numerically small source of sheriff candidates during the Southern Sung.
Although there is little information on the use of examination graduates
as sheriffs during this period, Tu Cheng (1166-after 1218) did submit
a memorial recommending the use of new doctors-in-letters for long-
vacant sheriffs' posts. 8 Occasionally men were also appointed as sheriffs
after special (as opposed to regular) examinations.9
Between 1140 and 1206 the sheriffs were also drawn from men who
received their official ranks in the facilitated examinations. These facil-
itated examinations were special, relatively simple examinations given to
men who had repeatedly (and unsuccessfully) taken the regular civil
service examinations. When such men reached a specified age and had
taken the regular examinations a specified number of times, they might
be permitted to take the easier facilitated tests. Ostensibly the Court
wanted to reward these men for their persistence in studying the canon-
ical texts; the system also served to defuse the resentments of those who
had studied long without success. The authorities recognized that such
men might be of marginal competence, but because they were elderly,
they would not long be a burden on the state. Graduates might be
presumed to be thoroughly familiar with the Confucian Classics. Un-

(1005), 68.14b (1008), 71.25a (1009), 114.11b (1034), 204.14b (1065), 403.2b (1087).
7 The figures on which this very rough estimate are based are from Thomas H. C. Lee,
Government Education and Examinations in Sung China (Houg Kong: Chinese University
Press, 1985), pp. 279ft0.My only intent here is to provide a very approximate picture of
the relationship between examination graduates and potential candidates for the post of
sheriff. The picture can be only sketchy, because the size of the bureaucracy changed
over time and the figures for examination graduates are defective in various ways.
8 Tu Cheng, Hsing-chan Vang-kao, SKCSCP ed., 6.naff.
9 See HCP 276.9a.

255
 Law and order in Sung China
fortunately, at least in the opinion of many regular officials, they also
were not too bright.
The use of such men as sheriffs increased during the Southern Sung,
which may reflect the growing importance of this type of examination
during the Southern Sung, when the regular examinations became more
highly competitive. 10 All the reports we have about this practice are
complaints about the problems caused by using such incompetent per-
sons or directives that they not be appointed if over sixty, but the
repetition of such orders probably means that in practice they con-
tinued to be used.11 Thus political reality led the Sung authorities to
draw on a group whose qualifications for the position of sheriff were
suspect at best.
Despite the importance of examinations as a route to official rank, the
emperors did not surrender the power to circumvent this relatively
meritocratic system. Official status was frequently granted as a matter
of imperial grace. This served to create a body of men whose status
depended entirely on the goodwill of the throne. Unlike their colleagues
who had graduated from the examinations, such men derived their
power from, and were indebted to, their immediate sponsors and the
emperor. The continued use of large numbers of such men in active
government service was simply one more example of the Sung rulers'
determination not to allow any group to have exclusive or nearly exclu-
sive control over any vital political function.
At times, appointments to the post of sheriff through the benevolent
dispensation of the emperor came in response to the meritorious actions
of the men so honored. Occasionally the appointees were men who had
submitted proposals (usually dealing with suppressing disorders) that
had received the attention of the emperor, and were given the position as
a reward. Such attention might catapult commoners into office,12 or it
might merely lead to a job for an unemployed examination graduate. 13
Such men had at least demonstrated a theoretical interest in and
understanding of the problems of law enforcement. Others, who had no
claim to such qualifications, might be chosen because it was socially and
politically expedient. Thus grace was also extended to certain imperial

10 See Chaffee, The Thorny Gate of Learning in Sung China.

11 SHY, chih-kuan 48.71b (1140), 48.79a (1180), 48.81a (1185), 48.81 a-b (1190), 48.83a
(1206); HCP 120.8b; SS 35.672 (1181).
12 See the case of a commoner from Chen Prefecture who was ordered appointed as a
recorder or a sheriff because he had submitted an ingenious plan for taking over some
troublesome southern barbarian areas (HCP 236.3a—b).
13 In 1080 a doctor of letters named Liu Tang was appointed sheriff of Hsiao District in
Hsu Prefecture because he had sent up a memorial containing ten suggested regulations
for controlling bandits (HCP 304.15a). See also a case dated 1075 (HCP 2506.21b).

256
 Personnel selection
relatives. Such people were automatically accorded official rank, follow-
ing specific rules. 14 At times they were actually awarded substantive
offices, including that of sheriff,15 at least during the reign of the Emperor
Shen-tsung (r. 1065-78), who approved a proposal that the most distant
group of imperial mourning relatives (those with the three months of
mourning obligation) be used as sheriffs.16
Men were also appointed through the "protection" (yin) system,
under which high officials could gain official status for relatives. This
was a very important tool for controlling the bureaucracy. In the words
of Winston Lo:

The prospect of being able to confer the yin privilege on one's heirs was one of
the most powerful incentives that the Chinese civil service had devised. From
the Han period on, it had become a standard part of the personnel practice
of successive dynasties. If we may characterize the relationship between the
emperor and his ministers as a contractual relationship, thejym privilege must be
considered an important part of this contract. It was an integral part of the
remuneration for which ambitious men were prepared to make the utmost
sacrifice to enter to emperor's service and gain his good graces.17

Before 1002, high officials exercising their yin privilege could ask that
appointments (including appointments as sheriff and recorders) be
awarded to their protegees and in-laws. In 1002 it was decreed that
officials could seek appointments through yin only for relatives with the
same surname. 18 Despite such limitations, however, the yin system was
the principal device, in terms of numbers, for selecting government
functionaries. Such men outnumbered those who came in by other
routes, though they did not outweigh them in terms of political power.
According to Winston Lo, people qualified only through yin may have
suffered discrimination in competing for posts calling for literary abili-
ties. And in the Northern Sung at least, they seem to have been under-
represented in the higher levels of the civil government. The situation
changed, however, during the Southern Sung. Figures for 1213 indicate
that at that time 18 percent of those who had entered the civil service
through the examinations had advanced to the elite administrative class
of the bureaucracy, whereas 82 percent were in the lower executive class.

14 SS 159.4b-5a. See also Yang Chung-liang, Hsu tzu-chih t'ung-chien ch'ang-pien chi-shih
pen-mo (Taipei: Wen-hai, 1967), 35.4a. The Sung system was very like that of the T'ang.
See Robert des Rotours, Traite des examens (Paris: E. Leroux, 1932). pp. 223-25.
15 HCP 239.2b.
16 HCP 239.2b.
17 Lo, An Introduction, p. 103. Much of the following discussion of the civil service position
of the sheriff is based on information from Professor Lo's book.
18 HCP 53.16b.
257
 Law and order in Sung China

Thirteen percent of those who had entered the service through yin were
members of the elite administrative class, and 87 percent were in the
executive class. These figures suggest that entering by way of the exam-
inations was not so important during the latter period. 19 Given the large
number of men who entered through the yin privilege, the sharply taper-
ing pyramid of power in the Sung bureaucracy, and the predominance of
examination graduates in higher offices, at least in the Northern Sung
(and perhaps to some extent in the Southern Sung), the majority of yin
entrants were likely to spend a good part of their careers in low offices
such as that of sheriff. Such people might be intelligent and fully con-
versant with the Confucian Classics and the values they enshrined. But
because they had not gone through the examination process, the degree
to which they were conversant with the cultural core of the Chinese
tradition was problematic, though it might be supposed that such can-
didates, who came from elite families, shared the ruling class's values
and so had gained some feeling for politics and bureaucratic processes as
a result of their upbringing.
Various sorts of officials with irregular credentials were also appointed
as sheriffs. Su Ch'e, writing in 1086, reported that in Kuang-nan, irreg-
ular officials (she-kuan) with two formal recommendations from the
circuit authorities could be used as recorders or sheriffs and could
become regular officials if they had no faults during two terms of office.20
Irregular officials of this sort were peculiar to the far south, where the
normal practices of the rest of China sometimes did not apply. Because
of the scarcity of qualified officials from other regions willing to serve in
this region, the central authorities made an exception to the general rule
and allowed locally prominent men to move directly into government
positions through the device of "irregular office."
During the Northern Sung, locally resident functionaries who were
"outside the career" (liu-wai), that is, not in the regular civil service,
were also used as sheriffs. During the T'ang, such men had occupied
positions of considerable responsibility, but by the Sung the position of
this group had declined. Most such men seem to have served in sub-
ordinate staff positions in government offices, doing routine work and

19 See Lo, An Introduction, p. 85. The figures are from Li Hsin-ch'uan, Chien-yen i-lai chao-yeh
tsa-chi (Taipei: Wen-hai, 1966), vol. 2, I4.i5a-b. There is one caveat to the implication
of these figures: In the Northern Sung a significant number of those who initially gained
entrance to the civil service through the yin privilege later succeeded in the
examinations. I have no data on how frequently this happened in the Southern Sung,
but presumably it did on occasion. Thus the figures given by Li Hsin-ch'uan, which
pertain to the initial method of entrance, may not accurately reflect the impact of the
examinations on career success.
20 HCP 368.9a.

258
 Personnel selection

having little opportunity to exercise independent judgment. They were,

however, occasionally appointed to responsible local posts. In response
to a complaint in 109721 the central authorities in 1098 stopped such
functionaries from being appointed sheriffs in districts in which they
would have to serve concurrently as recorders. 22
Sheriffs were also sometimes chosen from men with clerical back-
grounds. Clerks were the workhorses of the traditional bureaucracy. A
typical Sung district, with from one to four civil service officials, might
range in population from fewer than a thousand households to several
tens of thousands of households. The civil government was headquartered
in the district town, where the civil service officials spent most of their
time. It is inconceivable that they could have exercised any degree of
control at all over the affairs in their jurisdiction without the support of
non-civil service subordinates who functioned as gatekeepers, wardens,
secretaries, and clerical functionaries. Such persons certainly out-
numbered their civil service superiors by as many as a hundred to one.
The civil service functionaries, whatever their actual method of entry
into service, were presumed to have been socialized into the higher ethics
of governmental work. The non-civil service clerical functionaries,
however, whatever the ethical level of their actual behavior, were
systematically stigmatized as venal and inferior. In the documents of the
period, written almost exclusively by civil service officials, clerks are
stereotyped as dissolute and corrupt. At the beginning of the Sung, none
of them seems to have been regularly salaried, and even after the reforms
of the late eleventh century, most were not paid a living wage. They
therefore had to gain their livelihood by squeezing payments from those
who needed their services and support. The rabid denunciations of
clerical venality that are sprinkled through the writings of Sung officials
can be traced in part to actual abuses and excesses by clerks. It is
equally certain that they reflected the fear and frustration of officials who
recognized the passive power of the clerks, without whose efforts the
government's daily operations would have been impossible.
The civil service group not only denounced the improprieties of clerks.
They also sought to place strict limits on clerks' upward mobility.
Although from early in the Sung, a tiny handful of clerks were allowed
to transfer into the regular civil service, this was possible only after
extremely long and meritorious service. The use of such transferred
clerks as sheriffs was forbidden by an edict of 1132,23 but a report dated

21 HCP 491.7a. See also SHY, chih-kuan 48.62b (1024).

22 HCP 494.6b; SHY chih-kuan 48.66a. The order was repeated in 1100. For the
background of this order, see HCP 491.7a (1094).
23 Liu Cheng, Huang Sung chung-hsing, 11.4b.

259
 Law and order in Sung China
1206 states that when selecting sheriffs the authorities "frequently
nominate men who have qualified through the facilitated examinations
or who are ex-clerks." 24
Other irregular appointees were drawn from among men with official
rank through "contribution," that is, men who had been granted rank
because in times of great need such as famines they had made gifts to the
state. 25 Men were also drawn from the relatives of sheriffs killed in
the line of duty. 26 And at the beginning of the dynasty, sheriffs were
appointed from ex-officials of the southern states conquered by the
Sung. 27 All these irregular groups were less than ideal, being deficient in
either experience or capacity or having dubious values and loyalties.
Finally some men were appointed from those who had served previously
in local offices - from district magistrates who were being demoted or
from sheriffs and recorders being moved to new districts. These men
were presumably capable, albeit not outstanding.

Civil versus military credentials

These various sorts of candidates, with their predominantly civil back-
grounds, might at best have had some qualifications for dealing with the
bureaucratic side of a sheriff's duties. Ideally they would be gentle-
men. The actual law-enforcement facets of the position, which called
for military skills, required a different set of talents. For the military
the state needed heroes, men with the training and inclination to use
violence. Although men with the various civil backgrounds just des-
cribed probably supplied the majority of sheriffs, from the 1070s to the
end of the dynasty many sheriffs had military credentials. These men
might be graduates of the military examinations, 28 men who apparently
gained merit in battle without being members of the military officer
corps, 29 or most frequently servitors minor (hsiao shih-ch'en) or servitors
major (ta-shih ch'en) whose dossiers described them as militarily talented. 30
One quite large group of men classified as servitors were not in fact
military men but, rather, those who had won their military officer

24 SHY, chih-kuan 48.83a. See also HNYL 904.

25 See an approved memorial dated 1219 in which the memorialist noted two men so
chosen and asked that in future such men be used only as tax officials (SHY, chih-kuan
48.86a-b).
26 HCP 122.1 a, 130.5a, 2400.7b, 2400.24a, 140.6a, 1412.14b, 152.11a, 172.8b.
27 HCP 22306.4a (971).
28 SHY, chih-kuan 48.760-773, 48.81a, 48.83a, 48.85b, 48.87a-b; Liu Cheng, Huang-Sung
chung-hsing, 47.15a.
29 SHY, chih-kuan 48.71b, 48.69b.
30 SHY, chih-kuan 48 et passim; HCP 290.5a.

260
 Personnel selection

classification through the protection {yin) of relatives who were mem-

bers of the imperial clan. Such appointments, which we might see as
nepotism, were in the Sung context an expression of the emperor's
familial concern and a device for counterbalancing the power of the
examination graduates. There is no reason to assume that these men had
military skills or interests. The second largest group of servitors, after the
yin beneficiaries, was the men who had transferred from regular military
officer positions.31 Moreover, the sources suggest that when men were
chosen from among the servitors for law-enforcement posts, they were
selected mostly from those with military backgrounds. The sources often
specify that "men with military talents" were wanted for such posts.
Those with other backgrounds may have clustered in nonmilitary jobs
like those of the service agents who oversaw tax collection depots. 32
The split between civil and military sheriffs forms the heart of a long
continuing dispute within the bureaucracy over the qualifications needed
for appointment as a sheriff. Although all parties recognized that a
sheriff had to be physically vigorous and possess military prowess, there
were sharp differences over the importance of military versus civil skills.
In part this dispute reflected the various problems in the empire. In
areas where serious violent crime was a major problem, there was a
tendency to favor military men as sheriffs. In regions where such major
violence was less of a problem, Sung officials tended to stress the dangers
of using military officers, whose commitment to appropriate values was
questionable and who were moreover serving in a post that required
many bureaucratic skills.33 Lack of bureaucratic skills was a very serious
problem. A reading of the materials suggests that many of the servitors
minor were barely if at all literate and thus would be unfit for the
nonmilitary aspects of the sheriff's role.34
This dispute did not arise until the latter part of the Northern Sung.
Because the sheriffs had originally been established as one means of
reducing the powers of the military, there does not seem to have been a
question of appointing military men to these offices before the reign of
Shen-tsung (r. 1067-85). Then, perhaps as one part of Wang An-shih's
31 Winston Lo, "A New Perspective on the Sung Civil Service," Journal of Asian History 17
(1983): 121-35.
32 SHY, chih-kuan 48.71b, 73a-b, 74b, 129a; Yuan Shuo-yu, Tung-Vang chi, SKGSCP,
8.20a.
33 At times the bureaucratic treatment of these two types of sheriffs differed. In 1176 it
was ordered that military sheriffs with five years of tenure be granted advancement
under the rules governing acting police commissioners (SHY, chih-kuan 48.77b-78a).
34 SHY, chih-kuan 48.68-69a (1127), 48.69a (1127), 48.69a-b (1132), 48.71b-72a
(1141), 48.72b (1145), 48.74b (1163), 48.74b (1163), 48.75a (1168), 48.75b (1170),
48.76b (1173), 48.79b (1181), 48.80b (1184), 48.81b (1195), 48.81b-82a (1201), 48.82b
(1202), 48.83a (1204), 48.85b-86a (1215).

261
 Law and order in Sung China

general reorganization of some areas of the government, the state began

to appoint military men. In 1077 there were requests that military
officers be sent as sheriffs to certain northern border districts and to
districts in Ching-tung Circuit that had suffered natural disasters, an
area that corresponded roughly to parts of present-day Shantung and
Ho-pei. 35 Two years later we find a list of eighteen districts along the
northern and western borders for which the Bureau of Lesser Military
Appointments was told to select servitors minor as sheriffs.36 This policy
seems to have been due in part to a conviction that only military officials
could do a competent job of giving military training to the men of the
newly established pao-chia militia system. 37
In 1082 this practice of appointing military men as sheriffs was
extended to all heavy-penalty areas (chung-fa-ti), places where certain
types of criminals were subject to especially severe punishments. 38
Within a few years this policy was bitterly attacked by Su Ch'e, who
was particularly worried about the insufficient ethical socialization of
military men:
According to the old system, sheriffs were always selected from among civil
executory officials [hsiian-jen]. Recently, because the people have become impov-
erished, bandits have become numerous. Those who deliberated on affairs did
not understand how to solve the problem. Therefore they asked that in heavy-
penalty places, military men be used as sheriffs. I have not heard that after this
change the number of bandits decreased as a result. Moreover, the military men
used are corrupt and harsh. They do not fear the laws and act oppressively. The
archers do not have their hearts in law enforcement and bedevil the villages,
troubling the people. I feel that in the art of capturing bandits, the most
important step is first to gain the archers' commitment. Next in importance is to
secure the aid of the villagers, for they spread the net. If beforehand the method
is understood, it will be easy to complete the job. Under the old system in which
civil candidates were appointed, even if they did not completely follow this, still
they respected the laws and had good relations with the people. It was not
necessary that they themselves be able to shoot and ride, but still they could
capture the criminals.
The reason for Su's objections is noteworthy. The military are espe-
cially criticized for behavior that reflects their incorrect values, whereas
the civilian candidates are praised for the propriety of their behavior in
office. The contrast is between the gentleman, albeit not a perfect gentle-
35 HCP 280.9b, 282.2a.
36 HCP 290.5a; SHY, chih-kuan 48.64b.
37 SHY, chih-kuan 48.67a. See Wu Ching, Chu-chou chi, n.8b-ioa, for a spirited
argument that the position of sheriff was basically military by both nature and
tradition.
38 HCP 328.2a; SHY, chih-kuan 48 65a.

262
 Personnel selection

man, and the overbearing hero. Clearly the government had failed to
devise institutional restraints that could control the police officials,
or policies that would lead to effective indoctrination of its military
personnel.
Su Ch'e's request that civil officials again be used was accepted in
part, and it was ruled that military officials be used only in border
regions. 39 Such men continued to be used in border regions during the
remainder of the Northern Sung, 40 although civil sheriffs were also
sometimes appointed to these areas. 41
This reemphasis on civil officials was reversed during the period
of troubles that began with the rebellion of Fang La in 1120-21
and culminated in the Jurchen invasions of the middle and late 1120s.
Military men were sent to previously civil posts, 42 and for a brief period
(1131—34) there seems to have been a general policy to increase the
number of military sheriffs.43 Needing military sheriffs for their law-
enforcement talents, the authorities still sought some way to counter-
balance their bad qualities. A memorial by the official Wang Yang
(doctorate of letters degree 1123) even suggests that there may have been
a deliberate policy to create a dual system, with "a civil sheriff to guard
the laws, and a military sheriff to press the attack." 44 Whether or not
there was a policy to create a dual system, there was without doubt a
move to increase the number of sheriffs, so that districts often had two,
and occasionally more than two, sheriffs. Indeed, the Southern Sung
official Yiian Yiieh-yu suggests that the accepted policy was two sheriffs
per district. 45
As order was restored in the 1130s, the emphasis on military sheriffs
declined. There was even a decree that they be replaced by civilian
sheriffs,46 but during most of the Southern Sung, the state employed a
mixture of civil and military men as sheriffs. The prefecture described by
Gh'en Yuan-chin (doctorate of letters degree 1211) was probably typical,
having some districts with only civil sheriffs and some with one civil
sheriff and one military sheriff.47 This, like the increasing number of
patroling inspectors, is another facet of the militarization of law enforce-
ment during the Sung.
39 HCP 385.3a; SHY, chih-kuan 48.65b-66a.
40 SHY, chih-kuan 48.66a, 67b.
41 SHY, chih-kuan 48.67a ( n 15), 48.67a (1116), 48.67b ( n 17).
42 SHY, chih-kuan 48.68a and passim.
43 SS 26.19b, 27.17b.
44 Wang Yang, Tung-mou chi, 9.19b-20a.
45 Yiian Yiieh-yu, Tung-t'ang chi, 8.23a.
46 SHY, chih-kuan 48.70b, 70b—71a; Liu Cheng, Huang Sung chung-hsing, 14.18a, 15.13a;
HNYL 1257.
47 Ch'en Yuan-chin, Yu-shu lei-kao, SKCSCP i.2b~3b.

263
 Law and order in Sung China

In general, military men - sometimes limited to those younger than

fifty years of age - were frequently used along the borders and the sea-
coast or in other regions where there were serious problems of lawless-
ness. 48 After 1159, graduates of the military examinations were apparently
appointed as sheriffs in some border posts, 49 but civil officials were
generally preferred in other areas. There continued to be the same sorts
of complaints against military sheriffs. Ch'en Yuan-chin, reporting on
some places in Chiang-nan West Circuit that had military sheriffs, said
that such men were out of touch with the people, were wholly at the
mercy of the clerks, were in some cases completely illiterate, were toadies
to the local powers, and were abusive to their subordinates. 50
This dispute, never completely resolved, led in the Southern Sung to a
complicated system under which the districts seem to have been placed
in certain categories that determined the order of preference for different
types of candidates. In a given district, preference might first be given to
civil officials of the executory class. If none was available, the next choice
might be servitors major, and if servitors major were not available, the
choice would be servitors minor. 51 Within these categories of men, pre-
ference would be given to those characterized in certain ways in their
official dossiers.
There are many limited and partial reports on selection that provide
tantalizing glimpses of the criteria and procedures used. Fortunately, a
more complete report dated 1174, in combination with more scattered
evidence, gives a more coherent picture of selection in operation. The
initial nomination from among men on lists of candidates was to be
made from men in certain specified civil service categories, by specific
officials, within a set period of time (five days for initial nominations at
the capital). If no men were found within the most preferred category,
then the selection would pass on to the next most preferred group, and
so on. The candidates generally had to be younger than sixty years of
age. This seems to have been true from early in the Sung and remained
so at least into the thirteenth century. 52 In certain cases, in which
military talents were particularly needed, the age limit was fifty years. 53

48 SHY, chih-kuan 48.74b (1163). See also Yuan Yiieh-yu, Tung-t'ang chi, 8.23a; Tu Cheng
(1166-after 1218), Hsing-shan t'ang-kao, 6.i6a-b.
49 HNYL 181.3005.
50 Ch'en Yuan-chin, Yii-shu lei-kao, i.2b~3b.
51 SHY, chih-kuan 4 8 ^ - 7 7 0 .
52 HCP 102.8b (1028), i 3 3 . i a - b (1041), 12899.14b (1042); SHY, chih-kuan 48.63a
(1041), 48.63a (1042), 48.64b (1103), 48.74a (1164), 48.79a (1180), 48.81b (1192),
48.87a (1219). But see the anecdote in Wang P'i-chih, Sheng-shui yen-Van lu, Chih-pu tsu
chai ts'ung-shu ed., 9.3a, about a seventy-five-year-old sheriff.
53 SHY, chih-kuan 48.73a-b (1156), 48.74b (1163), 48.82b (1202).

264
 Personnel selection

They could not be found (on the basis of a physical examinations) to be

ill, disabled, or otherwise incapacitated. They had to have records free
from certain types of offenses. Men might also be excluded from service
if their dossiers recorded convictions for public misdemeanors (kung-lsui),
infractions committed in the line of administrative duty and not for any
personal reason, or private infractions (ssu-tsui) in which the violation
implied a desire for personal advantage. 54 Literacy became a problem
during the late Northern Sung when military officials, many of whom
were barely literate, were more and more frequently nominated; so in the
Southern Sung tests of the candidate's ability to read and write were
introduced. 55 In 1178 it was decreed that candidates for the position of
sheriff or supervisor of a market town had to pass a test on reading the
laws in order to demonstrate their literacy. 56 There were also problems
in the Southern Sung with using supernumerary officials, men who
held official credentials at the lowest levels of the hierarchy but who
ordinarily were not active members of the civil service. They were said to
be ill qualified.57 Finally, it seems that preference was given to men who
had already entered on an official career (ch'u-shen), who were incumbent
in office,58 who had had experience in police or military matters, 59 and
who had been sponsored by higher officials.60
Selection ordinarily took place at the capital, where nominations
of civilians, and sometimes of military officials also, were handled by
the Bureau of Civil Nominations (Liu-nei ch'iian) within the Ministry
of Personnel (Li-pu). At the beginning of the dynasty when military
officials were not used, all central nominations were handled through the
Ministry of Personnel; later in the dynasty the Bureau of Military Affairs
entered the process. When the initially preferred candidates were civil
officials but no acceptable civil candidates were available, the choice
moved on to military candidates. The Ministry of Personnel still had

54 SHY, chih-kuan 48.62b (1027), 48.63a (1042), 48.63b (1068), 48.69b (1133), 48.83a
(1204); HCP 12899.14b (1042). There are scattered instances in which men were
rehabilitated and appointed; HCP 73.4a (1010), 80.2a (1013).
55 SHY, chih-kuan 48.76b (1174), 48.79a (1178), 48.880-893 (1178), 48.850-863 (1215).
56 SHY, chih-kuan 88b-89a.
57 Yuan Ytieh-yu, Tung-Vang chi, 8.23a.
58 SHY, chih-kuan 48.63b (1068), 48.71b (1140). However, the central authorities also
chose men who were free and awaiting appointment; SHY, chih-kuan 48.82a-b (1201).
59 SHY, chih-kuan 48.71b (1140), 48.83a (1204).
60 In 1068 men were ordered chosen (for posts in the Kaifeng area) from among those
who had already begun a career, who had already served one term in office and had
thus collected three annual merit ratings without committing any public or private
infractions punishable by a sentence of penal servitude or worse, and who had already
had three sponsors (SHY, chih-kuan 48.63b).

265
 Law and order in Sung China
61
charge of the selection, but the Bureau of Military Affairs was some-
times asked to evaluate the nominations. 62 At other times, when mili-
tary candidates were involved, the selection itself was apparently the
responsibility of the Bureau of Military Affairs or the Bureau of Lesser
Military Assignments (San-pan yuan). 63
In some cases the selection seems to have been made on the local level
and was then forwarded to the capital for confirmation. Local appoint-
ment of local officials had a long history in China, but its role had been
greatly diminished under the Sui (589-618) and T'ang (618-907) gov-
ernments. The Sung, perhaps because the post of sheriff was at the same
time lowly and important, sometimes permitted both permanent and
temporary local appointments. At times during the Northern Sung there
appears to have been a standard policy that local persons might be
appointed by the intendants' offices,64 and a case from n 34 indicates
that in the early Southern Sung, intendants were also able to make such
appointments under certain circumstances. 65
The local authorities' prerogatives were greatest in cases in which they
had been ordered to investigate incumbent sheriffs. If they judged these
men unfit, they often could replace them with local candidates of their
own choosing and have their choices reported to the Court for confirma-
tion. 66 One report dated 1133 implies that such substitutes were subject
to more stringent rules than were ordinary candidates, for it says
that men could not be so used if the were "guilty of any crimes." 67
Irresponsibility was discouraged, by making the sponsors liable for
punishment if their recommendations were unwarranted. 68 Under the
rules of 1174 the first nomination was to be from among men described
as "militarily talented and close to the people" (ts'ai-wu ch'in-min), 69
61 See SHY, chih-kuan 48.81b (1192), 48.73a-b (1156), 48.83a (1204),
(1169), 48.74b (1163).
62 SHY, chih-kuan 48.73a-b (1156).
63 In 1117 there was an order that in border areas where the police commissioners and
sheriffs were military officials, the Bureau of Military Affairs was to select those who
had already served in border posts (SHY, chih-kuan 48.67b). For the Bureau of
Military Nominations, see HCP 290.5a (1078), 432.8a-b (1089); SHY, chih-kuan
48.64b (1078), 48.79a (1178), 48.76b-77b (1169).
64 SHY, chih-kuan 48.64a; HCP 218.1a.
65 SHY, chih-kuan 48.70b—71a. See also SHY, chih-kuan 48.69a. Sometimes the local
authorities merely recommended local candidates; SHY, chih-kuan 48.69a (1127); HCP
I43.i9a-22a (1043).
66 HCP I43.i9a-22a (1043), 275.4b (1076), 365.1a (1086); SHY, chih-kuan 48.64b
(1076), 48.68b (1127), 48 69b (1134), 48.70a (1133), 48.69b (1134), 48.71a (1135),
48.74b (1163), 48.79a (1180). In 1127 the fiscal intendants in Liang-che Circuit were
told to depute nominees "within one month" (SHY, chih-kuan 48.68b).
67 SHY, chih-kuan 48.69b.
68 SHY, chih-kuan 48.69b (1134).
69 SHY, chih-kuan 48.76b. In a personal communication, Winston Lo suggested that

266
 Personnel selection

Service agents with a military bent were the second choice.70 The num-
ber of possible combinations of qualifications was obviously very large.
Because we lack a complete record, we will probably never be able to
reconstruct this part of the system even for one period, much less trace
its evolution through the Sung,71 but the general thrust of the policies
is clear, a balance of priorities and abilities, in descending order of
desirability.

Functional importance and bureaucratic unimportance

Although the post of sheriff was very low in the bureaucratic hierarchy,
it bore heavy responsibilities. The importance of maintaining local order,
combined with selection policies that filled many of these slots with
either bureaucratic beginners or the aged and incompetent, led to an
almost continuous litany of criticisms. Wu Ching (1125-83) is typical.
Complaining of the appointment of new civil officials, he observed:
Such men, if not doctoral graduates, are the sons and brothers of the local elite
[shih ta-fu]. We use such civil officials to handle military matters, placing in lowly
positions men who must handle weighty problems, appointing subordinate book-
worms without administrative experience, men who because of their back-
grounds are unfamiliar with the difficulties of rural life.72
The response of the central administration was a succession of orders
that local administrative officials examine and test incumbent sheriffs so
as to weed out those unfit for service, and the design of the detailed
policy on selecting candidates just described.
A system of physically testing sheriffs and patroling inspectors seems
to have been introduced early in the dynasty. It is not clear whether all
the men appointed to this office were so tested or only men in those
regions from which officials had submitted complaints about the physical
incompetence of police. Beginning in the 1040s a sequence of directives
set higher standards for the original selection of sheriffs73 and ordered

these two terms may have been technical terms, with ch'in-min probably indicating that
the person in question had served in general administrative posts.
70 SHY, chih-kuan 48.76b; see also SHY, chih-kuan 48.70a. Such men were at least
sometimes graduates of the military examinations; HCP 328.2a (1082); SHY, chih-kuan
48.81a (1185).
71 This system was planned down to its most minute details. For example, we learn from
one document that in the circuits of Huai-nan, for which either civil or military sheriffs
might be nominated, if two men - one civil and one military - were nominated on the
same day, the post was to go to the civilian (SHY, chih-kuan 48.76b-77b).
72 Wu Ching, Chu-chou chi, u . 8 a - i o a .
73 For example, that men to be appointed to these posts had no convictions for rapacity,
were under sixty years of age, and so on (see HCP 133.1a—b).

267
 Law and order in Sung China
superior officials to examine incumbent sheriffs and patroling inspectors
and replace the incapable. Initially these inspections were handled by
the circuit intendants. 74 During the Southern Sung, perhaps because the
bureaucratic level of the patroling inspectors had been lowered, the
investigations were frequently handled by prefectural officials75 and only
occasionally by the intendants. 76 Men judged incompetent as sheriffs
were not necessarily cashiered; frequently they were merely transferred
to positions requiring less physical prowess, such as recorder or service
agent (chien-tang).77

Motivation
Sung authorities might devise guidelines for choosing men competent to
be sheriffs (or patroling inspectors) and mandate tests to evaluate their
physical fitness. A more difficult problem was how to encourage the men
so chosen to accept the offices and perform the duties entailed without
abusing their powers. In particular, in its law-enforcement services the
Sung faced a perennial problem: how it could encourage its officers to
carry out their sometimes dangerous tasks actively and aggressively
while preventing them from exceeding their bounds.
The Sung authorities responded to these potential problems in a
manner typical of traditional Chinese bureaucratic empires, which we
outlined when describing the reward system. They created large num-
bers of regulations that specified appropriate procedures and behaviors.
Like all others aspects of Sung government, law enforcement was gov-
erned by a bewildering plethora of rules. No major set of such rules is
now extant, but scattered through the sources we do have examples of
the names of collections, such as the "Rules on Seizing Bandits" (Pu-tsei
t'iao) of 962. 78 This pattern is a characteristic governmental device of the
Chinese states. By specifying permitted actions, it allowed superiors to
monitor the actions of subordinates and permitted the easy and early
identification of dangerous tendencies. This forest of rules was charac-
teristically Chinese in another way. By establishing these regulations the
authorities created the appearance of a government of rules; by making
the rules so numerous they also almost guaranteed that ordinary officials

74 SHY, chih-kuan 48.63a (1041), 48.65b (1086); HCP 144.2b (1043), 271.14a (1075),
305.13a (1080), 335.15b (1083), 365.1a (1086).
75 SHY, chih-kuan 48.71a (1135), 48.81a (1185), 48.83a (1206), 48.135a (1211); HNYL
84.1383 (1135).
76 SHY, chih-kuan 48.73a (1149).
77 SHY, chih-kuan 48.70a (1133), 48.83a (1206), 48.135a (1211); Hu Su, Wen-kung chi, pp.
85-86.
78 HCP 3.14a.

268
 Personnel selection

would regularly be in violation, so that their professional welfare would

in fact depend on their personal relationships with those with the power
to sanction breaking the rules. A government of laws in theory became in
practice a government of personal connections.
One key element in the Sung strategy for securing effective police work
was a set of rules defining not only the proper and improper performance
of duty but also the rewards (and punishments) meted out to those who
did well or ill. This policy is simply a specific example of a general
traditional Chinese practice of using regulations and laws to control
subordinates by limiting their discretion and defining their tasks, and
associating these restrictions with rules for rewarding proper perform-
ance and penalizing failures. The original Sung rulings, established in
962 and 963, simply followed the "previous regulations," presumably
inherited from the T'ang and Five Dynasties.79

Rewards and punishments

Like informal and semiformal agents, formal police agents could benefit
from the system of rewards for capturing criminals or suffer penalties for
failing to do so. The system was by no means always fair in practice.
The Southern Sung official Wang Chih-wang (d. 1170) wrote to com-
plain about a recent order penalizing certain local officials when in fact
they had just succeeded in capturing major bandit gangs. 80
Like most aspects of Sung government, these rules were periodically
collected for more convenient reference by the authorities. These reward
rules, which were governed by a type of law in the Sung called ko, might
refer to specific areas or groups. In 1086 the Ministry of Finance com-
piled the "Reward Laws for the Capture of Bandits in Cheng and Hua
Prefectures,"81 and in that same year the Bureau of Military Affairs
petitioned successfully to have the "Rules on Rewards for Capturing
Bandits Applicable in Shensi and Ho-tung" recompiled. (The bureau
was dissatisfied with the existing compilation because it had been
created during the Reform period.) 82
The rewards offered to officials usually took the form of advancement
within the civil service but might also include gifts of money or goods;
the penalties might include loss of salary, demotions, expulsion from
office, or even criminal punishments. In a bureaucratic world, bureau-
cratic advancement was the most coveted prize. Unfortunately, there are

79 SHY, chih-kuan 48.60b.

80 Wang Chih-wang, Han-p'in chi, SKCSCP 1975, 7.24a-25a.
81 SHY, hsing-fa 1.13a; HCP 373.17b.
82 HCP 368.22b-23a; see also HCP 307.5a.

269
 Law and order in Sung China

still so many obscurities surrounding the actual working of the Sung

personnel system that at times it difficult to evaluate the meaning of the
rewards offered to ambitious sheriffs.

Promotion
Movement upward through the civil service hierarchy was affected by
tenure (usually measured in terms in office, jen), by annual merit ratings
(k'ao), by recommendations from higher officials, and by the rewards
and commendations earned by the official himself. The speed of the
promotion was also affected by the method through which the man had
originally entered the civil service and, in some cases, by special exam-
inations for advancement (including examinations in law).
Of all the factors influencing his bureaucratic advancement, the one
most directly under the control of the sheriff or the patroling inspector
was the accumulation of merit through the system of rewards and com-
mendations. Ideally, sheriffs would spend their terms without outbreaks
of banditry and would be rewarded. In the more usual case, sheriffs (and
in a rather similar way, the patroling inspectors) could earn rewards if
they killed or captured bandits (i.e., armed and violent robbers, not
thieves) within the time limits set by law. The rewards and punishments
were determined by the number of men captured, the percentage of the
gang captured, and the dangerousness of the criminals. Conversely,
those who failed to capture wanted men in the set time limits suffered a
graded set of sanctions. 83
Under the 962 rules there were three time limits of twenty days each
for capturing armed or murderous bandits. Apparently this limit was
later extended to thirty-three days. The general pattern of three equal
limits of approximately one month each remained in use during the rest
of the dynasty. 84 Under the system inaugurated in 962, a sheriff who
captured any of the wanted bandits within the first time limit was to be
"forgiven one selection" (chien i hsuan). This presumably would have
meant that unlike an official proceeding in the normal pattern, he would
not have to wait for such an extended time on furlough before becoming
eligible for another appointment. 85 If he captured some of the bandits

83 SHY, chih-kuan 59.1b. Because the Chinese government was thoroughly

bureaucratized, hapless officials might also face the opposite problem. Men of merit
were sometimes denied rewards or even penalized as a result of the nature of the reward
procedures (see P'eng Kuei-nien, Chih-t'ang chi, n.135-36).
84 HCP 3.i4a-b; SHY, chih-kuan 48.60a. This system of three time limits does not wholly
agree with the rules preserved in the Sung Code; SHT 28.1b. For the continuing use of
three limits, see SHY, chih-kuan 48.61a (986); HCP 97.11b (1021).
85 See Lo, An Introduction, p. 133.

270
 Personnel selection

within the second limit, he would be "moved forward one step in

seniority" (ch'ao i tzu). If he captured some within the third limit, he
would advance one subgrade in stipendiary office {chieff). If within these
limits he seized the majority of the gang, the aforementioned reward
would be doubled - he would be advanced two subgrades, forgiven two
selections, and so on. If the sheriff or district magistrate personally
fought with the bandits and captured them all, he was to be rewarded
with a gift of dark red ceremonial clothing, clothing ordinarily restricted
to officials of the sixth rank or higher. If the sheriff had done the fighting,
he was to be appointed a magistrate. A magistrate who had fought was
to be advanced in office. The meritorious officials were granted an
additional two steps in seniority. If they personally fought with and
captured half of the gang or the leader, they were to be "rewarded and
commended" (ch'ou-chuang),86

86 Unfortunately, some parts of these rulings on promotion and demotion are more easily
read than understood. During the T'ang dynasty, men newly entering the civil service
and men finishing terms in office were required to take special examinations for
placement in a new post. These examinations were called hsiian. See Rotours, Le Traite
des examens, pp. 213 ff. By Sung times this system appears to have disappeared
completely. Men were selected on the basis of the information included in their
dossiers, without having to be examined. Indeed, after 972, executory-class officials did
not even have to go to the capital to receive a new appointment (see Kracke, Civil
Service, pp. 86-87).
What would it have meant to a Sung official to be forgiven one (or two) "selections"
or moved forward one or two steps in seniority? The answer can be found in part in the
context of the Sung system of an advancement through a determined sequence of posts
on the basis of fixed qualifications. First according to the Sung dynasty history, under
ordinary circumstances, in order to be nominated as an executive inspector of a
prefecture (lu-shih ts'an-chun), a sheriff who had entered office through the examinations
had to accumulate two terms in office and four annual merits ratings; that is, he would
have had to serve two terms in office, each of several years' duration. Sheriffs who had
entered in some other way had to accumulate two terms in office and five annual merit
ratings, and sheriffs of irregular status (she-kuan) had to accumulate three terms of office
and seven annual merit ratings. Second, the importance of hsiian is connected with
what Winston Lo called the "lapse factor." After serving a term in office, most officials
in the executory level of the bureaucracy had to spend a number of years on extended
furlough before they would be eligible for another post. Being forgiven a hsiian would
spare them this furlough.
During the Sung these measures of service had to be accompanied by separate formal
letters of recommendation from superiors, for a candidate to be qualified for certain
kinds of promotions. If these men had recomendations from four sponsors, or from two
sponsors who were qualified to serve as commissioners, they might be nominated as
district magistrates.
The exact meaning of being moved forward one step in seniority remains somewhat
unclear, but we may assume that it must have been a change that increased a man's
chances for advancement, albeit less than "being forgiven a selection." Possibly we
might argue by analogy here and suggest that in these reports a "step in seniority" was
the equivalent of one annual merit rating. Being demerited one selection would mean
condemning an incumbent official to an extra term before he might become eligible for
advancement.

271
 Law and order in Sung China

According to Wang Yung, a decree of the 960s had stated that if a

sheriff were himself wounded when pursuing bandits but seized the
whole gang, he would be allowed to wear red clothing. If he caught two-
thirds of the gang, his waiting time for promotion would be cut by three
"selections" (hsiian). For one-third of the gang it would be cut by one
"selection," and he would be promoted one subgrade. If a magistrate
captured the whole gang, he would be promoted to capital office status
and given higher-grade clothing. If such men were killed in the line of
duty, their sons or younger brothers would be given appointments. 87
After 986 a sheriff's success (or lack of success) in capturing wanted
men could theoretically be determined by examining the calendrical log
(/zf), on which the sheriff was supposed to note the date on which the
criminals first became active, the times of their captures, and the judg-
ments rendered. 88 The system introduced in 971 was refined when it was
decreed that captures were to be entered on the official merit examina-
tion record. If a violent criminal were seized within the first time limit,
this could be used to expunge two failures to capture other criminals. If
the wanted man were seized within the second time limit, this would
clear the official's record of one failure. However, when a capture was
used in this fashion to expunge other failures, it could not also be used to
claim rewards under the system of meritorious achievement. 89
During the eleventh century, various refinements were added to this
system of rewards. Before 1017, if a magistrate or sheriff seized a com-
plete gang of ten or more, or suffered injury while seizing a complete
gang of fewer than ten, he would be rewarded and commended (ch'ou-
chuang). In 1017 the rewards were increased when it was ruled that a
magistrate or sheriff who personally led a group of archers in killing or
capturing ten or more bandits was to be rewarded and commended, even
if this was not the whole gang. They were to be equally rewarded and
commended for the capture of an entire gang of from seven to ten, or of a
particularly violent gang of fewer than seven. 90 Ten bandits seems to
have been a key figure in these calculations; when more than ten men
were involved, regular army troops apparently were dispatched to seize
them. 91
Other reports dating from the 1020s make clear that a sheriff who
personally captured whole gangs of ten or more wanted men - depend-
87 Wang Yung, Sung-ch'aoyen-i i-mou lu, i.3b-4a.
88 SHY, chih-kuan 48.61a; see also HCP 144.6b.
89 SHY. chih-kuan 48.6a. As was the case with most other Sung officials, sheriffs (from
962) were subject to annual merit ratings (k'ao) by their superiors; SHY, chih-kuan
48.60b; see also HCP 34.14b.
90 HCP 90.10a; SHY, chih-kuan 48.62a.
91 HCP 273.11a (1076). See also HCP 348.ioa-b (1084), 386.3b (1086).

272
 Personnel selection

ing on the sheriff's qualifications as recorded in his dossier - might hope

to be rewarded with a new appointment to a higher substantive post,
with honorary ceremonial clothing and an increase in titular status, 92
or merely with advancement in the lists of seniority. Seniority in the
personnel lists was a key factor in selecting candidates for most posts
(including the post of sheriff).93
According to the law of 1029, which remained in force thereafter, it
was generally necessary for a sheriff (and presumably also a patroling
inspector) to have been involved personally in the capture before he
could claim rewards. 94 (There is however one decree, dated 1036, which
says that men particularly adept at laying plans should be rewarded
even if they did not take part in the actual captures. If their plans led to
the capture of a complete gang of seven or more, an imcomplete gang of
ten or more, or a group of three of more particularly violent criminals,
they should be rewarded.) 95 Numerous examples in the sources indicate
that successful sheriffs were in fact awarded promotions for having
captured bandits. 96
Because such successes were important to their careers, sheriffs some-
times responded by falsely claiming to have been present when dangerous
criminals were seized, though as Ts'ai K'an commented, "they often
had sent out other men and were not themselves involved in the captur-
ing." 97 The government's answer was a set of rules designed to show
that a sheriff had displayed courage and military skill in personally
seizing bandits, the ma-ch3ien san-pu (literally, "three paces in front of the
horses") system.98 Despite such rulings, however, the government con-
tinued to be plagued by false claims. In 1177 Chou Pi-ta, after citing the
rules promulgated in the Northern Sung, stated:
92 A decree of 1029 ruled that if a sheriff who was qualified for consideration for the post
of district magistrate had captured an entire gang often or more men, he should receive
the ceremonial clothing of a fifth- or sixth-grade official, with a fish pouch and titular
status as a capital official (ching-kuan); SHY, chih-kuan 48.78a-b; see also HCP
337.15a. For the clothing, see SS 153.2b. In some cases at least these men were granted
dark red banners.
93 SHY, chih-kuan 48.62b (1024), D U t n could be "skipped over"; SHY, chih-kuan 48.62b
(1027).
94 SHY, chih-kuan 48.60b, 48.78a.
95 HGP 118.13a.
96 Compare HCP 216.3a, 13b, 299.io,b-2oa, 303.3b. A sheriff whose seniority and merit
ratings qualified him for consideration as a district magistrate was to be appointed a
prefectural judge in a regional command prefecture or a regional supervisory
prefecture; SHY chih-kuan 48.62-a-b; HCP 95.7b. See also individual cases: HCP
4.15a, 83.1a, 94.7a (1019), 299.i9b-2oa (1078), 337.15a, 303.3b; or SHY, chih-kuan
48.60b (963), in which a sheriff was appointed as a district magistrate "as a reward for
his merit in capturing bandits."
97 Ts'ai K'an, Ting-chai chi, 1.20a.
98 SHY, chih-kuan 48.84b-85a.
273
 Law and order in Sung China

Now the original system established in the T'ien-sheng period [1023-32] after-
ward became daily more and more corrupted. Generally, the sheriffs depended
on the archers to capture the required seven men, but their memorials always
said that they personally captured the bandits in accordance with the ma-ck'ien
san-pu rules, submitting these to be treated according to the precedents for
rewarding men who have achieved military merit in seizing violent bandits. Thus
they want to receive advancement to the administrative class."

Chou Pi-ta complained that the sheriffs frequently resorted to bribery in

pressing their claims. 100 They might also abuse the system by seizing
innocent people in order to fill out their required quota, by listing
nonviolent criminals as bandits, or by counting men seized by their
predecessors.101
A Sung official suggested that abuses could be reduced by returning to
the practice of granting seniority rather than changes in official grade.
Presumably this would have lessened the incentive to cheat. 102 However,
the Sung authorities chose to maintain the system of the 1020s, merely
adding some qualifications designed to make cheating more difficult.
Superior offices were required to submit documents vouching for the
sheriffs in question, 103 and rewards were delayed until the sheriff's term
was completed. At that time their records were examined. Only if they
were free of faults were the men rewarded. 104 The ma-ch'ien san-pu system,
presumably with the modifications just described, continued in use at
least into the 1220s.105
Any sheriff who was rewarded under the system of rewards and
commendations (ch'ou-chuang) could advance more rapidly. A com-
mended sheriff in his first term could jump ahead in seniority and
become eligible for a post as the administrator of a district, and a
"commended" sheriff in his second term might be nominated for a post
as a regular magistrate. 106
The most coveted prize in this game of bureaucratic advancement was

99 SHY, chih-kuan 48.78a-b. It seems that Chou Pi-ta is referring to kai-kuan -

advancement to the administrative class. The phrase he uses (kai-tzu teng-kuan) may in
fact be the full term, kai-kuan being elliptical.
100 SHY, chih-kuan 48.78a-b; see also SHY, chih-kuan 48.84b-85a.
101 SHY, chih-kuan 48.78a, 79a, 84a-b, 840-853. See also HCP 364.2a (1086).
102 SHY, chih-kuan 48.79a. Although this suggestion was approved, it was not enforced.
See SHY, chih-kuan 48.870-883 (1220).
103 SHY, chih-kuan 48.78b, 8ob-8ia.
104 See a document dated 1024 that recommended that sheriffs be commended and
rewarded if, when they were being replaced after having finished their term in office,
they left no "bequeathed deficiencies"; SHY, chih-kuan 48.62b. See also SHY, chih-
kuan 48.67b.
105 SHY, chih-kuan 48.87b.
106 SHY, chih-kuan 48.78a-b.

274
 Personnel selection

promotion from the executory class to the elite administrative class. This
move was called "changing office" (kai-kuan) . 107 Gaining "rewards and
commendations" seems to have been crucial to securing this promotion.
The Sung dynasty history says that to make this jump, a sheriff had to
accumulate three terms in office, six annual merit ratings, recommenda-
tions from higher officials, plus personal citations of merit. 108 The rules
governing these promotions changed occasionally, so this description
should be taken as exemplary rather than definitive. For instance, the
rules were reissued under Shen-tsung, at which time it was ruled that a
sheriff who had received a first-class award under the system of com-
mendations and who also had five annual merit ratings might advance to
the administrative class. 109
We do not have any examples of these claims for the reward of
advancement to administrative class, but a report from 1177 gives some
icjea of the number of sheriffs who claimed rewards for capturing ban-
dits. Five such applications were submitted in 1171 and 1172, eight in
1173, eighteen in 1174, sixteen in 1175, and thirteen in 1176,110 this at a
time when there were probably approximately eight hundred sheriffs in
the empire. To put these figures in perspective we should recognize that
in the late twelfth century the annual total number of executive-class
officials who were permitted to advance to administrative class was
limited by law to about eighty.
Furthermore, we should stress the extraordinary character of such
promotions. Normally, to advance from executory to administrative class,
a candidate had to pass through a laborious procedure, acquire consider-
able seniority, and collect at least five sponsors' endorsements, a process
that usually took ten or more years. 111
From the beginning these rewards were coupled with negative sanc-
tions for failure. Events in the decade after 962 had convinced the Sung
leaders that the original system of penalties for failing to capture bandits

107 Winston Lo, personal communication. See also Lo, An Introduction, pp. 212-13.
108 SS 169.15a-b. The sources give numerous examples of men promoted for success
against bandits. See, for example, Lin I-chih (1136-85), Kang-shan chi, SKCSCP ed.,
4.5b- 7a, 1 ib—13a. These rules lend particular cogency to a suggestion of Professor E.
A. Kracke, Jr. that at the time when the Sung regulations were being formulated, the
government writers were using hsiian ("selection examination") as a synonym for jen
("term in office"), because in the T'ang a hsiian had been taken for each term in office.
This regulation would thus be saying that a sheriff who had acquired merit enabling
him to "be forgiven one selection" was in fact being credited with an extra term of
service, so that he might be advanced more quickly to a higher office.
109 Chin Chung-she, "Pei-Sung chu-kuan chih-tu yen-chiu" (A study of the official
organization of the Northern Sung), Hsin-ya hsueh-pao, vol. 9, p. 251.
110 SS 158.2a.
111 HCP 290.15a.

275
 Law and order in Sung China

was too harsh and rigid. In 971 an order was issued that if the officials
who exceeded the time limits for seizing bandits eventually did capture
the wanted men, this was to be entered on a register and could be used
to reverse the punishments of the responsible officials. However, a man
who had been expelled from office under the system of demerits (tien-
chiangfa) could not seek relief under this rule. 112 The authorities under-
stood that sometimes locally responsible officials did not have sufficient
force and so had to call on their superiors for help. However, if the
law-enforcement authorities had enough armed strength to oppose the
bandits but failed to do so - delaying so that the bandits escaped - then
they were to be punished. Moreover, if the third time limit passed and
no bandits had been caught, the sheriff was to be fined one month's pay,
or the magistrate, half a month's pay. (The latter rule perhaps refers to
districts in which there was no sheriff.) When the sheriff had been fined
three times for this reason, or the magistrate four times, he was to be
demerited one selection (Hen i hsiian). Three such demerits could lead to
expulsion from office.113 A report of 1067 clarifies the penalties for failing
to capture the wanted men within one hundred days, and some of the
practical problems of administering the system:
If it has been a bad year, then hungry people will gather to rob. If the jurisdic-
tion reports this as banditry, then the officials charged with capturing the
robbers will be bound by the precedent on capturing violent bandits within the
time limits. If they do not capture them, then in accordance with the rules, they
will be investigated and reprimanded, including the separate punishment of
demotion in grade. Yet if they do capture the chief robbers and are eligible for
commendations, the Court will take into account the impact of the natural
disasters, and that the people were stealing food, and reduce the degree of the
commendations. . . . I fear that under this system, it will be difficult to motivate

112 HCP 22306.2a. I think we must assume here that an official was simply returned to
the status quo ante in these cases. If, for instance, it was a first failure for a sheriff and he
had been docked a month's pay, it would be paid to him, and so on.
113 The most complete report on this set of rules in found in SHY, ping 1 i.ib-2a, but it is
repeated in somewhat abbreviated form in SHY, chih-kuan 48.60a-b, and HCP
3.i4a-b. For an example of such rewards, see HCP 4.15a (963). More rules were set
up in 968; HCP 9.2a; Wang Ying-lin, Yu'-hai (Chia-ch'ing blockprint edition), 66.21b.
When serving as a prefect, Su Shih called in the sheriff to discuss some uncaptured
bandits. He is said to have told the sheriff that if the bandits were captured, he would
report it to the Court but that if they were not captured, he would see to it that the
sheriff lost his post; Hsieh K'o, Chu'-yu chiu-wen (Chih-pu tsu chai ts'ung-shu ed.)
10.4a—b.
114 SHY ping n.27b-29a.

276
 Personnel selection

Remuneration
Still, such negative sanctions were probably minimally effective. As the
Southern Sung official Ch'en Ch'un (i 153-1217) observed about the
ineffectiveness of the system of rewards, "How could the idea of someone
posthumously receiving silver and silk overcome men's fear of death?" 115
Motivation was also affected by the level of the police officials' remuner-
ation. At the beginning of the dynasty, as Wang Yung remarked, the pay
of these officials was very low. Recorders and sheriffs received only 3
strings, 570 coins per month (with magistrates receiving no more than 10
strings). 116 In 1024 concurrent sheriff-recorders in small places (fewer
than two thousand households) were in the 7-string-per-month salary
range. This compares with the 10 strings paid to magistrates in these
backward areas. 117 We can get some idea of the value of such a salary by
noting that the archers who served under the sheriff were paid about 30
strings a year, or 2.5 strings per month.
In addition to money, Chinese officials also received remuneration in
kind. A ruling on grants of office land dated 1039 divided these pro-
perties into three types depending on regional prices. Sheriffs received
50, 100, or 150 bushels per year depending on the region (their superiors
the magistrates received three times these amounts). 118 We may give
these figures more meaning by noting that in the Sung the subsistence
ration for people in times of dearth was half a pint per day and the
ration for a regular soldier was slightly more than two and a half pints
per day. Thus even the smallest of these figures, 50 bushels (five thou-
sand pints, or more than thirteen pints per day), should have provided
an ample rice ration for a household of half a dozen people for a year.
In 1043 w e a r e given rates for hsien-kuan t'ien (limited office fields)
under which sheriffs in districts with ten thousand or more households
were limited to three ch'ing* (one ch'ing* was one hundred mou, approxi-
mately fifteen acres). Those in districts of from five thousand to ten
thousand could have two ch'inga fifty mou, and those in districts with
fewer than five thousand households could have two ch'ing (with magis-
trates having allotments twice this size). 119 According to the rules of
1073, recorders and sheriffs, in districts with ten thousand or more

115 Ch'en Ch'un, Pei-ch'i ta-ch'iian chi, 4 3 . 5 a - b .

116 Wang Yung, Sung-ch'aoyen-i i-mou lu, 2.2b.
117 H C P 102.8b. Their actual salaries were paid from regularly issued funds until 1043
and from specially issued current funds thereafter. H C P 141.11a, 146.3a.
118 H C P 123.3b.
119 H C P 145.14b-15b.

277
 Law and order in Sung China
households, were to receive three ch'ing* (approximately forty-five acres)
of office land; for districts with from five thousand to ten thousand
households, two and a half ch'ing; and for those with fewer than five
thousand households, two ch'ing* (magistrates received twice these
amounts). 120 Allotments in 1073 thus ranged from two hundred to three
hundred mou (about thirty to forty-five acres). These allotments, where
provided, would presumably be rented out to tenants, with the official
receiving the rents in money or in kind. We can get some feeling for this
allotment by noting that under the equitable field-land distribution sys-
tem used in the early T'ang, under which peasants were supposed to
receive an allotment of land sufficient to support themselves, an adult
male peasant was supposed to be allotted one hundred mou. The land
allotment of the lowest grade of sheriffs was thus perhaps not especially
munificent but, when added to his other incomes in kind and in money,
should have made possible a comfortable life for a sizable household of
family members and retainers.
Southern Sung gazetteers also give us more specific information on
income in kind. In Tang-t'u District in Liang-che Circuit, the concurrent
sheriff-recorder received annually 33 ounces of silk floss, 4 bushels (shihb)
of wheat, and 23 bushels of unpolished rice. 121 In Ch'ang-shou District
in Liang-che, the sheriff's office received 284 bushels, 2 pecks (toub), 2
pints (sheng) of rice, and the two patroling inspectors received only 74+
and 123+ bushels, respectively.122 Thus, although sheriffs and patroling
inspectors were not well paid by official standards, they were well off by
the standards of ordinary people. Like many traditional Chinese officials
they no doubt often set aside part of their income to increase their
patrimonies, in preparation for the day when they would leave office.
Even so, the more noteworthy sheriffs at times contributed from these
incomes to repair their own offices.123

Conclusion
The distinctiveness of the post of sheriff derived from a combination of
functional importance (since the maintaining of law and order was sine
qua non to the stability of the state) and bureaucratic unimportance
(since these men were almost at the bottom of the bureaucratic hier-
archy). The evolution of the personnel policies governing this post reflects
both the interaction of these two traits and the more general political

120 HCP243.i4a-b.
121 Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, p. 2843.
122 Sun Ying-shih, Ch'in ch'uan chih, pp. 2707-8.
123 Yuan Fu, Meng-chai chi, 17.251, 18.258-59.

278
 Personnel selection

history of the Sung. At the beginning of the dynasty the rulers faced two
overriding problems, enforcing order and restricting military power. The
sheriffs were to play key roles in dealing with both of these problems, so
the central government sought highly qualified men as candidates. They
much preferred men from the elite group of examination graduates.
Presumably highly indoctrinated, well-motivated, and civil rather than
military in their attitudes, such men were the perfect choice, but they
were few in number. Moreover, the combination of low bureaucratic
standing with dangerous duties caused many candidates to decline
appointments to the post. As early as 989 there was a complaint that a
district with a population of five thousand households had been unable
for seven years to secure a sheriff, having only a recorder serving con-
currently. 124 In 1021 it was reported that there were 643 vacant posts for
recorders and sheriffs,125 and in 1023 t n e Bureau of Civil Nominations
reported that some of these posts had gone unfilled for more than seven
years. 126 The Sung never did solve this problem. The Southern Sung
official Tu Cheng (1166-after 1218) remarked that in his day there were
sheriffs' posts that had been vacant for fourteen or fifteen years. 127
Again, not filling posts, like leaving sheriffs in their posts for very short
periods, can only have weakened the effectiveness of the law-enforcement
system. The growth of this problem coincided with the decline in the
powers of the military and the rise in the number of men who lacked
offices. The authorities responded by lowering the standards for appoint-
ment, filling many of these jobs with men who came from "outside the
career" (liu-wai), or with graduates of the facilitated examinations.
This in turn created a new problem. Many of these men were aged
and incompetent. Therefore they could not effectively maintain order,
and as a result there was an increase in lawlessness. These growing
disorders caused some government leaders to favor a harder approach to
the lawless, symbolized by increasing the quotas of archers and introduc-
ing the use of military men as sheriffs. This practice of using military
sheriffs began on the frontiers, where the conditions perhaps required
such officers, but it spread to other regions, where its usefulness was
questionable.
The authorities placed themselves in a dilemma. Military sheriffs were
supposed to be more effective in suppressing disorder. But insofar as
they succeeded and peace reigned, the duties of the sheriff became more
bureaucratic, and the incumbent officers less qualified to handle them.

124 HCP 30.7a.

125 HCP 97.11b.
126 HCP ioi.5a-b.
127 Tu Cheng, Hsing-shan t'ang-kao, 6.naff.

279
 Law and order in Sung China

The result was a compromise rather than a solution. This compromise,

and the ruling that led to it, provide in miniature a picture of some of the
problems of law and order in Sung China, and the devices that the state
designed to deal with them. An elaborate scheme was created, specifying
who, where, and in what order men were to be chosen, a scheme that
also varied among regions, depending on the authorities' judgment
about the need in an area for a "gentleman" or a "hero."
Through this whole history run two threads, the steady increase in the
numbers of police personnel and the increase in the number of military
men relative to civilians in the law-enforcement process. At the begin-
ning of the dynasty, order was maintained by one sheriff per district and
one patroling inspector for a group of prefectures; by the middle of
the Southern Sung some districts had two sheriffs, and many had two
or more patroling inspectors as well. For example, in 1133, Ch'ang
Prefecture in Liang-che Circuit, which had four districts, had a total of
twelve patroling inspectors. 128 In 1181, Ning-tu District in Chiang-nan
West Circuit had two sheriffs and four patroling inspectors. 129
No doubt some of this increase was simply a function of population
growth - if a district's population doubled, it would presumably need
more police to maintain order. But beyond this demographic factor we
may also see an era marked by a growing disrespect for the law and for
those charged to uphold it. The state was investing more revenue in law
enforcement. The group involved in law enforcement might have been
pressing for more resources as a matter of self-interest, or the state might
have been seeking to reduce the level of lawbreaking either as a result of
its own conceptions of the importance of local security or because local
interests were more effective in demanding protection. Still, the simplest
and intuitively most reasonable explanation for the greater size of the
law-enforcement system, beyond that needed to keep up with simple
population increase, is an actual deterioration in the standards of local
order.
The Confucians always insisted that the fundamental problem of
governance was the selection of good men, not formal structures or
accepted policies. Good men would do what could be done to create
good formal structures and suggest appropriate policies, and they alone
would be capable of using in a humane way whatever structures and
policies there were. In the abstract this is a strong position. The problem
is how to get enough men of proper character and ability. Given the
qualifications of the good official, the policies for selecting and inducting
officials, the pool of possible candidates, and the right of a nominated

128 SHY, chih-kuan 48.70a. 129 SHY, chih-kuan 48.79b.

280
 Personnel selection

candidate to refuse appointment to an offered post, could the state

authorities find enough men for the available positions? With approxi-
mately one thousand posts for sheriffs in the empire and average terms
in office of about two years, the state had to appoint about five hundred
sheriffs a year. Of course, some men did serve longer than two years, but
many served less, and some men might be reappointed for a second or
third term to a post as sheriff. Even so, there was clearly a need for many
new men. At its most productive, the examination system could provide
only a few hundred potential officials per year. Even if all of them had
gone to sheriffs' posts, there would still have been a need for other sorts
of men. Moreover, the higher graduates from the examinations did not
go to sheriffs' posts but to more prestigious jobs. Numbers were a
problem.
The state authorities could have listed the desired qualifications for
men to serve in law-enforcement positions. If they had done so, two
problems would have been immediately apparent. First, some qualifica-
tions were not commensurate with others. The position of sheriff in
particular, as compared with the position of patroling inspector, had a
variety of responsibilities that called for somewhat different personality
traits. The sheriff was both a bureaucrat and a law-enforcement agent.
He needed both intellectual competence, for his investigatory and mis-
cellaneous duties, and physical competence, for the pursuit and capture
of criminals.
The second problem was more general. The sheriff occupied a position
almost at the bottom of the hierarchy of status and prestige in the civil
service, but the post brought with it difficult and potentially dangerous
duties. There was, for all the offices in the civil service, a limited pool of
truly qualified manpower. When supernumerary officials who gained
office through grace or contribution are counted, the Sung authorities
certainly had a great many men whom they might in theory select, but a
great many of these men were by no stretch of the imagination qualified
for actual office. Furthermore, given the limited numbers of men in the
pool of those with real qualifications for office and the availability of
better positions, the more highly qualified men would certainly tend to
avoid serving in an office with low prestige and burdensome duties. It is
hardly surprising that the search for good sheriffs was never ending and
only partially successful.
Therefore the state leaders had to recognize that they could get the
services of only a few of the best-qualified men. They had to be satisfied
with men of lesser qualifications in many posts and to leave many posts
unfilled. With this situation facing them, the central authorities tried to
create the best possible arrangement for securing the best level of per-
281
 Law and order in Sung China
formance that could reasonably be expected. They created an elaborate
system of rewards and punishments to encourage sheriffs (and patroling
inspectors) to work effectively and to discourage abuse of office. Given
the constraints within which they worked and the problems they faced,
we should not be surprised that they sometimes failed but, rather,
amazed that they succeeded as often as they did.

282
 9
Urban crime and
urban security

Introduction
The preceding chapters have described the law-enforcement system of
the Sung dynasty, principally as it operated in rural areas and small
towns. The pattern of law enforcement, and the related aspects of main-
taining security in the larger cities of Sung times, differs in a number of
significant ways from the pattern in rural areas.1 The most striking
change initiated by the Sung was a renewed emphasis on the use of civil
officials, mainly the sheriffs, in maintaining order in rural areas. They
were responsible for law enforcement in both the district towns and the
surrounding countryside. However, in the cities of the Sung there was
greater continuity from the practices of the T'ang and earlier and a
concomitant reliance on the military as the principal agents in maintain-
ing law and order.
District towns were often quite small. In them the sheriff was the most
important official concerned with law and order; the prefectural seats,
usually larger urban centers than the district towns, also were the head-
quarters of army units that could be called on for law enforcement.
Civilian sheriffs might also contribute, because the prefectural city was
usually also the location of a district government with a sheriff on its
staff, but the military were the dominant group. In the Sung some of the
prefectural cities were quite large, forming part of a hierarchy of urban
places culminating in the capitals, which were among the largest and
most sophisticated cities in the world of that time. Much of our informa-
tion on urban crime and law enforcement pertains to these capitals,
Kaifeng and Lin-an. Fortunately, the scattered evidence we have for

1 An earlier version of the material appearing in this chapter can be found in Brian E.
McKnight, "Urban Crime and Urban Security in Sung China," Chinese Culture 29
(December 1988): 23-66.

283
 Law and order in Sung China

other cities suggests that the patterns characteristic of the capitals were
reproduced on a less elaborate scale in other urban centers.

Crime and criminals

The city, like the countryside, produced the usual crop of crimes. Fraud,
extortion, homicide, arson, robbery, adultery - the whole spectrum of
wrongdoings were committed, though some with a peculiarly urban
twist. Robbery and theft by individuals were endemic in large Sung
cities. Writing in the thirteenth century, an official described the situa-
tion in the capital city of Lin-an:
Recently both inside and outside the city, we have had those who in broad
daylight seize foodstuffs and those who rob women of their jewelry. There are
also cutpurses who cruise the streets at night, taking people's valuables and even
causing injuries. The reports are amazing. People are most disturbed. This is no
way to serve the four quarters and repress evildoers.
Individual robbers preyed on individuals; organized groups of robbers
sought out larger targets. Sometimes robbers formed gangs to prey on
the shipping that was the lifeblood of the capital. A report from the end
of the Northern Sung said that
in the capital the bandits often travel on the Three River boats, so that the men
charged with arresting them do not dare to enter and investigate. From now on,
both inside and outside the capital, all boats of the various rivers may be
investigated, whether they are government or private boats. All previous com-
mands forbidding such entering of boats for purposes of investigations are hereby
changed. However, anyone who in a disorderly manner boards ships, not
because of bandits, and causes trouble, destroying public or private goods or
seeking bribes, is to be tried for violating an imperial order.2
The problem of gangs of predatory criminals could be traced to
certain aspects of the Sung penal system. Perhaps the most common
severe penalty in the Sung was penal registration in certain units of the
provincial armies for a fixed term of labor. After having served the term
of labor, those so punished had the option of continuing on in other units
of the provincial armies. There they also would provide various sorts of
labor or other services for the state, though they were only rarely used as
soldiers. In addition, irregularly but frequently, young able-bodied men
from the provincial armies were taken into the regular imperial armies as
soldiers. Thus, in most prefectures, including the capital, there were
rather large bodies of convicts who were serving out their sentences

2 SHY, ping i2.2ib-22a.

284
 Urban crime and urban security

and other men in the armies who had first entered the armies as con-
vict laborers. This clearly posed security problems. A decree of 1009
observed:
We have heard that in the capital there are stalwart young men who because
they have committed banditry, have been tattooed. Frequently, because they
are controlled loosely and without restraint, they cause trouble to the people.
Kaifeng Prefecture is ordered to write out their names, seize them, and divide
them up for registration in the outer prefectures.3
These problems were understandably acute in the capital. On the one
hand the capital government needed substantial numbers of provincial
army units for labor. In addition, a number of state enterprises head-
quartered in the capital needed large numbers of workers, who often
were convicts, and also there were large imperial army garrisons.
To meet this need for labor, the state would have had to bring in large
numbers of convicts, but because it was the capital, there was a desire to
be especially careful to preserve security. Early in the dynasty — continu-
ing the pattern set in the Five Dynasties - the Sung had all those
sentenced to penal registration sent initially to the capital. 4 From there,
presumably, those not needed in the capital were sent elsewhere. Early
in the eleventh century, however, this policy was changed, and men
could be sent from the prefecture of their conviction to their place of
registration. Still, a large number of such men were needed in Kaifeng.
Criminals from Kaifeng itself were often sent elsewhere to serve
sentences of forced labor, perhaps in order to prevent men from continu-
ing their old, bad, associations. An edict of n 34 that deals with condi-
tions in the Southern Sung capital of Lin-an states:
The situation in Lin-an Prefecture is like the situation in Kaifeng. If there are
those from this prefecture who are liable for penal registration, they should be
provisionally registered in the neighboring prefectures. Accordingly, when there
are men liable for registration in neighboring prefectures, do not register them in
the capital.5
Such men were not supposed to be sent back to the capital on assign-
ments. 6 In the capital itself, those men who were serving as laborers
were, after 1083, rotated monthly among the major military organiza-

3 SHY, ping 11.6a.

4 See, for example, comments in WHTK 166.1443, SHY, hsing-fa 4.4a-b.
5 TFSL 526. An ordinance found in the same work also says that generally "for those who
are supposed to be registered in the prefecture of conviction or the units of the provincial
armies in that prefecture, in the case of the capital, they should be registered in
prefectures near the capital" (TFSL 520—21).
6 SHY, hsing-fa 4.21b.

285
 Law and order in Sung China

tions in the capital, from service in the Capital Command, to service in

the Metropolitan Cavalry Command, to service in the Metropolitan
Infantry Command, and back again. Presumably this was done to help
resolve the problems that might arise if such men stayed too long in one
unit and thus became overly familiar with its staff and operations. 7 In
1168 the prefect of Lin-an remarked:
Recently in the prefectures and districts, there have been petty thieves who have
joined together in small numbers to form gangs that rob travelers. Basically this
is because the men who have been judged and sentenced to penal registration
escape before even reaching their places of registration, or they have already
reached their places of registration, but the official charged with controlling
them has let them escape, or they have been divided and sent to the garrisons in
the various jurisdictions, and the armies have neglected to control them so that
they escape. When they have no place to return to, they gather together to form
gangs.8
The situation in the urban areas created by the presence of large
numbers of registered criminals and ex-criminals serving in the armies
was probably not amenable to any lasting solution. The convicts were
certainly the greatest problem, but even ordinary soldiers who were not
deserters could cause problems by their riotous behavior. 9 They were,
after all, like the convicts, in their being young, frequently unattached
males with limited akills and limited prospects. The combination of such
men and a group of convicted criminals in close association was a
dangerous one. It is thus not surprising that the authorities shifted back
and forth, sometimes bringing convicts to the capital but periodically
ordering that certain categories of criminals from the capital be sent to
various different prefectures away from the city itself.10
The typical urban criminal was young, male, and possibly a recidivist
or a deserter from the army. This would mean almost certainly that he
came from the laboring groups in Sung society. But not all criminals
were from the underprivileged strata of the Sung. A report dated 1133
describes how members of the imperial clan, or young men who because
of their relatives' positions had the privilege of protection (yin), were
opening up shops. There they sold privately produced liquor at night
and robbed their customers. 11 And a century later, after describing a
crime wave in Lin-an, an official recommended:

7 SHY, hsing-fa 4.28b.

8 SHY, hsing-fa 4.52a.
9 SHY, hsing-fa 7.1a.
10 SHY hsing-fa 4.44a, 47~48a, 50a.
11 SHY, hsing-fa 2.147a.

286
 Urban crime and urban security

In Lin-an Prefecture more police posts and fire stations should be established.
There should be a strict system of patroling. A clear system of rewards and
punishments must be established governing the capture of fugitives and deter-
mining the time limits for their capture, so that the capital can again be secure.12
Like the countryside, the cities also had their share of oppressive and
extortionate bullies. In the capital city in the early eleventh century, a
case was prosecuted that resulted in the demotion of a number of
officials and the punishment of the principal criminal, a man named
Ts'ui Pai, and his son:
Ts'ui Pai's family lived in the capital. He was simply overbearing and coerced
people, seizing their valuables to become rich. Formerly there was a man named
Man Tzu-lu who was aggressive and abusive. His name was sent to the auth-
orities. Also, a man named Chao Chien was incorrigible about submitting to the
law. Ts'ui Pai said to people, "Man Tzu-lu is my follower. Chao Chien is my
man. You others are not fit to deal with them." Liang Wen-wei was a neighbor
of Pai. Pai wanted to force Liang to sell him his house, but Liang Wen-wei would
not do so, and so Pai often verbally abused him. When Liang Wen-wei died, he
left a wife, nee Chang, and two small children. Pai sent men, day and night, to
throw stones and tiles at her house in order to frighten her. Mrs. Chang was
unable to stop them and moved out. She then set a price for the house of thirteen
hundred strings of cash. Pai made an offer of nine hundred strings and bought it.
Mrs. Chang brought a suit against him in the capital court and subsequently
was paid an additional three hundred strings. Then Pai secretly reduced the
rental valuation [of the property], and sending a male slave as a witness, he
brought suit in the prefecture against Mrs. Chang. Furthermore he heavily
bribed the clerks. He toadied to [the erudite of the National University, Han]
Chiu, who was the acting responsible official. Han Chiu therefore sent the case
for trial to Ho Chiu-kung. Chiu-kung tried Mrs. Chang for having wildly inflated
the rental value and had her beaten. Pai then bragged to his neighbors about the
influence he had used in the case. An official of the Capital Security Office
(Huang-ch'eng ssu) reported the matter, and a decree was issued that Pai be
arrested and turned over to the Censorate for interrogation. The censors found
out the truth. As a result all the men were found guilty.13
Adultery and its aftermath was both an urban and a rural problem. A
case from the 980s reflects the suffering that might result and the ways in
which the urban investigation and adjudication system worked. The case
is far from typical, as it came to the personal attention of the emperor,
but for that reason we have a relatively detailed record of what happened.
We are told that a widow surnamed Liu committed adultery. Then
fearful that she might be found out, she became ill. Furthermore, afraid

12 SHY, ping 3.11b—12a; See also SHY, ping 3.23b-24a.

13 SHY, hsing-fa 4.70a.

287
 Law and order in Sung China

that her son might accuse her, she ordered her serving maid to accuse
the son of having poisoned his mother's food. While the mother was ill
but not yet dead, the case came in some undescribed fashion to the
attention of the right army patroller. When he could not get at the truth,
he sent the left army patroling judge (tso chun-hsun t'ui-tien) to investigate.
This man took a bribe from the widow to deal with the son. The mother
eventually died, and although the son tried to maintain the truth, he was
held for a number of months, without the matter's being settled. At this
point the Records Office (Ssu-lu ssu) was sent to investigate the two
army patrolers. As a result the false accusation again came to light. The
emperor was disturbed by what seemed to him obscurities in the charge
that the son had poisoned his mother and ordered that the son be spared
the death penalty, beaten, and punished with penal servitude in Kaifeng.
When he was about to be beaten, the son himself demanded that the
officials and clerks return the bribes that he had given them. Another
investigation ensued. The son had another accusation sent in for imperial
attention. As a result the case was transferred to the Censorate, which
investigated thoroughly. In the end, the abusive officials were punished
and the son was exonerated. 14
Because the capital was such a sensitive area, the authorities there
were also very concerned about people who spread disturbing stories or
otherwise stirred up or misled the people. In 1012 in Kaifeng there was a
scare involving men who were said to practice gruesome magic that gave
them superhuman powers. The authorities were ordered to suppress
them swiftly.15 Six years later there was a scare in the western capital
because people believed that a witch had entered the city and was
practicing cannibalism. The state's concern about this can easily be
understood. As the report says, "the army camps were greatly dis-
turbed," a worrisome situation at any time. As a result, several people
were seized and eventually executed. 16 About a century later there was a
similar scare in Kaifeng, where members of a heretical religious sect
were disturbing the peace. Again, heavy rewards were established for
those who cooperated in ridding the city of these people. 17
Finally, some of the guidebooks to Lin-an give us a less official picture
of the kinds of underworld figures who lived and worked in the great
capital city. The official Chou Mi (1232-after 1308), in his Wu-lin chiu-
shih, said:

14 SHY, hsing-fa 5.2a-b.

15 STCLC 735.
16 HCP 92.4a.
17 SHY, hsing-fa 2.65a-b.
288
 Urban crime and urban security

In the crowded neighborhoods there are a great many people; idlers who are
crafty and villainous are really numerous. There is a gang called the Bureau of
Beautiful People (which uses male and female prostitutes as beautiful concubines
to entice young men). There is also a gang called the Gambling Chest Bureau
(which depends on gambling and uses its skills to cheat people) and a Bureau of
Watery Merit that creates disturbances [using the excuses of] seeking official
positions or recommendations for official positions, of doing acts of mercy, of
making transfers, of taking care of legal cases, or of carrying out commercial
transactions, and then extorts money. One such gang would be too many. There
also are merchants who substitute false goods for genuine, for example, using
paper in place of cloth in making clothes, or copper or lead in place of gold and
silver, or earth and wood in place of fragrant oils and drugs. They are super-
naturally skilled in making substitutions. We might call them men who rob in
broad daylight. If there are market gates, then there will be cutpurses, who are
called "children who look for what is attached." There are, in addition, burglars.
Each has a recognized chieftain. These fellows, such as the "Street Blocking
Tiger" or the "Nine Striped Dragon," cause great harm in the marketplace.18

General security problems

Preserving security in urban areas of any kind necessitates dealing with
several problems, including welfare needs and fire protection as well as
the maintenance of law and order. Because of the crowding together of
buildings and people, cities face a number of problems. Their density
of settlement, particularly if the buildings are made of wood, makes
them especially vulnerable to fire. Special fire-fighting systems must be
created, which stress early detection and a rapid response by people
with appropriate equipment and skills. Because of their large, usually
heterogeneous, and frequently socially disorganized populations, cities
also often have public welfare problems. For these same reasons they
may also have special problems with investigation and law enforcement.
They need to have some judicial mechanisms for adjudicating the crimes
committed. Finally, in the case of the Sung, law-enforcement problems
were compounded by the fact that the Sung used convicts as a major
source of common labor.

Police personnel
In a large city, the authorities need a considerable number of men if they
are to carry out these security functions. One major policy issue, then, is
18 Chou Mi, Wu-lin chiu-shih, in Meng Yiian-lao, Tung-ching meng-hua-lu, wai-ssu pien, 6.444.
See also the Tung Ssu-kao, Hsi-hu pai-yung, Wu-lin chang-ku ts'ung-pien ed., shang.
iaff., for descriptions of prostitution in Lin-an.

289
 Law and order in Sung China

how the state is to find such men. The traditional Chinese answer to
this problem touches on another characteristic of city life. Because of
their great political and economic importance, cities, and especially the
capital city, need to be defended against domestic and foreign military
enemies. Therefore, all cities must have available some sort of military
force. Under premodern conditions of communication and transporta-
tion, it was not possible to station these units away from the urban areas.
Army units - of varying sizes depending on the degree of insecurity and
the importance of the place - were always stationed at or near cities in
traditional China.
The presence of these units of fighting men meant that the authorities
already had available a group of men who were of the proper age, who
had had some professional training, who were already receiving state
pay, and who were available for use in the security system. It was both
logical and simple for the state in traditional China to draw on its
existing military forces for those needed to man security posts. The
changes that occurred in Chinese urban life during the T'ang-Sung
transition in no way affected this situation.

Urban security during the T'ang

Understanding the complex and changing patterns of urban law enforce-
ment in the Sung would perhaps be easier if we described the simpler
situation prevailing during the T'ang. The use of the military as suppliers
of men for the security system was clearly the pattern in the T'ang cities
about which we have information. There were relatively few large cities
in the T'ang. The major growth of cities of all sizes occurred toward the
end of that dynasty and in the ensuing Five Dynasties and Sung periods.
Thus, almost all our information on T'ang cities and their security
problems refers to the capital at Ch'ang-an. That city was divided into
two districts by the north-south-running Red Bird Gate Street. The
area to the east of this dividing line was called the Left Streets. The area
to the west was called the Right Streets. Each area was divided into fifty-
four walled wards (fang).19
During the T'ang period, Ch'ang-an officials from the Censorate had
charge of investigations (a responsibility occasionally given them also in
the Sung), 20 and the Chin-wu Guard had special responsibilities for
policing. The Comprehensive Mirror reports that "if there are deaths in the
streets, the Chin-wu Guard will submit a report on them; if there are

19 Sung Min-ch'iu, Ch'ang-an chih, chap. 7.iaff.

20 HCP 233.12a.

290
 Urban crime and urban security

deaths inside the wards, the Left and Right Patroling Offices [tso-yu
hsun-shih] will submit reports on them." 21 According to a note:
The Chin-wu Left and Right Street officers divide responsibility for investigating
and patroling in the Six Streets. Generally, within the city walls, in the wards
and neighborhoods, there are military posts. One hundred guardsmen and
cavalrymen guard the main city gates. Thirty men are stationed at each major
police post. Twenty men are stationed at each of the lesser gates, and at the
small posts there are five men. At sunset they beat the drum eight hundred times
and close the gates. During the night hours the cavalry patrols the major streets
on horseback and summons military officers to investigate. At two minutes past
the fifth watch, the drum begins inside; the drums in the streets take it up; and
the gates of the wards are opened. They beat the drum three thousand strokes
and then stop.22
Once the gates, including the ward gates, were closed, it was illegal to
move about the streets. In 808, in a famous case, an official who violated
the curfew while drunk was beaten to death. This was extraordinary
punishment, since the rule from the T'ang Code merely called for twenty
blows of the light rod for those who violated the curfew without cause.
The night guards were to be punished if they allowed men to move
about who should have been forbidden to do so (though they were also
liable for sanctions if they denied permission to those who had legitimate
reasons for being excused from the curfew). 23 Only at the time of the
Lantern Festival, that is, on the nights of the fourteenth, fifteenth, and
sixteenth of the first month, were the gates left open and the curfew
suspended. 24
The T'ang also sought to promote urban security by forbidding the
carrying of arms in the city. The T'ang Code of 737 specified that people
who shot arrows in the streets, byways, or buildings could be punished
with sixty blows of the heavy rod. Even those who merely threw rocks or
tiles could receive forty blows of the light rod. If as a result of their
actions someone was killed, they could be punished with the penalty for
killing in an affray, reduced by one degree. 25 After the An Lu-shan

21 Miyazaki suggests that the Chin-wu guards were especially concerned with problems in
the network of streets separating the walled wards of Ch'ang-an, whereas the ten
censors rotated for monthly terms as left and right patroling officers in charge of
problems occurring within the walled wards themselves. See Miyazaki Ichisada,
"Kandai no ri-shi to Todai no bo-shi" (The li system of the Han dynasty and the fang
system of the T'ang dynasty), Toyoshi kenkyu 21 (1962): 27-50.
22 Ssu-ma Kuang, Tzu-chih Vung-chien (Beijing: Ku-chu ch'u-pan she, 1956), 239-2ob-2ia.
23 TLSI, 4.36.
24 See Wu Po-lun, Hsi-an li-shih shu-liieh (Sian: Shensi jen-min ch'u-pan she, 1980), pp.
166-67.
25 TLSI, 4.31.

