Вы находитесь на странице: 1из 189

Chinese empires were established by force of arms, but sustained by religious

rites and intellectual theory. The four centuries from 206 BC to AD 220
witnessed major changes in the state cults and the concepts of monarchy,
while various techniques of divination were used to forecast the future or to
solve immediate problems. Michael Loewe examines these changes and the
links between religion and statecraft. While both mythology and the
tradition nurtured by the learned affected the concept and practice of
monarchy throughout the period, the political and social weaknesses of the
last century of Han rule bring into question the success that was achieved by
the imperial ideal. Nevertheless that ideal and its institutions were of prime
importance for the understanding of Han times and for the influence they
exercised on China's later dynasties.
University of Cambridge Oriental Publications 48
Divination, mythology and monarchy in Han China
A series list is shown at the back of the book
Divination, mythology and
monarchy in Han China
Cambridge University
Published by the Press Syndicate of the University of Cambridge
The Pitt Building, Trumpington Street, Cambridge CB2 lRP
40 West 20th Street, New York, NY 10011--4211, USA
10 Stamford Road, Oakleigh, Melbourne 3166, Australia
Faculty of Oriental Studies, University of Cambridge, 1994
First published 1994
Printed in Great Britain at the University Press, Cambridge
A catalogue record for this book is available from the British Library
Library of Congress cataloguing in publication data
Loewe, Michael.
Divination, mythology and monarchy in Han China / Michael Loewe.
p. cm. - (University of Cambridge oriental publications; 48)
Includes bibliographical references and index.
ISBN 0 521 45466 2 (hardback)
I. Religion and state-China. 2. China-History-Han dynasty,
202 BC-220 AD I. Title. II. Series.
BL65.S8L64 1994
299'.512177'09014-<lc20 93-28327 CIP
ISBN 0 521 454662 hardback
List of figures page x1
Pref ace xm
Acknowledgements xv
List of abbreviations xvi
Introduction: the history of the early empires 1
1 Man and beast: the hybrid in early Chinese art and literature 38
2 Water, earth and fire: the symbols of the Han dynasty 55
3 The Han view of comets 61
4 The authority of the emperors of Ch'in and Han 85
5 The term K'an-yii and the choice of the moment 112
6 Imperial sovereignty: Tung Chung-shu's contribution and his
predecessors 121
7 The cult of the dragon and the invocation for rain 142
8 Divination by shells, bones and stalks during the Han period 160
9 The oracles of the clouds and the winds 191
10 The Almanacs (Jih-shu) from Shui-hu-ti: a preliminary survey 214
11 The Chiieh-ti games: a re-enactment of the battle between
Ch'ih-yu and Hsiian-yiian? 236
12 The failure of the Confucian ethic in Later Han times 249
13 The imperial tombs of the Former Han dynasty and their
shrines 267
List of Han emperors 300
Glossary 302
Bibliography 317
Index 343
Map of the Han empire 195 BC; reproduced from The
Cambridge History of China, vol. I, p. 125. page xviii
2 Map of the Han empire AD 140; reproduced from The
Cambridge History of China, vol. I, pp. 252-3. x1x
3 The silk manuscript from Ch'u; from Barnard (1972-3), vol. II,
folded sheet in rear cover pocket. 43
4 The twelve peripheral figures of the Ch'u silk manuscript; from
Barnard (1972), p. 2. 44
5 Examples of Tongue and Antler figures; from Barnard (1972),
p.18. 47
6 Decorative figures from the coffins of tomb no. 1, Ma-wang-tui;
from KK 1973.4, p. 249, fig. 2. 48
7 The Queen Mother of the West, with suppliants; reproduced
from TOho gakuhO number 46 (Kyoto), March 1974, p. 63,
fig.20. 50
8 Pairs of birds and other animals in the art of Ch'u; from
Barnard (1972), pp. 14-15. 51
9 (i) J'he lacquered screen from Wang shan; from Barnard (1972),
p. 17; (ii) Reliefs from I-nan, Shan-tung; from Finsterbusch
(1966-71), table 95. 53
10 Illustrations to the Classic of the Mountains and the Lakes;
from an edition of 1893, which reproduces woodcuts of before
1667. 54
11 and 12 Illustrations of comets, from the silk manuscript from
Ma-wang-tui; from WW 1978.2, plates 2, 3. 63/4
13 Figures seen in the clouds, from the silk manuscript from
Ma-wang-tui (for source, see p. 192 note 2). 194
14 Figures seen in the clouds, from Chan yiin ch'i shu (for source,
seep. 198 note 25). 199
15 Transcription of strips nos. 730-42, from Shui-hu-ti; transcribed
by Dr Howard Goodman and reproduced from Asia Major,
third series, vol.I, part II, 1988, p. 6. 219
16 The chiieh-ti as portrayed in the San ts'ai t'u hui. 239
xu List of figures
17 Scene from a relief from a Han tomb in Nan-yang, interpreted
as chiieh-ti; from WW 1973.6, 19 and 21, fig. 3.
18 The imperial tombs of the eleven Former Han Emperors; after
Liu Ch'ing-shu and Li Yu-fang (1987), p. 2, fig. 1.
I am glad to express my thanks to the Publications' Committee of the Faculty
of Oriental Studies, Cambridge, for the opportunity to reprint these articles.
They concern three themes that recur in most aspects of China's early imperial
history, and which drew the attention of many of the leading men of the day,
i.e., the ever present call of mythology, the prevalence of divination in public
and private life and V1e development of concepts of imperial sovereignty. To
these I have added an introductory chapter which seeks to show how the study
of this period of history has developed in the last few decades, and in doing so
to acknowledge my deep debt to those scholars from Asia, America, Australia
or Europe who have made such developments possible. It is a matter of
satisfaction that a number of the articles that are reproduced here owed their
origin to invitations to contribute to volumes published in honour of some of
those colleagues. Tributes to Werner Eichhorn, Karl Bunger, Derk Bodde,
Anthony Hulsewe and Tilemann Grimm are thus included here as chapters 2,
4, 7, 11 and 12.
As each of the following chapters was written for publication independent-
ly, there is necessarily some degree of duplication, which has not been
removed in the process of editing for inclusion in this volume. The chapters
thus remain as separate studies; and while the later ones build on themes set
out at earlier stages, they are not dependent on one another and may be read as
individual items. New information or references that have become available
since the original publication have usually been added to the notes within
square brackets; on a few occasions extra information has been placed
separately in an addendum to a chapter. As different conventions and
methods of reference had been required for different publishing houses, it has
been necessary for the sake of consistency to redraft all the footnotes, and to
provide a complete bibliography.
The following have kindly given permission to reprint from books or
periodicals that they have published or edited:
The Editor of Asia Major, Princeton
Attempto Verlag, Tubingen
E. J. Brill and the Editors of T'oung Pao, Leiden
! .
xiv Preface
The Editor of Early China, Chicago
Gesellschaft fi.ir Natur-und Volkekunde Ostasiens, Hamburg
Otto Harrassowitz, Wiesbaden
Hong Kong University Press
The Editor of Numen, Bremen
Ostasiatiska Museet, Stockholm
The School of Oriental and African Studies, London
Where it is necessary to distinguish between homophones, letter references are
appended proper names terms, both in the text and the glossary. These
are usually at the first occurrence of a term in a chapter, but they are
not necessanly repeated where ambiguity is unlikely; nor are they included for
well-known expressions such as dynastic titles. The names and titles that are
given in the bibliography are not repeated in the glossary.
The author thanks the editors and publishers of the following books and
jol,lmals in which the articles collected in this volume have previously
1. Numen, vol. 25, fascicule 2 (1978), 97--117.
2. Nachrichten der Gesellschaft fur Natur- und Volkerkunde Ostasiens/
Hamburg, vol. 125 (1979), 63-8 (this article was dedicated to Werner
Eichhorn on his eightieth birthday).
3. Bulletin of the Museum of Far Eastern Antiquities, vol. 52 (1980), 1-31
4. Dieter Eikemeier and Herbert Franke (eds.), State and law in east Asia;
festschrift Karl Bunger (Wiesbaden, 1981), 80--111.
5. Early China, vol. 9/10 (1983-5), 204-17.
6. S. R. Schram (ed.), Foundations and limits of state power in China
(London and Hong Kong, 1987), 33-57.
7. Charles le Blanc and Susan Blader (eds.), Chinese ideas about nature and
society: studies in honour of Derk Rodde (Hong Kong, 1987), 195-213
8. T'oung Pao, vol. 74 (1988), 81-118.
9. Bulletin of the School of Oriental and African Studies, vol. 51, part 3
(1988), 500-20.
10. Asia Major, 3rd series, vol. 1, part 2 (1988), 1--27.
11. W. L. Idema and E. Zurcher (eds.), Thought and law in Qin and Han
China: studies dedicated to Anthony Hulsewe on the occasion of his
eightieth birthday (Leiden, 1990), 140-57.
12. Peter Kuhfus (ed.), China Dimensionen der Geschichte: Festschrift fur
Tilemann Grimm anliisslich seiner Emeritierung (Tiibingen, 1991 ), 179-
13. T'oung Pao, vol. 78 (1992), 302--40.
The following abbreviations are used in the notes and bibliography
Asia Major
Bulletin de !'Ecole Franr;aise de !'Extreme Orient
Bulletin of the Institute of History and Philology
Bulletin of the Museum of Far Eastern Antiquities
Bulletin of the School of Oriental and African Studies
Michael Loewe, Crisis and Conflict in Han China 104 BC to
Ch 'un-ch 'iu fan-lu
Ch 'ien-ju lun
Cambridge History of China, vol. I
Committee for the Preservation of Ancient Monuments
Chin shu
Early China
Fan Sheng-chih shu
Bernhard Karlgren, Grammata Serica Recensa
Homer H. Dubs, The History of the Former Han Dynasty
Hou Han chi
Hou Han shu
Hou Han shu chi-chieh
Harvard Journal of Asiatic Studies
Han shu
Han shu pu chu
Journal of the American Oriental Society
Journal of the Royal Asiatic Society
Kaogu yu wenwu
K'ao ku hsueh pao
Li chi
Lun heng
List of abbreviations xvii
Lu shih eh 'un-ch 'iu
Edouard Chavannes, Les memoires historiques de Se-ma
Mitteilungen das Seminars fur Orientalische Sprachen
Po hu t'ung
Joseph Needham et al., Science and Civilisation in China
Shan-hai ching
San kuo chih
Ssu min yueh ling
Ssu pu pei yao
Ssu pu ts'ung k'an
Shih san ching chu shu
Tzu-chih t 'ung-chien
T'oung Pao
T'ai-p'ing yu-lan
Ts'ung shu chi ch'eng
TOyoshi kenkyu
Wen wu
Yen-t'ieh lun
Zeitschrift der Deutschen M orgenlandischen Gesellschaft
Figure 1 Map of the Han empire 195 BC.
Figure 2 Map of the Han empire AD 140.