291
 Law and order in Sung China

rebellion (756-63) there were several prohibitions against bearing arms.

In 778 there was an order forbidding the carrying of weapons and in 792
a further order concerning the storing of weapons. Finally, in 806 there
was an order for the arrest of men who, without cause, carried weapons
in the streets of the city.26
Within the wards, maintaining local order was the responsibility of the
ward chiefs (fang-cheng), who apparently were in some way subordinate
to the censors rotated to serve as left and right patroling officers. Un-
fortunately, the sources tell us very little about how these functionaries
operated or how effective they were. We do know that the richer and
more powerful households had their own retainers to guard against
intruders.
During the late T'ang, the government's control over urban areas
broke down. Strict enforcement of the walled-ward system began to
decay, and security became more and more difficult to maintain. In
urban police work, the imperial guards were overshadowed by regular
army troops. During this same era the older army units of the emperor's
guards, including the Chin-wu Guard, were less powerful than the new
Army of Divine Strategy, which like the Ghin-wu Guard became active
in law enforcement.27
The patterns of organization used in this capital-based army were
mirrored in the areas outside the capital in the armies of the military
commissioners (fan-chen), regional military leaders who came toward the
end of the T'ang dynasty to exercise almost independent power in some
regions of the empire. The man who displaced the T'ang and founded
the Liang, Chu Wen, derived his power from his position as commander
of one of these armies. Through it the patterns of command and organ-
ization characteristic of the T'ang were passed on to the succeeding Five
Dynasties. 28

Urban security during the Five Dynasties

The scattered information we have from the Five Dynasties (907-60)
shows how these T'ang patterns persisted. Military units, organized
according to patterns derived from the armies of the regional com-
mandants, continued to dominate law enforcement. The armies in the
capital of the Liang (907-23), like those in the T'ang capital, were

26 Wu Po-lun, Hsi-an li-shih shu-liieh, p. 167.

27 See Muronaga Yoshizo, "Godai jidai no gun-jun in to ma-ho in no saiban"
(Adjudication by the army patrol office and the cavalry and infantry office in the period
of the Five Dynasties), Tqyoshi kenkyu 24 (1966): 17-18.
28 Wang Yin-lin, Yii-hai 139.5a.

292
 Urban crime and urban security

divided into wings {hsiang*). According to Lii Tsu-ch'ien, there were four
such wings, which "dealt with fire-fighting problems in the city and the
suburbs." 29 This pattern of dividing the capital forces into wings and
assigning them responsibility for security in sections of the city and
suburbs persisted through the rest of the Five Dynasties and into the
Sung. 30
Chu Wen formally displaced the T'ang in 907 and declared himself
the emperor of the Liang, with his capital at Pien (later Kaifeng). Only
two years later the police structure of Kaifeng was reorganized, dividing
the capital into two large jurisdictions. The Collected Documents of the Five
Dynasties (Wu-tai hui-yao) reports that "in the tenth month of 909 the
left and right army patrol officers [tso-yu chu'n hsiin-shih] were established,"
and a note indicates that before this, although the various armies were
supposed to dispatch men to maintain security in the capital at Kaifeng,
in fact the men were often sent on different business, and enforcement of
the system was lax. Therefore at this time these new offices were estab-
lished, the Left Office being responsible for security to the north of the
Pien Canal and the Right Office in charge of security south of the
canal. 31 In some cases at least, these officers might be concurrently
appointed patroling inspectors.32 The areas of responsibility were called
the Left hsiang* and the Right hsiang*. Because by this time this word had
really come to be a territorial designation rather than the name of a kind
of military unit, the word hsiang* might best be translated as "borough."
The Later T'ang (923-36), the successor to the Liang, attempted to
imitate more closely the usages of the T'ang, of which it claimed to
be the legitimate successor. The Later T'ang rulers reassigned to the
censors responsibility for law enforcement. In name they also revived
the Chin-wu Guard; in fact the city was divided into Left and Right
Boroughs, and the Chin-wu Guard differed only in name from the army
patrol system that had been in use under the Liang. 33 Despite the use of
the term Chin-wu Guard, in Loyang, the capital of the Later T'ang, the
police patrolers were called army patrols (chun-hsun). An edict of 927

29 Lii Tsu-ch'ien, Lit Tung-lai hsien-sheng wen-chi, Chin-hua ts'ung-shu ed., 20.
30 Thus, by the Sung period, the T'ang term "street" (chieh) for the major divisions of the
capital had been replaced by the term hsiang. The term hsiang originated as a term in
military administration. Lii Tsu-ch'ien says that "the left and right wings (tso-yu hsiang)
originated in the T'ang. They were originally used in the military arrangement of Li
Ching. The various armies were each divided into left and right wings for
administration." This description is confirmed by the Yu-hai. Initially, the word was
used only for military administration; it was not applied to the divisions of cities.
31 Wang P'u, Wu-tai hui-yao (hereafter cited as WTHY) (Shanghai: Shanghai ku-chi
ch'u-pan she, 1978), 24.389.
32 Habu Kenichi, "Godai no jun-ken," pp. 45-46.
33 WTHY 12.204. See also Muronaga, "Godai," p. 19.

293
 Law and order in Sung China

stated: "We have heard that in the markets of the capital city and
the military encampments there are those who deliberately violate the
regulations, by killing cattle and selling the meat. We expect the pre-
fectural and district army patrols to increase their investigations." 34
A few years later the Left and Right Army Patroling Offices in Loyang
again appeared in the records, when they submitted a complaint about
the ways in which private households were encroaching illegally on the
public streets in the boroughs and wards. 35 The functionaries in these
offices were also given some responsibilities for controlling abuses in
trade. 36 The Southern Sung encyclopedia Shih-wu chi-yiian, quoting the
Collected Sung Documents, confirms the origins of this patroling system: "At
the end of the T'ang this position was first established. During the Later
T'ang it was divided into Left and Right." 37 Military personnel thus
continued to be the principal figures in law enforcement.
During the Five Dynasties, only the Later T'ang had its capital at
Loyang. The other dynasties had their capitals at Pien (later Kaifeng),
which was strategically located on the canal that carried rice north from
the agriculturally rich region of the delta south of the Yangtze River.
The Shih-wu chi-yiian indicates that the office of army patrol also was used
under the other Five Dynasties in the city of Kaifeng. In that city,
during the Five Dynasties and the early Sung, the position was filled by
lieutenants (ya-chiao) , 38
Army patrols were active in law enforcement in Kaifeng. In 939 an
official concerned about the severe rules governing people who caused
injuries by riding recklessly in the crowded streets of the capital asked
that before such rules were rigidly enforced, the army patrols and the
populace in general be informed about the nature of the law. Thereafter
the army patrols might enforce the law without favoring the children of
the powerful, who had often caused trouble because of their recklessness. 39

Sung administrative hierarchy in Kaifeng

Although there are still some obscurities regarding the structure of the
urban administration of Kaifeng in the Northern Sung, the overall pat-
tern is clear. In line with general traditional Chinese practice, the city
was part of the empirewide pattern of territorial administration. Kaifeng
and a large surrounding area were under the control of a metropolitan
34 WTHY 9.156.
35 WTHY 26.411.
36 WTHY 26.415.
37 Kao Ch'eng, Shih-wu chi-yiian (Taipei: Yee Wen, 1967), 6.25b.
38 Ibid.
39 Wang Ch'in-jo, Tse-fuyuan-kuei (Hong Kong: Chung-hua shu-chu, i960), 7362.

294
 Urban crime and urban security

\
\ NORTH NO. 2 NORTH NO. I
t Word 2 Words
\

\
WEST \ RIGHT NORTH LEFT NORTH EAST
NO. 3 \ II Words 9 Words / ' NO. 3
\ 7,000 hth 4,000 hsh /
\
2 Wards I Ward
II
RIGHT 2 LEFT 2
16 Wards
WEST RIGHT WEST"" Pol.
15,900 hsh. LEFT EAST EAST
NO. 2 26 Words HMHH
9 Words NO. 2
I W. 7 0 0 h.
8,500 hsh RIGHT I LEFT I 26,800 hsh
8 Words 20 Words
2 Words j 7,000 hsh. I Word
8,950 hsh. Y

WEST EAST
NO. I JL \ -L NO. I
/ RIGHT SOUTH LEFT SOUTH \
/ 13 Words 7 Words \
\
2 Words 9,800 hsh 8,200 hsh \ I Ward
-HI II ^ i ii
\
SOUTH \
2 Words \

Figure 9.1. Kaifeng about 1021: schematic arrangement. Reproduced by

permission from John W. Haeger, Crisis and Prosperity in Sung China (Tucson:
University of Arizona Press, 1975), © Arizona Board of Regents, 1975, fig. 3.1.

prefect. Among the districts constituting Kaifeng Prefecture, two -

Kaifeng District and Hsiang-fu District - included the main urban
nucleus of the capital city. 40 For purposes of administration, urban
Kaifeng was divided into the old city (the inner city) and the new
city (the outer city). The inner and outer cities were then subdivided
(in different ways for different purposes) into boroughs {hsiang*) (see
Figure 9.1).
For most of the Sung period the boroughs remained the basic large

40 Kracke, "Sung Kaifeng," pp. 61-62.

295
 Law and order in Sung China

unit for most sorts of urban administration, not just in the capital, but
in other large Sung cities as well.41 These boroughs were divided for
administrative purposes into wards. The Sung, typically, continued to
use the old term from T'ang times, even though in the Sung these units
were no longer walled enclosures but simply areas.
During the early Northern Sung the population of the city appears to
have grown rapidly, especially in the suburbs. A single sheriff was the
only official responsible for law enforcement in this suburban area. The
authorities finally became convinced that the problems of law enforce-
ment in these burgeoning suburbs were becoming too great for one
official to handle. In 1008 eight outer boroughs were established for the
capital, with functionaries who were drawn from the capital prefecture. 42
In the following year this was increased to nine outer boroughs. 43 Like
the boroughs in the city proper, there were regular complements of low-
level functionaries, whose number was determined by the size of the
population in that borough. The general rule seems to have been that in
these outer boroughs, for every five hundred households, the authorities
selected four aides (so-yu), three street men (chieh-tzu), four street officers
(hang-kuan), and one ward scribe (hsiang-lien). For areas with fewer than
five hundred households, the number of functionaries was somewhat
lower.44

Military command structure

Integrated into this system of urban units and subunits was a military
command structure. In most parts of the Sung empire, the smallest
urban areas with significant numbers of troops stationed in or near them
were the prefectural seats. In these towns, perhaps the most common

41 At the very end of the Sung these boroughs began to be displaced by units called
precincts (yii). Initially these were special-purpose units for fire fighting, but in the
Yuan period (1260-1368) they became the general urban unit of administration. For
an excellent description of the evolution of this system, see Sogabe Shizuo, Chugoku oyobi
kotai Nikon ni okeru kyoson keitai no hensen (The evolution of the form of villages in China
and ancient Japan) (Tokyo: Yoshikawa kobunkan, 1963). See also the older but still
useful work of Kato Shigeshi, Chung-kuo ching-chi shih k'ao-cheng (Beijing: Commercial
Press, 1962). Many of these materials have been assembled in Cheng Shou-p'eng,
Sung-tai Kai-feng fu yen-chiu (Taipei: Kuo-li pien-i kuan Chung-hua ts'ung-shu pien
tsuan wei-yuan hui, 1980), pp. 343-416. For a general description of urban security
works, see Michael Finneagan, "Urbanism in Sung China" (Ph.D. diss., University of
Chicago, 1976).
42 SHY, ping 3.1a; HCP 70.19b.
43 SHY, fang-yii 1.12b. I think that we must assume that these outer wards were located
outside the new city, in fact outside all the city walls.
44 SHY, ping 3.3a. This passage gives detailed population figures for the various hsiang
and fang of Kaifeng and their complements of urban functionaries.

296
 Urban crime and urban security

pattern was to have a prefect who acted concurrently as the commander

of the local military forces. In these prefectural cities the policing seems
to have been managed, on the military side, by military unit officials.
The thirteenth-century encyclopedia of institutions, Wen-hsien t'ung-k'ao,
explains:
The prefectural military directors in chief [tu-chien] were staffed by servitors
major and minor [ta-hsiao shih-ch'en]. These officers had charge of the garrison
troops, including their training and dispatching, and concurrently served as the
patroling inspectors within the cities. Those with lesser qualifications were called
supervisors of militia [chien-ya]. Earlier they sometimes used civil officials, but
later only military.45
Supervisors of militia seem to have been widely used in urbanized
areas in the Sung, providing the equivalent of the patroling inspectors in
the rural areas. The late Northern Sung official Pi Chung-yu comments
on their use in the capital, and material from other sources indicates
their use in other cities. The Collected Sung Documents says that "the
various sorts of prefectures all had supervisors of militia to exercise
control over the central city as well as over the troops stationed in
military agricultural colonies." 46
These low-level military officials used army troops as their police
force, a practice that periodically caused problems because of the troops'
tendency to abuse the people. A memorial from 1123 probably reflects a
recurring issue:
Recently soldiers have been sent into the markets to seize wrongdoers. They
oppress the ordinary people. . . . I ask that there be a special order to the officials,
to establish a law saying that in the prefectural seats, except for the borough
patroling and arresting troops [hsiang-hsun pu-ping], other troops be strictly
forbidden to go into the markets on the excuse of seizing wrongdoers.
The proposal was approved. 47

Command structure in the capital

In Sung Kaifeng and probably in some other larger cities, a different and
much more elaborate command structure was used. The Sung founder,
Chao K'uang-yin (r. 960-76), took over and modified for his own
45 WHTK, 59.541. See also SHY, ping n . i o a - n a , in which the fiscal intendants are
assigned the responsibility for assessing the qualifications of "urban patroling
inspectors." Also see a decree in 1012 that assigns responsibility for fire and law
enforcement in Kuang prefectural town to the "military administrator" (ch'ien-hsia), a
title frequently held concurrently by a civil or a military official; HCP 79.7a.
46 Pi Chung-yu, Hsi Vai chi, SKCSCP ed., I . I O - I I ; SHY, chih-kuan 49.iaflT.
47 SHY, hsing-fa 2.63b.

297
 Law and order in Sung China

Bureau of Military Affairs acting with) Ministry of War

Metropolitan Metropolitan
Palace Command
Infantry Command Cavalry Command

Commander in chief or Commander in chief or

(supervised seizure deputy deputy
of witnesses and (serving concurrently as) (serving concurrently as)
investigations) chief patroling inspector chief patroling inspector
of the New City of the Old City

East South West North) North North South South

East West East West

Borough deputy commander Bureau deputy commander

in chief (concurrently Four in chief (concurrently
Quadrant patroling inspectors?) patroling inspectors)

Patroling Patroling Patroling Patroling Patroling Patroling Patroling Patroling

inspector inspector inspector inspector inspector inspector inspector inspector

(dealt with regular police patroling and control)

Figure 9.2. Command structure in Kaifeng after the mid-eleventh century

purposes the command structure that had been evolving since the early
Five Dynasties. At the beginning there were two offices at the top of this
military command structure, the Palace Command (Tien-ch'ien shih-wei
ssu, frequently abbreviated to Tien-ch'ien ssu), and the Metropolitan
Cavalry and Infantry Command (Shih-wei ch'in-chiin ma-chiin pu-chiin
ssu). In the middle of the eleventh century this was increased to three,
when the Metropolitan Cavalry and Infantry Command was divided
into the Metropolitan Cavalry Command (Shih-wei ch'in-chiin ma-chiin
298
 Urban crime and urban security

ssu) and the Metropolitan Infantry Command (Shih-wei ch'in-chun pu-

chiin ssu). Taken together these three offices were known as the Three
Commands (San-ya). These commands were headed by commanders in
chief (tu chih-hui shih) and then deputy commanders in chief (fu tu chih-hui
shih). They were under the supervisory control of the Bureau of Military
Affairs (Shu-mi yuan), acting jointly with the Ministry of War (Ping-
pu). Although the command structure changed in some minor ways
during the Sung, the basic pattern outlined here seems to have remained
relatively constant (see Figure 9.2).
It seems probable that initially the commanders in chief had responsi-
bility, through their regular chain of command, for the capital's security
and that they used regular troops from the imperial armies (chin-chun) for
policing and other security matters. 48 The highest officials in the Palace
and Metropolitan commands apparently had final control over policing
powers in the capital for the whole of the dynasty (with the exception of
one brief period in the 990s). Under the pattern that emerged in the
middle of the eleventh century, the key to this supervisory control was
concurrent service, with the heads of these high organs holding at the
same time offices at the top of the regular police system. The commander
in chief of the cavalry served concurrently as the chief patroling inspec-
tor of the old city (chiu-ch'eng tu hsun-chien)^ and the commander in chief of
the infantry (tu chih-hui shih) served concurrently as the chief patroling
inspector of the new city (hsin-ch'eng li tu hsun-chien).49
The Southern Sung encyclopedia Shih-wu chi-yuan describes this sys-
tem: "Since the beginning of the dynasty, the four borough deputy
commanders in chief have supervised and controlled the various policing
activities within the capital city. At that time they were called borough
48 This system seems to have been modified somewhat in the late 980s. The great
annalistic history of the Northern Sung known as the Ch'ang-pien reports that in 988 the
Court established the four-borough commander in chief of the Cavalry and Infantry
Dragon Spirit Guard (Ma-pu chiin lung-shen wei ssu-hsiang tu chi-hui shih), and the
four-borough commander in chief of the Palace Command Supportive Heavenly and
Martial Guard (Tien-ch'ien feng-jih t'ien-wu ssu-hsiang tu chih-hui shih). This was
simply a clearer recognition of the important role of the commanders in chief in
maintaining security in the capital.
The emperor, however, was still dissatisfied with the arrangement. In 994 an attempt
was made to return to a pattern more like that of the high T'ang. The capital was to be
divided, as Ch'ang-an had been, into right and left streets. A special elite guard force of
about six hundred men was selected and divided into four encampments. They might
have been divided among the four boroughs, but in any case they were to serve, as the
Chin-wu guard had in the T'ang, as the key policing force in the city. This experiment
seems to have been short-lived.
49 For the practice of not appointing commanders, see a report from the Ch'ien-tao period
(1165-73) cited in SS, 166.3929. During some periods it was apparently the practice
not to appoint commanders in chief, and at such times their roles were filled by the
deputy commanders in chief.

299
 Law and order in Sung China

leaders [hsiang-chu]. In cases of banditry or fires they then sent their

subordinates to go to the sites." 50 One report suggests that each of these
officials was aided by one inspector in chief {tuyu-hou) and that the next
level of command was a varying number of borough patroling inspectors
(hsiang hsiin-chien) who had responsibility for the actual daily mainten-
ance of urban security.51
The borough patroling inspectors were presumably the top level of the
police apparatus. A decree in 1009 indicates that in the capital the
system of superior patroling inspectors (tu hsiin-chien) and patroling
inspectors (hsiin-chien) was already in use at that time. 52 At least after
1014 there were four patroling inspectors appointed for the capital city,
selected from lesser intendant servitors (ch'i-p'an shih-ch'en).53 These
officials, the four quadrants patroling inspectors (ssu-mien hsiin-chien),
were in charge of the four suburban sectors - east, west, north, and
south - surrounding the city. They were given regular army troops to act
as police.54
The importance of the office of patroling inspector increased over
time, so in 1060 it was ordered that those men serving in it must be
chosen from those who had had previous service in posts classified as
"close to the people," who had no record of corruption, and who cur-
rently held the title of audience usher (ko-men chih-hou) or higher.55
The four official suburban subdivisions also had headquarters offices
that were referred to as the police offices (chiin-hsiin yuan). These offices
were headed by left and right military inspectors (chun-hsiin shih) who
were assisted by left and right administrative assistants (p'an-kuan).56
Although the generally approved practice was to use soldiers, we know
that sometimes the authorities used less savory urban types. Before 1009,
homeless men had been registering with the borough offices. The offices
gave them some money and used them to patrol and seize wrongdoers.
These flunkies used their positions to cheat people, who did not dare to
report the abuses to higher authorities. 57 In that year a decree ordered
these men not to disturb the people.

50 Kao Ch'eng, Shih-wu chi-yuan, 6.30a.

51 SS 166.3929.
52 SHY, ping 3.1 a - b .
53 HCP 83.7a. In times of crisis the number of these patroling inspectors could be
temporarily increased. See Sun feng-chi, Chih-kuanfen-chi, 35.43a; HCP 182.13b, 183.6a.
54 HCP 83.7a, 519.1b.
55 HCP 191b; SHY, chih-kuan 48.128b.
56 SHY, hsing-fa 6.77b. These administrative assistants had miscellaneous responsibilities,
for example, for seeing that the rules on the proper construction of prison fetters used in
the police office jails were obeyed.
57 SHY, ping n.6a.

300
 Urban crime and urban security

Soldiers
Under the patroling inspectors, there were apparently several sorts of
lesser officers, as well as the common soldiers who formed the backbone
of the policing and investigation systems. 58 Although the Sung had
inherited a system in which the military provided the police, there was at
first some conflict with the prefectural civil authorities about the status
of these men, reflected in disputes about which office was to maintain
discipline among them and to punish those who abused their positions. 59
This jurisdictional problem eventually seems to have been solved by
involving both offices. According to a decree of 1007:
In the capital city, in the various inner and outer boroughs and fire-fighting
tracts, imperial army soldiers are dispatched to conduct investigations. In
general they are to investigate theft and robbery. If it is learned that on the
excuse of "spying out cases," they have taken goods, the Kaifeng prefectural
authorities are to discover, apprehend, and severely punish them. The system
will be supervised by the Palace Command [Tien-ch'ien shih-wei ssu].60
Under a system established two decades later, the persons involved in
investigations were to have their work examined monthly. They were
required to confine their function to writing down the facts of the
original crimes and then turning over their information to their superiors
for rechecking and action. 61
These soldiers filled dual functions, fire fighting and law enforcement.
The system is well described in the guidebook to Kaifeng called the
Dreams of the Eastern Capital (Tung-ching meng-hua lu):

In each ward, about every three hundred paces, there is an army patroler's post
[chun-hsunpu-shih], with five post soldiers. At night they patrol and receive cases.
Also in high places there are brick fire-watch towers. Men keep watch from the
tops of the towers. Below are several spaces of offices, with a hundred or more
garrison soldiers, and equipment, like large and small buckets. . . . When there is
a fire the cavalry quickly takes the news to the borough chief. The Cavalry and

58 Like other participants in the law-enforcement process, the soldiers serving under these
capital patroling inspectors were rewarded for capturing wanted men. Under a rule
established in 1018, captured deserters from the military were worth two hundred coins
per man, and bandits five hundred; SHY, ping 3.2a.
59 SHY, hsing-fa 7.3b. Between 1004 and 1007 the palace command (Tien-ch'ien shih-wei
ssu) complained that when its men (who were charged with seeking witnesses involved
in legal cases) committed infractions, they were arrested by the Kaifeng prefectural
authorities. The palace command asked for and was given the power to arrest, try, and
punish such offenders when their crimes were minor. In more serious cases, a memorial
had to be submitted to the Court; see SHY, hsing-fa 7.3b.
60 SHY, ping 3.1a.
61 SHY, ping 34b~5a (1025).

301
 Law and order in Sung China

Infantry commands, the Palace Command, and the prefectural authorities all
lead soldiers to extinguish the fires. They do not burden the ordinary people.62
The Dreams of the Eastern Capital describes Kaifeng as it was near the end
of the Northern Sung. The institution of police posts, however, had
existed earlier. If we may judge from an order in 1068, which resulted in
the closing of eighty-six of these posts, the average post was almost
exactly as described, because in 1068 the number of soldiers averaged
six.63
The use of soldiers as night patrols was an old practice. We are told
that early in the dynasty those so posted had to arm themselves, buying
equipment or borrowing it from others. Then in 1012, an order was
issued that the state issue them truncheons. The authorities were to use
a register to keep control of the weapons. 64
Patroling inspectors had supervisory control over both the fire-fighting
forces and soldiers serving in law-enforcement jobs. The patroling
inspectors may have had some disciplinary control over their own
soldiers; they also sometimes, improperly, judged and punished other
sorts of people. In 1019 additional army chiefs (chiin-chu) and inspectors
in chief (tuyu-hou) were given charge of the fire-fighting soldiers. 65
The fire-fighting forces were a source of continuing trouble to the
authorities. Looting commonly followed arson, 66 and the fire fighters
themselves were well known for looting buildings while supposedly fight-
ing fires. An order of 1024 sa*d: "If the fire-fighting soldiers or the men
from the water guild loot valuables while fighting fires, the patroling
inspector should immediately seize them and investigate. If the accusa-
tions prove true, the arrested men should be tried and sentenced as
bandits by the Kaifeng prefectural authorities." 67 The authorities
wanted people to begin fighting fires immediately yet hesitated to let the
soldiers start before their superiors arrived, for fear of looting. In the end
the dangers represented by uncontrolled fires forced the authorities to

62 Meng Yiian-lao, Tung-ching meng-hua-lu, wai ssu-chung (Shanghai: Chung-hua shu-chu,

1962), 3.22.
63 SHY, ping 3-5b-6a.
64 SHY, chih-kuan 4.123b. This pattern was also used in other urbanized areas, even
quite small ones. In 1019, for example, an official remarked that south of the prefectural
town of Ping Prefecture there was a market with more than two thousand households.
There was also a military camp there, but only four provincial army soldiers patroled
the area. The official asked that more men be assigned duties as police, including
making night patrols (see SHY, ping 3.2a-b).
65 SHY, ping 3.3a.
66 See a report on this problem from 1020; SHY, ping 3.2b; see also SHY, ping 2.i7a-b.
67 SHY ping 3.4b.

302
 Urban crime and urban security
permit people to begin at once to fight fires, without waiting for the
officers to arrive. 68

Jails
There were a number of jails in the capital in which people might be
held after being arrested. During the late tenth century there seem to
have been problems because of the lack of rules about what units could
properly imprison people, so an order was issued in 997 stressing that
the various offices in the capital city not be allowed to incarcerate people
without authorization. 69 Many central offices, however, were allowed to
establish jails. In addition, the local government entities of Kaifeng, such
as the districts of the capital area and the prefectural government, had
jails. The most important capital jails seem to have been those of the
prefectural government itself and the jails of the borough patroling
inspectors' offices. Early in the dynasty the jail (or jails) at the Censorate
were important, and the Court of Judicial Review had a jail, but this was
abolished, only to be revived during the reign of Shen-tsung (r. 1067-85),
abolished again, and then again revived. 70 Finally, the armies undoubtedly
had their own jails in the capital as they did elsewhere in the empire.
From the beginning of the dynasty the emperors expressed concern
aboui the treatment of those held in jails, a concern that was reflected in
ait Increasingly elaborate set of rules governing jail personnel. A typical
edict was issued in 969, when the emperor ordered the senior officials in
the various prefectures of the two capitals to investigate the condition of
those imprisoned, supposedly every five days. The jail cells were to be
washed and swept, and the fetters of the prisoners washed off. Prisoners
who were poor and unable to support themselves were to be given
food and, if ill, medicines. When the crimes involved were minor, an
immediate disposition was ordered. From this time on, the text says,
such an order was issued every summer, and "it became a regular
custom." 71 Prisoners in the capital (and in other areas) were also sup-
posed to be given coal and warm clothing in the winter. 72 As described,
in Chapter 12, the authorities' concern was also reflected in rules govern-
ing the various sorts of prison instruments. Fetters, chains, manacles,

68 SHY, ping 3-ia-b, 4b; HCP 110.1b.

69 SHY, hsing-fa 6.52a.
70 WHTK 166.1445, 167.1449-50; HCP 295.8a-b; SS 201.5021, 5023.
71 SS 199.4968. See also, for example, HCP 61.lib; SS 199.4992. On the continuing
concern about providing food and medicine, see orders (with regard to the capital area
as well as the outer areas) from the Southern Sung; SS 199.4993.
72 SHY, hsing-fa 6.56b-57a.
303
 Law and order in Sung China

and the cangue all were governed by rules that determined their size,
weight, and composition. And in fact, we know of cases in which respon-
sible officials in the capital city were demoted for having improper
instruments.73
Concern about abuses in the capital also prompted the emperor to
order the appointment of a judicial intendant for the area in 1006.
The Southern Sung encyclopedia Shih-wu chi-yiian says that during the
previous summer the authorities of Kaifeng Prefecture had interrogated
a doctor of letters, judicially torturing him and ripping open his back,
without having the case statement supposedly necessary before torture
could be legally applied.74
All this concern, however, did not prevent numerous deaths among
prisoners in jails. The response of the authorities was to establish quotas.
The rule for the capital said that if seven men died in the jails of the
Kaifeng borough patroling inspectors' offices, the jail personnel there
would be liable for sixty blows of the heavy rod. In places other than the
capital, the functionaries would be punished if two men died in a year.
The larger quota in the capital is presumably due to the larger numbers
of prisoners. When in 1005, the emperor examined the prisoner records
for Kaifeng Prefecture, he found more than two hundred persons cur-
rently being held in jail. 75 As we noted, the judicial intendants of the
jurisdictions, including Kaifeng, had to submit an annual report to the
Court, listing the number of prisoners who died in the jails under their
jurisdiction.76

Adjudication

T h e way in which adjudication of minor cases in the capital was handled

early in the dynasty is n o t clear. It seems, however, that borough
patroling inspectors h a d been j u d g i n g some minor cases. T h e Southern
Sung encyclopedia Shih-wu chi-yiian observed:

In the past, in Kaifeng, lieutenants [ya-chiao] were chosen to act as [magistrates].

In 973 this changed. For the first time, literati - Li O and Chao Chung-heng -
were appointed to this post. In his San-ch'aopao-hsun, Lii 1-chien says, "In 973 the
emperor spoke to his advisers, saying, 'Until now only lieutenants [ya-chiao]
have been appointed to serve as the Kaifeng Prefecture left and right army
patrols and the various prefectural cavalry and infantry magistrates. Punishment

73 SHY, hsing-fa 6.79a; SS 199.4993.

74 Kao Ch'eng, Shih-wu chi-yiian, 6.19a.
75 HCP6i.nb.
76 SHY, hsing-fa 6.56a-b; SS 201.5021. The WHTK 167.1448 gives the figure as ten
deaths. See also a modification of this system in 1093; SHY, hsing-fa 6.57a-b.

304
 Urban crime and urban security

deeply affects the people's lives. From now on we should appoint literati to
doit.'" 77
The judgment of minor cases by the borough officers continued at least
until 1029, w n e n it was forbidden. The report says that "at the begin-
ning of the Chih-tao period [995-97] the generals in charge of regular
army troops in the garrisons, and the borough officers, had, in a dis-
orderly way, been judging legal suits and incarcerating men." 78 In 1029
the Court issued an order: "When there are fights or disputes in the
capital, those involved are to be sent to the prefectural authorities for a
hearing. The patroling inspectors of the borough in question are not
allowed to render judgment." 79 In 1033, perhaps because the press of
cases was overburdening the prefectural authorities, the Court ordered
that judges be established in each borough, to handle minor cases. 80
This arrangement seems to have remained in use until the reforms under
Shen-tsung.

Heavy-penalty system
The Sung authorities also sought to increase security in the capital by
creating special systems of laws and punishments for use there. The
Sung commonly compiled sets of decrees and rules applicable to certain
regions, including the capital. From the Northern Sung we know, for
example, of the two-chapter "Hsi-ning Period Decrees Concerning the
pao-chia System of Kaifeng Prefecture" and the ten-chapter "Recompila-
tion of Decrees from the Hsi-ning Period Relevant to Kaifeng Prefecture"
and, for the Southern Sung, of a twelve-chapter compilation of gener-
ally applicable laws for the eastern capital, which included decrees,

77 HCP 12306.17b; Kao Ch'eng, Shih-wu chi-yuan 6.25b. Lii I-chien was in a position to be
well informed about these matters. In 1006, when the Kaifeng Prefecture judicial
intendant's post was established, he was its first incumbent. See Kao Ch'eng, Shih-wu
cki-yuan, 6.19a. A decree from the following year extended this system to the outer
prefectures. The report states: "Before this, the prefectures in the various circuits had
appointed lieutenants [ya-chiao] as cavalry and infantry inspectors in chief concurrently
serving as magistrates for judging cases. They often were incorrect. In the autumn of
974 this was stopped by decree. The Cavalry and Infantry Office was changed to be the
Public Order Office [Ssu-k'ou yuan]. Administrators for public order [ssu-k'ou ts'an-
chun\ were chosen from new graduates with the degree of chin-shih, nine Classics, or five
Classics, or from among executory officials [hsuan-jen] who had the proper seniority."
See HCP 12306.18b.
78 SHY, ping 3.1b.
79 SHY, ping 3.2a; HCP 83.5b. See also HCP 99.i4a-b, which records a decree specifying
certain times during which the palace cavalry and infantry commands were not allowed
to pass judgments.
80 WHTK 63.568.
305
 Law and order in Sung China
ordinances, regulations, specifications, and clarifying decrees. 81 There
were also special rules with special punishments for crimes committed in
the palace area. 82 More importantly, however, the Sung introduced a
system that increased penalties in the capital region, the "heavy-penalty"
system, which will be described in Chapter 15.
Although the system of law enforcement followed in the Southern
Sung capital was patterned after that in Kaifeng, it is not clear whether
the heavy-penalty system continued in general use. We do have one
order from 1132 specifying that certain types of robbers or arsonists were
to be "judged according to the laws used in Kaifeng." 83 People could
also be punished with special severity by subjecting them to military
law, which for most crimes called for beheading, a practice used in the
Southern Sung against some arsonists. 84 In general, however, it seems
that the ordinary laws were applied.

Reforms in the late Northern Sung

During the reign of Shen-tsung (r. 1067-85), in this as in so many facets
of governmental organization, there were significant reforms. The basic
trend was to create a more complicated apparatus, by dividing up
among the newly created offices some of the functions associated with
maintaining security. Fighting fires, patroling, investigating crimes,
measuring boundaries (which might be important to lawsuits regarding
property), taking depositions, and adjudicating cases were divided
among several offices.
A separate system for adjudicating minor cases was the first step in
this division. In the fifth month of 1070 the provisional prefect of
Kaifeng, Han Wei, memorialized asking that the four borough servitors
be abolished. To replace them he asked for the appointment of four civil
officials of relatively high rank (capital or court officials, ching-ch'ao kuan),
who had had experience as prefectural controllers-general or district
magistrates, to act as judges for minor cases in the capital. These men
were to have charge over the left and right boroughs of the new and old
cities. They were to have the power to decide legal cases in which the
penalties were sixty or fewer blows of the heavy rod. They were also to
handle cases involving debts and marriages, plus some other miscel-
laneous duties.
Under this system the previously appointed servitors, who had per-

81 SS 204.6a, 7a; SHY, hsing-fa 1.38b.

82 HCP 493.3a.
83 HNYL 917; SHY hsing-fa 2.109b.
84 See HNYL 1185.
306
 Urban crime and urban security

formed a variety of tasks, were to be abolished. 8 5 Four months after

submitting his initial proposal H a n Wei again memorialized. It had
become clear that the new arrangement was overwhelming the remain-
ing officials. In his memorial H a n Wei implied that the various services
performed in the past by the servitors had now passed to others. T h u s ,
the dredging of canals had become the responsibility of the Metropolitan
Water Office (Tu-shui chien). Fire fighting and relief work had become
the responsibilities of the patroling inspectors' offices (hsiin-chien ssu). T h e
miscellaneous duties, once the responsibility of the servitors, had fallen
to the newly created magistrates, who were supposed to manage them in
accordance with the practices of the previous servitors. However, H a n
Wei said, these servitors had been responsible for conducting inquests
and reinquests, measuring boundaries, investigating criminal cases, and
distributing relief to the poor. They were, in short, functionaries with a
variety of duties not appropriate to the newly appointed capital and
court officials. H e asked that the servitors be appointed again, to handle
such miscellaneous duties, leaving the newly appointed magistrates to
concentrate on settling legal cases. T w o months later, after an official
noted that the number of cases to be settled was really quite small,
two of the positions were abolished, leaving two officials to handle the
magistrate's duties. 8 6
Borough servitors continued to be involved in relief work. A decree
dated 1070 said:

It has been snowy and cold in and around the capital city. There are those who
beg because they are old, ill, young, or solitary and have no sources of support.
The authorities of Kaifeng Prefecture are ordered to apportion management to
the four Bureaus of Fortunate Fields (Fu-t'ien yuan), who will note such people
outside the regular quotas, in order to give them food. As before, we expect the
administrative assistant servitors of the four boroughs, in accord with the old
rules of the Bureaus of Fortunate Fields, to examine this process on a daily basis
and, according to the amounts to be given to those within the quotas, to
distribute money and save lives.87

In the following year the man who followed H a n Wei as prefect of

Kaifeng was dissatisfied with the existing system and asked that the
position of right borough magistrate be abolished. His ostensible reason
was that the existence of two magistrates meant that people were being
sentenced differently for the same crimes, though the sources suggest

85 SHY, chih-kuan 37.ga-b.

86 HCP 2ii.i5a-b, 2i5-i3a-b, 225-7b-8b; SHY, chih-kuan 37.ga-b, cjb-ioa; SS
166.3942; WHTK 63.568. There are some minor contradictions in the information
given in these various sources.
87 SHY, chih-kuan 37.9b-10a.
307
 Law and order in Sung China

that the question had been affected by the bitter partisan politics of that
time. His idea was that the adjudicating functions be returned to the
prefectural authorities and that the remaining single borough magistrate
concern himself with the miscellaneous law-enforcement duties that had
in the past been handled by the servitors. His request was turned
down. 88
These changes had been prompted principally by concerns about
settling minor legal cases. The servitors who acted as patroling inspec-
tors had proved indispensable as a group, though there seem to have
been occasional changes in their numbers. The texts suggest that some
patroling inspector positions had been eliminated some time in the
1070s, since an order of 1081 called for the reestablishment of two
patroling inspectors in the four outer areas around the city.89 In 1086
this number was raised by dividing the jurisdictions within the new city
into left and right. The provisional prefect of Kaifeng requested the
following:
In the new and old cities of the capital, there are only the two boroughs. When
at night there are incidents, the distances to be covered are great. I ask that in
the new city the borough be divided into left and right, to make a total of four
boroughs. Select two additional civil officials for these posts, and increase the
number of clerical personnel.
His suggestions were approved. 90
Four years later the statesman Su Ch'e wrote:
At the beginning of the Yiian-yu period [c. 1086], when Han Wei had control of
the government, Hsieh Ching-wen was selected prefect of Kaifeng Prefecture.
Han Wei had previously been prefect of Kaifeng and had divided the city into
two boroughs for administration. Hsieh Ching-wen wanted to flatter Han Wei,
and so he asked that the city be divided into four boroughs. This policy was
useless, indeed pernicious, and the Court recently has abrogated it.91
Su Ch'e was alluding to the abolition of the four-borough system in
1089, D U t m IO 94 ^ w a s reestablished. 92 This change might have been
connected with an important, but undated, memorial by Pi Chung-yu.
Pi Chung-yu was a well-known scholar-official who served, among
other important posts, as judicial intendant in Ho-tung Circuit. In his
memorial he said:

88 HCP 225.7b-8b.
89 HCP 311.5a.
90 HCP 381.23a; see also SS 166.3942; SHY, chih-kuan 37.10a.
91 SS 166.3942; SHY, chih-kuan 37.ioa-b.
92 SHY, chih-kuan 37.ioa-b; SS 166.3942.

308
 Urban crime and urban security

I have noted that the regulations for controlling banditry in Kaifeng Prefecture
are very numerous, and yet major parts of the plans for stopping banditry are
incomplete. If, for example, we look at this in terms of capturing bandits in outer
circuits, we find that within the [outer prefectures'] cities themselves there are
military directors in chief [tu-chien] and supervisors of militia [chien-ya], whereas
outside the cities there are patroling inspectors and sheriffs. The arrangement of
these officials cannot easily be changed. Nowadays, outside the city [of Kaifeng],
the system of patroling inspectors and sheriffs in the outer prefectures is gener-
ally similar. Within the city, however, the duties of patroling inspectors are sub-
ordinate to the cavalry and infantry generals [shuai-ch'en] and the four borough
chiefs. Even if the emphasis is on patroling in the capital, in the search for
ordinary bandits in the past they have been ill prepared. Hitherto, under the old
precedents of Kaifeng, the servitors were ordered secretly to pursue and arrest
men. If they seize bandits, they will get rewards, but if they fail to do so, they
will not be punished. Men therefore take their duties lightly. On the contrary,
they often even have dealings with the bandits. The various borough servitors
minor, although they are supposedly responsible for investigating bandits, are in
reality involved in family welfare activities. They take legal depositions from
people who are ill, take part in inquests, help with fire fighting, and perform
similar miscellaneous duties. They are not really officers who capture bandits. 93

Pi Chung-yu recommended that six borough patroling inspectors be

appointed, two in the old city and four in the new city, with the Kaifeng
prefectural authorities having control. Thus, the servitors who served in
the boroughs would continue to be selected by the prefectural authorities
according to the pattern first established under Jen-tsung (r. 1022-63)
and reaffirmed during the Hsi-ning period (1068-77). This is confirmed
by an approved memorial dated 1099.94
Occasional grave concern about the state of urban security and law
enforcement was reflected in the authorities' willingness to devote large
amounts of their scarce resources to police needs. There were attempts
to improve the physical facilities available to the law-enforcement
apparatus. The boroughs, for example, had their own jails. Because of
an increase in the number of men being held, an official in 1082 said that
because "the left and right boroughs have many men held in detention
and their jails are overcrowded, we ask that more jail cells be added." 95
In 1116 the problem was the barracks for the men serving as police.
The prefect of Kaifeng wrote:

93 Pi Chung-yu, Hsi-t'ai chi, I . I O - I I (17b-18a). Military directors in chief and

supervisors of militia also seem to have had charge of law enforcement in other cities.
See Ou-yang Hsiu, Ou-yang Wen-chung kung cki, 12.16.
94 HCP 511.2a. Apparently at about this time the authorities also added investigators
(Van-shih jen) to the eight boroughs. Their number increased until eventually the total
reached sixteen men, but the posts were abolished in 1100; HCP 520.22a.
95 HCP 323.3a.

309
 Law and order in Sung China

In the four boroughs of the capital city that are under my control, there are
many wards. There wards all have a number of patroling soldiers, who dwell in
barracks. These barracks have been deteriorating for many years. Some are no
longer in existence, and so there is no way for the men to shelter from the wind
and rain or to avoid the cold or heat. The soldiers all live in other places. Often
they huddle beneath people's bamboo fences. . . . Bandits can act without any
fear.

As a result, a major rebuilding program was initiated. 96

Civilian law-enforcement functionaries

I n addition to these military functionaries, there were also some civilian

law-enforcement personnel in u r b a n areas. T h e capital city itself included
two districts, which like districts elsewhere, h a d civilian sheriffs (hsien-
wei) in c h a n g e of law enforcement. I n 1055 t h e authorities appointed a
second sheriff in Hsiang-fu District of Kaifeng Prefecture, a n d a second
one w a s a p p a r e n t l y also a p p o i n t e d in t h e other district of the capital. 9 7
A n order of 1081 reduced t h e c o m p l e m e n t to a single sheriff in each
district, because t w o patroling inspectors h a d j u s t been reestablished for
each q u a d r a n t of the s u b u r b s . 9 8 Cooperative action for law enforcement
in s u b u r b a n areas between t h e sheriffs, w h o h a d charge of the whole
district, a n d t h e b o r o u g h military, w h o h a d u r b a n responsibilities, w a s
also characteristic of other S u n g u r b a n i z e d a r e a s . 9 9
U r b a n districts, like districts in general, h a d a c o m p l e m e n t of archers.
T h e n u m b e r s of these constables varied from district to district, d e p e n d -
ing in p a r t o n t h e n u m b e r of households. F o r Kaifeng itself we know t h a t
Hsiang-fu District h a d sixty archers in 1081. 1 0 0 F o r a brief period t h e
n u m b e r of archers w a s lowered, u n d e r a policy that confined t h e sheriff's
concern to u r b a n i z e d areas, thus leaving t h e rural areas solely to their
patroling inspectors. I n t h e early 1080s in t h e capital, therefore, each
sheriff h a d only twenty constables. 1 0 1 Some years later, however, t h e
n u m b e r elsewhere w a s increased, in Kaifeng as in other regions. 1 0 2
Cities h a d their o w n sorts of drafted s u b b u r e a u c r a t i c functionaries,
which corresponded to t h e village officers of rural areas. D u r i n g t h e
T ' a n g period, p a r t of the b u r d e n of u r b a n law enforcement fell o n labor

96 SHY, ping 3.6a.

97 WHTK 63.573.
98 HCP 311.5a; SHY, chih-kuan 48.64b. See also HCP 331.13b-14a.
99 An approved memorial in 1031 ordered such cooperation in the suburbs of Ting
prefectural town in Ch'ing-hu North Circuit (SHY, ping 11.13b).
100 HCP3ii.5a-b.
101 HCP 311.5a; SHY, chih-kuan 48.64b.
102 HCP 364.2a (1086).

310
 Urban crime and urban security

service officials called ward chiefs (fang-cheng). Such men continued to be

drafted during the Sung but seem to have been concerned with tax
collection rather than law enforcement. In addition, in the districts that
included Kaifeng there were also the usual village officers charged with
maintaining law and order. Because these elders and stalwart men had
received no particular training or support, it is hardly surprising that
they are said to have been incapable of dealing with bandits. 103 An
official remarked in 1070 that even though there were elders, stalwart
men, and other village officers in the area around the capital, they were
in fact too weak to resist bandits, and even if they did sometimes succeed
in capturing them, this just caused the other members of the bandit
gangs to come together to aid their fellow bandits, causing untold
trouble. 104
In 1070 in the capital area, the new pao-chia militia system was
instituted. Although this device was most developed in rural areas, its
units also seem to have overlapped the suburban sprawl surrounding the
capital. A decree of 1084 states that in Kaifeng Prefecture there were at
that time 280 superior guards (tu-pao), with a total of 68,863 men. 105
Unfortunately, the materials do not usually distinguish the rural areas
from the city, and so it is difficult to determine how actively the urban
pao-chia participated at this time in law-enforcement activities.
In theory, all urban populations were to be part of some such sys-
tem, but in practice there were problems with solitary and unattached
individuals in the cities. This is made clear in an order in 1082 by the
prefectural authorities in Kaifeng:
Those prisoners who are old, young, ill or disabled, as well as those who are
alone and have no one to pay their commutations, are to be given surety and
released. This prefecture daily must judge cases. Of those eligible for commuta-
tion, many are solitary or poor. Also, they lack membership in neighborhood
guards. The borough patroling form should be used as the essential document to
serve as surety so that they may have a convenient guarantee document.106

Southern Sung
The patterns of urban law enforcement just described were the model for
the systems used during the Southern Sung. In 1132, when the refugee
Court had recently arrived in Lin-an, an official submitted a memorial
on the security problems of the new capital:

103 SHY, ping 2 .6b-7a (1070).

104 SHY, ping 2.6b~7a.
105 HCP 346.12a.
106 HCP 323.5b.

311
 Law and order in Sung China

Within the city and the district of Ch'ien-t'ang, there is a large area with many
bandits. Also, in the aftermath of military problems, there are many refugees.
Because they often are living in thatched huts, controlling fires is a very serious
problem. Although there are two men who are supposed to serve as the left and
right ward patroling inspectors, the system is defective. The men are appointed
in name only. I ask that you have the Bureau of Military Affairs [Shu-mi yuan]
order the Cavalry and Infantry Commands [Ma-pu chiin-ssu] to take responsi-
bility for drawing up laws for patroling inside and outside the city. Divide the
urban area of Ch'ien-t'ang into four wards, each having its own patroling
inspector and provisionally dispatching talented commanders [Chih-hui shih]
from the armies to fill these posts. In each borough, assess its area and decide on
a certain number of police posts, assigning six imperial army soldiers and a
commander. At night they will use a water clock to note the time and strike the
drums. They will prepare the fire equipment. For every two posts, appoint an
adjutant (chieh-chi), and for each thousand soldiers, assign a military officer to
have general control over them. When legal cases are brought to the patroling
inspectors, they are to be forwarded to the Lin-an prefectural authorities. As
before, a daily security report should be sent to the Cavalry and Infantry
Commands [Ma-pu chiin-ssu]. If in the area under their control there are
bandits, then the patroling inspector and his subordinates should be punished in
accordance with the law of the capital. 107

This proposal was approved. In this era Lin-an was organized accord-
ing to the pattern used previously in Kaifeng (see Figure 9.2). The post
of superior patroling inspector in the new city (outside the old walls)
being filled concurrently by the man who was the head of the Infantry
Office, and his colleague, the head of the Cavalry Office, served con-
currently as the superior patroling inspector of the old city. 108 The
pattern of four boroughs outside the city is indicated in a memorial
dated 1167, in which an official commented that because of the large size
of the boroughs, it was difficult for the sheriffs and patroling inspectors
to handle law enforcement adequately. He asked that another patroling
inspector headquarters be set up west of the city. 109 The Gazetteer of Lin-
an from the Hsien-ch'un Period lists nine boroughs for the city: the Palace
Borough, the Left No. 1 South Borough, the Left No. 1 North Borough,
the Left No. 2 Borough, the Left No. 3 Borough, and the Right Boroughs
Nos. 1 through 4. 110 As areas for law enforcement, these boroughs over-
lapped areas that were the responsibility of the men from the patroling
inspectors' forts. The Meng-liang lu says that there were sixteen patroling
inspectors' forts, with over thirteen hundred soldiers, but this passage

107 SHY, ping 3.7b-8a.

108 SHY, ping 3.9a.
109 SHY, ping 3.3a.
110 Ch'ien Yiieh-yu, Hsien-ch'un Lin-an chih, 19.4070-71.

312
 PLAN OF HANGCHOW (Lin-an)
AFTER THE RECONSTRUCTION BY A.CMQUIE
C. 1274

AtUrforihc .
.••: sadrices of t/>&
•southern suburb

Map 9.1. Plan of Hangchow (Lin-an). Redrawn from Arthur Christopher

Moule, Quinsai: With Other Notes on Marco Polo (Cambridge: Cambridge Univer-
sity Press, 1957).
313
 Law and order in Sung China

seems to include forts not in the city or suburbs but in their environs (see
Map 9.1).111
The central military and security authorities in the capital worked
together with the prefectural authorities in staffing this police system. In
1132 the Palace Cavalry and Infantry Command (Tien-ch'ien ma-pu
chiin ssu) asked that the left and right patroling inspectors work together
with the Lin-an Prefectural Office of the military director in chief (tu-
chien ssu) to evaluate the imperial army soldiers manning the 102 police
posts in the capital. 112 In fact, 115 police posts were set up in the capital,
with 673 men, the Three Commands, and the prefectural troops each
providing half of the complement.113 In 1152 this was raised to 150
posts, with more soldiers to be used for night patrols. 114 By 1171 there
were 232 posts, each with 4 dispatched soldiers and 1 post jailor (ya-pu).
An official writing at that time commented that the original order had
called for the prefecture, acting in concert with the Three Commands
(san-ya), to appoint the soldiers. He complained that the Three Com-
mands were not doing a satisfactory job and suggested that the prefec-
ture take sole charge of making these assignments from the imperial
army soldiers under its command. Two years later the Three Commands
and the prefecture again were given joint responsibility for making the
assignments. Although there were some occasional changes in policy
after this, the most common pattern was for these two offices to share the
responsibility.115
The prefectural authorities and the Three Commands, acting jointly,
selected the soldiers to be sent for tours of duty in these army patrols. 116
These tours of duty seem to have been one month long, though there was
a complaint in 1167 t n a t m e n were not being replaced. As a result, the
men became "thoroughly familiar with the streets and alleys, and fre-
quently they got into trouble. Those repeatedly arrested have been
numerous." The official asked that the superiors of these men be made
to provide sureties of joint responsibility for the behavior of their sub-
ordinates, that only well-behaved men with families be chosen, and that

111 Wu Tzu-mu, Meng-liang lu, in Meng Yiian-lao, Tung-ching meng-hua lu, wai ssu-chung,
189, 214-15.
112 SHY, ping3.8a-b.
113 HNYL, 51.900.
114 SHY, ping 3.9a-b. SS 166.3944 s a v s t n a t there were 148 posts, but the reference is not
dated.
115 SHY, ping 3.10b. An official had said: "I ask that in Lin-an Prefecture, in the alleys
and byways, you again increase the number of police posts and fire stations. There
should be a strict system of patroling. A clear system of rewards and punishments
must be established governing the capture of fugitives and fixing the time limits for
their capture, so that the capital can again be secure." SHY, ping 3.11b-12a.
116 SHY, ping 3.9b.

314
 Urban crime and urban security

they always be rotated at the end of their monthly tours of duty. The
exceptions to this rule were to be the areas east and west of the city,
where the men were to serve two-month tours. 117
The Southern Sung guidebook to Lin-an, the Meng-liang-lu, describes
the situation:
Lin-an city is very large and has a huge population. People's houses are like a
forest. Their eaves almost touch each other. There is not an inch of open space.
The canals and ditches are obstructed, and the streets narrow, being no more
than paths. Many are in poor repair. The authorities have established army
patroling posts about every two hundred paces, with a small number of soldiers
assigned to each post. At night they patrol to control robbery and fires. If there
are disturbances or other illegal activities, they make arrests, and then the
boroughs conduct investigations, turning over their prisoners to the prefectural
authorities for the hearing of the cases. There also are fire watches and people to
stand guard at official buildings, to protect them against bandits. Generally, the
authorities are very concerned about fires and, in the wards, have established
security buildings with a staff of regular soldiers. In the wards they have built
watch towers. From dawn to dusk they rotate, and troops keep watch. If they see
smoke anyplace, they use flags to indicate its location and name. At night they
use lanterns.118
As in the Northern Sung, the police were also involved in the city's
welfare system. A decree of 1164 on the distribution of rice to the needy
assigned responsibility to the prefectural authorities, working with the
borough officials, to investigate and certify that those being aided really
were poor. 119 And in 1181, in an order dispensing medicines and food
during an epidemic, some of the personnel used were from the borough
offices.120 Note that this use of borough personnel in welfare activities
was not confined to the capital, as documents indicate a similar practice
in the city of Shao-hsing at the beginning of the Southern Sung. 121
In addition to these regular army soldiers working as police within the
city, Lin-an also had substantial numbers of "local soldiers" {t'u-ping)
serving in forts in the city's environs. Some of the forts where these men
were stationed seem to have been some distance from Lin-an, but of the

117 SHY, ping 3.9b-1 ob.

118 Wu Tzu-mu, Meng-liang lu 10.214-15. Later in the dynasty the administration of
Lin-an became even more differentiated. The boroughs continued to be the law-
enforcement units in the capital (and presumably in other cities also). However, fire
control was increasingly administered through new units called precincts (yu). We
should add, however, that one passage in Chu Hsi, Chu wen-hung cheng-hsun, TSCC ed.,
p. 4, suggests that precinct officials might play a role in maintaining the law.
119 SHY, shih-huo 60.12b-13a.
120 SHY, shih-huo 58.14b.
121 SHY, shih-huo 5g.22b-23a, 60.8a.

315
 Law and order in Sung China
nine forts listed in the Gazetteer for Lin-anfor the Hsien-ch'un Period (Hsien-
ch'un Lin-an chih), five were located at the city gates. Two of these forts
had 120 men; two had ioo; and one had 68. 122 Such men, even if not
spending their time as police, formed part of the security system, as in
their own way did the regular army soldiers stationed in the four major
military camps at the capital.
The authorities thus relied heavily on soldiers for police work. They
also used other sorts of people, even ex-criminals. In his book of remi-
niscences about Lin-an, the Wu-lin chiu-shih, Chou Mi remarks that
among the law-enforcement personnel used by the borough authorities,
many of those who were bravest and had the most fearsome reputations
"originally came from bandit gangs." 123
Perhaps because of its huge population, the capital also faced special
problems in adjudicating legal cases. The rural system in which the
district magistrate was the customary judge of first instance was simply
not workable in a city of perhaps a million people. Therefore, during the
Southern Sung as in the Northern Sung, additional judges had to be
appointed. The Sung History stated that the men appointed as borough
officials to hear people's complaints were to be selected from among
capital and court officials "who were close to the people" and possessed
the appropriate degree of seniority. 124 The Southern Sung encyclopedia
Shih-wu chi-yuan says that such men could deal with cases calling for
punishment with the light rod. 125 A decade later, in 1141, the Northern
and Southern boroughs outside the city were divided into left and right,
with personnel selected according to the previously used criteria. They
were given the power to hear cases calling for sixty blows or fewer of the
heavy rod. 126 In 1156 a censor stated:
Lin-an Prefecture initially followed the precedent of Kaifeng Prefecture. Outside
the city wall it established a Northern and a Southern borough for administra-
tion. Recently, within the city, we added Left and Right boroughs and two
officials, out of a desire to reduce the caseload of court cases in the city. This
prefecture each day has seventeen or eighteen law cases. All are sent for judg-
ment to the two boroughs. The clerks in the boroughs play with the facts and
distort the law. They pursue men without reason, in a wholly improper way
bringing them in. Lawsuits are numerous. Those the borough officials cannot
handle simply do not get judged. The prefecture does not decide the cases, and
the two boroughs do not handle them, so that the people are disturbed. As a

122 Ch'ien Yiieh-yu, Hsien-ch'un Lin-an chih, 57.4403.

123 Chou Mi, Wu-lin chiu-shih, 6.445.
124 SS 166.3944.
125 Kao Ch'eng, Shih-wu chi-yuan, 6.30a.
126 Ch'ien Yiieh-yu, Hsien-ch'un Lin-an chih, 19.4071.

316
 Urban crime and urban security

result, the situation in Lin-an Prefecture is not so simple as was the situation in
Kaifeng of old. This prefecture has many officials, but at present the offices in
the Northern and Southern boroughs are unfilled. I ask that two officials be
taken from the Left and Right boroughs within the city and transferred to the
Northern and Southern boroughs outside the city. Lawsuits within the city
should be sent to the prefectural authorities for judgment, according to the
former practice.
The suggestion was approved. 127
In the Lin-an area the pao-chia militia may have continued to perform
some of their old security functions during the first new decades of the
Southern Sung. In 1162 an order was issued provisionally abolishing
their night patrols in Lin-an Prefecture. They were to be replaced by
patrols of soldiers. 128 Two years later, however, the prefectural auth-
orities submitted the following memorial:
There recently was an order abolishing this prefecture's pao-chia guard organiza-
tions. This year the Great Rites will be performed. However, except for places in
the city with separate guild organizations [hang], in the streets and neighbor-
hoods of the northern and southern sections of the city there are no means for
investigating evil robbers. We ask that in this prefecture, outside the city proper,
you provisionally use the pao-chia guard organization so that they can patrol and
investigate [in the suburbs].129
Less than a month later the prefectural authorities submitted another
memorial, asking that the system be extended to cover the city itself:
Within the city there is a multitude of people. We desire that the old system be
provisionally revived. The people of the city should be organized into guards,
with ten families forming one small guard and one hundred families making up a
large guard, with the post of large guard head to be filled in rotation. Each night,
in rotation, one of the small guards should patrol. If they see within their area a
household in which there has been a crime, or if at night they meet men who are
moving about, they should interrogate them, enter their names in order on a list,
and give this list to the large guard head of the section where they live, for
controlling.130
Both memorials were approved, so we can probably assume that for at
least some time after this the pao-chia played an active neighborhood
watch role in maintaining urban security.
Other mutual security units were organized in the Sung. A report
from 1127 mentions patroling societies (hsiin-she). The organizational

127 Ibid. See also HNYL 175.2891. For a general description, see SS 166.3944.
128 SHY, ping 3.9b; see also SHY, ping 2.420-433.
129 SHY, ping 2.433-430.
130 SHY, ping 2.43b.

317
 Law and order in Sung China

scheme of these societies resembled that of the pao-chia, though the

societies had five levels of organization rather than three. Ten men
formed one unit, with a unit chief (chia-chang). Five units formed a
squad, with a squad chief (tui-chang);fivesquads formed a section, with a
section chief (pu-chang); five sections in turn formed a society, with
a chief and an assistant chief (she-fu-chang); and five societies formed
a group (tu) headed by two officers. The members guaranteed one
another's behavior and were charged with investigating banditry. 131
Though the original societies seem to have been organized in rural areas,
the report says that "all patroling societies organized among urban
people should follow the pattern of the rural societies."
Finally, the memorial from 1164 just quoted provides a tantalizing
hint that the city's guilds (hang) helped maintain law and order. 132 In
the T'ang these guilds had been sections of wards in which the practi-
tioners of a particular craft or dealers in a particular commodity were
concentrated. By the Sung period, with the breakdown of the controlled,
walled-ward urban pattern, the guilds had become organizations of those
engaged in a craft or in selling a commodity, regardless of the location in
the city of their place of business. The Sung government used these
organizations to collect certain taxes. In later dynasties such guilds did
exercise discipline over their members. The brief passage that recom-
mends reestablishing pao-chia units in the city - because no other organ-
izations for law enforcement existed "except for those places in the city
where there are separate organizations of guilds" - provides a rare
indication that this later pattern already existed in the Sung.
Sheriffs and archers continued to have a role in urban law enforce-
ment during the Southern Sung, though they were overshadowed by the
patroling inspectors. Some idea of their relative roles can be deduced
from an official's complaint dated 1169 that says that the districts of Lin-
an had a total quota of 1366 local soldiers for service under the patroling
inspectors (though there were in fact at the time only 911 in actual
service). The quotas for archers in these districts was slightly more than
half the quota for local soldiers. There were supposed to be 782 archers,
though there were currently only 737. 133 There sometimes were not
enough archers. The well-known official Fan Ch'eng-ta reported in 1170
about an incident in which the archers of Yu-ch'ien District (partly in
urban Lin-an) had tried to seize some salt smugglers, but because their
forces were inadequate, they had been wounded or killed. 134

131 SHY, ping 2.503-51 a.

132 SHY, ping 2.433-43!).
133 SHY, ping 3.27a.
134 SHY, ping 3.27a-b.

318
 Urban crime and urban security

It is important, however, to see these sheriffs and archers as parts of a

larger security apparatus, in which they were closely tied to the military
side of security work. It is certainly not accidental that the sheriffs'
offices of the two urban districts of Ch'ien-t'ang and Jen-ho were adjacent
to the large East and West military camps. (There were four major
military camps in Lin-an, the North, South, East, and West camps.)
Furthermore, the superior patroling inspector's office for the city's
Western Borough bounded the office of the sheriff of Ch'ien-t'ang District
to the east and north. This complex of agencies was further integrated by
assigning legal cases arising in the East and South military camps to the
city's Southern Borough offices, and those arising in the West and North
camps to the city's Northern Borough offices.135

Conclusion
Over several centuries there were striking changes in the patterns of
Chinese urban life. During the late T'ang and Five Dynasties, the
character of Chinese cities was radically transformed. Ch'ang-an, the
prototypical city of the T'ang, was laid out on a grid pattern, divided
into walled wards, and subjected to strict curfew and other controls. It
was the model of a tightly controlled urban area. However, in the
late T'ang and thereafter these patterns began to change. The system
of walled wards broke down. Individual houses now opened onto the
streets. Curfew restrictions were relaxed. The state no longer attempted
to control so strictly the locations of business activities. The attempts to
regulate prices of goods were less and less effective. As city life became
increasingly free from state interference, it seems that the security prob-
lems would have changed and that as the security problems changed, the
state would have had to change the ways in which it tried to deal with
urban crime, fire fighting, adjudication, and related problems.
We must remember, however, that these changes took place piecemeal,
over many decades. From the point of view of those alive at the time,
they must have seemed quite gradual. Because the transformation came
about in this way, the state responded by making a sequence of relatively
minor modifications in its inherited repertory of policies for urban
security and control. A large number of men were still needed to staff the
security positions, and it was still true that the military was the obvious
source for such men. The Sung understandably continued to depend on
men from its military forces to enforce order in its cities. These were
balanced to some degree by civil officials and constables, but in the

135 Ch'ien Yiieh-yu, Hsien-ch'un Lin-an chih, 19.4072.

319
 Law and order in Sung China

major cities the emphasis certainly was on the military. The personnel of
the law-enforcement system thus did not change radically from previous
patterns.
With the growth of great cities and the opening up of closed neighbor-
hoods, a regular underworld also formed. I think that we may also
assume, judging from the evidence of contemporary developments, that
"one cause of an increase in criminality is the increased urbanization of
the population." 136 Associated with the urbanization during this era was
a general rise in prosperity. But affluence can contribute to crime as
well. In the words of Wilson and Herrnstein:
In the individualistic environment of the cites that grow along with a nation's
economic development there are targets for theft in some rough proportion to the
general level of affluence. .. . But this may not be the whole story. The suggestion
has also been made that it is not only affluence that fosters crime, it is relative
deprivation, o r . . . inequity. Wealth tends to be accumulate unequally, and to
those not possessing it, it may seem (and may in some sense be) inequitable as
well as unequal. The contrast between the haves and the have nots becomes
more, not less, palpable . . . as affluence grows, if people are separated by wider
gaps in wealth, especially if they live side by side in cities. This is even more so if
the family and community ties of traditional society have been dissolved in
the impersonal exchanges of commercial society, and wealth has become the
measure of individual worth.137
Wilson and Herrnstein are discussing the rise in crime in recent times,
but all that they say also is relevant to the Sung. It may well be that the
urban society of the Sung was not so individualistic and fragmented as
urban society is today; it is certainly true that Sung urban society was
more individualistic and fragmented than Chinese society as a whole had
been in earlier times.
The problems facing law-enforcement officers also were much more
difficult, given the freedom with which people could move around Sung
cities. The state reacted to these changes by increasing the size of
the groups concerned with security maintenance and by introducing a
certain amount of specialization in function and in territorial units. Still,
the authorities acted rationally, not by radically altering their urban
policies, but by gradually adapting old ways to new conditions.

136 Wilson and Herrnstein, Crime and Human Nature, p. 411, speaking of increases in crime
rates in the modern world.
137 Ibid., p. 446.

320
 10
The Sung penal system

Introduction — functions of penalties

The product of an effectively functioning law-enforcement system is
captured suspected criminals, but this product is merely a means to a
further end, the reduction to a tolerable level of socially threatening
deviance. Insofar as the presence of law-enforcement agents actually
discourages the commission of crimes, it directly contributes to the larger
goal. The captured criminals pose a different problem: Assuming that
they are found guilty, how should they then be treated? What sort of
treatment will contribute most to the general aim of creating a society
not threatened by serious deviance? More exactly, given the array of
options that the ruling group can envision as both possible and accept-
able, and within the constraints set by organizational skills, technology,
fiscal limitations, and competing state goals, what responses will be most
useful? For the majority of those captured and convicted, this comes
down to the question of how they are to be punished. The answers given
in a specific culture are bounded in part by contemporaries' visions of
their own past history; in trying to understand their own time they look
at the evolution from their cultural past to the present.
The levels of punishment exacted in traditional China reflected a
vision of the past and how the current practices evolved, the ideas of
state leaders about the uses of penalties, and a perception of the current
historical situation. Penalties are seen as useful in both internal and
external terms. Internally, they might create a more stable and peaceful
society. By reflection the resulting good order helps in the competition
with external enemies. A Sung official quoted with approval a passage
from Mencius: "If a ruler . . . takes advantage of times of peace to explain
the laws to the people then even large states will certainly stand in awe
of him." 1
1 Yuan Hsieh, Chieh-chai chi, 3.33. For the passage from Mencius, see Lau, Mencius, p. 81.

321
 Law and order in Sung China

Punishments also had to be adjusted to the conditions of the times. In

the Sung, many officials accepted the traditional wisdom, that heavy
penalties had to be used in times of disorder and that lighter penalties
were appropriate in times of peace. In times of good order, excessively
harsh penalties brutalize the people; in times of disorder, the use of
inappropriately light penalties would not suffice to suppress lawbreaking.
The eleventh-century official Wen Yen-po expressed a widely shared
view when he said that heavier punishments were fitting in times
of troubles but in times of peace would actually promote disorder. 2
Adherents of this view argued for appropriate measures by pointing to
historical examples of rulers who persisted in using heavy punishments
without in fact reducing disorder. 3 Some of these men went further and
attempted to analyze how this situation had come to pass. In the words
of Wen Yen-po:
The system of penalties has a history. Thus the systems transmitted by the
successive dynasties have differed. Minor penalties are used to restrain deprav-
ity; serious penalties to suppress disorder. I have investigated the situation of
long ago and found that it was not then as it is now. In high antiquity people
were by nature simple and pure. When those above made manifest a desire for
peaceful transformation, their inferiors were not contentious. . . . The government
was extremely simple. The people were naturally orderly. The penalties were
extremely light, yet men did not transgress. After this the Sages, basing them-
selves on Heaven's punishing of the guilty, established the Five Punishments.
Above this reached to the use of military force. Below it reached to beatings and
whippings. Thus the five degrees of severity were clarified. This is the origin of
the penalties.4
Thus, to be appropriate, penalties had to fit with the nature of the
people and their times. Simple people need simple laws. We should note,
however, that there were a minority of officials who argued to the
contrary. Whatever might have been the case among the ancients, in
more recent times the people were anything but simple and pure. Good
order, these critics maintained, existed when the penalties were severe
and the laws detailed. Disorder marked eras in which the laws were too
lenient and not sufficiently detailed to guide behavior. 5
Both the advocates and the critics of strict laws had a vision of how
this early system had deteriorated in later times. Some discussed the
causes for this deterioration. The eleventh-century official Liu Chih, in a
memorial concerned with the recompilation of the laws, wrote:

2 WHTK 167.1449; HCP 2 i 7 . 9 a - b .

3 Fan Tsu-yii, Fan t'ai shih chi, 22.11 a.
4 Wen Yen-po, Lu-kung wen-chi, SKGSCP 1976, vol. 295, 9.8b.
5 See, for example, the arguments of Tseng Pu. Also HCP 214.19b.

322
 The Sung penal system

The laws [fa] are the great Commandments [ming] of all under Heaven. When
the [exemplary] Former Kings [of the early period] systematized the laws, their
intention was to cause men to change themselves and avoid penalties and to
make crime difficult. Therefore the laws were extremely simple and extremely
correct, and yet they were adequate to fulfill the principles of Heaven and Earth.
When later times systematized the laws, they were merely fearful that the guilty
might sometimes escape. Therefore their laws were numerous, extensive, and
closely meshed, so that the people did not know where they might put their
hands and feet.6

The determination of the state authorities not to let anyone who was
guilty escape the net of the law led paradoxically to a situation in which
many ran afoul of it. The government itself contributed to the people's
loss of a state of simple purity. By enacting a bewildering multitude of
laws, it guaranteed a steady supply of lawbreakers. The presence of these
lawbreakers then faced the authorities with the question of how they
should be treated. What in fact were the basic goals of punishments?
The system of penalties was a partial answer to the question, what are
we to gain from inflicting punishments on wrongdoers? The articulate
part of the population in traditional China responded with answers that
are common elsewhere. Deterrence, retribution, social defense, and
rehabilitation all were viewed in one way or another and to one degree
or another as goals in the Chinese conception of punishment.

Deterrence
The criminologists Wilson and Herrnstein remark that "there is . . . good
though not yet entirely conclusive evidence that the behavior of offenders
can be altered by altering the consequences attached to their behavior." 7
One way to alter such consequences is to link the behavior with negative
sanctions. Like most people, the Chinese believed in the deterrent func-
tion of punishment. If men cannot be socialized to be good, then we
must make them fear to do evil. The idea that punishments deter is so
widespread in China and so casually accepted that it is rarely discussed
in any detail. The basic premise, however, is implied in literary sources
that most traditional Chinese scholars assumed stemmed from the most
ancient times. The section called the "Punishments of Lu" from the Book
of Historical Documents opens by remarking that "when the king had
enjoyed the throne till he was of the age of a hundred years, he gave
great consideration to the appointment of punishments, in order to

6 Liu Chih, Chung-su chi, 6.81. See also HGP 373.2a.

7 Wilson and Herrnstein, Crime and Human Nature, p. 396.

323
 Law and order in Sung China

restrain the people of the four quarters." 8 This attitude is echoed later in
various writings. The legalist text called The Book of Lord Shang simply
states, in words echoed by the legalist Han Fei Tzu, that "the nature of
men is to like titles and emoluments and to dislike punishments and
penalties." 9
It is hardly surprising to find these sentiments occurring repeatedly in
the writings of Sung authors. At its simplest, this conviction reflected the
belief that "if the laws are severe, there will be few bandits; if the laws
are lenient, there will be many bandits" 10 or that "punishments are used
to end the use of punishments; the death penalty is used to frighten
men." 11 This simple view was a minority view. Most Sung officials had a
more sophisticated understanding of the relationship of the weight of
penalties to the maintenance of good order, believing, again, that dif-
ferent sorts of eras required different levels of penalties. In either view,
the emphasis was on deterrence and was coupled with an underlying
assumption that the greatest effect would be achieved when the severity
of the penalty corresponded as closely as possible to the seriousness of
the crime.
During the middle of the twelfth century Emperor Kao-tsung (r.
1127-62) spoke for deterrence and appropriateness when he debated the
current pattern of reducing most death sentences to penal registration.
He said that he desired that the penalties fit the crimes and that if "we
act only in a lenient manner, ignoring the law and using precedents [to
reduce the penalties], then men will not be fearful. This is no way to use
penalties to restrain evildoers."12 Despite the emperor's arguments, how-
ever, the practice of commuting most death sentences continued, in part
because the influential Sung officials felt that the empire, despite its
external troubles, was internally relatively peaceful.
The men who argued against the emperor were also committed to
appropriate penalties and to deterrence. However, they differed in their
understanding of what was appropriate. The official Hsueh Ghi-hsiian,
after having commented on the rise of the punishments of penal servi-
tude and exile (in place of death sentences), noted that these punish-
ments "served as warnings to draw people to the good and get rid of

8 Legge, The Shoo King, p. 588. What is important here is that later Chinese believed that
the "Punishments of Lu" was an early and authentic document. H. G. Creel argues
that the document in fact was a late forgery, but Sung writers certainly did not think so.
See H. G. Creel, The Origins of Statecraft in China (Chicago: University of Chicago Press,
1970), p. 161.
9 Duyvendak, The Book of Lord Shang, p. 241; Liao, Works of Han Fei Tzu, vol. 1, p. 47.
10 Fu Ch'a, Chung-su chi, 6.74-75; HCP 214.20a.
11 HCP 43.1 i a - b .
12 HNYL 2972; see also HNYL 2944.

324
 The Sung penal system
13
evil." Deterrence was of central importance, but to be effective the
deterrent penalties had to fit the offense committed. Therefore, in
peaceful times, sparing men execution and sentencing them to the lighter
penalties was appropriate.
Indeed, the key to the debate was the concept of appropriateness, not
severity. Kao-tsung was in no sense advocating indiscriminately harsh
penalties. He simply felt that death was a just penalty for those who had
committed homicide, and anything less was not. He himself recognized
that the problem was not one of simple severity. The emperor, like many
astute observers on penal systems, recognized that penalties felt to be
excessively harsh often would not be enforced.14 On the other hand,
officials were mindful of the Chinese tradition of mercy in punishing. In
the Northern Sung the official Fan Tsu-yii cited the examples of both
Confucius and the paragon of emperors, T'ang T'ai-tsung (r. 627-50),
in opposing harsh punishments. Fan noted that when the ruler Chi
Kang asked Confucius about "killing the unprincipled for the sake of the
principled," Confucius replied:
Sire, in governing, why should you use killing at all? Let your evinced desires be
for what is good, and the people will be good. The relations between superiors
and inferiors is like that between the wind and the grass. The grass must bend
when the wind blows across it.
Fan went on to note that when T'ang T'ai-tsung had first ascended the
throne, the officials who discussed suppressing banditry asked that harsh
penalties be established. "T'ai-tsung sighed, saying, 'People become
robbers because taxes are high and the demand to provide public ser-
vices is onerous and because officials are corrupt. . . . If the people have
sufficient food and clothing, they will not be robbers. Of what use are
heavy penalties?" 15 Men who insisted on harsh penalties risked brutaliz-
ing the people to the point that they would no longer be awed by
severity.16

Retribution, rehabilitation, and social defense

Deterrence was the key; retribution, rehabilitation, and social defense
were minor but often present themes. The men of Sung China believed
in punishment as a form of retribution. Wronged men demanded ven-
geance. Arguing against the issuing of an amnesty and after listing a

13 Hsiieh Chi-hsiian, Lang-yu chi, 16.23a-24a.

14 Liu Cheng, Huang Sung chung-hsing liang-ch'ao sheng-cheng, 22.20a.
15 Fan Tsu-yii, Fan-t'ai shih chi, 22.11a.
16 Wen Yen-po, Lu-kung wen-chi, 9.9a.

325
 Law and order in Sung China

variety of serious crimes, an official pointed out that if the state would
"requite these crimes with the regular penalties, then the anger of the
people would be appeased." 17 By punishing the guilty itself, the state
lessened the temptation of victims to seek private vengeance. State
punishment not only appeased the living, it also might appease the dead.
This is most apparent in cases of homicide. The state had to try to
placate the angry spirits of the dead so that they would not plague the
innocent living. Emperor Chen-tsung (r. 997-1022) was very concerned
about the numerous bodies found in the Pien Canal that flowed through
the capital. Some, he was sure, were those of murder victims. To soothe
their ghosts, he ordered the Buddhist and Taoist temples in Ssu Prefec-
ture, where the canal entered the Yangtze River and where many of the
bodies first came to notice, to pray for the dead. 18 This idea of averting
the anger of the spirits, as well as the belief that good order needed to be
reestablished by a reversal of fortunes, seems to have been connected
with the retributive use of the lex talionis. In arguing against lenient
treatment of criminals, writers repeatedly said that "those who kill men
should die; those who injure men should themselves be injured," 19 and
again in the early Southern Sung Emperor Kao-tsung wondered, "How
can it be the intent of the law that a killer not suffer death?" 20
The rehabilitation of wrongdoers was also a goal of punishment. Like
most members of the articulate part of the Chinese elite, Sung writers
were deeply committed to a voluntarist view of individual psychology.
This attitude is implicit in the position of Ch'en Liang, that Sung
penalties were "designed to instruct and based on benevolence." 21
Penalties were lessons designed to instruct and inform. Such a view
necessarily implies that those so informed are educable, that they can
learn and be transformed as a result. Wrongdoing was voluntary be-
havior under the control of the culprit. One implication of this was
optimistic. Wrongdoers had the capacity to reform themselves - in
the Chinese phrase that appears over and over in Chinese amnesty
documents - they could "renew themselves" (tzu-hsin).
Some Sung officials saw no conflicts or even tensions among these
various functions of punishments. Su Sung (1020-1101), after advocat-
ing the revival of the T'ang penalty of exile with added labor, remarked
that his system would not only punish the offender but would also "be a
sufficient warning" to others. He went on to say that his reforms "would

17 Huang Huai, Li-tai ming-ch'en tsou-i, n 8 . n b - i 2 b .

18 HGP 79.12a.
19 Compare HCP i24-3b-4a.
20 HNYL 1637.
21 Ch'en Liang, Lung-ch'uan wen-chi, TSCC ed., 27.327.

326
 The Sung penal system

suffice to open up the road to self-renewal, to transform the evil into

good people, and to change the stupid and vulgar through education." 22
Individuals were in control of their own behavior, but to the tradi-
tional Chinese, individuals were inextricably interwoven with and indeed
defined by their relationships with others. That being so, individuals
necessarily were sensitive to the reactions and attitudes of others. A
modern sociologist concluded that among the population at large, the
fear of informal sanctions, specifically the loss of reputation, is a more
powerful deterrent than is the fear of formal sanctions.23 In traditional
China, as in all traditional peasant societies with low horizonal mobility
and great social and economic interdependence, the social group on
which the individual depended for livelihood and emotional support
could exercise enormous pressure for conformity. Such pressure was
especially great in the family. This was true of all premodern peasant
societies, but perhaps nowhere more true than in traditional China.
Therefore, even when the individual himself was not overly concerned
about his reputation, he could be pressured indirectly by actions against
the family. The idea that men were subject to change through social
pressure was also embodied in practices that shamed their families.
During the late eleventh century the Sung authorities used a system
under which

after a criminal has been judged, a sign is to be nailed on the gate of his family,
describing the crime and the punishment. If such a criminal, liable for penal
servitude or worse, gives information that leads to the capture of two thieves
liable for penal servitude or exile, or of one robber who has used force, or, if a
criminal liable for a beating with the heavy rod gives information that leads to
the capture of one thief liable for penal servitude or exile, his family may be
spared the posting of the notice.24
Like the scarlet letter in Puritan New England, such a sign was both a
part of the punishment and a reminder to others of the cost of wrongdoing.
Social defense - the use of penalties not to deter or to reform the
criminal or to avenge the injured, but simply to protect society against
further abuses - was also part of Sung penalties, but its presence is
shown in Sung practice rather than in theory. Like other Chinese dyn-
asties, the Sung regularly used a system of fettering dangerous criminals,
which was both a punishment in itself and a device to prevent them from
committing more criminal acts. In addition, criminals in the Sung were

22 Su Sung, Su-wei wen-hung chi, 18.1 i b .

23 C. R. Tittle, Sanctions and Social Deviance: The Question of Deterrence (New York: Praeger,
1980).
24 HCP 398.12a-13b.