The history of the early empires
The place and development of the subject
Up to 1949, western sinologists had concentrated almost exclusively on two
periods or aspects of Chinese civilisation, one very early and one almost
contemporary. Beginning with the Jesuit scholars of the sixteenth and
seventeenth centuries, pioneers in the subject had set themselves two tasks; they
needed both to converse in the spoken language and to read China's classical
texts. Only fluent command of the dialect of the officials and influential families
of Peking would enable them to talk to their hosts on terms of equality; only a
familiarity with their hosts' written tradition would demonstrate that the
foreigners were men of culture. They therefore determined to learn how to read
classical writings so that, in the course of learned conversation, they would
demonstrate to their hosts the serious nature of their studies and their good
faith in claiming an interest in the products of Chinese civilisation.
The early missionaries thus embarked on the arduous study of the earliest
stages of China's philosophy and religion for which information and
instruction would be available, and in so doing they necessarily depended on
the guidance provided by their Chinese teachers. These latter had been trained
to a man so as to satisfy the demands of the imperial civil service; since
boyhood they had been imbued with a deep respect for the China which, they
had learnt, had preceded the establishment of the first of the empires in 221
BC. Following their teachers' examples, the early fathers directed their efforts
to elucidating the texts of those early centuries; they strove to understand the
precepts and ideas of China's first recorded thinkers, formulated up to two
thousand years before their own time.
Such a precedent laid its mark on the training in Chinese studies which the
much more numerous and varied band of western visitors of the nineteenth
century received. But by then major changes had taken place, in the growing
frequency and complexity of contacts between the countries of the West and
the Ch'ing empire, and in the emergence of new demands and new ambitions
that the visitors harboured. A new type of foreigner, with different aims and
needs, was settling in the missionary houses of the interior, the Legation
Quarter of Peking, or the offices and godowns of the Treaty Ports. The
2 Divination, mythology and monarchy
newcomers were engaged in persuading local officials to allow them to preach
the word of God; or they were acting out the niceties of international
diplomacy at the capital city; or they were seeking the most favourable terms
for the conduct of their business. As consuls or journalists they were also in
duty bound to keep the West informed of the state of the country and its
prospects. For such men and women a study of China and its culture was of a
more practical nature than that which the earlier missionaries had espoused.
Their interest lay in the way oflife that they saw practised around them, in the
institutions and legal prescriptions of the existing government or the oppor-
tunities for trade that they espied in China's ports and markets.
In such circumstances attention to the developments of the two thousand
and more years that intervened between the Chou period (c. 1045-221 BC)
and the days of the Dowager Empress (1835-1908) tended to be neglected. As
a result western observers were largely oblivious of the greater part of China's
imperial history, and the foreign policies adopted by their governments were
sadly inhibited by such ignorance. Diplomats accepted the conditions they
encountered in the latter decades of the nineteenth century as the norm; they
took the apparent weakness of government to be usual, and they saw no
reason to suppose that China would be capable of organising cohesive
policies, of mustering strength to see them implemented or of acting in full
confidence of the glories of the past. There can be little surprise that the
emergence of a strong united China in 1949 startled the corridors of power of
the western world; for they had not learnt of previous occasions when
comparable developments had taken place.
By then a fresh impetus had developed in Chinese studies. Although the
greater part of the new effort was being devoted to an assessment of the
contemporary scene of the mid-twentieth century, academic circles had at last
woken up to the realisation of a basic truth; that if China's history was to be
understood at all, the two thousand years of the empires demanded a detailed
study and a systematic appreciation; and that without such an appreciation
many of the motives and activities of the twentieth century could not be
explained satisfactorily. For some fifty years a number of distinguished
centres oflearning in the West have fostered a:r;i interest in such studies, and a
review of their achievements is now due. The following pages will be
concerned with the attention paid to the history of the Ch'in and Han
dynasties, between 221 BC and AD 220, both by western scholars and their
colleagues of East Asia. An attempt will be made to summarise their main
work and to point to some of the problems that are now calling for research.
The early efforts of the West
Fifty years ago textual and philological enquiries characterised Chinese studies
in the West. The magnificent work of scholars such as Couvreur and Legge
1 Couvreur (1913) and (1914); Legge (1861-72).
History of the early empires 3
translating the classical texts had been followed by the young Arthur Waley,
exceptional in that he turned his attention to translating poetry of the Han and
rang periods, and sawsomeofhis work in print by 1918.
But even before then
one of the earliest ventures of a western scholar to delve into the history of the
early empires had begun and borne fruit. This was the work of Edouard
Chavannes, whose prime interest had indeed been in the philosophy of the
pre-imperial period. But in 1888 he had been persuaded that he would be well
advised to turn his attention elsewhere, in view of the impact made on the
western world by Legge's work in that subject. Chavannes' monumental
translation of forty-seven chapters of the Shih-chi, fully annotated, followed
with surprising speed between 1895 and 1905.
The output of a pioneer who was
one of the exceptions of his time, these five volumes were completed in
accordance with the highest professional standards; their treatment of the
subject may be described as the West's first exercise in analytical criticism of a
period of Chinese history. The introduction brought into question the
authority or validity of the original work; the translation and notes presented
the Shih-chi in a manner that would engage the attention and interest of the
learned world of the day. Most members of that world had been trained in
Greek and Roman philosophy, literature and history; their horizons were
being widened by the archaeological discoveries of the Ancient Near East,
Egypt and the Mediterranean lands; they were pondering the riches oflndia's
cultural heritage. Thanks to Chavannes they were now able to catch a glimpse
of China's Standard Histories.
To reach the high standard of his Memoires historiques, Chavannes had
been able to call on the fruits of nearly two thousand years of China's own
scholarship. This was of particular value in those early days of the West's
study of Ch'in and Han history, as it was the Shih-chi and the other two
histories of the period (the Han shu and the Hou Han shu) that had inspired
comments and annotation, beginning with Ying Shao (c. 140 to before 204)
and extending in Chavannes' own time to Wang Hsien-ch'ien (1842-1918).
The latter's awe-inspirjng work
(published in 1900) was hardly available to
Chavannes at the time when he was translating the Shih-chi; but Homer H.
Dubs, whose annotated translation of certain chapters of the Han shu
appeared between 1938 and 1955, did enjoy that advantage.
decrying the achievements of those two scholars, it may none the less be
remarked that they could not have been expected to utilise such work to the
full. For the Chinese commentators had been trained during the centuries of
the imperial age and they had penned their notes for the benefit of readers
who shared the same scholarly background. As yet westerners could hardly
be expected to be sufficiently familiar with the whole Chinese tradition to
evaluate that background, or to understand the inhibitions imposed on
officially sponsored scholarship.
Waley (1918). 3 Chavannes (1895-1905).
See under Pan Ku and Fan Yeh, for Wang Hsien-ch'ien's annotated editions of the Han shu
(1900) and Hou Han shu (1924). ' Dubs (1938-55).
4 Divination, mythology and monarchy
Certain aids were beginning to appear, which would both clarify the
attainments of the scholars of the Ch'ing period, and bring other material to
bear on an understanding of the early empires. Japanese scholars had been far
from dormant, producing a number of annotated editions of the early Chinese
texts that historians needed to read; teachers at Japanese schools and
universities were beginning to be able to call on a number of textbooks on the
subject; and several Chinese and Japanese journals were soon to carry essays
in historical criticism.
Meanwile Japanese adventures on the continent had
stimulated work on a few archaeological sites of the Han period, and several
illustrated monographs had appeared, for example, on Lo-lang (1930; in
Korea) and Ying-ch'eng-tzu (1934; in Liao-ning).
Some of China's own
scholars, freed from the duty of interpreting history so as to serve the needs of
the imperial system, were publishing highly critical articles on textual,
historical or philosophical questions in the Ku shih pien (1926-41). The
Harvard-Yenching Institute's indexes of the three Standard Histories ap-
peared in 1940, 1947 and 1949.
The discovery of manuscripts
In the meantime the first accounts were to hand of the fragmentary
manuscripts discovered by Sir Aurel Stein during his first two expeditions of
1900-1 and 1907--8. To the great credit of the author, the Trustees of the
British Museum and the Clarendon Press, Chavannes' volume of photo-
graphs, transcriptions and translations of the thousand wooden strips from
the military lines at Tun-huang was published in 1913; Wang Kuo-wei and Lo
Chen-yii's work on the same documents appeared in 1914; but the learned
world had to wait until 1953 for the posthumous publication of Maspero's
work on the manuscripts that Stein had brought to light in his third expedition
(_ of 1913-15. Maspero had himself perished in Buchenwald in 1945.
Two subsequent major developments, each bringing new evidence to bear
on existing problems, stimulated new research in Ch'in and Han history, both
in China and Japan and in the West. The first concerned the discovery of
manuscript texts in far greater volume, with more varied contents and in far
better condition than the fragments from Tun-huang; the second concerned
the evidence of newly exacavated archaeological sites, particularly from 1950
onwards, and the wealth of artifacts that they contained.
Exploring to the east of Tun-huang from 1927 to 1934, Sven Hedin had
come across further remains of the Han lines of defences, at sites known as
Chii-yen or Etsingol. Embedded in the walls, or abandoned in the rubbish
Kambun editions of the Shih chi may be found in Ni dai kanseki koku jikai (1919-20), and
Kambun taikei (1911 ); for general histories of China, see Ichimura (1939-50) and Wada ( 1950);
critical essays appeared in the Ku shih pien from 1926, and in journals such as the T6y6 gakuh6
from 1911 and the T6h6 gakuh6 (both Tokyo and Kyoto series) from 1931.
Harada and Tazawa (1930); Ying-ch'eng-tzu (1934); Rakur6 (1934); Rakur6 (1935).
Chavannes (1913); Wang and Lo (1914); Maspero (1953).
History of the early empires 5
pits, there survived extensive parts of the documents whereby the Han forces
had been controlled and administered. The subject matter of this new material
was of the same type as that found at Tun-huang; its form was identical; and
the dates mentioned in the inscriptions covered approximately the same
period, running from c. 100 BC to c. AD 100. As at Tun-huang the finds
consisted of dismembered parts of documents that had been inscribed on foot
long (Han feet: i.e., 23 cm) strips, mainly of tamarisk, but also of bamboo or
other woods. Originally such strips had been fastened together by cords thus
maintaining the documents in their integrity. When the cords became
unloosed or broken, or when they had rotted, the component parts were
dispersed, and many of them were snapped into pieces. Two features
distinguished these finds from those that Sir Aurel Stein had come across at
Tun-huang. In the first instance they were 'far more extensive, numbering
some 10,000 rather than 1,000 pieces; secondly the new finds included two
examples of documents which were still intact, with the cords that bound the
strips together still fulfilling their function.
Working in the most adverse conditions of war-stricken Ch'ungking, in
1943 Lao Kan published a set of mimeographed transcriptions of these
fragments; a printed version followed in 1949, but it was only in 1957 that
photographs of the originals became available for study, in publications from
Taipei and shortly from Peking (1959).
In the meantime the documents
themselves had suffered a strange experience, of the type that seems only too
frequently to dog the footsteps of unique manuscripts. Early in the 1940s they
had found their way to the United States of America, and for some thirty years
they languished in the Library of Congress, whose custodians lacked the
necessary authority to allow access to scholars. By about 1970 the strips and
fragments had been returned to the care of Academia Sinica, Taipei, packed in
the very same cases and wrappings in which they had left China some thirty,
years previously; and at long last they were available for inspection and study')
by approved scholars, on request. /
In the meantime considerable work had been accomplished mainly by
Japanese scholars including Fujieda Akira, Mori Shikazo, Nagata Hidemasa
and Oba Osamu, to name but a few.