327
 Law and order in Sung China

often tattooed, again as both a measure for immediate security and a

warning to those with whom they came into contact about their charac-
ter. Social defense is also explicit in the suggestions of some Sung
officials that hardened criminals not benefit from the liberal amnesty
practices of the Sung, so that "evil people would remain government
prisoners for years and could not harm the good people." 25 Wilson and
Herrnstein call such measures "incapacitation." They note that "as long
as legal punishment physically separates the convicted offender from the
public, it may protect the public, though perhaps not fellow offenders. . . .
In its pure form, incapacitation works, if it works at all, simply by
removing opportunities for crime." 26

Penalties in early China

A similar mixture of goals is present in the penalty systems of many
times and places. The specific institutional form of the punishments
designed to reach these goals is determined by different emphases on the
different goals and by ideas, ideals, and technology. In early China the
authorities speak about the desirability of reforming wrongdoers, but in
practice they aimed at deterring crime by making examples of criminals
so that their suffering could serve as a message to others and prevent
further abuse by the criminals themselves. "Legal punishment as public
spectacle strengthens noncrime by publicizing the risks of apprehension
and convictions, and it weakens crime by inculcating disapproval." 27 To
be effective as warnings, punishments had to be obvious, public, and of
appropriate severity.
Mutilating penalties provided vivid reminders of the costs of wrong-
doing. The earliest sources describe somewhat different forms of corporal
punishment. In the first Chinese records, the oracle bones of the Shang
dynasty (traditional dates 1750-1122), the eminent historian Hu Hou-
hsiian found evidence that the authorities often punished by cutting off
limbs. One graph appearing in several forms shows a man with one leg
shorter than the other, with the element meaning "saw" next to the
shorter leg. The written evidence is supported by archeological evidence,
the bones of a man whose leg had been amputated, a servant or slave
perhaps, buried with his master. 28
In slightly later times, ancient Chinese states employed a variety of

25 HCP 214.17b-18a.
26 Wilson and Herrnstein, Crime and Human Nature, p. 493.
27 Ibid., p. 495.
28 Hu Hou-hsuan, "Yin-tai te yiieh-hsing" (The penalty of amputation in the Yin
dynasty), K'ao-ku, no. 2 (1973): 103.
328
 The Sung penal system
gruesome penalties. Many sorts seem to have executed the victims by
slicing off their limbs. At the most extreme were penalties that seem to
call for first subjecting the victim to a variety of mutilations, including
tattooing, slicing off the nose, cutting off the limbs, and ending with the
exposure of the head, and making the flesh and bones into a kind of meat
preserve. This penalty was used against the highest ranks of society. 29
This belief in using mutilating penalties, implicit in much of the early
Chou Book of Historical Documents, is stated most clearly in the Discourses of
the States, written many centuries later: "The great punishments use
armor and soldiers; the next lower level uses the ax (for executions). The
medium punishments use knives and saws (for amputation); their next
lower level uses awls and presses. The minor punishments use whips and
rods, so as to warn the people." 30
This passage from the Discourses reveals the corporal nature of much
early punishment. From war to death to amputation to various tortures
to beatings, there are degrees of violence, but all aim at humiliating the
body. Having failed to dissuade men from committing crimes, we must
now make it more difficult for the culprit to commit more violations in
the future and hope that the punishment may also deter others. This was
surely part of the rationale for the mutilating punishments inflicted in
ancient China.
These penalties are intense, irreversible, and brief: When the head
falls in the basket, the execution is over. The man whose feet are
amputated is footless as soon as the saw has done its work, the noseless
man when the knife has finished. Only beatings occupy an intermediate
position, of the body, intense in their violence, but not necessarily
obviously and permanently disfiguring. All these penalties share a com-
mon trait. They do not involve the political powers in maintaining
physical control over their victims for any extended periods of time.
In the section of the Book of Historical Documents called the "Punish-
ments of Lu," the Five Punishments supposedly overused by the non-
Chinese Miao people are listed as death, cutting off the nose, cutting
off the ears, castration, and tattooing. The Five Punishments used in
the Chou period were death, castration, cutting off the feet, cutting off
the nose, and tattooing. 31 In all cases the results were irremediable. The

29 Tu Cheng-sheng, "Ku-tai hsing-yu tsa-k'ao" (Studies on the punishments of antiquity),

Chung-kuo shih hsin-lun, August 15, 1985, pp. 21-46.
30 Wei Chao, Kuoyii, SPTK ed., Lu Yu-shang.
31 For a description and comparison of the Five Penalties, with an analysis, see Uchida
Tomoo, "Sho-sho Ryo-kei no go-kei ni tsuite" (On the Five Penalties in the
"Punishments of Lu" of the Book of Historical Documents), Tohogakkai soritsu nijugo shunan
kinen tohogaku ronshu (1972), pp. 1 1 1 - 3 0 .

329
 Law and order in Sung China

body was central. Reverence for the spirits of the ancestors reinforced the
conviction that the body should not be mutilated. These punishments
therefore form the most striking demonstrations of the power of the ruler,
sending the loudest and clearest message to both men and spirits. They
were of special force in a society that believed in the continued life of the
spirit after death but that also believed that a mutilated body meant a
mutilated spirit.
Chinese practice seems to have been much more brutal than Chinese
theory. In the most important Chinese myth about the history of penal-
ties, the five classical mutilating punishments are said (in the Book of
Historical Documents) to have been used by the Sage Emperors 32 but were
supposedly applied only with great caution. The excessive and exclusive
use of such cruel punishments is blamed on the non-Chinese Miao
people, who in the end were destroyed because of their injustice.33 Thus
the Sages were always chary of mutilations. Although the penalties they
did inflict were often mutilations, the Sages are reputed to have always
sought circumstances that would justify the substitution of less brutal
punishments. They had a system of redemptions for cases with mitigat-
ing or doubtful circumstances and are also said to have used punish-
ments based on restricting liberty, the so-called Five Exiles.34

Rise of labor penalties

The Book of Historical Documents stresses the Sages' reluctance to apply
corporal punishments: such punishments were in fact common during
the Chou period (1027-256). Punishments calling for loss of liberty
combined with labor became common only late in the dynasty. Although
slave labor seems to have been used from very early on, few criminals
seem to have been punished by being confined and used as laborers
during the Western Chou and the early part of the Eastern Chou. But by
the late Eastern Chou (771—256), rulers were exploiting the labor power
of convicts.35
This momentous transformation is described in more detail in Chapter
12. In brief, during the late Chou the rulers of the modernizing states
found it technically possible and economically attractive to use convict

32 Legge, The Shoo King, p. 38.

33 Ibid., p. 591.
34 Uchida, "Sho-sho," pp. 111-30.
35 Li Li argues that penal labor was used in the early Spring and Autumn. The evidence
seems ambiguous, but the author does bring forward considerable evidence that labor
penalties were being used to some extent late in the Chou dynasty. See Li Li,
"Chung-kuo ku-tai t'u-hsing" (The penalty of penal servitude in ancient China) in
Chung-kuo hsiieh-tao pao 1 (1987): 3 7 - 3 8 .

330
 The Sung penal system

labor for certain tasks. The origins of this transformation thus predate
the unification of the empire by the state of Ch'in in 221 B.C. The
postunification Gh'in seems to have been a transitional period between a
greater dependence on mutilations as penalties and a system in which
the labor value of convicts dictated the wisdom of keeping them whole.
In addition to their distinctive clothing and shaved heads (shaving the
head being itself a mutilation), Ch'in convicts might be tattooed, have
their noses cut off, or have one or both feet amputated. 36
Although mutilations continued to be practiced under the early Han,
Emperor Wen-ti (r. 179-157) replaced the cutting off of the nose and
feet with beatings and the wearing of foot irons. Convict labor was
useful, and the authorities sought to avoid punishments that might
lessen its value. Tattooing, a mutilation but not one that diminished a
convict's labor value, was used during some of the dynasties that fol-
lowed the Han, but mutilating penalties as a whole were less in evidence.

Mutilating punishments during the Sung

From the Han onward, mutilating punishments were used rather
sparingly, though officials, including Sung officials, occasionally sug-
gested reviving them as a way of cutting down the incidence of crime.
They appealed to some as a solution to the law-and-order problem,
whereas others saw them as a workable and merciful alternative to
current practices. In recommending the revival of mutilating punish-
ments in 1070, the Sung official Tseng Pu placed them in the context of
the general utility of penalties:
Yang-tzu said, "Mutilating penalties are penalties indeed." But in accord with
the situation of the times, penalties need to be heavier or lighter. Thus lenience
and severity will have principles, and we will not lose the intention according to
which in antiquity the penalties were regulated. Generally, in a well-ordered
world, penalties are severe, and in a disordered world they are lenient. There-
fore, there were three thousand crimes under the Five Punishments and five
hundred crimes calling for the death penalty under the Institutes of Chou. During
the [disordered] era of King Mu, there were only two hundred capital crimes.
Today the class of capital crimes is large. If among the crimes with this death
penalty you seek those about which the facts are deserving of pity, and punish
the guilty using castration [one of the classic Five Punishments], many people
will be spared death. For military deserters who are due to be beheaded and for
bandits for whom the amount of the booty is sufficiently large that they are due
to be strangled, cutting off one foot would be a sufficient penalty.37

36 Hulsewe, Remnants of Ch'in Law, p. 15. 37 HCP 214.19b.

331
 Law and order in Sung China

Thus Tseng advocated mutilating penalties as a way of returning to the

mercy of the ancients, arguing that mutilations were merciful because
the alternative was execution.
I have heard that the Former Kings — who in their system of penalties did not try
things not rooted in humanity - practiced the cutting off of limbs and the carving
of flesh, reaching to exterminating punishments, which were not abolished. . . .
From [the Sage Rulers] T'ang and Yii and the Three Dynasties [of Hsia, Shang,
and Chou], for many centuries, in times of order and of disorder, of flourish-
ing and decline, of long reigns and short, still they had [the mutilating punish-
ments]. . . . I say that it would be appropriate, below the death penalty, to add
castration and the cutting off of the feet as substitutes for death when the
circumstances [of the crime] are not serious and that we should inflict the law of
tattooed exile in imitation of the ancient penalty of nose cutting and tattooing.
For the next level, men should be sent into exile according to appropriate
principles.38
In the debate occasioned by Tseng's proposal, the chief minister,
Wang An-shih, temporized. He agreed that the mutilating penalties
could be used under certain limited conditions, but even this less radical
position did not go unchallenged. Some officials argued that although
the mutilating penalties were ancient, they were hardly verities, since
they had arisen in times of disorder and were associated with ineffective
rulers. 39 In a debate of 1070 another official, Feng Ching, was an
outright, outspoken opponent of mutilations. These opponents of re-
introducing mutilating punishments continued to hold the rhetorical
high ground in the controversy. In the following century Ch'en Liang
commented:
Those who discuss the flourishing of the mutilating punishments believe that
they arose among the Miao people but that [the Sage King] Yao thought on
them and made use of them, requiting thefierceso as to deter people. They were
used to warn petty men but did not originate from the fundamental character of
the Sages. Therefore [the Sage King] Shun often did not carry them out, so as to
spare the people the penalties, and employed them only against those who
persisted in wrongful ways. . . . How could such punishments coexist with a
system designed to instruct and based on benevolence? Yet the Confucians of
today who speak of antiquity always say that "the well-field system, feudalism,
and the mutilating punishments all are the great principles and great laws of the
Sages and cannot be abolished." . . . In antiquity the Sages distinguished the
nature of men from the nature of beasts. . . . The mutilating punishments cer-
tainly existed in rough form, but as yet there was no system. The Miao people

38 HCP 2i4.i8b-2oa.
39 Wang Ling, Kuang ling chi, SKCSCP 1971, 3o.2b~3a. See also HCP 214.17b ff., 4.20a
(1069), 5.24b (1070), 292.5a, 328.1b.

332
 The Sung penal system

began to make instruments for killing people and used them without restraint.
Yao feared this might be taken as an example by the world. Therefore he
ordered and regulated them. . . . [The founding Chou] Kings Wen and Wu were
extremely cautious in the criminal trials of the people, and [their successors]
Kings Ch'eng and K'ang renounced and did not use [mutilating penalties]. . ..
When the Han commenced, it inherited the excesses of the Ch'in. The laws and
measures of the Former Kings had been completely abandoned, yet the mutilat-
ing punishments alone were preserved. [The Han] Emperor Wen (r. 179-157),
being moved by a woman's appeal, abolished them. Thereupon we can speak of
a Way of complete transformation. . . . Now, because [these punishments] have
been completely done away with, they cannot be revived. They were used only
with great care and occasionally by the Sages. If we tolerate their use, will this
not be reversing [the Sages'] system of order? If the people have a sense of
shame, then the current laws will suffice. If the people no longer care for their
lives, then even if the mutilating punishments are applied daily, they will still be
no law.40
In the end these opponents of mutilating punishments carried the day.
Mutilating penalties were not to be used. Convict laborers were a
key element in the state's labor economy. To have reduced their num-
bers or usefulness through mutilation would have caused enormous
inconvenience. Convict laborers were also men and so could be much
more easily controlled if they were left in a condition that would encour-
age them to renew themselves and become useful citizens.
The Sung debate over the mutilating penalties reflected a more gen-
eral debate over the appropriateness of Sung penalties. Did they fit the
circumstances of the Sung? Sung officials had a vision of the past history
of punishments and saw their own practice against this understood
background. In a long discussion of the evolution of the Chinese penal
system from classical times down to his own era, the late Southern Sung
official Lou Yuen (d. 1244) repeatedly stressed the importance of the fit
between the conditions of the times and the penalties applied. Having
brought his discussion from the Ch'in-Han through the T'ang and the
Five Dynasties, he continued:
As for the system used in our dynasty, it has become, day by day, more
generous. Thus, looking at the situation from today's point of view, men com-
ment on the severity of the laws of the Sung founder, T'ai-tsu. They fail to
understand that in his own time, he was seeking to transform the corrupted laws
of the Five Dynasties by making the penalties lighter.
In his own day, he thought, men had become too concerned about the
penalty's being overly severe, and the result was a tendency to memorialize

40 Ch'en Liang, Lung ch'uan wen chi, TSCC ed., 4.42-43.

333
 Law and order in Sung China
far too many cases, seeking from the central authorities a reduction in
the penalty. 41

Sung penalties: law and practice

For those who did commit crimes, the Sung state inflicted an elaborate
variety of punishments. These reflect a mixture of goals. Initially the
Sung used punishments inherited from the T'ang by way of the Five
Dynasties. The basic penal laws enacted at the beginning of the Sung
were contained in a code compiled two years after the founding of the
dynasty. This code, the Sung hsing t'ung, was largely a reedited edition of
a code from the Later Chou dynasty (951-60), which in turn reproduced
the T'ang Code of 737. The Later Chou Code and its Sung successor
simply added to the T'ang Code a variety of laws that had been enacted
after 737.
This Sung Code of 962 remained in use for various legal purposes
under the dynasty. It was used in training legal officials, and when
the officials concerned with legal education objected to the presence of
passages that differed from the T'ang original, a new edition especially
for students was issued in 1029, purged of materials added after 737. 42
The first code of 962 also continued to serve other purposes in the legal
system and continued to be of concern to statesmen. During the reign of
Shen-tsung (r. 1068-85), a period noteworthy for the large number of
legal compilations issued, the powerful statesman Tseng Pu oversaw a
revision of the text, supposedly eliminating items not currently being
enforced by the Sung and clarifying the commentaries. 43
The Sung had inherited and enshrined in the code of 962 the tradi-
tional Five Punishments: death (ssu) (by beheading and strangulation),
exile (liu) (in three degrees of three thousand, two thousand, and one
thousand lid), penal servitude (t'u) (in five degrees of three, two and a
half, two, one and a half, and one year), beating with the heavy rod
(chang*) (from one hundred blows to sixty blows by groups often), and
beating with the light rod (ta) (from fifty blows to ten blows by groups of
ten). 44 The code also contained a set of equivalent fines, under which
people convicted of crimes could in certain circumstances convert the
ordinary penalties to payments to the government. There does not seem
to have been any large-scale use of fines per se by the Sung state. Such

41 Lou Yiieh, Kung-k'uei chi, 22.333.

42 Wang Ying-lin, Yu-hai, 66.28a-b.
43 SHYhsing-fa 1.8b.
44 SHT 1.1-5. For the T'ang system, see Wallace Johnson, The T'ang Code: Vol. 1, General
Principles (Princeton, NJ: Princeton University Press, 1979), pp. 55 ff.

334
 The Sung penal system
collections always took the form of conversions from the regular penalties
permitted by the authorities.
Although in 962 the Sung government adopted this code based almost
entirely on the T'ang Code, at almost the same time it altered the
functioning of the penalty system by introducing a system of conversion.
Early in 963 all penalties below death were ordered converted to beat-
ings with the ordinary penal heavy rod, sometimes followed by penal
labor. The following chart shows this early Sung conversion scheme: 45
Code penalty Converted penalty
Exile with added labor 20 blows on the back plus three years of labor
Exile to 3000 lid 20 blows on the back plus one year of labor
Exile to 2500 lid 18 blows on the back plus one year of labor
Exile to 2000 lid 17 blows on the back plus one year of labor
Penal servitude
3 years 20 blows on the back
2.5 years 18 blows on the back
2 years 17 blows on the back
1.5 years 15 blows on the back
1 year 13 blows on the back
Heavy rod
100 blows 20 blows on the buttocks
90 blows 18 blows on the buttocks
80 blows 17 blows on the buttocks
70 blows 15 blows on the buttocks
60 blows 13 blows on the buttocks
Light rod
50 blows 10 blows on the buttocks
40 blows 8 blows on the buttocks
30 blows 8 blows on the buttocks
20 blows 7 blows on the buttocks
10 blows 7 blows on the buttocks

This policy seems to have remained in effect until the closing years of
the Northern Sung; in 1118 the schema was revised somewhat, to lighten
some of the penalties. Under this policy the revised penalties were 46
Penal servitude
2.5 years 17 blows
2 years 15 blows
1.5 years 13 blows
1 year 12 blows

45 HCP 4.6a-7a.
46 WHTK 167.1452.
335
 Law and order in Sung China

Light rod
50 blows 10 blows
40 blows 8 blows
30 blows 7 blows
20 blows 6 blows
10 blows 5 blows
This system of equivalencies continued to be used during the Southern
Sung. Material in the early thirteenth-century code Ch'ing-yiian t'iao-fa
shih-lei suggests some of its uses. Those who were eligible to commute
penalties could use the system in calculating payments. The conversion
policies were also useful in cases in which the person convicted had
already suffered some sort of penalty. Conversion allowed the officials
to assess how the penalty already suffered should be measured when
deducting it from the amount of punishment still due. 47 The system of
conversion put into place in the 960s was modified and refined, so that
during the latter part of the dynasty Sung authorities could assess
penalties according to the traditional categories of the Five Punishments
and then mete them out in terms of penal registration, with each degree
of a traditional penalty being comparable to a level of penal registration.
They could also substitute some penalties for others. For example, men
who had labored could use a certain number of days of work to lower the
number of blows that they were to receive as a part of their sentence.
Although the system was very useful for conversions, it seems that
some of its provisions were not always used. Under its terms, the light
rod should have disappeared entirely as an actual tool for punishment,
yet it continued to be used regularly for punishments in the military.
Furthermore, in 1108 the authorities again passed a law on beating with
the light rod, under which the smaller rod was to be used in carrying
out punishments, with a nominal ten blows actually being five, twenty
nominal blows being seven, thirty being eight, forty being fifteen, and
fifty being twenty. The heavy rod was not to be used for analogical
conversion.48 The mid-thirteenth-century text of forensic examinations,
the Washing Away of Wrongs (Hsi-yiian lu), also indicates the continued
use of the light rod, saying with regard to deaths resulting from corporal
punishment that "the marks from beating with light rod on the left side
[of the buttocks] should have a crosswise length of three and a half
inches and a breadth of three inches." 49

47 See, for example, TFSL 515.

48 WHTK 167.1452.
49 McKnight, The Washing Away of Wrongs. Because the military was deeply involved in the
Sung penal system, it is certainly possible that such scars might have resulted from
discipline inflicted on men while they were serving in military units.

336
 The Sung penal system

Names and realities in Sung punishments

Actual penal practices often ignored both the code and the changes
brought about by the conversion policy. The Sung took over the edicts
and rules from the Five Dynasties, under which the penalties for many
crimes were heavier than those found in the current code. For example,
the code's punishment for falsifying official papers was exile to two
thousand lid but under current laws outside the code the penalty was
strangulation. 50 More generally, for many of those who under the code
would have been liable for penal servitude or exile, the actual penalty
inflicted was penal registration with added labor, a punishment that
combined some of the characteristics of both of the earlier penalties.
Despite the existence of these complicating factors, Sung officials con-
tinued to use the terms exile and penal servitude in their discussions,
because these terms served as measures of the seriousness of crimes, and
as noted, these traditional categories also provided the standards for
converting penalties through commutation or privilege. 51 Thus, a report
on the newly compiled Collected Decrees of the Ch'ing-li Era, indicating
which categories of penalties were larger and which were smaller than
those in the previously collected edicts of the T'ien-sheng era (1023-32),
notes an increase of eight edicts calling for capital punishment, an
increase of fifty-six calling for exile (liu), a decrease of sixteen calling for
penal servitude (t'u), and increases of thirty-eight, eleven, and three,
respectively, for the heavy rod, the light rod, and penal registration.
Edicts in the category "death sentences or less that call for the sub-
mission of a memorial with an ensuing imperial order" were reduced by
twenty-one.52
As this report makes clear, the Sung edicts might specify exile and
penal servitude when in practice these would be converted to beatings or
penal registration. Under Sung law, penal servitude was considered to
be equivalent and was converted to penal registration in one of the
prison citadel commands of the provincial armies at five hundred lid or
more. Exile was equivalent to penal registration in one of the prison
citadel commands located at one thousand or more lid from the place of
jurisdiction. 53 Penal registration thus in practice took the place of these
earlier penalties (see Chapter 12).

50 SS 201.5009.
51 On privileges, see Brian E. McKnight, "Song Legal Privileges," Journal of the American
Oriental Society 105 (1985): 95-106.
52 HCP 160.2b. See also Wang Ying-lin, Yu-hai, 66.29b; SS 204.6a.
53 HCP 214.27a; see also HCP 322.3b.

337
 Law and order in Sung China

Figure IO.I. Judicial torture. From Lo Kuan-chung, Ch'uan-hsiang San-kuoyen-i.

Beatings as penalties
As a form of punishment, beatings have several advantages from the
state's point of view. If sufficiently public, they not only punish the
offender but also serve as a deterrent. Moreover, they avoid the expense
and difficulty of controlling wrongdoers for an extended period of time.
Beating may leave scars, but they are not so disfiguring as the mutilating
penalties are, which makes it easier to reintegrate the victim into regular
society. Being beaten might be thought of as shameful, but its effects
were supposed to be temporary, and the degree of shame attached was
diminished by the common use of beatings during interrogation. Accord-
338
 The Sung penal system

Figure 10.2. Judicial torture. From Li Pao-chia, Kuan-ch'ang hsien-hsing chi.

ing to a Later Chou decree reproduced in the early Sung Code, it was
legal to beat someone up to fifteen blows on the back with the heavy rod,
"in the course of public business." 54 Such beatings during judicial
torture were to be administered by dividing the blows among the but-
tocks, thighs, and bottoms of the feet55 (see Figures 10.1 and 10.2).
Under Sung law all sentences more serious than beating with the
heavy rod had to be reviewed at judicial levels above the district.
Beatings, however, could be mandated and carried out at the district
level, the higher authorities requiring only that periodic reports be sub-
mitted regarding such punishments. The severity of beatings could also
be adjusted to the severity of the crime. For these reasons, beatings were
widely used in traditional China to punish relatively minor crimes. In
Sung China, it was the sole penalty for many less serious crimes, includ-
ing such misdemeanors as the illegal taking of religious orders or the
taking of family goods by coresident juniors, most assaults, driving
carriages or riding horses recklessly in crowded areas, and being drunk

54 SHT 1.21. 55 TFSL500.

339
 Law and order in Sung China
56
and disorderly. They might also serve as supplementary penalties.
Convicts were often beaten before being sent off to do penal labor, and
for some crimes, like thievery, the criminals were also tattooed before
being released. 57
Numerous regulations governed these beatings, including their timing,
who was to administer them, and the dimensions of the rods used. For
example, beatings were not supposed to be inflicted while prisoners were
ill or when they were still bearing wounds from previous beatings. 5 8
Such beatings were generally inflicted during the daytime, perhaps be-
cause when officials did carry them out at night, those sentenced to the
punishment often hired substitutes. 5 9 When this happened, according to
the rules found in the early thirteenth-century code Ch'ing-yiian t'iao-fa
shih-lei, both the person for whom a substitute suffered the penalty and
anyone who enabled the substitution were liable for one hundred blows
of the heavy rod. 6 0 Local officials were also said to have sometimes
illegally increased the number of blows in carrying out sentences. 61
Because they were concerned about the victims' suffering from the
heat, the authorities specified that the beatings were to be carried out
before seven in the morning or after seven in the evening (though if there
were a great many cases, they might be carried out until nine in the
morning and after five in the afternoon). If the beatings continued into
the night, those functionaries responsible were liable for a penalty of one
hundred blows of the heavy rod. 6 2 O n e document suggests that the
beatings were carried out in groups - originally every ten days, but after
a ruling in 1011, every three to five days. 6 3 Rules also specified the days
when penalties could be inflicted. Apparently they were not supposed to
be carried out on certain holidays, though after 1022, as a result of an
official's request, beatings with the heavy rod could be inflicted on
imperial birthdays. 6 4
During the Sung, as under the T'ang, the regular senior officials were
supposed to attend interrogations at which judicial torture (in the Sung,
legally restricted to beating with the heavy rod) was used. This rule was

56 SHT, passim.
57 Wang Yen, Shuang-ch'i lei-kao, SKCSCP 1972, 23-5b-8b.
58 TFSL 499, 500.
59 HCP 76.10a.
60 TFSL 507.
61 HCP 82.16a (1014).
62 TFSL 499, 500.
63 HGP 7 6 . i o b - n a ; WHTK 166.1446 ( i o n ) .
64 HCP 99.4a. Note that Li Tao quotes a now-lost source as saying that "under the old
system it was forbidden to carry out penalties on imperial birthdays. From the
beginning of the T'ien-sheng period [1023] it was allowed to inflict the heavy rod"
(HCP 99.14b). The fang in this passage should be read as chang.

340
 The Sung penal system

presumably intended to prevent the abuse of prisoners by lower func-

tionaries, though in practice the higher officials often left this duty to the
clerks.65 Clerks, on the other hand, except for those actively serving as
judicial clerks, were not allowed to go to the place of torture. These
beatings were inflicted by special functionaries called heavy-rod wielders
(chang-chih) (supposedly while being observed by the unit's chief officials).
These functionaries were stationed at the prefectural and district seats.
When needed by officials on tour or in offices without such people on
their staffs, they were called up for temporary service, after which they
were to return to their posts. 66 If such men increased the number of
blows beyond what was called for, or added materials to the surfaces of
their rods to make them more damaging, they could be punished with
two years of penal servitude. When administering the beatings that
preceded a sentence of penal servitude, exile, or registration, if they
deliberately beat the convicts in such a way as to kill them, the heavy-
rod wielders were to be tried for deliberate homicide.67 The Sung retained
from earlier dynasties the rule that the men administering the beatings
could not be changed during the beating. 68
From the beginning of the dynasty, regulations specified the length
and width of the several sorts of rods used for various penal and judicial
purposes. An edict of 963 states that in the past these devices had been
governed by the penal officers' ordinances {yii-kuan ling). When the law
of rates of converting penalties was passed, the regular official heavy rod
was ordered to follow the specifications of the Later Chou rules of 958,
being three feet five inches long, with a thickness at the large end of two
inches and, at the small end, a diameter of not more than nine-tenths of
an inch. The light rod was not to exceed four feet five inches in length
and to have a large-end diameter of not more than six-tenths of an inch
and a small-end diameter of not more than five-tenths of an inch. The
dimensions of the heavy rod remained unchanged during the dynasty, at
least until the early thirteenth century. Those of the light rod remained
almost unchanged, though in the code of 1202 this rod was ordered to be
four feet five inches long, with a lower-end thickness of four-tenths of an
inch. 69 There also were separate rules for the rod to be used in inter-
rogation.70 Although these early rules set the other dimensions, not until
1029 w a s ^ decreed that the weight also be regulated. The standards

65 Chang Nieh, Tzu-wei chi, 25.15a.

66 TFSL 500.
67 TFSL 499.
68 TFSL 500.
69 TFSL 504; see also WHTK 166.1444.
70 HCP4.6a-7a.
341
 Law and order in Sung China

LONG FETTER

Figure 10.3. Prison instruments. Redrawn from Wang Ch'i, San-ts'ai t'u-hui.

were to be stamped on the face of the instruments, to show that they

were regularly inspected by the responsible officials.71 (The rules found
in the Ch'ing-yiian t'iao-fa shih-lei called for the incumbent officials to
inspect them once a month and to certify their correctness by having
them branded.) Moreover, in his early twelfth-century book of advice for
district magistrates, Li Yiian-pi advises those newly arrived at a post to
have someone weigh the judicial instruments to make sure that they
conformed to the regulations. The weigher should be made to sign

71 HCP io8.8b-ga.
342
 The Sung penal system

his name or make his mark on them. The instruments would also be
branded at this time to certify their correctness.72 The penalty for
possessing illegal instruments was sixty blows of the heavy rod 73 (see
Figure 10.3).
Nonetheless, abuses were a perennial problem. A regulation found in
the Ch'ing-yuan t'iao-fa shih-lei shows that the authorities continued to try
to suppress such violations. As noted, those guilty of actually killing their
victims could be tried for homicide:
If those who administer beatings clandestinely increase the number of blows or
on the surface of the heavy rod they add other materials, deliberately acting in a
cruel manner, then they will be liable for two years of penal servitude. If
motivated by rapacity or personal hatred, they beat men liable for penal servi-
tude or exile, sentence [the functionaries responsible] to penal registration in
their own prefecture. If they deliberately beat people to death, the guilty func-
tionaries are to be tried for deliberate homicide.74
A twelfth-century official commented:
Written rules governing the system of heavy and light rods exist. Their weight
and size are not to be privately altered. However, in recent years, functionaries
cruelly destroy the humane intent. In cases of putting prisoners to the question
using the penalty rod, each application is not to exceed thirty blows, and the
total is not to exceed two hundred blows. This is the intent of the law. But
nowadays the prefectures and districts do not use the penalty rod but, rather, use
the split rattan or double the penalty rod, binding two together as one, or they
whip the haunches and feet. They sometimes apply as many as three hundred or
five hundred blows. No punishment is more cruel than this.75

Fetters
Social defense was not the main focus of premodern Chinese penalties,
but fetters should be considered in this light, as devices that not only
constituted punishments in themselves but also hindered known malefac-
tors from committing new outrages. Those held for crimes with penalties
greater than beating were often confined in various sorts of fetters, a
common practice in premodern China. The basic rules for their use in
the Sung, cited in the Sung Code of 962 from the penal officers' ordi-
nances (yu-kuan ling), are repeated with minor modifications in the
collection of approved memorials and edicts called the Collected Sung

72 TITC 1.4b.
73 TFSL 499, 503.
74 TFSL 499.
75 WHTK 167.1454-55; see also TFSL 499.

343
 Law and order in Sung China

MMMSM
Figure 10.4. Wearing the cangue. From Feng Meng-lung, Ku-chin hsiao-shuo.

Documents and were reaffirmed in the early Southern Sung under Kao-
tsung (r. 1126-62). 76
According to the rules in the Collected Sung Documents, men guilty of
capital crimes were to be kept in fetters and manacles. Even men liable
only for a beating with the heavy rod were sometimes required to wear
fetters. Fetters, including both a special cangue and chains, were also
used on prisoners during transport. 77 Fetters might in themselves be

76 On Southern Sung, see HNYL 2730.

77 See, for example, SHY, hsing-fa 4.6b~7a, 6.17D-18D, 6.78b.

344
 The Sung penal system

used as a minor punishment. In 1005 it was ordered that men who

entered lawsuits that were none of their business might be beaten and
put into a cangue for ten days (see Figures 10.3 and 10.4). 78
Some categories of lesser criminals or disadvantaged persons were to
be exempted from wearing fetters. Those eighty or older or ten or
younger, the disabled, the pregnant, or dwarves did not have to wear
fetters even if guilty of capital crimes. 79 There were also circumstances
under which prisoners in general were spared the fetters. Before being
shackled, prisoners were supposed to be examined by a medical practi-
tioner to see whether they were ill, disabled, or otherwise not able to wear
the bonds. From the beginning of the dynasty, prisoners who were
seriously ill were supposed to be freed from their fetters. 80 Such rules
existed throughout the dynasty. The code of 1202 further indicates that
prisoners in transport who were under the care of a doctor did not have
to wear restraints. 81 If the prisoners were women, it was necessary to
determine whether they were pregnant. Pregnant women, like the aged,
ill, and disabled, did not have to wear fetters even if guilty of capital
crimes.
The dimensions of these fetters, which were to be inscribed on their
surfaces, were governed by the category of laws called specifications
(shih); the general rules for their construction, laid down as ordinances
(ling), were occasionally reconfirmed. From the beginning of the dynasty
the specifications of most sorts of shackles were fixed by law. However,
until 1045 t n e interrogation cangue (p'an-chia) worn by prisoners in
transit was not regulated, even though it was in regular use. A memorial
in that year asking that rules be established was approved. Judging from
an edict dated 1119, the weight of the instrument was not to exceed ten
pounds. 82 The fetters were supposed to be made of dry wood; the joints
were to be shaved off; and no nails or other protuberances were to be
added.
The most complete description of such fetters comes from the late
Northern Sung official Li Yuan-pi. He advised Northern Sung magis-
trates that for serious offenders they should have artisans build "four-
layer cangues," with two layers of wrought iron and two layers of
uncured thick cowhide attached while wet, each layer to be three fingers
wide and thick. For lesser prisoners the cangue could be made simply of
two layers of wrought iron. On both types, soft hemp ropes were to be

78 SHY, hsing-fa 3.12b-13a.

79 SHT 2g.2b-3a; SHY, hsing-fa 6.51a; TITC 5.25b.
80 SHT 29.471.
81 TFSL 530, citing the tuan-yii ling.
82 SHY, hsing-fa 6.77a~78a, 78b.

345
 Law and order in Sung China

fastened tightly through the front and back clasps. In the left, extended
end of the cangue a hole three ringers large was to be drilled, so that at
night when the prisoners had returned to their cells, a long iron chain
with bells on it could be run through the hole and attached. Serious
offenders were to be fettered with this iron chain, which was to be eight
Chinese feet long, with a foot shackle at one end, made of good-quality
wrought iron and as thick as the t h u m b . T h e shackle was to be pounded
shut around the prisoner's foot but was to be wrapped in soft cloth so
that it would not chafe. When the prisoner was released, a special tool
could be used to open it. 8 3
T h e Sung also kept in force an earlier regulation that during the sum-
mer the fetters were to be washed every five days. This rule appeared in
the Later T ' a n g in 923, was included in an edict of the Later Chou in
955, was incorporated in the Sung Code of 962, and was repeated in
later memorials and edicts. 8 4 O n the day when the fetters were to be
washed, one official was supposed to go in person to check the jails and
their equipment, including fetters and other judicial instruments. 8 5
There was concern from the beginning that officials might be either
overly lenient or overly harsh. They might allow prisoners to avoid
wearing the required fetters, use ones that were excessive and illegal, or
fetter prisoners who should not have had to wear them at all. 8 6 This
concern seems to have been prompted on the one hand by fears that
officials might take bribes to spare their charges of wearing fetters. T h e
early Sung Code says that officials who exempted prisoners liable for the
heavy rod from wearing fetters should themselves be liable for a penalty
equivalent to beating with the light rod. When the crimes called for
penalties of penal servitude or higher, the responsible officials should be
liable for a penalty heavier by one degree. If they merely switched the
kind of fetters in question, allowing those who were supposed to wear the
cangue to wear chains instead (or vice versa), their punishment would
be reduced one degree. 8 7
Officials might also abuse prisoners by using excessive or improper
fetters. Higher officials were well aware that wearing fetters could easily
lead to infection. Late in the Northern Sung a warning was issued that
officials who detained in fetters men who should not be so detained were
liable for a penalty equivalent to one hundred blows of the heavy rod. 8 8
83 TITC 2.9b.
84 See WTHY 161; SHT 29.471; SHY, hsing-fa 6.51a, 77a-b, HCP 10.7b.
85 WHTK 167.1453.
86 HCP477.ib-2a.
87 SHT 29.465-466.
88 HCP477.ib-2a.

346
 The Sung penal system
Despite the threatened penalties, the problem was chronic, and officials
occasionally were admonished to adhere strictly to the rules. 89 Concern
about men being shackled too long was also one spur to attempts to
speed up judicial process, especially at the review level. 90
Officials are accused repeatedly of using illegally heavy or injurious
fetters. They might be perforated or of knobby, rough, or raw rather
than dry wood.91 A report of 1118 says that the clerks in the detention
houses sometimes used iron fetters and neck fetters. 92 In response, the
central government ordered the chief authorities in each jurisdiction
periodically to investigate the legality of the instruments being used. 93
According to the Ch'ing-yuan t'iao-fa shih-lei, senior officials were to make
such inspections every month, to see whether the instruments conformed
to the specifications. The instruments were branded to indicate their
certification.94
When such investigations were conducted, abusive officials actually
might suffer punishment. In the capital in 1142 a number of officials
suffered penalties because the instruments in their jurisdictions were
either under- or overweight or were improperly constructed, and in 1149
senior officials were again told to inspect the instruments in person and
to certify their conformity with the rules. 95 The penalties inflicted on
officials in these special cases were relatively severe; under the regular
rules, as preserved in the Ch'ing-yuan t'iao-fa shih-lei, those having illegal
fetters were liable for only sixty blows of the heavy rod. 96
The problem of controlling the system was caused partly by con-
fusion about who was responsible. Even when there were rules for such
responsibility, they sometimes assigned it to people who lacked the
power to see that the rules were obeyed. It is an example of the classic
problem of bureaucratic systems in which those who know what is wrong
do not have the power to correct it, and those who have the power to
correct it are ignorant. In some areas, the difficulty of preventing the
building and use of illegal instruments was compounded by the policy of
appointing archers rotated for one-month terms to be in charge of
them. Such men were hardly in a position to resist pressure from the

89 HNYL 2588.
90 HCP 65.16b (1007).
91 SHY, hsing-fa 6.77b.
92 WHTK 166.1452.
93 SHY, hsing-fa 6.51a, 77a~77b, 77^, 78b-7ga.
94 TFSL 503.
95 SHY, hsing-fa 6.79a; SS 199.4993; HNYL 2588.
96 TFSL 499.

347
 Law and order in Sung China

Table 10.1. Sung fetters

Length Width Diameter Weight

Cangue (chia) 5'-6' 3"
25 chin (death)
20 chin (exile/servitude).
15 chin (heavy rod)
Interrogation cangue Wchin
(worn during
transport also)
Press (chia) 2'5"-6" 3"
Manacles (ch'ou) l'6"-2" 3" 1"
Collar (chHen) 1'-1'5" 8 Hang-1 chin
Lock/chain 8'-10'2"

Note and sources: SHY, hsiang-fa 6.77a; SHT 29.466. Initially there was no rule
fixing the weight of the cangue used when dealing with recalcitrant prisoners
accused of crimes calling for beating with the heavy rod, but in 1007 this was fixed
at fifteen pounds. See also Wang Yii-ch'eng, Sheng-shuiyen-Van lu, cited by Shen
Chia-pen, Shen ch'i-i hsien sheng (Taipei: Wen-hai, 1964), p. 514; Chiang Shao-yii,
Huang-ch'ao leiyuan 20.7b, 21.10b; WHTK 166.1445; SHY hsiang-fa 6.77b; HCP
67.18b; WHTK 166.1445, 167.1453.

permanent jail functionaries or the officials to permit the use of illegal

instruments. 97
In central government offices, with their larger staffs, the authorities
sought to avoid such problems by designating those who were to keep
the fetters. For example, in the late 1080s, in the Censorate and other
analogous organs, the storehouses for the fetters were under the direct
control of civil service officials called recorders 98 (see Table I O . I ) .

Judicial tattooing
Judicial tattooing also served as a mechanism for social defense. Tattoos
were felt to be a form of mutilation. The knights-errant of Chinese
tradition often had themselves tattooed as a permanent sign of their
defiance of ordinary social conventions and political authority. A tattoo
on a criminal, however, served to warn those whom he encountered of
his status. They also served as a control device and a deterrent, since
convict tattoos were felt to be shameful. Their deterrent effect, however,
was reduced by the general Sung practice of tattooing soldiers, so that a

97 Sun Ying-shih, Ch'in-ch'uan chik, 2709. 98 HCP 374.15b.

348
 The Sung penal system

significant minority of the Sung adult male population bore tattoos of

one sort or another. Furthermore, the tattoos used on criminals were, in
many instances, deliberately designed to resemble military tattoos, in
order to facilitate the transfer of ex-convicts into military units.
Sung criminals guilty of crimes calling for more than beatings might
or might not be tattooed. If they were, the size, placement, and text of
the tattoos might differ. In using such markings the Sung government
was following an ancient tradition. In both the Book of Historical Docu-
ments and in the Autumn Officer section of the Rites of Chou (c. 200 B.C.),
one of the regular penalties was tattooing on the face. Here, the intent
was to inflict a permanent stigma on the convict, shame the guilty,
deter the tempted, and defend society against a potential lawbreaker by
revealing his character. This penalty was supposedly eliminated by
Han Wen-ti (r. 179-157), but it reappeared during the Southern and
Northern dynasties (317-589). Under Ming-ti of the Liu Sung (r. 465-
73), criminals were tattooed on the face and leg, and under Wu-ti of the
Liang (r. 502-50), on the face. The Western Wei (535-57) tattooed men
and women escapees. During this same period the practice of tattooing
army troops also began, probably under the Southern Ch'i (479—502).
Under the T'ang, private tattooing was practiced but was not a
regular punishment, nor was it used in the military. However, during the
closing years of the T'ang, several rebel leaders revived the pre-T'ang
practice of tattooing their troops. This practice remained current during
the ensuing Five Dynasties period and was inherited by the Sung."
Although the general traditional practice had not been to tattoo men
sentenced to exile or penal servitude, the tattooing of criminals also
reappeared in the Five Dynasties during the T'ien-fu period (936-43) of
the Later Chin, when criminals sentenced to penal registration were
tattooed on the face.100 The Sung inherited and retained this practice.
We are told in an order of 1013 that unless there was an imperial order
to use large characters, those sentenced to penal registration should
follow the precedent of the small-character army tattoos. If the criminal
already had a tattoo indicating that he belonged in an army command,
the authorities could merely add the place of registration.101
Although the tattooing of large characters was said to have often
seriously injured the criminals,102 for some very serious crimes large
characters were still used. In 1024 t n e authorities in the capital area

99 See Sogabe Shizuo, "Sodai no shi-hai ni tsuite" (On penal registration in the Sung),
Bunka 29 (1965): 1-23.
100 WHTK 168.1459.
101 SHY, hsing-fa 4.6b.
102 SHY, hsing-fa 4.7a.
349
 Law and order in Sung China

asked that evil robbers be tattooed on both cheeks with large characters
so that it would be difficult for them to remove traces of the tattoos with
drugs or cover them over with burns. 103 A decree of 1085 specified that
men guilty of robbery were to have a ring tattooed behind their ears,
apparently with the character "bandit" (tsei) inside. 104 If the penalty
were penal servitude or exile (liu), a square was to be tattooed, presum-
ably inside the ring, and if the penalty were beating with the heavy rod,
this was to be a circle. For those recidivists guilty of three crimes calling
for beating with the heavy rod, the tattoo was to be moved to the face. In
size it was not to exceed five-tenths of an inch (fen).105
The same pattern continued during the Southern Sung. Wang Yen
(1138-1218) noted that thieves who had been beaten with the heavy rod
were tattooed with a circle and released. 106 During this period, for
those to be sent for registration, the general pattern seems to have been
to tattoo a phrase that included the place and unit of registration. A
typical example would have read, "Penally registered in X prefecture, Y
Military Agricultural Unit, for heavy labor." 107 There were occasional
changes in the phrasing. Apparently before the early 1080s the phrase
read "tattooed and made subject to X Command for penal registration
in the army" (tz3u-ch3ung mou chih-hui pei-chun), but because the emperor
was worried t h a t the soldiers might feel s h a m e d , this was changed to
read simply " X C o m m a n d miscellaneous l a b o r e r " (mou chih-hui tsa-i).W8
For S o u t h e r n S u n g criminals, these characters were usually larger
t h a n those used on a r m y troops, b u t if such m e n were for some reason
freed a n d wished to r e m a i n in the provincial armies, the authorities
could simply m a k e relatively minor alterations in the characters. 1 0 9 An
O r d i n a n c e on J u d g i n g C r i m i n a l Cases - which dates from the N o r t h e r n
Sung b u t is preserved in the Ch'ing-yuan t'iao-fa shih-lei - states:

For all military [personnel] who are shifting their place of registration when the
name of the unit is not the same, and for those who are being demoted and
penally registered, the characters with which they are to be tattooed are not to
exceed two-tenths of an inch. For deserters or men penally registered in the base
citadels, it should be four-tenths of an inch. For prison citadel [commands],

103 SHY, hsing-fa4.na.

104 HGP 468.7a-b.
105 HCP 362.11b; SS 201.5018; WHTK 168.1460.
106 Wang Yen, Shuang-ch'i lei-kao, 23-5b-8b.
107 See, for example, SHY, hsing-fa 4.47a, 447a-48a, WHTK 168.1460.
108 HCP 334.10a.
109 See, for example, an order of 1189 that after an amnesty if there are registered men
who are old, ill, or otherwise incapable of supporting themselves, they might on
request have their tattoos altered and remain in the provincial forces (SHY, hsing-fa
b-59a).
350
 The Sung penal system

five-tenths of an inch; for distant evil prefectures or for Sramana Island, seven-
tenths of an inch. 110

As this indicates, large characters were used on certain very serious

criminals. One large group was composed of robbers who would ordi-
narily have been sentenced to death but who had been spared. Accord-
ing to an order of 1167, they were tattooed with large characters. 111
An order of 1181 went further. Such men were to have the character
"forceful robber" (chiang-tao) tattooed on their foreheads, with the other
characters divided and tattooed on their cheeks. (If their foreheads were
already tattooed, they could be spared being tattooed again.) 112

Conclusion
Penalties are messages, messages to those who suffer, to those who
observe, to those who execute, and to those who know of them only by
rumor. Penalties inform both spirits and men, both the powerful and the
weak, for good or ill. The messages may be deliberately spectacular and
open, like the public executions of Europe (and China) until recent
times; the closed sanitized executions of the contemporary world send
with their silence as loud a message. Penalties are messages about
power, about the capacity of the punisher to demean the punished.
Penalties tell the victim of his powerlessness. They remind the bystanders
of their own weakness. They even warn the executioner. They demon-
strate resolve coupled with authority. The use of force is a message about
the power to rectify and the power to cleanse. It is a necessary attribute
of sovereignty, but in Chinese political mythology the Chinese sovereign
was properly the sovereign of all, not just of the Chinese. To rectify
meant to straighten out not just those within who were warped but also
those without. Because those without were less immediately under the
power of the sovereign, he could not punish them with police power.
Greater force was needed. War was distinguished from punishment
principally by its scope and intensity. War involved groups of men
acting against other groups, using different methods to be sure, but still
aiming to chastise the deviant. War was criminal punishment carried
one step further; criminal punishment was war against those so power-
less that their resistance did not call for the use of great force.

110 TFSL 520. This ordinance, like many of the rules found in this code, is from the
Northern Sung, as is indicated by the mention of Shamen Island, which was lost to the
Sung in the 1120s.
111 SHY, hsing-fa 4.51b.
112 SHY, hsing-fa 4.56a; WHTK 167.1455; see also TFSL 520.

351
 Law and order in Sung China

The voices of conservative traditional Chinese advocates of the strict

and severe enforcement of penalties sound familiar. The official Liu An-
shih remarked in 1089:
I have heard that in antiquity, punishments were used to constrain those who
injured the state by being disloyal or unfilial. Therefore, when the crime was
great, the punishment was heavy. When the crime was minor, the punishment
was light. It is also said that when the punishments fit the crimes, then evil will
be stopped. This is the Way of great fairness. The practice of this was the same
in antiquity as now. It has been called "awe inspiring authority," whereby to
buttress virtue and to restrain evildoing.l ] 3
The great eleventh-century statesman Ssu-ma Kuang drew the political
lesson even more sharply, saying to the emperor that the failure to
enforce the law strictly would endanger imperial power. "If those who
kill men are not executed, if those who injure men are not punished,
under such circumstances even the great Sage Emperors Yao and Shun
would not be able to govern." 114
The penalties to be used ran the gamut from horrifying executions to
beatings, the use of fetters, and tattooing. All could serve as either
separate or supplementary punishments. The Sung authorities clearly
understood their penal character. Their role as deterrents and as instru-
ments of social defense was recognized by thoughtful observers. Another
facet of Sung penalties, which to us seems clearly punitive in effect if not
in intention, was not so recognized - the time spent in jails.

113 HCP 427.7b-ga. 114 SS 201.5011.

352
 11
Jails and jailers in
the Sung

Introduction - imprisonment and punishment

The policies governing imprisonment in the Sung dynasty reflect a
mixture of motives determined by the special uses of jails, by current
views of the nature of criminals, and by some general characteristics of
the traditional Chinese state. Jails were used primarily for those awaiting
either trial or punishment. They might also hold witnesses or convicts in
transport. Imprisonment, however, was not thought of as a punishment
in itself. The widespread conviction that most people, including most
criminals, were capable of acting properly conflicted with a tendency to
dismiss those in jail as bad people who deserved their fate. After all,
their being in jail was proof of something. In his book of advice to local
officials, the thoughtful and compassionate Ch'ing magistrate Huang
Liu-hung comments: "It is all too easy to say that criminals are wicked
people who deserve nothing except death. Alas, Heaven creates tigers,
leopards, wolves, wasps, and lizards, and they are all endowed with
lives. How can human beings be deprived of an opportunity to live?"1
Concern for the lives of convicted and suspected criminals was part of a
wider human concern. Furthermore, not all those held in jails were in
fact either convicted or suspected criminals. Under premodern condi-
tions (as indeed is still all too true under modern conditions) the dif-
ficulties of getting witnesses to the courtroom at the proper time are very
serious. The traditional Chinese answer was to hold certain witnesses
in jail. The willingness of the authorities to do this may have been
buttressed by a feeling that some of the witnesses, though perhaps not
all, shared in the guilt of the accused, especially if they were relatives or
neighbors, who had obviously failed to prevent the lawbreaking. Given

1 Huang Liu-hung, A Complete Book Concerning Happiness and Benevolence, trans. Djang Chu
(Tucson: University of Arizona Press, 1984), p. 313.

353
 Law and order in Sung China

Figure I I . I . Chinese jail. From Feng Meng-lung, Hsing-shih heng-yen.

the difficulties facing criminal investigators under the conditions of the

time, and the state's desire for correct verdicts, it is understandable that
the authorities often found it advisable to detain in jail those who had
evidence as well as the accused, though this practice might also have
worked to weaken the system by making people less willing to come
forward with evidence for fear of being imprisoned.
This situation may have contributed to the practice of not treating
time spent in jail as part of the penalty. If punishments were inflicted on
the guilty to deter others, to inflict retribution, to cause the guilty to
reform, and to defend society against further depredations, how could
time in jail count as punishment, when on occasion the innocent were
354
 Jails and jailers in the Sung
also held? Significantly, no Sung official seems ever to have suggested
reducing the time of the sentence by the amount of time spent in jail
awaiting trial.
Despite the authorities' desire to limit time spent in jail, stays were not
always brief.2 People might be held for extended periods, under condi-
tions that were at best uncomfortable and shameful and at worst danger-
ous (see Figure I I . I ) . One report on conditions in 1172 says that some
men accused of crimes calling for penal servitude had already been in
jail for more than three months, with no decisions on their cases,3 and an
anecdote preserved in the Sung casebook Che-yu kuei-chien says that a
clerk accused of theft had been kept in jail for three years without having
his case resolved.4 Another report, from the early eleventh century, says
that a man who had already been sentenced to registration had to wait
more than forty extra days after his sentencing while a report on the trial
was sent to the circuit intendant, even though the prefecture involved
was close to the intendant's headquarters. 5
Such lengthy stays ought not have to have occurred if proper pro-
cedures had been followed. Although they probably formed only a small
percentage of Sung cases, in absolute numbers they no doubt affected
many prisoners. The higher authorities, considering them to be highly
unusual and improper, did periodically complain about instances that
were brought to their attention. Despite such official concern, however,
the abuses continued. Thus, although imprisonment was not considered
a form of punishment, the deterrent value of such extended "nonpenal"
stays in jail should not be discounted.
The authorities knew how dangerous time spent in jail could be, that
many prisoners died, that some of them were merely witnesses, and that
others, although accused, had not yet been convicted. The state auth-
orities were, therefore, interested in trying to make life in jail bearable
and the stay there brief.
The thought of jail might be a deterrent to some criminals, and the
time spent there in fact a punishment of the true criminals confined, but
the authorities had practical as well as ideological reasons for not using
imprisonment as a major form of punishment. The traditional Chinese
states were able to secure only a small proportion of the total social
product for their use. The income secured by the state had to be

2 For a good brief description of Sung jails and jail policies, see Tai Chien-kuo, "Sung-tai
te yii-cheng chih-tu" (On the system of jails in the Sung dynasty), Shanghai shih-fan
ta-hsiieh hsiieh-pao, no. 3 (1987): 89-96.
3 SS 200.4994.
4 CYKC 68.
5 SHY, hsing-fa 4.10b.

355
 Law and order in Sung China

committed first to the support of the officials who staffed the government
and to the national defense. Despite theoretical commitments to support
various social services and to achieve other goals, in practice the tradi-
tional governments rested lightly on society. Then, as now, imprisonment
was expensive. Given the state's limited resources, the widespread use of
imprisonment as a penalty would have been a serious drain on revenues.
Thus in the early twelfth century when the authorities, citing ancient
precedents, attempted to use prisons to punish certain classes of crim-
inals, the new policy was soon abandoned. 6

Physical facilities
To help ensure continued decent treatment of prisoners, the state
enacted policies that required officials to inspect the physical condition
of these jails on a regular basis. An approved memorial of 1097 states:
"As to the jails in the prefectures and districts, the incumbent respon-
sible officials are to come once each half-year to conduct personal inspec-
tions and to see to necessary repairs. It is important that these things be
done." 7 Analogous rules seem to have remained in force throughout the
dynasty. The rule was incorporated in the early thirteenth-century code
Ch'ing-yiian t'iao-fa shih lei, which says that the incumbent officials in the
prefectures and districts were to go twice a year to inspect the physical
facilities of the jails and the goods provided to the prisoners. Regular
allotments of money were earmarked for needed repairs and materials. 8
We can get some idea of the physical conditions of these jails from the
occasional comments in Sung sources. In 1075 the emperor ordered that
all units in the capital where men were imprisoned should build beds for
those held. 9 Four years later the emperor, concerned about conditions in
the jails in Kaifeng, ordered that the rooms be increased in size or some
prisoners moved elsewhere. As a result, the jails are said to have been
enlarged by some one hundred bays (chien - an architectural unit defined
as the distance between two upright pillars). 10 About two decades later a
memorial submitted by the Ministry of Justice and the Court of Judicial
Review asked that
all jails build airy towers with cooling windows. Gruel to drink and mats for
sleeping should be provided. The inmates from time to time should be allowed to
bathe and to eat. This is the regular pattern in warm weather. When it is cold
6 WHTK 167.1452.
7 HCP 485.3a.
8 TFSL 538.
9 HCP 270.3b.
10 HCP 246.15b.

356
 Jails and jailers in the Sung

they are to be given appropriate amounts of firewood. The poor may borrow
clothes. In hot weather their fetters are washed once every five days.11
During the early Southern Sung, P'eng Kuei-nien (i 142-1206)
described the completion of a new jail in Wu-ning District. It was "a
building of a number of columns, with a number of jail cells running east
to west, with a separate place for interrogation. The kitchen, bathroom,
and living quarters are in the back." 12 In 1236 Liu Tsai described the
repair of the jail in Chu-chi District in Liang-che Circuit, which
opened up and widened the site, built up the foundations, and strengthened the
walls with stone so that water could not wear them away. The walls were so
constructed that men could not leap over them or cut through them. There were
five cells. Men and women were separated, yet it was all spacious. In the center
was the interrogation room. There were four separate storerooms for cloth-
ing, rations, and prison implements. On the surrounding sides were the living
quarters. Also, the earth was covered with planks to keep out dampness, and the
beds were varnished so that lice could not get in.13
These records were written especially to celebrate the building of proper
jails; unfortunately they also are testimony to their unusual character. A
report from the late Northern Sung probably describes more typical
conditions. A circuit intendant reported that in his jurisdiction, in the
districts located away from the prefectural seat, the local authorities
often merely turned over their prisoners to the nearest units of the
provincial armies for detention, or if there were no units available, they
held people in such places as stores or shops. 14

Locations of jails
All districts in theory had a jail, associated with the offices of the sheriff,
though clearly some districts did not in fact have jails of their own. 15 On
the level of the prefecture the office of the district whose headquarters
was in the prefectural seat had a jail, and the office of the administrator
for public order (ssu-li tsan-chun) had one and sometimes two jails. Chien-
k'ang Prefecture in Liang-che Circuit is said to have had six prefectural
jails, including two associated with military encampments and two dis-
trict jails attached to the district offices located in the prefectural city. 16

11 SS 18.347.
12 P'cng Kuei-nien, Chih-t'ang chi, 10.129.
13 Liu Tsai, Man-t'ang wen-chi, 23.27a-29a.
14 SHY, hsing-fa 2.63b.
15 SHY hsing-fa 2.63b.
16 Chou Ying-ho, Ching-ting Chien-k'ang chih, 1057-58.

357
 Law and order in Sung China
Any circuit intendants in the prefecture also had jails where prisoners
might be held for interrogation during case reviews. 17
The imperial capital, Kaifeng, had no central jail. Rather, at various
times a number of central government organs had their own jails. 18
The chief urban police organs, including the Palace and Metropolitan
Cavalry and Infantry commands (Tien-ch'ien ma-pu chun ssu) and
the various patroling inspectors' offices, had jails attached to their
offices. Judicial organs, like the Court of Judicial Review and the
Censorate had jails, and so did such nonjudicial nonpolice offices as the
Bureau of Water Transport (where they held transport laborers accused
of crimes), 19 the Four River Transport Offices (Ssu p'ai an), 20 the
Finance Office,21 and other units. 22 Furthermore, the offices of the sheriffs
of the metropolitan districts presumably had jails, as did the prefec-
ture. Kaifeng Prefecture's jails included jails in the General Executive
Inspector's Office (Ssu-lu ssu) and the Left and Right Courts of Military
Inspection (Tso-yu chiin-hsun yuan). 23 We may also assume that the
armies stationed around the capital had jails.
These Kaifeng jails had been of special importance at the beginning of
the dynasty. At that time the Court of Judicial Review did not have its
own jail, so its prisoners were held in these prefectural jails. Because of
their growing numbers, the problems of crowding and illness in these
Kaifeng jails increased. In 1078 Emperor Shen-tsung ordered the High
Court jail reopened. It was to be staffed by one chief, two assistants, and
four aides, who were to have sole control over interrogations. In addition,
this jail had associated with it one recorder and two law-checking officials
(chien-fa kuan).24 The jail was again closed in 1088 because of a variety of
abuses, 25 only to be resurrected in 1096 (when it was organized after the
pattern of the Censorate's jails). 26 The capital also had the east and the
west jails of the Censorate, which in 1079 were located close to the Altar
of Grain. 27 These capital jails were supposed to be inspected periodically
and in rotation by officials from the Ministry of Justice and the Court of

17 HCP 71.19b; Wen T'ien-hsiang, Wen-shan hsien-sheng ch'uan-chi, Wan-yu wen-k'u ed.,
9.292-293 (1259).
18 SS 201.5021.
19 HCP 52.17a (1002).
20 HCP 52.17a.
21 HCP 65.17a.
22 HCP 52.17a, 295.8a-b; WHTK 167.1450.
23 HCP 57.9a, 12306.17b, 324.9b.
24 SS 201.5021; HCP 295.8a-b; WHTK 167.1449.
25 HCP 418.2b; WHTK 167.1450.
26 HCP 410.6a; SS 201.5023.
27 HCP 2 9 8.3b- 4 a.

358
 Jails and jailers in the Sung
28
Judicial Review. The great number of jails was advantageous in that
no single one needed to be too large for reasonable security. But the
large number could also cause problems, since it allowed officials worried
about their supervisors' visits to conceal prisoners by transferring them
to another jail.

Competing concerns and jail policies

The policies specifically concerned with the structure and functioning of
jails seem to be products of the sometimes competing concerns for
security, the humane treatment of inmates, and the desire to limit
corruption and abuses by jail functionaries and officials. Because those
in jail might be dangerous and desperate, the authorities tried to design
facilities that would prevent escapes, run them according to policies
that would ensure decent treatment, and restrict the opportunities and
temptations for officials and jail functionaries to abuse their powers.
Central government officials repeatedly expressed their concern about
the welfare of those held in jails. Some were innocent witnesses, who
certainly did not deserve to suffer. And the underlying voluntarist view
of human nature common to most members of the Sung elite led them to
believe that many criminals could be salvaged, could in the standard
phrase "renew themselves." If this were true, it behooved the authorities
to treat prisoners with compassion, in hopes of their redemption.
This recognition of the inmates' moral potential was also reflected in
the policy of separating men from women and supposedly of separat-
ing minor offenders from hardened criminals. 29 Women were kept in
separate cells with female guards. 30 By law these women prisoners were
allowed to have a female servant girl accompany them. 31 The text says
literally a "singing girl," that is, a prostitute. Possibly the authorities felt
that such women would be better able to deal with the surroundings of
the jails than would ordinary servants. The policy of separating women
from men was continued throughout the dynasty. 32
In his twelfth-century book of advice for local officials, Li Yuan-pi
underscores the importance of detaining minor offenders in a different
place from serious criminals. He advises the district magistrate to build
what he calls "upright cells" (li-hsiab) for petty offenders. These prob-
ably were like the "tiptoe" cells of the Ch'ing dynasty (1644-1912),

28 SS 199.4992, 201.5020.
29 SHY, hsing-fa 6.51a, 52b.
30 SHY, hsing fa 6.51a, 52b.
31 TFSL 537.
32 TFSL 537.

359
 Law and order in Sung China

I
Figure 11.2. Tiptoe cell. From Wu Hsiang-chu, Tien-shih chai hua-pao.

from which the head of the offender protruded (see Figure 11.2).33 The
separation of hardened criminals from minor offenders, however, was not
always observed. A report of 1006 suggests that the brutality in prison in
ancient China was all too similar to that in our own prisons. We are told
that the hard-core criminals "coerce younger prisoners to feed them their
food, so that the younger inmates are often hungry and thirsty. Without
asking whether or not the crimes were serious, both sorts of criminals are

33 TITC 2.9a.
360
 Jails and jailers in the Sung
held in the same jails." This complaint was followed several months
later by an order that serious and petty criminals be kept in different
cells.34 We can only assume that the kinds of physical and sexual abuses
found in modern prisons also occurred in Sung jails. Perhaps the fetter-
ing of serious prisoners reduced the opportunity for predatory behavior.
In any event, the sources are silent about this side of prison life.
In later dynasties some of these problems were mitigated by having
several sorts of facilities for holding prisoners. Huang Liu-hung, writing
in the seventeenth century, says that in his time, jails by tradition were
divided into four levels, from the "soft prison" for relatively minor
offenders, through the "outer prison," the "inner prison," to the "black
dungeon" that housed bandits awaiting execution. 35 In the eighteenth
century the official Fang Pao, in his "Notes on Prison Life" also refers to
dividing the prison of the Ministry of Justice into levels, to segregate
prisoners, ostensibly on the basis of the seriousness of their crimes,
though in practice in part according to the amounts they paid to the
jailers as bribes. 36
The concern among Sung officials about abuses in jails, as well as
about theft by jail personnel, led to a policy of keeping a register to
record the new inmates' clothing and personal possessions, possessions
that would be returned to them (and checked off the register) when they
had been judged. Under the rules found in the Ch'ing-yuan t'iao-fa shih-lei,
when a criminal arrived at a jail bringing goods with him, the officials
were to inspect them and make a list giving the names of the prisoner
and his possessions. They were then to be sealed and put into a store-
house for safekeeping. If the accused was found innocent, the goods were
to be returned on that same day. They were not to be used as rewards or
confiscated by the authorities. If the jail functionaries on their own
authority distributed goods that had not been inspected by officials, they
were to be punished with eighty blows of the heavy rod. And if they took
advantage of the situation to steal the goods, they could be tried for
theft.37
We do not have for the Sung the same sorts of descriptions of jail life
that we have for the Ming and later, but occasional comments by
officials do hint at the quality of prison life. The general practice seems
to have been that when they could afford to do so, the family of the
prisoner provided food, drink, and other necessities, a practice that

34 HCP 63.10b; SHY, hsing fa 6.52b.

35 Huang, Complete Book, p. 311.
36 Derk Bodde, "Prison Life in Eighteenth Century Peking," Journal of the American Oriental
Society 89 (1969): 311-31.
37 TFSL 537.

361
 Law and order in Sung China

continued under later dynasties. T h e prison officials were responsible for

seeing that the materials actually reached the prisoner. 3 8 Indigent
prisoners were to be provided for by the state. According to an edict of
the Later Chou ( 9 5 1 - 6 0 ) , which was incorporated in the early Sung
Code, indigent prisoners were to be given two pints of official rice per
day. 3 9 This was reaffirmed by a decree of 986. 4 0 Dangerous prisoners
might have food brought to them in their cells. Less serious prisoners
might be let out to eat. 4 1 T h e provision of food to the indigent was in all
probability the one service to prisoners that was provided regularly. In
descriptions of prison life from the Ch'ing, commentators do mention the
regular system of rations. Even in this respect the Sung decrees in regard
to feeding prisoners may not in fact have been enforced in all jails,
and where food was provided there were still complaints that the func-
tionaries substituted inferior grain. 4 2 Toward the end of the Northern
Sung, Li Yuan-pi advised potential magistrates to "inspect the prisoners'
food with special attention. Reductions are not to be allowed. T h e
provisions in the jail should be checked once every few days. Extra
[rations of] beans and rice should not be provided." 4 3
It seems possible that during the Southern Sung, prisoners were sup-
posed to be given not only grain but also supplementary money for salt
and vegetables. T h e Southern Sung handbook for local officials called
the Chou-hsien t'i-kang says that prisoners were to be given two pints
(sheng) of rice daily and ten coins of "salt and vegetable money." This
amount of rice was about the standard Sung daily ration. It seems
possible that the rice was regularly provided but less likely that prisoners
always received the supplementary money. According to a decree of
1143, indigent prisoners in the capital were to be given twenty coins per
day for food (which would have totaled more than seven strings of cash
per year), and those in the outer prefectures were to be given fifteen
coins (about five strings per year). 4 4 Since the local constables were
being paid only about thirty strings per year, it seems improbable that
such funds were in fact disbursed on a regular basis. Rice, on the other
hand, probably was regularly provided. Information from K'un-shan
District in Su Prefecture in 1251 suggests that in that district the admin-
istration's income in grain included a portion earmarked for prisoners. 4 5

38 SHY, hsing-fa 6.51a.

39 SHT 29b.
40 STCLC 742.
41 CHTK 3.27.
42 HCP 52.17a; SHY, hsing-fa 6.6oa-b.
43 TITC 2.26.
44 SS 199.4993. For a report on providing food, see HNYL 2589 (1149).
45 Ling Wan-ch'ing, Yufeng chih (SYTFCTS ed.), 3814. A matter of some 90 shih, 4 sheng.

362
 Jails and jailers in the Sung

Prisoners were also to be given on occasion appropriate drinks to pre-

serve their health. In line with the current government policy, Li Yuan-
pi in the late Northern Sung advised magistrates to see that prisoners
were kept warm in winter and cool in summer and allowed to bathe
regularly. In winter, when they were going back into their cells, they
should be given a cup of hot boiled water to drink and in summer a
draught of freshly drawn water. 46
The authorities were also responsible for providing for the other needs
of indigent prisoners. The Sung Code of 962 contains a rule stating that
prisoners were to be given mats, drugs for illnesses, and water for
washing, 47 as well as clothing, including, in winter, padded jackets,
robes, trousers, stockings, and gloves. In addition to food and clothing,
prisoners supposedly were given firewood or coal during the coldest
weather, though there are complaints that the prefectures and districts
often ignored this rule. 48 Despite these rules and the examples from some
local areas, it is clear that sometimes the prisoners were not adequately
fed or always provided with other necessities. In some cases prisoners
were said to have been deliberately denied food by the responsible
personnel in efforts to get them to confess,49 a practice that sounds
remarkably like the treatment of prisoners in the People's Republic
described in Prisoner of Mao.™ It seems probable that hunger contributed
to, if it did not itself cause, some of the prisoners' deaths in jails. Early in
the thirteenth century an official complained that
the prisoners in the districts suffer bitterly from a lack of food. This is at its worst
in the district subordinate to the city. In law the officials are allowed to spend
transportation money [on prisoners' needs], but often the districts do not dare to
do so and so violate the precedent. Often those who share control in the jails
privately take food from the households of corvee laborers, dividing it among the
incarcerated men. When the food comes in, the men grab for small pieces. How
pitiful!51

Health care
Health care was a particular concern. In 932 under the Later T'ang an
edict had ordered each prefecture to set aside a jail hospital room (ping-
ch'iu yuan) for serious offenders. The Sung reaffirmed this policy at the

46 TITC 2.32, 2.33, 2.34, 2.36.

47 SHT 29.8a; see also, from 986, STCLC 742.
48 SHY, hsing-fa 6.51a, 56b~57a; SS 18.347; HNYL 2728 (1154).
49 SS 200.4996—97.
50 Bao Ruo-wang (Jean Pasqualini), Prisoner of Mao (New York: Penguin Books, 1976).
51 WHTK 167.1455-56.

363
 Law and order in Sung China

beginning of the dynasty, in 996, and again in 1101. According to the

request of 996, local authorities were to open clinics for prisoners being
held for armed robbery calling for sentences of penal servitude or worse.
Lesser offenders were to be released on bond for treatment outside
the jail 52 (though at times it is said that some were unwilling to go
out). 53 The policy was again affirmed in the Southern Sung and indeed
remained standard practice under later dynasties. 54
According to the jail functionary ordinances cited in the early Sung
code Sung hsing-t'ung, when a prisoner was ill, the functionary responsible
was to inform the chief local official, who was to make a personal
inspection. If the prisoner really was ill, then the official was to see to the
provision of medicines and the care of someone with medical knowledge.
Prefectures were supposed to have three such medical attendants, and
districts one. These men seem to have been local individuals serving in
this capacity as one form of requisitioned state service. We are probably
safe in assuming that most were professional medical practitioners. 55
These men were to keep a log describing the symptoms of the prisoner's
illness and its daily progress. 56 If seriously ill, the prisoner was to be
freed from fetters. One member of the prisoner's family was to be
allowed to enter the jail to help with the care. 57 An early twelfth-century
order specified that ill prisoners were to be fed gruel made from fresh
white rice. Local authorities were forbidden to substitute cheaper
grains.
The authorities' continuing interest in this problem is revealed in a
decree of the mid-twelfth century that all the jails of the Court of Judicial
Review and the capital prefecture should receive one hundred strings of
cash to use for prisoners' medicines. Other prefectures were to receive
sixty strings, the Three Commands (san-ya) (the Palace Command, the
Metropolitan Cavalry Command, and the Metropolitan Infantry Com-
mand, which shared responsibility for security in the capital) were to
receive fifty strings, large districts thirty strings, and small districts
twenty strings. 59 These rules, phrased in terms of general health care,
also probably were designed to help prisoners who were suffering the
aftereffects of judicial or interrogatory torture.

52 SHT 3.51; HCP 48.8b; WHTK 166.1445; see also SHY, hsing-fa 6.52a-b.
53 Li Kuang, Chuang-chien chi, i2.2ia-b.
54 HNYL 2619. For later dynasties, see Huang, Complete Book, p. 311.
55 Liang K'o-chia, San-shan chih, i4.hsien-i jen.
56 TITC 5.28b.
57 SHT 29.472; HNYL 2619; SS 199.4972.
58 SHY, hsing-fa 6.60a-b.
59 HNYL 2639. The money for such purposes was to be drawn from various sources,
including money from fines or from regular official funds; see TFSL 74.

364
 Jails and jailers in the Sung

Of course, we cannot assume that all such rulings were always enforced.
For such systems to work they required the support of energetic and
sympathetic local officials. In the late Northern Sung an official noted
that "local officials may not be thorough in carrying out Your Majesty's
merciful intentions" and, in typical bureaucratic fashion, asked that the
circuit intendants be required to submit annual reports on the number of
prisoners who had died of illness.60 Those outstanding local officials who
promoted the humane care of prisoners were held up as exemplars. The
gazetteer for T'ai Prefecture in Liang-che Circuit described the building
in 1211 of prefectural facilities for housing sick prisoners:
Before this, sick prisoners had no place to stay. They were scattered to sleep at
places in the city buildings such as temples or Buddhist dwellings, and many
died. Huang [Shou-sun] thereupon built these facilities so that they would have
a place to sleep. All their daily needs were provided for. . . . They received rice,
salt, medicine, and other foodstuffs, all paid for with money collected in fines. In
cold weather they were given clothing and firewood.61
The most complete set of rules on the treatment of sick prisoners is
found in the Ch'ing-yiian t'iao-fa shih-lei, which indicates that when a
prisoner became ill, this was to be reported immediately to the officials,
who were supposed to check out the situation in person. They were also
supposed to recheck the situation every three days. Relatives of the
prisoners were also to be informed so that one of them could come to
care for the prisoner. If the prisoner were an official of the fourth rank or
higher, or a woman holding official rank of enfeoffment with the district,
two persons could come, including wives, sons, or grandsons. The jails
were to have registers listing the names of their medical practitioners.
Substitutes were not allowed. The jails kept a record in which they were
to note the ill prisoner's symptoms, the names of the medical practi-
tioners used, and whether the patient was cured or died. This rule also
covered those who had been released on bond for treatment outside the
jail if they died within ten days. If those let out for treatment died after
this period, their names, the dates, and the cause of death were to be
recorded. The incumbent officials signed these registers, which were
reviewed annually. If these rules were not followed and many prisoners
died, the responsible personnel were to be punished according to the
rules for abusing prisoners. The authorities were all too aware of the
possibilities of abusing prisoners and ruled that if the circumstances of
an incident of illness seemed very serious, an uninvolved official should

60 Mu-jung Yen-feng, Li wen fang chi (Sung ming-chia chi-hui ed.), 10.2b.
61 Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, 7104.

365
 Law and order in Sung China

be dispatched to bring a medical practitioner under guard to perform an

examination. 62
The system portrayed in this early thirteenth-century code is similar
to the system found in early Sung materials, with a few significant
differences. During the early Sung, criminals liable for penal servitude or
worse were treated in jail, but by the early thirteenth century they could
be allowed out of jail on surety, unless the circumstances of their crimes
were quite serious. The Southern Sung materials also do not mention the
special hospices mandated in the Northern Sung, which may help ex-
plain why such prisoners were allowed to leave the jails for treatment.
Finally, one partial answer to prisoners' health problems was public
health measures, such as giving prisoners reasonably cool and suf-
ficiently large quarters and trying to avoid the more obvious causes of
infection. In this regard, the Sung always continued the policy from
earlier times of periodically washing down the fetters in which men were
being held. 63 An early description of this system, from the summer of
969, says:
The emperor thought about the great heat and the sufferings of those imprisoned.
He issued a decree that in the various prefectures of the western capital the chief
officials who managed the jails were to make an inspection every five days, to see
to the sweeping of the cells and the washing down of the fetters. Those [prisoners
who were] poor and unable to sustain themselves were to be given food and
drink, and those who were ill were to be given medicine. Those accused of minor
crimes were to be dealt with immediately so that there would be no delay. From
this time on, every summer, during mid-summer, there was a clarification about
this decree to warn officials [not to forget it].64
Some years later similar orders were issued regarding prisoners held in
the Four River Transport Offices'jails {ssu-p'ai an)65 and orders regard-
ing sweeping jails, washing fetters, and serving food were occasionally
repeated. We have examples from 981, 986, and 1003.66 The system
established for the prefectures was extended to the circuits when in 1005,
they built centers for interrogating prisoners. 67 In 1090 Fan Tsu-yii
reported:

62 TFSL 512-13.
63 For this earlier practice, see SHY 29.472 (932).
64 HCP 10.7b; STCLC 740; SS 199.4968.
65 HCP 52.17a.
66 HCP 55.11b. HCP 56.12b says every three days, but I suspect that this is simply a
miscopying for the word Jive (see also STCLC 740, 742).
67 HCP 60.16b. This passage suggests that the pattern was every ten rather than every
five days, as it says that the practice should involve "washing them every ten days after
the pattern of the district jails."

366
 Jails and jailers in the Sung

In accordance with the edicts of the Yiian-yu period [1086—93], during the hot
months of summer the fetters on prisoners are to be washed once every five days.
The prisoners are to bathe at this time. . . . I have looked into the ancestral
system. Every year, in winter and summer, decrees were sent down ordering that
prisoners be treated with mercy in punishing. From 986 during the time of T'ai-
tsung [onward] this has been done continuously. In 1070 the Bureau for the
Compilation of Edicts memorialized asking that each year, from the fourth
month to the tenth month, the various circuits' judicial intendants inspect
and certify that the various prefectures' chief officials [have complied with this
system]. 68

Finally, one source suggests that prisoners were allowed out of their cells
one day in ten. This would allow the authorities to clean the cells, to
examine the health of the prisoners, perhaps to find men who ought to
be released, and presumably to allow the prisoners an opportunity for
exercise.69

Abuse of prisoners
The authorities worried about abuses and corruption. Writing in the
seventeenth century, Huang Liu-hung complains about a problem that
had never been solved:
Once a prisoner is sent to jail, his life is in the hands of the jailers. Some
prisoners are killed by inhumane treatment at the hands of jailers who try to
extort bribes without success. Other prisoners are killed by their accomplices
who conspire to get them out of the way in the hope that their own cases will be
settled sooner. Some prisoners are simply murdered by jailers who have been
bribed by their enemies. There are also cases in which rapacious officials murder
the prisoners to silence them and to appropriate their stolen property. When a
notorious and dangerous criminal is jailed, officials sometimes fear that he might
escape and cause further trouble, and so they simply have him killed, and report
to their superiors that he died of illness.70
Sung prisoners, like these prisoners during the Ch'ing, were sometimes
deliberately killed by their jailers. 71 The Sung History says that the auth-
orities themselves would
break the hands and feet (of prisoners) and chain them in the sheriff's stockade.
The powerful could bribe the clerks to charge ordinary people falsely and then
have them killed in prison. It reached the point that all trials on civil matters
68 HCP 443.21b—22a. The official Li Kuang (1078-1159) advocated a similar system: See
Li Kuang, Chuang-chien chi, I2.2ia-b.
69 C H T K 3 . 3 1 .
70 Huang, Complete Book, pp. 308-9.
71 SS 200.4995, 4997; Chang Kang, Hua-yang chi (SPTK ed.), 14.1a.

367
 Law and order in Sung China

also were oppressive. Food and drink were not given, and so the prisoners died of
starvation. They had no way to seek help against clerks who so maltreated them
that they died. There were cases of suits involving two parties in which, because
of bribery, one of the prisoners would be cruelly beaten to death. Fearing
disclosure, the clerks would first send up an illness report, saying that the
prisoner was ":under a doctor's care" when he was already dead; then they
would say that [the prisoner] had "died of illness" when in fact he had been
murdered. 72

Abuses seem to have been common, even in the capital city. One
special torture device used in the capital is said to have so crippled its
victims that after they had been tortured they were unable to raise their
hands. 73 These Sung practices are similar to those so vividly described
by the seventeeth-century magistrate Huang Liu-hung. He even gives
the slang names for the abuses. A warden beating a new prisoner with
help from the other jailers was called "sharing a community dish."
Pouring water on the ground and forcing a prisoner to sleep in it was
called "wearing a wet shirt." Hanging a prisoner by his feet with his
head hanging down was called "climbing a tall tower." 74 Although we
have less information from the Sung, the names used for Sung tortures
suggest that the jailers were no less inventive in their tormenting of
prisoners.
Jailers certainly had ample opportunity to abuse prisoners, since they
administered both the pretrial beatings, which were legally permitted
during the investigation, and the beatings used during the trial. The
legal interrogatory beatings were limited to twenty blows of the heavy
rod, with civil service personnel as observers. In practice they often left
this matter to the clerks, who abused their powers.75
Sometimes the clerks inflicted beatings without informing the officials.
An order of 1023 says that jail personnel who beat prisoners without
having informed the senior officials could be tried for having violated
regulations, which carried a penalty of two years of penal servitude.
However, if those so beaten turned out to be guilty, this would mitigate
the punishment. Should innocent people die as a result of such abuse,
the responsible functionaries would be liable for prosecution for delib-
erate homicide.76
The central authorities periodically reaffirmed the importance of hav-
ing the officials personally present during interrogations with torture. 77
72 SS 200.4997.
73 SHY, hsing-fa 5-2b-3a.
74 Huang, Complete Book, p. 310.
75 HGP 335.10b, 376.5b-6a; HNYL 2544, 2857.
76 HCP ioi.8b-9a.
77 Compare STCLG 741, 742, 744; WHTK 166.1444.

368
 Jails and jailers in the Sung

Officials who were not personally present but had their clerks do the
torturing could be liable for a penalty equivalent to two years of penal
servitude. 7 8 In the mid-twelfth century the central authorities even
ordered the circuit intendants to have the rules on judicial torture carved
on placards that were to be hung in front of prefectural and district jails,
perhaps to alert those who might be subjected to torture of the limits on
its application, an order that also indicates the high level of literacy
among Sung men. 7 9
Sung central authorities saw corruption and the abuses associated
with it as primarily a problem among the clerical-level jail functionaries,
who could manipulate the jail situation in their search for bribes. T h a t
is, they might deliberately involve innocent people in cases, in order to
squeeze money from them.
T h e broader problem of abuse was not confined to clerks. It also often
involved civil service officials. In the case of officials it was perhaps more
often associated with excessive brutality toward inmates than with seek-
ing bribes. T h e Northern Sung official Mu-jung Yen-feng wrote:

In the circuits and in the capital the servitors who are charged with capturing
bandits, although they have the duty of capturing bandits, sometimes because of
momentary doubts seize witnesses who are not guilty and bind them. Under
interrogation, many out of fear make false confessions and are sent to be jailed in
the offices. Initially on the basis of the confession statement, they are fettered and
subjected to judicial torture. Because of the time consumed in transmitting
documents and so on, they are held for months.. . . Eventually they are released
because they are innocent or convicted of some minor crime. 80

T h e situation was said to be even worse in the early thirteenth century.

T h e censor Hsu J a n complained that "in recent years the officials in
the prefectures and districts have been treating the jails like execution
grounds. Those implicated in capital crimes or witnesses in violent
robberies are bound and fettered. They are illegally subjected to judicial
torture. They are arbitrarily imprisoned. This is brutal oppression." 8 1
In this same era it was the rule that officials who had been guilty of
abusing prisoners could subsequently be appointed only to minor service
agent positions. (Service agents, who oversaw local tax collection depots,
were at the very bottom of the official status pyramid.)
According to the general rule, if prisoners died as a result of deliberate
abuse by officials, the officials could be tried for deliberate homicide.
Even if the deaths were not deliberate and the officials were simply
78 WHTK 167.1453.
79 HNYL 2832.
80 Mu-jung Yen-feng, Li-wen t'ang-chi (Sung-ming-chia chi-hui ed.), 10.6b-7a.
81 SHY, hsing-fa 546a-b.
369
 Law and order in Sung China

trying to elicit information, they could still be tried for deliberate homi-
cide, with the resulting penalty reduced by three degrees.82 Those who
assigned others to investigate or did not get their fellow officials to
participate in the interrogation but simply conducted it themselves and
did not take into account the evidence, were liable for a penalty of sixty
blows of the heavy rod. 83 However, some officials felt that the rule was
insufficiently exact, so in 1158 it was decreed that officials who from
personal motives jailed the innocent, caused prisoners to die, beat them,
broke the laws on making judgments, or illegally used the fetters or other
judicial instruments so as to cause injury or worse all were to be held
personally responsible. 84
Reports on abuses only occasionally tell why functionaries abused
prisoners. Presumably it is partly a question of the kind of petty tyranny
that happens in jails everywhere; it also often seems to have involved
attempts to extort money from prisoners or their families. And some-
times it is said that enemies of those jailed paid the jailers to injure the
prisoners.

Deaths of prisoners
Despite expressions of official concern, the problem of illness and death
among jailed prisoners was a serious and continuing one. The higher
authorities were understandably troubled about the abuses. Deaths
among prisoners was one of the categories of deaths that automatically
called for an inquest. Starting in 1034 the judicial intendant was sup-
posed to investigate cases in which it was suspected that prisoners had
died from mistreatment by jail personnel. The authorities seem to have
been trying seriously to stop such abuses. They offered large rewards,
one hundred strings of cash, to nonofficial informants who reported such
abuses, or an additional full term in office equivalent (jen) to officials
who did so. 85 At least after 1067, annual reports were to be compiled at
year's end indicating the number who had died in jail. 86 From the
Ch'ing-yuan Viao-fa shih-lei we learn:

Officials should be dispatched to conduct inquests on all dying people who have
no relatives of the ssu-ma degree or closer [i.e., relatives who were to do three
months' mourning, the lowest group in the hierarchy of mourning grades] and
on all dead prisoners. (Clerical personnel may be sent if the prisoners are slaves

82 SHT 29.478.
83 SHT 29.474.
84 HNYL 2973.
85 HCP 114.16b; SHY, hsing-fa 6.55a.
86 SS 14.267.
370
 Jails and jailers in the Sung

and oral depositions have already been made.) In the cases of the deaths of
prisoners and deaths from unnatural causes, a reinquest should be performed.
When the inquest and the reinquest have been completed, the corpse may be
buried. If relatives are known to live in other places, they should be informed. 87

The problem, however, continued. Early in his reign, Shen-tsung (r.