Working on the basis of the published
photographs, they succeeded in correcting some of the readings that had been
suggested and in solving a number of problems of interpretation, particularly
of technical terms. As a result it became possible to start to build a convincing
picture of the organisation of the Chinese forces at these remote parts of the
Han empire between c. 100 BC and c. AD 100. It was also possible to suggest
how some of the fragments could be assembled together as parts of one and
the same original document, and to establish some of the routine procedures
9 Lao Kan (1949), (1957), (1959) and ( 1960); Chii-yen Han chien chia pien ( 1959) and Chii-yen Han
chien chia i pien (1980).
' Fujieda (1955); Mori (1975), Oba (1982) and Nagata (1989) include reprints or summaries of
earlier studies.
6 Divination, mythology and monarchy
whereby official documents were drafted, prepared for despatch and distrib-
These fragments were shortly to be supplemented by even richer discoveries
of texts written not only on wood or bamboo but in some cases on silk. It is
difficult to overstate the significance of these finds for the history of the period.
In the first place, they derived not only from sites in the north-west, such as
Wu-wei (Kan-su),
but also from graves excavated in the interior of the Han
empire, such as Ma-wang-tui (Hu-nan), Chiang-ling (Hu-pei), Shui-hu-ti
(Hu-pei) and Yin-ch'i.ieh shan (Shan-tung); secondly, the subject matter of the
newly found documents was often of an entirely different type from that of the
strips from the north-west; and thirdly they were for the most part complete, if
being sometimes in a rather poor state of preservation. In addition, further
work at the site of Chii-yen (from 1972) had revealed even richer finds than
those made by Sven Hedin, including a few more examples of multi-strip
documents that were still intact.
Preliminary accounts of these discoveries soon appeared in the Chinese
periodicals; and while a number of splendid monographs, with photographs
tracings and transcriptions, have been published,
it has still not been
possible to make all the texts available in this way. They include literary and
philosophical works and historical annals; copies of the statutes and ordi-
nances of the kingdom ofCh'in and the Han empire, and legal case-histories;
almanacs, and documents that served the needs of divination. There are also
manuals of medical practice and military strategy; tables drawn up by
astronomers and a few copies of the calendar, the document that was
indispensable for all aspects of administration and whose preparation and
circulation was a closely guarded prerogative of imperial government.
Some of these manuscripts are copies of literary works for which a received
text, with voluminous commentaries, has long weighed down a librarian's
shelves. By vindicating their accuracy to an astonishing degree, the manu-
scripts lend considerable strength to the authority of much of China's early
literature. In some cases, such as the two copies of the Lao-tzu, where the
manuscripts differ from the traditional versions in some important respects,
considerable light has been shed on problems of textual transmission.
Sometimes the authenticity of a piece of writing that had been suspect has
been proved. Of especial value are the copies of texts hitherto unknown to
scholarship, such as the philosophical essays that precede or follow the
Lao--tzu, and that are thought to derive from the Huang-Lao school.
of the almanacs which were written on wood included information set out in
diagrammatic or tabular form. In addition to a few choice paintings on
Loewe (1967), vol. I, chapter 2.
Wu-wei Han chien. 13 Loewe (1986b).
Ch'ang-sha Ma-wang-tui i hao Han mu vol. 1, pp. 130--55, vol. 2, plates 270--92; Ma-wang-tui
Han my po shu; Yin-eh 'iieh shan Han mu chu chien; Yiin-meng Shui-hu-ti Ch'in mu; Loewe (1977)
and (1981). 15 Henricks (1989).
Jan Yiin-hua (1977); for further references, see Loewe (1977), pp. 120.
History of the early empires 7
religious themes,
the finds included the earliest examples of Chinese maps,
on wooden boards (dated c. 239 BC), proto-paper (180--150 BC) and silk (c.
Up to 1939 archaeological work in China had been concentrated on
pre-historical and pre-imperial periods. Together with Sinanthropus Pekinen-
sis, for long reckoned to be the earliest of man's progenitors, there had been
revealed a series of sites of the neolithic ages, distinguished as yet into the two
principal groups or stages of Yang-shao and Lung-shan; and the series of
eleven tombs at An-yang, with their hoard of magnificent bronzes, was being
correctly identified as the cemetery of the Shang-Yin kings. As yet archae-
ological work had been largely organised without official participation, being
sponsored in many cases by persons or organisations that lay outside China,
and being led by European or Japanese specialists. In a few notable instances,
Chinese palaeontologists and archaeologists such as Tung Tso-pin, Li Chi or
P'ei Wen-chung had taken a major part in the work of the 1920s and 1930s.
But apart from the illustrated catalogues printed by traditional Chinese
collectors and antiquarians, little attention had so far been paid to sites and
artifacts of the imperial ages. Bernhard Laufer's work on potteries (1909) and
jades (1912), Wilma Fairbank's study of the Wu Liang shrines (from 1941)
and Chavannes' investigation of sculpture and inscriptions (1893) formed the
principal exceptions, together with the Japanese monographs on the sites
which they had excavated.
Two major changes then intervened, the one concerning sponsorship of the
work, the other regarding its extent. Resentful of the manner in which some of
_ the rich treasures of jades, bronzes and ceramic wares had already been
removed from China to adorn the galleries and museums of both the West and
Japan, the new Chinese authorities of 1949 imposed controls to prevent such
exports. Responsibility for excavation devolved on a series of committees and
other organisations, established either at the capital city or in the provinces. In
the early days a few Russian experts were called in to assist; but it was the
Chinese authorities who allocated resources and organised the work.
At the same time the scope of archaeological investigation widened beyond
expectation. Regional bodies began to undertake work in areas that had so far
not been subject to investigation. It was realised that sites which dated from
the long centuries after the kings of Chou were well worthy of study, and that
the contents that they might yield could be of just as great a value in tracing the
17 Hsi Han po hua; Ch'ang-sha Ch'u mu po hua; for paintings of the Chan-kuo period see WW
1 KK 1975.1.53; WW1975.2.35fand 43f; 1976.1, 18fand 24f; 1976.6.20f; 1989.2, 1-11, 12--22and
31, and plates III, IV.
19 Chavannes (1893); Laufer (1909) and (1912); Fairbank, Wilma (1972); for the Japanese
monographs, see note 7 above.
8 Divination, mythology and monarchy
achievements of the people of China as those of the neolithic, Shang and Chou
ages. The new impetus was in part due to reasons that were in no way
academic. For as the work of national reconstruction gathered force, so were
the builders and engineers, the miners and the farmers lighting more and more
frequently on the material evidence of China's past. Faced with the slogan of
'Let the past serve the present', they were in duty bound to report such
discoveries to the local committees, who would in turn alert the provincial or
central authorities of the higher levels.
As projects for laying down railway lines or establishing irrigation facilities
moved on apace throughout the People's Republic, the number of finds that
were reported and the sites that were investigated was little less than
staggering. The bulldozer and the spade set to work with no considerations of
stratigraphy; a high proportion of the evidence that they unearthed dated
from the imperial ages, amounting, as it may be estimated, to well over 10,000
graves for Han times alone.
Such were the results of what was basically rescue archaeology, and the
ensuing embarras de richesse presented its own problems; there were not
nearly enough specialists to examine the new discoveries fully; conservation of
fragile materials, sometimes requiring control of temperature and humidity,
could not always be assured; preparation of catalogues of the finds could
involve specialist and skilled labour that simply did not exist (for example, the
two royal tombs of Man-ch'eng, Ho-pei, included over a total of 4,200
and publication of reports, with the necessary illustrations was
costly. All such work was in any case limited by the financial shortages or
other problems attendant on the growth of the People's Republic; at best it
was subject to interruption or abandonment; at worst to the deliberate
destruction of material evidence during the so-called cultural revolution.
Regular reports of these discoveries began to appear in journals and
monographs from 1950. As the years passed the inferior standard of the
illustrations gave way to line-drawings and half-tones of greater clarity and
quality, and eventually to colour plates. Reports were soon showing the
results of applying modern techniques, such as radio carbon 14 tests and
thermoluminescence, to the newly found artifacts. Quite soon news was
forthcoming of graves and their furnishings that could be dated in the Ch'in or
Han periods, and at times the occupant of a grave could be identified by name
or date. Other work concerned newly found stone monuments and epitaph
inscriptions that augmented those studied by antiquaries of the Sung period
e.nd later; a few remains of city walls or buildings were identified.
\ Many of the graves which were now being revealed had been constructed
~ f n g l y or perhaps for a man and his wife; but in addition a number of sites
Man-ch'eng Han mufa-chiiehpao-kao, vol. I, p.450.
Hotaling; Bielenstein (1976); for further reports on Ch'ang-an, see KK 1987.10, 937; 1989.1, 33;
1989.3, 261; 1989.4, 348; KGYWWl981.1, 123; for Lo-yang, see KK 1990.3, 268; for traces ofa
city in Fukien, see KK 1990.12, 915 and 1990.12, 1107.
History of the early empires 9
,,,,,:,,,,,,were found which included large numbers of graves, in a site that was
doubtless chosen owing to the belief that it would convey numinous blessings
-0n the deceased persons. A different type of cemetery, seen only rarely,
consisted of graves laid out neatly in grid fashion, for convicts or criminals,
;with scant attention to the niceties usually provided for the obsequies of their
The style and type of graves varied considerably, both in place and in time.
Some of the dead were buried in large timber chambers buried deep
J,i.nderground in pits; others were placed in clefts in the rock, or in chambers
fiewn out therein. Brick built chambers were probably the norm for officials or
other leading individuals of the Later Han period, sometimes extending into
several compartments and bearing decorations impressed before the brick had
dried. From early days, Chinese archaeologists had realised the potential
value of examining the assembled groups of such brick built graves. For from
such evidence it became possible to draw up schemata which showed the
sequences of different designs of the tombs on a secure basis. Thus the 225 Han
graves at Shao-kou (Lo-yang) could be set out in six major periods, ranging
from the middle of Former Han to the later part of Later Han; and the
schemata thus established for the style of these tombs continue to serve as
yardsticks for dating tombs found at other sites throughout the country.
\. Simultaneously the assemblies of large numbers of tombs at one and the
r same site made it possible to draw up schemata for the artifacts buried with the
j deceased persons, thus demonstrating the development of artistic and
, religious motifs and changes in technological skills. Many of the tombs
included valuables, of jade or bronze; vessels used for the sacred purposes of
prayer or purification; symbols of status that displayed the rank or function of
j the deceased person; musical instruments to be played for entertainment;
I equipment that might be needed to maintain a livelihood or ward off enemies;
jars that held consumable supplies of food and drink; cases of raiment; and a
.supply of coins. New criteria thus became available for dating objects of these
Previously it had often been impossible to authenticate objects said to be of
the Ch'in and Han periods, proudly exhibited though they were in the
museums or in the collectors' catalogues. One of the more important
differences to note is the new confidence with which newly reported discove-
ries can be accepted as being genuinely derived from an identifiable site. Of the
large number of sites of the Ch'in and Han periods, some have been of little
less than spectacular significance owing to the new types and the quantity of
the evidence that they have yielded. They have included Ma-wang-tui
(Hu-nan; c. 168 BC), known not only for the library offifty-two items but also
for the successful preservation of the body of the Countess of Tai for 2000
years; Man-ch'eng (Ho-nan; c. 113 BC), whose cliffs contained the tombs of
KK 1974.2, 2, plates IV, V. 23 Lo-yang Shao-kou Han mu.
10 Divination, mythology and monarchy
the King and Queen of Chung-shan, with the first known examples of jade
suits used for the burial of the highest in the land; Shih-chai shan (Yiin-nan; c.
108 BC) whose highly decorated drum-heads have revealed something of the
religious, musical and military activities of the non-assimilated peoples of that
region; Holingol (Inner Mongolia c. 160--70), whose murals painted a vivid
picture of official and military life at a somewhat remote distance from the
capital city; and I-nan (Shan-tung; perhaps c. AD 250) whose subterranean
tomb had been laid out in palatial style, with a rich profusion of carving that
embellished the pillars.
Perhaps the best known and most widely publicised
of all such sites is that of the tomb of the First Ch'in Emperor (died 210 BC).
Here a series of trial pits that were opened up at the perimeter of the
surrounding park disclosed the presence of the army of terra-cotta figures,
several thousand strong. Neither that tomb nor those of any of the Han
emperors, many of which have been identified, have been excavated fully.
Preliminary reports of these discoveries appeared in journals such as
Wenwu, Kaogu and Kaogu xuebao which were published in Peking. Since 1979
English abstracts or translations of some of the articles have been published in
Chinese Studies in Archaeology; and three regional journals have provided
room for further discussion.
In due course fully documented and illustrated
reports followed for the more important sites and finds. For readers who do
not wish for detail, several separate volumes are devoted to giving short
summaries of China's new archaeological work and discoveries, written in
encyclopaedic style.
The support of other disciplines and the value of scholarly exchanges
Along with the discovery of manuscripts and archaeological finds, progress
achieved in other aspects of Chinese history and in other disciplines has had a
marked effect in stimulating advance in the study of the Ch'in and Han
periods. For the pre-historic period, work by scholars such as Cheng
Te-k'un, Chang Kwang-chih and Yiian K'o in anthropology and mythology
has shed a light on the background to which many aspects of Han religious
practice must be related.
There has followed a deeper understanding of Han
poetry, and mystical or religious literature, such as parts of the Ch'u tz'u, and
24 Ch'ang-sha Ma-wang-tui i hao Han mu; Man-ch'eng Han mu fa-chueh pao-kao; Yun-nan
Chin-ning Shih-chai-shan ku-mu-ch 'un fa-chueh pao-kao; Ho-lin-ko-erh Han mu pi-hua; Tseng
Chao-yii (1956).
Of the many reports on the First Ch'in emperor's tomb, see Lederose and Schlombs (1990); for
Former Han imperial tombs, see Liu Ch'ing-chu and Li Yii-fang (1987). For a report of
excavations carried out at the site of the tomb of Han Ching-ti, see WW 1992.4, lf.
26 Jiang Han kaogu; Kaogu yu wenwu; Zhongyuan wenwu.
Hsin Chung-kuo ti k'ao-ku fa-hsien ho yen-chiu; Wen-wu k'ao-ku kung-tso san-shih nien;
Chung-kuo ta pai-k'o-ch 'uan-shu: k'ao-ku-hsiieh. For an analytical account of the artifacts, see
Hayashi (1976).
2 Cheng Te-k'un (1933); Chang K. C. (1983); Yiian K'o (1960) and (1985).
History of the early empires 11
, ;bf the iconography chosen to decorate Han tombs. In so far as Ch'in and
'.:]Ian political institutions and procedures drew on precedent, examination of
lbe states of the Ch'un-ch'iu and Chan-kuo (for example, by Hsu Cho-yun)
,periods has been enlightening. In the same way the analysis that has been
,undertaken for some of the later, and better documented, periods of Chinese
history (for example, T'ang, by Pulleyblank, and Twitchett)
has been of
similar value in framing a major chronological context within which the Han
achievement should be placed. The great advances made in the study of
Taoist and Buddhist religious practice (for example, by Demieville, Zurcher
and others)
have raised questions about spiritual aspirations, beliefs in an
.after-life and the trust placed in divination in Ch'in and Han times.
Work in other fields of enquiry has likewise affected the study of the early
empires. Bibliographical analysis (by van der Loon);
work on the relation-
;&hip between scholarship and political decisions (McMullen);
the meticulous
attention paid to the post-Han material from Tun-huang (Maspero, posthum-
ously, and Fujieda);
analysis of social structure (Ebrey);
and legal practice
(Johnson, and Bodde and Morris)
have all played their part in clarifying
much of Han history; for it is to the precedents of the Han age that many of
these developments may be traced, and in subsequent ages that the full effect
of those precedents may be assessed. Furthermore the study of science and
technology, by Needham and his collaborators, has opened up a new
dimension in Ch'in and Han history, as it has for other periods, stimulating
fresh approaches to age-old and well-savoured evidence, and forcing a new
consideration within a new context.
Simultaneously work on philology and
linguistics (Karlgren and Pulleyblank)
has ensured that attention must be
paid to the fundamental questions that lie behind all scholarship.
As in other aspects of Chinese studies, so with Han literature and history the
publication of Morohashi Tetsuji's monumental Dai Kanwa jiten (preface
dated 1955) had an immediate impact on research; the basic search for
evidence could be undertaken much more speedily and its results could be
more comprehensive. Other research aids that have been of especial value
have included the newly made and re-issued indexes and concordances of
texts, in particular the index of the Hou Han shu, published by the Jimbun
kagaku kenkyujo in 1960-2.
The reprint of basic sources of Chinese history also had a marked effect on
the subject. Publication of punctuated texts of the histories by the Chung-hua
shu chii, of the Tzu-chih t'ung-chien by the Ku chi eh 'u-pan she, eased the task of
29 Hsu Cho-yun (1965); Yang K'uan (1955).
Pulleyblank (1955); Twitchett (1963).
" Schipper (1982); van der Loon (1984); Barrett (1986); Demieville (1986); Lagerwey (1987).
van der Loon (1952). 33 McMullen (1988).
Maspero (1953); Fujieda (1955). " Ebrey (1974), (1978), (1983).
Bodde and Morris (1967); Johnson (1979).
Needham (1954- ).
Karlgren (1951), (1957); Pulleyblank (1984), (1991).
12 Divination, mythology and monarchy
those embarking on these studies considerably and provided editions which
scholars have been ready to cite and of which copies may be obtained easily.
Similarly, reprints from Taiwan and elsewhere of the standard edition of the
Thirteen Classics, with Juan Yiian's notes, are now readily available. Some of
the monographs written by modern Chinese scholars, for example on Han
poetry or institutions, or their critical editions of texts such as the Yen-t 'ieh lun
or the Ch'ien-fu lun have added significantly to the subject, as have the
publications of collected articles, mainly by Japanese scholars.
The growth of Ch'in and Han studies also owes much to corporate work.
The first meeting of the Junior Sinologues, which was held in Cambridge in
1948, gave promise that some measure of co-operation would be forthcoming
among the small and gallant band of young scholars, embarking on research
on various aspects of Chinese studies with the help oflibraries that were as yet
stocked somewhat meagrely. These early meetings provided a welcome and
perhaps essential venue for an exchange of ideas and preliminary reports on
research plans, and for sharing information about library holdings. Thanks to
the initiative and expert guidance of a few scholars (Balazs, Herbert Franke,
Wolfgang Franke, Haloun, Prusek, Seuberlich, Simon and van der Loon) and
the benevolent policies and grants of a few institutions, well-equipped
collections, including reprints of ts'ung-shu, microfilms of unique documents
or runs of periodicals, and Japanese works on subjects of sinology became
available in most of the major centres of Chinese studies in Europe. The
arrival of the Xerox machine left scholars with little excuse for failing to
consult material relevant to their work; and as teachers they could now
distribute copies of a text to a class, instead of obliging their students to make
their own transcripts by hand.
Further opportunities for the exchange of information, discussion of
research plans and debate on major topics were soon forthcoming in the
conferences that were being called in Europe, North America and Japan.
Some of these concentrated on a major theme or subject that ran through the
centuries of Chinese civilisation, such as that on historiography (School of
Oriental and African Studies London, 1956); or the three that focussed on
Chinese thought, sponsored by Fairbank, Wright and Twitchett and held in
America from 1951 onwards.
Specialists in the early empires both contrib-
uted to the sum total that was achieved by these projects and deepened their
own understanding of Ch'in and Han ideas and institutions, by seeing the
place that they came to occupy in later developments.
39 These punctuated editions of the Shih-chi, Han shu and Hou Han shu were published in 1959,
1962 and 1965; the Tzu-chih t'ung-chien was produced in 1956.
40 See under Ruan K'uan and Wang Fu, for critical editions of the Yen-I 'ieh fun and Ch 'ien-fu fun,
by Wang Li-ch'i and P'eng Tuo; for collected articles, see, for example, Kanaya (1960),
Kurihara (1960), Nishijima (1961) and (1966), Oba (1982), Hamaguchi (1966), Yoshinami
41 See Fairbank (1957), Wright (1953) and (1960), Nivison and Wright (1959), Wright and
Twitchett (1962).
History of the early empires 13
Other conferences or workshops of a different type which were equally
enriching focussed on particular aspects of early imperial history. Here
scholars would examine a subject in the light of disciplines or techniques that
were only recently being applied to Chinese studies, for example, palaeogra-
phy at one end of the spectrum matched by sociology at the other. Han studies
thus formed one of the subjects of the meeting held in Leiden in 1975, under
the title The state, ideology andjustice.
The workshop held in the University
of California, Berkeley in 1979, to consider the recent manuscript finds from
Ma-wang-tui drew contributions from specialists in palaeography, religion,
linguistics, philosophy, textual criticism and art history. A further instance,
which was a far cry from the initial ventures of the 1950s and was possible
thanks only to the developments of the intervening years, was the colloquium
held in the University of Chicago in 1991, under the title of 'Moment and
momentums in Han life'. This included papers on the place of the ju chia and
Ii" in Han society; the concept of empire; disciplines required by hygiene, as
seen in a Han manuscript; and the reflection of social structure in Han art. On
a number of occasions these meetings were enriched by the attendance of
colleagues from the People's Republic, Taiwan and Japan.
In the meantime it had become regular practice for western scholars to
spend long periods of sabbatical leave in East Asia, being engaged in field
work or consultation with their opposite numbers in the centres oflearning of
China, Taiwan, Hong Kong and Japan. Chinese language training on a new
scale of intensity, that had not been feasible previously, was ensuring that
newly emerging scholars, unlike some of their predecessors, would be capable
of direct communication on scholarly and other matters with their Chinese
colleagues. In addition, the new generation of western scholars possessed one
advantage that had not been so easily available to their predecessors, in their
ability to command written, and perhaps spoken, Japanese. Sinology had
received a bonus, in the form of a by-product of the Second World War, when
it had been necessary to train a large number of young persons to handle
modern Japanese language. As a result, westerners were now able to
contribute in a marked degree to conferences held in East Asia, while Chinese
and Japanese visitors to the West were leaving their stamp on the scholarly
projects of the West.