1068-85) complained about the number dying. He attributed this to
corrupt jail personnel and to the lack of thorough investigations of such
incidents. The emperor issued an order that if in the prefectural jails two
prisoners sickened and died, or if in a prefecture with five or more
districts three men died in a year, or if seven men died in the offices at
the capital, the jail personnel were to be beaten sixty blows with the
heavy rod. This was to be increased by one degree for each additional
death, to a maximum of one hundred blows. The prison officials were
responsible for investigating. If there were two such offenses, they them-
selves were to be tried for "breaking regulations" (wei-chih), a crime
carrying a penalty equivalent to two years of penal servitude. 88
According to a report of 1093, before that time the local units did not
have to report deaths of less than 5 percent of their prisoners. This
practice was ended by decree. 89 It seems that thereafter the officials and
clerks would be penalized if the death rate were 1 percent. However, the
maximum penalty, one hundred blows of the heavy rod, was said to be
insufficient to discipline the clerks. 90 The seriousness with which higher
authorities viewed this continuing problem of high death rates among
prisoners is suggested by rules such as the one promulgated in 1174
under which those functionaries to be punished could not commute their
sentences by the surrender of their offices, nor could they benefit from
amnesties. 91 Even so, such abuses and problems continued until the end
of the dynasty, despite reported increases in the penalties used against
the clerks.92
Because officials could be penalized for having too many deaths in jail,
they were tempted to conceal them. Clerks would report prisoners as ill,
and the districts would hide their deaths. 93 From early in the dynasty it
was ordered that the officials conducting prison inquests must be from
other offices.94 We know that later in the dynasty in the capital area,

87 TFSL 513.
88 SS 201.5021; WHTK 167.1448; SHY, hsing-fa 6.s6a-b.
89 SS 201.5023; WHTK 167.1451.
90 Chang Kang, Hua-yang chi, i4.ia-2a.
91 SHY, hsing-fa 6.70a; TFSL 74.
92 SS 200.4997; SYTC 92.12b. The penalties inflicted were reportedly increased in 1195.
93 Compare Chang Kang, Hua-yang chi, i4.ia-2a (1083-1166).
94 HCP 76.10a. This order initially applied specifically to the prefectural Public Order
Office (Ssu-li yuan) jails but presumably was followed elsewhere.

371
 Law and order in Sung China

Censorate officials were sent to conduct such inquests. 95 Like inquests in

general, on prisoners presumably involved a physical examination of the
corpse in the presence of the officials and the writing of an official report
to be signed by those responsible.
The death rates were sometimes frightful. From 1135 we have a report
in which the worst prefecture is said to have had a death rate of 26
percent. 96 Among criminals accused of violent crimes, the death rate
might have been even greater, perhaps because of abuse by jail func-
tionaries. In 1139 a censor reported on a case of piracy in Lei Prefec-
ture in Kuang-nan West Circuit for which seven commoners had been
arrested. Five of them had died in jail. 97 In another incident from the
early Southern Sung, thirty-four of forty-two men seized and accused of
robbery died in jail. "The circuit intendants felt that the clerks had
accepted bribes to practice coercion, imprisoning men who had been
forced into joining the bandits, and causing them to die. They then
altered the case reports to claim that these men had been among the
principal bandits." As a result of the scandal caused by this case, a law
was enacted under which prisoners who died before the final disposition
of a case could not be counted as captured criminals for purposes of
calculating rewards (thus giving jail functionaries an incentive to keep
prisoners alive). 98 We do not have any complete reports on the numbers
who died, but one document from the thirteenth century does give us
some idea of both the number of prisoners handled by units on an
annual basis and the number who might die. Because the figures were
being used among other things to demonstrate how well jails could be
run and to specify the ratios of deaths that would lead to penalizing the
responsible officials, we cannot assume that the figures are typical, but
they are illustrative (see Table 11.1).
The higher authorities did try to keep a check on the number who
died. In the late Northern Sung the official Mu-jung Yen-feng suggested
a policy under which at the end of each year the intendants' offices were
to receive a complete list of those who had been ill or died. They would
then compare the records of the prefectures and districts under their
jurisdiction, demoting the officials in those units with the highest death
rates, promoting those with the lowest,99 and reporting the number of
prisoners who died in their jurisdictions. 100 A similar proposal was

95 HCP 475.1b.
96 SS 200.4993.
97 SS I99-4993.
98 HNYL 1630.
99 SS 200.4992; see also Hsiieh Ying-ch'i, Sung-Yuan t'ung-chien, 105.3b.
100 SS 201.5021; HNYL 1419; WHTK 167.1455.

372
 Jails and jailers in the Sung

Table 11.1. Numbers of prisoners and of deaths from disease

Date Place Number of prisoners Deaths from disease

1135 Hsuan
Hsiian Prefecture
Prefecture 355 0
Ch'u Prefecture 618 0
Su-sung District 7 1
Wu-i District 72 4
1136 Prefecture
Lin-an Prefecture 1634 0
Chiang-yin Prefecture 74 4
Yang Prefecture 122 12
1137 Prefecture
Fu Prefecture 682 0
Wu-p'ing District 40 2

Source: Information from SHY, hsing-fa 6.65a-66a.

suggested in a memorial by Chang Kang (1083-1166), 101 and we know

that at least from the 1130s the prefectures and districts were supposed
to submit to the circuit monthly reports giving the names and the
number of men held in their jails.

Security in jail
As noted, the control of prisoners was made easier by keeping the more
dangerous prisoners in fetters. Men guilty of capital crimes were to be
held in fetters and manacles, though some categories of disability could
excuse their use. When prisoners returned to their cells at night, these
dangerous criminals would have their fetters chained to a support. Bells
were attached to these chains, so that the prisoners' movements could be
heard by the guards. Li Yuan-pi tells us more about the security and
daily living conditions in such places when he suggests that privies
should not be set up in the jail area because criminals could escape
through them. He suggested simply providing buckets for the prisoners. 102
The district magistrate was advised to see that the inner and outer walls,
doors, and windows were regularly and carefully inspected.
For security reasons, paper, writing brushes, drinks, metal knives, or
clublike objects were not allowed in the jails, nor were prisoners allowed
to wear belts (perhaps to prevent suicide by hanging). 103 Such restric-

101 See Chang Kang, Hua-yang chi, n . i a - 2 a . Chang himself traced the general idea for
such a system to the Han emperor Hsiian-ti.
102 TITC 2.9b.
103 SHT 29.472.

373
 Law and order in Sung China

tions continued in force throughout the dynasty. According to the rules

found in the Ch'ing-yuan t'iao-fa shih-lei, prisoners were not allowed to
have gold, knives, wine, paper, writing brushes, money, porcelains, or
staffs with them in prison. 104 The porcelains and knives were presum-
ably feared as potential arms, the paper and writing brushes as enabling
communication with outside gang members, and the gold and money as
materials to be used in bribing jail personnel.
The authorities also worried about security issues involving jail per-
sonnel. Li Yuan-pi indicates that there was a policy of avoidance. If a
prisoner were related to member of the a jail staff, he was to inform his
superiors and would be excused from management. 105 The authorities
were also concerned about the staff s becoming lax in performing their
duties. These functionaries were not allowed to drink. And we are told
that they were not allowed to use mosquito netting, the officials perhaps
fearing that if the functionaries were too comfortable at night, they might
fail to be on their guard. When the jails were closed at night, a single
tablet was to be used, listing the names of those currently incarcerated.
Each evening a report was to be turned in verifying that the prisoners
had been secured in their cells. Overseeing this lockup was supposedly
the personal responsibility of the local administrator. 106 Prisoners were
supposed to be returned to their cells at three rounds after the beginning
of the first watch (which started at 7 P.M.), with the cells being opened
again at five rounds after the beginning of the fifth watch (about 6:30
A.M.). Functionaries who broke these rules regarding lockup could
receive eighty blows of the heavy rod. 107

Jail staffing
The staffing of these jails followed the pattern of other lower-level gov-
ernment services during the Sung. Staffed in the past by unpaid people
levied from the local population, the jails gradually came to be staffed by
long-term personnel paid either by the state or with money squeezed
from those who needed their services. The case of T'ai Prefecture is
illustrative. Early in the dynasty the prefectural jail had drafted house-
holds to provide members to serve as adjutants (chieh-chi) for two-year
terms. During the reforms of the Hsi-ning period (1068-75) these men
104 TFSL 537.
105 TITC 5.28b.
106 TFSL 537.
107 TFSL 537; WHTK 167.1454. In the winter, from the tenth month to the second
month, the jails were opened at the third round after the fifth watch. I am presuming
that the Hen in this passage refers to the five rounds that the watchmen were supposed
to make during each two-hour watch period.

374
 Jails and jailers in the Sung
were displaced by men from fourth-grade households (in the five-grade
system, i.e., not ordinary peasants, but not the more wealthy, either)
who were paid to serve one-year terms. During the anti-reform period
that followed the death of Shen-tsung in 1085, the jails used "unpaid
volunteers," probably long-term servicemen who squeezed a living out of
their charges. This policy remained in force until the early years of the
twelfth century when for a brief period, men, rotated in for tours of
duty from other clerical posts, were paid "heavy salaries." When this
salary system was stopped during the late Hsiian-ho period (1119-25),
the system presumably returned most of the time to using "unpaid
volunteers," though there is an order dated 1171 that capital jail func-
tionaries be paid a total of ten strings per month plus six pecks (toub) of
rice. 108
When these jail functionaries were paid by the state, the expenses
must have been quite heavy. In the early Southern Sung the jail quota in
T'ai Prefecture was seventy-three functionaries; at the time the gazetteer
was written (1223) there were still seventy men employed. 109 If these
jailers were paid approximately as well as were similar functionaries in
the capital, their salaries would have totaled more than seven hundred
strings per month, exclusive of payments in kind. These prefectural jail
personnel were presumably divided among the several jails, usually
three, in the prefecture, so there would have been about twenty-four
functionaries per jail, which is close to the figures given for jails in Fu
Prefecture in Fukien. 110
Some of these men served as regular guards for the prisoners. When
numerous prisoners guilty of serious crimes were being held, the jail
functionaries were not allowed to leave the jails. In case of emergency,
they had to secure the permission of their superiors before leaving.
According to the Ch'ing-yuan Viao-ja shih-lei, the rule was that if the
responsible clerical personnel and jail functionaries did not stand guard
at night, they were to be punished with eighty blows of the heavy rod.
The clerical personnel were, however, permitted to stand guard by turns.
If at night any men escaped or killed or injured others, the men off duty
were not to be penalized. These guards were also responsible for seeing
that the prisoners did not kill or injure themselves. If prisoners injured
themselves, those responsible were liable for eighty blows of the heavy
rod, a penalty that would be increased by two degrees (to one hundred
blows) if any of these prisoners died as a result. 111

108 SHY, hsing-fa 6.62a-b; SS 200.4994.

109 Ch'en Ch'i-ch'ing, Chia-ting ch'ih-ch'eng chih, pp. 7201 ff.; SHY, hsing-fa 6.62a-b.
110 Liang K'o-chia, San-shan chih, pp. 7737, 7738.
111 TFSL537.

375
 Law and order in Sung China
When the security of the jail seemed threatened by such problems as
large numbers of prisoners, some jails could have special units of soldiers
with noncommissioned officers acting as night guards in addition to the
regular jail functionaries. There was also a general rule that when a jail
held five or more men accused of banditry, men presumed to be violent
and dangerous, a round-the-clock guard of soldiers was to be main-
tained. 112 According to an order of the early eleventh century, if there
were in the capital numerous dangerous prisoners, the jails were to have
fifteen guards under an overseer. 113 Except for the actual jail personnel
and the dispatched guards, other people were not permitted to enter the
jail area. 114

Length of time in jail

The authorities were also concerned about the problem of prisoners
being held for extended periods because of delays in the processing of
cases. This flaw in the system is repeatedly discussed. One high official
spoke in 1007 of one-hundred-day delays being common among cases
under examination by officials in the capital. 115 Late in the century the
emperor himself complained about the delays at the capital, noting that
more than five hundred cases had accumulated that called for review but
that had not yet been decided. 116 Sometimes the delays were deliberately
created by jail functionaries, who tried to implicate large numbers of
people in hopes of getting bribes. 117
This problem was exacerbated before 1033 because all men accused of
crimes calling for registration had to have their cases memorialized for
review and, while awaiting replies, were held in jail. After an official
complaint, it was ordered that such cases be divided into serious and
minor. 118 Presumably thereafter, some cases could be disposed of with-
out such long delays.
The state also tried in other ways to underscore the official recognition
of the suffering of those held for long periods in jails and to alleviate
that suffering. In the mid-1030s, for example, jail functionaries were
forbidden to feast or entertain while there were many people held in
their jails. 119 Such prohibitions served to emphasize the concern sup-

112 SHY, hsing-fa 6.51a, 53b- 54a.

113 SHY, hsing-fa 6.53b~54a.
114 Hu T'ai-ch'u, Chou-lien hsu-lun.
115 HCP 65.16b.
116 HCP 296.9b.
117 Chang Kang, Hua-yang chi, 14.6b-7a.
118 HCP 112.14b.
119 HCP 115.15a.

376
 Jails and jailers in the Sung
posedly felt by the authorities for those incarcerated. More importantly,
the state set time limits for processing cases 120 and for forwarding
reviews,121 which were designed to limit terms in jail.
Despite these measures, however, the problem persisted. In the late
1070s, officials were disciplined for delaying in a case that had dragged
on for a year without being settled. 122 An anecdote from the middle of
the twelfth century tells of a woman held in jail for five years (during
which there had been two major empirewide amnesties) without being
released or having her case settled. 123 The "Treatise on Punishments" of
the Sung History says that the situation was particularly serious under Li-
tsung (r. 1224—64), 124 though clearly it had always been a problem.
Investigating this situation was one of the major tasks of the cir-
cuit intendants. In the mid-twelfth century they supposedly submitted
quarterly reports on the number and kinds of prisoners in the jails under
their jurisdiction. 125 They were also to prepare an annual report giving
the names and posts of those offices in the jurisdictions under their
circuit in which there had been delayed judgments or complaints of
judicial abuse. 126

Number of prisoners
We also occasionally find reports that suggest something about the
number of prisoners held. With its large population and numerous jails,
the capital was home to the largest number of prisoners - early in the
dynasty, they numbered at least in the hundreds. In the autumn of 1005,
when the emperor examined the prisoner records {ch'iu-chang) for the
capital, he found that there were over 200 people being held. 127 (These
records were presumably similar to the prisoner rolls - ch'iu-pu128 - that
the emperor examined in 1007 or to the jail statements -ju-chuang - that
all jails in the capital were ordered to provide in 1077.) 129 Later in
the dynasty this number rose dramatically. In 1084 Emperor Shen-
tsung, disturbed to learn that the jails of the Court of Judicial Review
and Kaifeng Prefecture held more than 1000 men, ordered a speedier
120 HCP 378.1a, 429.14a.
121 HCP 2i9-3a-b.
122 HCP 3o8.5a-b.
123 HNYL 2808. Amnesties freed most prisoners in jails but probably affected an even
larger number of men serving sentences of penal registration.
124 SS 200.4996, 201.5015.
125 HNYL 2475.
126 TFSL81.
127 HCP 61.nb.
128 HCP 65.16b.
129 HCP 270.3b.
377
 Law and order in Sung China
130
processing of cases, but another report seven years later again reported
that i ooo were being held in the capital itself and its surrounding dis-
tricts. 1 3 1 T h e situation in the early Southern Sung was similar. In 1135 it
was reported that the capital prefecture held a total of 1634 prisoners. 1 3 2
In the outer prefectures the number of men being held at any given
time was much lower. In 1003, after issuing an empirewide ordinary
amnesty, when the emperor personally examined the lists of men held,
he is said to have freed 4606 men (at a time when there were approxi-
mately 300 prefectures and the empire's population was probably be-
tween 35 million and 40 million). 1 3 3 Even if only a relatively small
proportion of prisoners were freed, this still means that the number held
was rather small.
W e also have more specific information on some of the prefectures.
In 1159 the judicial intendant of Fukien, inspecting Fu Prefecture,
discovered and released 147 innocent people in the jail of the police
inspector. 1 3 4 In 1135 all the officials in Hsiian Prefecture were awarded
advancements because the judicial intendant had reported that during
the previous year there had been 355 prisoners in the jail and not a
single death from illness. 1 3 5 In 1135 Ch'ii Prefecture in Liang-che Circuit
was reported to have held a total of 618 people. During the following
year Chiang-yin Prefecture in Liang-che held a total of 74, and Yang
Prefecture in Li-chou Circuit, a total of 122. 136
Of the areas about which we have information, Fu Prefecture was the
outer prefecture with the largest number of prisoners. A figure for 1137
indicates that Fu Prefecture in Fukien had an annual total of 682
prisoners. If we assume that there were three jails in Fu Prefecture, each
would have held about 230 prisoners during the year 1137. A rough
calculation suggests that there would have been less than two dozen in
each jail at any one time.
T h e districts naturally held fewer prisoners than did the larger units.
It was reported that during 1135 Wu-i District in Liang-che Circuit had
held a total of seventy-two men, and Wu-p'ing District in Fukien had
held some forty prisoners. 1 3 7 Perhaps the best we can say is that in the

130 HCP 349.6a.

131 HCP 459.1a.
132 SHY, hsing-fa 6.6 5 b-66a.
133 HCP 55.1b. The figure for the total population is only approximate. On Sung
population statistics, see Hans Bielenstein, Chinese Historical Demography, A.D. 2-ig82
(Stockholm: Bulletin of the Museum of Far Eastern Antiquities, no. 59, 1987), pp.
46-78.
134 HNYL 3035.
135 HNYL 1498.
136 SHY, hsing-fa 6.65b-66a.
137 Ibid. Tai Chien-kuo uses these figures to try to calculate approximately how many

378
 Jails and jailers in the Sung
districts under ordinary circumstances, there seem to have been very few
prisoners held at any given time, the number for a year being fewer than
one hundred. In a prefecture, during one year, the prisoners might have
numbered in the hundreds, but we must remember that the prisoners
presumably would have been divided among at least three jails, and so
the numbers per facility may have been rather like those in the single jail
in most districts.
The capital contained many more jails. It had two basically urban
districts, each with a jail. There was also the prefectural government
with three jails, and the several patroling inspectors' offices each with a
jail. When added to the jails run by the armies, the Censorate, the Court
of Judicial Review, and a number of other capital offices, these would
bring the total of capital jails close to twenty, in either Kaifeng or Lin-
an. Thus even one thousand prisoners would have averaged only about
fifty per facility (though no doubt the distribution was far from even).
This figure of one thousand prisoners is for men being held at a specific
time. Even though these are only estimates, they do seem to indicate that
jails in even the more important outer prefectures could be relatively
small; perhaps some of the jails in the capital were larger.

Supervising jails
The condition of the prisoners in the jails was also supposed to be
monitored by the local officials (or in the capital by officials from the
central judicial organs) through a regular system of frequent inspections.
In 981 the central authorities ordered district officials to inspect the jail
and its prisoners once every five days. They were to write a report on the
number of prisoners held and released and submit this to the prefecture.
They also kept a log, presumably listing the prisoners and their time of
entrance and containing jail statements giving the crime and date of
incarceration (yti chuang).138 The prefectural jails themselves were to
have a separate log that would be inspected by the chief officials every
few days. Every month a report was to be sent up to the Ministry
of Justice. 139 In 984, because "conditions are again peaceful," the
local jurisdictions were allowed to submit such jail statements every ten
days. 140 We can get some idea of the nature of this process from Li

prisoners might be expected to be in a jail at any one time. See Tai Chien-kuo,
"Sung-tai," p. 90.
138 HCP 22.i6b-i8a. For the twelfth century, see also HNYL 2950.
139 HCP 270.3b (1070).
140 WHTK 166.1444 gives the date of this order as 982. It is given as 984 in WHTK
167.1453; SS 199.4969; and STCLC 741.

379
 Law and order in Sung China

Yuan-pi's early twelfth-century book of advice to district magistrates

which suggests that "each section on local prisoners, which includes
people released under local bond, should list [its prisoners] one by one
and provide explanations. The executive secretaries are to verify all of
the above." 141 Such reports dealt with ordinary prisoners. If officials or
the wives of officials were being held, they were to be covered in separate
14-9

reports.
These inspections by resident local officials were supplemented by
visits from outside commissioners who came to examine the cases on the
local docket, seeking to eliminate delays in sentencing. After Emperor
T'ai-tsung in 985 remarked to his ministers that he was concerned about
the condition of those being held in jails, he dispatched servitors to go to
the various circuits to check on delays. 143 This is an early example of a
practice that grew over time into a regular system of periodic inspections
of prisoners and cases by the circuit intendants or their appointed
agents. A report from the closing years of the eleventh century describes
the prevailing policies for inspections by the circuit authorities:
Concerning the stipulated systems for being merciful in punishing, in accord
with the ordinance of the Yiian-feng period [1078—85], the judicial intendants'
offices during the first fortnight of the fourth and the tenth months are to inspect
[the jails] to see that the various prefectural chief officials have followed these
systems and to report [on their findings].144
The circuit intendants interviewed prisoners, issued summary deci-
sions in many cases, and in fact often freed prisoners. 145 This practice
began in the Northern Sung but appears to have fallen into disuse late in
that period, since Li Kuang (1078—1159) memorialized early in the
Southern Sung asking that it be reestablished. Perhaps, as was the case
with so many practices, it had lapsed during the troubles of the 1120s. 146
Reports indicate that it was in regular use during the later Southern
Sung. 147 (Local officials who were holding people illegally sometimes
evaded these inspections by temporarily moving such prisoners to other
quarters during the intendants' visits.) 148
The Ch\ng-yuan tyiao-fa shih-lei gives a detailed description of these
141 TITC 1.31.
142 SS 199.4992.
143 HCP 26.3a.
144 HCP 485.4b.
145 See Brian E. McKnight, The Quality of Mercy: Amnesties and Traditional Chinese Justice
(Honolulu: University of Hawaii Press, 1985), pp. 86—91.
146 See Li Kuang, Chuang-chien chi, I2.2ia-b.
147 Liu Tsai, Man-Vang wen-chi, SKCSCP ed., 23.27a; HNYL 2741.
148 HNYL 2741. This same charge is repeated early in the twelfth century; see SHY,
hsing-fa 5.46a-b.

380
 Jails and jailers in the Sung

tours. At that time, in addition to quarterly examinations, the judicial

intendants at year's end were to tour their jurisdictions, checking on
and releasing prisoners when appropriate. The code even specifies the
number and types of their retinues, and a system of graded rewards for
having saved men unjustly held. 149 The reports for these journeys even
followed a particular format that included the number of prisoners, their
crimes, and the names of the subunits where the intendant had reviewed
and reversed cases. At the end it had a section in which the intendant
could record the number of good clerks he had recommended. 150 In
theory such a system should have kept higher authorities well informed
of the current state of those in jails, but in practice, especially in the
more distant areas, such duties were not carried out carefully.151

Empty-jails system

From early in the dynasty, having empty jails was considered worthy of
praise a n d reward. 1 5 2 According to the Collected Sung Documents:

Whenever in the various prefectures the jails were empty, under the old regula-
tions, decrees of commendation were to be issued. If the jails attached to the
prefectural offices or to the Office of Public Order (Ssu-li yiian) were empty for
three days or more, in all such places meditation halls were to be established.
For the necessary daily provisions, official money was to be given out. 153

Meditation halls were Buddhist religious buildings used for meditation

and ceremonies for the dead. T h e establishment of such a n institution
honored both the locality a n d its officials. Although we d o not know the
date of these " o l d regulations," we d o know that by 982 it was already
an accepted practice for local officials to submit reports when their jails
were empty, a n d from 992 we have a document that outlines regular
procedures for reporting such empty jails, for verifying these claims by
higher officials, a n d (apparently) for automatically granting rewards
without requiring a special imperial decree. 1 5 4
This system was modified during the eleventh century a n d seems to
have reached its final form by the end of the Northern Sung. By that
time the judicial intendants were to submit monthly cumulated a n d
verified reports on the claims of empty jails sent to them by the s u b -

149 TFSL 86.

150 TFSL 87; HNYL 1664.
151 HCP 6.60a.
152 For more information on this system, see Brian E. McKnight, "A Sung Device for
Encouraging Speedy Trial," Journal of the American Oriental Society 93 (1975): 483-85.
153 SHY, hsing-fa 4.85a.
154 SHY, hsing-fa 4.85a; HCP 23.13b; WHTK 166.1445.

381
 Law and order in Sung China
ordinate prefectures and districts. These claims were investigated by the
Ministry of Justice and the Censorate in the capital. If approved, the
reports were passed on to the Institute of History for possible inclusion
in the state history. The Bureau of Academicians (Hsueh-shih yuan)
would then issue an edict commending the responsible officials, who
would also be given various rewards. 155 This system continued in opera-
tion during the Southern Sung.
Some idea of how uncommon it was to have empty jails can be found
in an essay written in 1259 by the late Southern Sung official and patriot
Wen T'ien-hsiang, to commemorate the excellent record of a prefect of
Chi Prefecture. Wen reports that according to the general memory of the
older people of Chi Prefecture, the jails had been empty on only two
occasions since the beginning of the Southern Sung in the late 1120s.
And yet, says Wen, this present prefect has had empty jails in two
separate months within only two years. "In the period of more than a
century there were only those two former occasions. Now this gentleman,
having held office for three years, has already been able to match that
record." 156
Such a system invited attempts to abuse and misuse it. Local auth-
orities are said to have moved prisoners around among a group of jails,
so that inspectors always found empty jails when they visited. Local
authorities are also accused of having hurried reviews and executions so
that their jails would be empty. 157 Such problems were merely one more
example of the endless struggle between the central authorities' seeking
to concentrate their power and the local authorities' seeking to have
more control over the affairs in their jurisdiction.

Conclusion
The Sung system of jails and jailers was the result of the interplay of
complex factors. Fiscal considerations, if not ideology, prevented the use
of imprisonment for extended periods as a major penalty. The need for
security, however, combined with the practical difficulties of investiga-
tions to spawn a system in which substantial numbers of people ended
up in jails, whether as accused or convicted criminals or as witnesses.
The need to hold and keep secure relatively large numbers of prisoners,
at a bearable cost, dictated some aspects of physical conditions of the jail

155 SHY, hsing-fa 4.85b, 86a, 87b, gob-gia. The system continued to the end of the
dynasty, despite the order of 1113 that the system be abolished, at least in the capital;
SS 21.392; and see also HNYL 2578.
156 Wen T'ien-hsiang, Wen-shan hsien-sheng ch'uan-chi, 9.292-93.
157 HCP 72.16a.

382
 Jails and jailers in the Sung

system. However, we shall miss one of the keys to understanding Sung

justice if we fail to see the impact of ideology and especially of the deep-
seated belief in the capacity of at least most people to control their
behavior and renew themselves, to cease being deviant and to become
law-abiding, productive participants in the larger society. Of course, the
practice never lived up to ideals, except perhaps temporarily under the
guidance of especially concerned and effective officials. There is also no
doubt that the low-level workers in the system, like the low-level workers
in our penal system today, being exposed daily to the realities of the jail
system and their clientele, were under pressure to become disillusioned
and callous.
The concern for security promoted the building of strong walls, sys-
tems of guards, the use of manacles and fetters, and all sorts of physical
systems of restraint and control; the concern for the lives of people and
their possible later reintegration into the outer society promoted policies
for provisioning jails, offering health care, and maintaining hygiene. The
interaction of these concerns was complicated by the fiscal problems of
paying for such care and the tendency among many jail personnel and
officials to view the prisoners as evildoers and the characters of those
men attracted to service as jail functionaries.
The bitterness of life in jail should not be minimized. No doubt it was
usually frightening, dirty, and shameful, as well as extremely danger-
ous. Death, disease, and degradation were common currency. Life in a
Sung jail was probably almost inevitably a brutalizing experience. The
Southern Sung writer Chen Te-hsiu describes it:
Sometimes the prisoners' food is insufficient, and their clothing inadequate so
that hunger and cold beset them. Sometimes the fetters are excessively heavy and
are not washed clean so that their necks become inflamed and abscessed. Some-
times their cells' roofs are leaky and go unrepaired, so that the wind and rain
penetrate them. Sometimes the beds in the jails are crowded together so that in
no time they suffer from lice. Sometimes the privies are inside the cells and have
no covers, so that the stench is dreadful. When they fall ill because the prisoners
sometimes do not receive early medical treatment, they die. Sometimes even
though they are guilty of minor crimes, the prisoners are housed together with
men guilty of capital crimes. These kinds of problems are simply innumerable.158
However, even if this is a reasonably accurate description, when com-
pared with most other systems of imprisonment in traditional (and many
modern) societies, the Sung system and the Chinese premodern systems
generally had relatively good records. There is a tendency to belittle the
impact of state rules in traditional China, to assume that many of them

158 Chen Te-hsiu, Hsi-shan cheng-hsiin, p. 7.

383
 Law and order in Sung China
were ignored if inconvenient for administrators. Yet when we read con-
temporary records by outside observers regarding the punctiliousness of
Chinese bureaucrats, even in periods thought of as having relatively
weak central control, the striking thing is the degree to which the rules
were in fact followed carefully. I think that we must assume that the
rules set down by the central authorities set a top standard for the
treatment of prisoners and that in practice the jail policies fell short of
the ideal but did not ignore it entirely.
The key to this was the very widely shared conviction among the
members of the Sung elite who participated in administration that pris-
oners were people, most of whom could remake themselves, people who
could be salvaged to begin again in some fashion as integral members of
society. In his essay on prison life in eighteenth-century China, Derk
Bodde was struck by the unflagging faith of the prisoner Fang Pao in the
basic goodness and worth of humankind. We should not be surprised at
this attitude, which is closely interwoven with the fundamental tenets of
the worldview of most members of the elite in late imperial China. Even
in revolutionary China, this basic conviction underlies the avowed goal
of punishment. In a culture in which threatening behavior is seen as
willful and the elite sees only one correct path, the resulting action by
the state authorities may be brutal, but it may also include earnest
attempts to woo back the deviant through a process of education. As the
warden of a contemporary prison told reporter Robert Elegant, "We
take great pains to educate our prisoners - not only so they'll learn their
errors, but also so they'll learn trades. Above all we treat them as human
beings." And as for hard cases, "it may take a long time, years even, but
almost everyone finally reforms." 159 The reality of Chinese jails has often
been far more brutal than suggested by this self-serving assessment, but
at their best this is what Chinese jails might be.

159 Robert Elegant, "Why the Chinese Stop Crime Better Than We Do," Parade, Sunday,
October 30, 1988, pp. 6—7, quoting Han Shizhang, vice-warden of the Beijing
Municipal Prison.

384
 12
Penal registration

Introduction
The Sung government, like governments of all dynasties during the
Chinese empire, frequently used deprivation of liberty plus forced labor
as a penalty. The Sung punishment of forced labor resembled in many
ways the forced labor the systems used before (and after) the Sung, but
during the Sung these inherited practices were modified to create a dis-
tinctive pattern. During the Sung most laboring convicts were registered
in special units of the army stationed alongside regular army units. Some
of these regular units might be actual fighting troops, like the imperial
armies (chin-chun). Many others were, like the convict units, actually
labor battalions, the so-called provincial armies (hsiang-chun). Men serv-
ing in the nonconvict labor units of the provincial armies received
distinctly better treatment than did the men in the convict units. The
convicts were encouraged to behave well by the knowledge that they
would have opportunities to move up into these better-treated units after
completing their term of convict labor or after one of the frequent Sung
acts of amnesty. They were encouraged by the constant presence of
examples of how their lot might be improved and yet were kept under
military discipline and control by the presence of other army men, who
always outnumbered those in the convict units.

Liberty penalties in early China

The earliest Chinese sources describing punishments emphasize corporal
punishments, making the body suffer; sources from later periods mention
penalties based on the restrictions on liberty. The "Punishments of Lu"
section of the Book of Historical Documents, a late Chou document that
purports to be from the early Chou, speaks not only of the five corporal
385
 Law and order in Sung China

punishments but also of the Five Exiles.1 Exile is profoundly different

from these corporal punishments. It does not disfigure, nor does it
overtly distinguish the sufferer as an outcast from normal society. It does
mean that the sufferer is not free to return home or to travel. Exile is not
immediate or violent, but it endures. It stretches out the time of the
suffering. To the convict every new day is a new demonstration of his
powerlessness. The exile is humiliated in his relationship to the power of
the ruler but is not necessarily shamed before ordinary people. Because
he has not been disfigured, reintegration into the ranks of those fully free
can be a matter of moments.
Long before the "Punishments of Lu" was written, slaves and pris-
oners were used as laborers; indeed, forced labor was as old as the
Shang. 2 But labor as a punishment for having committed crimes, as
opposed to having been captured in warfare, does not seem to have been
widely used at the time the first sections of the Book of Historical Documents
were written. Perhaps there was some use of criminals as laborers, but
exile - the penalty most closely linked to penal servitude among the Five
Great Inflictions mentioned in early Chou sources - may in practice
have been a penalty for the elite that did not entail labor.
There does not seem to be any reliable evidence in the early materials
that the men who were mutilated as punishment were used in large
numbers as laborers, but by the late Eastern Chou, rulers were already
exploiting the labor power of convicts. This was a momentous trans-
formation. Why should the states of that time have shifted from pri-
marily relying on mutilating punishments to primarily relying on penal
servitude? In explaining the change from the spectacular (and corporal)
punishments of early modern Europe to the systems of punishments now
in use, Michel Foucault speaks of a change from punishment as a symbol
of the sovereign's vengeance to punishment as an act of social defense by
the dominant class. 3 But this can hardly explain the comparable Chinese
phenomenon. The very earliest Chinese sources to comment at length on
punishments stress the merciful penchant of the ruler, his hesitation to
inflict punishments, his search for the circumstance that mitigates, the
condition that explains. Indeed, the early tradition blames the common
use of the mutilating penalties on a foreign tribe, a group beyond
the bounds of civilized behavior. Furthermore, the transformation of
penalties - their metamorphosis into forms of state-supervised labor - is

1 For the dating of the "Punishments of Lu," see H. G. Creel, The Origins of Statecraft in
China.
2 See D. N. Keightley, "Public Work in Ancient China: A Study of Forced Labor in the
Shang" (Ph.D. diss., Columbia University, 1969).
3 Michel Foucault, Surveiller et punir: Naissance de la prison (Paris: Gallimard, 1975).

386
 Penal registration

coincidental with the growth of sovereign power, with the emergence -

both in theory and in practice - of autocratic rule.
In Ch'in, the state which has stood for two millennia as the symbol,
the quintessence, of despotic institutions, the use of penal labor as the
primary penalty appears before the unification of China in 221 B.C. The
very names of the types of penalties are suggestive. Those sentenced
to the heaviest penalty were called "wall builders" if men or "grain
pounders" if women. The next most serious offenders were employed as
gatherers of "firewood for the spirits" (if men) or as sifters of "white
rice" (if women). Less serious yet was condemnation to service as a bond
servant. 4
Why this change? The answer lies in both the aim of the punishment
and the organizational technology of the time. The aim of the penalties
was to prevent as far as possible further violations, to give criminals an
opportunity and incentives to reform, to deter others by example, and to
do all of these at a minimal cost to the state.
Mutilating penalties might deter and might be relatively inexpensive,
and some would certainly make further violations more difficult for the
culprit, but they would hardly be conducive to reformation. Reform
through labor not only provided an opportunity for repentance; it also,
by promising the convict eventual redemption and reintegration into the
general community, offered incentives for good behavior. The penalties
of earlier times, with their emphasis on physical (and usually highly
visible) mutilation, had among their effects the permanent exclusion of
the criminal from ordinary life. To have one's feet or nose cut off was
deeply shameful as well as disabling. Even a person who wanted to
return to normal life could not do so. In the Sun Tzu book we are told
that the ruler of Yen made a cruel mistake in his attack on Ch'i, because
he cut the noses off his prisoners. Knowing this, "the men of Gh'i were
enraged and conducted a desperate defence." 5 In the biography of Sun-
tzu in the Historical Records, Ssu-ma Ch'ien wrote that when the ruler
of Ch'i wanted to make the great strategist Sun Pin (who had been
mutilated as punishment) the commander of his armies, Sun Pin refused.
It would not be fitting, he said, for a man so mutilated to hold such a
position. Mutilating punishments were designed to shame the crim-
inal and to outcast him from society.6 Mutilating punishments had
another purpose in a society that was loosely spread across a largely

4 A. F. P. Hulsewe, Remnants of Ch'in Law (Leiden: Brill, 1985), p. 14.

5 Samuel B. Griffith, Sun Tzu: The Art of War (Oxford: Oxford University Press, 1971), p.
75-
6 For the story of Sun Pin, see Ssu-ma Ch'ien, Shih-chi (Beijing: Chung-hua shu-chu,
I959)> 65.2163.
387
 Law and order in Sung China
empty landscape: They were permanent and relatively inexpensive. The
technology - fiscal, physical, and political - for punishing large numbers
of men through long-term deprivation of freedom did not exist. One way
to establish permanent control over a man was to deprive him of the
means of free action, physically by amputating his feet or socially by
tattooing him or cutting off his nose.
The increasingly widespread use of labor penalties coincides with the
growth of the state's ability to exploit large numbers of laborers for
extended periods of time. In the Ch'in these labor penalties were accom-
panied by mutilating penalties as supplementary punishments, but
thereafter mutilating penalties become less common. The wider use of
labor penalties also coincides with a new military technology that made
large quantities of labor a valuable asset, and a political ethos that saw
men as useful tools. How stupid to blunt a useful tool deliberately. How
much more sensible to make use of that tool. Reform through labor, to
use the jargon of our own age, was cost effective. The exact form of the
penalty was determined in part by the influence of past examples, by
some aspects of technology (particularly weapons, transportation, and
communications technologies, which affected the cost and the security
level of the penalty), by the labor needs of the authorities, by the costs
associated with the use of the labor involved, and by the need to
maintain security.
Technologies and economies determined; ideology explained. The
earlier condition of society and economy may have encouraged the use of
mutilating penalties, but their use always ran counter to some of the
deep-seated axioms of political understanding. If the ruler was indeed
the parent of the people and was responsible to Heaven for their welfare,
how could he bear to mutilate his children? If criminal behavior resulted
because these children had not been properly raised and cared for, then
was it not an unfortunate reflection on the ruler if he used such dis-
figurements, which would remain permanent reminders of his lack of
success? Furthermore, one of the sovereign's attributes was his capacity
to forgive, to cleanse the outcast of his stigma, and to take him back into
the social fold. But if the outcast were permanently disfigured, how could
he be so reintegrated? The answer was simple. The outcast could be
temporarily set apart, temporarily disfigured. But the possibility would
remain for his reincorporation into normal society. The typical criminal
— young, male, and having a short time horizon so that he tended to
discount heavily both future rewards and future punishments - could
be exploited through a system that controlled his behavior without
permanently damaging his economic worth.
With these reforms the rulers could at the same time gain a large and
388
 Penal registration
useful labor force, using policies that temporarily set these laborers apart
from normal people and yet reduced their temptation to resist discipline
by holding out the possibility, indeed the probability, of a normal future.
Thus, in the Ch'in (and in some late dynasties), all convicts had their
hair and beards shaved off. Given traditional Chinese beliefs about the
body, these were mutilations, but not permanent ones. Convicts wore
distinctive clothing and might be fettered, both temporary and quickly
reversible disfigurements.
The Ch'in seems to have been a transitional period between a greater
reliance on mutilations as penalties and a system in which the labor
value of convicts dictated the wisdom of keeping them whole. In addition
to their distinctive clothing and shaved heads, Ch'in convicts might be
tattooed, have their noses cut off, or have one or both feet amputated. 7
Under the early Han, some mutilations were still practiced, but the
Emperor Wen-ti (r. 179-157) transformed the penalty system by replac-
ing the cutting off of the nose and feet with beatings and the wearing of
foot irons. Castration continued to be used as a rare punishment under
the Han but seems not to have been thought of as belonging in the
same class as the other mutilations. 8 Knowing the usefulness of convict
labor, the authorities avoided punishments that might reduce its value.
Being masterful politicians they naturally attributed the reforms to the
benevolence of the emperor.
Loss of liberty continued to be the mainstay of traditional Chinese
penal systems after the Han; indeed, the system that was to be enshrined
in the influential T'ang codes evolved from these Han practices. The
Han system was described by A. F. P. Hulsewe:

Hard labor in various forms was the principal punishment during the Han
dynasty.. . . During the Han dynasty there existed the punishment of hard labor
in the following series of gravity: ch'eng-tan (if the ancient explanations are
correct: "building walls or fortifications and standing guard from early dawn")
with in addition: having the head shaved, wearing an iron collar, undergoing the
bastinado, and wearing legirons; the duration being five years, the bastinado
varying between 500 and 300 strokes (as from 167 B.C.) or 300 and 200 strokes
(as from 144 B.C.).... The iron collar and the shaven head were the main
characteristics of this punishment.. . . Below the group of shaved and collared
ch'eng-tan there is a group of hard labor punishments . . . involving shaving off the
beard.. . . The ch'eng-tan served for four years. The three-year punishment was
called kuei-hsin, "(collecting) firewood for (the sacrifices to) the spirits.". . . The
two-year punishment was called ssu-k'ou, robber-guard. The one-year punish-
ment was called . . ."penal labor in garrisoning." The general term for convict,

7 Hulsewe, Remnants of Ch'in Law, p. 15.

8 Hulsewe, Remnants of Han Law, pp. 127-28.
389
 Law and order in Sung China

specifically hard-labour convict, is Vu. [Note: This term was later used to mean
penal servitude.] 9

Penalties of forced labor continued to be common under the states that

ruled in China after the collapse of the Han empire. As under the Han,
these penalties were differentiated in part according to the length of the
labor demanded by the state and also might call for shaving the beard,
wearing fetters, and receiving a preliminary beating. 10 The T'ang
dynasty (618-907) inherited such a system from the Sui (581-618) and
passed it on in outline to the Five Dynasties.
One characteristic of some of these emerging systems was the division
of levels of penalties according to the distance to which the convict was
sent. Reflecting on this trait, Sung commentators assume incorrectly that
the use of different distances originated because men found it a greater
hardship to be farther away from their home locations. In fact, the use
of distances was connected with the traditional Chinese practice of
vendetta. As the bloody and honor-bound society of early Chou times
evolved into the later imperial order, the tradition of the vendetta was
rationalized and codified, so that by Han times the duty to avenge
wrongs done to one's relatives or sworn master reached to only a certain
distance. If the enemy lived beyond that distance, the aggrieved party
was not honor bound to seek revenge. Thus, in setting distances, the
codifiers were seeking to reduce the frequency with which private ven-
geance would be sought. Men of the T'ang and Sung maintained the old
form for new reasons, without understanding why it had originally been
created.

Penal registration in the Sung

The Sung was simply following common practice when it extensively
used penal labor and restrictions on liberty. The dynasty retained in
name the T'ang punishments of penal servitude and exile, but it used in
practice a complicated system of penal registration that could grade the
punishment in a variety of ways, by assigning types or lengths of labor,
providing differing amounts of support to convicts, adding supple-
mentary punishments, and registering convicts in different kinds of
places. The penalty used in the Sung - which was said to have been
developed under the Later Chin during the T'ien-fu reign period (936-
944) - called for penal registration in the armies. An official who traced

9 Ibid., pp. 128-30.

10 For example, consider the punishments under the code of the Liang dynasty (502—56);
see Etienne Balazs, Le Traite juridique du "Souei-ckou" (Leiden: Brill, 1954), p. 37.

390
 Penal registration

the history of the penalty indicated that the Sung added a beating to this
inherited penalty. 11
This penalty, which usually included labor, was in fact a carefully
graduated punishment combining features of penal servitude, the old
penalties of exile, and a beating. The beating served as both a deter-
rent and an act of vengeance. The labor was important to the state's
economy. The exile removed the convict from his old, possibly criminal,
social companions; it took him from a setting in which he might be the
agent of or the victim of vengeance and, by removing him from his
home, penalized him emotionally. In the words of an early thirteenth-
century official, "In our state we know that men do not want to be
separated from their homes. Therefore we have the penalties of the
various forms of penal registration." 12 There is some indication that men
so punished, like convicts in earlier times, may also have suffered the
temporary disfigurement of having their hair cut off. Because the men
subjected to this penalty were grouped together in special provincial
army units, which were physically located near other army units, the
convicts were under the more or less watchful eye of the military.
Perhaps more importantly, because better-paid service in the armies was
one of the possibilities held out to those who behaved well, this constant
exposure to the better conditions of others fitted well with the probable
general character of the convicts. The short time horizon of the typical
criminal could benefit from the daily exposure to the benefits of good
behavior.
The Southern Sung official Hsiieh Chi-hsiian compared this penalty
with the T'ang penalty of exile with added labor:
The Former Kings established the laws on penal servitude and exile in order to
control evil and to avoid executing people. They entered convicts' names on the
military registers as a means whereby to control the violent nature of the
criminals and restricted their manner of living in order to make their hearts
obedient. These penalties serve as warnings to people to draw them to good
behavior and keep them from evil. . . . The Emperor Shih-tsung of the Later
Chou (r. 954-60) gathered together the unruly and desperate men of the empire
to form a brigade, controlling them through the use of military law. This was an
effective measure. . . . Military law restrains their violent natures. Army labor
makes their hearts obedient. If you follow this pattern, then you will move them
toward good behavior and guide their merit, and they will not again cause
disorders among the people.1^

11 WHTK 168.1461.
12 WHTK 168.1462.
13 Hsiieh Chi-hsiian, Lang-yu chi, 16.23a-24a.

391
 Law and order in Sung China

This penalty was used for many sorts of crimes, including those whose
nominal punishment was penal servitude, exile, lesser penalties such as
beating made heavier by special imperial order, and (quite commonly)
nominal death penalties in which the convicted person was spared by
special imperial order. Most Sung officials discussing this penalty
thought of it as especially connected with the crime of banditry in which
the circumstances seemed to justify reducing the penalty from death. 14
In the sources now available to us, banditry — that is, organized robbery
with violence — is by far the most frequently mentioned crime, but
registration in the armies could also be used for many other sorts of
crimes, such as recidivism by thieves, arson, some homicides, rape, and
witchcraft.15
If punishments were to achieve some of their goals, especially deter-
rence and retribution, they had to be carefully adjusted to fit the offense.
Punishments were messages to potential wrongdoers, to the people
wronged and those around them, to the spirits of the dead, and, indeed,
to the complex interacting natural world within which human actions
occurred. Improper messages could cause all sorts of difficulties, so the
necessity of adjusting punishments to fit the crime was a constant theme
in traditional Chinese thinking about penalties.
The penalty of penal registration was well suited to this purpose.
The heaviness of penal registration could be adjusted to fit the circum-
stances of the crime, varying from a form of (nominally) lifetime exile at
extremely unpleasant places where the convicts, after being beaten on
the back and tattooed on the face, were permanently assigned to do
heavy labor, to what appear to be forms of house arrest at varying
locations away from the convict's home. The heaviness of the penalty
could also be adjusted by changing the location of registration, the type
of labor demanded and its length, the physical facilities and support
given the convicts, the strictness of supervisory control, and the use of
ancillary penalties. The determination of many of these was primarily a
function of the attempt to make the penalty fit the crime, but they were
also affected by the state's labor needs and desire to maintain security.

Treatment and classification of those penally registered

The physical treatment of the convicts was used to distinguish groups.
Different amounts of food and clothing might be provided, with the
better-treated groups receiving dependent allowances. In this case those
14 For some of the rules governing this system, as preserved in the Ch'ing-yiian Viao-fa
shih-lei, see especially p. 502.
15 See the materials in SHY, hsing-fa 4, passim.

392
 Penal registration

who were being punished more severely might actually receive better
treatment than those supposedly receiving lighter sentences. A decree in
the second month of 1009 indicates that before that time those convicts
who had been tattooed had received regular remuneration in cloth and
grain. But those without tattoos had been getting only a prisoner's daily
rations, and so their dependents had been suffering grievously. The
emperor ordered that all men in the army's convict labor units be
remunerated equally. 16 However, this kind of disparity continued in
some cases. Those people who were sentenced to the Sung equivalent of
house arrest never seem to have received government support payments.
The Sung government set rules for dividing men liable for penal
registration into categories that were based on the seriousness of their
crimes. From at least the early 1070s, there were four such grades. 17
Rules from the Yiian-feng period (1078-85) also divided convicts into
four grades, in descending order of seriousness, from those who were
"not to be transferred" (to more desirable locations, presumably after
amnesties had been issued that would benefit other convicts), to those
who were "not to be freed," to those who "might be transferred," and,
least serious of all, those who "might be freed."
A few decades later, during the Cheng-ho period (1111 —17) four
different categories were used: those with "serious circumstances,"
"relatively serious circumstances," "relatively light circumstances," and
"light circumstances." As described by an official in the late twelfth
century, according to this schema a man who was categorized as some-
one with "serious circumstances" was covered by the rule on "not
being allowed to be transferred or released." Men classed as "relatively
serious" were to be tattooed only on the temple (rather than the face)
and placed in the category of those "to be penally registered for ten
years." The treatment of those classified under "relatively light circum-
stances" was governed by the regulation (ko) for those spared tattooing
on the face and allowed to transfer their place of registration or to return
to their homes. Those classified as having "light circumstances" fell
under the rules on "resident labor" (chii-i). They had separately fixed
terms of labor and could be released after completing this term. 18
In the 1180s an official indicated that there was a schema using
fourteen categories. He suggested a somewhat simpler version, based in
principle on a suggestion by the great eleventh-century statesman Chang
Fang-p'ing. According to his scheme (which docs not appear to have
been adopted) there would have been eleven categories.
16 HCP 73.6a.
17 HCP 247.23a.
18 WHTK 168.1461.
393
 Law and order in Sung China
1. Those never to be freed to return.
2. Those who must labor for their whole lives beyond the sea. (This probably
referred originally to registration on Sramana Island off the north coast of
Shantung, but in the Southern Sung, when Shantung was not under Sung
control, it presumably referred to lifetime registration in Kuang-nan, on
Hainan Island, or the prefectures on the coast opposite to Hainan.)
3. Those who must labor for eight years in Kuang-nan's distant evil places.
4. Those who must labor for seven years at 3000 or 2500 /f.d (From the
criminal's home, or in the case of a soldier, from his permanent base camp.)
5. Those who must labor for six years at 2000 or 1500 li.d
6. Those who must labor for five years at 1500 li.d
7. Those who must labor for four years, who have been specifically ordered to
be registered in neighboring prefectures.
8. Those who must labor for three years in their own prefecture's base citadel
commands.
9. Those who must labor for two years and are not tattooed on the face.
10. Those who labor for one year who are not tattooed on the face.
11. Those who must labor at an appropriate place and who, even if they
repeatedly meet with amnesties, are not to be freed.19
It is not clear whether any version of this scheme was adopted at this
time, but it can serve as an example of the kinds of principles used in
differentiating the degrees of the penal registration system. 20

Role of the army

In its Sung incarnation the penalty of penal registration was in most
cases being enrolled on the registers of certain units of the provincial
armies (hsiang-ping), being transferred a fixed distance from the place of
conviction, being forced to provide labor for a fixed term, having to live
in guarded barracks, and possibly being tattooed. 21
The Sung army was composed of four types of soldiers - the imperial
armies (chin-chun), the provincial armies (hsiang-chun), the foreign auxil-
iaries (fan-chun), and the militia (min-chun). The smallest major unit into
which these soldiers were divided was the command. The command
(chih-hui) was the building block of the Sung armies, and the unit most
important for our purposes. According to Wang Tseng-yii, the com-
mands of the provincial armies were supposed to average five hundred
men. 22
19 WHTK 168.1461.
20 SS 201.5020.
21 On the provincial armies of the Northern Sung, see Chiba Hiroshi, "Sodai no sho-gun"
(The provincial armies of the Sung dynasty), Toho daigaku tanki daigakubu 2 (1982-83):
1-24.
22 See Wang Tseng-yii, Sung-ch'ao ping-shih ch'u-t'an (An initial investigation of the Sung
military system) (Beijing: Chung-hua shu-chu, 1983), p. 30.

394
 Penal registration

Most penal registration sentences called for registration in either the

prison citadel (yu-ch'eng) commands or in the base citadel (pen-ch'eng)
commands, a term that seems to refer to all provincial army commands
other than the prison citadel commands. 23 The rule preserved in the
Ch'ing-yuan t'iao-fa shih-lei explains: "Whenever it says 'register' but the
facial tattoo does not specify the name of the army of registration,
the men should be registered in a prison citadel command. When it
says base citadel, this refers to the various army encampments at which
various sorts of men are stationed." 24 As the official Sun Shih said in
1015, "The Court considers registration in base citadel commands (pen-
ch'eng) to be the light penalty, and registration in the prison citadel
commands (yu-ch'eng) the heavy penalty." 25
These provincial armies, however, were not simply composed of men
being punished for criminal offenses. Men were also inducted into the
provincial armies in hard times. These armies thus served as a combina-
tion of welfare institution and a labor pool. On rare occasions the
provincial troops might bear arms, but they were primarily laborers. In
the words of the thirteenth-century gazetteer of Shao-hsing Prefecture,
"Provincial soldiers provide labor and that's all." 26
The state thus might act as an employer of last resort during difficult
times, providing some sort of minimal support in return for largely un-
skilled labor. Such service may also have provided a haven for foreigners
without Chinese ties or sources of support. 27 There were a bewildering
variety of provincial army units. The Sung History says that there were
some 223 different unit titles. Their members worked in monopoly sales
systems, transportation, the postal service, roadwork, bridge building,
horse herding, levee work, dike guarding, and a host of other tasks. 28
From the late Northern Sung such men were assigned in substantial
numbers to tasks associated with agriculture, and in the Southern Sung
many were assigned directly to military agricultural colonies. 29 It
appears that many men who had initially entered the provincial armies
as convicts remained voluntarily after their sentences were completed,
because the armies provided them with a livelihood.
23 HCP 256.2b. Men serving in the Chung-ch'ing commands were also doing penal labor.
Some were tattooed and some were not (HCP 73.6a). For a good brief description of
this system, see Saeki Tomi, "Sodai ni okeru ro-jo gun ru oite" (The establishment and
transition of the ex-convict corps in Sung China), in Kinugawa Tsuyoshi, Yin Tzu-chien
Hakase so-shi kenkegd rouso, pp. 267-92.
24 TFSL 520.
25 HCP 85.9a.
26 Shih Su, Chia-t'ai k'uai-chi chih, p. 6214.
27 See SS 189.4641.
28 SS 189.4640.
29 Chiba, "Sodai no sho-gun," p. 19.
395
 Law and order in Sung China

It is impossible to estimate the percentage of the men in these prov-

incial army units who entered initially as convicts. T h e prison citadel
commands themselves were in most instances (though not always) com-
posed wholly of convicts, but they represented only a small fraction of
the total number of provincial army troops in a given jurisdiction. T h e
men in other units seem to have been of mixed origins, though most of
them may have begun as men serving in penal registration. Like most of
the other units included in these provincial armies, the convict units
were forced to provide labor. T h e men serving in them lived in their own
camps, though these might be adjacent to other provincial army camps. 3 0
Although the provincial army units were not ordinarily used in
combat, men from them might be transferred to combat units, or men
guilty of crimes that ordinarily would call for registration in the prov-
incial armies might be ordered to be inducted directly into the imperial
armies, if they were young, strong, and talented. A typical order of this
sort, dated 1034, specified:
Men from the areas in Huai-nan that have suffered from natural disasters, who
have turned to banditry but have not yet killed anyone, if they turn themselves in
and confess within two months, will be forgiven their crimes. Those who are
young and able-bodied should be tattooed and registered in the base citadel
[commands of the provincial armies, which were being used here as a form of
welfare, as the men had already been forgiven their crimes]. If they have military
talents, send them to the capital to be registered in the imperial armies/ 1
This policy was especially favored in times of military crisis. An approved
memorial of 1040, at the beginning of the wars against the Hsi-hsia state,
asked that able-bodied bandits from a number of circuits be formed into
commands of the regular army, given military training, and, when there
were disorders, sent forward to the front ranks, where they could "use
their lives to gain merit." 3 2
T h e Sung gazetteers indicate that the men penally registered in prison
citadel commands formed only a fraction of the total of provincial troops
in any given location. Table 12.1 indicates the relative proportions of
different sorts of troops.
Convicts were greatly outnumbered by other sorts of troops, including
regular imperial army and other troops. 3 3 Hui Prefecture was probably
30 Shih Neng-chih, Hsien-ch'un p'i-ling chih, p. 3533. See also Gh'en Gh'i-ch'ing, Chia-ting
ch'ih-ch'eng chih, p. 7205; Ch'en Kung-liang, Yen-chou t'u-ching, 1.23.
31 HCP 115.1a. See also another example from 1040, in which the order specified that
bandits registered in the prison citadels or other army units, if under forty years of age
and strong, be put in the imperial armies (HCP 127.5b, 15b).
32 SHY, hsing-fa 4.19b-20a.
33 Chiba feels that the number of provincial troops was greatly cut during the reign of
Hui-tsung, with hired laborers being used as substitutes, but the evidence presented is

396
 Penal registration

Table 12.1. Sample figures for prison citadel troops and other military units
(number and percentage)

Imperial and
Date Prefecture Prison citadels Other provincial local armies

1139-40° Yen 49 11% 871 89% Unknown

1170* Lin-an 824 9% 8793 91% Unknown
1175C Hui 150 7% 450 22% 977 and 500
47% and 24%
1223*" T'ai 218 9% 2200 91% Unknown
1259" Ming 52 4% 1200 96% Unknown
1265-74^ Lin-an 824 10% 7000 90% Unknown
120H Hu 194 16% 977 84%

"Ch'en Kungliang, Yen-chou Vu-ching, p. 6932. This prefecture had an original

quota of two hundred prison citadel soldiers. The information in this table comes
from gazetteers describing conditions in Chekiang. It is possible, though I think
it unlikely, that conditions elsewhere differed significantly.
*Chou Ts'ung, Ch'ien-tao Lin-an chih, SYTFCTS ed., 2.34. There was one prison
citadel command out of a total of twenty-six provincial army commands in the
city.
c
Lo Wiian, Ch'un-hsi hsin-an chih 1.22a. It is possible that the proportion was
larger in Chiang-nan Prefecture, where two of the five provincial army units were
convict units, but we have no figures for the number of men involved (Chou
Ying-ho, Ching-fing Chien-k'ang chih, p. 1057).
""Cheng Yao, Ching-ting Yen-chou hsii-chih, p. 7025.
T o Chiin, Pao-ch'ing Ssu-ming chih, p. 5154. These figures for Ming Prefecture
overstate the number of convicts in this area. Many of the men in the "convict"
units were in fact army retirees. This may have been a common practice
elsewhere, but we have no evidence.
^Ch'ien YCieh-yu, Hsien-ch'un Lin-an chih, p. 4403. Since the figure for the number
of men in prison citadel commands is identical with the figure for 1170, it is
possible that the compilers simply copied the figure. In 1265-74 there was still
only one prison citadel command out of a total of eighteen provincial army
commands in the prefecture.
^T'an Wiieh, Chia-t'ai wu-hsing chih, p. 6776.

typical. There the 900 provincial troops (of which only 150 were in the
prison citadel command), were outnumbered by the 977 imperial army
soldiers and the 500 fort soldiers. This situation reflected government
policy: The regular army troops might serve as role models for convicts;
they certainly served as their guards to prevent further abuses. 34

sufficient only to show that there were such reductions in certain limited areas (see
Chiba, "Sodai no sho-gun," pp. 21-22).
34 SHY, hsing-fa 4.380-393.

397
 Law and order in Sung China
Prison citadel commands were structured like other provincial army
units. For example, when the No. 13 Prison Citadel Command had 218
men, it also had 10 commanders (chiang), 38 adjutants (chieh-chi), 10
inspectors general (chiangyu-hou), 10 controllers (ch'eng-chu), and 10 dis-
cipline officers (ya-kuan).35 The high ratio of officers to men no doubt
contributed to the unit's ability to exercise internal discipline.
The men serving in this unit received 1 bushel (shihb), 5 pecks (toub) of
unpolished rice per month (approximately 80 U.S. dry pints). In addi-
tion, every spring and winter each man was given 1 length (p'i) of cloth
(about 10 feet) and 1150 coins for "cloth-cutting money." To understand
the value of these payments, note that half a pint per day or 15 pints of
grain per month were considered subsistence rations. During the Ch'ing
dynasty (1644-1912), 1 pint per day was the ration of grain given to
each prisoner. Thus a grain income of about 2§ pints of grain per day
would have been more than adequate to support a single man. It is not
clear whether a prisoner who was accompanied by his family received
extra rations for them, but this seems probable.

Registered control and detained control

Most of those sentenced to penal registration in the Sung were sent to
provincial army units. There were, however, two other sorts of penal
registration, "registered control" (pien-kuan) and "detained control" (chi-
kuan), used mainly to punish officials, which seem not to have included
registration on army rolls. The Yuan dynasty guidebook for clerical
matters, Li-hsiieh chih-nan, defines arranged control as "to be registered in
exile without being tattooed," but this is too general. 36 Clearly, these
forms of punishment were less serious than registration in the criminal
units of the provincial armies. The heads of those so sentenced were not
shaved, nor did these persons have to wear fetters. On the other hand,
they also were not given rations by the state. The penalties of registered
control and detained control appear to have been forms of house arrest
in exile; that is, the people so punished ordinarily did not have to live in
army-style barracks, but on the other hand, they did not receive support
from the state. They were supposed to report monthly to the local auth-
orities, rather like people on probation in our legal system. Occasionally
other functionaries were supposed to oversee them. 37 The prefectures
were to keep annual registers of those held, and the local authorities

35 Ch'en Ch'i-ch'ing, Chia-ting Ch'ih-ch'eng chih, p. 7205.

36 See Hsu Yiian-jui. Li-hsiieh chih-nan, Chu Sung pi-yung shih-lei ch'iian-chi ed., chap. 15,
P- 325.
37 HCP 289.16b.
398
 Penal registration

submitted monthly reports on such convicts to the Department of State

Affairs. 38 Like those under some forms of probation in our own times,
the Sung convicts serving registered or detained control were not free to
travel. T h e ChVing-yuan t'iao-fa shih-lei gives an example of a report form:

"Registered Control" - so many persons, name; number of years [from original

registration to the present]; jurisdiction of origin; category of person; current
crime; crimes in which the person was an accessory. For such and such a crime,
on X day, Y month, Z year, from jurisdiction A, this person has been sent for
registered control; the convict has or does not have the notation "perpetually
never to be transferred or freed" on his record or "the command may assess and
transfer this person," in which case it is permitted to indicate the jurisdiction of
transfer.39

In 1053 an official asked that no more than ten such men be sent to
any one prefecture, for fear of collusion, nor should they be sent to the
borders. T h e sources do not indicate whether the suggestion regarding
numbers was enacted, but the Court did forbid sending these men to the
borders. 4 0 These convicts might be asked to perform services, though not
apparently physical labor. From early in the dynasty, demoted officials
might have to serve as office managers (ya-ch'ien).41 Although this text
does not say so explicitly, they almost certainly were being held under
"registered control." 4 2 T h e Ch'ing-yuan Viao-fa shih-lei describes them as
having to provide "services." 4 3 Using officials to do official work while
being punished was a natural response of the state authorities to the
need for skilled workers and continued to be used during later dynasties.
Such convicts were only lightly restrained. Naturally, as with other
forms of penal registration, some convicts escaped from registered control
and detained control, though the relatively lenient conditions of their
punishment probably encouraged good behavior. 4 4 Such people were
covered by the rules on amnesties, though we do have complaints that
local authorities sometimes did not release them. 4 5 As with those held
under other forms of registration, the sentencing documents could specify
that the person sentenced was not to benefit from amnesties, though
amnesties frequently explicitly overruled such provisions and included

38 Lu Tsu-ch'ien, Huang-ch'ao wen-chien, SPTK ed., 7i.i3a-b; SS 31.577; HNYL 2683.

39 TFSL 525.
40 SHY, hsing-fa 4.23b.
41 SHY, hsing-fa 4.4b. See also HCP 110.4a, which says that when in 1031, a group of
officials was punished, several of them were disenrolled from the civil service registers
and "registered in Kuang Prefecture as office managers."
42 See also an officer so registered as a supernumerary official; HCP 126.4a.
43 TFSL 521-22.
44 SHY, hsing-fa 4.52a; quotation from SHY, hsing-fa
45 HCP 380.17a.
399
 Law and order in Sung China

those previously ordered excluded. 46 It is clear, however, that this pro-

vision about not giving amnesty was sometimes enforced. Early in 1018
an amnesty document spoke specifically about releasing people who had
been held in detained control for more than five years, 47 people who had
apparently not been affected by the three Great Acts of Grace or the
numerous other amnesties that had been issued during the preceding five
years.
There were also complaints that some men were being improperly
treated under this penalty. 48 One observer in the mid-twelfth century
even reported that in some places such men were being detained in
fetters.49 Although some were given rations, apparently the general rule
was that those subject to registered and detained control did not receive
government support. From the mid-twelfth century there was an order
that such men would be allowed to volunteer for service in the provincial
army. This ruling was issued after the wife of one such man submitted
an appeal saying that her family needed the income from such tattooed
labor service. The emperor remarked that he had himself seen men
sentenced to registered control begging in the marketplaces. 50
These forms of penal registration were used most often to punish
officials,51 sometimes as special mitigations of the death penalty. 52 The
use of these penalties for officials perhaps explains the rule that such
persons were not to be registered in any prefecture within five hundred
lid of the capital. 53 Nonofficials occasionally were also punished in
this way. An order of 1105 refers to doctoral graduates who had been
sentenced to "registered control" but who were now to be released. (The
order says that their good behavior after they had reached their homes
was to be guaranteed by relatives and that if thereafter they committed
crimes calling for exile or worse, left the prefecture, or slandered people,
these relatives were to be considered equally guilty.) 54 If not officials,
doctoral candidates were at least qualified to be officials, but these
penalties were also sometimes applied to people without qualification for
offices, including foreigners. An order of 1041 that sought to stem the
outflow of Chinese coinage, stated that if those seized for exporting coins

46 HCP 408.13a.
47 HCP 92.5a.
48 HCP 217.6b (1070).
49 HNYL 2683.
50 See an order that such men were not to be incarcerated (SS 31.577; HNYL 2739). See
also SS 31.581; WHTK 160.1460.
51 See SS 201.5017, and innumerable other examples throughout the sources.
52 HCP 185.7a.
53 SHY, hsing-fa 4.50a.
54 HCP 25.18a.

400
 Penal registration

were not Chinese, they were to be sentenced to "registered control." 55

Ordinary male commoners also could be so punished. A decree of 1029
ordered that those in Ho-pei who were recidivists at damaging embank-
ments (men who were hardly likely to have been officials) be sent for
registered control. 56 In 1031 the punishment was applied to a commoner
who had killed the man who had earlier beaten his father to death.
Though convicted, the murderer had been freed by an amnesty, and the
son had taken vengeance. The ordinary penalty for such a homicide
would have been death, but the sentence was reduced because of the
circumstances. 57 During the Southern Sung, the system was also occa-
sionally used for commoners 58 and at least in the 1180s seems to have
been the standard punishment for horse thieves in some areas. 59
People who had been sentenced to these registration punishments
might be forbidden to return to their homes after release from penalty.
The rule in these cases for people who had been sentenced under the sys-
tem of joint adjudication was that after six years in the area of registra-
tion, they were to be entered on the household registers as regular
commoners. These registers were to indicate the clauses under which
they had been convicted. The village officers were responsible for keep-
ing track of their whereabouts. Officials were not to trouble them. And
having registered in one place, they might under some circumstances be
allowed to transfer their registration to another prefecture, after having
secured official approval. 60

Women in the penal registration system

The preceding descriptions of penal servitude deal largely with penalties
inflicted on men. We know less about the treatment of women. But
according to the T'ang and Sung codes, "The law on women says that
they are not to go alone into exile. Therefore if they commit a crime
punishable by exile, they are not sent into exile but remain at home.
They are beaten with the heavy rod and perform labor at a fixed
place." 61
An order of 993 states that women guilty of crimes up to penalties of
exile were to be excused from registered labor. 62 However, women who

55 HCP 131.1a.
56 HCP 108.14b.
57 HGP 110.12b.
58 HNYL 2624.
59 SHY, hsing-fa 4.56b.
60 TFSL 523.
61 SHT 3.52. For the T'ang, see Johnson, The T'ang Code, p. 163.
62 SHY, hsing-fa 4.3b. WHTK 168.1459 dates this to 989.

401
 Law and order in Sung China

were convicted of having produced or taught others to produce black

magic poison (ku) were to be "expelled to the frontiers of the empire in
order to cut off this evil." 63
For the first sixty years of the Sung dynasty some women were sen-
tenced to labor in the capital's workshops. The wives of men exiled to
the Sung maximum-security facility on Sramana Island off the north
coast of the Shantung peninsula were sent to such workshops from the
beginning of the dynasty, though this was said to have been stopped at
some unspecified date. We also know that as late as 1018, soldiers' wives
guilty of adultery were being held as workers in these workshops. In that
year it was ordered that such women no longer be sent there and that
the 157 women currently being so held were to be released.64
The punishment of arranged control was also applied to women,
primarily when they had been convicted under the provisions for joint
adjudication of household members of certain classes of criminals. An
order of 1083 abolished the system under which the wives and depen-
dents of counterfeiters had been subject to arranged control. 65 Wives of
robbers from heavy-penalty places, who because of their husbands'
crimes had been sentenced to "arranged control" in neighboring prefec-
tures under the edicts of the Yiian-feng period (1078-85), were ordered
released during the antireform period, but after 1098 we have a report
indicating that they were still being held. 66 According to the rule pre-
served in the Ch'ing-yiian t3iao-fa shih-lei, women who were liable for such
penalties were to be beaten twenty blows on the back with the heavy rod
but were spared penal labor at their place of registration. 67 Women were
occasionally punished in other ways under the joint adjudication rules.
In 1070, for example, the wife of a criminal who had been executed by
slicing was registered with a soldier who had no family.68

Basic factors affecting the use of penal registration

Three factors were particularly important in determining the form of the
penalty. First, the state attempted to make the level of punishment fit the
crime. Second, the state used the penal registration system as a primary
supplier of labor, so that the location of registration and its length were

63 See Johnson, The T'ang Code, p. 163; SHT 3.52.

64 SHY, hsing-fa 4.1a; HCP 91.7b; SHT 3.51.
65 HCP 333.7a.
66 SHY hsing-fa 4.30a (1088); HCP 499.16b-17a.
67 TFSL 520.
68 HCP 2i3.i9b-2oa.

402
 Penal registration

partly determined by labor needs. And third, the state tried to achieve
the two preceding goals without creating serious problems of public
security. Because these goals were often in conflict, imperial policy was a
compromise in which competing aims were partially but never wholly
fulfilled.

Labor needs as a determinant of policy

T h e system of penal registration, with its prison and base citadel com-
m a n d s , was designed to punish convicts a n d to deter others and also to
provide the state with large amounts of labor at an acceptable cost,
without endangering internal security. Labor might or might not be
d e m a n d e d of penally registered convicts and, if demanded, might differ
in kind a n d length. T h e kinds of labor involved are indicated by a prison
official ordinance (yii-kuan ling) found in the early Sung Code:

If the various convicted criminals who are liable for registration and labor are in
the capital city, they are to be sent to the Directorate of Palace Buildings
[Chiang-tso chien]. 69 Women in the capital should be sent to the Directorate
for Imperial Manufactories [Shao-fu chien] 70 to work as seamstresses. In the
provinces they are to do official labor at their supervising jurisdiction. If the
jurisdiction has no enterprise that employs official labor, then allow them to be
held in that prefecture to work on the city walls, in the granaries, or as miscel-
laneous servants in the public offices.71

An order dated 1023 g i v e s further information: Convicts serving in the

capital city were to work in one of the eight crafts offices that provided
the government with such services as plastering, painting, varnishing,
stoneworking, tileworking, masonry, well building, and working with
b a m b o o . These m e n would labor during the day and at night return to
their camps, where the c o m m a n d in which they were registered was
responsible for controlling and feeding them. 7 2 According to the Ch'ing-
yiian t'iao-fa shih-lei, they were allowed one day of rest in ten and seem
also to have been given time off on several Sung holidays. 7 3
C o m m o n labor for use around the prefectural towns might have been
hired on the market or required of other groups as a service; the presence
of the provincial army units meant that the state h a d less need to tax

69 An organization responsible for palace construction and maintenance, under the

Ministry of Works. See Hucker, Dictionary, no. 708.
70 Hucker, Dictionary, no. 5098.
71 SHT3.51.
72 SHY, hsing-fa 4.10b.
73 TFSL 144.

403
 Law and order in Sung China

people to pay for hired labor or to exact corvee, though of course the
taxpayers bore the burden indirectly by supporting the cost of the
provincial armies. The need for a pool of laborers was also increased in
the Sung by the founding emperor's decision to staff the postal sys-
tem using men drawn from the provincial armies rather than corveed
laborers drawn from the general population. Convicts were also often
used as workers in mints. 74 In 1154 the Ministry of Works successfully
petitioned to have convicts transferred to certain mints that were short of
labor, 75 and the Ch'ing-yiian t'iao-fa shih-lei indicates that heavy labor
might mean work in the mints. Although most convicts performed un-
skilled or semiskilled labor, the miscellaneous work assigned to convicts
might include clerical tasks, sometimes quite responsible ones. The
Ch'ing-yuan t'iao-fa shih-lei reports that those who were not physically
capable of physical labor might be used as underlings, 76 and this seems
to have been a common way of forcing officials serving sentences of
registration to contribute their labor.
Prefectures in need of laborers sometimes successfully petitioned to
have the various circuits send criminals liable for registration in certain
categories. An order in 1072, for example, states that except for certain
categories of criminals, when convicted of crimes calling for tattooed
penal registration, farmers from an enumerated list of circuits were to be
sent to Hsi Prefecture in Ch'in-feng Circuit. This rule followed an appeal
by an official who cited the region's need for skilled farmers to exploit
unused lands. Similarly, permission was granted in 1082 for sending to
Lan Prefecture in Ch'in-feng Circuit all miscellaneous criminals sen-
tenced to registration at one thousand to two thousand lid for registra-
tion in the base citadel commands there. 77 General policies governing
the location of registration can thus only partially explain what actually
happened, as individual imperial orders were also issued that sent crim-
inals from certain jurisdictions to other specified jurisdictions, either in
response to pleas for more labor or because of security considerations. 78
Convicts already registered might also be transferred elsewhere when
labor was needed.

74 SHY, hsing-fa 4.320-333.

75 SHY, hsing-fa 4.35a-b.
76 TFSL 520.
77 HCP 239.15a-3b. See also HCP 228.2b-3a with regard to the need for laborers in the
capital, and HCP 330.1 ib for the 1082 request from Lan Prefecture. See also a request
dated 1121, that men be sent to Yen Prefecture, because many of its prison citadel
troops had been killed or injured or had fled during the disturbances (of the rebellion of
Fang La) (SHY, hsing-fa 4.39a).
78 SHY hsing-fa 447a~48a.

404
 Penal registration

Security as a determinant of policy

In these policies we can see reflected the not always commensurable
desires to punish the prisoner appropriately and to provide the state with
labor where it was needed, but these goals had to be balanced against
the need to maintain internal security. The convicts, after all, were
criminals. They could cause severe problems wherever they were sent,
especially as the typical criminal subjected to penal registration in the
Sung seems to have been a young male. Such men were needed for
their labor but had to be handled with care. Convicts seem often to
have become involved in further crimes, for which they were sometimes
brutally punished.79
Security considerations are clearly reflected in policies restricting
penal registration in certain sensitive areas, in some cases forbidding it
entirely and in others limiting the number or sorts of criminals to be sent
for registration. Such considerations are no doubt the background to an
approved memorial in which local officials asked that dangerous crim-
inals no longer be sent to the northwestern frontier, where they often
escaped and joined tribal groups that preyed on the people.80
The Ch'ing-yiian t'iao-fa shih-lei includes a rule that when the people
liable for registration were from prefectures on the border or one in from
the border, they should be registered in the next prefecture farther away
from the border.81 This is an interesting change from T'ang practice.
Convicts in that dynasty were apparently always sent to the borders.
Perhaps the Liao and Chin dynasties that bordered the Sung to the
north were thought to be more attractive to potential turncoats than had
been the Turks and other frontier tribes of T'ang times. Perhaps men
during the Sung felt less strongly about staying within China's borders.
Or perhaps the Sung state was more fearful of defectors. This concern
about security on the borders was naturally greatest during times of foreign
wars. Thus, in 1043, during the conflict with the state of Hsi-hsia, the
Court ordered that "criminals from the various circuits from now on are
not to be registered in the border prefectures of Ho-pei."82 Eventually
there seems to have been a policy of moving criminals from the center of
the empire outward toward the periphery. This policy could have con-
flicted with the authorities' concern about sending dangerous criminals

79 SHY, hsing-fa 4.64b-65a, in which officials are said to have put such recidivists into
the cangue, chained their feet, and left them in jail until they died.
80 In 1143 the policy was modified slightly to allow convicts from a small group of
prefectures to be registered in Shensi and Szechuan. It was decreed that men should
not be registered in the border prefectures in these areas; compare HCP 142.7a,
145.17a.
81 TFSL 520-21. 82 HCP 142.7a.

405
 Law and order in Sung China

to militarily sensitive border areas, but it could be followed if men were

sent to less sensitive peripheral regions, including the south or west.

Capital region
There was also a continuing concern about security in the capital area.
At the beginning of the Northern Sung, all serious criminals had been
sent initially to the capital. From there, those not needed in the capital
would be sent to other areas for registration. An official at this time
compared the Sung system unfavorably with that used in antiquity,
because of the security problems presented by having large numbers of
criminals in the capital:
All prisoners from distant places are sent to the capital for registration as
workers. This is extremely inappropriate for the sacred territory where the
emperor dwells. How can we have exiled prisoners crowd together as laborers? I
hope that from now on the criminals of outer jurisdictions will not be permitted
to be sent to the capital or detained for work in the capital workshops.83
The Sung state faced a dilemma. The capital, the largest city in the
empire and the home of the majority of bureaucrats, needed a large
number of convict laborers, but meeting this labor need posed serious
security problems. Such problems in the capital were increased by the
frequency with which prisoners escaped while being transported. The
closer a convict came to the capital, the more likely he was, after
escaping, to travel to the capital and join other criminals there. The
Sung authorities tried to solve these problems. In 982 an edict issued in
response to such critics said that convicts were to be sent directly to
the appropriate prison citadel command for registration. They were no
longer to be sent first to the capital. 84
Later policy reflected the continuing tensions between the desire for
labor and the fear about security. There are repeated orders about not
sending convicts to Kaifeng, but clearly they were not followed. Despite
an order in 982, a report in 1001 notes that men guilty of robbery or of
armed assault that had not resulted in death were still being sent to the
capital. It was ordered that such men be beaten, tattooed, and registered
at other base citadel commands at five hundred lid or more. 85 Again,
despite this order some men still had to be sent to Kaifeng for review
and distribution. 86 And in 1013 the authorities decreed that serious

83 HCP 23.17b.
84 HCP 23.17b; SS 199.4969.
85 HCP 49.5a.
86 HCP 63.13b.

406
 Penal registration

criminals from the outlying circuits of Kuang-nan, Fukien, and Szechuan

were to be sent to the capital for labor; lesser criminals could be regis-
tered elsewhere by the circuit intendants. 87
Similar problems plagued the Southern Sung. Concern for security in
the capital presumably was behind the practice before 1143 of not
registering some kinds of convicts in the capital. And a rule from the
Ch'ing-yuan t'iao-fa shih-lei says that residents of the capital who were
sentenced to be registered in their home prefecture base citadel com-
mand were to be registered in neighboring prefectures. 88

Other sensitive areas

No doubt security concerns also explain the Northern Sung rule found in
the Ch'ing-yiian t'iao-fa shih-lei, which says that men convicted of piracy
were not to be penally registered in prefectures by the sea. This same
ordinance also forbade registering men in the capital or in the border
areas of the three circuits of Ching-chi, Ching-tung, and Ching-hsi
surrounding the capital, or in the circuits of present-day Szechuan and
Shensi. It also gives a short list of prefectures where especially evil
criminals were not to be registered. These prefectures are located in the
Ching-hu North and South circuits, areas of central China with large
non-Chinese tribal populations. 89

The quota system and the breaking up of gangs

The concern for security was also reflected in other general policies by
which the authorities sought to prevent excessive numbers of convicts
from being registered in a single prefecture. One such policy was estab-
lished as early as 1027. ^ n t n a t v e a r t n e prefect of T'ing Prefecture in
Fukien complained that currently there were 359 convicts in his jurisdic-
tion. He felt this to be a dangerously large number and asked that some
of them be transferred elsewhere. The central authorities approved his
request and stipulated that "in general, when the various prefectures
memorialize that they have too many convicts, they may follow this
order as a precedent." 90 Prefectural quotas were also fixed, because each
prefecture had a set number of provincial army units, and these units -

87 SHY, hsing-fa 4.6b.

88 TFSL 520.
89 TFSL 522-23-This rule presumably dates from the Northern Sung, as it specifies the
three circuits that were lost to the Sung in the 1120s.
90 SHY, hsing-fa 4.14b.

407
 Law and order in Sung China

mainly the prison citadel commands - had quotas limiting their numbers
of personnel.91
Under the ordinance found in the Ch'ing-yuan t'iao-fa shih-lei, when the
number of men registered in a jurisdiction exceeded its quota by 50
percent, this was to be reported to the judicial intendant, who would
investigate to verify the facts. He would report to the Department of
State Affairs (Shang-shu) and the Ministry of Justice, so that penal
registrations in that prefecture could be suspended. The prefecture was
also responsible for submitting an annual year-end report giving the
number and names of men currently held in the various forms of penal
registration.
The quota system limiting the number of men sent to certain areas
was closely related to a policy under which gangs were to be broken up,
their members scattered in various locations. 93 Too many convicts in an
area spelled trouble. The essence of the problem was captured in the
vivid phrase of an official who wrote in 1191 to complain about the
policy of sending to Hainan Island all the violent criminals who had
been spared execution. It was, he said, like "putting a hundred tigers
together on one hill." 94
From at least the 1040s the judicial intendants were charged with
overseeing the numbers of such dangerous criminals and diverting them
from overcrowded prefectures to less crowded ones. 95 Under the rule
found in the Ch'ing-yuan t'iao-fa shih-lei, members of gangs who had been
convicted of robbery, gang killings, arson, and theft were not to be
moved as groups, nor were they to be registered in the same prefecture. 96

Documentary controls
The Sung authorities also sought to bolster their control over the con-
vict population by instituting an elaborate system of documents that
indicated the number and characters of the convict population in each
jurisdiction. In addition to the periodic, including annual, reports sent
up from each prefecture enumerating those held in all forms of penal
registration, 97 reports were sent up to superior agencies on the day when
new prisoners arrived in a jurisdiction. In the cases of the more serious

91 For example, a report indicates that the quota for the prison citadel command in Yen
Prefecture was two hundred men (see SHY, hsing-fa 4.39a).
92 TSFL 522.
93 SHY, hsing-fa 4.25a-b.
94 WHTK 168.1461.
95 SHY, hsing-fa 4.21a.
96 TFSL 523.
97 SS 200.4992.
408
 Penal registration

locations for registration, separate reports were supposed to be sent to

the Ministry of Justice within five days of the prisoners' arrival. Deaths
or escapes were also to be reported in the same way. A form for
reporting on the convicts in a jurisdiction has been preserved in the
Ch'ing-yuan t'iao-fa shih-lei:

Have Received for Control:

For registration on the army rolls, so many men tattooed on the face.
Convict's name; number of years (Note: This means the number of years since
registration); name of jurisdiction; sort of person (indicate which army, and
then write that the origin of the person was as a volunteer or someone penally
registered for service; the nature of the convict's current offense) (Give a brief
description, e.g., if the convict were convicted of stealing, write "with force" or
"a thief"; write whether the convict was the principal or an accomplice,
whether or not he was armed, and whether or not the criminal act was
actually consummated. For homicides, merely note whether the crime was of
the principal categories of homicide or was classified as a miscellaneous
offense, whether the convict was the principal or an accomplice, and whether
in fact anyone was killed.98 The formats for reporting on the types of penal
registration called arranged or detained control, and condemnation to slavery
are similar."
In accordance with the sentence (Note: This means the light rod, heavy rod,
penal servitude, or exile), the prisoner has been sentenced to penal registration
in the prison citadel commands or other provincial army commands of such-
and such a prefecture or in such-and-such a jurisdiction for heavy labor in the
army. 100

Locations of registration and the transportation of

prisoners
Location of registration was certainly one of the most important ways of
grading the severity of a penalty. A convict's sentence was also graded
according to the distance from his home (or in the case of military
personnel, from their permanent camp). 101 For transient commoners or
soldiers without an assigned camp, the authorities used the jurisdiction
in which they had been convicted. 102 The least serious penalty involved
registration in the home prefecture. Next in degree of severity was

98 In T'ang and Sung law, both attempted homicide and actual homicide were covered
in the state statutes, hence the phrase "whether in fact anyone was killed."
99 This is the only example known to me in which there is any suggestion that during the
Sung, people were registered as slaves as a punishment.
100 TFSL 524-25.
101 TFSL 520-21, 526. Transients used the prefecture of conviction.
102 TFSL 526.

409
 KUANG-N AN
WEST C I R C U I T

"DISTANT EVIL"
PREFECTURES

CH'ANG-
HU
A HAINAN
( ISLAND
\ X N WAN AN
1U-YA1N»

Map 12. i. "Distant evil" prefectures

 Penal registration

SRAMANA
ISLAND

TENG ^
) PREFECTURE

Map 12.2. Sramana Island

registration in a neighboring prefecture, then at prefectures approxi-

mately 500, 1000, 1500, 2000, 2500, or 3000 lid away.
According to the regulations, when choosing a site for penal registra-
tion, the authorities were not supposed to exceed the assigned distance
by more than three hundred li. If there was no appropriate prefecture
within this limit, they might register the convict in the nearest prefecture
411
 Law and order in Sung China
103
beyond the limit. Very serious crimes might be punished by registra-
tion at one of the thirteen "distant evil" prefectures, all but two of
which were on Hainan Island or the adjacent coast of Kwangtung, or at
the most dreaded place of all, Sramana Island off the north coast of
Shantung (see Maps 12.1 and 12.2).
The Sung government established an elaborate system for moving
prisoners, individually or in groups. Each prefecture through which the
prisoners passed was responsible for detailing imperial army soldiers,
servitors minor, and clerical personnel to move the men from one border
of the prefecture to the other. The use of soldiers for this duty was
traditional; the Sung inherited the practice from earlier dynasties and
continued it from the beginning of the dynasty to the end. 104 Stages of
march were fixed, with prisoners expected to move about fifty lid (about
seventeen miles) per day. 105 An elaborate documentary system was used
to avoid abuses and to allow superiors to monitor their subordinates'
performance. Many of the rules governing this system have been pre-
served in the Ch'ing-yuan t'iao-fa shih-lei. One ordinance cited in that work
gives some idea of the way in which the system worked:
After people have been sentenced to penal registration, arranged control, or
detained control, record on a form their crime, the names of their accompanying
family and dependents, the amount of their valuables, and their home. (Note: If
the principal criminal is an active official, do not record the wealth of family and
dependents.) This form is to be turned over to the men detailed to transport the
prisoners. A calendar-ledger should also be provided, on which the districts and
market towns [chen] should record the date [on which the prisoners arrived]. If
prisoners become ill, then there should be a certificate of verification. If it is
necessary to spend money [from the prisoners' valuables], this is permitted. The
functionaries in charge should themselves enter the amount on the travel docu-
ment and stamp it, turning it over successively. It will be examined to compare
receipts. When those who have crossed the boundary into another circuit reach
the prefecture or district that is their destination, they should inform the judicial
intendant of the circuit, who will check out [the document]. After they have
reached the prefecture where the prisoner will be registered and have been
accepted there, a report should be sent back to the office that originally pro-
nounced sentence.106
The office that had pronounced the sentence also kept a log indicating
which prisoners had been sent where and when they should arrive at
103 TFSL 520-21.
104 Although this seems clearly to have been the practice, the only specific statement I
have found on this point merely indicates that the rules on the system existed before
1133. See, for example, SHY, hsing-fa 4.43b—44a.
105 SHT 3.45.
106 TFSL 521.

412
 Penal registration

their destination, and on receipt of the notification of arrival (or of death

or illness), it was to note this on its records. If the expected report did
not arrive - indicating that for some unexplained reason a prisoner sent
out from the office had not reached his destination in the time originally
projected - the office was to inform the circuit intendant so that the
problem could be investigated.107
When for some legitimate reason, such as illness or bad weather,
prisoners did not keep to their projected schedule of movement but
instead were detained at some point en route, the guards accompanying
them were supposed to inform the authorities of the situation. The
officials in this jurisdiction had the power to permit such delays but were
to check on the matter daily. If in a group of prisoners one man fell ill,
he might be allowed to delay his movement while his companions con-
tinued their journey. 108 Men might also be granted compassionate
delays. According to the early thirteenth-century code, if en route to
registration, men learned of the death of a parent or grandparent, they
were to be granted a delay. They could also be permitted delays if there
were illnesses or deaths among their accompanying family members or
dependents or if someone in the party gave birth. 109
Early in the dynasty there was supposedly a practice of men going into
exile being moved together at the end of each season,110 but in practice
the men seem to have been moved on an irregular schedule. By the end
of the first quarter of the eleventh century, because of concern that delays
in transport caused suffering among the prisoners, there were special
orders regarding speedy movement.111 However, during the dead of
winter and in the period of greatest heat during the summer, prisoners
were kept at the place where they had been convicted and were moved
only when the weather made this feasible.112 An ordinance found in the
Ch'ing-jyuan t'iao-fa shih-lei says that in winter (from the eleventh month to
the end of the first month), men who were to be transported for penal
registration in the armies were to be detained for labor either in the
jurisdiction in which they were convicted or in the jurisdiction at which
they had arrived. When the second month arrived, they were to be sent
on. (If they volunteered to go on, this was allowed.) When the circum-
stances of their crimes were grave, when they were to be registered in the
far south (in Kuang-nan), or when they had already entered the circuit
107 TFSL521.
108 TFSL 530.
109 TFSL 523.
110 SHT3.45ff.
111 SHY, hsing-fa 4.11 b-12a.
112 SS 201.5018. See also an order of 1071 that men in winter be held until the fifteenth
day of the second month of spring (SS 15.280). See also WHTK 168.1460.

413
 Law and order in Sung China

to which they were being sent, this ordinance was not to be applied. 113
Prisoners were allowed to bring along family members and dependents
(though this rule did not apply to men sentenced to registration on
Sramana Island). The policy of allowing family members to accompany
convicts had very practical benefits. It was perhaps connected with the
ideal of reformation and reintegration. Having a family or dependents
at the place of registration gave the convict a system of support, an
immediately present motive for good behavior, and a place to which he
could turn on finishing his sentence. If the family or dependents were in
another jurisdiction than the one in which the prisoner was convicted,
the convicting jurisdiction was to send a document to the place where
they were, indicating that they were allowed to proceed to the place of
registration. If the convict died, those who had accompanied him were to
be a allowed to return to their homes. 114 At the beginning of the dynasty
it appears that dependents and family could not avoid accompanying the
prisoner; an order from the early eleventh century said that when men
were being sent for registration to a prefecture other than their home
prefecture, if their wives and children did not want to accompany them,
they did not have to do so. 115
The convicts being moved - and after 1026, those accompanying them
- were supposed to be given provisions at post stations and granaries.
(Children under three years of age did not receive a ration.) This same
rule applied to those who for some legitimate reason had been delayed
en route. 116 A system of ration tickets, which could be exchanged for
food, was used. A decree of 1027 mentions a "Document for Enumerat-
ing Goods Needed en Route" which each successive escort group was
supposed to check carefully and transfer to the next group of guards.
Disbursement of provisions was to be entered on this register. 117
According to the rules, the state was supposed to pay for provisions,
but as other sources indicate, this rule was often ignored in practice, and
the men sometimes had to beg for their food.118 The sources that refer to
the system in which guards might be authorized to spend some of the
prisoners' funds presumably means using such funds to provide for the
convicts' needs. The plight of accompanying families and dependents
seems to have been particularly grievous. The guards sometimes limited
their task to moving the prisoners themselves, "leaving the dependents

113 TFSL523.
114 TFSL521.
115 HCP 107.3a.
116 TFSL 530; HCP 104.20a.
117 SHY, hsing-fa 4.14b.
118 SHY, hsing-fa 4.i7b-i8a (1031).

414
 Penal registration

to be scattered and lost on the roads so that few live to return." 119 Their
problems were especially desperate if the man they were accompanying
died en route. An order of 1004 indicates that in such cases the family
and dependents should be given passports allowing them to return
home, but nothing is said of giving them food or money for such a
trip. 120
All prisoners were accompanied by guards. Under most circum-
stances, such guards were responsible for moving their prisoner or
prisoners from one boundary of their jurisdiction to the other, where
they were to turn over the prisoners and their accompanying documenta-
tion to a new group of guards. Thus, the prisoners were passed from
hand to hand until they reached their final destination. The guards were
to wait at the exchange point if their replacements had not yet arrived
and, when the exchange had been completed, were to return to their
base camps. 121
An exception to this practice of having guards stay with a group of
prisoners for only short distances was made in the cases of violent and
dangerous criminals being sent for registration in Kuang-nan or Hainan.
At least during the thirteenth century the guards for such men were to
accompany them, not from one prefectural border to another, but from
the prefecture that was the headquarters of one circuit intendant to the
prefecture that was the headquarters of the next circuit intendant along
the specified route, before turning the prisoners over to the next relay of
guards. 122
The number of guards per prisoner varied depending on the danger-
ousness of the criminals, as reflected in the places to which they were
being sent. Thus, according to an order of the late Northern Sung, it
appears that there was to be one guard for criminals being sent to
locations of less than two thousand lid away, two guards for places listed
as more than two thousand lid away, three guards for those destined
for Hunan and Kuang-nan, and five for those being sent to the sea
islands. 123 In each prefecture one civil service official was supposed to
be in charge of transporting the prisoners, acting jointly with a local
military official. They were to prepare a register listing the names of the
imperial army soldiers who would be acting as guards. These men would
remain in their encampments, awaiting orders.
The dangers associated with such transporting of prisoners varied

119 HCP 101.9a.

120 SHY, hsing-fa 4.3b.
121 TFSL529.
122 SHY, hsing-fa 4.65a-66a.
123 SHY, hsing-fa 4-37a-b.

415
 Law and order in Sung China

depending on the characters of the convicts and their number. The rule
was that no more than ten men were to be moved together as a group.
When there were more than ten men, they would be moved in separate
groups of ten, and those not yet sent would be held in jails. 124 If more
than seven were being sent as a group (not counting women or males
fifteen years old or younger), if the criminals were considered dangerous,
or if the area were near the borders, the guard of imperial army troops
was to be accompanied by a superior such as an officer (chiang-chiao),
an office chief (ya-ch'ien), or an inspector (yu-hou)}25 Even with such
precautions, however, moving prisoners could be dangerous. Huang
Kan (i 152-1221) mentions a fort soldier sent to escort two robbers to
registration who was attacked by them and badly beaten. 126
Although the guards sometimes were abused, it seems that abuse
by the guards was far more common. Because the guards could be
penalized if they failed to meet their time schedule, they naturally
sometimes abused and beat their prisoners in order to make them move
more quickly.127 They also sometimes used their prisoners as beasts of
burden, having them carry bundles or packages.
The authorities also set penalties for functionaries who did not fulfill
their responsibilities in this system. Those who should have transported
prisoners but did not, or did but avoided the regular post-roads, illicitly
appropriated the goods of those under their charge, failed to show up at
the meeting place, or sent insufficient numbers of guards were to be
punished. 128

Fugitives
Probably the most widespread problem was prisoners escaping. Such
escapes, from jails, while in transport, or after having reached the place
of registration, seem to have been extremely common. 129 After noting
that several thousand men each year were sent for registration in Kuang-
nan and Hunan, one commentator observed, perhaps with some rhetor-
ical exaggeration, that "some escape en route and some after they
arrived. Those willing to reform and serve as laborers in the prison
citadel commands are not 10 to 20 percent." 130
Escape was a constant problem because, given the technical condi-
124 TFSL 529-30.
125 TFSL 529-3°-
126 Huang Kan, Mien-chai chi, 33.42a~43a.
127 HCP 83.16b.
128 TFSL 528, 529.
129 SHY hsing-fa 4.14b (1027).
130 Wu Ghing, Chu-chou chi, 2-5a-6a.
416
 Penal registration

tions of the time, it seems to have been relatively easy. There are
repeated complaints about escapees who sought to return to the center of
empire, and especially to the capital region. 131 In the early Southern
Sung an official remarked:
Men subject to registered control all bribe their escorts, so that often they do not
arrive [at their place of exile]. Sometimes they just go out the gate [of the place
from which they are being sent]. Sometimes they go halfway and then turn back.
Even though we have the law on using imperial army troops as escorts, because
the regulations of the Shao-hsing period [i 131—63] do not establish rules for the
rewards to be given to informers, this leads the guards to follow their own
feelings and let people go free, so that the miscreants, lost en route, return in
droves to the capital or enter the markets to cause trouble.132
The general legal response to escapees was to increase their original
sentences by some degree and to penalize the responsible officials.133 It
was common to sentence those responsible to the penalty to which the
criminal in question had been sentenced, usually reduced somewhat in
severity. Chinese codes traditionally contained rules specifying such
punishments. The T'ang Code and the Sung Code of 962 say: "If the
responsible officials inadvertently allow prisoners to escape, they will be
given penalties equivalent to that to which the prisoners had been
sentenced, reduced by two degrees." 134 And a rule that the Sung had
inherited from the Later Chou states:
Those responsible local officials who have custody over prisoners and in-
advertently lose them - if they themselves or their relatives and dependents
recapture the escapees - may ask to be spared the penalty for allowing escape
through inadvertence. If other people capture them or the prisoners themselves
die or voluntarily surrender and confess, then the culpable officials may appeal
to be tried for the penalty assessed against the convicts [but] reduced by two
degrees.135
Guards as well as officials were covered by rules on escapees. Accord-
ing to a report of 1098, when prisoners escaped because of their guards'
negligence, the guards were to be tried for their escapees' crime, with
lesser penalties. Thus if the criminals had been sentenced to tattooed
registration, the guards would be punished with two years of penal
servitude. Otherwise they would suffer one hundred blows of the heavy
rod. However, if they had deliberately allowed robbers liable for registra-

131 HCP 5.24b (1070); SHY, hsing-fa 443b~44a (1133).

132 SHY, hsing-fa 443b~44a.
133 HCP468.7a-b (1091).
134 SHT 28.460.
135 SHT 28.461.

417
 Law and order in Sung China

tion to escape, the guards would be subject to registration in a neighbor-

ing prefecture if they were soldiers and to registration in their own city if
they were civilians. 136 (Later Sung codes included rules specifying
rewards to be given to men who rightly accused others of having allowed
prisoners to escape.) 137
Such escapees jeopardized the whole system of law enforcement. Such
criminals often returned to the scene of their crimes to take vengeance on
those who had accused or captured them. Moreover, if this happened
frequently, the local people would naturally become afraid either to
inform on robbers or to help in their capture, thereby leading to an
increase in their numbers. 138 Understandably, the authorities felt a
need to gather reliable information on the dimensions of the problem.
Eventually a system was instituted under which each season a report
had to be prepared about escapees, with the judicial intendants to
manage it.
The authorities were well aware of the difficulty of this task. Beyond
threatening guards with severe punishment, they also specified rewards.
A document from 1154 records different levels of rewards and penalties
for guards, depending on the difficulty of their tasks. If the servitors
minor who were used to move prisoners from the capital to Kuang-nan
or Hainan had twice completed such assignments, they would be
rewarded with a one-year reduction in the number of case reviews
(mo-k'an) through which they had to pass in order to be eligible for
promotion. If en route many prisoners died or if such deaths happened
on a second trip, they were to have an extra year added to their case
review requirements. Those guards who had twice accompanied pris-
oners to registration at one thousand lid or more, were on arrival to have
their case review requirements reduced by one-half year, but if there had
been many deaths or such deaths happened on a second trip, they were
to have a half-year added. Performances better or worse than those
named were to be treated analogically, with the increases or decreases to
14-0

stop at two years.

136 HCP 499.4a—b. Under some circumstances the penalties might vary. Thus it would
seem that those guarding prisoners held over for the winter were liable for only sixty
blows of the heavy rod when prisoners escaped because of their guards' negligence.
Perhaps the sentence in such cases was lighter because the prisoners had not originally
been their responsibility.
137 TFSL521.
138 See SHY, hsing-fa 447a~48a.
139 SHY, hsing-fa 4.52a.
140 SHY, hsing-fa 4.4a-b.

418
 Penal registration

Release of convicts
Registered convicts were only partly restrained by threats of punish-
ment. Probably equally important was the promise that good behavior
would lead to improvements in their lot. Penal registration was rarely a
perpetual punishment. The earliest Sung code includes a T a n g edict
that says that "all those sent into exile may be freed to return after six
years." 141 This edict expresses a general attitude embodied during the
Sung in a regular set of rules, which apparently were created piecemeal
and were not always enforced but which nonetheless set a general pat-
tern for releases.
Men were released for a variety of reasons. Those registered in the
penal provincial army units who had reached age seventy or were ill
might be mustered out according to rules laid down in edicts. 142 Some-
times such men were sent back to their homes and even given traveling
money. An order of 1105 freed men subject to registered control and
detained control to return to their homes, where their relatives were to
be responsible for guaranteeing their good behavior. There was some
concern among officials that this pattern of releases might reduce the
deterrent effect of forced labor, especially among hardened criminals. An
official complained in 1192 that
although from the time when the law on penal registration was established, there
have been time limits on releasing men, recently these rules have been ignored.
After one or two years, men are freed according to convenience. This causes men
to make light of the law, considering penal registration to be just like ordinary
life.143
In 1204 the Sung formally adopted a more general policy embodying
these principles, though such a policy may well have been common,
if informal, earlier in the dynasty. Registered criminals were divided
into two groups, minor criminals, whom the memorialist describes as
"villagers who in a single incident commit assault, kill or injure someone
or clerks guilty of corruption . . . who even if they escaped would not
cause trouble," and hardened criminals, "habitual criminals, who rob,
join together to deal in illicit goods, and have already killed or injured."
Members of the first group were to be registered for heavy labor in their
home prefectures. When they had completed their term of labor, they
were to be given regular registration certificates and released to become
commoners.
141 SHT3.8b.
142 SHY, hsing-fa 1.6b.
143 WHTK 168.1461.

419
 Law and order in Sung China

Men in the second group were considered too dangerous to release

into civilian society. As early as 1032 an order was issued that hardened
criminals in Szechuan were not to be allowed to return home, presum-
ably to prevent their taking revenge on those who had helped in their
prosecution. 144 Such dangerous criminals were to be registered for labor,
but when their terms were over they were to have their tattoos altered
and to be inducted into the regular army. 145 The official Wang Yen
(1138-1218) argues for using bandits guilty of a first offense or thieves
with several offenses as soldiers not only to relieve the shortage of troops
but also to increase village security. 146 Induction of young, strong con-
victs into the imperial army was a regular practice in the Sung, although
it appears that each case was in response to a specific request rather
than a result of a general rule. 147 In addition to these more or less
systematic policies for releasing convicts, the sources also note occasion-
ally that central government officials were ordered to review the lists of
men held in penal registration, in preparation for an order from the
emperor releasing some of them.
Perhaps as important as the possibility of being freed to return home
was the possibility of moving up in the grades of the provincial armies.
Many of those convicted were young men, presumably with few skills
and no reliable source of livelihood. It was a general practice that a man
being released from custody as a registered criminal could volunteer to
join one of the other provincial army units. The men in these units had
better living conditions and less burdensome forms of labor than did
those in convict units. For a young man with no certain future, the offer
of a steady and secure if not prestigious job must have been reassuring.
Thus these units could serve as a welfare system, providing subsistence
for men due to be released who had no place to return to and no pros-
pects of employment. The alternative, in the opinion of some officials,
would be to condemn the old and the weak to starvation and the young
and strong to a renewed life of crime. 148

Punishment - Sramana Island

The severest grade of registered labor was reserved largely for those who
had been spared the death penalty. During much of the Northern Sung

144 HCP 111.8a, but see also HGP 104.1b; SS 201.5017.

145 SS 201.5020.
146 Wang Yen, Shuang-chH lei-kao, SKCSCP 1971, 23-5b-8b.
147 On the induction of men in the regular armies, see, for example, HCP 128.6b—7a
(1040); SHY, hsing-fa 4.49a-b (1160).
148 SHY, hsing-fa 4.58a~5ga.

420
 Penal registration

such men were sent to Sramana Island, some ten kilometers off the north
coast of Shantung 149 (see Map 12.i). Registration here seems to have
been especially feared, perhaps because of the hard labor of salt produc-
tion and probably because of the isolated location, which made escape
difficult. The earliest recorded use of the island as a place for registering
prisoners dates from the Northern Han dynasty (947-50), about ten
years before the founding of the Sung. An officer of the Northern Han
army, tried for having lost a city, was "exiled to Sramana Island in Teng
Prefecture. This was the origin of this practice." 150 The Sung began
using the island in 961 after a deserter who had illegally carved an
official seal was seized and beheaded. Perhaps feeling that the sentence
had been excessively harsh, the emperor issued a decree that soldiers
guilty of some serious offenses calling for the death penalty should be
spared and instead registered at the island. 151 The island thus began as a
place for exiling certain military personnel, and it continued to be so
used until the end of the Northern Sung.
The sources mention specific cases of exiling military men to Sramana
dated 961, 1007, 1020, 1021, 1026, 1037, 1040? 1071, 1078, 1083, and
1088, most of which involved officers or noncommissioned officers, not
common soldiers. The men exiled here had been sentenced for rapacity
(meaning primarily venality in office), for killing suspected criminals
without authorization, for beating subordinates to death, for provoking
disorders among troops, for deserting while convoying official goods,
and, in one case, for attempting to murder a fellow officer while within
the palace precincts. 152 The number of violent crimes is striking, reflect-
ing perhaps a milieu in which the physical coercion of subordinates was
commonplace. Such military personnel may have composed as much as
half the island's convict population.
These cases also indicate that the island was used exclusively for
criminals originally sentenced to death who had been spared execution
and sent to Sramana for labor. It was not a place to which men were
sent under ordinary sentences of exile. Nor was it a place for ordinary
people. Commoners and common soldiers sentenced to death were per-

149 For more detailed information, see Brian E. McKnight, "Maximum Security in the
Sung: The Facilities at Sramana Island," Bulletin of Sung and Yuan Studies, no. 16
(1980): 8-22.
150 Li Shang-chiao, Chin-shih hui-yuan (Shou shan ko ts'ung-shu ed.), 5.9b.
151 SHY, hsing-fa 4.1b. See also Li Ghih, Huang Sung shih-ch'ao kang-yao (Important events
during ten reigns of the Sung) (Taipei: Wen Hai, 1968), 1.6b. There are some other
references that date the emperor's decree to the third year (see SHY, hsing-fa 7/1 a; SS
1.12a; HCP3.8D).
152 HCP 67.12a, 96.6a, 226.9a, 340.10a, 373.ib-2a, 408.19a; SHY hsing-fa 4.13b, 6.10b,
15b, 7.17a; SHY, ping 12.12a.

421
 Law and order in Sung China
haps more likely to suffer the actual capital penalty, whereas officers,
more likely to be spared, ended up on Sramana. Many cases say specifi-
cally that the criminal was spared by special imperial order.
From the early years of the dynasty, the island was also used for
officials guilty of capital crimes, whose lives had been spared. A report
from 964 says that the "commissioner of the Court of Artistic Ornaments,
Ch'ang Ch'en, was beaten, tattooed on the face, and registered at
Sramana Island." 153 Other cases of officials so punished are recorded for
966, 974, 982, 1015, 1016, 1026, 1039, 1044, 1048, 1053, 1083, 1107, and
1126. These officials came almost entirely from the low and middle levels
of the bureaucracy. Although in 966 the assistant commissioner of
military affairs (a very high post) was so exiled, ali the other cases
involved officials of rank 6B or below (in the nine-rank scale). There are
some cases with strong political overtones and a few that involve central
government bureaucrats, but the majority involved local administrators
guilty of various forms of corruption in office, the crime called rapacity.
During the earlier part of the dynasty these officials, like military crim-
inals, were beaten and tattooed before their exile. In 1069 it was ordered
that officials no longer be subjected to the humiliation of beating and
tattooing, but in fact they continued on occasion to be tattooed. 154
Officials were involved in slightly fewer than half the cases that I have
found. Again as with military officers, we may assume that they formed
a large percentage of the prisoner population. Although we cannot prove
in every case that the crime carried the death sentence, all the crimes
described could receive such a sentence, and in most of the cases we are
told that the officials were spared for exile. That a large proportion of
these officials had been in local posts when they committed their
crimes is noteworthy. Local officials comprised only about one-third of
the total body of Sung civil servants 155 but are a majority among those
exiled to Sramana. This disproportion probably reflects both the greater
opportunities for venality afforded by service in local posts and a concern
by the government about the quality of local administration.
The paucity of Sramana cases involving neither officials nor higher
military personnel probably reflects the lesser chance of nonprivileged
people of escaping actual execution. Most of the few cases we have are
recorded because of their peculiar circumstances - a son who committed
153 HCP 5.16b.
154 For cases concerned with exiling officials, see HCP 5.16b, 6.13a, 23.6b, 86.6b, 104.5a-
b, 147.3a, 174.9b, 339.6b, 340.11b; SS 2.4a, 3.6b, 8.12b, 24.11a; Li Chih, Huang Sung
shih-ch'ao kang-yao, 17.2b, 21.2a; SHY, hsing-fa 4.71b, 6.16a, 25a; SHY, chih-kuan
64.23a, 28a, 28b, 39a, 65.2b, n a , 66.26a, 26b, 68.14a.
155 For the proportions of officials in local versus central posts, see McKnight, Village and
Bureaucracy, p. 8.

422
 Penal registration

armed robbery on orders from his father, a man who anonymously and
falsely accused a low-level official of plotting rebellion, and a violently
anti-Buddhist doctoral graduate (chin-shih) who wrote a tract attacking
the religion and used Buddhist writings for his bedclothes. 156
From 1093 we have a more detailed description of the kinds of crimes
that might result in exile to Sramana. At that time, those on the island
were listed in two groups. The most serious offenders included those
who had committed violent robbery, killed others, been incendiaries,
masterminded rapacity involving fifty strings of cash, or committed
repeated acts of rapacity with a total value of three hundred strings
or one act involving booty valued at more than one hundred strings,
violently raped relatives, been convicted on two capital counts of fighting
and causing injuries, killed with premeditation, plotted or given the coup
de grace in an attack that resulted in the death of the victim, com-
mitted one of the Ten Abominations for which the sentence was death,
or made and fatally administered ku poison. The second group, guilty
of slightly less reprehensible crimes, included accomplices in armed
robbery in which a victim had been killed, and men implicated in
rapacity with a value of two hundred strings but not originally involved
in the plotting. 157
Other individual rulings indicate that men might be exiled there for
such disparate crimes as breaking the blue-white salt laws, being ring-
leaders among embezzlers of granary funds of ten thousand cash, being
guilty on two counts of hiding bandits, or being recidivists among some
kinds of bandits. Even noncommissioned officers who repeatedly refused
to drink the imperial health and call out "ten thousand years" were
threatened with exile to Sramana. There seems to be no thread on which
to hang all these crimes except that the state viewed them all as pro-
foundly threatening to its control of society and so declared them capital
crimes.
From the time of their convictions, the men sentenced to exile at
Sramana were a group apart. They were identified differently, treated
differently, and even became the objects of a distinctive system of paper-
work. Whereas most ordinary offenders seem to have borne tattoos
not larger than two-tenths of an inch, and escapees and those to be
registered at the citadel commands bore tattoos of five-tenths of an inch,
the men sentenced to Sramana (and to "distant evil places" in Kuang-
nan and on Hainan Island) bore tattoos of seven-tenths of an inch. 158

156 HCP 7.4b, 341.10a, 345.9b; SS 2.3a; SHY, hsing-fa 6.13b, 14a, 16b.
157 By far the fullest version of this list is given in HCP 468.6b, but there are other, more
or less abbreviated versions in SHY, hsing-fa 4.31a; WHTK 168.1460.
158 TFSL 520.

423
 Law and order in Sung China

Unlike some types of Sung convicts, the Sramana Island exiles were
forbidden reductions of sentence length in return for suffering fixed
numbers of blows with the heavy rod. Nor could their families accom-
pany them into exile. Indeed, at the beginning of the dynasty their wives
were also sent for registration, not at Sramana, but at the "needleworks"
(chih-chen, presumably a capital workshop where seamstresses worked).
This punishment of wives was soon stopped. 159
Special documents were drawn up by the convicting jurisdictions for
Sramana exiles, listing the criminals' native places, ages, crimes, the
statutes cited, and the judgments. These documents were forwarded to
the authorities in Teng Prefecture by the relays of guards that took the
prisoners to their destination. An error in these or their improper dis-
closure was punishable. 160 When men destined for Sramana Island or
Kuang-nan had completed their passage through a given prefecture, the
prefectural authorities were to report this to the convicting jurisdiction.
When the criminals finally reached the prefecture where they were to be
registered, the courier service was to be used to notify the authorities of
original jurisdiction. 161 The convicting jurisdiction, within a month of
sentencing someone to Sramana Island, Kuang-nan or "distant evil"
prefectures, mostly located on Hainan Island and the adjacent section of
the Kuang-nan east coast 162 (see Map 12.2), was to send to the Ministry
of Justice a brief memorandum indicating the nature of the punishment
and giving the day and month when the criminal was sent. When the
convict reached the prefecture where he would be registered, the auth-
orities there were to inform the ministry within five days. If some of the
men involved escaped or died en route, the prefectural authorities of the
place where this happened were also to send a report to the ministry
within five days. 163 Such escapes among men bound for Sramana Island
are said (in a report from 1025) t o have been common because all the
guards were lax, 164 though we might hypothesize that the convicts en
route there were also especially desperate to escape registration in such a
dreadful place.
A separate, permanent set of guards was stationed on the island itself,
possibly housed with resident families. We do not know how many
guards there were, but a decree of the early twelfth century increasing

159 TFSL 504, 521; SHY, hsing-fa 4.1a.

160 TFSL 529.
161 TFSL 530.
162 TFSL 520. This list of thirteen prefectures is taken from the early thirteenth-century
code. It is possible that the list of prefectures so classified changed over time, though
there are reasons to think that the Hainan prefectures were always included.
163 TFSL 522.
164 SHY, hsing-fa 4.i2a-b.

424
 Penal registration

the n u m b e r by two hundred men suggests that they must have been
n u m e r o u s . 1 6 5 Early in the dynasty these guards appear to have been
under the immediate control of the commissioner of military colonies
(t'un-ping shih).166 Later (1019) the fort on the island was under the
control of a supervisor of militia (chien-ya). Possibly officers of this title
had succeeded to the commissioner of military colonists, or it was merely
an alternative designation or a concurrently held title. 1 6 7 In that year the
abuse of prisoners on the island led the Court to assign to the military
servitor of the five islands (t'i-tien wu-tao shih-ch'en) the task of investigat-
ing such mistreatment. 1 6 8 T h e administrator of T e n g Prefecture and the
fiscal intendant of Ching-tung Circuit also participated in discussions of
internal problems on the island. 1 6 9 At the capital, as befitted an area
staffed by military personnel, affairs concerning the island were over-
seen by the Bureau of Military Affairs, although some central judicial
agencies might participate in decisions regarding particular problems. 1 7 0
Although under the general supervision of military personnel, the
criminals themselves seem in practice to have been housed with the
civilian families on the island. O n e report from 1058 says that there
were eighty households on the island at a time when there were more
than twice that n u m b e r of criminals (and an indeterminate n u m b e r of
guards). T h e families (probably mostly engaged in salt production)
a p p e a r to have provided shelter and food to those prisoners housed
with them and used the convicts as laborers. 1 7 1 After 1004 the resident
families were freed from the burden of taxes, not in return for services to
prisoners, but rather in return for taking care of the harbor facilities used
by ships bringing J u r c h e n horses across the gulf from Liao-tung. 1 7 2
Conditions on the island and the treatment of the prisoners provoked
repeated protests by the more h u m a n e officials. Sometimes direct, per-
sonal abuse of convicts came to the attention of higher authorities. In
1018 or 1019 the then military commissioner in charge of the fort on
S r a m a n a Island had two convicts murdered. O n e of these men had been
an assistant secretary in the Bureau of Editors. His son (who, given his
father's position, was probably both literate and familiar with official
channels) beat the message d r u m (with which ordinary citizens could

165 SHY, hsing-fa 4.33a.

166 HCP 21.31b; see also WHTK 168.1459.
167 HCP 93.2b; SHY, hsing-fa 4.9a-b.
168 HCP 93.2b.
169 HCP 188.10a, 246.6b-7a; SHY, hsing-fa 4.23b-24a, 26b.
170 HCP 188.1 oa, 5.24b, 246.6b-7a; SHY, hsing-fa 4.5a, 19b, 23b-24a, 26b.
171 SHY, hsing-fa 4.23b-24a.
172 SS 1.15a; HCP 4.20a. See also Chiang T'ing-hsi, Ku chin t'u-shu chi-ch'eng (Kuang-hsu
block printed), chih-fang tien, 273.a, K'ao 2a.

425
 Law and order in Sung China

lodge a complaint that would receive high-level scrutiny) and submitted

an accusation. The commissioner seems to have killed the men because
they offered him insufficient bribes. Unfortunately, with the key figures
dead, it was not possible to hold a proper investigation. In order to avoid
similar problems in the future, the military intendant of the five islands
was made responsible for investigating abuses of prisoners, and a decree
was issued that the fort military commissioners were "not permitted to
kill men sentenced to exile for reckless private motives." When prisoners
died as a result of beatings administered in the ordinary course of
custody, with no personal malice involved, the higher authorities did not
consider the responsible functionaries guilty of any serious offense.173
Prisoners deliberately killed by their overseers were probably greatly
outnumbered by those who died of starvation and neglect. Before IOIO
no provision had been made for giving rations to the exiles. In that year
a commissioner visited the island and reported that countless prisoners
had died of starvation there. A ration system was ordered established
but does not seem to have been properly enforced.174 In 1036 another
official asked that the prisoners be allotted one pint (sheng) of rice per
day, but in 1058 we are told that the prisoners "lacked any clothing and
food" and that as a consequence many died. 175 Perhaps as a result of
this inquiry, at some time during the ensuing decade, the state began to
provide rations for three hundred prisoners. 176 Unfortunately, too many
men were sent there for the quota of food provided.
The number of men sentenced to exile on Sramana seems to have
risen slowly during the Northern Sung. We have no figures for the early
years of the dynasty, but the report from 1058 states that at that time
there were 180 convicts on the island, 177 and the number continued to
grow. In 1069, after discussing with the Bureau of Military Affairs the
large number of convicts there, Shen-tsung ordered it to consult with
the Secretariat Department about solutions to the problem. Apparently
nothing substantive was done: One report in 1073 says that 650 crim-
inals were on the island, and another gives the figure as 401. 178
The report of 1058 notes that 200 to 300 men a year were sentenced to
exile there, which would mean 2000 to 3000 over a decade. Yet in that
year there were less than 200 convicts on the island. Several decades

173 HCP 93.2b (1019); SHY, hsing-fa 4.9a-b.

174 SHY, hsing-fa 4.5b; HCP 74.15b; SS 7.24b.
175 SHY, hsing-fa 4.19b, 23b-24a, 26b, HCP 188.10a, 2346.6a-b.
176 SS 344.23b.
177 HCP 188.10a; SHY, hsing-fa 4.23b-24a.
178 HCP 5.24b (1069), 246.i6a-b; SHY, hsing-fa 4.26b.

426
 Penal registration

later, at a time when there were between 400 and 650 convicts, only 93
men had been on the island four years or more. 179
The official reporting in 1058 blamed the small number on the appal-
ling death rate among the prisoners, and certainly that is a part of the
answer. Indeed, the situation had already become so notorious by 1036
that an unsuccessful attempt was made to close down the facilities. 180
However, the high death rate alone does not explain the relatively small
number of men on the island at any given time. Rather we must look,
during the early years of the Sung, to one of the key institutions in the
traditional Chinese legal apparatus, the amnesty system, and, during the
eleventh and twelfth centuries, to a combination of the amnesty system,
the prisoner quotas, and the transfer of prisoners as a result of specific
requests (sometimes for service in army combat units). 181
Ordinarily an amnesty resulted in the release of prisoners or reduc-
tions in their sentences. Exile to Sramana stood at the far end of the
penal spectrum. Beyond it lay only death sentences. If Sramana were
fitted into the regular amnesty pattern, then at least some of its inmates
would have been transferred following each amnesty to less fearsome
places of registration. This by itself would have helped keep the popula-
tion on the island at a tolerable level and would have lowered the
number of old-timers. But did Sramana fit into the regular system? We
do know that some amnesties applied to the convicts there. In 1013, for
example, we have a decree that the criminals on Sramana Island, except
for those who were to be sent to the capital, should have their cases
reviewed as a result of an amnesty. If the circumstances of their crimes
were not heinous, they were to be sent for registration to nearby terri-
tories on the mainland. 182 There were also some men exiled to Sramana
by decrees that specifically declared them ineligible for amnesty; that is,
they were never to leave the island. 183 Taken together, these reports
indicate that Sramana did fall within the regular amnesty system, which
contributed to lowering the numbers of men on the island.
From the 1070s and 1090s we have descriptions of the amnesty pro-
cess. The granting of transfers under amnesties depended on the length

179 HCP 188.10a, 246.i6a-b; SHY, hsing-fa 4.26a, 26b.

180 WHTK 168.1460; SHY, hsing-fa 4.19b.
181 In 1042 it was decreed that "the Bureau of Registers of Military Chiefs (Chiin-t'ou
ssu) may select strong and robust men from among the criminals being released for
return from Sramana Island and place them under the Commandancy of Stalwarts
Returning from Afar in the neighborhood of the capital; see SHY, hsing-fa 1.20b; SS
12399.8a (1042). In 1060, criminals were again shifted; see HCP 191.6b; SHY,
hsing-fa 4.24a. The amnesty system will be described in greater detail in Chapter 14.
182 HCP 80.6a; SHY, hsing-fa 4.7a; SS 8.6b.
183 SS 2.4a; SHY, hsing-fa 6.16a, 25a; HCP 339.6b, 373.ib-2a.

427
 Law and order in Sung China

of time that a prisoner had been on the island, his record while there, his
age and physical condition, and the crime he had committed. Under an
amnesty granted in the early 1070s, prisoners who had been on the
island for four or more years were divided into two groups according to
the seriousness of their crimes. Those whose crimes had been less serious
were transferred to the mainland. 184 In 1080, more men were freed who
had been held on Sramana for a number of years. From 1093 we have
our most elaborate description of the amnesty system as it affected those
on Sramana Island. The decree of that year first lists those particularly
heinous crimes that rendered their perpetrators ineligible for amnesty,
then the crimes that permitted transfer to Kuang-nan, and finally the
crimes that permitted transfer to Ching-hu North or South, or to Fukien.
If the men transferred to Kuang-nan exceeded the quota for that circuit,
the extra men could be sent to the "distant evil prefectures," on Hainan
and the south coast. If the quotas for Ghing-hu North and South or
Fukien were exceeded, the extra men were to go to Kuang-nan. Even
those criminals ordinarily excluded from transfers might be shifted if
they were sixty years old or older and had been on the island five or
more years (they could be shifted to Kuang-nan). If they had been
on the island for ten or more years, such men could be treated as
if they were ordinary criminals. The dangerously ill or those seventy
years old or older might be shifted to "nearby prefectures" (near Teng
Prefecture on the Shantung coast opposite Sramana Island). Even men
"perpetually" forbidden to move might leave, but they had to serve two
years beyond these conditions.185 Similar policies prevailed during the
latter years of the Northern Sung as well.186
The convict quotas mentioned in the decree on amnesty also affected
Sramana during the late Northern Sung. No formal quota seems to have
been set for the island until early in the reign of Shen-tsung. Then, in the
early 1070s, a quota of two hundred men was established, apparently
following a request by the administrator of Teng Prefecture. Any con-
victs beyond the quota were to be shifted to other places of exile, literally
"across the sea," referring to the nearby mainland of Shantung. This
policy was immediately attacked by the Bureau of Military Affairs and
the Judicial Control Office, which warned that it would weaken
the deterrent effect of punishment. These agencies asked that if men had
to be shifted, they be sent to Yai Prefecture, Tan b Prefecture, Ch'iung
Prefecture, or Wan-an Prefecture, that is, to the "distant evil prefec-
tures" located on Hainan Island. As a consequence of these criticisms,
184 HCP 236.24b; SHY, hsing-fa 4.26a.
185 HCP 468.6b (1093). See also SHY, hsing-fa 4.31a.
186 WHTK 168.1480 (1096); SHY, hsing-fa 4.34a ( n 12), 36b (1117).

428
 Penal registration

the quota was raised to three hundred, and apparently the number of
guards was also increased. 187
The orders issued as a result of this debate did not slow the flow of
convicts to Teng Prefecture and, in any case, were only partly to the
point. The larger quota of convicts and guards merely meant that more
men moved to the island without any corresponding improvement in the
already inadequate supply of housing or provisions. In the seventh
month of 1073 the administrator of Teng Prefecture submitted a new
proposal. He asked to be allowed to forward monthly to the Bureau of
Military Affairs a list of the prisoners on Sramana, giving their names,
homes, and crimes. The bureau could establish a register in which it
would record the information from these reports. As long as the Sramana
quota was filled, the bureau would not permit any more criminals
to be moved to Teng Prefecture (the prefecture that had jurisdiction
over Sramana Island). Men over the quota would be appropriately
transferred.
His recommendations were followed only in part. Although the in-
crease in guards was to stand, the authorities in Teng Prefecture were to
memorialize concerning convicts sent there by other areas. However,
those sent to the island by special imperial decree were not included in
this system. Because many of the men on the island were originally
sentenced to death and then spared by imperial order, this latter proviso
must have seriously weakened the impact of the general policy. 188 Some-
time later, one more attempt was made to deal with these numbers. A
system was set up under which Teng Prefecture was to inform the
judicial intendant when its annual quota was half full. The intendant
would investigate the report and pass it on to the Ministry of Justice.
When the quota was full, this was to be immediately reported to the
ministry, and at year's end the prefecture would send the judicial
intendant a report showing the names and number of men currently
held. 189
According to one gruesome and suggestive if exaggerated story, these
formal quotas had been preceded by an informal and brutal de facto
quota system. At some time before the reign of Shen-tsung, perhaps in
1058, a quota had in effect been set by providing for only three hundred
men. In the years just before the accession of Shen-tsung, the military
commissioner, Li Ching, was "solving" the problem of excess convicts
by throwing all men over the quota into the sea. In two years he had
killed seven hundred men. When Ma Mo was appointed administrator
187 HCP 245.14a, 246.6b-7a; SHY, hsing-fa 4.26b; WHTK 168.1460.
188 HCP 246.6b-7a; SHY, hsing-fa 4.26b.
189 TFSL 522.

429
 Law and order in Sung China

of Teng Prefecture, he was appalled by the practice, criticized Li directly,

and prepared to submit an impeachment. Li then hanged himself, and
Ma sent in a twenty-item reform proposal on running the island facil-
ities. In particular he suggested that when the quota was exceeded,
the oldest convicts who had not committed any offenses since arriving
should be repatriated to Teng Prefecture. As a result many lives were
saved.
Afterward, when Su Shih was administrator in Teng Prefecture, the
elders are said to have approached him in the street to ask if he would
govern them with the same care as Ma Mo had shown.190 In later times
a legend grew up around this story. According to Chao Shan-liao's
Collection of Self-admonitions (Tzu ching pien), shortly after Ma Mo had
secured Shen-tsung's assent to his proposed reforms, Ma was sitting in
his hall.
Suddenly it grew dark. As if in a dream he saw a gentleman coming through the
air, accompanied by a boy and girl. When the gentleman arrived in front of Ma
Mo, he cried out in loud voice: "I am from the Eastern Peak. The Sage Emperor
has a Heavenly Order for Ma Mo. You have no descendants. Now in light of this
business about transferring criminals from Sramana he has especially decreed
that you shall have a son and a daughter." Then taking the two children, he
mounted a yellow cloud and left. Ma rose in alarm. Later, as a result, he had a
son and a daughter.191
In the end the problems of the island were solved not by the Sung
authorities but by foreign invaders. When the Jurchen armies swept
into Sung territory in the mid-1120s, the government transferred all
the prisoners elsewhere or freed them and ordered the judges to send
men who would have been sentenced to Sramana to registration in the
"distant evil prefectures" in Kuang-nan, on a "temporary" basis. 192 The
Sung never returned, and the Chin do not seem to have used the island
as a place of exile. The brutal chapter in the island's history was over.193

Registration in "distant, evil" prefectures

After S r a m a n a Island as feared places of penal registration were the so-
called distant evil prefectures of K u a n g - n a n . T h e s e included, according
to a Southern S u n g listing, some thirteen prefectures. Four were on
H a i n a n Island - Yai, C h ' i u n g , T a n , a n d W a n - a n prefectures. T h e

190 SS 344.23b.
191 Cited in Chiang T'ing-hsi, Ku-chin t'u-shu chi-ch'eng, chih-fang tien 280, tsa-lu 2a.
192 TFSL 520. Men from Hainan who were being punished with registration were to be
sent to the "distant evil prefectures" in the mainland part of Kwangtung.
193 SHY, hsing-fa 4.36b; TFSL 526, 527.

430
 Penal registration

nine other prefectures were almost all on the mainland of Kuang-nan

opposite Hainan. All shared a general reputation for being unhealthy.
Malaria was endemic. No doubt the Sung authorities pictured the lives
of convicts in these places as nasty, brutish, and short. 194 Unlike con-
victs elsewhere, the men sent to the prefectures on Hainan Island seem
to have been used as soldiers to fight against the indigenous Li b tribes-
people, which led periodically to security problems caused by escapees
who joined forces with the natives. 195

Distance of registration and regional transfers

Below registration in " d i s t a n t evil prefectures" came penalties measured

in distance, at three t h o u s a n d , two thousand, one thousand, a n d five
h u n d r e d li.d T h e actual distances were modified by m a n y factors a n d ,
according to a Sung ordinance, could vary legally as m u c h as three
h u n d r e d lid from the specified distance. 1 9 6 T h i s division by distance
c a m e to be integrated with policies determining the transfers of prisoners
from region to region.
A n a t t e m p t was m a d e to move m e n guilty of serious crimes to regions
with which they would be unfamiliar, for example, moving southerners
to the north, a n d vice versa. H a r d e n e d criminals from the area known as
Ling-nan (corresponding generally to K w a n g t u n g ) were supposed to be
registered in areas north of the divide (in Ling-pei). 1 9 7 Serious criminals
from Ling-pei were often registered south of the divide (in Ling-nan). A
decree clarifying the terms of a n amnesty dated i o 18 gives some idea of
the way that the system then worked. T h e amnesty h a d reduced death
sentences to exile or penal servitude, b u t there was some confusion a b o u t
w h o w a s to be sent where. T h e decree stated:

Violent bandits who have injured men and were to be registered at Sramana
Island, if from Kuang-nan are to be registered at Ch'iung, Tan, Yai, or Wan-an
prefectures [i.e., the "distant evil prefectures" on Hainan Island]; if from I-chou
or T'zu-chou circuits, at Shang, Chun, Chin, Hsiang, or Teng prefectures; or if
from Li-chou or K'uei-chou circuits, in Ching-hu South Circuit. All are to be
held in the prison citadel commands. After being tattooed on the face, those who
did not injure people may be registered at one thousand lid plus citadel com-
mands. Where the penalty is less than exile, register them in their own cities. 198
[See Map 12.3.]
194 For a fine description of conditions on Hainan Island, see Edward H. Schafer, Shore of
Pearls (Berkeley and Los Angeles: University of California Press, 1970).
195 SHY, hsing-fa 4.i4a-b, 33b, 5gb-6oa.
196 TFSL521.
197 SS 199. 4971.
198 HCP 91.13b (1018).

431
Law and order in Sung China

HAINAN
ISLAND!
 Penal registration

HAINAN r> *)
ISLAND f C

Map 12.3. Sung circuits. CY = Ching-yuan circuit (1001); FY = Fu-yan circuit

(966); HC = Huan-ch'ing circuit (966). Hsi-ch'uan (I-chou and Tz'u-chou) was
a circuit by 969. Hsia Lu (Li-chou and K'uei-chou) was a circuit by 994 at the
latest. Ho-pei was a single circuit as early as 1004 and was four circuits from
1048.

Another decree, from the early eleventh century, provides information

on the working of this system:
From now on, all the bandits who are liable for tattooed registration at prison
citadels at a distance of one thousand li,d if they are from Ho-pei or Ho-tung all
should be registered south of the Yellow River; those from prefectures in Shensi
should be registered east of the T'ung Pass; those from prefectures in Ching-hu
South should be registered in Ling-nan; those from Hupei should be registered in
areas across the Han River; those from Chiang-nan and Liang-che should be
registered in areas north of the Yangtze River; those from Szechuan, in areas
outside the borders of that region; those from the prefectures of Kuang-nan near
the mountains, in the region north of those mountains; those from other parts of
Kuang-nan, in the Eastern and Western circuits should exchange prisoners or
transfer them to Fukien; and those from Fukien should be registered in Kuang-
nan or Chiang-che. 199 [See Map 12.4.]

199 SHY, hsing-fa 4.4a.

433
Map 12.4. Transfers of amnestied convicts: this page, north and central China;
facing page, south China

434
435
 Law and order in Sung China

Choice of prefecture for registration

Policies determining regions had to be integrated with policies for deter-
mining the specific prefectures to which men would be assigned. Choos-
ing the appropriate prefecture for registration was a complex process,
except for convicts to be registered at their home prefecture or at
maximum-security locations. The sentencing document might say that a
particular convict should be registered at a prefecture a certain number
of lid away, but decisions about where convicts were to be sent were only
partly determined by the distance noted in the sentencing document. In
theory this would have meant that any of a ring of prefectures at the
appropriate distance from the convicting prefecture could be used, but in
practice this was not so. There seem to have been some general policies
not now mentioned in the extant documents but reflected in the specific
examples preserved in the sources. These policies came to be embodied
in precedents (lie) that stemmed from individual imperial orders. 200
These policies apparently took two forms. One form, just described,
specified that criminals (sometimes only certain sorts of criminals) from
certain large geographical sections of the empire were to be sent to
specified other regions. The other sort of order indicated that convicts
from Prefecture A were to be sent to Prefectures B, C, or D if their
sentencing document specified punishment at a X li,d but to Prefectures
E, F, and G if the document specified Y lid and so on.

Problems and solutions

Although in many ways the system of penal registration remained
basically unchanged during the dynasty, some aspects were altered.
Such changes resulted from the continuing effort to make the level of
punishment fit more closely the crime committed and from the continu-
ing growth in the numbers of crimes for which penal registration was the
penalty, but the major impetus seems to have been the desire to limit
both the costs and the suffering of the convicts, their families, and their
guards.
These changes can be seen in the historical evolution of the system.
The Sung had inherited from the Five Dynasties the system of penal
registration, with its disparate and sometimes conflicting aims. These
five dynasties, four of which had their capitals at Kaifeng, controlled
only north China. Because no region under their control was excessively
far from the Kaifeng and because large amounts of labor were needed in
the capital, the general policy was that all serious criminals were to be
200 SHY, hsing-fa 4.50b (1163).

436
 Penal registration

sent there for sentencing. The Sung took over this system from the Later
Chou, and before 981 it appears that all criminals guilty of crimes calling
for penalties heavier than beating with the heavy rod had to be sent in
fetters to the capital for sentencing before being sent to their place
of registration, which often was in the northwest. Their families and
dependents accompanied them. At the beginning of the Sung, when the
state was confined to part of north China, such a policy was perhaps
understandable. By the 980s, however, the Sung had successfully retaken
south China. As the empire grew, the burden on the prisoners, their
dependents, and the accompanying guards became much heavier. A
report in 981 said that 60 to 70 percent of the convicts sent died en route.
This estimate may be exaggerated, but clearly the old system was caus-
ing problems.

Early reforms of the system

The need to reform the old system was obvious. The purposes of punish-
ment were to deter others, to provide the state with labor, and to
chastise the guilty while encouraging them to reform and giving them
the opportunity to do so. Dead convicts could hardly reform themselves
and become useful citizens. A system that exacted excessive penalties,
even if only incidentally, was contrary to the state's long-term interest.
Nor was the suffering of the accompanying dependents justifiable.
Finally, as the empire grew, the number of convicts to be sent to the
capital increased, adding to the security problem.
In 981 an official proposed reforms designed to resolve these problems.
First, he suggested that the prisoners be interrogated on arrival at
Kaifeng, to uncover any abuses by those bringing them. Second, he
asked that only the principal criminals be automatically sent to the
capital. Families and dependents should await a separate imperial order.
Those other than the principals involved in the crime were to have their
cases reviewed by the circuit intendants, and if sent to the capital, they
were to be exempt from wearing fetters. This official had himself ruled in
one case that the families need not go to the capital. The record says that
"from this time on, the number of criminals sent from Chiang-nan was
reduced by more than half."201 The policy of excusing dependents from
coming to Kaifeng with the prisoners was aimed at alleviating the
suffering of families that accompanied men who came to the capital only
to be sent on for registration elsewhere. It was not aimed at permanently
separating families. Orders in 1012 and 1013 show, however, that the

201 HCP 22.16b-17b. On the Sung inheriting a policy of sending men to the northwest,
see SS 201.5016.

437
 Law and order in Sung China
families and dependents of some criminals still accompanied them to the
capital. 202
Initially, the Sung continued the Five Dynasties policy of sending
most serious criminals for registration in the northwest by way of the
capital. This policy may have provided labor where it was needed, but it
also created security problems. Such men often escaped and joined forces
with non-Chinese tribesmen. When the Sung conquered the south, a
policy was inaugurated of sending some serious criminals for registration
to the newly acquired southern regions. 203 As with the capital, of course,
labor was still needed in border prefectures, and so some criminals
continued to be sent there. A decree of 1006 ordered officials to send to
the borders those recidivist robbers from Szechuan and Shensi who were
liable for penal servitude and whose crimes were serious. 204 No doubt
there were other similar orders, now lost.

Reforms under Jen-tsung

The new policy of the 980s eliminating the mandatory sending of all
those convicted of serious crimes to Kaifeng was targeted at ending
certain abuses and at reducing costs and suffering. Later attempts to
modify the system betray the desire to balance the maintenance of
control by the central authorities, the need for imperial security, the
capital's need for labor, and the costs of long-distance transport. Because
the Court did not want to lose any control over the disposition of its
forced labor units, the local officials were ordered to keep the central
government informed about the convicts that they dispatched. However,
some prefects apparently took advantage of the changed rules and sent
men away for registration without reporting this to the capital. Shortly
after the accession of Jen-tsung (r. 1023-63), he ordered local officials to
stop this practice. The Ministry of Justice was to receive reports and
review the sentences. Because of the growing problems of law and order
in the mid-eleventh century this order, and the other procedural safe-
guards added, seem not to have been effective.205

Reforms under Shen-tsung

Problems with the system persisted. In the late eleventh century, when
the Court attempted to deal with a host of accumulated state problems,
202 HCP 78.14a; 80.4a.
203 SS 201.5016. Although it is said that the policy of sending men to this region
(Ling-nan) was abrogated in 1008, this does not in fact seem to have happened.
204 HCP 63.6b.
205 SS 201.5016-17.

438
 Penal registration

the system of penal registration also came under review. Emperor Shen-
tsung (r. 1067-85) was concerned about the malfunctioning of the
system, in particular about the expenses of transporting prisoners and
the loss of life en route. The official Su Sung recommended that
we follow the ancient practice, establishing prisons. When it is time to punish
those liable for exile, shave their heads and chain their ankles. During the day
they can work where they are being held, and at night they can be kept in these
prisons. When they have served three years, they can be released. If there should
be an amnesty before they have completed their terms, do not free them. When
they have been released, send them home and keep them under surveillance. If
after three years, they have no violations, let them become as before.206
Shen-tsung did not fully follow Su Sung's suggestions, but his basic
position was restated by others. An official of the Secretariat Department
(Chung-shu) criticized the current system for not achieving its goals.
Under it, he pointed out:
Good people, if they happen to transgress the laws, are beaten on the back,
rejected by the multitude, and feel ashamed for life, whereas evil people, although
tried and beaten, are at peace with themselves in a few fortnights, forget about
their pain, and feel no shame, so that the punishment is not adequate to restrain
their evil. If with regard to those sentenced to penal servitude and exile whose
circumstances were not evil, you were to order that we return to the ancient law
of confined labor, reducing its length if an amnesty occurred, then the good
people could avoid injury. When their terms of labor were completed, they could
return to being commoners, having renewed themselves. Evil people would
remain government prisoners for years and so could not harm the good people.207
Eventually Shen-tsung reinstituted a system of keeping convicts in their
home jurisdictions for miscellaneous labor. According to a note in the
Ch'ang-pien:

In the beginning, Shen-tsung felt that the men transported for exile who were
separated from their homes sometimes became ill and died en route. Also, the
imperial army soldiers charged with convoying convicts lacked appropriate
training and had to suffer labor and expense. Therefore he imitated the ancient
penalty of exile, added a beating and tattooing, and had the convicts appended
to the armies (in their home areas) for heavy labor.208
This policy, enacted during the Yuan-feng period (1078-85), appar-
ently was used only in cases in which the crimes were not considered
heinous. 209 It was reversed in the mid-1080s after the rise of the anti-
206 HCP 209.7a.
207 HCP2i4.i7bff.
208 HCP 360.1a; see also WHTK 168.1460.
209 HCP 334.10a.

439
 Law and order in Sung China

reformers. Their basic position was stated in an order that "men liable
for penal registration be registered after transport in accord with the
previous system." 210 An edict of 1086 suggests that in practice, for
several years there was a mixed system in which some convicts were
moved and others were not:
Miscellaneous laborers who have been registered in the army from the prefec-
tures of the various circuits are to be registered in their current prefecture's
prison citadel commands. In the capital those who were originally liable for
registration in Kuang-nan may be divided for registration under the capital's
East and West Pottery Works; those liable for three thousand lid registration
may be registered in the cart encampment works. Those liable for two thousand
lid registration may be divided and registered under the Kuang-ku command.
From now on, all men guilty of crimes calling for penalties heavier than beating
with the heavy rod should be transported for registration according to their
original sentences.211
This mixed system apparently remained in use for a number of years.
Then, in 1095, w n e n the reformers again controlled the government,
they attempted to revive the Yiian-feng policy of limited movement of
convicts, only to abandon it again soon thereafter.212

Reforms under Hui-tsung

When the reform partisan Ts'ai Ching came to power in the early
twelfth century, he attempted to institute radical reforms of the system,
ordering that jails be established in various prefectures for men guilty of
robbery but spared death. They were to labor during the day and return
to jail at night. When their sentences of labor were completed, they
could enlist in the army. The system was tried twice, for a few years in
each case, before being finally abandoned in mo, 2 1 3 but the problems
associated with the alternative, transporting large numbers of prisoners,
had not really been resolved.

Southern Sung reforms

Many of these same arguments again surfaced in the 1180s when the
system again came under serious review. An initial criticism in 1184 was
followed by an order that the Ministry of Justice and the Court of
Judicial Review discuss the problem, but after three years they had
210 SS 201.5018; HCP 334.10b; WHTK 168.1460. See HCP 359.1 i a - b for details.
211 HCP 379.18a. See also 4O2.yb-8a for more details.
212 HCP 334.10b.
213 SS 201.5019-20.

440
 Penal registration

reached no conclusions. The problem came under more general debate,

during which one official commented, with regard to the idea of keeping
men in their home areas to labor:
If men are merely forced to labor and are not moved from their home areas,
some of them will take advantage of this to do evil. This certainly will not serve
to stop evildoing. If we completely follow the system of registration and are not
considerate about tattooing, then who, having been once shamed by the tattoo
on their face, will be willing to return to be enrolled [as a commoner]? Stalwart
people, facing exile to a far place and the threat of the tattooing knife, having
transgressed, would see no reason to reform. I have examined the regulations
[ko] of the Ministry of Justice from the YCian-feng period [1078-85], under
which all men to be registered were classified as "not to be transferred," "not to
be freed," "may be transferred," and "may be freed." The regulations from the
Cheng-ho period [1111-17] on registration also have the categories of "serious
circumstances," "relatively serious circumstances," "relatively light circum-
stances," and "light circumstances." Suppose we follow the [earlier] regulations,
with some revisions and additions. For example, if there are cases with "serious
circumstances," then in imitation of the older system we could tattoo the faces of
the convicts and apply the regulations on "not allowing transfers" and "not
freeing." The next level would be "relatively serious," under which we could
tattoo the temple [rather than the face] and employ the regulation on registra-
tion for ten years. When the circumstances are "relatively light," we could spare
them the tattooing and apply the regulation on "not tattooing and releasing to
return when labor is completed." If there are cases of "light circumstances,"
then command the convicts to do labor at a fixed place, and separately establish
a regulation governing the time limit for their release. 214

Although it is not clear how many of these proposals were enacted, they
do preserve a good picture of the detailed ways in which the Sung
authorities attempted to use registration and its variations to deal with
the problem of men guilty of similar crimes with widely varying circum-
stances. The policies of the period seem to have been generally like those
followed during the Northern Sung, with some modifications. During
some periods of the Southern Sung, it was common to register criminals
in the newly established military agricultural colonies, where the soldiers
were in theory part-time farmers who could be called on to fight in times
of emergency. 215 The Southern Sung authorities also registered small
numbers of convicts in units of the regular army rather than in separate
units of their own. These convicts then provided heavy labor for the
regular troops. 216

214 SS 201.5020.
215 See, for example, SHY, hsing-fa 4.50b, 51a, 52a-b, 52b, 53a, 53b.
216 SHY, hsing-fa 4.55a~56a, 6 i a - b .
441
 Law and order in Sung China

The resulting system of penal registration, if indeed it can be called a

system, was the most complex pattern of punishments ever employed in
traditional China (and probably anywhere else), but it did allow a
maximum of flexibility to the authorities. The major role it played
among Sung penalties can perhaps be suggested by noting the greater
number of crimes calling for penal registration as the penalty. In 1008-
16 there were 46; in 1041-48, 70; in 1070, more than 200; and in
1174-89 there were 570. 217

Conclusion
The Sung system of penal registration is a superb example of the crea-
tivity with which the Sung leaders approached political problems. They
did not theorize about them; rather, they molded the institutions and
practices inherited from earlier dynasties to fit their views of contem-
porary needs. The greatest problem was excessive power in the hands of
the military, especially but not exclusively the local commanders. The
Sung rulers built on the solution begun by their predecessors by bringing
the very best soldiers into the imperial armies concentrated near the
capital. This posed a new problem: What was to be done with all the less
capable soldiers not so selected? The answer was to find new uses for
them. The provincial armies in which they were registered were in a
sense welfare institutions providing a living for them. In this light the
provincial armies could also serve as homes for the homeless in times of
disaster. But idle hands are the devil's tools. Sung authorities avoided
the dangers of an army with nothing to do by using the provincial armies
as laborers. Thus the government at the same time kept potentially
troublesome men busy and relieved ordinary commoners of having to
perform many sorts of local tasks.
The Sung dynasty also inherited and altered the traditional policy of
registering some criminals in the army. Here again they did not create a
new policy but changed the meaning of the ancient one by greatly
increasing the number of men so used. The era preceding the Sung had
been marked by severe laws harshly enforced. It was especially noted for
the number of capital crimes on the books and the frequency with which
criminals were executed. The Sung did not try to abolish these laws
completely. Instead, the authorities initiated a policy of routinely sparing
a great many of those nominally subject to the death penalty and
lowering their actual penalties to penal registration. The men so spared
provided the labor to relieve commoners of such tasks as running the

217 SS 201.5020; HCP 214.17b-18a.

442
 Penal registration

Table 12.2. Remuneration of soldiers in provincial army units

Name of command Remuneration

Hsiung-chieh One bushel and five pecks of polished rice per month
Command # 6 480 coins per month
Each spring, two lengths of coarse raw silk plus 1700 coins
for tailoring
Each winter, two lengths of coarse raw silk, one-half length
of thin silk, twelve ounces of cotton [to be used in
padding winter clothing], plus 850 coins for tailoring
Wei-kuo One bushel and two pecks of polished rice per month
Command # 6 600 coins per month
Each spring, two lengths of coarse raw silk cloth, one-half
length of thin silk, plus 1844 coins for tailoring.
Each winter, two lengths of coarse raw silk cloth, one-half
length of thin silk, twelve ounces of cotton, plus 1844
coins for tailoring
Ch'ung-chieh One bushel and five pecks of unpolished rice per month
Command #31 Each spring, four lengths of coarse raw silk cloth, plus 1500
coins for tailoring
Each winter, two lengths of coarse raw silk, one-half length
of thin silk, twelve ounces of cotton, and 850 coins for
tailoring
Prison citadel One bushel and five pecks of unpolished rice per month
command Each spring, one length of coarse silk and 1150 coins for
tailoring
Each winter, one length of coarse silk and 1150 coins for
tailoring

Source: Ch'en Ch'i-ch'ing, Chia-ting ch'ik-ck'eng chih, 7204ff.

postal system. The desire of state authorities to achieve the goals of

punishment at an acceptable cost explains both the structure and the
evolution of the Sung system of penal registration. The goal of deterrence
was aided by the very public character of the criminals' punishment.
They served as laborers in the army camps located at most Sung urban
centers, where their ill-rewarded labors must have been regularly before
the eyes of the public.
This form of punishment also served the ends of social defense and
rehabilitation. The men had been separated from their old evil associates
as well as from their homes. They were put into a new controlled
environment, where they were given the opportunity to make a new life
for themselves. If they did their work well and did not cause any more
trouble, they might look forward at worst to a move up into a more
443
 Law and order in Sung China

desirable unit of the provincial armies, with better conditions and treat-
ment and more liberty. At best they might well be released to return
home, though they had the option of volunteering for service in the
provincial armies or, if young and strong, in the better-paid imperial
armies. Because of the deep-seated belief in the ability of men to control
their own behavior and to reform bad habits, the state tried to design a
system that would provide incentives for good behavior. We can gain
some idea of how the state set before the convicts the prospect of a
gradual improvement in their lot by comparing the remuneration of
convict and nonconvict units of the provincial army stationed in T'ai
Prefecture in Liang-che Circuit in 1223 (see Table 12.2).
The intuitive brilliance of this Sung system is striking, especially in
light of recent criminological research. Wilson and Herrnstein remark
that
there have been countless demonstrations of the effect of changed reinforce-
ments on the deviant behavior of individuals in schools, prisons, and clinics.
Virtually all these demonstrations have involved various forms of behavior
modification.. . . These methods involve a variety of techniques . . . all of which
have in common making rewards (and occasionally punishments) contingent on
displaying desirable (or undesirable) behavior and ignoring, for the most part,
the subjective state of the offender and his past history.
The authors caution, however, that the improvements in behavior have
not been shown to be lasting. Once the changed individual leaves the
institutional setting, most effects seem to fade away, and the people so
treated may well return to crime. 218 But removal from the institutional
setting was just what the Sung system did not do. Many ex-convicts
remained in an institutional (provincial army) setting that was very like
the one they had left behind. In addition, the example of their recent
(criminal unit) setting, and its drawbacks, was before their eyes each
day.
The use of the provincial armies as the locus of such rehabilitation was
particularly desirable from the state's point of view. By means of their
labor, the convicts were kept from the dangers of idleness and, at least as
important, gave back to the state a good portion of the cost of their
punishment. The Sung government could not have functioned without
them. The penal system provided these laborers at relatively low cost. It
was, in short, a brilliantly conceived apparatus.
The Sung authorities were not so fatuous as to think that all men were
likely to reform. They therefore built an elaborately flexible system of
penalties within which there were many different ways to make penalties
218 Wilson and Herrnstein, Crime and Human Nature, p. 381.
444
 Penal registration

more severe. Maximum-security facilities were in places where condi-

tions were harsh and dangerous, and the convict's life was likely to be
nasty, brutish, and short. Labor on Sramana Island, in the cold waters
off the north coast of Shantung, with bitter winters and the drudgery of
the salt works, was the next penalty below death in the Northern Sung;
the malarial wild country of Hainan Island filled the same role in the
Southern Sung. Yet even here the men sent for punishment were not
without hope. The state still publicly followed policies that could lead to
their reintegration, at least to some degree, in the larger society.
The Sung government balanced its labor needs and its desire to
promote reformation against its needs for security and social defense.
These various aims often conflicted, however. The compromises used to
deal with these conflicts varied with the time and place. The Sung
leaders argued about the proper balance, and the mixture of policies
changed as the personnel in the administration and the current situation
changed. But some underlying and defining characteristics remained the
same. Perhaps most important was the belief, which seems to have been
shared by all significant leaders, that most criminals were capable of
contributing to the overall social endeavor, if only as closely watched
and not wholly trusted soldiers. By designing a carefully graded hier-
archy of control policies that extended beyond the end of a convict's
sentence, the Sung leaders demonstrated their great common sense and
avoided the pitfalls of the modern policy of simply releasing many ex-
convicts into the larger society where they must find their own way. The
Sung system was anything but neat and well ordered, but if we judge
from the success of the Sung in avoiding dangerous unrest in local areas,
it was remarkably effective.

445
 13
The death penalty

Introduction
The death penalty as it was generally practiced in traditional China
was the ultimate demonstration of the ruler's power and the ultimate
humiliation of the powerless victim. Like executions in early Europe,
it was a profound and loud message to all concerned and therefore
required a public ceremonial. The victim was executed in public, and
sometimes his head would be publicly displayed. His personal suffering
at the same time shamed his family and his ancestors. One early Chinese
term for execution, still used in Sung texts, means, literally, "cast-
ing away in the marketplace" (ch'i-shih). The victim suffered the final
ostracism, being cast out from this world. The ceremony reinforced the
message about the relationship between criminal punishment and war
and reiterated the role of the criminal as a foreign attacker of society, for
the ax or the sword was in the hands of a soldier.
In antiquity, Chinese states used a wide variety of forms of execution,
frequently brutal and sometimes public. Burning, cutting in two at the
waist, dismembering, being torn apart by chariots, slicing, the list goes
on and on. The sources, however, give the impression that the most
gruesome penalties (with the notable exception of death by slicing) were
far more common before the creation of the unified empire in the third
century B.C. During the Ch'in and the Han there was a movement
toward a more limited repertoire of forms of execution, strangulation and
beheading being the most common.1 These two penalties continue to be
the most typical forms of execution for the rest of traditional Chinese
history.

1 For an excellent summary of materials drawn from the standard traditional sources, see
the work of the very great late Ch'ing jurist and scholar Shen Chia-pen, Li-tai hsing-chih
k'ao (Beijing: Chung-hua shu-chu, 1985), fu-chi i-wen tsun, vol. 1, chap. 17.

446
 The death penalty

Sung death penalties

When the Sung dynasty inherited from the T'ang the traditional system
of the Five Punishments (wu-hsing) - beating with the light rod, beating
with the heavy rod, penal servitude, exile, and death - it initially con-
tinued to use the two forms of death penalty that had been most com-
mon under preceding dynasties, strangling and beheading. These two
penalties embody fundamental Chinese views about the world, and their
difference was important not only to the suffering victim but also to
the onlookers who received their message. Beheading was the ultimate
mutilation. The Chinese believed that for the proper passage of the soul,
the body must be whole. To dismember the body, and particularly to
sever the head, was to extend the person's punishment beyond the
grave. 2 Beheading thus was much more to be feared than strangulation.
The Sung was not distinctive in its forms of execution, but it was
distinctive in that several of its nominal death penalties were often
commuted to penal registration, and it frequently spared other sorts of
criminals sentenced to death, using a variety of devices described in
Chapter 14.

Beheading and strangulation

T. T. Meadows described a group of beheadings in Canton in the mid-
nineteenth century. We have no reason to believe that the procedure
differed in its essentials from that practiced in the Sung, so in the
absence of contemporary Sung descriptions, we shall quote Meadows at
some length:
Having heard, on the evening of the 29th July, 1851, that thirty-four rebels or
bandits were to be executed on the following day between eight and ten o'clock, I
went to the ground at about half past eight with two English residents at Canton,
who had not previously witnessed any execution. . . . The criminals were brought
in, the greater number walking, but many carried in large baskets of bamboo
attached each to a pole and borne by two men. We observed that the strength of
the men so carried was altogether gone, either from excess of fear or the
treatment they had met with during their imprisonment and trial. They fell
powerless together as they were tumbled out on the spots where they were to die.
They were immediately raised up to a kneeling position and supported thus by
the man who stands behind each criminal. The following is the manner of
decapitation. There is no block, the criminal simply kneels with his face parallel
to the earth, thus leaving his neck exposed in a horizontal position. His hands,

2 It is presumably for some such reason that in the earliest literate Chinese dynasty, when
men were sacrificed, the bodies and the heads were often buried in separate places.
447
 Law and order in Sung China

Figure 13.1. Beheading. From Tsang Mao-hsun, Yuan ch'u hsiian.

crossed and bound behind his back, are grasped by the man behind, who, by
tilting them up, is enabled in some degree to keep the neck in the proper
position The executioner stands on the criminal's left. The sword ordinarily
employed is only about three feet long The executioners . . . are taken from
the ranks of the army The sabre is firmly held with both hands, the right
hand in the front, with the thumb projecting over and grasping the hilt. The
executioner, with his feet firmly planted some distance apart, holds the sabre for
an instant at the right angle to the neck about a foot above it in order to take aim
at a joint: then, with a sharp order to the criminal of "Don't move!" he raises it
straight before him as high as his head, and brings in rapidly down with the full
strength of both arms, giving additional force to the cut by dropping his body
perpendicularly to a sitting posture at the moment the sword touches the neck.

448
 The death penalty

Figure 13.2. Beheading. From Feng Meng-lung, Tien-shih chai hua-pao.

He never takes a second cut, and the head is seldom left attached even by a
portion of the skin, but is severed completely. 3 [See Figures 13.1 and 13.2.]
We also do not have contemporary descriptions of the other standard
Sung death penalty, strangulation. Again, we may assume that it was
carried out in the Sung as it was during the Ming and Ch'ing periods.
During strangulation the victim was tied to a post while two soldiers
slowly turned a rope twisted about his neck (see Figure 13.3). This
procedure was obviously a far more painful death than simple behead-
3 T. T. Meadows, "Description of an Execution at Canton," Journal of the Royal Asiatic
Society 15 (1856): 55.

449
 Law and order in Sung China

Figure 13.3. Strangulation. From Ch'uan-shan hsienpao-chia chang-ch'eng.

ing, but because it left the body whole, it was regarded as less serious.
According to traditional Chinese beliefs, the severing of the head from
the body destroyed the spiritual life of the victim. Preserving the body
whole, even at the cost of a very painful death, was preferable.

Irregular death penalties

When driven to it by especially frightful crimes, the Sung authorities
might set aside the standard penalties and inflict even more gruesome
ones. Occasionally the Sung employed the ancient punishment of cutting
in two at the waist. 4 In 1020 one group of criminals, convicted of crimes
pertaining to the state, were crucified. After three days their hands and
feet were cut off. Then they were finally executed. Several others con-
victed in this same incident were punished one step less harshly. They
were not crucified, but their hands and feet were cut off before they were
executed.5 The sources also mention men who, as a formal penalty, were
"beaten to death." 6
4 SHY, hsing-fa 4.2a (980); HCP 127.4b (1040), 157.4a (1045), SS 200.4998 (1075);
HNYL 183 (1127).
5 HCP 96.11 a.
6 HCP 21.9a, 99.9b, 2i3.i9b-2oa, 217.4b, 341.8b.

450
 The death penalty

Figure 13.4. Death by slicing. From Chin-shan hsien pao-chia chang-ch'eng.

The best known of these irregular punishments is the punishment

usually called death by slicing (ling-ch'ih). The earliest records of its use
can be found in the section on punishments in the standard history of the
state of Liao (907-1125) founded by the Khitan. 7 In Sung times, the
victim's limbs were first sliced off, and then his throat was cut8 (see
Figure 13.4). Although the first Chinese state to employ the penalty was
probably the Later Chou (951-60), 9 the Sung apparently borrowed it
directly from the Liao, since the Later Chou ended in 960, and the
earliest reference to death by slicing in Sung sources dates from i o n . In

7 See McKnight, "Death by Slicing"; T'o T'o, Liao shih (Taipei: San-min shu-chu, 1965),
113.2a, 3b, 5a, 6a; 114.3a. The term ling-ch'ih is probably of foreign origin. The attempts
to find Chinese roots for it are unconvincing.
8 WHTK 167.1447.
9 In the Treatise on Punishments of the Old History of the Five Dynasties (Chiu wu-tai shih)
there is a description of an execution using a short knife. Although the term ling-ch'ih is
not used, it seems clear that this was the punishment used. See Ch'iu Han-p'ing, Li-tai
hsing-fa chih (The legal treatises from successive dynasties) (Taipei: San-min shu-chu,
1965), P- 374-
451
 Law and order in Sung China

that year an official in Ching-tung Circuit named Yang Shou-chen asked

that the punishment be used for executing bandits. 10 And in 1014
another official asked that a murderous bandit be executed by being
"sliced" (luan-hai). The emperor in this latter case questioned the use of
such irregular forms of execution.11 "The Five Punishments are the
regular system. How could we do something so cruel?" Permission was
refused. Four years later the official Yang Shou-chen again asked to be
allowed to use death by slicing, when he was in Shensi in charge of
suppressing bandits. And again permission was refused.12
Under Chen-tsung's successor Jen-tsung (r. 1023-63) permission was
granted to use death by slicing in executing the devotees of a religious
sect that practiced human sacrifice.13 Various sources suggest that the
punishment was used during this early period to execute those guilty of
treason, of "plotting great sedition" (mou ta-ni) (violations of the temples,
tombs, or palaces of the reigning house), or of murdering senior family
members. By 1032 it was used as a means of executing bandits. In that
year, when a district magistrate reported having used death by slicing to
execute ten bandits, the court ordered that in the future such cases were
to be reported before the executions were carried out. 14 In 1034, how-
ever, its use was approved for executing certain bandits. The decree
stated: "If among the bandits captured, there are men who have already
been found guilty of homicide and who have repeatedly committed
banditry in which the circumstances of the crimes are truly evil, they
may be executed by slicing." 15 And in the Hsi-ning period (1068-74) it
was approved for some sorts of robbery and for sedition. 16 During this
era of bitter political struggle, the penalty is said to have been used with
some frequency in cases with political overtones, especially in cases in
which the sentences were set by imperial decree, with the culprits being
found guilty of "having acted in defiance of right." 17

10 SHY, ping n.6a; HCP 75.9b.

11 HCP 83.12a.
12 SHY, ping 11.7a; HCP 85.14b; see also WHTK 167.1447. This material is cited in
Shen Chia-pen, Shen Chi-i hsien-sheng i-shu, chia-pien, vol. 1, p. 47. See also the
comments of Niida Noboru, Chugoku hosei-shi kenkyu (Studies on Chinese legal history)
(Tokyo: University of Tokyo Press, 1964), keiho, p. 157.
13 WHTK 167.1447. This is the date usually cited as the beginning of the punishment
under the Sung. See Derk Bodde and Clarence Morris, Law in Imperial China
(Cambridge, MA: Harvard University Press, 1967), p. 94.
14 HCP 111.19b.
15 SHY, ping 11.15a.
16 SS 200.4998; HCP 267.10b-1 ia.
17 WHTK 167.1449.

452
 The death penalty

Procedural safeguards
Despite the increased use of death by slicing and the occasional use of
other especially cruel forms of execution, Sung writers frequently praise
the Sung rulers for their mercy toward prisoners accused of capital
crimes. The T'ang dynasty was regarded as a very lenient era, but the
era of the Five Dynasties that preceded the Sung was seen as a period of
great cruelty. The Southern Sung historian Hung Mai remarked that
"during the Five Dynasties the rulers treated killing men as an amuse-
ment, and men's lives were as grass." 18 Sung emperors are thus credited
with trying to correct some of the Five Dynasties' excesses.
Because execution could not be undone, there was special concern that
no errors be made in sentencing. The result in Sung times was a judicial
system that combined a legal commitment to openness with a lack of
openness in practice. The openness was intended to help avoid injus-
tices, as well as to defend officials from charges of abuse; the lack of
openness in practice resulted from the illegal but often tolerated actions
of the clerks and guards, who were not members of the elite.
The nominal commitment to openness included a trial system in
which at the end of the trial the accused was to have his case record read
to him and had to acknowledge its accuracy before his sentence could be
given. This practice, which is mentioned in the classical Book of Historical
Documents, continues to be a characteristic of Chinese justice today.
According to Sung law, the record was supposed to be read at a plenary
meeting of the headquarters officials and the clerks involved in the case.
In practice, however, clerks are accused of "plucking out the paper,
carelessly reading it out, abbreviating its wording, pronouncing words so
they are incomprehensible, and having done all this hurriedly, stopping
and calling on the prisoner to sign his name." 19
A similar openness was also supposed to characterize the actual execu-
tion itself. There are several brief descriptions of some of the steps in
carrying out capital sentences. Various officials, including both those
who had been involved in the case and other colleagues, were to par-
ticipate in interrogating the condemned person immediately before his
execution. He was to be questioned about his home, his age, and his
name, to be absolutely certain that these did not differ from those in the
records. He was to be given wine and food, allowed to speak his final
words to the spirits, and to visit with his relatives. Both his relatives and

18 Hung Mai, Jung-chai sui-pi, fifth collect., p. 65.

19 WHTK 167.1455.

453
 Law and order in Sung China

the condemned convict were to have his criminal record read to them.
The man to be executed was not to have his mouth or his ears stopped
up, nor was his face to be covered. This was to permit him to question
the record as it was read. For the same reason the record was not to be
read in a hurried manner. 20
Nonetheless, the process was sometimes so rushed that the wrong
prisoners were punished. 21 And at the time of actual execution, accord-
ing to the eleventh-century official Lin Tan, condemned prisoners were
frequently abused:
When people are to be executed publicly, if there are those among them who
have been bound and whipped, they are carried in baskets [to the place of
execution] and afterward are freed to walk, with their mouths and ears plugged.
Also, paper money is used to cover their heads thickly. The prison runners from
the army command number about a hundred. They crowd before and behind the
prisoner, enclosing him. They use iron hammers to strike the fetters. They call
out, drumming up a hubbub and not stopping for a moment. Even if he wants to
call out that there has been judicial abuse, the criminal has no way of doing so.22
Although there were exceptions, capital sentences generally took place
in the jurisdiction where the case had been tried. 23 The execution itself
was to be carried out between one and five o'clock in the afternoon.
Executions were supposed to be carried out only at the execution
ground. No added injuries were to be inflicted on the body.
This whole process of public execution was so frightful and demeaning
that prisoners often sought ways to avoid it. This was by no means a
new problem. The T'ang Code (in a passage copied into the early Sung
Code) reports:
When the case reports of criminals convicted of capital crimes have been com-
pletely investigated and the prisoner's relatives deliberately kill him, or hire
someone else to kill him, both the relatives and the actual killer are to be tried for
the original death penalty handed down for the prisoner, with their penalties
reduced by two degrees [to penal servitude].
The statute goes on to list factors that might affect the severity of
the sentence against those who arrange for the death of a condemned
prisoner, but in all cases it is clear that such homicides would be
prosecuted. The determination of the Sung government to discourage
such actions is reflected in a decree of 1046 that promised huge rewards

20 TFSL 500; WHTK 167.1454.

21 WHTK 167.1454.
22 HCP 376.5b-6b. Dr. Charles Benn has suggested that the practice of gagging prisoners
on the way to execution may have originated during the reign of Empress Wu.
23 See HCP 9.7b for an example of an exception.