In these ways, and for these reasons, it has been possible to develop a deeper
and wider understanding of most aspects of the Ch'in and Han heritage, with a
far greater degree of cohesion than had been possible previously. In the
following pages it is possible to do no more than cite some of the principal
names, works and achievements, chosen either because of their own para-
mount importance or as being representative of new scholarly endeavours in
this field of learning.
The proceedings of this meeting were not published.
14 Divination, mythology and monarchy
The literary, intellectual and religious background
Textual studies and bibliography
Attention to historiography
has been concerned with the sources from which
the received texts of the Standard Histories were drawn and the manner in
which they had been made up. It has shown the weaknesses to which the
literary sources are prone, and the clear existence of omissions, inconsistencies
or errors; it has also brought to attention any reason that there may be to
doubt the authenticity of certain sections of these works.
At the same time a
few - all too few - of the newly found manuscripts which carry passages of
imperial decrees validate the accuracy of the Histories in this respect and
thereby lend authority to their other parts.
Archaeology supports the credibility of the Histories in a further way. The
rich finds of some of the major tombs bear out the textual descriptions,
carried in the Han shu and the Yen-t'ieh lun, of the style of burial accorded to
the highest in the land.
Material evidence likewise vindicates some of the
allegations that a highly luxurious style of living was being practised by
the rich; such statements were formerly suspect as being due to exaggeration,
but they may now be regarded as acceptable.
Thus the careful arrangements
to reconstitute the scene of a banquet, for example, in tomb no. 1 at Ma-
wang-tui, and similar representations on stone or brick in the tombs of East
China lend credence to some of the more fanciful statements of the sources.
Some of the sites have also served to elucidate the meaning of certain technical
expressions, such as the Huang ch'ang, or barricade, style of burial, now
available for inspection at the site of Ta-pao-t'ai.
Studies of textual and bibliographical history
have produced a clearer
comprehension of the value of the catalogue that forms chapter 30 of the Han
shu. Long conscious of the loss of77 per cent of the 677 items that are entered
in that list, scholars have been ready- perhaps too ready - to identify some of
the recently discovered manuscripts with the titles included there. In doing so
they have come to re-assess the work achieved by Liu Hsianga (79-78 BC) and
his son Liu Hsinb (?46 BC to AD 23), whose collation of existing documents
and classification ofliterature served so long as a means of distinguishing the
literary and philosophical categories of traditional China. It is now possible to
43 Chavannes (1895-1905) vol. I, (introduction: eh. l); Bielenstein (1954), pp. 9f; Hulsewe (1961).
44 Hervouet (1974), Hulsewe (1975).
45 Loewe (1967) vol. II, p. 230; for other decrees, see Wu-wei Han chien.
46 See, for example, Han shu pu chu 68, l laff, for the tomb ordered for Huo Kuang (68 BC); for
more general terms, see Yen-t'ieh lun 29 (Wang Li-ch'i ed. pp. 2067); Ch'ien-fu lun 12 (P'eng
Tuo ed. p. 134); for an example ofa 'barricade' style tomb, comparable with that prescribed for
Huo Kuang, that was built at Ta-pao-t'ai, see WW 1977.6, 23f, 30f.
47 Yen-t'ieh lun 29; Ch 'ien-fu lun 12.
48 For example, see Pirazzoli-t'Serstevens (1991) for the evidence from Ma-wang-tui; and WW
1972.10, 62 for a representation ofa banquetting scene from tomb no. 1, Ta-hu-t'ing, c. AD 200
(Ho-nan province). 49 For example, van der Loon (1952).
History of the early empires 15
'1;ummarise the main textual developments for some sixty or more works of
pre-Hano:. Han origin.
Writers such as Tjan Tjoe Som (1949-52), Kramers
(1950), Fujikawa (1968), and Anne Cheng (1985) have analysed the growth of
the different scholastic groups that arose from a concentration on the
Confucian Classics and the part that those works played in academic or
political controversy.
In the process of this work it has become apparent that the influence of
scholarly pressure, as supported by the government of the day, was far stronger
in Laterthanin Former Han. More may be said of the composition and content
of some of the texts that include the key to Han thought. Karlgren's works on
the authenticity of the Tso chuan and the language of the Lun-heng have been
followed by studies such as that of the 'Ten wings' of the I ching, the
and the T'ai-hsuan ching.
As a result, there is now a deeper
appreciation of the part played by those and other texts in the cumulative
growth of ideas in the pre-Han and Han periods; and there is less tendency to
give credence to the existence of separate discrete schools, unaffected by one
another's thought, and unaffecting each other's development.
of the manuscripts has resulted in a much clearer picture of the way in
which documents were prepared, drawn up and circulated, with the use of
wood as the principal medium of writing.
Such information concerns not
only the reports, orders and periodic returns that formed the stuff of central
and provincial administration; it also concerns the way in which both literary
texts and unofficial pieces of writing were drafted and copies were distributed.
It is now far easier to understand how short passages of text could be lost or
and how the sections of a work could have been re-arranged
dunng the process of transmission.
Literature and historical writings
A of works have led to a deeper appreciation of the development of
poetic forms and the place of the Ju in Chinese literature. Hightower's
monograph on the Han shih wai chuan (1952), Waley's translation of The
Nine Songs (1955) and Hawkes' Ch'u tz'u: Songs of the South (1959) were
among the early evaluations and critical studies of Han literature, to be
followed shortly by Dieny's translation of the Nineteen Old Poems (1963).
Later work included Knechtges' study of The Han Rhapsody (ju) of 1976 and
Birrell's Popular Songs and ballads of Han China (1988). Selections of Han
literature were also treated in translations of the Wen hsuan by von Zach
(1927 and later) and Knechtges (1982-7). Literary studies were further
advanced thanks to Hervouet's two volumes on Ssu-ma Hsiang-ju (1964 and
See the contributions by a number of scholars in Loewe (1993)
Tjan (1949-52); Kramers (1950); (1968); Cheng (198S).
(1926) and (1951); Yamashita (1961); Shchutskii (1980); Le Blanc (1985); Nylan and
S1vm (1987); Roth (1992). 53 Loewe (1967), vol. I, eh. 2.
16 Divination, mythology and monarchy
1972), of which the latter presented an annotated translation of a chapter of
the Shih chi.
In his supplementary volume to Memoires Historiques, Demieville (1969)
included translations of chapters 48 to 52 of the Shih-chi which Chavannes
had left in manuscript. Burton Watson's translation into English of forty-
eight chapters of that work (1961) and his inclusion of parts of the Han shu in
Courtier and Commoner (1974) have done much to introduce those histories to
non-specialist readers, as had Watson's earlier biographical study of Ssu-ma
Ch'ien (1958). De Crespigny's series of translations from the Tzu-chih
t'ung-chien (1969 and later) provide a chronological account of dynastic
history for the last decades of the Later Han period.
Other studies and
translations of parts of the Standard Histories will be mentioned below in
connection with specialist subjects.
New, punctuated, editions, of1955and1976, of the Hsi Han hui-yao of1221
and the Tung Han hui-yao of 1226 provide a handy, but by no means complete,
guide to the subjects treated in the Standard Histories.
Critical notes on the
Shih-chi and Han shu by Yang Shu-ta (1955) and Ch'en Chih (1959; second
edition 1979) have been acompanied by Wang Shu-rain's notes on the Shih-chi
(1965-82)56 and Mansvelt Beck's critical analysis of the treatises that have
long been included as part of the Hou Han shu (1990).
The intellectual framework and pattern of change
Thanks to the strength of the Chinese tradition, the deliberate message of
Chinese political writers of the imperial age and the rudimentary state of
knowledge in the West, there was for long a tendency to draw a clear-cut
distinction between three major approaches to life or modes of thought,
known respectively as 'Confucianism', 'Taoism' and 'Legalism'. Such a view
was in part due to the needs of Liu Hsiang and Liu Hsin; for in compiling the
lists of writings included in the imperial library at the end of the Former Han
period they had no option but to formulate categories. But such general
typologies cannot be regarded as satisfactory. For they fail to distinguish
between the different principles included under these general and amorphous
terms, and they imply that philosophical writings or political decisions rested
exclusively on the teachings of one of these three schools. The attempts that
have been made to place the Hsiin-tzu in a particular category illustrate the
point. Such a simplification has given way to the acceptance that Chinese, no
less than other thinkers, could draw eclectically on existing theories and
contribute by formulating a synthesis of ideas to suit the times and circum-
stances in which they lived. A clear example may be seen by identifying the
different elements in Tung Chung-shu's (c. 179 to c. 104 BC) own writings.
It has been accepted by most historians now that imperial Han government
54 De Crespigny (1969) and (1989).
Hsii Tien-Jin.
56 Yang-Shu-ta (1955); Ch'en Chih (1979); Wang Shu-min (1983).
History of the early empires 17
and its protagonists had little choice but to depend on a compromise of
principles; some of these drew on the ethical and humane approach to
humanity voiced by Confucius and his disciples; some rested on a recognition
of the practical need for an impersonal discipline ascribed to the leaders of
Ch'in. The description of the Han period, particularly of Wu ti's age (141to87
BC) as the 'Victory of Confucianism' requires modification.
In such circumstances it has become possible to suggest a framework within
which major changes came about in ideology, religion and imperial policies.
For Former Han such changes can be seen to have taken place in a highly
consistent manner and it is possible to chart the chronological stages in which
they developed. It may be seen that an initial period of some sixty years (c. 202
to c. 141 BC) was marked by consolidation, re-inforcement of the central
government's power, retrenchment on expenditure and a negative policy
vis-a-vis the potential enemies of the north. Such an initial stage gave way to a
period (c. 140 to c. 90) of positive and constructive development which saw
deliberate attempts to control the economy, an initiative in expanding the
Chinese sphere of influence in Central Asia and the inauguration of state cults
designed to augment the power of the dynasty. But the pace was too severe,
and such efforts became too demanding to be sustained for long. The final
century of Former Han saw a move to reject the forward-looking policies of
the previous decades, to eliminate extravagant expenditure and to relax the
controls that the government had imposed on parts of the population. A
defensive foreign policy led to withdrawal from some of the recently
penetrated territories, and the cults of state were directed to new ends.
In such circumstances a major change had been taking place in ideological
terms; from a view of the Han empire as the logical successor to Ch'in by right
of conquest, to a claim that the Han emperors were the true heirs of the kings
of Chou and their ideals. One of the achievements of Wang Mang, long vilified
in the Chinese tradition as the usurper who had ousted the Han house of Liu,
was to formulate the new ideal of empire that few dynasties who followed his
own short-lived regime of Hsin (AD 9 to 23) could afford to jettison.
pretension and practice were different matters. However eloquently the
spokesmen of the Later Han period mouthed the precepts of the Confucian
ethic or praised the glories of the house of Chou, such ideals failed by and large
to promote politicaland social stability or to restrain ambition or rapacity.
Public life was marked by the rise and fall of different families of imperial
consorts and the growing influence of the eunuchs in the palace. Officials
trying to implement the will of the central government saw the emergence of
semi-independent estates in some of the provinces, amid a marked attention to
the acquisition of wealth. Short victorious expeditions to the north-west, and
a partly successful policy of dividing China's potential enemies into several
See Loewe (1974a) for some of these themes.