454
 The death penalty
to those identifying any clerks who provided poison or otherwise killed
those waiting for execution. 24
One night after the execution, the relatives or friends of the victim
were to be allowed to claim the body for burial. If there were no relatives
or friends to claim the body, it was turned over to the authorities for
burial. Because the days on which men were executed were inauspicious,
officials in the jurisdictions involved were forbidden to listen to music,
and the emperor continued a traditional practice of abstaining from
music and eating simple food.25

Timing of executions
During the Sung the consideration of capital cases and the execution of
criminals were governed by policies regarding the appropriate timing.
The ideological roots of these rules are very old. In the Chinese written
record they can be traced back to the "Monthly Ordinances" section of
the Book of Rites, according to which capital punishments were to be
carried out in the late autumn, to accord with the nature of the season of
death: "In the second month of Autumn .. . orders are given to the
proper officers to revise with strict accuracy [the laws about] the various
punishments. Beheading and [the other] capital executions must be
according to [the crimes] without excess or defect. Excess or defect out of
such proportion will bring on itself the judgement [of Heaven]." 26
During some earlier dynasties this had led to a pattern of holding
men sentenced to die until the arrival of autumn, when they would be
executed. This policy was meant to fit the actions of the emperor with
the characteristics of the season of the year. This same policy was
suggested to the Sung emperor but was rejected as a general policy
because it would have resulted in suffering for those men being held for
extended periods in jail. 27 The Sung emperors temporized, sometimes
delaying executions. The alternative policy of not delaying executions
was probably more generally practiced during the Sung and was designed
to show the emperor as a benevolent ruler seeking to avoid undue
suffering, even that of convicted and condemned persons. The Northern
Sung official Su Sung [1020-1101] summarized both traditions:

Our state uses humane concern in ruling the world. The ruling ancestors were
extremely careful in their use of penalties.... From the Court above to the
24 HCP 158.13a.
25 SHT 30.494; TFSL 500.
26 James Legge, trans., Li Chi: Book of Rites (New Hyde Park, NY: University Books,
1967), vol. 1, p. 288.
27 SS 1994974-
455
 Law and order in Sung China

prefectures and districts below all honor the decrees and do not violate them.
Even if it is merely a matter of a beating with the heavy or light rod, if it is not in
accordance with the law, it is not done. When there has already been a convic-
tion, additional investigations are carried out by the officials. . . . It is only the
timing of death penalties that has not yet been settled. . . . The [Classic history
entitled the] Tso chuan says, "Reward them in the spring and summer. Punish
them in the fall and winter." This was the timing of the Three Dynasties [of
Hsia, traditionally dated 2056-1752, Shang, traditionally 1751-1122, and Chou,
1122—256]. In spring and summer, executions were never carried out. The
histories record that during the Ch'in [221-207], men were executed in all
seasons, and Wang Mang [r. 9—25] killed men in the summer. These really were
evil governments. Under the Han system, condemned criminals were regularly
executed in the three winter months. They were not executed in spring and
summer [literally, "when yang flourished"].. . . From the Eastern Han [27-220]
this system has sometimes been practiced and sometimes not. In recent times it
has not been used. . . . Generally this has been because people were concerned
about convicts being held in jail for long periods of time. 28

The emperors also wanted to fulfill their filial obligations to their

ancestors and to show their concern by observing days of mourning. The
early Sung Code preserves a T'ang dynasty decree dating from 833
that says that on days commemorating the deaths of emperors and
empresses, drinking wine or listening to music was forbidden. Only the
most minor penalties could be carried out.29
Like other dynasties, the Sung also enacted various policies that pre-
vented the punishment of prisoners on inauspicious days. Failure to
observe the proper timing could injure the functioning of the natural
order. Some taboos it inherited from earlier times, but others were
introduced or reintroduced by Sung officials. In 991 an official asked
that there be no capital sentencing on the first day of the new year or on
other ritually important days such as the eighth day of each month and
the first and fifteenth days of the first, seventh, and tenth months.30
Another early important document on this policy was submitted in 1002.
An official cited the Classical precedents for avoiding executions during
certain times of the year. He noted that during the early and middle
T'ang these rules were followed but that they had fallen into disuse
during the late T'ang and Five Dynasties. He cited an edict from 942
during the Later Chin, retained in the early Sung Code, that ordered that
executions were not to be carried out on overcast days, in rain or snow,

28 Su Sung, Su Wei wen-kung chi, i8.8a-b. Su Kung himself favored a return to the policy
of not executing people (except those guilty of certain terrible crimes like rebellion)
during the spring and summer.
29 SHT 26.406.
30 WHTK 166.1444.

456
 The death penalty

or on the days called "Beginning of Spring" (li-ch'un, i.e., the fourth or

fifth of the second lunar month, or about the beginning of March),
"Beginning of Summer" {li-hsia* i.e., the sixth or seventh day of the fifth
lunar month or about the beginning of June), the days of great imperial
sacrifices {ta-chi chi), the day of mid-winter (cheng tung), or the import-
ant Buddhist holiday called the Cold Food Festival. 31 He argued that
because the Later Chin had been a time of great troubles, it had let the
two days of Beginning of Spring and Beginning of Summer represent the
whole of spring and summer. The Sung had reestablished peace and
order and should extend this period. Because the year was composed
of twenty-four periods of fifteen days each, the authorities could re-
introduce some measure of conformity to the seasons by avoiding execu-
tions for fifteen days at the time of Beginning of Spring and Beginning of
Summer. Some cases, such as those involving the Ten Abominations,
would not be covered by this rule, nor would cases in the border areas
under military law. 32 Whether this suggestion was fully accepted is not
clear. We do know that five years later the authorities ruled that no docu-
ments concerning capital punishments were to be sent up on the day
before either a major sacrifice (ta-ts'u) or a major anniversary (ta-chib),
though the processing of minor criminal cases was not to be delayed. 33
During the reign of Chen-tsung, in 1012, an order forbade carrying
out punishments for seven days during the holiday called "Heavenly
Felicity" (t'ien-ch'ing), commemorating the dates of the receiving of the
"Heavenly Books," which fell on the third day of the first month, and for
one day on the sixth day of the sixth month, another time when the
Heavenly Books had been received.34 Another document from later in
that same year indicates that there were similar suspensions of penalties
on the holidays marking imperial birthdays. 35
Again, in 1017 the throne approved a request by the Court of Rites
and Ceremonies:
As to the suspension of punishments on the various holidays, we now ask that
the edicts of the past be classified as precedents and promulgated, under which
all death penalties are to be temporarily suspended for seven days before and
three days after the holidays of celebrating the receiving of the Heavenly Books
on the third day of the first month [t'ien-ch'ing], the imperial birthday [hsien-t'ien]
that falls on the first day of the seventh month, the imperial birthday that falls on

31 SHT 30.495. Officials who carried out executions at some inappropriate times were
liable for fifty blows of the light rod (SHT 9.150).
32 HCP 53.9a-10b.
33 HCP 67.14b.
34 HCP 77.9b.
35 HCP 79.8b.

457
 Law and order in Sung China

the twenty-fourth day of the tenth month [ch'u-sheng], the day called "Receiving
the Mandate" (ch'eng-t'ien), for one day each on the sixth day of the sixth month
[t'ien-k'uang] when the Heavenly Books were received, and the holiday called
"Heavenly Omens" [t'ien-hsiang]. For sentences of penal servitude, exile, or less,
four days before the holidays arrange the place [for carrying out the punish-
ments], temporarily transfer the prisoners to these other locations, and send
officials to execute the judgments three days before the holidays. Those criminals
who have passed through the day of the holiday may be punished on the
following day. On the holidays, for crimes of the heavy rod or less, when the
circumstances are not serious, free the culprits. 36
T h e general concern of the authorities not to carry out death sentences
during certain seasons is also demonstrated in an edict of 1020:
From now on, in regard to those in the empire who have been sentenced to death
because they have been guilty of the Ten Abominations, banditry with homicide,
premeditated and deliberate homicide, or homicide during an affray, arson,
armed robbery, official judicial abuses, rapacity by officials, the making of false
official seals, heresy, the transmitting of witchcraft, the making of poison, or
deserters from the imperial army [chin-chun] who have become bandits after
deserting, if the twelfth month arrives, they are to be held temporarily in the
jurisdiction of judgment. After the holiday of the third day of the first month
[t'ien-ch'ing] is over, they may be executed. Those guilty of other capital crimes
should not be sentenced and punished during the twelfth month or during spring
and summer. They should be held, and a memorial should be submitted asking
for a decision. Also, it was decreed, in the past those sentenced to death have
been treated in accordance with an edict of beheading. From now on, except for
the crime of contumacy [o-ni] in four degrees, which should follow the statutes
concerning punishment, in cases of other crimes calling for beheading during the
spring and summer those convicted should simply be beaten and held. When the
"autumn equinox" [ch'iu-fen, i.e., the ninth day of the eighth month or about
September 23] has passed, then they may be executed.37
This rule was relaxed somewhat some months later because of an
official complaint that its strict enforcement was resulting in men's being
held for long periods in jail. In "serious cases," the authorities could
continue to carry out the penalties. 38 Two years later it was further
ordered that certain officials in the capital were not to give death sen-
tences on a certain set of holidays, including the days beginning each
season. 39 Again, it is not clear how closely these rules were followed,

36 HCP 90.16b-17a. I have not yet been able to learn when the holiday called t'ien-hsiang
fell. The term itself refers to the mineral drug realgar and was perhaps connected with
the practice of drinking a special drink containing realgar at the time of the Dragon
Boat races to ensure good health for the rest of the year.
37 HCP 95.140-153.
38 HCP 96.22b.
39 HCP99.i4a-b.

458
 The death penalty

though their observance was not always strict. In 1024 t n e r u l e s w e r c

again changed. Previously it had been forbidden to memorialize capital
cases for five days before and after the Five Holidays [wu-chieh] and the
third day of the first month. This was now cut to two days. Other
sentences were not be to carried out on the days themselves. 40 Later that
same year there was an order that in Kaifeng, the execution of penalties
was to be suspended for three days on from the first day of the first
month and the winter solstice, and one day on the fifth day of the fifth
month. 41
In 1028 the chief minister commented that under Chen-tsung (r.
997-1022), there supposedly had been a system that forbade punish-
ments and the hearing of cases for three days on either side of the dates
of imperial mourning, and forbade officials to listen to music for five
days on either side, but the system had gradually been abandoned. As a
result of the minister's remarks, there was a new decree that punish-
ments were not to be considered for two days before and after the
holiday, and for three days music was to be banned. In that same year
there was another order that there should be no executions for two days
before and after the holidays called "Origin of Heaven" (ch'ien-yuan) and
"Extended Tranquillity" (chang-ning). Other penalties could be inflicted
except on the actual holidays. 42 In 1041 a decree indicated that execu-
tions were suspended only on the third day of the first month, on the day
of "Heavenly Foundation" (t'ien-ch'i), the first day of the seventh month
(hsien-t'ien), and the imperial birthday on the twenty-fourth day of the
tenth month (ch'u-sheng). They were also suspended on the day of the
"Origin of Heaven" (ch'ien-yuan), as well as the day before and after this
holiday. 43
Such rules continued in force throughout the dynasty. The early
thirteenth-century code, Ch'ing-yuan t3iao-fa shih-lei, contains an undated
ordinance that states:
Sentences of death are not to be carried out on imperial birthdays, on the third
day of the first month [t'ien-ch'ing], during the holiday marking the first four days
of the New Year [k'ai-chi], the first day of the seventh month [hsien-t'ien], or the
twenty-fourth day of the tenth month [ch'u-sheng]. [In each of these cases, this
means the three days before the holiday as well as the day after.] Nor are such
penalties to be executed on the holidays of the sixth day, sixth month, com-
memorating the receiving of the "Heavenly Letters" [t'ien-shih], on the "Heavenly

40 HCP io2.i7b-i8a.
41 HCP 102.19b.
42 HCP 106.2b, 6a.
43 HCP 130.6a.

459
 Law and order in Sung China

Foundation" [t'ien-ch'i], on ting-mao days [fourth days in the sexegenary cycle of

dating], on wei-tzu days [twenty-fifth days in the sexegenary cycle], on the first
day of the New Year [yuan-cheng], the Cold Food Festival [the sixteenth day of
the second month, or about April 5], the winter solstice [the eleventh day of the
twelfth month, or about December 22], the Beginning of Spring [the fourth or
fifth day of the second lunar month or about the beginning of March], the
Beginning of Summer [li-ksia, the sixth or seventh of the fifth lunar month, or
about the beginning of June], the holiday of the star Jupiter [t'ai-sui], the first
and fifteenth days of the months [san-yuan], the holiday of the Spring Festival [ta-
t'ung], and the days commemorating imperial deaths. [For each of these holidays
this means no executions on the day itself.] Nor are executions to take place on
days when it is raining or snowing and not clear. 44
This passage suggests that there were more than fifty days a year
when executions were forbidden, without considering days excluded
because of inclement weather.
The concern about appropriate timing demanded that the penalty be
fitting, not necessarily that it be delayed. Fittingness could also cut the
other way. When the timing was propitious for executions, they could
sometimes be carried out with fewer procedural safeguards than in
normal times. An order of 1008 says that after the first day of the ninth
month - after the beginning of a period during which capital punish-
ments might be carried out without disturbing the proper order of things
- the Judicial Control Office (Shen-hsing yuan) and the authorities of
the capital prefecture were not to memorialize cases involving death
penalties. The Secretariat (Chung-shu) could itself give final rulings. 45

Reviews of death sentences

The final act in this drama, the execution of the criminal, was the last
step in what could be a complicated and drawn-out process. From the
beginning of the Sung the general pattern in capaital cases was for
the district authorities to prepare a preliminary finding, including a
recommended sentence. The prefectural officials reviewed this document
to see whether the crime documents were clear and the sentence appro-
priate, and then they reported their recommendation to the circuit level.
The circuit authorities reviewed the documents and either approved the
execution (if it seemed fitting) or forwarded the case to the capital for
further review.46
From the early Sung period onward, some cases were being memo-
rialized to the central authorities. A law on the review of capital sen-
44 TFSL 501. For the dates of hsien-t'ien and ch'u-sheng, see HCP 79.8b.
45 HCP 69.12b.
46 SHY, hsing-fa 4.57a-b.

460
 The death penalty

tences was enacted in 962, two years after the founding of the dynasty.
The commentary says that initially the emperor ordered the memorializ-
ing of all death sentences to the capital, but then (presumably in the law
of 962) ordered the prefectural officials to make at least the initial
determination. 47 By an order of 970, after a death sentence was passed,
the case was to be reviewed and a report written on it in red ink giving
such information as the law under which the case had been judged and
the officials' names. 48
Although a policy of reviewing cases was revived, it does not appear
that all such cases had to be reviewed. Only if the circumstances
demanded compassion or the law was not clear were reviews mandatory.
When Emperor Chen-tsung (r. 997-1022) came to the throne, he exam-
ined the reports on capital cases and was disturbed by the many death
sentences he found. He asked his officials to explain why the T'ang sys-
tem of three automatic reviews of death sentences had been abandoned.
In the ensuing discussion, one official complained that the number of
capital cases in the Sung was a hundred times that of the T'ang. He then
went on to say that cases in the Sung capital jurisdiction were reviewed.
However, when local officials had cases for which the facts were not
wholly clear or there appeared to be mitigating circumstances, and they
memorialized them for a ruling, they were often upbraided and punished
by the central authorities for having acted improperly. An order of 987
had prescribed penalties for local officials who memorialized cases that
they should have decided themselves. 49 Apparently the officials were
worried about violating this rule. The official in this instance asked that
all capital cases be reviewed. His stand was disputed by another official,
who argued that "if all the capital cases of the empire are sent up for
review, those who may be liable for death sentences will fill up the
prisons and for long periods of time will not be sentenced." As a result,
an imperial edict was issued expressing the emperor's concern for the
multitude of people who became enmeshed in the law and stating that in
the future, cases were to be sent up for review "when the circumstances
are deserving of mercy or the law is doubtful." 50 Clearly, the edict of
1025 did not visualize the memorializing of all cases.
The general Sung rule for such cases thus seems to have been that a
case was to be memorialized if the case involved doubtful points of law
or mitigating circumstances, but not otherwise. 51 Generally in these

47 WHTK 166.1443.
48 WHTK 170.1474.
49 STCLC 742.
50 WHTK 170.1474-75; SS 199-4974-75-
51 STCLC 739; see also HCP 104.7b.

461
 Law and order in Sung China

cases the Court of Judicial Review (Ta-li ssu) rendered a sentence that
was reviewed by the Ministry of Justice (Hsing-pu), 52 though at times
the two units acted together.53 Cases without mitigating circumstances,
for which the law was clear, were not supposed to be reviewed beyond
the circuit level.54
There were also special orders excluding certain types of cases from
the need for review. An order of 1002 on the beheading of deserters
who committed banditry suggests that the punishment was inflicted by
the jurisdiction involved without prior review,55 and a decree of 1043
gave the fiscal intendant of Kuang-nan the authority to behead "men
registered in the armies who are guilty of repeatedly committing crimes
with serious circumstances." The authorities were thereafter to send in a
report. 56 A similar system was used in the late 1080s in those border
areas where banditry was rife,57 and materials describing the earlier part
of the reign of Shen-tsung suggest that in certain sensitive border areas,
the military intendants had considerable leeway in executing criminals
(and in imposing other penalties). 58 Summary justice was also re-
introduced as a temporary measure in 1127 at the height of the Chin
invasions.59 However, there also were orders requiring the memorializing
of certain types of capital crimes, for instance, crimes of theft in which
the booty was large enough to call for the criminal's execution.

Numbers of reviews
Evidence giving the numbers of cases sheds some, unfortunately con-
tradictory, light on this review procedure. A report dated 1033 says that
the desk at the Ministry of Justice that reviewed capital cases processed
about 200 per month, or 2400 per year. 60 This rate of review is also
mentioned in regard to an incident recorded in the book of miscellaneous
notes, Sheng-shui yen-Van lu, of Wang Pi-chih. Wang reported that the
Court of Judicial Review said that although it ordinarily dealt with
several hundred cases a month, for a period of more than a month no
cases at all had been sent up. 61

52 For the earliest indication of this system, see HCP 5.30-43 (964).
53 HCP 12307.10b (957).
54 WHTK 170.1477.
55 HCP 53.9a.
56 HCP 142.10a. For other examples, see HCP 72.16a; WHTK 166.1445.
57 HCP 376.7a.
58 SS 1994979.
59 HNYL 234.
60 See HCP 113.13a, SS 1994976.
61 Wang Pi-chih, Sheng-shuiyen-t'an lu, TSCC ed., 1.2.

462
 The death penalty
However, evidence from the late eleventh century indicates that the
policy had changed. Only a small fraction of cases were being memo-
rialized. An official reported in 1086 that between 11/23/1084 an<^
11/23/1085, 264 cases calling for capital punishment had been memo-
rialized, at a rate of 22 per month. Between the latter date and the end of
the second month of 1086, in less than one hundred days, 154 cases had
been memorialized, slightly more than twice the rate of the preceding
year. 62 Even at this higher rate, the central authorities would have seen
only about 500 cases in a year. And yet in the twelve months between
December 1084 and December 1085, during which 264 cases were
memorialized for review, 2066 people were sentenced to death. In 1086,
during which at the rate current early in the year, there might have been
500 cases to review, some 5787 people were sentenced to die. 63 It seems
clear that in the late eleventh century only 7 to 9 percent of the cases
were being reviewed at the capital.

Records of executions
The state maintained careful records of those executed. These reports
may have been derived from reports compiled as a result of an imperial
order of 974 that the prefectures were to report all death penalties. 64 The
supervising authorities in the circuits were charged with maintaining
records of the number of men sentenced to death in their subordinate
prefectures. A decree of 1092 says that the judicial intendants were to
submit to the Ministry of Justice quarterly reports of the number of
capital cases reviewed, 65 and from 1091 there is a report indicating that
reports were submitted in spring and autumn on the number and types
of executions, which would be compared with the record of the preceding
year. 66 Under the code of 1202 at the end of each year, the intendants'
offices were to submit reports on the number of capital sentences handed
down in their circuits, classified by the total number actually sentenced
to die - those to die by slicing, by beheading, by strangulation - and
those liable for death who had been spared. This enumeration was also
to show the number of capital cases in which the men were not convicted

62 HCP 370.8b. SS 201.5012 and WHTK 170.1477 give the figure as 264 cases.
63 HCP 363.19a, 393.29a.
64 SS 3.50.
65 HCP 349.1a. HCP 479.2b indicates that there was also an annual year-end report by
the intendants in the circuits and the Censorate in the capital, to the Ministry of
Justice, which then would report to the Department of Ministries the jurisdictions with
the largest numbers.
66 HCP 465.15b.

463
 Law and order in Sung China

in the jurisdiction because their cases had been memorialized, and the
number executed on special imperial order for crimes not ordinarily
capital. 67 The form reads:

Specification [jfo"Aa] on Sentences in Criminal Cases:

Numbers of capital sentences:
The judicial intendant's office in such-and-such a circuit now records that the
prefectures in this circuit in such-and-such a year have carried out so many
capital sentences, classified as follows:
Memorials sent — so many; death sentences - so many; death by slicing — so
many; beheading - so many; strangulation - so many; spared death — so
many; men not executed by this jurisdiction — so many (Note: These men
that the jurisdiction was not competent to execute, as well as those spared,
need not be listed again. The others should be enumerated in accordance
with the list of those sentenced to execution after the cases have been
memorialized.); executed on special orders for crimes not ordinarily capital
- so many; memorialized cases - so many; sentences carried out in the
jurisdiction - so many. (Note: Write for each person a report, giving the
crime and the circumstances, including what jurisdiction handed down
the sentence and whether or not the matter had already been charged.).
The preceding itemized report as given above has been respectfully pre-
pared and submitted.
Department of Ministries, Ministry of Justice form.
Such-and-such a year, month, and day. (Note: the various prefectures are to
follow this format in reporting to the offices of the judicial intendant.) 68

Numbers of executions
Sung sources generally do not tell us much about the number of crim-
inals in the legal system. Fortunately, there are some figures concerning
capital crimes, presumably stemming from a record system like the one
just described. Although Li Yen, writing in 1060, said that before that
time there had not been a regular system of annually reporting the
number sentenced to die, in fact such figures were collected at least
sporadically. The earliest information is a brief passage that says that
between 969 and 975 some 4108 men were spared from death sen-
tences. 69 In 990 a report states that in that year 800 were sentenced; in
1001, around 3200; and in 1025, 2 43^ persons were "sentenced to death"
(tuan ta-pi).70 Although these are the only early records, beginning in

67 TFSL501.
68 TFSL501.
69 HCP 12307.7b.
70 HCP 104.7b.

464
 The death penalty

1060 there is a long sequence of such reports, covering most of the years
until the end of the eleventh century and based on the system of annual
reports of the number of death sentences established in 1060 after a
request by Li Yen.71 As Table I3«i(a) shows, there was considerable,
and unexplained, variation in the number of persons.
There is also a series of reports from the early Southern Sung, found in
the annalistic history Chien-yen i-lai hsi-nien yao-lu. The extraordinary
decrease in the number of men sentenced to death suggests that there
had been a radical alteration either in punishing serious criminals or in
reporting punishments during this era. See Table i3.i(b).
How are these discrepancies to be explained? There had not been
sufficient significant change in the laws during the brief interval between
1099 and 1129 to cause this extraordinary decrease. The empire had
certainly shrunk in size and population after the loss of the area north of
the Huai River to the Chin invaders, but this loss was not enough to
explain the huge drop in the number of capital sentences. Rather, it
seems to be that the figures cover two separate categories. Although the
Northern Sung figures are for total numbers of capital cases, most such
cases were not memorialized but, rather, were finally decided on the
circuit level. The Southern Sung figures probably excluded men dealt
with on the circuit level and included only those actually executed after
review at the capital. This possibility is suggested by a comment of the
Northern Sung official Fan Ch'un-jen, who in 1086 noted that in the
previous year 264 capital cases had been sent up for review, but in
the end only 25 persons had actually had their sentences confirmed and
been executed.72 This figure is close to the figures given for the Southern
Sung. Thus, in the absence of dramatic changes in judicial procedure, it
seems plausible that the figures are for these two different categories of
persons. 73

71 SS 199.4977-
72 SS 210.5012.
73 However, there is one passage that raises the possibility that in the Southern Sung all
cases were memorialized, with only a small fraction of the people involved eventually
being executed so that in fact the number executed during the Southern Sung may have
been lower than during the Northern Sung. WHTK 167.1455 says only 181 cases were
tuan out of 1811 capital cases in 1201.
These Southern Sung figures are remarkably close to figures given for the T'ang. In a
memorial dated 1060 an official noted that in the year 630 there had been only
twenty-nine capital sentences, and only fifty-eight in 738; see HCP 104.7b. During most
of the T'ang period, all capital cases were supposed to be memorialized to the emperor
for review. It is not clear how well this was enforced during the T'ang, but the figures
do suggest that the T'ang figures are for the number of memorialized cases ending in
execution. This policy of memorializing capital cases was not followed in the succeeding
Five Dynasties period.

465
 Law and order in Sung China

T a b l e 13.1. Numbers of persons sentenced to death

(a)
Date Number of persons condemned Source

968-73 4108 Hsu t'ung-tien 117.2a

996 800 WHTK 166.1445
1001 3200 c. HCP 48.15b
1023 2700 Yin Shu, Ho-nan 2.5b
1024 2200 Yin Shu, Ho-nan 2.5b
1025 2436 HCP 104.7b
1060 2560 HCP 191.9a-b
1062 1683 HCP 197.16a
1063 1606 HCP 198.18a
1064 2493 HCP 203.17b
1065 1736 HCP 206.27a
1066 1832 HCP 208.20b
1070 3523 HCP 218.22a
1071 3699 HCP 228.20a
1072 3792 HCP 241.13b
1074 3509 HCP 258.17b
1076 758 HCP 279.24b
1077 389 HCP 286.7a
1078 1104 HCP 295.13b
1080 1212 HCP 310.16b
1082 2815 HCP 331.22a
1083 2671 HCP 341.19b
1084 2365 HCP 350.14a
1085 2066 HCP 363.19a
1086 5787 HCP 393.29a
1087 5573 HCP 407.21a
1088 2915 HCP 420.16a
1089 5405 HCP 436.22a
1090 4261 HCP 453.16a
1091 4801 HCP 468.27b
1092 4191 HCP 479.12b
1098 2043 HCP 504.21a
1099 1395 HCP 519.10a
(b)
Date Number of persons Place in HNYL

1129 324 594

1142 24 SS 30.558
1144 26 2460
1145 91 2497
1147 35 2548
1148 32 2574
1149 31 2602
1150 25 2625
466
 The death penalty

Table 13.1. (cont.)

Date Number of persons Place in HNYL

1151 22 2650
1152 16 2676
1153 25 2708
1154 19 2739
1155 21 2799
1156 30 2900
1157 19 2950
1158 47 2997
1160 31 31,326
1162 41 SS 33.621
1202 181 SS 200.4995; W H T K 167.1455
1260 33 SS 46.892
1261 35 SS 46.896

Note and sources: In 1001, 800 cases had been reported in the first three months, so
we can project about 3200 cases. WHTK gives 1832 for the year 1065 (WHTK
167.1448). There may well be something wrong with the figures for the years
1076 to 1078. Po chaipien 10.3b-4a says that in the three years from Hsi-ning 8
through 10 there were 5182 capital sentences. It also says that in this period
there were 2592 officials convicted of crimes calling for tattooed registration down
to commutation. Note that the WHTK 166.1444 says that between 969 and 975
some 4108 men were spared death. WHTK 167.1452 and 1455 seem to imply
that tuan means execution.

To put the figures in perspective, we should remember that in tradi-

tional China the initial and the final sentence for a crime might not be
the same. Even if the initial sentence were reviewed and upheld at higher
levels in the judiciary, this did not necessarily mean that the person
convicted actually suffered the sentence. For a variety of reasons, the
sentence might be altered or the punishment reduced. In the case of
capital sentences, the final outcome was affected, among other things, by
the legal process of conviction and the review of cases. In the figures
given for 1086, only io percent of those whose cases were memorialized
were in the end executed (though we should note that Fan Ch'un-jen
submitted his memorial in order to complain that under a new set of
rules used in 1086, far more cases were being sent up, and a much
greater percentage - 6o percent - was being executed). 74 Another report,

74 SS 210.5012. Note that Fan Ch'un-jen memorialized in order to complain about the
change in procedures, saying that during the current year more cases had been
memorialized and that the rate of confirmation of executions was much higher, 40
percent versus the 10 percent of the preceding year.

467
 Law and order in Sung China

from the early twelfth century, says that in 1116 and 1117 there had
been 132 capital cases in the eleven prefectures making up Huai-nan
East Circuit. Of those tried in these cases, only 12 persons (9 percent)
had eventually been executed. 75 Another report that sheds more light on
this situation dates from the beginning of the twelfth century. It says
that in one year the empire had 1811 capital cases but only 181 persons
were actually convicted. 76
Insofar as the smaller numbers for executions in the Southern Sung is
not an artifact of the different processes used in collecting the figures, the
reduction might be explained by some policies introduced during the
military confrontation of the late 1120s and early 1130s. An order of
1127 said that when there were death sentences in cases with dubious
or extenuating circumstances, the local jurisdictions themselves could
lower the penalties. When peace returned, the jurisdictions were to
return to the former practice, but no specific order that they do so can
be found in the sources. 77 Moreover, in 1131 the central government
issued an order that because of current problems, all men subject to the
death penalty should have their sentences reduced to exile. Again, this
order was supposedly temporary, but we do not know when it was
superseded. 78 The local clerks may also have played a role. They are
accused of delaying cases and memorializing improperly, 79 and because
the majority of those who were lucky enough to have their cases mem-
orialized were spared, such a development would have tended to lower
the rates of execution. Also, in 1133 the Court reaffirmed the importance
of memorializing cases in which the relevant law seemed excessively
harsh or there were mitigating circumstances. 80 Finally, the tendency of
local officials in the Southern Sung to memorialize all death sentences,
regardless of their circumstances, was repeatedly criticized by con-
temporary officials.81
Most of those who were fortunate enough to have their cases sent up
were not executed. In Sung law there was a special category of punish-

75 WHTK 167.1452.
76 SS 200.4995.
77 HNYL 241. Of course, such an order would not have been necessary, but in similar
cases specific orders were often issued.
78 HNYL 801.
79 HNYL 1185, 1314, 1672, 2972. It is difficult to square the implications of these
complaints with a report which says that most of the cases memorialized had serious
circumstances and lenient legal provisions (HNYL 1165).
80 Liu Cheng, Huang Sung chung-hsing, 14.17b; see also HGCH 23-i9a-b. See also a decree
of 1156 instructing officials not to memorialize in a disorderly fashion asking for
reductions (SS 31.585).
81 See, for example, HNYL 1314 (1134); SS 201.5014 (1135); WHTK 170.1477 (1156),
170.1478 (1170); Hung Mai, Jung-chai sui-pi, fifth collect, p. 147.

468
 The death penalty

merits called "[nominal] death penalty [with automatic] memorializing

for a ruling." The number of crimes in this category was small relative
to the number in the regular death penalty category. Fan Ch'un-jen
reported that only 25 of the 264 cases reviewed in 1085 (9 percent) and
57 of the 254 cases reviewed between the eleventh month of 1085 and the
end of the second month in 1086 (22 percent) had their death sentences
confirmed.82 The rest had their penalties lowered to penal registration.
Possibly his report concerns only crimes in this smaller category. Note
that the percentage of cases memorialized in this way in these periods
was approximately the same, 7 percent in 1085 and 10 percent in 1086.

Types of capital crimes

One report that cites the total number of people sentenced to death also
gives us information on the kinds of crimes involved. In 1060, out of the
total of 2560 people sentenced to die, 140 cases (5 percent) involved
people who had killed relatives, including mothers, fathers, uncles, and
brothers. It also included wives who had killed husbands or concubines
and husbands who had killed the parents of their wives. Of these cases,
110 (4 percent) were sexual crimes involving homicide; 970 (38 percent)
were cases of forceful robbery; and 1300 (51 percent) were cases of
homicide - deliberate, premeditated, or during fights.83

Conclusion
In the opening section of his book Discipline and Punish Michel Foucault
describes the execution of Robert Francois Damiens in 1757. The execu-
tion lasted some days and involved various sorts of burnings and an
unsuccessful attempt to have the living body torn in pieces by horses.
(The executioners finally had to saw the limbs partly through so that the
horses could do their job.) 84 Even at their most barbarous, Chinese state
executions after the founding of the empire were no match for the
ferocity of the eighteenth-century French. If a capital case had been
handled according to law, a Sung criminal would have been treated
carefully if not kindly. He would have passed through a preliminary

82 SS 201.5012.
83 HCP i9i.ga-b; see also WHTK 167.1448.
84 Foucault, Surveiller et punir, pp. 9-13. We should not be so foolish as to think that this
sort of barbarity could not happen in this country or was confined to the distant past.
In the mid-nineteenth century a group of slaves in Missouri who escaped after killing
their master were brought back to Columbia after their capture and there were hanged,
drawn, and quartered and finally burned in the presence of the other black slaves of the
neighborhood.

469
 Law and order in Sung China
hearing at the district level, resulting in a finding and a recommendation
for sentencing that would be sent up to the prefecture for review. After
review the prefecture would forward the material to the circuit level for a
further review. A final decision could be reached at this level if the
circumstances of the case were clear and did not indicate that a more
lenient or more severe sentence was appropriate. If the circuit felt that it
could not render a final decision, the case might be sent for review at the
capital.
Convictions for serious crimes of the kinds calling for capital punish-
ment were not ordinarily possible without a confession by the criminal.
This procedural characteristic was both the cause for and the justifica-
tion of the system of judicial torture. But even this grim process was
supposedly governed by laws determining the kinds of torture that
could be legally applied, who was to apply them, how they were to be
done and with what, and the total amount of torture that could be
administered. At the end of the process the criminal was given an
opportunity to hear the case against him, read aloud in the presence
of the incumbent local officials. The accused could at this time raise
charges of judicial abuse, guaranteeing a new trial in a different venue. If
he did not protest, he had to signify his assent by signing or making his
mark on the trial statement.
After all the reviews, if the final decision was for execution, the order
was sent down to the original jurisdiction. The actual execution was also
to be carried out according to set rules, which governed time, place, and
manner. Executions could not be carried out during certain inauspicious
periods or during inclement weather. The victim was to be given food
and wine, time to meet with his relatives and to make his peace with
his gods, and a last opportunity to protest any judicial abuse. This
opportunity was his until the moment of execution. Even to the dead the
state provided a final service, burying those without friends or relatives.
In practice the death penalty process was both better and worse than
the rules surrounding it. Many of the rules supposedly governing the
process were ignored or broken with some frequency. No doubt the food,
medical care, and other facilities given to condemned prisoners often did
not live up to the regulations. Equally important, the system of inves-
tigation and adjudication may often have deviated from the prescribed
patterns. The executions, too, were sometimes carried out despite the
regulations, not according to them. Perhaps most seriously, the system
seemed to follow a kind of class justice, in which the rich and powerful
often were able to escape with their lives and the poor and weak suffered
execution. The Northern Sung official Yin Chu (1001-46) wrote:
"Regarding the crimes of the present day, nine out of ten of those
470
 The death penalty
condemned to death are poor. Those who are not poor do not fear the
law." 85
There were thus ways in which the practice was worse than the
theory. There were also, however, ways in which the practice was more
benevolent than the theory. A bewildering variety of factors might lead
to sparing condemned persons. In addition to the many death sentences
routinely commuted to penal registration, certain groups of people also
were accorded privileges that might exempt them. At times, because of
local problems, the authorities might issue amnesties that lowered the
penalties of some classes of criminals. Two other parts of the Sung
judicial system were probably even more important: the frequent issuing
of empirewide general amnesties and the trial review process. Statistics
show that of those fortunate enough to have their cases reviewed at the
capital, 10 percent or less were eventually executed. This might suggest
that local authorities appealed cases only when there were serious prob-
lems, but it might also be a reflection of clemency. In either case, from
the criminal's point of view, the result was the same.
The same was true of amnesties, which will be described in the
following chapter. Amnesties in the Sung were issued with great fre-
quency. They varied in type and coverage, but all of them lowered
penalties to one degree or another. Taken together, these devices -
amnesties, commutations, privileges for various groups, and so on -
greatly altered the actual face of Sung justice. As a result of these various
devices for reducing penalties, Sung justice may have been, in practice,
the most merciful in traditional China. The state retained the power to
execute, to demonstrate its ultimate control over people, but it combined
a formal legal pattern that enshrined this power with a set of practices
that in many cases mitigated its impact.

85 Yin Chu, Honan hsien-sheng wen-chi, SPTK ed., 2.5b.

471
 14
Modifications of penalties

Introduction — reasons for modifying penalties

The Sung dynasty was noteworthy for the combination of practices and
policies that it used to modify the penalties suffered by those convicted of
crimes. The various modifications seem to have been adopted either to
strengthen security or to encourage proper behavior. Proper behavior,
which for those guilty included personal reformation, was encouraged by
enacting policies that allowed the alteration of actual punishments so
that they would fit the particular circumstances of individual crimes and
the particular characters of those who committed them. There was a
search for justice, in the sense of fittingness, combined with the ever-
present possibility of mercy from the sovereign. As we noted, during the
Sung, many men sentenced for crimes carrying a nominal death penalty
were actually punished using penal registration, thus providing the state
with a pool of laborers and perhaps prompting in those so spared a sense
of indebtedness to the ruler.
From very early in their history the Chinese ruling group emphasized
the importance of fittingness. If the rulers of the state were to do violence
to some of their subjects, the level of punishment had to be chosen for an
appropriate reason, and the penalty had to be fitted carefully to the
crime. However, in the most widely shared Chinese view, if an error was
to be made, it should be on the side of mercy. The search for justice
was almost always the avowed justification for modifying penalties, but
justice tempered with mercy. Sung officials knew well the injunction
from the classic Book of Historical Documents: "Rather than put to death an
innocent person, the ruler should run the risk of irregularity and error." 1
The Sung is noteworthy for maintaining severe penalties in theory while
1 The passage is from the Book of Historical Documents. See Legge, The Chinese Classics, vol. 3,
The Shoo King, p. 59. For a Sung example of the citing of this passage to make a point in
favor of mercy in punishments, see Chang Kang, Hua-yang chi, 14.6b.

472
 Modifications of penalties
reducing the actual penalties by using amnesties, grants of commutation,
and rights to privileged treatment. These devices were by no means
unique to the Sung, but they seem to have been used more often in the
Sung than in other dynasties. The penalties were modified to take into
account the criminal's relationship to the victim, the role of vengeance
by those related to the victim, people's obligations to support others,
the criminal's mental or physical incapacity, the criminal's supposed
importance in the social order, and the criminal's desire to repent and
reform. Administrative practicality interacted with an ideology that
sought first to achieve justice by fitting the recompense to the injury
(according, naturally, to the views and values of the ruling group) but
that demanded that the merciful alternative be chosen in all ambiguous
cases.
Punishments also were modified because it would be too expensive, or
simply not feasible, to employ them as written. Of course, the rulers of
states rarely acknowledged their inability to enforce penalties for institu-
tional or fiscal reasons, but then most rulers are adept at deception and
misdirection. Sung leaders were often extremely capable men, and they
recognized that when penalties could not be enforced, the state should
somehow try to take credit for having refrained from using them. They
also were aware that when the general consensus was that a specific
penalty was excessively harsh for the crime, it would be very difficult to
get convictions under the law.
Penalties were also altered to mobilize political support. Immunity
from the law, or special treatment by the law, are two very common
commodities traded by states for the support of the powerful.
Because penalties are messages and communication is at the heart of
social life, its correctness is vital. An unjust penalty is a disaster. From
the earliest written tradition in China the authorities repeatedly stress
the importance of measure. Only the appropriate can rectify; only the
fitting can set right. Penalties could be made heavier if the need for
fittingness demanded. The "Punishments of Lu" section of the Book of
Historical Documents, which was believed by traditional Chinese literati to
date from the Early Chou period (1027-256), sets the pattern for this
change: "When [the crime] should incur one of the lower penalties but
there are aggravating circumstances, apply to it the next higher." 2
The same source was the locus classicus for the precedent of reducing
penalties when that was appropriate. The passages were well known to
the officials of the Sung. During the war against the state of Hsi-hsia in
the 1040s, the statesman Fan Chung-yen cited the Book of Historical

2 Legge, The She King, p. 607.

473
 Law and order in Sung China

Documents to justify the institution of a system of redemption in certain

troubled border areas in the northwest:
I have read the "Canon of Shun" where it says "Money may be received for
redeemable crimes." Moreover, the "Punishments of Lu" says, "If the Five
Punishments do not meet it, let them adjust the case to one of the five redemp-
tion fines." In this way Yu, Shun, and the Duke of Chou used the law on
redemptions.3
This tradition continued from the Chou to the Sung. The rule was
that cases might be sent up for a ruling if the facts and the law for the
case did not seem to fit together, either because the circumstances
seemed too heinous for the penalty prescribed by law or the penalty
seemed too severe under the circumstances. For example, in one case
from the late eleventh century, a military man had beaten his wife to
death, a crime that called for tattooed penal registration. However, he
had been provoked to act when his wife had cursed his mother. In light
of what the traditional Chinese authorities ruling on the case felt were
mitigating circumstances, his sentence was reduced to a beating and
demotion. 4 Such rulings, nominally acts of the sovereign, could there-
after serve as precedents for extending mercy in other cases with similar
circumstances.

Heavy penalties
Penalties were made heavier for two reasons, because the acts were
considered particularly heinous or the crimes were committed in regions
where maintaining security was especially important. Penalties could be
made heavier in a variety of ways. One way of increasing penalties was
to order that certain crimes be "treated in accord with military law" (i
chun-fa). This phrase occurs most often in the sources covering the
war crisis of the late 1120s and early 1130s, though it does appear
sporadically at other times. It is simply another way of ordering that
those guilty were to be beheaded. 5
3 HCP 143.27a; Legge, The She King, pp. 38, 602.
4 HCP 332.11b.
5 For the general topic of military law in the Sung, see McKnight, "The Military and the
Law in Sung China." Penalties were also changed for formal reasons, as a simple way of
amending previous laws. One common way of amending the law was to state in an edict
that the penalty for a particular crime was to be the same as the penalty for another
crime, increased or decreased by a fixed number of degrees. Such rulings entered the
general body of the law and from that point on were not modifications. There were,
however, a variety of conditions that clearly called for modifications of the general
system. See also a memorial of Han Yiian-chi (1118-after 1178) suggesting that military
law be used against the "youthful banditti" who were disrupting communications with
Yen Prefecture in Liang-che Circuit (Han Yiian-chi, Nan-chien chia-i kao, 10.15a).

474
 Modifications of penalties
The best-known Sung system for increasing penalties, the "Law on
Heavy-Penalty Jurisdictions," made more severe the punishments for
certain kinds of crimes (most importantly robbery) in certain areas. 6 The
creation and spread of this system is associated with attempts to main-
tain security in critical regions. Although the phrase occurs in Sung
sources as early as 1013, when an official "used heavy penalties" to
punish corruption among some clerks, the creation of the system is
usually traced to a decree of 1035 t n a t redetermined the penalties for
robbery (chiang-tao). At the end of the report on this decree the annalistic
history known as the Ch'ang-pien states that there was "also a decree that
in the capital city, if thieves who were armed got booty worth four
thousand coins, they were to be tattooed and listed in the army. From
this time on, the law on theft in the capital was especially heavy." 7
Unfortunately the sources spell out most of the penalties for robbery
and theft, but not the penalty for armed theft in areas outside the
capital. We can only assume that the authors of our sources knew that
this newly ordained penalty for use in the capital was heavier than the
penalties inflicted elsewhere. As the system was developed later in the
century, it was aimed mainly at robbery, for which it set especially heavy
penalties.
This rule for the capital may have been discontinued. The "Treatise
on Punishments" in the Sung History reports that in 1062 the heavy-
penalty system began in the districts of the capital city, as well as in
nearby Ts'ao Prefecture in Ching-tung West Circuit, Hua Prefecture in
Ching-hsi North Circuit, Shan Prefecture in Hopei, and P'u Prefecture
in Ching-tung West Circuit, all regions lying to the north and east of the
capital. The annals of the Sung History and the Ch'ang-pien explain that
this system was extended in 1071 to Hua Prefecture in Ching-hsi Circuit,
Su b Prefecture in Huai-nan Circuit, Ch'an Prefecture in Ho-pei Circuit,
and ten prefectures in Ching-tung Circuit. 8 In the summer of 1077 the
system was extended to ten new prefectures and to parts of three more
prefectures in Ho-pei East, Ching-hsi North, and Huai-nan prefectures. 9
When these prefectures are plotted on a map, it becomes clear that at
first the system was created and extended to ensure security in the areas
near the capital. The farthest prefecture was the exception, lying on the
Yellow River some 250 miles northwest of the capital. The new areas
added in the middle of 1077 were located within or on the edges of this
fan-shaped region (see Map 14.1).

6 SS 199.4978 gives a good brief description of the system and its history.
7 HCP 117.4a; SS 199.4976; WHTK 167.1447.
8 HCP 219.6b; SS 15.278, 199.4978.
9 HCP 282.2b, 283.2a.
475
 Law and order in Sung China

1 Kaifeng
2 Hua
3 K'ai-te
4 P'u
5 Ts'ao
6 Kung
7 Ch'i
8 Yung
9 Yenb
10 Chib
11 I
12 Tan
13 Ying-tien
14 Su
15 Hsu
16 Huai-yang
17 Ta-ming
18 Mingb
19 Hsing
20 Po
21 Te
22 Ti
23 Pin
24 Po b
25 Hao
26 Ssu

D Initial prefectures in 1062.

O Prefectures added in 1071.
A Prefectures added in 1077.
Map 14. i. Early expansion of the heavy-penalty jurisdictions

476
 Modifications of penalties

Local officials in areas affected by banditry began at this time to

submit memorials asking that the system be extended to their jurisdic-
tions, though the court remained cautious about extending the system.
In late 1077 the system first was applied to areas away from the capital,
being extended to four mountainous prefectures in interior Fukien. 10 In
1081 the pacification commissioner of Yung-hsing Chun Circuit (cor-
responding roughly to the present Shensi Province) complained of
bandits in the border areas and asked that the heavy-penalty system be
applied, but only to gangs of three or more. This same document says
that "Shensi Circuit followed this." 11 This 1081 memorial does not
indicate whether the system was to be extended to all of either circuit or
only to the border regions. However, because in the past the system
had been extended piecemeal, to individual prefectures and often to
individual districts, it probably was limited to specific prefectures. This
conclusion is strengthened by a memorial of 1086 that asked that the
system be extended to a district in Yung-hsing Chun Circuit. Such a
request would make no sense if the system had already been extended to
the whole circuit. 12
In the past, some scholars have assumed that the system was extended
to whole circuits, but this seems clearly mistaken. When the sources
provide detailed information about extending the system, they indicate
that it was always extended prefecture by prefecture or district by
district. 13
The pattern of using gangs of three as the threshold below which the
heavy-penalty system would not apply, which may have been based on a
precedent of 1075,14 was used when the system was applied to certain
other regions, for example, to the mountainous prefectures of Kuo and

10 HCP 286.7a; see also HCP 394.20a.

11 HCP 319.1a.
12 See HCP 375.1a.
13 In some of these cases, in some of the more abbreviated sources, are statements that,
taken alone, suggest extension to the whole circuit, but more detailed sources indicate
that these reports are inaccurate. For example, the annals of the Sung History say that in
1071 "they established the Ching-tung and Hopei Circuit heavy-penalty law for
banditry" (see SS 15.278). However, the Ch'ang-pien makes it clear that the law was to
be used only in a small minority of the prefectures in these circuits. Again, for 1081 the
sources quote a request that might be interpreted as a request that the system be
extended to the whole circuit of Yung-hsing Chun (see HCP 319.1a). But as we noted,
this can hardly have been the case, given a request later in the same year that the
system be extended to one of the districts in this circuit (see HCP 375.1a).
Furthermore, when in 1091 Fan Tsu-yii comments on the creation of the system in
1071, he gives exactly the same list of prefectures and districts provided in the report
available to us for that earlier year (see HCP 468.13a).
14 HCP 267.10b.

477
 Law and order in Sung China

Shang, and Ch'ien-yu District in Yung-hsing Chun Circuit in 1086,15

and Ju Prefecture in Ching-hsi North Circuit in 1087.16
As the name itself indicates, the law was specifically aimed at robbers
and thieves. The system of 1071 increased the penalties for robbery and
theft by punishing the criminal's family and confiscating his property. If
the sentence for a robber called for the death penalty, then his patrimony
was confiscated and used to reward informers. The wives and children of
the criminal were sent for registered control (pien-kuan) at one thousand
li.d They were not to be freed by amnesties. Robbers who themselves
benefited from an amnesty were to be penally registered in "distant evil
prefectures."
When the robbers were liable for penal servitude or exile rather than
death, they were to be sent to Ling-nan for penal registration. An
intervening amnesty might reduce their sentences to registration at
three thousand lid One-half of their patrimony was to be confiscated,
and their wives and children were to be subjected to an appropriately
reduced level of registered control. Thieves in these heavy-penalty areas
who were guilty of three crimes were to be beaten and penally registered
at five hundred lid or in a neighboring prefecture.
Those who harbored criminals from heavy-penalty places, even if they
themselves did not live in a heavy-penalty place, were to be severely
punished. Those who harbored a heavy-penalty robber who was liable
for death and whose crimes involved "serious circumstances" were to
be beheaded. Harboring other heavy-penalty robbers called for penal
registration in "distant evil prefectures." In both cases the harborers'
patrimony was to be confiscated. Harboring a robber liable for penal
servitude or exile was punishable with penal registration at five hundred
lid and confiscation of one-third of the patrimony. 17
The emphasis on those harboring criminals reveals part of the security
problem. Local people often protected their own, or they might aid the
bandits out of fear of reprisals. Concerned that the people in isolated
villages, especially those whose livelihoods were not secure, might
become bandits either voluntarily or through coercion, the Sung auth-
orities in 1097 adopted a policy of bringing such people into nearby
villages, organizing them, and training them for self-defense, a technique
widely used in other times and places.18
A document from 1083 implies that the penalties involved had not

15 HCP 375.1a.
16 HCP 424.8a.
17 HCP 344-2a-b; SS 1994978-
18 HCP 487.1a.
478
 Modifications of penalties
19
changed at that time. But, the system unfortunately was caught up in
the continuing struggle between the partisans of Wang An-shih and his
new laws and their conservative opponents. The heavy-penalty system
became associated with the partisans of reform, and so when the con-
servatives came to power after Shen-tsung's death in 1085, they lowered
the penalties. This change is reflected in the newly compiled edicts of the
Yiian-yu period (dated 1088). Most importantly, the wives and children
of robbers, thieves, and those who harbored them, were no longer to
be subject to registered control. 20 In 1092 the penalties were further
reduced. Only if there were five or more especially evil robbers were the
heavy penalties to be used. In other circumstances, ordinary law was to
apply, and the same rule was to be used for harborers. 21
Then in 1098 Ts'ai Ching, the uncompromising follower of Wang An-
shih, returned to power shortly after the accession of Emperor Hui-
tsung. Ts'ai Ching reinstituted the earlier, harsher laws. In 1098 the
Ministry of Justice successfully petitioned that the Yiian-feng edicts,
with their heavier penalties, be used and that no consideration was to be
given to the number of criminals. 22

Personnel policies in heavy-penalty jurisdictions

The system affected personnel policies in the heavy-penalty areas, as
well as criminal punishments. Beginning in 1071, military officials (wu-
ch'en) were selected as sheriffs in the heavy-penalty areas. 23 In 1082 it
was further ordered that in heavy-penalty places, only servitors (shih-
ch'en) were to be appointed sheriffs, and the sheriffs of neighboring
jurisdictions were to be military officials.24 Local officials, including
the prefects and controllers-general, were to be especially selected by
imperial appointment (rather than by the regular process of selection by
the Ministry of Personnel). 25
Two leading officials, Fan Tsu-yii and Su Ch'e, bitterly attacked the
system in the late 1080s and early 1090s. In 1086 Su Ch'e wrote:
19 HCP 344.2a-b.
20 HCP 409.20a; SHY, hsing-fa 4.30a.
21 HCP478.ua.
22 HCP 499.3b, 499.16b-17a, 500.23a. In 1103 it was ordered that in heavy-penalty
places where in the recent past there had been military sheriffs, the older, more severe
rules from the Yiian-feng period (1078-85) should be reestablished (see SHY, chih-
kuan 48.66a).
23 SS 199.4978.
24 SHY, chih-kuan 48.65a; HCP 328.2a; WHTK 63.574; Wang I-chih, Chih-yuan ts'o-yao,
pp. 19-24. In the preceding translation I have conflated several versions of this
memorial.
25 HCP 359.10b. On this process of selection, see also HCP 37o.29a-29a.

479
 Law and order in Sung China

Under the old system, district sheriffs all were chosen from among regular civil
service candidates [hsiian-jen]. In recent years, military officials [wu-ch'en] have
been used. Since this change the people have been impoverished, and there have
been many bandits. Those who discuss it do not understand, and so they ask
that in heavy-penalty places the sheriffs always be military men. But I have
not heard that after this change in the law, the number of bandits decreased
as a result. Moreover, the military men are corrupt and have no fear of the
regulations. . . . I feel that when capturing bandits the most important thing is
first to have good archers. Next you must get the aid of the villagers and then
spread your net. If beforehand you understand the methods, then it will be easy
to complete the job. Under the old system, when regular civil service candidates
were used, even if they could not follow [the way of capturing bandits] point by
point, still they were respectful of the laws, cared for them and used them. They
had good relations with the people. They did not necessarily themselves shoot
and ride, but they were able to apprehend [the bandits]. . . . I have heard that in
Ho-pei, Ching-tung, and Huai-nan circuits, where military sheriffs were used,
they caused grief to the people. I ask that you again tell the Ministry of
Personnel to use only regular civil service candidates." There was a decree that
except for border districts [where military men should be used] only regular civil
service candidates should be used.26

Five years later Fan Tsu-yii went further than Su Ch'e had and
argued that the entire system of heavy-penalty places be abolished. After
noting the establishment of the system in 1071 and its subsequent
extension, he said to the emperor:

Under Heaven what people are not your people? Now I see that the prefectures
and districts are treating the people as if they were barbarians. Twenty years
have passed since the system of heavy penalties was established. I have not
heard that the bandits have declined, much less disappeared. Rather, they have
become more numerous. The ancients sought the wellsprings of popular welfare
and established officials to instruct the people. This they did first... and after-
ward they used penalties.. . . Also, in the heavy-penalty places, when selecting
prefects and district magistrates, you have necessarily chosen valiant officials
determined to uphold the law and suppress banditry. These men look on the
people as the enemy, feeling simply that their official function is to use armed
force and pass penal judgments. How can such behavior fit a peaceful age?27

Fan Tsu-yii's plea fell on deaf ears, but he was not a man to give up
easily. Late in the following year he again memorialized and again
bitterly attacked the system of heavy penalties as discriminatory,
inequitable, and counterproductive:

26 SHY, chih-kuan 4 8.6 5 b-66a; WHTK 63.574; HCP 385.3a.

27 HCP 468.12b-13b.

480
 Modifications of penalties

I have heard that the king is like Heaven - there is nothing that does not respond
to his influence. That the king is like Earth - there is nothing that he does not
support. Within the Four Seas everyone is his child. No place is distant. All are
one. Nowadays, in heavy-penalty places it is true that if one man commits a
crime, the penalty includes his wife and children and destroys his patrimony.
This is just like the treatment of a rebel. Under the Former Kings the penalties
always caused people to renew themselves. I have never heard that they dis-
criminated among prefectures or other territories, practicing the law in an
uneven manner. I am worried that such behavior is not fitting for this sagely era.
Your majesty is about to sacrifice to Heaven and Earth. You will be going to the
tower to issue an amnesty. If the document of amnesty completely abolishes this
law, ending it utterly, this will signal the beginning of a new era, which will
succor the hearts of the people and give rise to a harmonious balance of the
elements.28
His plea was unsuccessful. The system did remain in use. But his
criticism might have triggered the reduction of penalties that was enacted
in the early 1090s.
The large-scale use of the heavy-penalty system apparently was dis-
continued at some time during the closing years of the Northern Sung or
the first few years of the Southern Sung. In 1133 it was ordered that
people who violated the tea and salt laws might be subjected to the
"heavy penalties." And in n34 an official asked that "in the capital
heavier penalties be used as they had been under the old system in
Kaifeng."29 It seems likely that this limited request was approved, but
the fact that an official had to ask for the application of such penalties
suggests that they were no longer in regular use. There is little evidence
that the system was used elsewhere during the Southern Sung. This
would certainly fit with the original intentions of the systems' creators,
who had intended it to be temporary, to be used because of particularly
serious security problems, and to be discontinued once those problems
were corrected.30

Reductions of penalties
The Sung authorities could and did increase penalties on occasion in
some regions; they also often reduced them. The precedent and justifica-
tion for doing so were drawn from early Chinese tradition. Sung thinkers
assumed that upright judges embodied a moral sense that accorded with
the natural order, but even upright judges could be misled by confusing

28 Fan Tsu-yii, Fan t'ai-shih chi, 24.2a; Huang Huai, Li-tai ming-ch'en tsou-i, 218.19b.
29 Liu Cheng, Huang Sung chung-hsing, 14.22a; HNYL 1243.
30 See, for example, HCP 402.12b, 424.8a.

481
 Law and order in Sung China

evidence or by the perjury of witnesses and accused criminals. They saw

judges as walking a narrow line and recognized that even the best must
sometimes put his foot wrong. In clear cases, it was necessary to rely
on the magistrate's judgment, but in unclear cases which way should
the j u d g e lean in handing down the sentence? Was it better to favor
excessive harshness or excessive mildness?
From the time of the earliest Classical writings in China in the early
first millennium B.C., there is no doubt about the standard Chinese
answer. If we must step off the path, it should be on the side of mercy.
T h e best known, and in later times the most influential, statement of this
attitude is found in the Book of Historical Documents. This work, preserved
because Confucius used it to instruct his disciples, was regarded by later
Chinese as a fountainhead of political wisdom. In the section entitled the
"Punishments of L u " - which was composed late in the Chou dynasty
but was believed by most traditional Chinese readers to have been
written early in the Chou - we are told: " W h e n there are doubts about
the infliction of any of the Five Punishments, that infliction should be
forborne. When there are doubts as to the infliction of any of the five
fines, it should be forborne." 3 1
T h e "Punishments of L u , " which lays out in detail the ways in which
and degrees to which mitigating circumstances should alter sentences,
exercised great influence over later Chinese penal practice. Although
during the Sung, penalties were sometimes increased, it was far more
common for them to be reduced or eliminated. During the reign of
Emperor Chen-tsung (r. 997-1022) after the emperor had attended a
lecture on the passage from the Rites of Chou, which had focused on the
passage " G r e a t famine, great disorder, light government, light penalties,"
an official commented that lightening penalties was a policy designed to
be used only with people who had erred through mistake, who in years
of poor harvest might be spared because of their poverty. T h e emperor
disagreed and opted for mercy. 3 2
Such grants of mercy took various forms. Occasionally men sentenced
only to beatings were simply freed on certain holidays, 3 3 or when natural
disaster struck, the authorities might order that all penalties in the
affected regions be reduced until the return of normal conditions. 3 4 T h e
"Treatise on Punishments" in the Sung History says:

In years of dearth, strong people would band together and take up weapons to
rob men and pillage storehouses. According to the law, they all were liable for
31 Legge, The She King, p. 604.
32 SS I99-4988.
33 HCP 99.14b, note.
34 HCP 251.28a.

482
 Modifications of penalties

execution. When the case records were sent up for review, they would be spared
death. During the reign of Chen-tsung [r. 997-1022] in Tsai Prefecture, there
were 318 men all liable for execution.... The emperor sent down a decree
sparing them. He sent an envoy to soothe the people of the circuit. He said to the
envoy, "When ordinary people find it difficult to feed themselves, the strong seize
food to preserve their lives. We cannot apply the law on robbery to judge
them." 35

Under Jen-tsung (r. 1022-63), this became a common practice. After

reporting on one such case from the year 1053 the historian Li Tao noted
that "subsequently this was written up as a precedent (/fe)."36 And a
decree from Shen-tsung's reign (r. 1067-85) calls for the reduction of
sentences by one degree for most sorts of criminals in prefectures suffer-
ing from natural disasters, with a further reduction of one degree if the
crime were stealing food.37
Legal conservatives like Ssu-ma Kuang (1019-86) argued against
such practices, saying that they were tantamount to giving permission to
steal. His colleague Yang An-kuo questioned the meaning of the passage
advocating mercy from the Rites of Chou, saying that it was intended to
cover only crimes through misadventure. The poor were to be pitied in
bad years, but if "a crowd takes up arms to attack a granary, to forgive
them is no way to restrain evil." The emperor disagreed. "All those in
the world are my children. There is hunger that the prefectures and
districts cannot assuage. Being so pressed, people steal. To arrest and
execute them, is this not extreme?"38

Vengeance
In ancient China, vengeance played a key role in requiting perceived
wrongs. Indeed, some Sung officials traced the emergence of a state-
administered system of legal punishments to policies established by the
early Sage Rulers, who had wanted to limit the possibilities of vendetta.
Lou Yueh (d. 1244) wrote:
Homicides have always occurred. If A killed B, then B's son or grandson would
kill A to avenge his father. When things were this way, then there would
be mutual vengeance without end. Therefore the Former Kings brought this
under control by systematizing penalties, and instructing their officials to stop
vendettas.39

35 SS 200.4987.
36 HCP 175.4b; see also HCP 171.8a.
37 HCP 267.10b-1 ia.
38 SS 200.4988.
39 Lou Yueh, Kung-kuei chi, 22.332.

483
 Law and order in Sung China

While defending the right of the state to be the sole agent in the
punishment of most crimes, the Sung also continued the traditional
practice of reducing the penalties of those who committed crimes to
avenge crimes committed against their elders. This principle was even
enshrined in the T'ang and early Sung codes:
If a grandparent or a parent has been assaulted and their sons or grandsons in
turn assault the attackers but do not cause them injury, the sons and grandsons
are not to be tried. If they do cause injury, they should be punished three degrees
less than in ordinary cases of assault with injury. If death results, follow the
regular statute.40
In practice the authorities were sometimes, though not always, even
more lenient than this statute. The popular attitude is perhaps reflected
in the comments of the writer Wang Yung, who records an incident in
which two sons avenged their father by killing his murderer. One official
wanted to have the sons executed, and another wanted to spare them.
Eventually the sons were executed. Wang Yung comments that the
lenient official "was an exemplary gentleman [chun-tzu] who took pleasure
in doing good for men." The harsh official "was a petty man [hsiao-jen]
who took pleasure in doing evil to men." He then goes on to cite a case
from the time of the second Sung emperor in which a boy avenged his
mother by killing a man. By special decree his life was spared, and he
was beaten and sent into exile.41 Again, during the reign of Emperor
Jen-tsung (r. 1022-63) a son killed the murderer of his father, when that
man had been spared by an intervening amnesty. The son was spared
death and sentenced to registered control. In another incident from the
reign of Shen-tsung (r. 1067—85) a boy's father was killed in a fight.
Because the boy was young, he could not avenge his father. But when he
grew older he killed the murderer with a knife, cut ofF his head and
limbs, and went to worship at this father's grave. He surrendered and
accepted the blame. Eventually he was sentenced to penal registration in
a neighboring prefecture. 42
Another incident, also from the reign of Shen-tsung, shows how kin
relationship might affect the treatment of vengeance. When his father
was killed by his father's brother's son, the son of the victim in turn
seized and killed the murderer. The Court of Judicial Review and the
other central government agencies that reviewed the case felt that the
young man should be executed for having committed "discord" (pu-mu),
one of the Ten Abominations which covered many crimes committed

40 TLSI 3.106; SHT 23.356.

41 Wang Yung, Sung-ch'aoyen-i tai-mou lu, 4.ia-b.
42 SS 199.4990; HCP 110.12a, 295.7b.

484
 Modifications of penalties
against relatives. The most powerful official of the time, Wang An-shih,
disagreed and wished to have the sentence reduced to exile with added
labor, with the stipulation that the man could be freed if an amnesty
were proclaimed. The emperor agreed with Wang An-shih and spared
the young man. 43 This case illustrates another facet of the reduction
of penalties in Sung China, the granting of individual pardons. The
emperors frequently granted such pardons, but their importance clearly
was overshadowed by the more general grants of mercy called amnesties.

Amnesties
The most striking, and in terms of its effects probably the most import-
ant, of the devices that modified the penalties of Sung criminals was the
system of state amnesties. 44 The Sung amnesty system clearly reveals the
interaction of ideology and practicality characteristic of Sung penal
practice. Amnesties, which were granted more often during the Sung
than during any other Chinese dynasty, were justified as acts that would
encourage criminals to reform themselves in response to imperial mercy;
the hard fact seems to have been that the police system put so many
people into the judicial and penal system that they had to be periodically
purged to make room for newcomers.
Amnesties had been granted during Chinese antiquity, and the
rationale for them is stated in the Book of Historical Documents. As a
system the practice of freeing criminals or reducing their sentences began
in the Han dynasty. From the reign of Han Kao-tsu (r. 206-195) the
government regularly issued decrees of amnesty. The Three Kingdoms
(220-80) and the Northern and Southern dynasties (317-581) con-
tinued the practice. Over time, the form and content of these amnesties
became increasingly fixed. By the T'ang, the system as it was to be used
in the Sung was largely complete.
From the T'ang also we have our first relatively complete texts of the
amnesty documents. These amnesties were basically of two sorts, what
might be called ordinary amnesties and Great Acts of Grace (ta-she).
Ordinary amnesties included both localized acts, which affected only
specified parts of the empire, and empirewide grants, as well as the
practice called "inspections of cases," which in theory was a device for
speeding the judicial process but in practice always resulted in the
lowering or forgoing of penalties for those inspected. 45

43 SS 201.5007.
44 For more on amnesties, see McKnight, The Quality of Mercy.
45 The Ch'ing-yiian t'iao-fa shih-lei contains a variety of rules governing the use of these
various amnesties. See, for example, TFSL 503, 518, 522, 527, and passim.

485
 Law and order in Sung China

T h e terms of ordinary amnesties cannot be simply described. They

came in different forms, but generally they were less generous than those
of the Great Acts of Grace. They usually did not apply to men guilty of
the group of crimes called the Ten Abominations - most homicides,
arson, and some other crimes that carried the death penalty 4 6 - or
to those who committed crimes deliberately hoping to be spared by
an expected amnesty. 4 7 Ordinary amnesties also often reduced some
penalties rather than forgiving them entirely. Like the more generous
Great Acts of Grace, such amnesties also applied to deserters and fugi-
tives from justice, who within a set period of time could surrender to the
authorities and confess, in return for which they would be spared or have
their penalties reduced. 4 8 In the words of one commentator, after an
amnesty men might be "reranked by degrees (if they were officials), have
the location of their punishment shifted, or be freed." 4 9
Great Acts of Grace always offered more generous terms than did
empirewide ordinary amnesties. In most cases a Great Act of Grace
freed all criminals, with certain very limited exceptions. 5 0 A typical
example, dating from the accession of the Sung founder in 960, states:

We confer a Great Act of Grace [ta-she] on the world. On the fifth day of this first
month, just before dawn, all criminals are to be pardoned, whether or not the
criminal cases have already been completed, whether or not the crimes have
been discovered, without distinguishing between the serious and the minor and
including those not forgiven under the terms of ordinary amnesties [ch'ang-she].
Officials who have been dismissed, degraded, censured, impeached, or expelled
from office are to be given grace. All those registered for penal labor, whether
men or women, are to be freed according to convenience. 51

Ordinary criminals liable for death, as well as all criminals sentenced

to lesser punishments, were to be freed and spared further prosecution,
whether or not their crimes had already been discovered and whether or
not their cases had already been completed. Because of this stipulation
the state forbade the bringing of cases based on crimes committed before
the last Great Act of Grace. Great Acts of Grace thus served as statutes
of limitations. Officials who dared to try such earlier cases were to be

46 See, for example, Ch'ien Jo-shui, T'ai-tsung huang-ti shih-lu, Ku-hsiieh hui-k'an ed.,
26.8a.
47 TLSI 4.85; SHT 30.490 ff.
48 Huang Huai, Li-tai ming-ch'en tsou-i, 218.11b-12b.
49 Wang Ying-chen, in Huang Huai, Li-tai ming-ch'en tsou-i, 217.9.
50 See Li Yu, Sung-ch'ao shih-shih (Shanghai: Ssu-k'u ch'iian-shu chen pen, 1955), chap. 2,
for some typical Sung examples.
51 Li Yu, Sung-ch'ao shih-shih, 19. These acts had relatively fixed formats. See also, for
example, Li Kang, Chien-yen shih cheng-chi, Sung ch'eng hsiang-li Chung ting kung
pieh-chi ed., chung. iaff., which quotes an amnesty with almost identical wording.

486
 Modifications of penalties
liable for the same penalty as the criminals involved were (though if the
criminal were liable for death, the officials were subject only to exile with
added labor). 52 Such amnesties both freed criminals and expunged the
record of their crimes. If a criminal freed by an amnesty was later
prosecuted for another crime, his previous crime could not be cited to
define him as a recidivist. These Great Acts of Grace also specifically
freed "those men not freed by ordinary amnesties."
Amnesties were the most important documents sent out from the
capital. They were carried by the postal service at the fastest possible
rate (at five hundred li per day), 53 and when they arrived at a local
jurisdiction they were made known with all due formality. A Sung local
official describes a ceremony for receiving a letter of amnesty that sounds
very like the ceremonies used in the T'ang. He "gathered together all the
officials, the Buddhists and Taoist priests, and the commoners to read it
to them." 54 After having arrived at a prefecture and been formally
received there, the amnesty letters were distributed to subordinate dis-
tricts for action. 55
After a letter of amnesty arrived, the local officials were supposed to
write reports on the prisoners currently in their jurisdiction. If these
people had good records of behavior and if the circumstances of their
original crimes were not judged especially evil, they were to be freed
"according to convenience." This last phrase, which occurs in most
amnesty documents, seems to mean that the local authorities had some
leeway in the timing of the release of their charges, though they could
not simply detain them indefinitely. They could, however, hold people
for long periods under certain circumstances. In 1156 it was reported
that a woman who had been accused of having killed her husband had
been in custody for five years, a span during which two Great Acts of
Grace had been issued, without her case's having been decided. Of
course, killing a husband was an especially heinous crime, and so it is
not surprising that she had not benefited from the amnesties, but the
length of her incarceration without judgment was felt to be shocking. 56
The circuit judicial intendant was responsible for seeing that amnesties
were properly administered, though frequent comments suggest that

52 See TLSI 4.6; SHT 24.373. There were a few exceptions to this - crimes not forgiven by
the rules of ordinary amnesties or crimes that had resulted in infractions of the rules on
personal status such as slaves marrying commoners.
53 HCP 457.3b.
54 Chang Yung, Sung-pen kuai-yai hsien-sheng chi (Taipei: Commercial Press, 1972), 10.28b.
For the T'ang practice, see the description given by Ennin. Edwin O. Reischauer,
trans., Ennin's Diary (New York: Ronald Press, 1955), p. 180.
55 Yang I, Wu-i hsin-chi (Pu-ch'eng i-shu ed.), 13.1 b£T.
56 HNYL 2808.

487
 Law and order in Sung China

fairly often officials were obstructive and dilatory in releasing pris-

oners. 57 Thus, although it is clear that not all those held benefited from
amnesties, it also seems clear that a great many were freed or at least
had the conditions of their punishment made less harsh. 58
Compared with other dynasties the Sung was most generous in its
granting of these various amnesties. During the 319 years of the dynasty
there were 95 Great Acts of Grace, an average of 1 every 19 months.
During the Northern Sung, there were 114 Great Acts, an average of 1
every 17 months; during the Southern Sung the average dropped to 1
every 25 months. The number of Southern Sung empirewide ordinary
amnesties, 18, was also markedly lower than the figure for the Northern
Sung, 39, though some of this change may well be due to the much more
limited sources for the later era. But even at this lower rate, the Sung
still was the most generous of dynasties in giving grace.
In addition to these empirewide ordinary amnesties and Great Acts of
Grace, there were also a plethora of localized amnesties during the Sung.
Some affected only a few districts, but others affected large areas of the
empire. There were, literally, hundreds of such grants, and the emperor,
of course, also had the power to pardon specific individuals.
Moreover, the Sung was also the high point for the use of another
device for reducing the penalties of criminals, the so-called inspections of
cases (lu-ch'iu) . 59 This practice of inspecting cases began in the Later Han
period, ostensibly as a way of speeding the judgment of cases, to spare
criminals long stays in jail. It was not originally intended as a way of
extending grace, but it rapidly became one. Succeeding dynasties con-
tinued the practice, which reached its height in the Sung. In the eighth
month of 966, the Sung founder personally reviewed and passed judg-
ment on cases from the capital city, freeing eighty criminals. Under his
successor Ta'i-tsung (r. 976-97), the inspection of cases became a

57 SHY, hsing-fa 4.27a, 45a-b, 46b~47a, 51a, 67a. Officials often argued against releasing
the truly evil and for carefully evaluating the others. See, for example, Yang Chieh,
Wu-weichi, SKCSCP 1974, I5.i7a-i8b.
58 There are two incidents from the 1070s that seem to contradict each other regarding the
numbers who might be freed. A 1076 report said that as a result of an amnesty, 1726
men out of a total of 2500 had had their sentences reduced or had been freed. However,
in 1078 after an order was sent down in the capital area for the freeing of prisoners, an
official reported that of the 70 men being held by the Finance Office, only 4 were freed,
and only 5 were freed of the 100 being held by the prefectural authorities. My
impression, on the basis of a wide variety of impressionistic data, is that the first
example is more typical than the second. Indeed, it may be that whereas the first refers
to a general amnesty, the second is a special order with far more restrictive terms (see
HCP 276.9b, 288.9a).
59 SHY, hsing-fa 5, is composed largely of reports on this system. It is also described
briefly in SS 199.4996 and WHTK 166.1445.

488
 Modifications of penalties

regular practice that was continued by his successors until the end of the
dynasty.
Such inspections came to be carried out not only in the capital city but
in all the circuits of the empire. 60 During these inspections the chief
figure, whether it be the emperor or in the outer areas a circuit intendant
(or his agent), apparently actually spoke face-to-face with the prisoners.
A collection of anecdotes records an instance in which one such prisoner
had the temerity to joke with the emperor about being a neighbor and as
a result was pardoned. And in the same collection, an anecdote is
recorded to illustrate the punctiliousness of one official. This man, it was
said, would never stoop to speaking in the common vulgar language. So,
at one time, when he was interviewing prisoners in his role as a circuit
intendant, he asked one of the prisoners if the man knew of any judicial
abuses. The prisoner could not even understand the questions he was
being asked until one of the other officials present translated it into
common parlance. 61
During some periods of the dynasty, these inspections were conducted
not just once but twice or even three times during a single year. The
earliest of these inspections in the Sung were held in the summer, to
show imperial concern for prisoners during the hot weather. 62 Later, a
regular winter inspection was added. By the early thirteenth century the
Sung authorities classified such inspections as just one more form of
amnesty. 63 These inspections certainly benefited most of those being held
in jail, though it is clear that many of the specific orders excluded very
serious sorts of crimes. The general pattern was for the sentences of
serious criminals to be reduced by one degree, for example, from death
to exile, or from exile to penal servitude, whereas men liable for lesser
penalties such as beating with the heavy rod (and sometimes penal
servitude) were freed.
On first reading these amnesty materials one may be tempted to
assume that the system was not strictly enforced. How could a state
function if it released so much of its criminal population? Yet there is
much evidence that the system was used, and although there were
exceptions, the general impression gained from reading the sources is
that the system did in fact free many men. The court did take note of
reports that some prefectures were not releasing men registered for labor
60 Compare Fan Tsu-yii, Fan t'ai shih chi, ig/5b-6a.
61 Fan Chen, Tung-chai shih-chi (Beijing: Chung-hua shu-chu, 1980), pp. 2, 21.
62 See Ts'ai Hsiang, Tuan-ming chi, 23.10b-12a; Chang Shou, P'i-ling chi, 1.14.
63 TFSL 226. Even in the Northern Sung, officials recognized this situation. See a
complaint by the Northern Sung official Ts'ai Hsiang that the traditional aim of such
inspections was to ensure speedy justice not to extend universal grace, as was being
done under the Sung. Ts'ai Hsiang, Tuan-ming chi, 23.10b.

489
 Law and order in Sung China

and ordered the judicial intendants to investigate and impeach the

responsible officials, revealing both the occasional failure of the system to
operate as intended and the court's concern to reduce such violations. 6 4
T h e system's actual effectiveness is also reflected in the continuing crit-
icisms of legal conservatives, who never ceased to rail at the evil effects of
amnesties. An official writing in the 1030s quotes what he asserted was
popular opinion:

The Three Kings sacrificed annually at the altar of earth, but amnesties were not
given out in a disorderly way. From the time in the T'ang when troops became
excessively powerful, the rites for serving Heaven were not regularly practiced,
and therefore Great Acts of Grace were used to rectify disorderly cases. And yet
those guilty did not necessarily renew themselves, even though they had been
treated leniently, and those who had suffered injury were not necessarily without
rancor. Those who do not renew themselves will again do evil; those victims who
are angered will turn from goodness. A single amnesty can turn men from the
good and encourage evil behavior. It is a great misfortune to good government. 65

This dislike of amnesties was one of the few points with which the great
reformer W a n g An-shih agreed wholeheartedly with his chief conserva-
tive opponent Ssu-ma K u a n g . 6 6 Despite all the opposition, however,
amnesties formed a necessary part of the Sung system of justice, because
the state lacked the capacity to handle its load of criminal cases and
convicted persons, and found it necessary periodically to lighten the
burden. T h e police systems overwhelmed the courts with accused crim-
inals, and the courts were sentencing more men than the penal system
could properly handle.
U p o n receipt of a letter of amnesty, the local jurisdictions were sup-
posed to send u p a document certifying that those who would benefited
had not committed further infractions of the law while in registration.
These reports also apparently indicated the seriousness of the original
offense. These documents went to the Ministry of Justice, which sent
down an order approving the convicts' transfer or release. 6 7 Because
grants of one sort or another came with such frequency and the judicial
process was time-consuming, it is probable that the only criminals who
ordinarily actually suffered their full sentences were those subject to
beatings. Disctict magistrates could on their own authority sentence
people to beatings with the heavy rod and could have the punishment
inflicted without the need for higher-level review. Therefore, such trials

64 SHY, hsing-fa 4.51a.

65 SS 201.5029. Such criticisms were frequent. See, for example, Ch'eng Pi, Ming-shui chi>
SKCSCP 1972, 2.3ia-33a; or Ch'en Shun-yu, Tu-kuan chi, SKCSCP ed., 7.2ib-23a.
66 Hung Mai, Jung-chai sui-pi,fifthcollect., p. 147.
67 SHY, hsing-fa 4.460-473; SS 201.5015.
490
 Modifications of penalties
could consume much less time than would trials for more serious crimes
in which the magistrate's recommendations had to be submitted to
higher authorities for review and confirmation. Clearly, this group of
amnesty policies changed the actual face of Sung justice, especially for
those subject to the heavier penalties.

Privileges
Another way in which penalties were reduced for large numbers of
people was through the elaborate system of privileged groups. 68 The
origins of privileges was traced by Sung writers to classical roots. In the
Rites of Chou is a list of eight categories of people with the right to
privileged treatment: those who were the relatives of the sovereign, those
who were old acquaintances of the sovereign, those who were of great
virtue, those who were of great ability, those who were meritorious,
those who were high officials, those who were exceptionally zealous in
performing government duties, and those who were the guests of the
sovereign (the descendants of the preceding imperial families). 69 In
practice, the only significant categories were relatives of the emperor,
holders of certain titles, and officials.
The system of legal privileges in use early in the Sung was virtually
identical with that in use in the T'ang. From the T'ang, the Sung had
inherited four distinct legal privileges, "deliberation" (i), "petition"
(ch'ingh), "reduction" (chien*), and "commutation" (shu). The highest of
these privileges, deliberation, carried the greatest benefits, covered the
most crimes, and was extended to the fewest people.
When a person with the privilege of deliberation committed a crime
that bore a sentence of exile (liu) or less, the penalty was automatically
reduced by one degree and could be commuted. For example, sentences
of exile were reduced to penal servitude and those of penal servitude to
beating, both being commutable to payments in copper. If anyone with
the privilege of deliberation committed a capital crime, the authorities
were required to submit a memorial to the emperor in which they
described the circumstances, indicated why the accused was entitled to
deliberation, and asked permission to consider the case. When they had
examined the case, they again had to memorialize the emperor asking
him to determine the sentence.
The only limitation on the crimes covered by deliberation was that it

68 For a more extended discussion of these privileges, see McKnight, "Song Legal
Privileges," pp. 85-106.
69 On the eight deliberations, see Johnson, The T'ang Code, p. 23; Ch'ii T'ung-tsu, Law and
Society in Traditional China (Paris: Mouton, 1965), pp. 177-79.