For a vindication of Wang Mang, see Bielenstein (1954), pp. 82f, 154f; for his religious ideals,
see HSPC 99b.4b.
18 Divination, mythology and monarchy
camps, could not guarantee China's territorial integrity. The despair of some
in high places is seen by their refusal of appointments to office; they were
unable to reconcile the call of their own consciences with the compromises
attendant on public service. In the final decades of the dynasty, critics were
calling openly not for a restoration of Confucian ideals, but for the imposition
of a political and social discipline that would restore a sense of unity and
suppress corruption.
Religious belief and practice
A clear view now emerges of some of the elements that contributed to religious
belief and practice. Traces of popular observance intrude in the Standard
Histories, however reluctant the compilers may have been to acknowledge the
existence of such cults. Many of these rested on the fears or hopes of the
activities of local spirits of the mountains or the lakes, and there is reason to
believe that the practices were at times of a crude or ignoble type. While the
texts refer to human sacrifices offered to the Lord of the River (Ho po), the
bronze drum heads that archaeology revealed from a site in Yiin-nan in 1956
carried vivid representations of some of the gruesome scenes that took place
among the unassimilated peoples of the south-west.
In a more stylised way,
and perhaps by way of entertainment, elements of mythology may perhaps be
traced in the formal or ritual enactment of a contest between the two
semi-divine heroes Huang ti and Ch'ih Yu. As Bodde shows, a complex series
of festivals to which literature refers drew on popular belief, a reliance on the
the practice Qf 0xorcism of ritual c.lance:

A variety of methods of divination, or of consulting an oracle, were
available both at popular and official levels, should the need arise to seek an
answer to a problem from occult powers. Signs that either encouraged or
discouraged a course of action might be observed in the behaviour of natural
objects or forces, such as comets, clouds or winds. Alternatively manipulation
of the yarrow stalks and consideration of the hexagram that had been created
in the process could provide symbols for a seer to interpret; others who were
beset by a problem might prefer to set the on an instrument (k'an yu), or
to go through the mechanical motions of consulting an almanac. The
professionals (fang shih) who practised some of these occult arts doubtless
included charlatans among their fraternity; the presence of manuals on these
subjects (shu shu) on the shelves of the imperial library testifies to the hold that
they may have exercised on some minds. Some of the esoteric signs or
expressions of the Ch'in and Han almanacs persist on copies of the calendars
" See Balazs (1964), pp. 187f, for the views voiced by Wang Fu, Ts'ui Shih and Chung-ch'ang
Tung; see also Ch'en Ch'i-yiin (1980) and (1986).
For the sacrifices to Ho Po, see Yiian K'o (1985), pp. 217f; for the practices of the south-west,
see the plates in Yiin-nan Chin-ning Shih-chai-shan ku-mu-ch'iin fa-chiieh pao-kao and The
Chinese bronzes of Yunnan.
Bodde (1975).
History of the early empires 19
drawn up by officials of imperial government; they may be seen today in the
calendars that adorn the walls of a bank in Taiwan or Hong Kong, or in some
of the manuals for guidance printed in Japan.
Initt(lHY the)mperial c.ults ofCh'i11 and J:fan had concentrated on the four,
and then the five, ti who were conceived as tutelary deities. within the cyclical
system that controlled all activity. From perhaps 114 BC it was intended that
the emperor should take a more regular part in these services than previously,
and that he should also attend in person the services which were being
inaugurated to two other objects of worship. These were T'ai i, first seen as a
spirit of one of the constellations, and Hou t'u, Sovereign of the earth. With
the increased attention that was being paid to the image of the kings of Chou
and their practices, from c. 31 BC the imperial cults came to encompass
services to T'ien, the god worshipped by Chou, who remained a cult figure
until the end of the imperial age.
J:!11ang.!i_had duly taken his place among the four .and.then .. the five. ti
W_9fQippegJrpmjlie (;h.'ih periods onwards. somewhat surprisingly, in Vtew
;;f the attention that this figure was receiving at the time, his cult was not
specifically re-inforced as part of the religious innovations of c. 115 BC. At
much the same time .. source or intermediary who
.. c9nJer t_h.e of(immortf!clity l to which end a number of other
techniques were being addressed. In Former Han these mainly included rituals
that sought a path to the paradise of the East, by way of P'eng-lai; in Later .
.!fll:I1 :dexotions.wete paicl.JoJIJv Qu.een ;Mother of i11: f
reachmg her .own long then, had \
appeared inva,rious .in II1Ythology.-Iiitlie iiieanHffie"many whowisfiea '
To bury their loved ones with the blessing of eternal felicity had been taking
symbolic steps, designed to place the deceased person correctly within the
major cycles of the cosmos that ran from birth to death and then to re-birth. 63
Distinctions between these ideas and between the different characteristics of
religious practice have become clear thanks to renewed study of well-known
texts in the light of archaeological discoveries and the newly found manu-
scripts. Such evidence has led to inferences regarding the underlying assump-
tions that may have been in the minds of those who practised these and other
rites. Other advances in an understanding of religious matters are seen in
Ziircher's study (1959) of the entry of Buddhism into China, and research in
the growth of the .Taoist. S,ects and.tlieir connection
Jnents of the last fifty .years of Han.
This has led to considerations of the
intellectual relationship of the two emerging systems
and it has stimulated
examination of the T'ai-p'ing ching which may be numbered among the more
enigmatic and frustrating pieces of Chinese literature. 66
Loewe (1974a), eh. 5. 63 Loewe (1979).
Ziircher (1959); Seidel (1969); Lagerwey (1987).
Kaltenmark (1969); Ziircher (1980); Demieville (1986 ).
Kaltenmark (1979); Mansvelt Beck (1980).
20 Divination, mythology and monarchy
Intellectual developments
From earlier ages, Ch'in and Han had inherited concepts framed for the very
different social and political conditions of the Warring States period. These
included the ethical ideals formulated by Confucius and his immediate
followers, or by Mo Ti; the mystical appeal to eternal values, as seen in the
Lao-tzu and the Chuang-tzu; and the principles for strengthening political
\, control of a population, as enunciated by Shang Yang and others.
), ideas of Yin and Yang and tbe ..
! )n .. the
and fomiulae of the Chou i, largely incomprehensible by Han
times, were being explained in terms of the ideas of staged change and
development that were current in the fourth and third centuries BC; and such
speculations had resulted in many of the texts now included in the Book of
Changes. Recent analysis of these texts and their ideas (for example by
Graham, Lau, Shchutskii and Hellmut Wilhelm) has formed a basis on which
a study of the next stages of intellectual development took place; in the new
imperial age these earlier concepts were subject to adoption, adaptation or
rejection, to suit the newly emerging moral and political order of the day.
Many of the ideas seen in Han times owed as much to a re-formulation of
existing concepts as to innovation.
uDitary sys,Wm.mth its. compone11t a11d interlocking estates of heaven, earth
and a or"reactfuilTrloiieofUiose
'tii;e;esfates to a particular type of activity taking place in one of the others;
anda belief in the capacity
, one type of beiiigto 11;,notnfr.Studies of the Huai-nan-tzu (completed 139
\ SfiOW1Tie"afiempt made in that book to explain the universe as operating on
\these principles, in the light of mythology, history and observation of the
'world of nature.
The work of Tung Chung-shu (c. 179 to c. 104 BC), often described as the
founder of Han Confucianism, has also drawn attention, in view of the
importance of his syncretic system for a number of centuries. His ideas, as seen
in the three memorials included in the Han shu, 'combined the humanistic
approach of Confucius, the o( and a belief in the <''
goodwill of heaven towarasmanKind. In is appeal to the records of human
history to illustrate the application of these principles, he saw the exercise of
imperial sovereignty as an essential and inescapable element of the system 9f
the universe, and offered his explanation for the occurrence of natural
Ideas of imperial sovereignty were taken further forward when the old
concept of the T'ien ming was being mentioned or invoked as a means of
providing a form of superhuman authority to which the power of the
Jii!:"' Ames (1983); Le Blanc (1985).
History of the early empires
emperors was ultimately due. When first established, the Ch'in and Han
owed their and recognition to the force of arms; but the l
need to display that they enjoyed the support of both religious and intellectual f
sanction had made itself felt quite soon. Documentation for this is seen most I
clearly in Pan Piao's essay on kingship, in the series of documents whereby i
Wang Mang asserted his claim to rule and in the procedure whereby the last od
the Han emperors abdicated his throne in favour ofTs'ao P'i, first king of the
Wei Dynasty.
From these and other developments it has become possible to discern the
wide religious and intellectual differences that separated Former Han from
Later Han. The change is apparent in the criticisms voiced, possibly by Ching
Fang the Younger (executed 37 BC) and certainly by Yang Hsiung (53 BC to
AD 18), who believed that the old system of sixty-four hexagrams was an
inadequate means of explaining the perpetual state of change in the universe;
Yang Hsiung evolved his own more subtle and complex system. 69 Expressing
ideas which partook of both a naturalist and a rationalist point of view, Wang
Ch'ung (AD 27 to c. 100) questioned the validity of contemporary beliefs and
fears about a life after death. He expressed his doubts regarding the efficacy of
divination and the concern that heaven was presumed to entertain on behalf of
human activities and destinies. He affirmed his belief in the spontaneous
nature of creation, as against the idea that the myriad objects of the seen
world, organic and inorganic alike, had been manufactured purposefully by
Wang Ch'ung was thus in no way to be classified with writers and thinkers
of the orthodox school, and for long his work attracted little attention. Alfred
Forke had indeed published his translation of the Lun-heng as early as 1907,
and he deserves the very greatest credit for achieving this at so early a stage of
western sinology, and without the aids that adorn a scholar's library today.
More recent studies have made it possible to relate this text more closely than
hitherto to the intellectual and political circumstances in which Wang Ch'ung
was writing. Pokora (1975) has shown something of the antecedents and the
part played by Ruan T'an (c. 43 BC to AD 28).70 In special chapters the
Standard Histories were reporting the occurrence of portents, but Wang
Ch'ung for his part refused to believe that such events carried any message for
the future of mankind; studies of these reports has suggested that on occasion
they may have been manipulated, perhaps for political purposes.71
For Wang Mang's pronouncements, see, for example HSPC 99A, 32a and 34a and 99B, 4bff,
Dubs vol. III, pp. 243f, 248f and 272f; for Pan Piao's essay (Han shu pu chu IOOA.
Sa) de Bary (1960), vol. I, pp. 1 for the accession ofTs'ao P'i, see Leban (1978). For the
changmg concepts of sovereignty, see Loewe in Twitchett and Loewe (1986), eh. 13.
Nylan and Sivin (1987).
' For paid to Wang Ch'ung in China at times when Marxist views were prevalent,
see T 1en Ch ang-wu (1958); for an analysis ofWang Ch'ung's thought, see Ch'en Kung (1968);
for an attempt to place Wang Ch'ung's views in context, see Loewe (1982).
Bielenstein (1950) and (1984); Eberhard (1957).