491
 Law and order in Sung China

could not be used for crimes falling in the category of the Ten Abomina-
tions, crimes that included rebellion, sedition, and treason, incest, lack of
filial piety, making the black magic poison called ku, some homicides, the
theft or counterfeiting of certain state goods, and a few other crimes
committed against superiors.
Petition, the second-highest privilege, was awarded to a somewhat
larger group and carried fewer benefits. Under it, all penalties of exile or
less were automatically reduced by one degree and could be commuted,
but in capital cases, after verifying the right of the accused to petition,
the officials tried the case and drew up a document recommending either
strangulation or decapitation. They then submitted a memorial stating
why the defendant was eligible for petition, giving the facts of the case,
and asking the emperor to endorse their suggested sentence. Petition
could not cover the Ten Abominations, other homicides, or the crimes of
officials who in their jurisdictions committed rape, robbery, kidnapping,
or accepted bribes to circumvent the law.
Reduction, the third privilege, was granted to a still larger group of
people but was far more limited. It could be used only for crimes calling
for exile or less. The punishments were to be reduced by one degree and
could be commuted. The lowest privilege, and the most widely distri-
buted, was the right to commute sentences of penal servitude or less into
payments in copper.
With the exception of this last, lowest privilege, these privileges were
granted only to those holding official or noble rank, their relatives, or
relatives of the imperial family. During the Sung the greatest privileges
were held by those with the greatest political power - functional officials
and close imperial relatives. The greater an individual's privileges were,
the further he could exend his protection to his relatives. The closer the
ties of kinship (as defined by the system of mourning grades) were, the
greater would be the potential benefits. 70
The implementation of these principles is clearly seen in the case of
imperial relatives. Obviously the emperor himself was the source of the
greatest degree and extent of privileges. All of his relatives to the Van-
mien degree of mourning (i.e., one degree beyond the regular mourning
circle) had the privilege of deliberation. The next most powerful figures,
the grand empress dowager and the empress dowager, could extend the
privilege of deliberation to their mourning relatives of the fifth degree
(ssu-ma), the lowest level of the regular mourning system. The wife of the
crown prince, by contrast, could extend the privilege of petition only to
her third-degree mourning relatives. Membership in the imperial clan,

70 See Figure 2.1, showing degrees of mourning.

492
 Modifications of penalties

no matter how distant, carried with it legal privileges. All clan members
had the right to commutation, and all female members could extend that
privilege (for crimes of exile or less) to their husbands and sons.
The other large group whose members had extensive legal privileges
in the Sung was the official class. In the early Sung Code, deliberation
was granted to active substantive officials of the third grade and up in
the nine-grade civil service system. Deliberation was also extended
to holders of certain honorific titles, whether or not they were active
officials. People who had prestige titles (san-kuan) (honorifics awarded to
certain people of high status or accomplishment) were granted delibera-
tion if the titles were of the second grade and up, as were those with first-
grade titles of nobility. (There were twelve grades of nobility in the Sung
that were given to some members of the imperial family and certain
distinguished officials.) Those granted this privilege could extend the
right of petition to all their one-year mourning relatives as well as
to great-grandparents, great-great-grandparents, great-grandsons, and
great-great grandsons in the main line of descent, and to paternal aunts
and paternal uncles and their wives.
Petition was granted to all active substantive officials in the fourth and
fifth grades, to those holding prestige titles in the third through fifth
grades, and to men with dignities or titles of nobility in the third through
fifth grades. Such men could extend reduction to paternal grandparents,
parents, brothers, sisters, wives, and sons and grandsons in the male
line.
Reduction was granted to all active substantive officials in the sixth
and seventh grades and to those with prestige titles and other sorts of
honorifics including "dignities" (hstin) or guard offices of the sixth and
seventh grades. All those with the right to reduction could protect their
paternal grandparents, parents, wives, sons, and grandsons in the male
line by extending the privilege of commutation.
In the T'ang and Sung the right to commutation was granted to all
active substantive officials from the bottom (ninth) grade and up. It was
thus the only one of these legal privileges shared by all officials.
Officials also benefited from policies that allowed them to substitute
for other punishments the deprivation of office, loss of official titles, or
surrender of honors. In the most serious cases those convicted would be
stripped of all the titles, honors, and ranks they had accumulated since
entering the civil service and would be returned to their original status.
This penalty, called "disenrollment," was considered equivalent to three
years of penal servitude. Those so punished might reenter the civil
service after a number of years and at a much lower rank. The second
form of official commutation, "resignation from occupied office," was
493
 Law and order in Sung China
much less harsh than disenrollment. An official so punished surrendered
his current active appointment, but not his other honors or dignities, and
he could resume office after a shorter period. Still less serious was
"replacement," under which the official might trade in his official titles
and honors without actually having to leave active official life. For
example, one office of the fifth rank or above could replace two years of
penal servitude for "private infractions," offenses committed with selfish
or corrupt intent; or one office of the ninth rank could replace one year of
penal servitude.
During the Sung the granting of privileges for officials was increased
by the simple device of granting to lower official grades the privileges of
a higher grade. In the late Sung code, the Ch'ing-yuan t'iao-fa shih-lei,
there is a rule saying that fourth-grade officials were to be granted the
legal privileges of third-grade officials, that is, deliberation, and sixth-
grade officials the privileges of fifth-grade officials, petition. Eighth-grade
officials were to have the privileges of seventh-grade officials, reduction.
However, although officials were in most ways privileged, they also
were liable for more severe penalties than ordinary people were for
certain sorts of crimes. According to the early Sung Code, an official who
committed robbery in his own jurisdiction was liable for a punishment
two degrees more severe than that due an ordinary person guilty of the
same crime. 71 Furthermore, many crimes were excluded from coverage,
and under some sets of defined circumstances, even crimes ordinarily
covered might be excluded.
The privileges of relatives or household members of officials also
multiplied during the Sung. In the late Sung, wives of the sons and
grandsons of officials in the eighth rank or higher were to receive the
same privileges as their husbands. Even female slaves might be protected
under some circumstances. Finally, there is one passage indicating
that girls betrothed to officials were covered by the privileges of their
intended husbands.
These groups enjoyed not only substantive privileges that reduced
their penalties but also privileges of procedure. Application of judicial
torture to officials was limited by law (though this was sometimes
ignored). Furthermore, some privileged persons were excused from hav-
ing to appear in person in court. Some Sung officials believed that this
exemption from personal appearance in court was a practice dating from
ancient times. 72
Sung legal privileges were closely related to the privilege of officials

71 SHT 19.304.
72 See, for example, Su Sung, Su-wei hung wen-chi, 17.14b.

494
 Modifications of penalties

called "protection" (yin). In its best-known form, protection allowed

many Sung officials to petition that certain designated assignees, usually
but not always sons or grandsons, be granted official rank. The device
served as a type of retirement benefit for the officials and helped cement
commitment to the dynasty. 73 The other side of this coin was the ability
of officials' relatives to escape punishment through the "protection" of
these officials. The book of miscellaneous notes called the Sung-ch'aoyen-i
i-mou lu reports:

At the beginning of the dynasty, all clerical workers were the sons and younger
brothers of literati who had been unable to establish themselves as civil service
officials. They felt ashamed. When guilty of crimes, they could use protection to
commute their sentence. This they depended on to do evil. In 1029 a clerk in the
Finance Office named Wei Shih-an committed a crime. He used the protection of
his grandfather, who had been a district magistrate. There was a decree that he
was especially to be punished. It was then decreed that from that point on clerks
were not to benefit from protection.74

Commutation

W e should n o t overestimate the n u m b e r s w h o could benefit from the

higher privileges. D u r i n g the Sung the great majority of those with
official status were in t h e eighth o r ninth grade. T h u s , d u r i n g m u c h of
the dynasty they would have h a d only the right to c o m m u t a t i o n . M a n y
of these m e n would not have accumulated a n y great n u m b e r of honors or
dignities that could be surrendered to defray penalties. T h e privilege of
c o m m u t a t i o n w a s gradually extended to other people w h o were per-
ipherally a t t a c h e d to t h e official class, including " t e m p o r a r y officials,"
some "irregular officials" (she-kuan, from i o o o ) , 7 5 a n d " s u p p l e m e n t a r y
officials" (pu-kuan, from 1139). Later in the Sung these lower privileges
were extended a t times to m e n having certain honorific military titles,
g r a d e d officials serving "outside the career" (i.e., not serving in posts
considered in t h e regular civil service), m e n with prestige titles even if
they d i d not have a service record, m e n w h o h a d gained official rank for
m a k i n g contributions in times of crisis or whose relatives h a d died
fighting for the Sung, a n d possibly even some m e n in clerical positions.
A decree of 1009 reportedly extended to all examination participants the

73 For an extensive discussion ofyin, see Lo, An Introduction, pp. 102 ff.
74 Wang Yung, Sung-ch'aoyen-i i-mou lu, 3.7a. See also an order dated 1046 forbidding the
use ofyin by relatives of officials if they had been found guilty of repeated crimes (HCP
159 4b).
75 Acting officials were men selected according to special examinations held in the
southernmost section of the Sung empire (Kuang-nan) for service there as local
functionaries (see Lo, An Introduction, p. 96).
495
 Law and order in Sung China

right to c o m m u t e penalties for crimes calling for beating with the heavy
rod. 7 6 Students in state institutions and examination candidates were
also allowed to c o m m u t e some crimes, with those in the more prestigious
schools being able to commute the penalties for more serious crimes.
Wealth could also purchase privileges. T h e privilege of commutation
also spread because the official credentials that carried with them a
degree of immunity were sold by the state. People could purchase these
official rank credentials by making contributions in times of need. T h e
titles they received would rarely qualify them for appointment to actual
office in the graded civil service, but they could secure limited legal
privileges. In 1127, during the military (and fiscal) crisis of the Chin
invasions, certain titles were sold for two thousand cash each. Holders of
these could c o m m u t e crimes calling for beating with the heavy rod, as
long as there were no injuries or robbery involved. 7 7
Finally, there is scattered evidence that at some times during the
Sung, various sorts of people, including commoners, were allowed to
c o m m u t e certain penalties into fines. According to a general rule in the
T ' a n g and Sung codes, those guilty of injuring or killing people by
accident were allowed to pay redemption. 7 8 In practice, officials in the
early Sung seem to have been extending the right to commute penalties
rather freely. A decree of 999 stated:

The senior officials in the prefectures have been indulging their personal feelings
in assessing the seriousness of crimes, by lightly punishing commoners or letting
them commute their penalties into payments in gold. From now on this is not to
be permitted. If wives are liable for penalties of the heavy rod or less, they may
have the seriousness of their penalties assessed and be released after paying
copper in commutation.

T h e Sung History adds that " E m p e r o r J e n - t s u n g [r. 1022-63] was deeply

concerned about people's ignorance and wanted to establish a law on
commutation with which to respond by minor penalties." T h e passage
continues with the text of a decree:
The Former Kings used light laws and simplified arrangements so that men
would understand the prohibitions and could easily follow the law. Later
dynasties established the prohibitions with regard to tea, liquor, and the salt
taxes, contending against the people for the profits, and so the punishments
proliferated. Now we have the compilations of edicts, which all are in addition to
the statutes. The number of such rules changes frequently. The officials do not
know them; how can the people do so? Once someone becomes enmeshed, even if

76 Yii Te-lin, P'ei-wei chai wen-chi (Taipei: Commercial Press, 1972), 19.5a.
77 HNYL 363.
78 SHT 23.358.

496
 Modifications of penalties

his circumstances are extenuating, the law does not allow him to commute
the penalties. How can we not seek to transform him through rites and music
and instead apply penalties and fines? The Han emperor Wen caused men
to carry grain to the borders as a way of receiving honor and of avoiding
punishments. . . . Some men imprudently seek profits and so violate the laws, but
others are pitiable, and so we should separately establish a law on commutation.
Villagers may use grain; city people may use money.
The suggestion was discussed by the officials, but it was attacked
because it meant, in the words of the historian, that the rich could
commute penalties, whereas the poor people would be punished. Even-
tually the idea was shelved, 79 but a passage from 1071 says that the
officials and people could commute penalties of the heavy rod "if the
circumstances were extenuating." 80 Eighty years later a censor remarked
to Emperor Kao-tsung:
"In antiquity, contributions in gold were used as a form of penalty. The Former
Kings had hearts that could not bear to see the suffering of others. When the
people knowingly transgressed, they could offer gold in requital of their guilt.
In later times this was put into the law code. There were clauses with regard
to fines in copper, from 1 pound to 120 pounds. Your Majesty's humanity
has revived the world, seeing the suffering of the people you wish to help....
Recently the prefectures and districts have aided those guilty of crimes calling for
beating with the heavy rod or penal servitude. They are ordered to pay money to
avoid the punishment. At the least they pay several tens or hundreds of strings.
At most they may pay 1000 to 2000 strings. Judged by the statutory rules, these
fines are excessive relative to the guilt. . .." The emperor said, "From now on we
should have a regular law system for this."81
But this same emperor had, a few years earlier, also expressed his
dismay at a legal system in which the wealthy were more and more often
able to escape penalties that the poor had to suffer.82 This problem, a
class division in punishments, was not confined to the Southern Sung. A
late Northern Sung official complained that when cases were reviewed
and the penalties were increased beyond the normal, those affected "are
often poor people. Such increases rarely affect the rich" because the rich

79 SS 201.5025-26. In 1070 a reform suggestion was submitted that resembled the

practice described for the Han. People could, in effect, gain exemptions from future
penalties by means of their current good behavior. Under the suggested (but not
adopted) system, officials were to issue certificates for especially upright and moral
behavior. If later the holders became involved in a minor crime, they could use their
certificates to commute the penalties (see WHTK 167.1448).
80 HCP 227.12b.
81 HNYL2604.
82 Liu Cheng, Huang-Sung chung-hsing, 25.12b.

497
 Law and order in Sung China

were able to bribe the clerks to present their cases in the most appealing
manner.
The privileges of the elite were further buttressed by stipulating espe-
cially heavy penalties for those who had the temerity to attack them.
Crimes against status superiors were punished more heavily than were
those against equals or inferiors. In the T'ang and early Sung codes, the
penalty for ordinary assault was forty blows of the light rod, but an
assault on an imperial relative of the t'an-mien degree of mourning (i.e.,
one degree beyond the ordinary outer limits of the mourning circle) was
punishable by one year of penal servitude. If injury resulted, the penalty
jumped to two years of penal servitude, versus sixty blows of the heavy
rod in the case of ordinary persons. There are many other specific rules
in the codes. 84

Indemnity payments and fines

T h e Sung system also allowed a practice that was reminiscent in some

ways of the European practice of indemnity payments to wronged parties.
According to a rule found in the early thirteenth-century Sung code,
when a person with the right to commute penalties committed battery,
the commutation money was to be paid to the family of the victim. 8 5
Fines in lieu of other penalties were also used in dealing with M a n
tribesmen a n d in 1090 were extended to the non-Chinese tribal peoples
in the border areas of L u Prefecture in Tzu-chou Circuit. T h e fines
ranged from three strings of cash for crimes calling for beatings with the
light rod to thirty strings of cash for capital crimes. T h e monies in these
cases went to the authorities, not to the families of the victims. 8 6 T h e
sources d o not indicate whether the Chinese authorities were merely
continuing traditional tribal practice or introducing a new system. How-
ever, given the widespread use of such payments, from the northern
peoples of early Europe, to the desert Arabs, to a n u m b e r of African
societies, it seems probable that the Chinese were simply sanctioning a
system of settling disputes that h a d been found workable by the non-
Chinese groups before they fell under Chinese domination. T h e Sung
writers generally looked down on these systems as if they in some sense
violated propriety b u t acceded to them as conventional among inferior
groups.
Although the details of the commutation system for H a n Chinese

83 SHY, hsing-fa 6.60b.

84 SHT2i.324ff.
85 TFSL 500.
86 HCP453.i3a-b.
498
 Modifications of penalties

offenders are not wholly clear, it does seem that the Sung government, in
this roundabout way, had created a system of fines as penalties, albeit
disguising them as commutations. The nominal penalties for infractions
were given in terms of the traditional Five Punishments, which did not
include fines. The widespread use of a commutation system meant that
in practice fines were a part of the penal system. Because these "fines"
were brought into the system as variations from what was seen as
standard practice and because they were used largely for minor crimes
that would not come to the attention of the higher authorities, they are
seldom discussed in the extant sources. The authorities were perhaps
uncomfortable with practices that deviated from the standard and came
close to the kinds of amends permitted to non-Han tribal groups.

Reductions because of family or personal circumstances

Sung law, like the law of earlier and later dynasties, also recognized as
just and appropriate the reduction of penalties in cases in which the
accused criminals were considered to be incapable or when the punish-
ment would work undue hardship on others. People were considered
incapable by virtue of age, illness, or other handicaps. Under the T'ang
and Sung codes, people who were aged or very young or suffered from
certain diseases or disabilities were either excused from penalties or had
their penalties reduced.
The early Sung Code stated:
If all those who are seventy years old or older or fifteen or younger, or critically
ill or disabled have been sentenced to exile or less, they will be permitted to
commute their penalties. If those who are eighty years old or older or ten years
old or younger, or those critically ill or disabled are guilty of rebellion or
homicide, in all cases when the penalty would be death, they may exercise the
privilege of petition... . When such people are guilty of robbery or batteries, they
may commute their penalties. . . . All other crimes are not to be adjudicated.
Those ninety or over or seven or younger, even if guilty of capital
crimes, were not to be punished. 87 For the same reason, members of
certain groups were also exempted from wearing fetters.
The commentary here traces these privileges back to the classical work
known as the Rites of Chou, which in the "Three Amnesties" section
speaks of granting lenient treatment to the young and weak, the aged,
and the incapable. The residual punishments to which these groups were
still subjected for the more serious crimes are explained as being neces-
sary because their crimes had necessarily created bitterness among
87 SHT 4.56ff.See also SHY, hsing-fa 3.10b.
499
 Law and order in Sung China

the victims and the victims' associates, bitterness that needed to be

assuaged. 88
Emperors also at times intervened to lighten the penalties of persons
whom they felt deserved special pity because of their age. During the
Ch'ing-li period (1041-49), a nine-year-old boy had killed someone in a
fight and had been declared liable for execution. Emperor Jen-tsung
said, "He was only a child fighting. He had no intention of killing
someone. Money should be paid to the family as a redemption." 89
Of course, those who deliberately tried to use the privileges they
derived from their condition might find themselves being punished. In
1025 t n e authorities in the capital prefecture reported to the emperor the
following:
The commoner Ch'en Wen-cheng and his wife Ah-tsung have been tried for
leading into prostitution the wife of a soldier in the Tiger Wing army. According
to the law, they should be liable for one and a half years of penal servitude. Both
husband and wife are blind and are eligible therefore to be forgiven, but Wen-
cheng trusted to his blindness to protect him and did evil. We ask that he be sent
to an outer prefecture for "registered control." From now on those who presume
upon age or illness and so deliberately commit wrongs when the circumstances
are difficult to forgive should have their crimes investigated and await an im-
perial order that they be transported to outer prefectures for registered control.
The emperor approved the suggestion of the prefectural government. 90

Religious and filial privileges

Buddhist and Taoist monks and nuns were also treated in a distinctive
way by the law. Under some conditions their religious vocation might
result in an increase in the ordinary penalty. For example, adultery
might result in heavier penalties than normal. In other cases the regular
law was applied only after the culprits had been laicized. However,
for relatively minor crimes such people might be punished by laicization
alone, thus escaping from the penalties inflicted on ordinary commoners. 91
This was one of the benefits of owning a certificate of ordination, certifi-
cates that were sometimes sold by the government to raise revenues. 92
Even if the criminal himself could not be excused on these grounds,
88 SHT 4.56. For the situation under the Ch'ing, see Derk Bodde, "Age, Youth, and
Infirmity in the Law of Ch'ing China," in Cohen et al., eds., Essays on China's Legal
Tradition, pp. 137-69.
89 WHTK 170.1475.
90 SHY, hsing-fa 4.12b.
91 SHT 26.424. See also SHY, hsing-fa 1.19b-20a; TFSL 515.
92 See Kenneth Ch'en, "The Sale of Monk Certificates," Harvard Theological Review 49
(October 1956): 307-27.
500
 Modifications of penalties

he might be spared if carrying out the penalty would result in undue

hardship for his parents or grandparents. According to the early Sung
Code, which was repeating the rule from the T'ang Code, "Petitions
may be submitted in all cases involving those who commit capital crimes
(other than one of the Ten Abominations) if the paternal grandparents
or parents are aged or infirm and require service and the family has no
adult relative within the second degree of mourning." Approval of such
petitions meant that these convicted persons were permitted to remain at
home to care for their elderly relatives. When the elderly persons died or
another relative of suitable mourning relationship came of age, another
petition would be submitted to the authorities so that they could decide
what to do with the original criminal. 93 The coverage of this rule is
restricted somewhat in the early thirteenth-century Sung code, which
specifies other crimes not covered by the rule:
Memorialize the cases of all those adults guilty of capital crimes (with the
exception of the crimes of the Ten Abominations, armed forceful robbery, and
premeditated murder or deliberate homicide when the victim has actually died),
whose elderly or ill grandparents and parents have no other adult mourning
degree relatives and so ought to be cared for.
This rule also might be applied to women: "Women who under the law
should be punished with registered control should be spared registration
when their husbands' grandparents or parents or their own parents or
grandparents are old or ill and need someone to care for them but lack
any other adult mourning-degree relatives." 94

Voluntary surrender and confession

Because the purpose of punishment is, among other things, to contribute
to the criminal's salvation by leading him to "renew himself," whatever
will contribute to that end may find a place in the penal system. T h e
Chinese early recognized that the first step on the road to renewal must
be the criminal's willingness to recognize and admit to others what he
has done. H e must begin by confessing.
Chinese legal codes allowed criminals to reduce their liability for
punishment by voluntarily surrendering and confessing. 95 In the words
of a Southern Sung official, "With regard to legal cases, the legal system
has stipulations concerning confessing, in order to make known to the
93 SHT 3.48. I am using the translation of the T'ang rule from Johnson, The T'ang Code, p.
152.
94 TFSL 527.
95 For confessions, see George Kennedy, Die Rolle des Gestandnisses im chinesischen Gesetz
(Berlin: n.p., 1939).
501
 Law and order in Sung China

people the road of their self-renewal" (and, he adds, to make it easier for
the judges to get the facts). 9 6 T h u s it was general practice to allow most
deserters from the armies to return without penalties within time limits,
usually two months from the day they deserted. This rule did not,
however, cover those who had committed armed robbery or rape, or
certain other sorts of banditry. 9 7 And an edict of 1042 made it possible
for bandits who had not yet killed anyone or who had not been the
original instigators of their crimes to return to their old occupations
within one hundred days without fear of prosecution. 9 8
T h e confession policy was one of the key legal issues in the most
celebrated case in the whole Sung dynasty, that of Ah Yiin. Ah Yiin was
a woman from T e n g Prefecture in Ching-tung East Circuit. While her
mother was still in mourning, Ah Yiin was betrothed to a man she
despised because he was ugly. She plotted to kill him but failed in
her attempt. When the investigation began, she voluntarily confessed.
T h e central government offices felt that the young woman should be
executed. They reasoned that betrothal was tantamount to marriage in
its legal effects. Therefore, Ah Yiin had attempted to kill her husband.
Those who disagreed with this legal stand, including the prefect who had
originally tried the case, questioned the reasoning of the central organs
and continued to affirm the applicability of the rules on confession. T h e
case became caught up in the bitter political struggle between Wang An-
shih and the conservatives and dragged on for a considerable period
despite repeated attempts by the emperor to impose a solution by decree.
In the end the opponents of execution won out, but only after bitter
wrangling. 9 9
As the edict makes clear, the policy of forgiveness was not open to
every criminal. It was especially intended for men who were "intent on
rectifying their faults and renewing themselves." 1 0 0 In 1085, officials
from the Department of State Affairs explained:

The statute allowing the reduction of the sentences of those who surrender and
confess refers to crimes that are not especially horrible and to criminals who wish
to reform. In regard to rapists and robbers, it is hard to classify them with other
crimes as deserving reduction on the basis of precedents. We ask that reductions
not be granted to robbers who have killed, rapists, robbers who having been
spared [again commit crimes], and armed robbers in groups of three or more

96 HNYL 2667.
97 HCP 398.12a-13b.
98 HCP 115.1a, 12399.25b, 355.2b.
99 SS 201.5006; WHTK 170.1475. See also Han Wei, Nan-yang chi, SKCSCP 1971,
26.ia-8a; HCP 3.15b.
100 HCP 398.12a-13b.

502
 Modifications of penalties
who know they are to be accused or who are already being investigated and
therefore confess.101
Psychological insight and common sense about the difficulties of effec-
tive investigation under premodern conditions came together to support
the continuing use of this practice for other lesser sorts of crimes. This
policy thus served several interrelated functions, encouraging rehabilita-
tion, easing the job of investigation, and, according to one observer,
helping break up gangs by sowing distrust among their members. 102

Conclusion
Like the leaders of all polities, the Sung authorities sometimes had to
choose between justice and mercy. The demands for justice, for a fit
between actions and responses, between crimes and punishments, were
rooted in fundamental Chinese ideas about communication and society.
Acts are messages and actors respond to them. Like actors on a stage the
men and women in traditional China had socially sanctioned roles to
play. If all actors played their roles properly and so sent the appropriate
messages to the right parties in the correct way at the proper times, then
the world would run well.
However, like actors on any stage the actors in the drama of tradi-
tional Chinese life could fail to perform their roles correctly, instead
acting willfully and for their own purposes or because others had failed
to train them properly. Penalties were necessary when actors, for what-
ever reason, did not fulfill their roles properly.
Mercy also had deep roots in the Chinese tradition. If those with
power could be certain that a given punishment was the appropriate
response to a given action, then of course they should apply it. However,
if there were any doubt, then the rulers were advised to err on the side of
mercy. Excessive penalties sent very unsettling messages to society at
large. Mercy also sent a message, because it gave the receiver both the
opportunity and the motivation to reform, and at the same time revealed
the power of the emperor to forgive.
The Sung penal system at all points reflects the interplay of these two
goals. Certainly the system was aimed at preserving the existing political
and social order. Those in power could and did use force and violence to
defend their prerogatives. Like ruling elites elsewhere they would no
doubt have looked on such actions as defenses of justice. If justice exists
when all persons fulfill properly the roles and functions appropriate to

101 SS 201.5011. 102 HCP 124.30-40.

503
 Law and order in Sung China
their condition and receive in return their proper deserts, then actions in
punishing those who refuse to fulfill their roles is just. The Sung elite
never doubted that society was by nature hierarchical, nor that the
imperial system of government was in some way attuned to the order of
the natural. Defending this hierarchy and the political system associated
with it was for them defending a natural system. Groups and individuals
who were oppressed by this system may have harbored doubts about the
meaning of justice, but few Chinese officials did. Each person had to act
out the role appropriate to his or her place in life.
This belief also emphasizes the importance of appropriateness and
measure. The punitive sanctions applied to actions that threaten the
given social and political order must themselves be appropriate. Since
rules are general, but situations are particular, it was clearly necessary to
provide some means for adjusting penalties to situations. The Sung
Chinese supplemented their system of penalties with elaborate policies
for modifying punishments, making them heavier or lighter.
The Sung officials called for penalties to be increased beyond the
normal degree of severity for two reasons, because the circumstances of
the crimes were especially outrageous or because maintaining security in
the area where the crime had occurred was especially important. From a
broader perspective the two traits coalesce. Acts were deviant because
they threatened the social or political order. Particularly heinous acts,
abuse of the old, killing the helpless, and so on clearly called for strong
responses. In especially sensitive regions, criminal acts were especially
dangerous.
The Sung authorities were even more concerned about reducing
punishments. The ideological justifications for such modifications of
penalties are spelled out in the earliest Chinese works on politics.
These ideological presuppositions interacted with the practicalities of
politics to shape the character and even the frequency with which mercy
might be extended. The amnesty system, which reached its height dur-
ing the Sung, was one of the most important means by which penalties
were reduced in traditional China. Amnesties were justified as imperial
acts of grace that gave people the opportunity to reform themselves and
demonstrated the power and the benevolence of the emperor. Though we
may be unduly cynical, it is also important to recognize that the police
system was putting so many suspects into the judicial system that it
became overburdened and that the judicial system was sending to the
penal system more men than it could handle. Periodically the numbers
needed to be lowered. Such reductions increased the convicts' labor
value by creating a turnover, and by holding out the possibility of

504
 Modifications of penalties
potential release at any moment might help keep men serving time from
becoming disillusioned and discouraged.
Most of those persons who benefited from these general amnesties
were probably men of low social status; most of those who benefited from
the system of privileges were people of high status and great power.
Some privileges were extended to the less powerful - women, the aged,
the disabled - but the most important privileges were extended to those
connected to the imperial family or to high officials. Privileges in China
or elsewhere are reflections of power. What was unusual in the tradi-
tional Chinese states was the bureaucratic and legally regularized way in
which these privileges were extended. The Sung is especially noteworthy
for the degree to which it tied the granting of privileges to substantive
officeholding. The Sung was the heyday of the bureaucrat.
The Sung was also a time when money seems to have talked with
special effectiveness. The wealthy have always been to some degree more
equal than others, but in traditional China their influence was over-
shadowed by the power of the bureaucrats. Still, even though the Sung
was in some ways the high point in the prestige and status of the
bureaucracy, it was also a time when the rich, at least for minor crimes,
could openly and legally trade money for exemption from penalties.
They might buy the ordination certificate of a monk, which could be
surrendered to spare the holder minor penalties. They might make
contributions to the government in times of emergency and so purchase
a nominal official rank that could be used to commute penalties. And
finally, they might also in some periods simply pay commutation fines.
Of course, as in all times and places, the wealthy could use their money
in illegal but often successful attempts to pervert the course of justice.
What is remarkable about the Sung is not the presence of corruption but
the number of ways in which money could be openly and legally used to
gain privileges.
The role of wealth per se should not be overs tressed. The circle of the
wealthy overlapped the circle of the bureaucrats. Many bureaucrats
were among the wealthy, and wealthy or not, their power was very great.
This bureaucratic elite believed that the existing social order, though far
from perfect in fact, was basically in accord with the order of nature.
Defending the existing order - including, of course their place within it -
was an act of justice. Such an attitude was self-serving, but it was also
tied to some of the elite's fundamental understandings about the nature
of man and the world. These views also led to a continuing emphasis on
the search for justice and the granting of mercy. It was important that
the responses to wrongdoing be fitting. It was proper to show kindness to

505
 Law and order in Sung China
those who after all also were people and so by definition were usually
capable of renewal. We should not be surprised that the system failed at
times or that it could be brutal. We should recognize the commitment of
many officials to trying to improve the system and the optimistic view of
the possibility of reform, which is one of the most striking and attractive
aspects of the Chinese judicial system of that era.

506
 15
Conclusion

The inevitable tensions between the incommensurable goals and desires

of different groups and individuals is a part of the dynamic core of any
society. These tensions work themselves out in a context of geography,
technology, economy, and tradition that shape the choices that the
groups and individuals in the society think that they have. Disparities of
values and circumstances, of patterns of action and experience, and of
expectations and understandings are sources of social energy; the institu-
tions, practices, beliefs, and patterns of living that emerge from the
interaction of all these factors are determined by both the actions of the
people involved and the world - physical, mental, cultural, and social -
within which they live and act. In successful societies these tensions
produce a healthy balance of interests, which makes it possible for many
members of the society to lead reasonably secure and fulfilling lives.
Such social tensions play this central role even in the simplest societies;
in states with a ruling elite, a new dimension is added to the interplay of
social tensions. While recognizing that their views and visions are not
shared by all others in the society, ruling elites are usually capable of
identifying their own views with the "proper" views. By so doing, they
define other views as deviant. These elites have the power to define
certain violations as crimes, to which they can legitimately respond with
force. Ruling elites can identify their own values with proper values
because in fact on many points they are in agreement with large seg-
ments of the societies they control. Some rules may be clearly concerned
with buttressing the power position of the elite, but many others embody
widely shared values. In enforcing compliance with rules framed to
support these shared views, the rulers are acting not just in their own
interest but in the interest of a much larger group. By enforcing these
rules, which often embody traditional values in legal form, the elite
generates support among other parts of the larger society.
In complex societies, special agencies are created to enforce such rules,
507
 Law and order in Sung China

which frequently are written down; special individuals exist to determine

whether and how the rules have been violated; and special institu-
tions are established to deal with offenders. The patterns of law, law
enforcement, adjudication, and punishment that emerge are conditioned
by both contemporary surrounding circumstances and inherited condi-
tions, understandings, and traditions.
Traditional Chinese culture from the founding of the empire in the
second century B.C. until the twentieth century was a conspicuous
example of a culture within which the tensions among values and pat-
terns of behavior contributed to a relatively stable pattern of develop-
ment. Certainly there were major changes during these two millennia in
the ways that people lived their lives and the things in which they
believed, yet these changes were far less radical than those occurring
over the same span of centuries in Europe. Throughout these centuries
in China, there was a sense of continuity and order that was missing in
Europe. This orderliness was certainly one of the attractions of the
Chinese way to the cultures that bordered China.
Certain human groups have been especially successful in creating
a semblance of order. They have been successful in the sense that
the societies based on their solutions to universal problems have been
capable of incorporating increasing numbers of individuals within their
cultures and successful in that their conceptions have influenced groups
far beyond their area of political control. Historically perhaps most
successful in encompassing a great number of people while displaying
remarkable continuity was Chinese traditional civilization. From its
beginnings in the river valleys of north China, the pattern of its charac-
teristic cultural discourse spread to cover much of eastern Eurasia.
From the time in the sixth century B.C. when the Chinese began to
reflect on and analyze their culture and society, they have appreciated
the importance of socialization and communication and recognized the
ways in which these can be affected by external conditions. The Classical
Confucian thinkers - who more than any other group were responsible
for shaping an ethos that dominated the thinking of the Chinese ruling
elite - recognized the fundamental role of primary-level socialization.
They also recognized the fundamental role of language communication
in transmitting culture among individuals and across generations. They
focused on reducing communication problems through proper social
training, especially by example. Confucians coupled their stress on the
role of socialization with the recognition that success would always be
only partial.
Transformation through teaching always had to be accompanied by
mechanisms for sanctioning and deterring those not properly trans-
508
 Conclusion

formed and taught. Education was always coupled with punishments.

With their aesthetic rather than logical sense of order, the Confucians
stressed the importance of harmony and balance and the role of models
and examplars but still acknowledged the inevitability of some deviance
and the necessity of using lawful force. Another group of Chinese thinkers
were even more concerned with these problems: The Legalists, with their
penchant for universalistic principles and for generalized and not par-
ticularistic ethics and justice, concentrated on the uses of law and force
in governance.
The opposition between these two approaches can be neatly phrased.
In practice and in the minds of many traditional Chinese, the tendencies
and concepts espoused by these rival strains of thought were thoroughly
mixed. During the Warring States period (403-221), strands of thought
from the various schools were shared in various ways. Even among those
ordinarily classified as "Confucian," we find an admixture of other
views. This eclecticism in practice begins early and remains charac-
teristic of the Chinese literate elite under the empire.
These legal beliefs were based only in part on the more general tenets
of the philosophers. They also drew on these thinkers' knowledge of the
existing systems of laws and legal institutions. By the founding of the
empire in the third century B.C. there already were elaborate codes of
laws, penalty practices, judicial systems, a stock of widely shared views
about the origins and nature of criminal behavior, and a repertoire of
responses thought to be effective, appropriate, and feasible. The history
of the evolution of the legal, law-enforcement, and penal systems that
built on this Classical heritage has yet to be written. Notable works have
already appeared that deal with the usages of the Ch'in-Han and the
T'ang, but because of the limited sources available for these periods,
these studies have had to leave many questions unanswered.
The Chinese under the empire themselves thought of their legal his-
tory as evolving from an era of simplicity to an era of excessive for-
malism. For the elite of the Sung, the golden age in law was the three
dynasties of Hsia, Shang, and Chou, but they also at times look back to
the T'ang dynasty as a more recent example of proper (i.e., lenient) law.
Their own dynasty they saw as an improvement on the brutality charac-
teristic of the Five Dynasties. 1 For a modern historian a knowledge of the
Sung system provides one benchmark against which to measure change
from a time for which there is a wide variety of sources extant. Such a
study can show how the responses to universal problems were shaped by

1 See, for example, Yin Shu, Ho-nan hsien-sheng wen-chi, 2.5b.

509
 Law and order in Sung China

specifically Chinese cultural elements and by particularities of the Sung

situation and experience.
The general problem facing the Chinese of Sung times was clear:
Deviance and crime persisted. People's conceptions of them were based
on both inherited wisdom and current experience. Sung authors at times
did differ concerning the origins and nature of the crimes of special
concern, but among most writers there was a basic consensus. Most
crime stemmed from the hardships suffered by the people and was
exacerbated by the abuses of officials. The Southern Sung official Liu
Yueh observed, "In recent years, the people being impoverished, bandits
have arisen. All have been provoked by abusive officials."2
Given this conception of crime, certain responses were predictable.
The best answer would have been to reduce the underlying causes of
crime, employ good officials, and thereby wean the people away from
bad behavior. Liu Yueh went on to say that since he had taken the office
of sheriff of Min District, he had tried to resolve the crime problem by
forbidding the clerical personnel from entering the countryside where
they had been abusing the people, by not allowing the locally responsible
mutual security agents to hire substitutes, and by warning his colleagues
to refrain from such abuses as using excessive punishments, being overly
harsh in collecting taxes, accepting bribes, and using their positions to
make money. 3
Some Sung officials even speak, perhaps hyperbolically, as if all ban-
ditry could be eliminated. The official Liu Chang wrote:
The world was suffering from bandits. Someone asked, "Can bandits be elim-
inated?" I answered, "How could it be that bandits could not be eliminated?
Those who are willing to become bandits do so for reasons. If you simply get rid
of the causes, how could we then suffer from banditry?" They asked, "What are
the causes of banditry?" I answered, "The inadequacy of food and clothing
causes banditry. Government exactions being inequitable causes banditry. When
we cannot effect transformation through education this causes banditry."
The consensus answer was clear. Again, Liu Chang:
In good times the reason that bandits are lacking is sufficiency. In times of good
government the reason for the lack of banditry is equity. In times when educa-
tional transformation is effective, the reason for the lack of banditry is harmony.
If you do not attend to providing sufficient daily necessities but attend to
suppressing bandits, this is like damming up water without stopping the source.
If you do not attend to the educational transformation of bandits but seek just

2 Liu Yueh, Yiin chuang chi, 17.14b.

3 Ibid., 17.14b.
510
 Conclusion

to suppress them, this is like tolerating the blaze while seeking to promote
harmony.4
Educational transformation was thought to be effective because its
effects persisted over time. It was seen as inexpensive because its success
cut the cost of the law-enforcement and penal systems. When such an
educational strategy worked, there was no obvious conflict between the
desirable and the feasible. Some Sung officials speak as if educational
transformation alone might end disorder. The Northern Sung official
Gh'in Kuan identified educational transformation as the sole means of
control used under the three great dynasties of Hsia, Shang, and Chou,
and punishments as the sole means of control of the hated Ch'in empire.
The Han and later dynasties tried to use a mixture of the two, but the
disorders continued. Some men, he believed, argue that the present day
and antiquity are fundamentally different, so that wishing to depend
only on educational transformation without using penalties is fatuous.
Such men, he observed, are "rotten Confucians who do not understand." 5
Despite such pleas, however, in the real world, educational trans-
formation clearly failed in many cases and crimes continued. Indeed, it
seems probable that the crime rate increased during the Sung. Wilson
and Herrnstein suggest that three main factors are responsible for long-
term shifts in crime rates:
First, shifts in the age structure of the population will increase or decrease the
proportion of persons — young males — in the population who are likely to be
temperamentally aggressive and to have short time horizons. Second, changes in
the benefits of crimes (the accessibility, density, and value of criminal oppor-
tunities) and in the costs of crime (the risk of punishment and the cost of being
both out of school and out of work) will change the rate at which crimes occur,
especially property crimes. . . . Third, broad social and cultural changes in the
level and intensity of society's investment... in inculcating an internalized com-
mitment to self-control will affect the extent to which individuals at risk are
willing to postpone gratification, accept as equitable the outcomes of others, and
conform to rules.6
We do not have sufficiently accurate demographic statistics to deter-
mine whether there was a change in the age structure during the Sung,
nor at present do we know enough about changes in Chinese social
investment in the socialization of people into appropriate roles. We can,
however, assume that the demographic changes associated with increas-
ing urbanism (characterized by both the growing number of large cities

4 Liu Chang, Kung-shih chi, 40.474.

5 Ch'in Kuan, Huai-hai chi, 14.2b-4b.
6 Wilson and Herrnstein, Crime and Human Nature, p. 437.
511
 Law and order in Sung China
and the vastly more numerous market towns) may have increased the
apparent benefits of crime.
How were ordinary people and the state authorities to react to such
wrongdoing? Given the level of communication and transportation tech-
nology, a concern about excessive military power and limited fiscal
means, the traditional Chinese administrations responded to the prob-
lem of disorder in part by attempting to foster local semiformal forces
that could help with crime prevention and law enforcement. Given
the choice between private and state-mandated or state-authorized
semiformal systems, the state clearly preferred the latter, because it
gave the state some sense of control in local areas. In contrast with
semiformal systems, which were often sponsored or approved by the
state, private agents were in theory illegal, though in practice the state
tended to overlook such things as the private ownership of weapons as
long as those involved caused no overt trouble and belonged to reliable
social groups.
Semiformal systems, especially mutual security groups, were fre-
quently used by Chinese states. From before the unification of the
empire in the third century B.C. they existed in some Chinese kingdoms.
Many imperial dynasties used such systems, and they continue in some
form in the present on the Chinese mainland. We might find intolerable
the compulsory registration in groups whose members are mutually
responsible for one another's behavior. In premodern China such feel-
ings were probably rare. In a relatively stable agrarian society like
traditional China, most of the members of a mutual security group
already knew one another well. The state was attempting to impose
liability on them, not the necessity of prying into the affairs of others.
Most of the misbehaviors of their neighbors were common knowledge.
The usefulness of such mutual security groups to the state was obvious,
but so was the difficulty of making them effective. According to the Sung
official Yuan Fu, "Banding together at the village well is fundamentally
difficult."7
Chinese states also tried to cut the costs of administration, including
law enforcement, by drafting local people to serve as village officers to
collect taxes, to provide some local services, and to control minor dis-
orders. The use of such agents, like the establishment of local security
groups, reflected a tension between the ideal and the possible. Such local
agencies were both potentially and to some extent in reality antagonistic
to the desires of the central authorities, because they embodied a neces-
sary but not a desirable decentralization of authority. The tension be-

7 Yuan Fu, Meng-chai chi, 3.32.

512
 Conclusion

tween the desire of central authorities to exercise control in local areas

and the desire of the local people (including some local administrators)
to have a greater degree of control over their own lives was always
present. In the case of local subgovernmental agents, the state simply
had no choice. The authorities could not afford to use formal, remu-
nerated state functionaries to perform all the tasks of local government.
So, although clearly cognizant of the potential dangers involved, they
always used local agents.
The same sorts of dangers were created when the government used
functionaries who were not themselves members of the civil service. Such
people had to be used, but their values and goals often brought them
into conflict with the larger goals of the state. The Sung official Wang
Chih-wang noted that a district had recently created a new group of
archers but that "the prefectures and districts cannot keep them under
control, and the people suffer."8 Even more dangerously, such func-
tionaries might fall under the domination of locally powerful families,
who could use them to conduct their own private vendettas. 9 And when
they were mustered out, such men were loath to return to farming and
might turn to banditry. 10
On the level of formal governmental institutions, above the village
level, the response of the authorities was shaped by sometimes conflict-
ing aims and a variety of intellectual and physical constraints. It has
become fashionable to downplay the importance of dynasties in Chinese
history and to focus on elements that transcend such divisions. Certainly
the importance of the dynasties should not be overestimated, but they
were in fact important. Dynasties, and especially the long-lived dyn-
asties, were important because they established a political and institu-
tional context within which economic, social, and cultural change could
occur. In particular, dynastic founders usually set a stamp on the shape
of things to come, by defining certain problems as paramount and
responding successfully to them. The success of their responses was
reflected in the length of the dynasty. Institutions and policies that
have been successful have great inertia. They tend to remain, at least
nominally, in place. When they are changed, it is often in incremental
ways. When no longer effective, they tend to be worked around or
bypassed rather than eliminated or replaced.
The choices made by early Sung leaders reflect both the real-world
context within which they worked and their assessment of the problems
of the recent past. Many factors formed the context of their choices. The
8 Wang Chih-wang, Han-p'in chi, s^b-ga.
9 Huang Kan, Mien-chai chi, 27-3b-ga, 29.8b-ioa, 32.ia-ioa.
10 Liu Chih, Chung-su chi, 6.73.
513
 Law and order in Sung China

Sung was a time of great economic and technological advance. A newly

emerging social group was seeking, within the received repertoire of
possible and acceptable means, to create a stable and prosperous society
in which their group could remain dominant. Their perception of this
repertoire was largely a question of having inherited from the past a
spectrum of institutional possibilities, along with a particular view of the
effects of different choices. The selection of key policies by the founders
of the Sung was most affected by their understanding of the recent
Chinese past, which they saw as an era marked by excessive power in
the hands of the military. This situation had resulted in an administra-
tion that was both inhumane and politically unstable. The Sung dilemma
was the need for a military force sufficient to reunify the Chinese
ecumene, but not so powerful that it could coerce the civil state.
Both the initial form and the later evolution of Sung law-enforcement
and penal systems can be understood only in light of this concern. The
reemphasis on civil officials in law enforcement clearly reflects this atti-
tude; the gradual increase in the role of the military, which is perhaps
the most striking change in the pattern of Sung law enforcement, may
reflect both an increase in the difficulty of maintaining law and order
and a decrease in the degree of government concern over the dangers
of military power. Paradoxically, a dynasty that at its inception was
characterized by a great mistrust of the military ended by having more
soldiers stationed in the Chinese countryside than did any other dynasty.
The Sung founders' reviving the office of sheriff was certainly an attempt
to lessen the powers of military men in law enforcement; the gradual
growth in the numbers of patrolling inspectors and the appointment of
military officials as sheriffs suggest that civil appointees were unable to
handle the growing problem of disorder.
For some purposes, civil officials were ideal as agents in law enforce-
ment. Many had passed the examinations. Many others, entering
through the protection privilege, though not examination degree holders,
had been reared in a milieu where the publicly avowed values stressed
loyalty and integrity, and they were beholden to the ruler for giving
them the opportunity to join the most prestigious group in Sung society.
But the civil officials had one serious drawback. They often seem to
have been incapable of leading or guiding military actions. The lack
of military abilities among civilian sheriffs is a common refrain. In
responding to this problem, the state authorities tried to compromise,
by appointing military men to the civil post of sheriff. But this was
insufficient. So in addition, the state increasingly used military agents in
law enforcement.
Even in the early Sung, the military were part of the law-enforcement
514
 Conclusion

process, especially in the cities but also increasingly so in the country-

side. Here what was desirable, an administration freed from any possi-
bility of excessive military influence, was faced with what was necessary,
the continuing reliance on the military as the institution best able to
mobilize a sufficient force at an acceptable cost. Chinese dynasties varied
in their response to the tension between the wen and the wu, the civil and
the military. The T'ang had gloried in its martial spirit; the Sung exalted
the civil. Yet the T'ang always relied on a basically civilian bureaucracy,
and the Sung could only reduce the presence of the military, not elim-
inate it. Even in its attempt to reassert civilian control of the law-
enforcement system, the Sung was drawing on military tradition. The
Sung founder reestablished the position of sheriff and staffed the post
with civil officials, yet the position itself had originated in the Ch'in and
Han periods as a military position.11
The Sung responded to this tension by creating a dual law-enforcement
system that included both civil and military, and over time, as the initial
concern with excessive military power abated (and as the policy of
lessening the power of the overweening military proved successful),
allowed a greater role for the military men. It is not clear to what degree
this was due to an increase in the amount and level of disorder, though
this was clearly a factor. By the end of the Sung, the majority of law-
enforcement officials were military, in both rural and urban areas, and
the absolute number of law-enforcement agents had risen. In choosing
between the hero and the gentleman, the state more and more often
turned to the hero.
Some of the increase in the number of law officers was probably
related to demographic changes. More people might not necessarily
mean more crime, but at the least they did pose greater problems of
control. To exercise the same degree of control over a larger popula-
tion, more agents were needed. Shortly after taking local office, the
Southern Sung official Hsueh Chi-hsiian suggested to his superiors that
an additional position of sheriff be established in the southern part of
the district where he was serving, because it was an area where
"merchants come and go because the trade route there is short and
convenient, and in recent years the area has flourished, yet there is no
government office."12
Greater prosperity might also have led to more criminals, by making
available more tempting targets for them. 13 And as the state responded

11 Wu Ching, Chu-chou chi, n.8b-9a.

12 Hsiieh Chi-hsiian, Lang-yu chi, 20.9b-11 a, 26.1 a fif.
13 For the current use of this rather counterintuitive suggestion as one explanation for the

515
 Law and order in Sung China

to new situations by creating new laws, it also in turn created new

crimes. The most noticeable example of this phenomenon in the Sung is
the enormous increase in the number of crimes punishable by penal
registration. The general Sung perception of the underlying problem was
explained by the Northern Sung official Liu Chih:
In my opinion the law [fa] is the great ordering of the world. The system of the
law under the Former Kings had as its intent making it easy to avoid mis-
behavior and difficult to commit transgressions. Therefore the system was simple
and straightforward, yet it sufficed to embody completely the principles of the
world. In regard to the system of law under later generations, they were con-
cerned only that those who were guilty might occasionally escape. Therefore they
increased the numbers of constraints so that the people did not know where they
might safely put their hands or feet.14
According to the general paradigm of law enforcement, the authorities
need to get sufficient numbers of adequately trained and armed, properly
led and motivated men to the right place in time to help stop crime or
seize lawbreakers. If there were an increase in crime, they might in
theory change any of these factors or any combination of them. The state
could provide more men, better armed, trained, and led, more effectively
imbued with appropriate attitudes, and so stationed that they could
react effectively. The actual Sung response to the growth of disorder was
to increase the number of men and create better control and supervisory
systems. The thirteenth-century official Hsu Ying-lung noted: "The key
to the suppression of bandits is the proper selection of personnel." He
then went on, however, to list three components of this strategy, not all
of which the Sung adopted in practice. For the proper selection of men to
be effective, the authorities must focus on the essential qualities of the
task and the man. They must keep the lines of authority simple and
clear. And they must react swiftly but surely.
In practice, beginning in the late Northern Sung, the men chosen for
law enforcement were more and more often regular soldiers led by
regular army officers, the patroling inspectors, and not civilian archers
led by sheriffs. The authorities perhaps hoped that they would be better
armed and trained than their civilian counterparts were. There are no
indications that the Sung authorities worked to improve the indoctrina-
tion of these men. Rather, they sought to improve coordination and
oversight by refining the supervisory system. During the Northern Sung,
in many areas the responsibility for supervising law enforcement and the

recent increase in American crimes see Wilson and Herrnstein, Crime and Human Nature,
pp. 412-13.
14 Liu Chih, Chung-hsuan chi, 6.81.

516
 Conclusion

legal process was shared by judicial and other sorts of intendants. This
pattern fitted the general Sung penchant for dividing up responsibilities
to prevent any one office from monopolizing control of any function.
However, during the Northern Sung, the role of the judicial intendant
became increasingly important, and in the Southern Sung, judicial
intendants had almost exclusive control over legal and law-enforcement
functions. This may have reflected a lessened concern about the con-
centration of control in one office; it probably also revealed the need for
a greater unity of command because of the greater number of large-scale
disorders.
The Southern Sung was noteworthy for its major bandit suppression
campaigns, most of which were supervised by judicial intendants. A
Southern Sung official described such a campaign that he directed
while serving as a judicial intendant. The bandits were really inland
pirates operating on the great lake called T'ai-hu. The band numbered
hundreds of men, armed with military weapons and using a large num-
ber of boats. The official describes how he moved larger numbers of
troops into the region in the least obtrusive way, built forts, and tried to
separate the criminal leaders from their followers by offering amnesty to
those who surrendered. When the final showdown came, the government
troops captured 380 men alive, plus 20 army deserters, and seized over
1900 pieces of military equipment and more than 200 boats. In addition,
he says, among the bandits there were "an unknowable number who,
aware of their crimes and seeing no way to escape, threw themselves into
the waters or strangled themselves among the hills." 15 When disorders
reached this size, the line between bandit suppression and military
action was thoroughly blurred.
The Sung authorities were aware of the kinds of problems raised by
their attempts to improve law enforcement. The numbers of soldiers
involved rose relative to the numbers of civilian archers, even though the
higher officials knew of the drawbacks of the more military system.
There were complaints that soldiers were ineffective because they were
ignorant of local situations and could not gain the cooperation of the
local people. Sometimes the soldiers made no pretense of trying to
control rural areas. Again, there was a tension between the desirable and
the feasible. The Sung army provided a ready pool of patroling inspec-
tors and law-enforcement soldiers, so that the numbers used could be
increased without difficulty. Increasing the number of archers and sheriffs
would have been far more difficult. Already in the Northern Sung there
were complaints that positions for sheriffs went unfilled for years because

15 Hsu Ying-ling, Tung-chien chi 8.16a-19b.

517
 Law and order in Sung China
qualified candidates refused to accept them. The job had heavy responsi-
bilities and low prestige, so men often shunned it. The Southern Sung
official Yuan Yiieh-yu wrote: "Among the official positions in prefectures
and districts, the lowliest and yet the most burdensome is that of the
sheriff. The sheriff cannot look down on men; men look down on the
office and will not fill it." 16 The patroling inspectors were already military
officers. To increase their number meant a mere shift in the duty loca-
tion of the officers and a detailing of regular soldiers to support them.
The crimes and criminals with which these law-enforcement agents
had to deal do not seem to have changed radically during the Sung. The
typical criminal seems to have been a young unattached male during
both the Northern and the Southern Sung, and in both periods, the
crime of greatest concern was banditry. This generalization needs to be
qualified by noting that the increase in city size and the change in urban
character spawned organized criminal organizations. (We have no evi-
dence that the kinds of organized crime described for late Southern Sung
Hangchow existed earlier in the dynasty.) Also, because control of the
coastal waters was more critical to them, the Southern Sung auth-
orities were more concerned with piracy. The state derived a substantial
amount of revenues from levies on water-borne commerce, and the
economy in general was much more oriented toward the sea. One re-
sponse was to use judicial intendants to coordinate large-scale pirate
suppression operations and to modify bandit-controlling policies orig-
inally designed for inland use to the special conditions created by the
coastal environment.
The overall aim of this system was providing a level of security
adequate to allow the continuation of normal life for most people and to
preserve stable political arrangements. The ideal might be to prevent
crime, but in reality the authorities combined crime prevention efforts
with systems for capturing or suppressing criminals. One product of this
system was captured criminals, which presented the authorities the
problem of how best to deal with them.
The Sung punished people in order to exact retribution, to deter
further wrongdoing, to defend society against the criminals, and to bring
about their rehabilitation. Like the system of law enforcement, the penal
system was the outgrowth of the goals sought, the means available, and
the historical context. The goals were not all sought equally, nor were
they always commensurable. A tension always existed between the need
to defend society and the desire to rehabilitate offenders, between the
search for security and the promise of redemption. Social defense implied

16 Yuan Yiieh-yu, Tung-t'ang chi, 18.3b.

518
 Conclusion
an exclusion of the criminals from society so that they could not harm it.
But rehabilitation demanded that these people not be so thoroughly
excluded that their later reintegration would be foredoomed. In addition,
the Sung leaders were aware that if laws were seen as excessively harsh,
they might not be enforced, and they felt that if they were enforced, such
harsh laws would result in brutalizing the people who would themselves
become cruel. 17 As with other aspects of the legal and law-enforcement
systems, the compromise resulting from the tension between conflicting
goals was determined by practical considerations and by the authorities'
understanding of historical experience. The Sung official Ts'ai K'an
repeated in a memorial the conventional wisdom concerning the utility
of punishments: "Executions are used in order to bring an eventual end
to them, and punishments are applied for a time so as to end them.
The kings of ancient times all cherished life, but they always executed
murderers." 18
The Sung dynasty had inherited the traditional system of the Five
Punishments, of beating with the light and heavy rods, penal servitude,
exile, and death. In a manner typical of the Sung approach to institu-
tional problems, the authorities maintained this traditional system and
even used it for some purposes, though for most purposes they used a
different system. The Sung had inherited from the immediately preced-
ing states a punishment called penal registration, and they used it as one
of the most important penalties. Its importance can be measured in the
steady increase in the numbers of crimes calling for registration and also
in the impressions of contemporaries.
Why did this penalty come to play such an important role? Labor
penalties per se had been extremely important since the Han period, so
in one sense the Sung was continuing in an old tradition. The use of the
labor penalty called penal registration stemmed from the political and
social conditions at the time of the dynasty's founding. When the Sung
founder succeeded in pacifying much of the empire, he found himself
with a large body of rootless vagabonds, ex-soldiers, and wanderers with
no fixed place in society and no home. He settled such men in the
military units of the provincial armies, a policy that served as a welfare
measure, a security measure, and eventually a labor measure. From the
point of view of these men, this arrangement provided a haven from their
previous rootless existence. It also reduced the dangers of their turning
to crime in order to survive and so contributed to the security of society

17 See, for example, the comments of Wen Yen-po, Lu-kung wen-chi, SKCSCP ed., 9.8b.
18 Ts'ai K'an, Ting-chai chi, 1.18a.

519
 Law and order in Sung China
as a whole. And finally, it gave the state a large pool of available
workers.
The development of this system under the Sung is one of the best
examples of the virtuosity with which the Sung created functionally new
systems by continuing old forms in name and changing them in essence
by altering their contents. During the Five Dynasties, there were a
multitude of crimes punishable by death. Penal registration was used as
a penalty for the few men convicted of these crimes who were spared.
The Sung retained many of the laws with their death penalties and the
practice of penal registration and yet radically altered the meaning of the
system by systematically sparing the great majority of those convicted
and sentencing them to penal registration.
The units of registration were thus used as places for convicted crim-
inals. Such criminals were registered in certain units of the provincial
armies. They lived in encampments generally located adjacent to the
encampments of other provincial and imperial army units. The men
in them were controlled by the same patterns of small-unit mutual
responsibility and military law applied to other military units. By locat-
ing the convict groups in the midst of other military units, the state in
effect used the violent to control the violent.
The empire was now at peace. Men who had initially been soldiers
were unlikely to be used for further military service. To keep them busy,
the authorities decided to use them as laborers. More than did the other
dynasties, the Sung used this labor force to perform a whole host of
tasks, from running the postal system to overseeing the herds of state-
owned horses to building local facilities, tasks that in most dynasties
were largely the responsibilities of the local commoners. This freeing
of the commoners from many burdens without question increased the
amount of labor power that they had available to pursue their private
interests and thus contributed in an unmeasurable but clearly significant
way to the economic growth of the Sung.
The rehabilitation process was the converse of the system of control.
The rootless men were now embedded in a military system, with fixed
homes, exemplars, and comrades. Furthermore, they had always before
them the promise of ameliorating their lot.
The Sung was especially noteworthy among Chinese dynasties for its
stress on ameliorating the lives of its convicts. In no other dynasty were
amnesties granted so frequently. There are repeated policy statements
about the importance of merciful punishments and even compilations of
the rules and orders concerning such mercy. Furthermore, although
there were a great many crimes that called for capital punishment, in
many cases those convicted seem to have been spared and condemned
520
 Conclusion

instead to penal registration. The concern of the state extended even

beyond the ending of their sentences. If they served out their time or if
they were forgiven their labors by means of amnesties, the men had the
option of reenlisting in other units of the provincial armies (or if they
were young and strong, even in units of the imperial armies). This meant
that men, young or old, who were mustered out of the convict units and
who had no other prospects did not have to face a bleak future. They
had always the opportunity to continue in the armies, under conditions
markedly better than those they had endured as convicts. This system
fit superbly with the fundamental personality traits of criminals such
as robbers and thieves. Modern criminological studies have shown
repeatedly that one trait shared by such persons is a lack of impulse
control and a very short time horizon. They discount heavily the future
possibilities of reward or future threats of punishment in favor of imme-
diate gratification. The Sung system put in front of these convicts, on a
daily basis, a possible pattern of life better than that they were enduring.
It held before them the virtual certainty that if they behaved reasonably
well, they could improve their lot at the end of their sentences and,
because of the frequent use of amnesties, also held out the promise that
they might see their status improved at any moment. Once out of their
original units, such men continued within an institutional context sup-
portive of their continued good behavior, and because they lived and
worked close to men still serving as convict laborers, they had always
before them a reminder of where they had been and where they might
return if they again committed crimes.
Death remained as the ultimate sanction. Its forms even increased in
ferocity. But this was counterbalanced by a tendency to spare men's
lives. This process, allowing crimes to continue on the books as capital
crimes but in practice sparing those convicted, may have had, indeed
may have been intended to have, the effect of making the convicts feel a
sense of having been given back their lives by the emperor, to whom they
were thereafter indebted. Men may have been spared so that the state
could exploit their labor; they may have been promised the possibility of
a gradual improvement in their lot because in practice it was difficult to
prevent the escape of any determined convict, but from the convicts'
point of view, the result was an extension of mercy.
I do not mean to imply that such mercy was extended equally. In both
theory and practice, Chinese law was marked by a sense of hierarchy
and privilege. Actions against superiors were more heinous than actions
against inferiors. For a commoner to harm an official, for a wife to harm
her husband, for a low-ranking officer to harm a higher-ranking officer,
for a junior relative to harm an elder, these were most serious crimes.
521
 Law and order in Sung China

For an official to harm a commoner, for a husband to harm his wife, for
a superior official to harm a subordinate, for an elder to harm a junior,
these might (or might not) be crimes, but they were almost always of
lesser seriousness.19
Privileges were also characteristic of Chinese law. Some groups - the
old, the young, the disabled, the insane, and women - benefited from a
conviction that they had a diminished capacity. Most privileges, how-
ever, were the fruits of power, especially political power. Officials and
relatives of the emperor by blood and marriage were privileged in various
ways.
This Chinese traditional system had respectable roots in the traditions
of the Classical period. The Sung official Su Sung described this system:
I have heard that in antiquity officials and their wives did not have to stand trial
in person. This was done to indicate the precedence of the honorable and to
cause the inferiors to suffer the trial so as to nourish a sense of shame and
promote appropriate behavior. Chia I said, "A sense of decorous shame and
of appropriateness of behavior is the way to cultivate people of exemplary
character. Therefore people might be 'granted death' [i.e., allowed to commit
suicide in private] and be spared the shame of public execution. Thus the
penalty of the cutting off of the nose did not extend to the great officers, who
were not far separated from the ruler himself."20
The later Chinese system of privileges, as it was enshrined in the
T'ang Code and inherited by the Sung, was peculiar primarily in the
formalism it used in awarding privileges. Specific ranks received specific
privileges. Even dispensations of mercy were surrounded by rules. We
cannot understand the working of the traditional political system with-
out understanding these networks of privileges extended to the politically
powerful.
More striking, however, and more typical of the Sung, was the seem-
ingly greater role of wealth per se in securing privileged treatment at the
hands of the law, not just extralegally, but also within the rules of the
game. The Sung rules on commutation often legally permitted the rich to
escape the penalties for minor crimes through legally permissible pay-
ments. Their wealth also allowed them to avoid punishment illegally
through collusion or bribery. The official Liu I-chih speaks of the
ways in which the powerful escaped penalties. They even committed
homicides and then importuned the officials to change the facts recorded

19 For this phenomenon as it affected relations between commoners and the demeaned
groups, see WHTK 167.1450.
20 Su Sung, Su-Wei wen-kung chi, 17.14b-15a. That the nobility in ancient China were not
always treated so leniently is beside the point as long as the men of the Sung thought
they had been so treated.

522
 Conclusion

in the reports or bribed the investigating personnel so that the testimony

of witnesses was m a d e unclear. 2 1 His colleague Ts'ai K ' a n criticized
both bribery and the abuse of the "protection" (yin) privilege by rela-
tives of officials. Because such abuses existed,

what is reported up to higher authorities is meaningless. They say that the

bandits did not kill or commit arson or rape, and so on. And when the case is
completed, they also say that the circumstances were not serious or that the
penalty called for by law is questionable. Thus they memorialize for a central
government decision so that subsequently the criminals avoid execution and are
merely sentenced to penal registration. Before they have been transported a few
post stations away, they escape from their fetters and flee to other areas where
they again turn robber. 22

At times the Sung provided mercy for the rich, punishments for the poor,
and justice for none.
For all their flaws, however, and the imperfections that inevitably
mark h u m a n institutions, the Sung systems of law enforcement and
punishment are remarkable for their combination of a sensitivity to
tradition, a grasp of practical psychology, an understanding of the
complex interplay of political reality and economic constraints, and a
willingness to retain the shell of hallowed institutions while changing
their contents. T h e ability of the Sung leaders to use old institutions in
new ways is one explanation of their success in staying within Chinese
tradition while adapting to a novel world.

21 Liu I-chih, T'iao-ch'i chi, SKCSCP ed., 12.9b-ioa.

22 Ts'ai Kan, Ting-chai chi, 1.18a.

523
 Glossary
Ah Yun MM
An (Prefecture)
An-fu (District)
An-fu shih
An-hsi (District)
An-su (District)
An-su (Prefecture)
chai jji
Ch'an (Prefecture) jf
changa fit
changb ti
Chang-ch'eng k'ou
chang-chia jltt]
chang-chih ft i t
Chang Fang-p'ing
Chang Rang
chang-kuan t'ing
Chang-ning ft¥
Chang Shih-ying
Chang Shou K T F
Ch'ang (Prefecture)
Ch'ang Ch'en ' S ^
Ch'ang-hua (District)
Ch'ang-kuo (District)
ch'ang-she #ifc
Ch'ang-shou (District)
Chao Chien ^ K
Chao Chung-heng
Chao K'uang-yin
Chao P'u m^
Chao Shan-k'uo H
Chao Shan-liao %
524
ch'ao i tzu @—^Sf
Che-yii kuei-chien
chen $|
chen-chiang Hitl
Chen-chiang (Prefecture)
Chen Te-hsiu
Ch'en (Prefecture)
Ch'en Ch'un
Ch'en Fu-liang
Ch'en Hsiao-sun
Ch'en Liang
Ch'en Tung
Ch'en Wen-cheng
Ch'en Yuan-chin
Ch'en Yuan WM
cheng IE
Cheng (District) ft(5
Cheng (Prefecture)
Cheng Hsing-i
cheng-tung
ch'eng-chu
ch'eng-li tu hsiin-chien
ch'eng-t'an ftJcJiL
ch'eng-t'ien 7pl:?c
Ch'eng Yi SEJi
Chia (Prefecture) ^
Chib (Prefecture) S
chi-kuan 111=?
Chi-lii (Garrison)
ch'i ft
Ch'i (Prefecture) Sf
ch'i-chang
 Glossary

ch'i-pan shih-ch'en Chin3 (Prefecture)

ch'i-shih |?Tf} Chinb (Dynasty) ^
chiaa ¥ chin-chiin ^ ^
chiab m chin-ping
chiac m chin-shih
chia-chang Ch'in (Prefecture) #
chiang S|# Ch'in-feng (Circuit)
Chiang-che tC$Jf Ch'in Kuan lHH
chiang-chiao Ch'in Kuei ^ f t
Chiang-nan ching-ch'ao kuan
chiang-tao Ching-chi (Circuit)
Chiang-tso chien Ching-hsi (Circuit)
Chiang-yin (Prefecture) £ Ching-hu (Circuit)
chiang yu-hou MSIfK Ching-tung (Circuit)
chiao-hua ifcft ch'ing3 m.
chieha m ch'ingb If
chiehb ffi Ch'ing-ming chi
Chieh (Prefecture) ffl Ch'ing-yiian t'iao-fa shih-lei
chieh-chi 8f5|5
chieh-tzu i^fi1 chiu-ch'eng tu hsiin-chien
chien3 M ch'iu-chang
chienb fgj ch'iu-fen
Chien-ch'ang (Prefecture) ch'iu-pu
chien-fa kuan Ch'iung (Prefecture)
chien i hsiian chou ji]
Chien-k'ang (Prefecture) Chou-hsien t'i-kang
chien-tang S # Chou Mi
chien-tang shih-ch'en S Chou Pi-ta
chien-ya Sniff ch'ou |3
Chien-yen i-lai hsi-nien yao-lu ch'ou-chuang
Chu-chi (District) fffi
ch'ien $B Chu Chin ^ H
Ch'ien (Prefecture) Chu Hsi %:M
ch'ien-hsia I^K chu-lu chien-ssu
Ch'ien-hsia ssu chu-lu chou-chun shang-fa
Ch'ien-t'ang (District)
Ch'ien-yu (District) chu-po
ch'ien-yiian chu-po tsu-tsei shih-ch'en
chih-chen E
chih-hui chu-pu ± f f
chih-hui shih Chu Tsung ^ K
chih-yii B « Ch'u (Prefecture) lH
ch'ih R ch'u-shen ffi#
Ch'ih (Prefecture) i ch'u-sheng ^ i g
chin Jf Chuan-yiin shih

525
 Glossary

Ch'uan-Shen Fu-yang (District)

chuang-ting JfctT hai tsei-li
Ch'un-hsi i-chou i-lu ch'ou shang- Han Ch'i
fa mm—wi-KiHttgc Han Chiu
Chung-ching (Command) ,1&#pf Han Wei
chung-fa-ti Han-yang (District)
Chung-shu Han Yiian-chi ^ X
Ch'ung-chieh (Command) hang fj
Ch'ii (Prefecture) if Hang (Prefecture)
chiian )jl$ hang-kuan fj'g'
chim f$ Hao (Prefecture) ^
Chiin (Prefecture) i% Ho Chiu-kung
chiin-chu Ho-pei (Circuit)
chiin-hsien Ho-tung (Circuit)
chiin-hsiin Hsi (Prefecture)
chiin-hsiin pu-shih Hsi-an (District)
chiin-hsiin shih W- Hsi-Ch'uan ffiJII
Chiin-hsiin yuan W hsi-pei chou-chiin chiu-shang
Chiin-t'ou ssu
chiin-tzu fJ-J* hsi-sha
E (Prefecture) Hsi-yiian lu
fa ft hsiang3 J|g
fa-chia £fe^ hsiangb %
fa-hui & # Hsiang (Prefecture)
fan-chen S$R hsiang-chu
Fan Ch'eng-ta hsiang-chiin
Fan Ch'un-jen ? Hsiang-fu (District)
Fan Chung-yen hsiang-hsiin chien
fan-chiin hsiang-hsiin pu-ping
Fan Chiin J hsiang-jen
Fan Tsu-yii hsiang-ping
fang i§ Hsiang-shan (District) 0I_
Fang (Prefecture) hsiang-tien iffiJft
fang-cheng hsiao-jen ^^A
Fang La Hsiao-shan (District) Hli|
Fang Pao hsiao shih-ch'en
Fang Shao hsieh %
fen ^ Hsieh Ching-wen
Feng Ching Hsieh Yen-hsiang ^ $
Feng-hua (District) hsien (^
fu ffi Hsien-chu (District) fllj
Fu (Prefecture) f§ Hsien-ch'un Lin-an chih
fu-ping fifgk hsien-kuan t'ien PS'g'ffl
Fu-t'ien yiian feBBI hsien-t'ien
fu tu chih-hui shih hsien-wei

526
 Glossary

Hsin (Prefecture) W I-(chou) (Circuit)

Hsin-ch'ang (District) §f H i chun-fa
hsin-ch'eng li tu hsiin-chien i-yung ftS
Jao (Prefecture)
hsinga M jen flE
hsingb T& Jen-ho (District)
Hsing (Prefecture) M Ju (Prefecture) ^
Hsing-kuo (Prefecture) Jfi Jung (Prefecture)
Hsing-pu fflfP kai-kuan B^'g'
Hsiu (Prefecture) If kai-tzu teng-kuan
Hsiu-ning (District) £fc¥ k'ai-chi ^flS
Hsiung-chieh (Command) K'ai-feng (Prefecture)
Hsii (Prefecture) ^ K'ai-te MM
Hsii Ching-heng rFJp;||f k'ao %
Hsu Hsu &M ko f&
Hsii Jan ftffi ko-men chih-hou
Hsii Ying-lung fFffifi k'o-hu
Hsiian (Prefecture) fi ku &
hsiian-jen 3gA ku-sha
Hsiieh Chi-hsuan g ¥ $ g K'uai-chi (District)
hsiieh-shih yiian K'uai-chi hsii-chih
hsiina ifr kuan-hu 1=f^
hsunb M Kuang (Prefecture) ^
hsiin-chien Kuang-hsin (Prefecture)
hsiin-she 3 Kuang-ku (Command)
Hsiin-tzu Kuang-nan (Circuit) jf M
hsiin-wei ? Kuei (Prefecture) It
Hu Su #j?g kuei-hsin %LW
Hua (Prefecture) f Kuei-yang (Prefecture)
HuaChen $|R K'uei (Prefecture) ®
Huai-nan (Circuit) K'uei-chou (Circuit)
Huai-yang itHf K'un-shan (District)
Huai-yin (District) Kung 1ft
Huang-ch'eng ssu kung-ping
Huang Kan ]!(!£ kung-shou
Huang K'o-chung kung-tsui £ f f
Huang Shou-hsiin Kuo (Prefecture) M
Huang-yen (District) Lan (Prefecture) jfjj]
Hui (Prefecture) M Lei (Prefecture) If
Hung Mai lia K
iib a
a
I (Prefecture) H lic (Prefecture) ^IJ
Ib (Prefecture) iff
I-chang (District)
I-ch'eng (District)

527
 Glossary

Liab
m lu-ch'iu 0 0
Li m Lu-shih ts'an-chiin
li-cheng MIE Lu YU mm
Li Chin ^& luan-hai ffiflj
Li-chou (Circuit) f IJ'/H lu ft
li-ch'un Lii I-chien S9IS8
li-hsiaa Lii Tsu-ch'ien
li-hsiab Ma Hsi JSE3
Li-hsiieh chih-nan ma-ch'ien san-pu
Li Kou Ma Mo MjR
Li Kuang ma-pu chiin lung-shen wei ssu-hsiang
Li Mi-hsiin tu chih-hui shih
LiO $ l f
li-pu ^SP Ma-pu chiin-ssu ( ^
Li Tao Ma-pu tu-chien
Li Yen $ Ma-pu yuan J|/
Li Yiian-pi Man ^
liang ^j ManTzu-lu
Liang-che (Circuit) Mao Fang ^ ^
Liang-ch'uan Meng-liang lu
Liang Tao ^ Meng Yuan-lao
Liang-tzu Mi (Prefecture)
Liang Wen-wei min-chiin
Liao En J p S ming pp
Liao Hsing-chih Minga (Prefecture)
Liao Kang JpBJ Mingb (Prefecture)
Lien (Prefecture) j mo-k'an
Lin-an (Prefecture) mou HA
Lin Tan mou chih-hui tsa-i
ling ^» mou-sha Wk$i
ling-ch'ih mou ta-ni Wk~tt8L
Ling-nan mu chia-tzu ^C^i?
Ling-pei W mu hsiang-ping ;fc
liu SE mu-hsieh ^Cft
Liu An-shih Mu-jung Yen-feng
Liu Ch'ang mu-lu ^ i i
Liu Chih Nai Te-weng iff#
Liu Hsi S Nan-feng (District)
Liu-nei ch'iian Ning-hai (District)
Liu-tu (Fort) Ning-kuo (District)
liu-wai Ning-tu (District)
Liu Yiieh o-hsiao 31'h
Lou Yiieh o-ni M$L
lu % Ou-yang Hsiu
Lu (Prefecture) pa-i

528
 Glossary

p'an-chia Shang-tang (District) _hH

p'an-kuan Shang-yii (District)
pao ft Shao-fu chien
Pao (Prefecture) ft Shao-hsing (Prefecture)
pao-chia ft^3 she-fu chang ttiljft
pao-chieh ft$| she-kuan W^
Pao-i ftft Shen-hsing yuan STPJSt
pao-wu ftffi sheng ft*
pen-ch'eng Sheng (District) M
pen-shen li Sheng-shui yen-t'an lu
P'eng Kuei-nien shiha ^C
P'eng Tao shihb 5
pi-lii tsu-tang shihc 7F
Pien "pt shihd R
pien-kuan shihe ±
Ping (Prefecture) Jf Shih (Prefecture) J6
Ping-ch'iu yuan shih-ch'en ^ E
Ping-pu ^SR shih-chia E[S
p'in Rp shih-fei ^ ^ ^
P'ing (Riding) # Shih-po ssu
Poa (Prefecture) If shih-ta-fu
Pob (Prefecture) 31 Shih-wei ch'in-chiin ma-chiin pu-chun
Po-chai pien
pu-chang fftJS: Shih-wei ch'in-chiin ma-chiin ssu
pu-ch'u fRffi
pu-kuan fillf Shih-wei ch'in-chiin pu-chiin ssu
pu-mu
pu-tao shih-wu ff*ffi
pu tsei t'iao Shih-wu chi-yiian
P'u (Prefecture) ft Shou (Prefecture) H
p'u-hu Mfi shu Si
San-ch'ao pao-hsiin H Ill| Shu-mi yiian
san-kuan feilf shuai-ch'en
San-pan yiian H; so H
San-shan chih H so-yu 0fS
San-ya H@f ssu ^E
san-yiian Ei7C Ssu (Prefecture) ffl
Sang Shih # » ssu-k'ou ^ I S
Seng Fa-tuan ssu-k'ou ts'an-chiin
Seng Shou-kung Ssu-k'ou yiian W|^
Shan (Prefecture) $1 ssu-li ts'an-chiin
Shan-yin (District) Ssu-li yiian
Shang (Prefecture) Ssu-lu ssu W]$gWJ
Shang Chih-chung
Shang-shu ffitt Ssu-ma Kuang

529
 Glossary

ssu-mien hsiin-chien t'i-tien wu-tao shih-ch'en

ssu-p'ai an
ssu wu hsieh t'ieh-kuo
Sua (Prefecture) tien-chiang fa
Sub (Prefecture) tien-ch'ien feng-jih t'ien-wu ssu-hsiang
Su Ch'e tu chih-hui shih
Su Shih g f t
Su Sung jfggg Tien-ch'ien ma-pu chiin-ssu
Su-sung (District)
Sui-an (District) Tien-ch'ien shih wei ssu
Sui-shou tsa-lu Tien-ch'ien ssu
Sun Sheng tien i-hsuan Wt~^
Sun Shih tien-chih
Sung-ch'ao yen-i i-mou lu t'ien-ch'i
t'ien-ch'ing
SungCh'i t'ien-hsiang
Sung hsing-t'ung t'ien-k'uang
ta § Ting (Prefecture)
Ta-an Ting-hai (District)
ta-chi ting-mao T^P
ta-chi chi T'ing (Prefecture)
ta-chia toua m
ta-hsiao shih-ch'en ^ toub 4
Ta-lissu ± l l ^ f tou-sha
Ta-ming (Prefecture) tsa-chu
ta-she ^ci5[ tsa-hu
ta shih-ch'en Ts'ai (Prefecture)
ta-ts'u Ts'ai Ching J I M
ta-t'ung Ts'ai Hsiang
T'ai (Prefecture) • Ts'ai K'an
t'ai-sui ^C^ ts'ai-wu ch'in-min
T'ai-p'ing (District) Ts'ao (Prefecture) W
Tana (Prefecture) tsei K
Tanb (Prefecture) Tseng Rung t »
t'an-mien Tseng Pu # ^
t'an-shihjen tso chiin-hsiin t'ui-tien
Tang-t'u (District) Tso-i tzu-chen
T'ang (Prefecture) tso-yu hsiang
tao ^ tso-yu chlin hsiin-shih
tao-tsei ^ K Tso-yu chiin-hsiin yuan
Te (Prefecture) ^> tso-yu hsiin-shih
Teng (Prefecture) tsu-tsei shih-ch'en
Ti (Prefecture) ^ Ts'ui Pai S S
t'i-chu tsu-tsei Ts'ui Tun-li
t'i-tien hsing-yii shih ts'un ^f

530
 Glossary

Wang Yen
Tu-ch'ang (District) f[$H Wang Yung
Tu Cheng It IE Wang Yii-ch'eng
tu-chien §[$i£ weila Ht
Tu-chien ssu fftSniRl weib S
tu-chih hui-shih fltfgjf {£ Wei (Prefecture) If
Tu Fan t t i £ wei-chih ^SftlJ
tu hsiin-chien i Wei-kuo (Command)
tu-pao Wei-nan (District)
Tu shui-chien wei-tzu
tu-t'ung hsiin-chien
tu yii-hou Wen (Prefecture) ffl
Tu Yen ftffi Wen-hsien t'ung-k'ao
t'u ft Wen Min-chung
t u-ping Wen T'ien-hsiang
tuan ta-pi Wen Yen-po ~$CM
t'uan H wu ft
tui-chang ^M wu-ch'en ftE
t'un-ping shih wu-chieh 51|f5
Tung-ching meng-hua lu Wu Ching ^ U
Tung-kang jRIS] wu-hsing 51 fj
T'ung-ch'uan (Prefecture) Wu-i (district)
t'ung-p'an 3ft £0 Wu-lin chiu-shih
t'ung-yung shang fa Wu-ning (District)
Tzu-ching pien Wu-p'ing (District) ft2?
Tzu-chou (Circuit) wu-sha W?:$t
tzu-hsin §§f Wu-tai hui-yao
Tz'u (Prefecture) $J ya-chiao
Tz'u-chi (District) tkM ya-ch'ien
tz'u-ch'ung mou chih-hui pei-chiin $'J ya-kuan
ya-p'u
Wan-an (Prefecture) Yai (Prefecture) S
Wang An-shih Yang (Prefecture)
Wang Chi-yii Yang An-kuo
Wang Chih Yang Shou-chen
Wang Chih-wang Yang Wan-li
Wang Ch'in-jo Yeh Yuan Iflx
Wang Rung Yena (Prefecture)
Wang P'i-chih Yenb (Prefecture)
Wang Shu yen-mei IfRU
Wang Ssu yin I*
Wang Tan 3iR Yin (District) IP
Wang Tseng-yii Yin Chu f*&
Wang-wu (District) Yinga (Prefecture)
Wang Yang Yingb (Prefecture)

531
 Glossary

yu-hsia yii-kuan ling

Yung (Prefecture) Yu-shan (District) I U 4
Yung-an (District) Yii-yao (District) §£M
Yung-chiu $Iii|$ yiian-cheng x I E
Yung-hsing-chiin (Circuit) Yiian-ch'eng (District)
Yiian-ch'ii (District) Sffi
yii-ch'eng Yiian-feng chiu-yii chih
Yii-ch'ien (District) Yiian Fu S"B[
Yii Ching Yuan Hsieh H'Jf
yii-chuang Yiian Ts'ai ^ ^
Yii-hai Yiian Yiieh-yu ftU:S
Yii-hang (District) Yiieh Fei fifR
yii-hou Yiin (Prefecture) i^
yii-kuan

532
 Bibliography

Works in Chinese and Japanese

Chang Rang §gffl. Hua-yang chi ¥ R § * . SKCSCP 1972 ed.
Chang Nieh 3R*H. Tzu wei chi #tjft||. SKCSCP 1975 ed.
Chang Shou 3RTF. Fi-ling chi S P t * . TSCC ed.
Chang Yung 3R|$. Sung-pen kuai-yai hsien-sheng wen-chi
Taipei: Commercial Press, 1972.
Ch'ang-sun Wu-chi H: J£ $t ^ . T'ang-lii shu-i JS fl^ i t Ml. Taipei: Commercial
Press, 1970.
Chao Shan-k'uo ffi#g. Ying-chai tsa-chu (tSFH^. SKCSCP 1975 ed.
Chen Te-hsiu RJgif. Hsi-shan cheng-hsun HU-lJ&JU. TSCC ed.
Hsi-shan hsien-sheng Chen wen-chung kung wen-chi ^lUJ9t^.M^C&&^C^.
SPTK ed.
Ch'en Ch'i-ch'ing HJ#JBI. Chia-ting ch'ih-ch'eng chih S ^ ^ t t S . SYTFCTS
ed.
Ch'en Ch'i-ch'ing K#?lP. Yiin-ch'uang chi ^ H f S - SKCSCP ed.
Ch'en Ch'un &£&. Pei-ch'i ta-ch'iian chi itW<?k%k. SKCSCP 1973 ed.
Ch'en Chun W&). Huang-ch'ao pien-nien kang-mu pei-yao fLMWb^Wi @ Iffiic.
Taipei: Ch'eng-wen ch'u-pan she, 1966.
Ch'en Fu-liang BftfgK. Chih-chai wen-chi lkWiCM. SPTK ed.
Ch'en Hsiang I^R. Chou-hsien t'i-kang 'HiUMM. TSCC ed.
Ch'en Kung-liang |$!2;^. Yen-chou t'u-ching RMHffi. TSCC ed.
Ch'en Liang | ^ ^ . Lung-ch'uan wen-chi flill^t^. TSCC ed.
Ch'en Shun-yii B l ^ i ? . Tu-kuan chi f ^ H . SKCSCP ed.
Ch'en Yuan BlffiH. Mo-t'ang chi S ^ ^ ^ . SKCSCP 1972 ed.
Ch'en Yuan-chin 1^x11. Yu-shu lei-kao ifeK^BS. SKCSCP ed.
Cheng Hsing-i f W S j . Cheng chung su tsou-i i-chi KfoM%Wtf[%. SKCSCP
ed.
Cheng K'o 155&. Che-yu kuei-chien iffJtftig. TSCC ed.
Cheng Shou-p'eng JUS lip 8£. Sung-tai K'ai-feng fu yen-chiu (A Study of Kaifeng
Prefecture in the Sung dynasty) T^ftMi+lSW^L. Taipei: Kuo-li pien-shih
kuan Chung-hua ts'ung-shu pien fa wei-yiian hui, 1980.
Cheng Yao %%. Ching-ting Yen-chou hsii-chih *5£|5'Jf|iMS. TSCC ed.
533
 Bibliography

Ch'eng Pi S B . Ming-shui chi ffi**. SKCSCP 1972 ed.