22 Divination, mythology and monarchy
For the Later Han period, examination of the so-called apocryphal texts
and the T'ai-p'ing ching has drawn attention to some of the heterodox ways of
thought, mention of which is ignored or suppressed in official accounts of the
period. Such work has been balanced by research in the growth of the
Confucian schools and their traditions from the Warring States onwards and
their effect on Han scholarship.
Dynastic administrative and political considerations
Dynastic instability
Traditional appreciations of Han history had long recognised that the Later
Han period had at times been rent by schism and rivalries, fought out largely
between the families of imperial consorts. By contrast it was implied that
Former Han had been a period of dynastic stability, thanks partly to the
sustained and successful reign ofWu-ti (141-87 BC). But such an assumption
requires modification. It is apparent that, between the Empress Lii's exercise
of power (187-180 BC) and the establishment of Wang Mang's dynasty (AD
9), so far from holding a monolithic and unquestioned sway over a mighty
empire, the house of Liu had perforce experienced a series of critical incidents,
some of which almost spelt its demise. These had included the challenges that
had preceded the accession of Wen-ti in 180 BC; the revolt of the seven kings
(154 BC), put down by the central government's initiative; the rivalry of two
families of imperial consorts, culminating in the outbreak of fighting in
Ch'ang-an and the suicide of the Empress and the Heir Apparent (91 BC); the
deposal of an emperor after a reign of twenty-seven days in 74 BC; and the
bitter contentions of the consorts' families, and the difficulty of finding an
imperial heir, that marked the reigns of Ch'eng-ti (33-7 BC) and Ai-ti (7-1
A clear analysis of disputes of this type, which characterised much of Later
Han history, has been completed by Bielenstein, who summarised the story
with the terse observation that 'the political history of this period is in large
measure a history of its factions'.
De Crespigny (1969 and 1989) has
provided specialist studies of the later decades of the period. Possibly the year
184, which witnessed the outbreak of the revolt of the Yell ow Turbans, should
be regarded as signifying the realistic end to the political and social cohesion of
the Han empire, rather than 220, when Hsien-ti solemnly enacted the
formalities of abdication.
It may be added that this was a time when a marked decline is noticeable in
the position and importance of the emperors; a number of these had been
enthroned while they were still infants; in several cases an incumbent did not
12 Dull (1966).
73 See note 51 above, and Ch'ien Mu (1958), Hsii Fu-kuan (1976).
74 Twitchett and Loewe (1986), pp. 136f, 14lf, 177f, 184, 214, 218; Loewe (1974a), chs. 2 and 8.
75 Twitchett and Loewe (1986), p. 277.
History of the early empires 23
survive to enjoy his manhood in that august situation. A major change may in
fact be traced in the function of the emperor over the four centuries of Ch'in
and Han rule; from their assumption of power as leaders of victorious armies
to their relegation as chief religious functionaries of the nation; from thei;
personal delegation of authority to their trusted supporters, to the manipula-
tion that they suffered at the hands of men and women of ambition; and from
the exercise of political leadership, to the role of acting out constitutional
forms. By 220 the person of the emperor had come to signify very little in terms
of imperial power.
Attention has also focussed on the contributions made by a few well-known
individuals of the Ch'in and Han periods, both to the achievements of their
own times and to the heritage bequeathed to later ages. Re-assessment of the
. traditional view of the first Ch 'in Emperor and Li Ssu on the one hand, and of
Wang Mang on the other suggests that, so far from being the villains that they
have been depicted, it was largely to these men and their colleagues that due
credit should be given for laying the foundations of successful imperial
government. Later regimes owed far more to the insistence on discipline
ascribed to Li Ssu and the adoption of Confucian models by Wang Mang than
they would ever have cared to admit. These results are due to the work of
Bodde and Bielenstein.
It has also been possible to assess the part played by Sang Hung-yang
(executed 80 BC) and Huo Kuang (died 68 BC) in co-ordinating economic
effort and stabilising political movement during the latter years of Wu-ti's
reign and subsequently.
The part played by Hsiao Wang-chih (suicide 46
BC) and K'uang Heng (died c. 30 BC) in the reaction which set in against the
intensive policies of Wu-ti's reign has also received greater recognition.
Balazs: of three prominent thinkers and critics of Later Han (i.e., Wang
Fu, Ts m Shih and Chung-ch'ang T'ung) drew attention to the extent of the
social imbalance and political instability of the last decades of Later Han.
Other such studies of individuals include a short monograph on Ch'ao Ts'o
154 BC);
Ch'en Ch'i-yiin's two volumes on Hsiin Yiieh (148-209)
and his ideology; and de Crespigny's assessment of Hsiang K'ai and his use of
portents as an instrument for criticising the contemporary state of the empire
in a famous memorial of 166.
The administration of the empire
Wang Yii-ch'iian's article on the central governmnent of Former Han was
first published in English in 1949, to be followed by Bielenstein's comprehen-
sive study, which covered the organs of the provincial as well as the central
administration. In addition to this analysis of official posts and their attendant
Bodde (1938); Bielenstcin (1986).
Kroll (1978); Wu Hui (1981); Loewe in Schram (1985), eh. 8. 78 Loewe (1974a), eh. 4.
Balazs (1964).
Ch'ao Ts'o chi ch'i chu-tso. 1 De Crespigny (1976).
24 Divination, mythology and monarchy
duties, a series of articles by Oba Osamu and others
set out to clarify some of
the details and distinctions of the system of officials. Yen Keng-wang's major
study (1961) of local administration covered Ch'in, and Former and Later
Han; de Crespigny (1966) had been concerned over methods of recruitment
and local officials of Later Han. Ebrey's close study (1978, 1980 and 1983) of
epitaph inscriptions and the history of one of the clans of the day illustrates the
relationship and hierarchies of the different strata of the civil service and the
importance of growing patron-client links. The documents from Tun-huang
and Chii-yen provided information about some of the lower-ranking officials
that are not available in other sources.
Research has shown the importance of Han's initial reliance on Ch'in's
practice of government; this was before the needs of empire required an
elaboration of the duties for some posts and a modification of others. It is also
possible to trace how Han's major compromise with existing forms, i.e., the
establishment of hereditary kingdoms committed mainly to the emperor's
sons, worked out in practice. The process saw a change by stages from an
initial situation in 202 BC, when these kingdoms occupied a good half of Han
territory, to the point when they had been reduced to small enclaves nestling
among the commanderies that lay directly under the control of the central
government (AD 2).
Such moves signified the increasing strength of the
Emperor's officials at Ch'ang-an. A further feature is seen in the attempt to
prevent an undue concentration of power in the hands of a few individuals, as
exemplified in the division of financial authority between separate offices of
state. It has also been realised that the distinction drawn at one time between
an inner and an outer court requires modification in the light of later analysis
of political procedures.
From the administrative documents ofTun-huang, Chii-yen and elsewhere
it has been possible to reconstruct some of the procedures for inaugurating,
handling and forwarding official documents.
These have included the texts
of a few imperial decrees, known hitherto only from the edited summaries that
were included in the Standard Histories. The same material has revealed
professional aspects of the work of both the civil.and the military officials,
working as they did according to their hierarchies. As in the services of other
nations, so here officials and clerks were engaged in preparing accounts with
meticulous accuracy, recording the receipt, issue or consumption of official
stores and maintaining records of times when certain routine actions took
place, or when an abnormal event required report to a senior authority. What
is now known, or can be inferred, about the creation of these administrative
documents has a wider bearing, in so far as it may be applied, with some
reserve, to the composition ofliterary texts found at other sites, and the whole
history of their transmission.
The Statutes (liib) and Ordinances (ling") of the kingdom and later the
Kamada (1949) and (1962); Oba (1982).
83 See maps ~ 9 in Twitchett and Loewe (1986). 84
Loewe (1967).
History of the early empires 25
empire of Ch'in had been drawn up as a means of maintaining security,
suppressing crime and conscripting manpower to work the land or fill the
ranks of the army. Some measures were intended to impose a measure of
regularity in agricultural or other work. Despite the traditional claim that,
immediately after its establishment, the founders of Han had simplified the
system, reducing the extent of its rules and mitigating the severity of its
punishments, it seems likely that the new masters of the empire introduced
little practical change beyond adapting these institutions for their own use.
Ch'in's administration had rested on the two complementary principles of
rewards and punishments, and the application of these has been subject to
consideration. The rewards for services that were likewise adopted in Han
included a series of orders ofhonour
that signified a rise in social status and
carried some material privileges, such as exemption from statutory obliga-
tions of conscript services, or reduction of punishments for those who ran foul
of the laws. In some cases services were rewarded by a grant ofland, either on a
personal basis or, less frequently, with hereditary tenure and the right to raise
taxation within prescribed limits. In particular cases the Han emperors
distributed bounties or compensated their subjects in other ways, if they had
required excessive exertion, for example, to supply the needs of an imperial
progress. Such bounties may also have acted as a means of validating an
emperor's claim that he was fulfilling his mission of encouraging the human
race to prosper. Han had perhaps developed here a somewhat new idea of
rewards, as compared with the more direct aims advocated by Shang Yang (c.
385-338 BC). The wide range of punishments to which criminals were
sentenced in the Han empire included the death penalty, carried out in various
ways, mutilation of the body, flogging, enforced service at the frontier,
confiscation of property or payment of fines.
Such conclusions result from Hulsewe's meticulous study of legal treatises
and the scant amount of archive material that has come to light in recent years.
His interpretation of the treatise on law of the Han shu involved philological
investigation, a familiarity with records of legal matters and a knowledge of
the procedures of government; in these and other respects the work of Chinese
scholars, particularly those of the Ch'ing period proved to be of immeasurable
Reconstruction of parts of the text and scope of the Han Statutes and
Ordinances depended initially on assembling short citations of those docu-
ments from the Standard Histories. Texts dating from the second century AD
onwards, such as the remarks of commentators to the histories, or other works
such as the Shuo wen, acted as corroborative or supplementary evidence.
Further examples of these laws were found in the documents from Tun-huang
and Chii-yen, together with some precious records of cases that had come up
for administration.
But a decisive change occurred with the discovery of
" Kamada (1949) eh. 2; Loewe (1960). 86
Hulsewe (1955) (1959) and (1979b).
26 Divination, mythology and monarchy
complete copies of some of the legal documents, dating from the Ch'in period
onwards. These derived principally from sites inside China (Shui-hu-ti and
Chiang-ling, in Hu-pei), and full publication of the texts is still awaited. As a
result a far more detailed assessment of the contents of these early imperiallaw
codes became possible, as may be seen in Hulsewe's annotated translation
(1985) of these enigmatic texts. In addition to information on the minutiae
that were involved, the few case histories disclosed the way in which officials of
central and provincial government set about implementing their orders.
It has also been possible to distinguish different categories of crime to which
some of the laws were directed and to learn something of the way in which
disputes were settled or litigation handled. It has yet to be shown how far these
laws included concepts of rights and obligations, apart from those of statutory
service, or how far they may be regarded as an instrument for protection
against official oppression.
Economic and social history
Nancy Lee Swann's annotated translation of those chapters of the Shih-chi
and Han shu that directly concern economic theory and practice (1950)
opened a new vista for historians of early imperial China. The book showed
the Ch'in and Han attitudes towards a number of aspects of production and
the great issues that were engaging the minds of the statesmen of the day.