Chiang Shao-yii tL^fil. Huang-ch'ao lei-yiian fi$|||%. Beijing: Chung-wen
ch'u-pan she, 1981.
Chiang T'ing-hsi M^WJ. KU chin t'u-shu chi-ch'eng ^ f ^ H W ^ ^ C . Kuang-hsii
blockprint, 1884.
Chiba Hiroshi ^PljI'A'. "Sodai no sho-gun" (The provincial armies of the Sung
dynasty) ^fi^Jffi^. Toho daigaku tanki daigakubu 2 (1982-83): 1-24.
Ch'ien Jo-shui IS:?? ^K. T'ai-tsung huang-ti shih-lu j&mMM'H^k. Ku-hsueh hui-
k'an ed.
Ch'ien Yiieh-yu i§ |ft S . Hsien-ch'un Lin-an Chih JiR ff US *?c ;£ . Ch'ien-t'ang
Wang-shih chen-ch'i-t'ang blockprint ed. 1830.
Chin Chung-shu ^ ^ f l l . "Pei-Sung chu-kuan chih-tu yen-chiu" (A study of the
official organization of the Northern Sung) it ^ M TET M SL W 3L. Hsin-ya
hsiieh-pao 9 (1969): 243-98.
Ch'in Kuan ^ffl. Huai-hai chi W&%. SPTK ed.
Ch'iu Han-p'ing Irjit2!5-. Li-tai hsing-fa chih (The legal treatises from successive
dynasties), Kftfflft/£. Taipei: San-min shu-chu, 1965.
Chou Mi M 85. Wu-lin chiu-shih ft tt H ¥ . In Meng Yiian-lao, Tung-ching
meng-liang lu, wai ssu pien.
Chou Ts'ung fflffc. Ch'ien-tao Lin-an chih ffiJIBI^^. SYTFCTS ed.
Chou Tzu-chih H ^ S . T'ai-ts'ang t'i-mi chi * # f ^ * . SKCSCP 1971 ed.
Chou Ying-ho JiJffi£\ Ching-ting Chien-k'ang chih Jt5!ftft/£. SYTFCTS ed.
Chu Hsi $tM. Chu-tzu ta-ch'uan * i F ± ^ : . SPPY ed.
Chu-tzu yii-lei ^ ^ | § ? R . Taipei: Wen-lu ch'u-pan she, 1986.
Chu wen-kung cheng-hsiin ^i&BcSH. TSCC ed.
Hui-an hsien-sheng Chu Wen-kung wen-chi MMJt^^AQ^tWz. SPTK ed.
Chu Pien yfc^p. Ch'ii-wei chiu wen ftiW^M. Chih-pu tsu-chai ts'ung-shu ed.
Fa-chih shih yen-chiu shih ffiSJ^Wf^L^. Chung-kuo ching-ch'a chih-tu chien-
lun (A brief discourse on Chinese police systems), ^S^^Wl^ffipffi. Beij-
ing: Ch'un-ts'ung ch'u-pan she, 1985.
Fan Chen ?Q M. Tung-chai shih-chi ^ W S£ • . Beijing: Chung-hua shu-chu,
1980.
Fan Chun ?ggt Hsiang ch'i chi WMM. TSCC ed.
Fan Ch'un-jen JEI^C Fan Chung-hsuan chi f&!&S.$&. SKCSCP 1978 ed.
Fan Tsu-yii mffliS. Fan-t'ai shih-chi faX$.$k. SKCSCP ed.
Fang Shao jj*}. Po-chai pien ffl^iji. Beijing: Chung-hua shu-chu, 1983.
Fu Ch'a ffiH. Chung-su chi &M&. SKCSCP 1973 ed.
Habu Kenichi ^ ^ . f t ^ . "Godai no jun-ken shi ni tsuite" (On the patroling in-
spectors during the Five Dynasties) E f t CO Steffi IZ gfcv^ X. Tohogaku, 29
(1965): 45-55.
"Hoso no jun-ken to hoko ho" (Patroling inspectors of the Northern Sung and
the pao-chia system). it^<DM^k^U^:A. Shien 92 (1962): 165-75.
Han Wei ©|§. Nan-yang chi jf fllfc. SKCSCP 1971 ed.
Han Yiian-chi i x ^ . Nan-chien chia-i kao ifjfiWZ/$!. SKCSCP 1975 ed.
Hino Kaisaburo 0 U MI HBP. "Godai chinsho ko" (A study of the chen-chiang in

534
 Works in Chinese and Japanese

the Five Dynasties) 2ftjRj|$%. Toyo gakuho 25 (1938): 54-85.

"Todai hanchin no bakko to chinsho" (Chen-chiang and the rampant power of
the T'ang fan-chen), Jf ft*0t£> SS/g £ H {ft. Toyo gakuho 27 (1940):
153-212
Hsieh K'o IlfjS. Ch'ti-yu chiu-wen ffiffiSM. Chih-pu tsu-chai ts'ung-shu ed.
Hsieh Shen-fu M 95flf. Ch'ing-yiian t'iao fa shih-lei ft 7C #6 j£ <P SI. Tokyo:
Koten kenkyu-kai, 1968.
Hsu Ching-heng f^JRffi. Heng-t'ang chi ft$jf&. SKCSCP 1975 ed.
Hsii Dau-lin &J1S5. Chung-kuo fa-chih shih lun-chi ^pW&V\^.lMl. Taipei:
Chih-wen ch'u-pan she, 1975.
Hsii Sung yf$t£. Sung hui yao chi kao S+cll'fcff^tj. Taipei: Shih-chieh shu-chu,
1964.
Hsu Ying-lung ffffift. Tung-chien chi ^MUk. SKCSCP ed.
Hsu Yiian-chieh &xj&. Mei yeh chi jUIflfc. SKCSCP 1975 ed.
Hsii Yiian-jui ^X^tft. Li-hsiieh chih-nan J^H^fe BJ. Chu Sung pi-yung shih-lei
ch'iian-chi ed.
Hsiieh Chi-hsiian § ? ^ J L . Ken chai hsien-sheng Hsiieh Ch'ang-chou lang yii chi
ftJff5fc£g#'JN*SI§*. Yung-chia ts'ung-shu ed.
Lang-yu chi feg§H. SKCSCP 1971 ed.
Hsiieh Ying-ch'i i ¥ B ^ . Sung-Yiian t'ung-chien ^ x S ^ . Taipei: Commercial
Press, 1973.
Hu Hou-hsiian 1$ If- M. . "Yin-tai te yiieh-hsing" (The penalty of amputation in
the Yin dynasty) KfWJJlWJ. K'ao-ku no. 2 (1973): 108-17.
Hu Su $?g. Wen-kung chi £ ^ H . TSCC ed.
Hu T'ai-ch'u M~kW. Chou-lien hsu-lun C H i i t l l . TSCC ed.
Hua Chen ¥£K. Yun-ch'i chu-shih chi a ^ g ± * . SKCSCP ed.
Huang Huai H j ^ . Li-tai ming-ch'en tsou-i K ft ^S E ^ ^ . Taipei: Tai-wan
hsiieh-sheng shu-chu, 1964.
Huang Kan J5&. Mien-chai chi MW*. SKCSCP 1971 ed.
Huang Kuan-chung HHii;. "Nan-Sung shih-tai Yung-chou te Heng-shan chai"
(Heng-shan Fort in Yung Prefecture during the Southern Sung) Bl^B^ft
l^WlfcUJtli. Han-hsueh yen-chiu 3 (1985): 507-34.
Hung Mai #£ S . Jung-chai sui-pi ^ ^ |5S ^ . Shanghai: Shang-hai shu-tien,
1984.
Johnson, Wallace. T'ang-lu shu-i yin-te (An index to the T'ang-lii shu-i) )SW®it
jff§|?|. Taipei: Wen Hai, 1964.
Kao Ch'eng MM. Shih-wu chi-yuan > f e | g g . Taipei: Yee Wen, 1967.
Kato Shigeshi Jp S 51 • Shina keizaishi kosho (Studies in Chinese economic
history) $ £P g ^ £ % f t . Tokyo: Toyo bunko, 1952. Translated into
Chinese as Chung-kuo ching-chi shih k'ao-ch'eng. Beijing: Commercial
Press, 1962.
Kawahara Masahiro ffij^lEW. "Sodai no sai-jin sai-ki ni tsuite" (On cultic hu-
man sacrifices in the Sung dynasty) 7 R f t ^ ® A ^ A ( - ^ V > " C . Hoseishi
kenkyu 19 (1969): 1-18.
Kawakatsu Mamoru )\\W^f- "Myodai jin-shi no shi-saku to jun-ken shi seido"

535
 Bibliography

(Market towns, water gates and the system of patrolling inspectors in the
Ming dynasty) WfWtrfr <*>*«£ jStftsJWK. Tohogaku 74 (1988): 101-15.
Kinugawa Tsuyoshi ?c ill 38 ed. Liu Tzu-chien hakase soshi kenkyu (Collected
studies on Sung history dedicated by professor James T. C. Liu in celebra-
tion of his seventieth birthday) S J ^ l t W i ^ S t W ^ f i . Kyoto: Dohosha,
1990.
Koiwai Hiromitsu 7 J N S#6^^&. "Ho-S6 no shishin ni tsuite" (On the shih-ch'en
of the Northern Sung) J t ^ O J S E ^ ' C ^ T . Shukan toyogaku, no. 48
(October 1982): 34-54.
Li Chih $"ft. Huang Sung shih-ch'ao kang-yao Ife^+iKllWi?. Taipei: Wen-hai,
1968.
Li Hsin-ch'uan $ <fr ft. Chien-yen i-lai chao-yeh tsa-chi 3*ifefcJt3fc®IIJSigE.
Taipei Wen-hai, 1966.
Chien-yen i-lai hsi-nien yao-lu ftj&W3fcSl^i?l|. Beijing: Chung-hua shu-
chu, 1956.
Li Hung-ch'i $ ^L ft. "Sung-tai kuan-yiian shu te t'ung-chi" (Statistics on the
number of Sung dynasty officials) S£ftTJfJiifc&^j$i!ft. Shih-huo yiieh-k'an
14 (September 1984): 227-39.
Li Rang $ M. Chien-yen shih cheng-chi ft fe Bvf Be IB. Sung ch'eng hsiang li
chung ting kung pieh Chi ed.
Li Kou ^ H . Chih-chiang Li hsien-sheng wen-chi WMHfe3c£.'$M. SPTK ed.
Li Kuang $}fc. Chuang-chien chi ffiffij^. SKCSCP ed.
Li Li $^J. "Chung-kuo ku-tai t'u-hsing" (On the punishment of penal servitude
in ancient China) 41 H "4f f^ fi! M . Chung-kuo hsiieh-tao pao 1 (1987):
37-38.
Li Liu-ch'ien $ j t § | . Tan-chai chi Jftflflfe. SKCSCP 1971 ed.
Li Mi-hsiin $3131. Yiin-ch'i chi ^ g S * . SKCSCP ed.
Li Shang-chiao $_h5f. Chin-shih hui-yiian j t f i ^ ^ X . Shou shan ko ts'ung-shu
ed.
Li Tao $ j H . Hsii-tzu chih t'ung-chien ch'ang-pien M S i n S ^ f t j i . Taipei:
Shih-chieh shu-chu, 1964.
Li Yu $f&. Sung-ch'ao shih-shih ^M^%. Shanghai: SKCSCP 1955 ed.
Li Yuan-pi $ x $ B . Tso-i tzu-chen f ^ g g ^ . SPTK ed.
Liang K'o-chia W%M. San-san chih HLJLJS*. SYTFCTS ed.
Liao Hsing-chih JgfjIt. Sheng-chai chi ^Sf|fc. SKCSCP ed.
Liao Kang %W\. Kao-feng wen-chi ^%.~^M. SPTK ed.
Lin I-chih Wfoit. Wang-shan chi « U 4 * . SKCSCP ed.
Ling Wan-ch'ing SHEK. Yii-feng chih 5S£;£. SYTFCTS ed.
Liu Ch'ang Wk. Kung-shih chi fc%L%. TSCC ed.
Liu Cheng ®JE. Huang-Sung chung-hsing liang-ch'ao sheng-cheng
l a S f t . Taipei: Yee Wen, 1967.
Liu Chih $ J & Chung-su chi ^ * * . TSCC ed.
Liu I-chih §U—Ik. T'iao-ch'i chi WM%k. SKCSCP 1971 ed.
Liu Tsai » J $ . Man-t'ang wen-chi ftSlilfe. SKCSCP 1979 ed.
Liu Yueh §IJ^. Yun-chuang chi B K S . SKCSCP 1971 ed.

536
 Works in Chinese and Japanese

Lo Chiin ! P # . Pao-ch'ing Ssu-ming chih JUKe3W;£. SYTFCTS ed.

Lou Yiieh 991. Kung-k'uei chi #jfe£. TSCC ed.
Lu Hsien S.M. Chia-ting Chen-chiang chih ^ ^ R f l j g . SYTFCTS ed.
Lu Yu R£j#. Wei-nan wen-chi MjS£H. SPTK ed.
Lti Tsu-ch'ien SPJLHL Huang-ch'ao wen-chien M.^R~iC&. SPTK ed.
Lii Tung-lai hsien-sheng wen-chi B^^Tfe^^C^. Chin-hua ts'ing-shu ed.
Ma T'ing-luan M S JR. Pi-wu wan-fang chi g g ? ^ ^ . SKCSCP 1975 ed.
Ma Tuan-lin ( | ^ | § . Wen-hsien t'ung-k'ao >C (ft S % . Kuo-hsiieh chi-pen
ts'ung-shu ed.
Mao P'ang ^ j § . Tung-t'ang chi 3 C ^ ^ . SKCSCP ed.
Mei Ying-fa 1ft(tf$. Pao-ch'ing ssu-ming hsii-chih ffjgHWiWS. SYTFCTS ed.
Meng Yiian-lao J x ^ . Tung-ching meng-hua lu wai ssu-chung ^ ^ ^ ^ ^ ^ f
IZ9HL Shanghai: Chung-hua shu-chu, 1962.
Ming-kung shu-p'an Ch'ing-ming chi ^5 £ H #lj jflf $j H • Shanghai: Chung-hua
shu-chu, 1987.
Miyazaki Ichisada *£ & l pf rf) /l!. "Kandai no ri-sei to Todai no bo-sei" (The Li sys-
tem in the Han dynasty and the Fang system of the T'ang dynasty) i l f t ^
I J l R j f f t ^ f & I J . Toyoshi kenkyu 21 (1962): 27-5.
"So-gen jidai no hosei to saiban kiko" (Law and judicial system in the Sung-
Yuan period), 5fc X %*it <VftftU£ Wt *'J ^F H . Toho gakuho 24 (1954):
115-227.
Mu-jung Yen-feng B $ 3 l ^ . Li-wen t'ang-chi ffli^H. Ch'ang-chou hsien-
sheng ts'ung-shu ed.
Li-wen t'ang chi J S ^ t ^ ^ . Sung Ming-chia chi-hui ed.
Muronaga Yoshizo ^ ^c 5? H . "Godai gunbatsu no keigoku kiko to setsudo
saibanden" (The system of criminal law of the military cliques in the Five
Dynasties) 31 f t * K <T> WiWMM t ifJtffiafilfll. (Kyushu daigaku bun-
gakubu) Toyo Shusaku 28 (1965): 61-75.
"Godai jidai no gun-jun in to ma-ho in no saiban" (Adjudication by the army
patroling office and the cavalry and infantry office in the period of the Five
Dynasties) 2Cftff$ft^>¥MIS t M#Ec<nWWl. Toyoshi kenkyu (Kyoto) 24
(1966): 16-38.
Nai Te-weng Wt%m. Tu-ch'eng chi-sheng f|5fe)c$£^. In Meng Yiian-lao, Tung-
ching meng-hua lu, wai-ssu chung.
Niida Noboru ^ # BB IS. Chugoku hosei-shi kenkyu (Studies on Chinese legal
history) ^^ffi^J^W^L. Tokyo: University of Tokyo Press, 1964.
Torei shui (Collection of T'ang ordinances) S ^ p ^ S . Tokyo: University of
Tokyo Press, 1964.
Okada Takehiko |5J|fflB? §£ . So-Min tetsugaku no honshitsu (The essence of
Sung and Ming philosophy) ^ B ^ ^ C O ^ R . Tokyo: Mokujisha, 1984.
Ou-yang Hsiu IftBRH*. Hsin Wu-tai shih f f K f ^ ^ t . Beijing: Chung-hua shu-
chu, 1974.
Ou-yang wen-chung kung wen-chi Kflf^^'fe&^t^. SPTK ed.
Pan Ku BEB. Han shu Slff. Po-na ed.
P'eng Kuei-nien ^ H ^ . Chih-t'ang chi ikSEifc. SKCS ed.

537
 Bibliography

Pi Chung-yu * f r i t . Hsi-t'ai chi ffi2*. SKCSCP ed.

Pien Shih & * . Yii-feng hsii-chih ilfcflfcg. SYTFCTS ed.
Saeki Tomi fl: ffi If. Sodai ni okeru ro-jogun ni oite (The establishment and
transition of the ex-convict-corps of the Song dynasty) 5^ft i- & * &t&^&£
I^V^iS V^T. In Kinugawa Tsuyoshi, ed., Liu Tzu-chien hakase soshi kenkyu
ronso, pp. 267-92.
Shen Chia-pen H ^ ^ . Li-tai hsing-chih k'ao KKf^TPJ/fe^. Beijing: Chung-hua
shu-chu, 1985.
Shen Chi-i hsien-sheng i-shu gfcfrS3fc£?t*. Taipei: Wen Hai, 1964.
Shih Chieh ~Bit. Ts'u-lai chi t i l * * . SKCSCP 1973 ed.
Shih Hao Jtffi. Mou-feng chen-yin man-lu g P * « H ? l ^ . SKCSCP 1971 ed.
Shih Neng-chih £ t g £ . Hsien-ch'un p'i-ling chih jRffHtR*. SYTFCTS ed.
Shih Su ifefM. Chia-t'ai K'uai-chi chih H f ^ # 8 f ; £ . Chia-ch'ing blockprint edi-
tion.
Sogabe Shizuo U S H P I P t t . Chugoku oyobi kotai Nihon ni okeru kyoson keitai
no hensen (The evolution of the form of villages in China and ancient
Japan) tf MR 0* "Sft B * K £ It & WttfiZMV &M. Tokyo: Yoshikawa
kobukan, 1963.
"Sodai no ketsu-kan" (Concerning feuds in the Sung dynasty) ^ f t ^ ^ ^ B B .
In Ishida Hakushi koki kinen jigyokai, ed., Ishida Hakushi shoju kinen
Toyoshi ronso. Tokyo: Ishida Hakushi koki kinen jigyokai, 1965.
"Sodai no shi-hai ni tsuite" (On the penalty of penal registration in the Sung)
5|5ft^fflKK-3l/>-C..Bunka 29 (1965): 1-23.
Sodai zaiseishi (Sung financial history) S i f ^ i t B c i . Tokyo: Dai Nippon insat-
su kobunshiki kaisha, 1941.
Ssu-ma Ch'ien ^1JSM. Shih-chi 5feoB. Beijing: Chung-hua shu-chu, 1959.
Ssu-ma Kuang W| J! }fe. Tzu-chih t'ung-chien Jf fa M $m. Beijing: Ku-chu ch'u-
pan she, 1965.
Su Ch'e jKtft. Lung ch'uan chi f|JI|^. SPTK ed.
Su Sung U^. Su-wei wen-kung chi Scgl5C&||. SKCSCP 1973 ed.
Sudo Yoshiyuki M jR ^ i.. "Godai no setsudoshi no shihai taisei" (The control
system of the chieh-tu shih of the Five Dynasties) S f t f i S K ^ ^ i E f f f t U .
Shigaku zasshi 61 (April 1952): 1-42 and (June 1952): 20- 39.
T6-S6 shakai keizaishi kenkyu (Studies on the social and economic history of
T'ang and Sung China) 0 7 R t t # S S ^ W ^ L . Tokyo: Daigaku shuppan kai,
1965.
Sun Feng-chi ^ 3 * ^ . Chih-kuan fen-chi fl»f ^IB. SPTK ed.
Sun Ying-shih WMM. Ch'in-ch'uan chih m\\M. SYTFCTS ed.
Sung Ch'i SfcSP. Ching-wen chi M ^ I I . TSCC ed.
Sung Min-ch'iu ^SSt^c. Ch'ang an chih S ^ i f e . Ch'ien-lung blockprint, 1787.
Sung ta-chao ling-chi TR^TIS^^C. Taipei: Ting-wen shu-chu, 1972.
Tai Chien-kuo IKftH. "Sung-tai te yii-cheng chih-tu" (On the system of jails in
the Sung dynasty) ^f^tfdWtWtMH. Shanghai shih fan ta-hsiieh hsueh pao,
no. 3 (1987): 89-96.
T'ai-p'ing lao-jen ^ C ^ ^ A . Hsiu-chung chin IJfflS. Hsiieh-hai lei-pien ed.

538
 Works in Chinese and Japanese

T'ang Keng IS H . Mei-shan T'ang hsien-sheng wen-chi H ill JS 5fe 4. ;£ H .

SKCSCP 1976 ed.
T o T'o W§L. Liao shih JgSfc. Taipei: San-min shu-chu, 1965.
Sung shih 3^£. Beijing: Chung-hua shu-chu, 1977.
Tonami Mamoru WM.WL. "Todai no ken-jo" (Hsien-wei in the T'ang) Jjfft^ M
ft. Shirin 57 (1974): 79-105.
Tou I SIR. Sung hsing t'ung ;$:}fiJ$S. Beijing: Chung-hua shu-chu, 1984.
Ts'ai Hsiang g $ . Tuan-ming-chi ^M^. SKCSCP 1973 ed.
Ts'ai Kan MWt- Ting-chai-chi /Sftfll. Ch'ang-chou hsien-che i-shu ti-i chi ed.
Tseng Kung # S . Yiian-feng lei-kao x f i l . SPTK ed.
Tseng Kung-liang # 2 : ^ . Wu-ching tsung-yao ftgfflff. SKCSCP ed.
Tso Ch'iu-ming £fi:W. Kuo Yii g | § . SPPY ed.
Ts'ui Tun-li gtfrii. Kung-chiao chi ^%M. SKCSCP 1972 ed.
Tu Cheng KIE. Hsing-shan t'ang-kao tt#^fil. SKCSCP ed.
Tu Cheng-sheng ttiEIH?. "Ku-tai hsing-yii tsa-k'ao" (Studies on the punishments
of antiquity) ^ft?PJSftl£%. Chung-kuo shih hsin-lun August 15, 1985.
Tu Fan tt®L Ch'ing-hsien chi JffWM. SKCSCP 1971 ed.
Tung Ssu-kao Mi5|$. Hsi-hu pai-yung ffiffiSlS. Wu-lin chang-ku ts'ung-pien
ed.
Uchida Tomoo p3 EH § W.. "Sho-sho ryo-kei no go-kei ni tsuite" (On the five
penalties in the 'Punishments of Lu' of the Book of Documents) psjilf Rjf'J<7)
Sff'J £ *3 V> X. Tohogakkai soritsu nijugo shunen kinen Tohogaku ronshu
(1972).
Umehara Kaoru ft IK ffi. "Sodai chiho shotoshi no ichimen, chin no hensen o
chushin toshite" (One aspect of small market towns in the Sung — On the
evolution of the chen) Tfcfttt^'hf&lfif <D — ffi, ^(D^Mi^^t It.
Shirin 41 (1958): 475-91.
Wang Ch'eng K 6 . Tung-tu shih-lueh ^ f P * i g . Taipei: Wen Hai, 1967.
Wang Chih B K . Hsueh-shan chi B U j * . TSCC ed.
Wang Chih-tao 3E^JI. Hsiang-shan chi ffllijUS. SKCSCP ed.
Wang Chih-wang 5.2M. Han-p'in chi MMM. SKCSCP 1975 ed.
Wang Ch'in-jo I #£ %. Tse-fu yiian-kuei ffi fff X ft. Hong Kong: Chung-hua
shu-chu, 1960.
Wang I-chih zEiSt^L. Chih-yuan ts'o-yao IftSflifti?. Hsii Chin-hua ts'ung-shu ed.
Wang Kung 3:36. Sui-shou tsa-lu H ^ $ 8 I I . Hsiieh-hai lei-pien ed.
Wang Ling 1 ^ . Kuang-ling chi R | ^ ^ . SKCSCP 1971.
Wang Fi-chih I H ^ . Sheng-shui yen-t'an lu i i * ^ ! ^ . TSCC ed.
Wang P'u 3£j$. Wu-tai hui-yao 51fti!"!?. Shanghai: Shanghai ku-chi ch'u-pan
she, 1978.
Wang Shih-p'eng i + J M . Mei-ch'i chi ftS*. SPTK ed.
Wang Tseng-yii i § J ^ . Sung-ch'ao ping-chih ch'u-t'an (An initial investigation
of the Sung military system) ^tft-^kMWfflz. Beijing: Chung-hua shu-chu,
1983.
Wang Ts'un EE#. Yuan-feng chiu-yu chih 7CB?lJ8;£. Taipei: Wen Hai, 1967.
Wang Yang 1 ^ . Tung-mou chi 3 C 4 ^ . SKCSCP ed.

539
 Bibliography

Wang Yen £jfc. Shuang-ch'i lei-kao WSkfflfc. SKCSCP 1972 ed.

Wang Ying-lin EEifiBf. Yii-hai 3E$i. Chia-ch'ing blockprint ed.
Wang Yung EE^C. Sung-ch'ao yen-i i-mou lu ^ f l ^ S I p ^ i ^ . Pai-ch'uan hsueh
hai ed.
Wei Chao $0S. Kuo yu H|g. SPTK ed.
Wen T'ien-hsiang 3fc ^ f¥ . Wen-shan hsien-sheng ch'iian-chi >C UL| jfe ^ ^ Hi.
Wan-yu wen-k'u ed.
Wen Yen-po i ^ i f . Lu-kung wen-chi $&&>CS. SKCSCP 1976 ed.
Wu Chien-fan ^ ft 31. "Sung-tai hsiin-wei liang-ssu chih" (The system of the
two offices of sheriff and patroling inspector in the Sung dynasty) 5*:f^3^11^
ffi"]ftlj. In Fa-chih shih yen-chiu shih, comp., Chung-kuo ching-ch'a chih-tu
chien-lun.
Wu Ch'ien ^IJH. Hsu kuo-kung tsou-i JrFBfiilli!. Pai-pu ts'ung-shu chi-ch'eng
ed.
Wu Ching J^fft. Chu-chou chi flr#H$$. SKCSCP 1973 ed.
Wu Po-lun jftftifii. Hsi-an li-shih shu-lueh (A brief history of Sian) ffi^Miift
IH§. Sian: Shensi jen-min ch'u-pan she, 1980.
Wu Tzu-mu ^ | § ft. Meng Hang lu j r §j£ ^ . In Meng Yiian-lao, Tung-ching
meng-hua lu, wai ssu pien.
Yang Chieh #<£. Wu-wei chi * g ^ . SKCSCP 1974 ed.
Yang Chung-liang % ft R . Hsii-tzu chih t'ung-chien ch'ang-pien chi-shih pen-
mo « « t e a * S i H i £ * * * . Taipei: Wen-hai, 1967.
Yang I 8MJg. Wu-i hsin-chi jBj^lfJflll. Pu-ch'eng i-shu ed.
Yang Wan-li H H S . Ch'eng chai chi WM%. SPTK ed.
Yin Chu f*&. Ho-nan hsien-sheng wen-chi Mffl^fe^^lfe. SPTK ed.
Yii ching ^4R. Wu-ch'i chi Kgllfe. SKCSCP ed.
Yii Son ^ ^ . Sung-jen hsiao-shuo lei-pien $K Adrift SI Hi. Beijing: Chung-kuo
shu-chu, 1985.
Yii Te-lin ^?^J85. P'ei-wei chai chi-wen fll^lS^IHI. Taipei: Commercial Press,
1972.
Yuan Fu ft ft. Meng chai chi USHfc. TSCC ed.
Yuafi Hsieh R'^. Chieh-chai chi WMM. TSCC ed.
Yuan Yueh-yu RtftS. Tung-t'ang chi ^ ! * S . SKCSCP ed.

Works in Western languages

Ames, Roger, and David L. Hall. Thinking Through Confucius. Albany: State
University of New York Press, 1987.
Austin, J. L. "A Plea for Excuses." In J. L. Austin, Philosophical Papers. Oxford:
Clarendon Press, 1961.
Balazs, Etienne. Le Traite juridique du (lSouei-chou." Leiden: Brill, 1954.
Bao Ruo-wang (Jean Pasqualini). Prisoner of Mao. New York: Penguin Books,
1976.
Bielenstein, Hans. "Chinese Historical Demography, A.D. 2-1982." Stockholm:
Bulletin of the Museum of Far Eastern Antiquities, no. 59, 1987.

540
 Works in Western languages

Bodde, Derk. "Age, Youth, and Infirmity in the Law of Ch'ing China." In
Jerome A. Cohen, R. Randle Edwards, and Fu-mei Chang Chen, eds.,
Essays on China's Legal Tradition. Princeton, NJ: Princeton University Press,
1980.
"Prison Life in Eighteenth Century Peking." Journal of the American Oriental
Society 89 (1969): 311-31.
Bodde, Derk, and Clarence Morris. Law in Imperial China. Cambridge, MA:
Harvard University Press, 1967.
Bol, Peter. "Culture and the Way in Eleventh Century China." Ph.D. diss.,
Princeton University, 1982.
Burkitt, F. C. The Religion of the Manichees. Cambridge: Cambridge University
Press, 1925.
Cahill, Suzanne E. "Taoism at the Sung Court: The Heavenly Text Affair of
1008." Bulletin of Sung and Yuan Studies, no. 16 (1980): 2 3 - 4 5 .
Carter, Thomas F. The Invention of Printing in China and Its Spread Westward. Rev.
L. C. Goodrich. New York: Ronald Press, 1955.
Chaffee, John. The Thorny Gate of Learning in Sung China. Cambridge: Cambridge
University Press, 1985.
Ch'en, Kenneth C. S. Buddhism in China: A Historical Survey, Princeton, NJ:
Princeton University Press, 1964.
"The Sale of Monk Certificates." Harvard Theological Review 49 (October 1956):
307-27.
Ch'ii T'ung-tsu. Law and Society in Traditional China. Leiden: Mouton, 1965.
Cohen, Jerome A., R. Randle Edwards, and Fu-mei Chang Chen, eds. Essays on
China's Legal Tradition. Princeton, NJ: Princeton University Press, 1980.
Confucius. The Analects. Trans. D. C. Lau. New York: Viking Penguin, 1979.
Creel, Herrlee Glessner. "Legal Institutions and Procedure During the Chou
Dynasty." In Jerome A. Cohen, R. Randle Edwards, and Fu-mei Chang
Chen, eds., Essays on China's Legal Tradition. Princeton, NJ: Princeton
University Press, 1980.
The Origins of Statecraft in China. Chicago: University of Chicago Press, 1970.
Davis, Richard. Court and Family in Sung China, g6o-i2jg. Durham, NC: Duke
University Press, 1986.
Dolby, William. "Early Chinese Plays and Theatre." In Colin Mackerras, ed.,
Chinese Theatre from Its Origins to the Present Day. Honolulu: University of
Hawaii Press, 1983, pp. 7-31.
Duyvendak, J. J. L. The Book of Lord Shang. London: Arthur Probsthain, 1928.
Ebrey, Patricia. Family and Property in Sung China: Yuan Tsai's Precepts for Social
Life. Princeton, NJ: Princeton University Press, 1984.
Elegant, Robert, "Why the Chinese Stop Crime Better Than We Do." Parade,
October, 1988, pp. 6-7.
Elvin, Mark. The Pattern of the Chinese Past. Stanford, CA: Stanford University
Press, 1973.
Engels, Frederick. The Origins of the Family, Private Property, and the State. New
York: International Publishers, 1972.

541
 Bibliography

Fairbank, John K., ed. Chinese Thought and Institutions. Chicago: University of
Clicago Press, 1957.
Finneagan, Michael. "Urbanism in Sung China." Ph.D. diss., University of
Chicago, 1976.
Fitzgerald, C. P. Barbarian Beds: The Origins of the Chinese Chair. New York:
Barnes, 1965.
Foucault, Michel. Surveiller et punir: Naissance de la prison. Paris: Gallimard, 1975.
Franke, Herbert, ed. Sung Biographies. Wiesbaden: Franz Steiner Verlag, 1976.
"Treaties Between Sung and Chin." Etudes Song, 1st series, no. 1. The Hague:
Mouton, 1970, pp. 55-84.
Fung, Yu-lan. A History of Chinese Philosophy. Princeton, NJ: Princeton University
Press, 1952.
Gernet, Jacques. Daily Life in China on the Eve of the Mongol Invasions. Stanford,
CA: Stanford University Press, 1970.
Golas, Peter. "Courier Transport System." In Harvard University, East Asian
Research Center, Papers on China. Cambridge, MA: Harvard University,
1966.
"Sung Financial Administration." Unpublished manuscript, n.d.
Goodrich, L. C. "The Development of Printing in China and Its Effects on the
Renaissance Under the Sung Dynasty." Journal of the Hong Kong Branch of the
Royal Asiatic Society 3 (1963): 36-43.
Goodrich, L. C , and Feng Chia-sheng. "The Early Development of Firearms in
China." Isis 36, pt. 1 (1946): 114—23.
Graham, A. C. "What Was New in the Ch'eng-chu Theory of Human Nature?"
In Wing-tsit Chan, ed., Chu Hsi and Neo-Confucianism. Honolulu: University
of Hawaii Press, 1986.
Griffith, Samuel B. Sun Tzu: The Art of War. Oxford: Oxford University Press,
I971-
Haeger, John W. "Between North and South: The Lake Rebellion in Hunan,
1130-1135." Journal of Asian Studies 28 (May 1969): 469-84.
Crisis and Prosperity in Sung China. Tucson: University of Arizona Press, 1975.
Hartwell, Robert. "A Cycle of Economic Change." Journal of Economic and Social
History of the Orient 10 (July 1967): 102—59.
"Demographic, Political, and Social Transformation of China, 750—1550."
Harvard Journal of Asiatic Studies 42 (1982): 365—442.
"The Evolution of the Early Sung Monetary System." Journal of the American
Oriental Society 87 (1967): 280-89.
"Markets, Technology, and the Structure of Enterprise in the Development of
the Eleventh Century Chinese Iron and Steel Industry." Journal of Economic
History 26 (March 1966): 29-58.
"A Revolution in the Chinese Iron and Coal Industries." Journal of Asian
Studies 21 (1962): 153—62.
Hightower, James. Topics in Chinese Literature. Cambridge, MA: Harvard
University Press, 1953.
Ho Ping-ti. "An Estimate of the Total Population of Sung and Chin China."
Etudes Song, 1st series, no. 1. The Hague: Mouton, 1970, pp. 33-53.
542
 Works in Western languages

Huang, Liu-hung. A Complete Book Concerning Happiness and Benevolence. Trans.

Djang Chu. Tucson: University of Arizona Press, 1984.
Hucker, Charles O. A Dictionary of Official Titles in Imperial China. Stanford, CA:
Stanford University Press, 1985.
Hulsewe, A. F. P. Remnants of Ch'in Law. Leiden: Brill, 1985.
Remnants of Han Law. Leiden: Brill, 1955.
Hymes, Robert. Statesman and Gentleman. Cambridge: Cambridge University
Press, 1986.
Johnson, David. "The Last Years of a Great Clan: The Li Family of Chao Chun
in Late T'ang and Early Sung." Harvard Journal of Asiatic Studies 37 (1977):
5-102.
Johnson, Wallace. "Criminal Procedure at the County Level in T'ang China."
Paper delivered at the conference "The Transformation of Chinese Law,"
Bellagio, Italy, 1980.
The T'ang Code: Vol. i, General Principles. Princeton, NJ: Princeton University
Press, 1979.
Kaplan, Edward. "Yueh Fei and the Founding of the Southern Sung." Ph.D.
diss., University of Iowa, 1970.
Keightley, D. N. "Public Work in Ancient China: A Study of Forced Labor in
the Shang." Ph.D. diss., Columbia University, 1969.
Kennedy, George. Die Rolle des Gestandnisses im chinesischen Gesetz- Berlin: 1939.
Kracke, Edward A., Jr. Civil Service in Early Sung China. Cambridge, MA:
Harvard Univeristy Press, 1953.
Kychanov, Evgenij. "Les Guerres entre les Sung du Nord et le Hsi-Hsia." In
Etudes Song, 1st series, no. 2. The Hague: Mouton, 1971, pp. 103-18.
LaFree, Gary. Rape and Criminal Justice: The Social Construction of Sexual Assault.
Belmont, CA: Wadsworth, 1989.
Lau, D. C , trans. Mencius. Baltimore: Penguin Books, 1970.
Lee, Thomas. Government Education and Examinations in Sung China. Hong Kong:
Chinese University Press, 1985.
Legge, James. The Chinese Classics: Vol. 1, The Analects. Hong Kong: Hong Kong
University Press, i960.
The Chinese Classics: Vol. 3, The Book of Documents. Hong Kong: Hong Kong
University Press, i960.
The Chinese Classics: Vol. 4, The She King or Book of Poetry. Hong Kong: Hong
Kong University Press, i960.
Li Chi: Book of Rites. New Hyde Park, NY: University Books, 1967.
Levy, Howard L. Footbinding: The History of a Curious Erotic Custom. New York:
Rawls, 1966.
Liao, W. K. (trans.). The Complete Works of Han Fei Tzu. London: Arthur
Probsthain, 1959.
Lieu, Samuel. The Religion of Light. Hong Kong: Hong Kong University Press,
1979-
Liu, James J. Y. The Chinese Knight Errant. Chicago: University of Chicago Press,
1967.
Liu, James T. C. China Turning Inward: Intellectual and Political Changes in the Early
543
 Bibliography

Twelfth Century. Cambridge, MA: Harvard University Press, 1987.

"An Early Sung Reformer: Fan Chung-yen." In John K. Fairbank, ed., Chinese
Thought and Institutions. Chicago: University of Chicago Press, 1957.
Reform in Sung China. Cambridge, MA: Harvard University Press, 1968.
Lo, Jung-pang. "The Emergence of China as a Sea Power." Far Eastern Quarterly
14 (i955): 489-5O3-
"Maritime Commerce and Its Relations to the Sung Navy." Journal of the
Economic and Social History of the Orient 12 (January 1969): 5 7 - 1 0 1 .
Lo, Winston. "Circuits and Circuit Intendants in the Territorial Administration
of Sung China." Monumenta Serica 3 (1974-75): 39—107.
An Introduction to the Civil Service of Sung China. Honolulu: University of Hawaii
Press, 1987.
"A New Perspective on the Sung Civil Service." Journal of Asian History 17
(1982): 121-35.
Ma, Lawrence J. C. Commercial Development and Urban China in Sung China (960-
1270). Michigan Geographical Publication no. 6. Ann Arbor: University of
Michigan Department of Geography.
Ma, Y. W., and Joseph S. M. Lau. Traditional Chinese Stories. New York:
Columbia University Press, 1978.
Macdermot, Joseph. "Charting Blank Spaces in Disputed Regions: The Problem
of Sung Land Tenure." Journal of Asian Studies 44 (1984): 13—41.
Mackerras, Colin, ed. Chinese Theatre from Its Origins to the Present Day. Honolulu:
University of Hawaii Press, 1983.
McGrath, Michael. "Military and Regional Administration in Northern Sung
China." Ph.D. diss., Princeton University, 1982.
McKnight, Brian E. "Administrators of Hangchow During the Northern Sung."
Harvard Journal of Asiatic Studies 30 (1970): 1 8 5 - 2 1 1 .
"Crime in High Places: The Case of Ch'en Chih-chung." In Kinugawa
Tsuyoshi, ed., Liu Tzu-chien hakase seshi kenkyu. Kyoto: Dohosha, 1990, pp.
499-5I5-
"Death by Slicing." Journal of the American Oriental Society, 93 (1973): 3 5 9 - 6 0 .
"From Statute to Precedent: An Introduction to Sung Law and Its
Transformation." In Brian E. McKnight, ed., Law and the State in Traditional
East Asia: Six Studies on the Sources of East Asian Law. Honolulu: University of
Hawaii Press, 1987, pp. 111-32.
"Maximum Security Under the Sung: The Facilities at Sramana Island."
Bulletin of Sung and Yuan Studies, no. 16 (1980): 8—22.
"The Military and the Law in Sung China." Second International Sinological
Conference, Transactions. Taipei: Chung-yang yen-chin yuan, 1990: 855-74.
"Patterns of Law and Patterns of Thought." Journal of the American Oriental
Society 102 (1982): 323—31.
The Quality of Mercy: Amnesties and Traditional Chinese Justice. Honolulu:
University of Hawaii Press, 1985.
"The Rebellion of Fang La." M.A. thesis, University of Chicago, 1964.
"Song Legal Privileges." Journal of the American Oriental Society 105 (1985):
95-106.
544
 Works in Western languages

"A Sung Device for Encouraging Speedy Trial." Journal of the American Oriental
Society 93 (1975): 4®3- 8 5-
"T'ang Law and Later Law: The Roots of Continuity. "Journal of the American
Oriental Society 112 ( 1 9 9 2 ) .
"Urban Crime and Urban Security in Sung China." Chinese Culture 29
(December 1988): 23-66.
The Washing Away of Wrongs: Forensic Medicine in Thirteenth Century China. Science,
Medicine and Technology in East Asia, vol. 1. Ann Arbor: University of
Michigan Center for Chinese Studies, 1981.
Village and Bureaucracy in Southern Sung China. Chicago: University of Chicago
Press, 1971.
Meadows, T. T. "Description of an Execution at Canton." Journal of the Royal
Asiatic Society 15 (1856): 55.
Miyazaki, Ichisada. "The Administration of Justice During the Sung Dynasty."
In Jerome Alan Cohen, R. Randle Edwards, and Fu-mei Chang Chen, eds.,
Essays on China's Legal Tradition. Princeton, NJ: Princeton University Press,
1980.
China's Examination Hell. Trans. Conrad Shirokauer. New York: Weatherhill,
1976.
Needham, Joseph. Science and Civilization in China. Cambridge: Cambridge
University Press, 1971.
Ng, Vivian. "Ideology and Sexuality: Rape Laws in Ch'ing China." Journal of
Asian Studies 46 (1987): 57-70.
Nivison, David S., and Arthur Wright. Confucianism in Action. Stanford, CA:
Stanford University Press, 1959.
Pelliot, Paul. Les debuts du Vimprimerie en Chine. Paris: Imprimerie nationale, 1953.
"Les Traditions manicheens au Fou-kien." Toung Pao 22 (1923): 193-214.
Pelliot, Paul, and Edouard Chavannes. "Un traite manicheen retrouve en
Chine." Journal Asiatique 18 (1911): 499-617 and 1 (1913): 261-392.
Peterson, Charles. "First Sung Reactions to the Mongol Invasion of the North,
1211 —1217." In John W. Haeger, ed., Crisis and Prosperity in Sung China.
Tucson: University of Arizona Press, 1975, pp. 215—55.
Philastre, P. L. F. Le Code annamite. Paris: E. Leroux, 1909.
Polo, Marco. The Travels of Marco Polo. Trans. Henry Yule. London: Murray,
1921.
Reischauer, Edwin. Ennin's Diary. New York: Ronald Press, 1955.
Rotours, Robert des. Traite des examens et traite de Varmee. Paris: E. Leroux, 1932.
Schafer, Edward. Shore of Pearls. Berkeley and Los Angeles: University of
California Press, 1969.
The Vermillion Bird. Berkeley and Los Angeles: University of California Press,
1967.
Schifferli, Cristoph. "Le Systeme monetaire au Sichuan vers le fin du Xe siecle."
T'oung Pao 72 (1986): 269-90.
Shiba, Yoshinobu. Commerce and Society in Sung China. Trans. Mark Elvin. Ann
Arbor: University of Michigan Center for Chinese Studies, 1970.
"The Commercialization of Farm Products in the Sung Period." Acta Asiatica,
545
 Bibliography

no. 19 (December 1970): 77-96.

"Urbanization and the Development of Markets in the Lower Yangtze
Valley." In John W. Haeger, ed., Crisis and Prosperity in Sung China. Tucson:
University of Arizona Press, 1975, pp. 13—48.
Sickman, Lawrence, and Alexander Soper. The Arts and Architecture of China.
Baltimore: Penguin Books, 1956.
Szasz, Thomas. The Myth of Mental Illness. New York: Harper & Row, 1984.
Tao, Jing-shen. "The Influence of Jurchen Rule on Chinese Political
Institutions." Journal of Asian Studies 30 (1970): 121-30.
The Jurchen in Twelfth Century China. Seattle: University of Washington Press,
1976.
"Political Recruitment in the Chin Dynasty." Journal of the American Oriental
Society^ (1974): 24-39.
Two Sons of Heaven. Tucson: University of Arizona Press, 1988.
Thompson, E. P., Douglas Hay, Peter Linebaugh, John G. Rule, and Cal
Winslow. Albion's Fatal Tree: Crime and Society in Eighteenth Century England.
New York: Pantheon, 1976.
Tittle, C. R. Sanctions and Social Deviance: The Question of Deterrence. New York:
Praeger, 1980.
Twitchett, Denis. "The Composition of the T'ang Ruling Class: New Evidence
from Tunhuang." In Arthur Wright and Denis Twitchett, eds., Perspectives
on the Tyang. New Haven, CT: Yale University Press, 1973.
"The Fan Clan's Charitable Estate, 1050-1760." In David S. Nivison and
Arthur F. Wright, eds., Confucianism in Action. Stanford, CA: Stanford
University Press, 1959.
Van Gulik, Robert. T'ang-yin Pi-shih: Parallel Cases Under the Pear Tree. Leiden,
Brill, 1956.
Wallacher, Benjamin. "The Chinese Offence of Homicide Through Horseplay."
Han-hsueh yen-chiu 1 (June 1983): 259-316.
Wang, Gung-wu. The Structure of Power in North China During the Five Dynasties.
Stanford, CA: Stanford University Press, 1963.
Wang Ling, "On the Invention and Use of Gunpowder and Firearms in China."
Isistf, pt. 2 (i947)- 160-78.
Widengren, Geo. "Mesopotamian Elements in Manichaeism." Uppsala
Universitats Arsskrift, no. 3 (1946): 7-198.
Wilhelm, Hellmut. "From Myth to Myth: The Case of Yueh Fei's Biography."
In Arthur Wright and Denis Twitchett, eds., Confucian Personalities. Stanford,
CA: Stanford University Press, 1962.
Wilson, James Q., and Richard Herrnstein. Crime and Human Nature. New York:
Simon & Schuster, 1985.
Wing, Tsit-chan, ed., Chu Hsi and Neo-Confucianism. Honolulu: University of
Hawaii Press, 1986.
Wittfogel, Karl, and Feng Chia-sheng. History of Chinese Society: Liao.
Philadelphia: American Philosophical Society, 1949.
Worthy, Edmund. "The Founding of Sung China: 950-1000." Ph.D. diss.,

546
 Works in Western languages

Princeton University, 1976.

Wright, Arthur, and Denis Twitchett, eds. Confucian Personalities. Stanford, CA:
Stanford University Press, 1962.
Perspectives on the T'ang. New Haven, CT: Yale University Press, 1973.
Wright, Hope. Geographical Names in Sung China. Paris: Ecole pratique des hautes
etudes, 1956.
Yang L. S. A Short History of Money and Credit in China. Cambridge, MA: Harvard
University Press, 1952.
Yoshikawa, Kojiro. An Introduction to Sung Poetry. Cambridge, MA: Harvard
University Press, 1967.

547
 Index
accomplices, 93, 102, 409, 423
accusations, 75, 82, 95, 126, 128, 302, 426
adjutants (chieh-chi), 312, 398
administrative misdemeanor (kung-tsui), 85
administrative officials, see capital and
court officials
administrative sanctions, 85
administrators for public order, 149,
3°5n77> 357
adultery, 82, 101—4, 287—88, 402, 500
Ah Yiin, 502
amnesties: crimes not covered, 76, 80, 163,
328, 371, 394, 478; effects of, 103, 326,
377, 401, 434-35, 486-87; "inspections
of cases," 112-13, 378, 488-89;
opposition to, 325, 490; processing of,
120, 244, 378, 399-400, 487-88; terms
of, 485—91; to induce surrender, 17,
88
amputation, 328, 329, 330, 331
An Prefecture, 235
An-hsi District, 150
An-su Prefecture, 152^3
Analects, 16
appeals of cases, 236
archers: abuses by, 165, 175-76, 184; as
criminals, 72, 142, 184; functions and
duties, 151, 170, 183; grading and
remuneration, 187-89; numbers, 145,
147; qualifications and training,
178-83; rewards for, 128
armed robbery, 89, 90, 119
arms, see weapons
army chiefs (chun-chu), 302
army patrols, 293—94, 304
arson, 122, 139, 523; background of, 104; in
cities, 95, 284, 302; in the code, 82
assault, 84, 95, 96, 98m 04; against status
superiors, 99, 498; laws against, 82, 484
548
bandit-capturing servitors, 150, 246
banditry, 85—94 and passim
barracks, 172, 186-87, 223, 3 O 9~ 1 0
base citadel, 85, 395
battery, 82, 96, 97, 99, 498
beating, 338-43 and passim; during
interrogation, 33-39
beheading, 90, 100, 132, 214, 306, 331, 334,
421, 458, 462, 463, 478; description of,
447-50
black magic poison (ku), 75, 139, 402, 423,
492
Book of Historical Documents, 12, 323, 329,
33°> 349, 3^5. 386, 453, 472, 473, 482,
485
Book of Lord Shang, 324
Book of Rites, 455
Book of Poetry, 6
booty, 90-94, 184,331,475
border officer, 144
borders, 109, 140, 163, 175, 201, 210, 215,
244, 264, 399, 416, 462; northern, 141,
151, 174, 217, 232, 262, 405, 477;
southern, 79-81
bribery, 101, 102, 169-70, 274, 284, 287,
417, 426, 510, 522-23; ofjailors, 346,
361, 367, 368, 369, 372, 376;
punishment of, 85, 288
Buddhism, 71, 77, 82, 91, 125, 326, 500
Bureau for the Compilation of Edicts, 367
Bureau of Academicians, 382
Bureau of Civil Nominations, 265, 279
Bureau of Lesser Military Assignments,
209, 210, 262, 266
Bureau of Military Affairs, 59, 148, 299,
425, 426, 429; in compiling legal rules,
168, 269, 312; in selection of law
enforcement officials, 207, 208, 217, 219,
265, 266
 Index

Bureau of Water Transport, 358 Chiang-che, 242, 433

burglary, 82, 89, 91, 289 Chiang-nan Circuit, 9, 47, 72, 153, 205,
butchering of animals, 104, 294 218,433,437
Chiang-nan East Circuit, 109, 114, 220
Chiang-nan West Circuit, 10, 109, 114,
cangue, 304, 344, 345746, 247, 264, 280
capital and court officials, 241, 253, 254, Chieh Prefecture, 114
306, 316 Chien-ch'ang Prefecture, 114
capital command, 286 Chien-k'ang Prefecture, 170m 23, 357
capital punishment, 71, 77; see also Chien-yen i-lai hsi-nien yao-lu, 465
beheading, death by slicing, Ch'ien Prefecture, 10
strangulation Ch'ien-t'ang District, 312, 319
capital region, 137, 141, 283, 347, 349~5°> Ch'ien-yii District, 478
362, 364, 371, 417, 461, 475; policing of, Chih-tao period, 305
104, 176-77, 201, 204, 208, 211, Ch'ih Prefecture, 114
289-303, 306-19; prisoners in, 112, Chin Prefecture, 431
3O3-4, 356"59> 379, 4°3> 406-7 Ch'in dynasty, 22, 122, 387, 511
capital security office, 287 Ch'in Kuan, 10, 73, 94, 511
castration, 329, 331, 332, 389 Ch'in Kuei, 42
Cavalry and Infantry Commands, 312 Ch'in Prefecture, 223
censorate, 303, 316, 348, 358, 372; Ch'in-feng Circuit, 405
investigative role of, 287, 288, 290, 372, Ching-chi Circuit, 153, 176,407
382 Ching-hsi Circuit, 153, 176, 407
censorship, 75 Ching-hsi North Circuit, 475, 478
Ch'an Prefecture, 475 Ching-hsi South Circuit, 109, 114, 167
Chang Fang-p'ing, 174, 175, 176, 393 Ching-hu Circuit, 242
Chang Kang, 373 Ching-hu North Circuit, 143, 197, 428
Chang Shou, 78, 145, 162, 170, 197, 246 Ching-hu South Circuit: criminals in, 80,
Ch'ang Ch'en, 422 114; geography and people's characters,
Ch'ang Prefecture, 177m45, 205, 207, 280 9; law enforcement in, 177, 205, 248;
Ch'ang-an city, policing of, 290—92 Man tribal people in, 80; penalties in,
Ch'ang-shou District, I77n45, 205, 428,431,433
22ini42, 278 Ching-tung Circuit: crime in, 112, 114; law
Chao Chung-heng, 304 enforcement in, 176, 262, 480; penalties
Chao K'uang-yin, see (Sung) Che-tsung: in, 129,407,425,452,475
T'ai-tsu Ching-tung East Circuit, 502
Chao P'u, 148 Ching-tung West Circuit, 475
Chao Shan-k'uo, 88, 216 Ch'ing-li period, 500
Chao Shan-liao, 430 Ch'ing-ming chi, 104
Chao Shih, 80 Ch'ing-yuan t'iao-fa shih-lei, 412, 413, 494:
Che-yu kuei-chien, 74, 95, 355 on corruption, 84, 85; information
Chekiang, 78, 96, 174 systems, 235, 241, 399; on penalties,
Chen Te-hsiu, 184, 383 336> 34°> 35°> 402, 404, 459; on
Chen-chiang Prefecture, 165, 223 prisoners, 342, 343, 356, 361, 370, 375,
Ch'en Ch'un, 277 380, 395
Ch'en Fu-liang, 135 Ch'iung Prefecture, 428, 430, 431
Ch'en Hsiao-sun, 168 ChouMi, 288,316
Ch'en Liang, 326, 332 Chou Pi-ta, 273, 274
Ch'en Prefecture, 248 Chou-hsien t'i-kang, 362
Ch'en Tung, 248 Chu Hsi, 58, 142; on human nature, 7, 24,
Ch'en Yuan, 9 55; on law enforcement, 109, 125, 153,
Ch'en Yuan-chin, 263, 264 177, 220, 221, 224
Cheng Hsing-i, 106 Chu-chi District, 357
Cheng-ho period, 153, 393, 441 Ch'u Prefecture, 144
Chi Prefecture, 10, 382 Ch'u Prefecture, 104, 378

549
 Index
Chun Prefecture, 431
Ch'un-hua period, 239
circuit intendant, 111, 153, 228—50;
inspections by, 188, 268;
law-enforcement duties, 198; personnel
duties, 209, 210, 214, 266; reviews, 437;
supervision of convict transfers, 413,
415; supervision of law enforcement,
147, 153; supervision of penal and
judicial work, 357, 365, 369, 372, 377,
380, 489; supervision of reporting
systems, 162, 355
Circuit Intendant Commendation and Reward
Laws, 168
circumstances, light or heavy, 61, 81, 93,
99-100
coastal areas, 106—7, X43> 212
Collected Decrees of the Ch'ing-li Era, 337
Collected Documents of the Five Dynasties, 293
Collected Sung Documents, 294, 297, 343—44,
381
Collection of Self-admonitions, 430
commendations, 240, 270, 271, 272, 274,
381
commissioner of overseas trade, 206
communication, 4, 5, 63, 109, 160—64
commutations, 324, 336, 337, 491—98; for
crimes by mischance or accident, 100;
for the disabled, 311; as an elite
privilege, 74, 85, 127, 371; origins of,
33O, 474
Comprehensive Mirror, 290
confiscation of property, 53, 85, 127, 236,
244, 478
Confucius, 16, 21, 55, 130, 325
contribution, 260, 495
controller-general, 66, 149119, 162, 172, 174,
246, 306
contumacy (o-ni), 458
convicts: labor functions, 395; numbers of,
396-97
cooptation, 14-15, 80, 106-7
correctness (cheng), 6, 14, 15, 18
corruption, 74, 84-85, 169, 214, 475
Council of State, 59
counterfitting, 73, 82, 94, 104, 402
Court ofJudicial Review (Ta-li ssu), 357,
358, 359, 364, 377> 44<>> 462, 484
Court of Military Inspection, 358
crime rates, 107—9
crimes: causes of, 9, 11, 12, 68, 80, 325,
510; locations of, 109-11; numbers of,
68, 107-11, 337, 442, 465-69;
responses to, 11, 14—15, 31, 88—89, 94,
217
550
criminals, numbers of, 79, 80, 106—7,
io7ni36, 109, 111-15, 111 n 156, 162,
272,517
crucifixion, 450
cutpurses, 284, 289
death by slicing, 204, 402, 451-52, 463
death penalty, 82, 85, 90, 95, 98, 446-71;
see also beheading, death by slicing,
strangulation
deliberation (1), 491
Department of State Affairs (Shang-shu
sheng), 399, 408, 502
depravity (pu-tao), 102
deserters, 486, 502; as criminals, 69, 70-72,
80, 110, 112, 201, 458; punishment of,
331> 35°>42i,46 2
detained control (chi-kuan), 398-401, 409,
412
deterrence, 15, 323-25, 328-30
deviance, 6-9, 13-15, 18, 22, 24, 31, 33,
9°>J 92
director-in-chief concurrently serving as a
patrolling inspector, 200
directors-in-chief, 209, 216, 297, 314
discord (pu-mu), 484
Discourses of the States, 329
disenrollment, 493
"distant, evil prefectures," 107, 351, 394,
412,423,428,430-31,478
divorce, 99, 102
Dreams of the Eastern Capital, 301, 302
E Prefecture, 197, 248
educational transformation (chiao-hua), 4,
14-15, 21,32,83,510-11
elders (ch'i-chang), 128, 134, 135, 138, 143,
emperor, role of, 16
"empty jails" policy, 241
escapees, 71, 81, 286, 406, 407, 409, 417,
423,421,523
examination system, 18-20, 130, 211,
254~56> 3°5n77> 495
executions, timing of, 455-60
executory officials (hsiian-jen), 253, 262, 264
exemplary gentleman (chun-tzu), 80, 193,
484
exile, 126, 128; for black magic, 76; for
corruption, 85; crimes requiring, 91,
337; for homicide, 100; for theft and
robbery, 90, 132, 142
exile with added labor, 326, 335, 391, 485,
487
 Index
extortion, 74, 82, 84, 90, 169-70, 224, 287,
289
facilitated examinations, 255-56, 260
false accusation, 82, 92, 101, 102, 128, 288
false confessions, 92, 125, 159, 204, 369
Fan Ch'eng-ta, 318
Fan Ch'un-jen, 465, 467, 469
Fan Chun, 9
Fan Chung-yen, 11, 57, 179, 473
Fan Tsu-yii, 18, 92, 325, 366, 479, 480
Fang Pao, 361
Fang Prefecture, 111
Fang Shao, 72
Feng Ching, 142, 332
fetters, 147, 343-48, 3^9* 37<>> 3^9. 4<>o>
523; construction and care of, 303,
345—46, 357; exemptions from, 364,
437; as social defense, 327, 343, 373
feuding, 96
fights, 73, 96, 98, 100, 104, 144, 160, 305,
423
Finance Commission, 59
fines, 79, 85, 213, 334, 365, 482, 496,
498-99
fire fighting, 95, 134, 287, 301-2, 309, 312,
315
fiscal intendant, 89, 231, 232—33m, 297n45,
425
Five Punishments, 331, 452, 482, 499;
definitions, 82, 329, 334, 447, 519;
origins of, 322
foreigners, 82, 395, 400
Former Kings, 22, 323, 332—33, 391, 481,
483,496,516
forts, 137, 145, 205-7, 220, 223-24, 312,
316
fraud, 82, 84, 289
frontiers, see borders
Fu Prefecture, 378
fugitives, 71, 111,416-18, 486; see also
escapees
Fukien: crime in, 17, 86, 106, 214, 247, 477;
law enforcement in, 179, 214, 218; as a
location for convicts, 407, 428, 433
gambling, 84, 104-5, X44> 289
Gangs, 109, 126, 213, 214; of bandits, 70,
72, 111,217, 269; dealing in illegal
goods, 74, 247; of pirates, 106-7; s i z e
of, 113—15, 163, 217; structure of, 12, 96
garrison towns (chen), 35, 207
garrison troop commanders, 35, 36, 37,
148, 172, 198
551
Gazetteer o/Lin-anfrom the Hsien-ch'un Period,
312, 316
General Executive Inspector's Office, 358
Generally Applicable Reward Law, 168
grave robbing, 82, 90
guarantors, 143, 311, 314, 317
guard (pao), 138
guard head (pao-chang), 138
guards for prisoners, 156, 376, 415, 417,
424
guest households (k'o-hu), 180
guild organizations, 317, 318
habitual criminals, 71, 419; .re* also
recidivism
Hainan, 79, 81, 394, 408, 412, 415, 423,
428,431
Han Ch'i, 137
Han Fei Tzu, 26, 123, 324
Han Wei, 306—7, 308
Han Wen-ti, 348, 389
Han Yuan-chi, 73
Han-yang Prefecture, 143
Hang Prefecture, 223, 254
harboring, 93, 184, 478
heavy-penalty places, 152^5, 163, 262,
305-6, 402, 474-81
heavy-rod wielders, 341
heresy, 82, 139, 169, 246, 288, 458
Historical Records, 123, 387
Ho-pei Circuit: law enforcement in, 141,
149119, 176, 201, 480; laws in, 129, 475;
registration in, 405, 435
Ho-pei West Circuit, 140, 150
Ho-tung Circuit, n o , 174, 235, 236, 240,
433; law enforcement in, 141, 151, 176,
211; laws in, 168, 176, 211, 235, 236,
240, 269, 433
homicide, 100-103 anc ^ passim; through
accident or misadventure (wu-sha), 99,
100; in an affray (tou-sha), 98—99, 100,
291, 458; deliberate (ku-sha), 100,
369-70, 458, 501; in horseplay (hsi-sha),
100; premeditated (mou-sha), 76, 100,
101, 102,423,458,501
horse thieves, 70, 79, 202, 401
house arrest, see detained control, registered
control
Hsi Prefecture, 404
Hsi-an District, 188
Hsi-ning period, 305, 309, 452
Hsiang Prefecture, 431
Hsiang-fu District, 177m 45, 204, 295, 310
Hsiang-shan District, 177,
 Index
Hsieh Ching-wen, 308
Hsien-p'ing period, 160
Hsing-kuo Prefecture, 109
Hsiu Prefecture, 223
Hsiu-ning District, 109
Hsu Ching-heng, 204
Hsu Hsu, 248
Hsu Jan, 369
Hsu Ying-lung, 247, 516
Hsiian Prefecture, 101, 378
Hsiieh Chi-hsCian, 109, 324, 391, 515
Hsun Tzu, 7-8
Hu Su, 179
Hua Chen, 25
Huai-nan Circuit, 242, 396, 475, 480
Huai-nan East Circuit, 72
Huai-nan West Circuit, 102, 107
Huai-yin District, 72
Huang Kan, 143
Huang Liu-hung, 353, 361, 367, 368
Huang-yu period, 223
Hui Prefecture, 207, 216, 396
human nature, 11, 24, 25
Hunan, 76, 222, 415, 416
Hung Mai, 453
I Prefecture, 223
I-chang District, 223
I-ch'eng District, 109, 205
I-chou Circuit, 431
illegal salt dealing, 73, i66nio9, 200, 481
illicit goods, 86n8o, 87, 94, 202, 286, 419,
481,496
imperial armies (chin-chun), 394-98 and
passim
imperial relatives, 99, 256—57, 261; as
criminals, 74, 286; privileges of, 498,
491,492
incest, 492
indemnity payments, 498—99
Individual Prefecture Individual Circuit
Commendation and Reward Laws of the
Ch'un-hsi Period, 168
information systems, 122-29, *39> J59>
243, 327; logs of activities, 134, 238,
272, 379, 412; registers of information,
152, I52n25, 153, 186, 238
informers, 89, 125, 128, 327, 417, 478
inquests, 155-57; personnel, 148, 238, 240,
307, 309; on prisoners, 370, 371, 372
inspections of cases, 485, 488-89
inspector in chief, 300, 302
inspectors general, 152, 218, 398
inspectors general for seizing bandits, 152,
200, 212, 2 1 3 , 2 1 5
552
Institute of History, 382
Institutes of Chou, 331
investigation, 157—60 and passim
irregular officials (she-kuan), 258, 495
jail functionaries, 157, 158, 184, 361
jails, 239, 303-4, 309, 353-84
Jen-ho District, 319
joint adjudication, 53, 102, 124, 401, 402
J u Prefecture, 166, 478
judicial abuses, 12, 238, 458, 489
Judicial Control Office (Shen-hsing yuan),
428, 460
judicial intendant, 233-50; supervision of
law enforcement, 80, 177, 207, 222, 308;
superivision of penal and judicial
action, 304, 367, 370, 378, 380, 381,
408, 429, 487
Jung Prefecture, 223
juvenile criminals, 70, 73, 95, 112, 124, 125,
166
Kaifeng, 38, 166, 283—320, 459; numbers of
prisoners in, 112-13, 285, 288, 303-4,
356, 358, 377. 379, 4o6, 437; policing of,
140, I77ni45, 201, 206, 211, 294-303,
305-11
kidnapping, 77, 119-20, 139, 244
K'uai-chi hsu-chih, 250
Kuang Prefecture, 206, 216, 219, 297n45
Kuang-hsin Prefecture, 152^3
Kuang-nan, 164; criminals from, 407, 428,
43°J 431» 433> 44°; a s a place of
registration, 95, 159, 395, 413, 415, 416,
423; policing in, 141, 258, 462
Kuang-nan East Circuit, 81; as a place of
registration, 412, 424; policing in, 106,
206, 207, 212, 216, 248
Kuang-nan West Circuit, 81; criminals in,
12, 372; policing in, 212, 223, 248
Kuei Prefecture, 223
Kuei-yang Prefecture, 114
K'uei-chou Circuit, 437
K'un-shan District, 168, 362
Kuo Prefecture, 477
large guard (ta-pao), 138
large guard head (ta-pao chang), 138, 317
Later Chin, 349, 390, 456
Later Chou, 35, 37; law code of, 75, 334,
339J 34!> 34^> 362; and the military,
J96, 391; penalties under, 92, 437, 451;
policing, 134
Later T'ang, 293-94, 346, 363
 Index

law-enforcement personnel: numbers of Lii I-chien, 304

officers, 200, 202, 204-5, 280, 281, 300, Lii Tsu-ch'ien, 293
308; numbers of subordinate personnel,
202, 220, 2 2 i n i 4 4 , 301, 302, 3 1 1 , 3 1 2 , Ma Mo, 429—30
3i453i8 ma-ch'ien san-pu reward policy, 273—74
laws, types of, 23—24 Man tribal people, 79, 222, 498
lawsuits, 13, 317 Manichaeanism, 11, 77-79
legal cases, numbers of, 316 Mao P'ang, 167
legalists, 21, 22, 24 market towns, 205, 2O5n52, 412
Lei Prefecture, 372 Mencius, 5, 6, 8, 21, 79, 321; on human
Li Chin, 248 nature, 7, 8, 24
Li Ching, 429 Meng-liang lu, 312, 315
Li Kou, 86 merchants, 73, 146, 206, 515
Li Kuang, 10, 12, 239, 280 Metropolitan Cavalry Command, 286, 298,
Li Mi-hsiin, 9 358, 364
Li O,304 Metropolitan Infantry Command, 286, 299,
Li tribal peoples, 79, 431 358, 364
Li Tao, 483 Mi Prefecture, 178
Li Yen, 464 military administrator (ch'ien-hsia), 216,
Li Yuan-pi, 189, 342, 345, 359, 362, 363, 2971145
373,379-80 military administrator offices, 242
Li-chou Circuit, 378, 431 military agricultural colonies, 395, 441
Liang Tao, 18 military commissioners, 36-37, 88, 1 4 8 ^ ,
Liang-che Circuit, 130, 144, 165, 278, 435; 292
crime in, 104, 114, 128; jail facilities, military examinations, 210, 212, 260, 264,
357, 365, 378; law enforcement, 167, 267^0
188, 206, 207, 218, 223, 280 military inspectors (chiin hsiin shih), 300
Liang-Huai, 107 military intendants, 89, 151, 232—33, 462
Liao Hsing-chih, 136, 178, 186 military law, 196, 306, 391, 457, 474
Liao Kang, 17 military sheriffs, i66niO9, 201, 262—64, 480
Lien Prefecture, 223 militia organizations, 136-45, 151, 394
lieutenant (jya-chiao), 304, 3O5n77 Min District, 510
light rod, 82, 90, 97, 154, 163, 291, 316 Ming Prefecture, 177, I77ni5i, 205, 207,
Lin Tan, 454 223
Lin-an, 73, 283—320, 379 Ministry of Finance, 168, 269
Ling-nan, 431,435, 478 Ministry ofJustice, i68ni2O, 440, 441, 490;
Ling-pei, 431 on imprisonment, 356, 361; reviews of
Liu An-shih, 352 cases, 103, 402; reviews of law
Liu Chang, 510 enforcement and judicial performance,
Liu Chih, 22, 516; on law enforcement, 180, 162, 235, 379, 382, 408, 409, 424, 429,
181, 223; on the origins of crime, 72, 438, 463; rule-making power, 168
110, 142 Ministry of Personnel, 128, 212, 265, 479,
Liu I-chih, 522 480
Liu Tsai, 357 Ministry of War, 299
Liu Yueh, 10, 121, 510 Ministry of Works, 404
local elites, 89, 135, 141, 142, 267; as minority peoples, 137, I48n4; see also Li,
oppressors, 122, 175, 513; role in Man tribal peoples
maintaining order, 130—31 misappropriation of funds, 85, 88
local soldiers (t'u-ping), 218, 219—20, 247, mitigation of punishments, 474, 482
315,318 mourning-degree relatives, 53, 99, 236, 370,
looting, 95, 302 492, 498, 501
Lou Yueh, 21, 333, 483 Mu-jung Yen-feng, 240, 242, 369, 372
Loyang, 293-94 Murder, see homicide
Lu Prefecture, 107, 498 mutilating punishments, 328-33, 387
Lu Yu, 9, 120, 222 mutual responsibility, 121 — 24

553
 Index
mutual security groups, 95, 105, 121-24,
512; as information-gathering agencies,
29, 101, 123; as law enforcers, 120; as
militias, 96, 142; as watchdog units, 70,
101, 123
Neo-Confucians, 24, 55
night patrols, 118, 302, 314, 317
Ning-kuo District, 109, 205
Ning-tu District, 280
non-Han criminals, 79-81
"Notes on Prison Life," 361
office chief (ya-ch'ien), 416
Office of Public Order, 381
Offices of Chou, 144
ordinance (ling), 61, 70
organized crime, 289
Ou-yang Hsiu, 26, 27, 79
"outside the career" (liu-wai), 258, 279, 495
pacification commissioner, 221
Palace Cavalry and Infantry Command,
3H> 358
Palace Command, 298, 301, 364
Pao Prefecture, 152^3
Pao-chia militia, 123, 136—45, 188, 305,
311, 317; background and evolution,
137-38; paramilitary aspects, 96, 177,
202, 262
Pao-wu militia, 124, 136, 143, 145
patroling, 118, 138, 151—55, 183—84, 206,
287, 291, 300, 309
patroling inspector (hsiin-chien), 198-218
and passim
patroling societies (hsun-she), 143, 317
peasantry, 131-32, 141, 142
penal officer ordinances (yu-kuan ling), 341,
343
P'eng Kuei-nien, 10, 86, 357
petition (ch'ing), 491-93, 499
petty man (hsiao-jen), 104, 484
physical examinations, 208, 265, 267
Pi Chung-yu, 297, 308, 309
Pien, 293, 294
Pien Canal, 206
Ping Prefecture, 151
piracy, 105—7, *43> 17811161, 214, 218, 407;
inland pirates, 107, 218, 517; numbers
of pirates, 106, 114, 517
Po-chai pien, 72, 74
poison, 10, 75, 82, 139, 188
police offices (chun hsunyuan), 300
police posts, 287, 312, 315
554
post stations, 71, 137, 147, 216
precedents (li), 62, 107; on rewards,
i68ni2o; in setting penalties, 107, 436;
use of to reduce or defer penalties, 324,
457,483
principal criminals, 93, 409
prison citadel commands, 337, 350, 395,
43i
prison instruments, 74, 239, 303, 357, 370
prisoners: abuses of, 157-60, 365, 367-72;
deaths of, 125, 126, 336, 365, 370-73,
383, 427; health care of, 345, 363-67,
383; numbers of, 112-13, 377-79
private infractions (ssu-tsui), 85, 265, 494
privileges, 337, 491—98
prostitution, 104, 289
protection (yin): as an entry into the civil
service, 208, 210, 257-58; of imperial
clan members, 261, 286; use of to escape
penalties, 495, 4 9 ^ 7 4 , 523
provincial armies (hsiang-churi), 394-98 and
passim
P'u Prefecture, 475
public order offices, 211, 3O5n77
punishments: conversion of, 335—36;
evolution of, 322—23, 333-34
"Punishments of Lu," 323, 329, 385, 386,
473, 474, 482
rape, 101, 103-4, i39> 392, 423, 502, 523
rations, 224
rebellion, 82, 423, 492, 499
recidivism, 204, 350, 4O5n79, 487;
ineligibility for mercy, 4 9 5 ^ 4 , 502;
registration as a penalty, 392, 401, 420,
423, 438, 478; state's response to, 71
recorders, 63, 148, 164, 165, 166, 212, 254,
268, 277
records, i n , 160-64, 240
Records Office, 288
redemption, 496, 500
reduction (chien), 491-93
reduction of penalty, 76, 324, 370, 424, 481,
489, 499-500
reformation, 8, 17, 20, 22, 323, 439
registered control (pien-kuan), 244, 401, 419,
484, 478, 500, 501
regulations (ko), 61, 168, 269, 393
rehabilitation, 325-27; see also reformation
religious groups, 75-79; see also Buddhism,
Manichaeanism, Taoism
remuneration, 277-78
reports, 239, 370, 373, 408
resident labor (chti-i), 393
retribution, 325-26
 Index
8 6
review systems, 96, 234-37, 339> 35 > 37 >
460, 462-63
Reward Law for the Prefectures of the Various
Circuits, 168
Reward Laws for the Capture of Bandits in
Cheng and Hua Prefectures, 168, 269
reward systems: amounts of rewards, 126,
138; for effective judicial performance,
245; for information, 126, 128;
law compilations on, 168, 269; for
law-enforcement agents, 89, 138, 213,
217, 269-71, 273-75, 309; to motivate
action, 93, 130, 144, 162, 163, 212-3,
240-41, 287; sources of rewards, 127,
138, 144
rites, 15, 20, 165,317
Rites of Chou, 349, 482, 483, 491, 499
robbery by strength (ckiang-tao), 89, 90
"Rules on Seizing Bandits" (Pu-tsei t'iao),
268
rural areas, 160; abuse of country people by
law enforcers, 154, 184; policing system
in, 138, 196, 200-201, 205, 206, 220
Sage Emperors, 25, 30, 352, 483
sailors, 219
salt smuggling, 247, 318, 423
San-ch'ao pao-hsiin, 304
San-shan chih, 250
Sang Shih, 166, 217
seacoasts, 107, 206, 219, 264
Secretariat Department (Chung-shu), 235,
426, 439, 460
Secretariat-Chancellery, 58, 59
sedition, 23, 82, 452, 492
self-renewal (tzu-hsin), 15, 22, 481, 490, 501,
502
Senior Officials' Court, 164
service agents, 209, 210, 212, 214, 267, 268,
369
servitors, 212, 307; law enforcement
positions of, 150, 151—52, 202, 207, 300,
369
servitors major, 260, 264, 297
servitors minor, 208, 260, 262, 264, 297,
3<>9>4i2
sexual crimes, 82, 101; see also adultery,
rape
Shan Prefecture, 475
Shang Chih-chung, 150
Shang dynasty, 386
Shang Prefecture, 431, 478
Shang Yang, 123
Shang-tang District, 235
Shantung, 243, 262, 394, 412, 421
555
Shao-hsing (city), 315
Shao-hsing period, 417
Sheng-shui yen-t'an lu, 462
Shensi, 203, 452; criminals in, 433, 438;
military in, 174, 180, 212, 223, 224;
regional rules, 168, 269, 477
sheriff (hsien-wei), 147-70 and passim
Shih-wu chi-yuan, 294, 304, 316
Shou Prefecture, 102
slaves, 98, 104, 370, 409
smuggling, 87, 200
social defense, 326
sorcery, 76
specifications (shih), 62, 63
sponsorship, 265n6*9, 266, 270, 27in86
Sramana Island, 132, 351, 402, 412,
420-30; conditions on, 243, 414,
423—26; types of criminals sent to, 169,
215,421,423,431
Ssu Prefecture, 71, 206, 326
Ssu-ma Kuang, 141, 352, 483, 490
stalwart men (chuang-ting), 128, 134, 135,
170,311
State Finance Commission, 112
statutes (lu), 61, 62
statutes of limitations, 486
strangulation, 331, 334, 447-50, 463;
crimes calling for, 84, 90, 98, 99, 100,
132, 214, 237; method, 449-50
students, 496
Su Ch'e, 9, 258, 262, 263, 308, 479, 480
Su Prefecture, 170, 362, 475
Su Shih, 55, 181, 430
Su Sung, 326, 439, 455, 522
Sui District, 167
Sui-shou tsa-lu, 184
Sun Sheng, 14
Sun Shih, 395
Sung Ch'i, 26
Sung Code (Sung hsing-t'ung): origins of,
334; on penalties, 484, 494, 499, 501; on
prisoners, 343, 346, 362, 364, 403, 417,
454
Sung History (Sung shih), 395, 475, 482, 496;
on crimes, 67; on judicial abuses, 367,
377; on law-enforcement personnel, 316;
on legal process, 234, 239, 367
(Sung) Che-tsung, 41; Chen-tsung, 38, 40,
273, 326, 457, 459, 461, 482, 483;
Hui-tsung, 41, 42, 60, 440, 479;
Jen-tsung, 40, 52, 233, 309, 438, 452,
483, 484, 496, 500; Kao-tsung, 42, 324,
325> 326, 344> 497; Li-tsung, 377;
Shen-tsung, 257, 334, 479, 484;
on incarceration, 371, 377, 426, 429,
 Index

(Sung) (cont.) tours of duty, 223, 314

438, 440; on law-enforcement towns, 152, 154, 200—201, 220
changes, 72, 176, 261, 303, 306, 479; township leaders (li-cheng), 133, 134
pao-chia reforms, 136, 137, 188; Traditional Rewards for the Northwestern
reforms under, 59, 139, 306; Prefectures, 168
T'ai-tsu, 31, 35, 36, 37, 230, 297, 333; training of law-enforcement personnel, 140,
Ta'i-tsung, 367, 380, 488; Ying-tsung, 141, 142, 202—203, 222
4i, 137 transfer of convicts, 242, 344, 345, 406,
Sung-ch'ao yen-i i-mou lu, 160, 495 412-13,439,523
superior guard (tu-pao), 136, 152, 311 treason, 452, 492
superior guard leaders (tu-pao cheng), 125, Tsai Prefecture, 483
153 Ts'ai Ching, 440, 479
superior joint patroling inspector, 151 Ts'ai Hsiang, 9, 179, 218
superior patroling inspectors, 199, 211 Ts'ai K'an, 12, 72, 184, 210, 248, 273, 519,
supernumerary officials, 265 523
supervisor of bandit capturing, 216 Ts'ao Prefecture, 475
supervisors of militia (chien-ya), 207, 209, Tseng Kung, 183
212, 215, 220, 297, 304, 425 Tseng Pu, 93, 98nio5, 152^5, 331-32, 334
supplementary officials (pu-kuan), 495 Tso-i tzu-chen, 111
surrender and confession, 417, 486, 501-3 Ts'ui Tun-li, 70, 72
Szechuan, 232; criminals from, 407, 420, Tu Cheng, 255, 279
433, 438; law-enforcement system in, Tu Fan, 109, 205
180, 212, 223, 224; legal process, 203 Tu Yen, 235
Tu-ch'ang District, 109, 153, 220
T'ai Prefecture, 114, I78ni6i, 223, 365 Tzu-chou Circuit, 431, 498
Tan Prefecture, 428, 430, 431
Tang-t'u District, 278
Taoism, 77, 82, 500 vagabonds, 73, 136, 300
tattooing, 393, 394; history of, 329, 331, vendetta, 390, 513
332, 349; of officials, 159, 215, 423; of vengeance, 123, 326, 401, 420, 483-85
soldiers, 174, 348, 349; sole punishment village officers, 127, 133-36, 151, 170, 236,
for theft, 340, 350; supplementary 401
punishment with registration, 159, 328, violated regulations: through mistake, 158;
349>35°>4°9> 43^441 for private motives, 158, 161
Ten Abominations, 423, 457, 458, 484, 486,
492, 5 O 1
Teng Prefecture, 114, 424, 425, 428, 429, Wan-an Prefecture, 428, 430, 431
43i>502 Wang An-shih: and the pao-chia, 137, 139,
terms in office, 170, 170m 23, 179, 209, 212, 142, 177; on penalties, 332, 479, 485,
223, 250, 270 490, 502; as a reformer, 41, 168, 180,
theft: cases, 91—92, 166; conditions leading 261
to, 104, 119; definition, 89, 90—93; WangChi-yu, 168
methods, 118; prevention, 118, 119, 139, Wang Chih, 11, 26, 96
243; punishments for, 90-91, 92, 142 Wang Chih-wang, 114, 269, 513
Three Commands, 299, 314, 364 Wang Hsiang, 17
Three Dynasties, 349, 390 Wang Kung, 184
Tien-fu period, 349, 390 Wang Pi-chih, 462
T'ien-sheng period, 274, 337 Wang Shu, 235
time limits: for arresting fugitives, 134, 163, WangSsu, 158
181, 270, 272, 276; for processing cases, Wang Tseng-yii, 394
377; for processing rewards, 127; for Wang Yang, 263
seizing witnesses for interrogation, 157; Wang Yen, 350, 420
for submitting police reports, 162; for Wang Yung, 104, 160, 272, 277, 484
surrender, 502 ward chiefs, 292, 311
T'ing Prefecture, 407 Washing Away of Wrongs, 336

556
 Index

weapons: of law-enforcement personnel, 96, Yai Prefecture, 428, 430, 431

138, 142, 171-72, 186-87; people's Yang An-kuo, 483
possession of, 119, 129, 131, 135; Yang Prefecture, 378
restrictions on, n o , 131-32, 291; theft Yang Shou-chen, 452
of, 72, 142; use in crimes, 89, 90, 97 Yang Wan-li, 207
Wei-nan District, 166 Yangtze River, 88, 107, 183, 218, 294, 326,
Wen Min-chung, 125 433
Wen Prefecture, 114, 128, I78ni6i, 223 Yen Prefecture, 73, I49n6, 167, 170m 23
Wen T'ien-hsiang, 382 Yii Ching, 10, 12, 26, 79, 80, 114, 213, 222
Wen Yen-po, 322 Yii-ch'ien District, 318
Wen-hsien t'ung-kao, 114, 297 Yu-shan District, 74
witchcraft, 288, 392, 458 Yuan Fu, 144, 145, 512
witnesses, 353, 355, 369, 523 Yuan Hsieh, 183, 186
wives, 92, 95, 99, 100 Yuan Ts'ai, 91, 118, 120, 129
women, 76, 345, 493, 494, 496; as Yuan Yiieh-yu, 263, 518
criminals, 101, 349, 357, 359, 387, 487; Yiian-feng period, 162, 380, 393, 402, 439,
as prisoners, 365, 377, 380, 401-2, 403, 440, 441
501; as victims of crime, 92, 103-4, I 2 ^ Yiian-yu period, 162, 308, 367, 479
Wu Ching, 26, 93, 109, I48n3, 161, 267 Yiian-yu Proportional Reduction Law, 162
Wu-i District, 378 Yiin Prefecture, 7 6 ^ 1
Wu-lin chiu-shih, 288, 316 Yung Prefecture, 212, 223
Wu-ning District, 357 Yung-an District, 217
Wu-p'ing District, 378 Yung-hsing chiin Circuit, 111, 114, 477,
478

557