Gale's translation of parts of the Yen-t'ieh lun (1931) had already alerted
readers to some of the subjects of controversy. Once again the documents
from the north-west provided further information; fragmentary as they were,
they included some evidence of the ways in which the government's measures
were operating.
Major issues
concerned land tenure, some parties actively wishing for this
to be as open as possible in the interests of increasing production and revenue,
others seeking to restrict the extent of holdings in the interests of social
equality. The basic point at stake here lay in the willingness, or right, of
imperial government to control the activities of individuals, and the same issue
arose in connection with other matters. These included a choice between
allowing coin to be minted privately, without restriction, or insisting that it
should be produced by the government's mints only, under official control.
The same issue, of public as against private interests, arose over the
establishment of state monopolies to control and exploit the production of
salt, iron and liquor.
In addition, the proper use of coin was itself sometimes brought into
question. Taxation was raised partly in goods and partly in coin. Occasionally
it was urged that the major financial transactions of the empire, such as the
payment of officials, should be made in kind rather than in cash, and the
87 Gale (1931); for attempts to analyse the issues at stake and the arguments, see Loewe (1974a),
eh. 3, Kroll (1978) and Loewe in Schram (1985), eh. 8.
History of the early empires 27
records from Chii-yen illustrate how officers at the frontier would sometimes
receive part of their pay in silken bales, and part in cash. Ingots of gold, to a
nominal value of 10,000 cash coins each, are quoted in the texts in connection
with major transactions, but it has yet to be determined how far they actually
circulated, and why Wang Mang took steps to call in gold resources from
throughout the empire.
In an early article, Yang Lien-sheng (19 57) showed that, whereas the usual
cry was for a reduction of public expenditure, there were occasionally those
who urged the benefit of increasing the consumption of expendables, in order
to stimulate production and the exchange of commodities. Opposing views
were taken by Han statesmen regarding one other principle; whether the
central government was entitled to take positive steps to co-ordinate produc-
tive effort, stabilise the price of necessities and organise their transport. The
plea that by so doing a government could alleviate public suffering and
eliminate profiteering in a time of shortage was met by resistance to a scheme
that allowed officials to play the part of merchants.
A number of Japanese scholars, including Kato Shigeshi (1952-3), Ut-
sunomiya Kiyoyoshi (1955), Nishijima Sadao (1966) and Hiranaka Reiji
chose to study these problems. Of the modern Chinese scholars, Li Chien-
nung (1957) set out his studies of different topics, such as agriculture, coinage
or taxation; Ch' en Chih ( 1958) concentrated on a number of specialised issues,
and Ho Ch'ang-ch'iin (1964) examined problems of land tenure. Documen-
tary evidence has produced some tantalisingly inadequate information about
the collection of tax;
archaeology has revealed the sites of some of the iron
foundries, with the moulds in which goods were manufactured, and examples
of certain key products such as crossbow bolts or ratchet wheels made to a
notable degree of precision.
Hsu Cho-yun's valuable collection and transla-
tion of documents that refer to agriculture (1980), preceded by Amano
Motonosuke's general history of the subject (1962), illustrates the working
conditions under which the farmers of Ch'in and Han laboured and solved
their problems.
Meanwhile Bielenstein had been attending to some of the more fundamental
conditions that could affect economic development. These included the
vagaries of the Yellow River and the consequent effect on production, and the
extent and distribution of the population, on which Lao Kan had worked
Corrected figures for the population may be studied alongside the
scant information available for the area of the arable land and the scale of
rations distributed to the armed force and their families. Such indications
suggest that in Han times China could not have produced sufficient grain to
Loewe (1967) vol. I, pp. 7lf, vol. II, pp. 295f; WW 1974.6, 44f.
For traces of iron foundries see WW 1979.4, 77; KK 1989.2, 149; KGYWW 1983.4, 27; for
moulds, goods and precision tools, see WW 1976.9, plate 8; KK 1983.3, 243, 320, 322; 1988.6,
576; KGYWW 1982.5, 68.
Bielenstein (1947) and (1954); Lao Kan (1935a) and (1935b); Twitchett and Loewe (1986),
maps 10 and 1 L
Divination, mythology and monarchy
feed the whole population satisfactorily, together with sufficient hemp for the
textiles needed for general clothing. The figures for the population (for AD 2
and 140) bring out the truth that as yet imperial China and the force of its
administration was largely concentrated north of the Yangtse River, where 90
per cent of the registered population were domiciled. By T'ang times this
proportionate imbalance had been reduced to 75 per cent; and by Sung the
move was well set whereby the south was achieving economic predominance.
The same conclusion may be drawn by comparing the size and distribution of
administrative units; the smaller, and more confined, commanderies of the
north, mainly along the Yellow River valley, were clearly subject to a more
rigorous degree of official control than those of the deep south or the farwest.
Marks of social distinction formed the subject of enquiry by a number of
scholars such as Kamada Shigeo, whose early essay of 1938 concerned the
orders of honour (chueh) of the Ch'in and Han systems.
At much the same
time Yang Shu-ta (1933) was assembling the evidence for ritual occasions in
the life cycle and the customs attendant on marriage or death. Clarence
Martin Wilbur's monograph (1943) on slavery in Former Han included an
early analysis of social structure that preceded his translations of all references
to his subject. For Later Han, Yang Lien-sheng (1956) was showing how
certain great families were achieving power and some measure of indepen-
dence with the support of their own bodies of retainers. To Ebrey's study of
one family and its extension into periods after Han, there should be added the
attention paid to the eunuchs of the day. Bielenstein has corrected the criticism
to which they have long been subject, by showing the key role that they played
in maintaining the survival of the house of Liu;
in a special monograph
Ulrike Jugel (1976) analysed political and social aspects of this group. Ch'ii
T'ung-tsu's comprehensive collection (1972) of source material that concerns
social distinctions is preceded by an introductory study of kinship, marriage,
the position of women, social classes and powerful families.
Relations with other peoples
Relations with the non-Chinese peoples of the north were of greater concern
to Ch'in and Han officials than those with the southerners; for it was from the
north that potential danger might arise. Early studies, by Hirth (1885 and
1917) and de Groot (1921-6) have now been supplemented by a new approach
and degree of comprehension. It has been realised that more allowance must
be given than formerly for the existence of a controlled and powerful
confederacy of the Hsiung-nu, no longer to be identified for certain with the
Huns. 94 For the Chinese side it is now appreciated that, so far from mounting
91 See Twitchett and Loewe (1986), maps 3, 5, 7, 8 and 9.
92 Reprinted Kamada (1949), eh. 2.
03 Bielenstein in Twitchett and Loewe (198fi), pp. 287f.
94 For Chinese relations with the leaders and confederacies of the north, see Yii Ying-shih (1967)
and Hulsewe (1979a).
History of the early empires 29
a continuous show of strength that backed a forward policy of expansion, the
Han empire could engage in no more than short campaigns into the recesses of
Central Asia or elsewhere, led, for example, by Huo Ch'ii-ping (d. 116 BC),
Wei Ch'ing (d. 106 BC), Pan Ch'ao (d. AD 102) or Ma Yiian (d. AD 49). The
success of the imperial effort depended on the ebb and flow of both Chinese
and non-Chinese strength, to a degree that had not been fully assessed.
The line of united defences that Han had inherited from Ch'in was not
strong enough to prevent incursion as far as the close vicinity of Ch'ang-an in
166 BC. The extension of the defence lines into Central Asia, which was
accompanied by the establishment of four commanderies in the north-west
(between 112 and 104 BC) allowed for the safer conduct of merchandise
principally in the silk laden caravans that were wending their weary way to h ~
West. It was also possible to exercise a more effective control over travellers
and potential deserters wishing to move in or out of Chinese territory; patrols
and observation of enemy activity could be maintained more regularly. These
and other activities are testified in the newly found documents, which include a
signals' code and signals' log, reports of patrols, and records of travellers and
goods admitted through points of control.
Such material makes possible a
new appreciation of the professional standards and procedures of the Han
forces, between c. 100 BC and c. AD 100.
The story of Chinese relations with the small units that were settled around
the rim of the Taklamakan Desert has also come up for review. Hulsewe's
annotated translation of the relevant chapters of the Han shu runs from the
pioneer expeditions of Chang Ch'ien (c. 123 BC) to Chinese attempts to
co-ordinate activities in these remote regions. The account concerns the
establishment of the first Protector General of the north-west (59 BC) and the
demise of that office (AD 23). Chinese records show various aspects of
Chinese diplomacy at work, resting sometimes on a display of force,
sometimes on a matrimonial alliance, sometimes on an exchange of hostages.
On. occasion, particularly in Later Han, the Chinese were able to adopt a
pohcy of divide et impera. One particular incident (36 BC) threw up the
dangers that a local Chinese commander would incur if he chose to take
military initiative without receiving full authority from the central govern-
ment. In a few other, somewhat ugly, incidents a sense of honour was not too
clearly apparent on the Chinese side.
Meanwhile a new situation was at times facing the Han government. The
normal units of administration, i.e., commandery and prefecture, that
satisfied conditions in the home areas could not operate effectively in lands
where a different climate prevailed and where inhabitants of different ethnic
origins were not engaged in the type of economic work that characterised the
heart of the empire. To control the newly penetrated areas it was necessary to
evolve new types of administration, such as the Dependent Kingdoms
Loewe (1967) vol. II, Documents MD 13, MD 17, UD 5, UD 7 and TD 8.
For the i n ~ i e n t ?f 36 BC, see Loewe (1967) vol. II, document UD 9; Loewe (1974a), eh. 7; for
Han relations with some of the leaders of the north-west, see Hulsewe (1979a), pp. 39f.
30 Divination, mythology and monarchy
(Shu-kuo). In such units some Chinese officials were posted to maintain a
Chinese presence; by way of compromise, native leaders were free to exercise
some measure of authority in their own terrain. The Chinese hope of retaining
the support of such leaders is also seen in other ways, such as the conferment
of imperial titles to those who had surrendered or made over to Chinese
authority, and even by the accommodation of groups of immigrants within
Chinese territory.
A number of scholars
have been engaged in examining the motives that lay
behind a Chinese wish to expand its influence or maintain its prestige in these
distant regions. Arguments have ranged over the priority given to defence or
to trade, and the extent to which trade was conducted by private venturers or
was subject to the sponsorship of the government. The principles behind two
types of policy, categorised in general terms as one of appeasement and one as
a tributary system, have been subject to discussion, and it has been shown that
the latter, rather than increasing the resources of the empire, involved China in
expenditure. The discovery of Chinese goods, for example, silks or bronze
wares, in sites that lay within the Hsiung-nu sphere of influence support this
Foreign relations of Later Han and their resultant military operations,
strategy and tactics have been studied in depth by de Crespigny (1984). Dubs'
speculation (1957) that Roman soldiers in flight from the Mediterranean
world found themselves at large in those parts of Central Asia that are
mentioned in the Han shu has not received support. The military and
diplomatic ventures which engaged Chinese interest in the north-east have
been studied by Gardiner and de Crespigny, whose case history of T'an-
shih-huai (1977) is of especial value. While the establishment of Chinese
commanderies and outposts in Korea during Former Han, as attested by both
literary accounts and archaeology, had as yet little effect on China's own
history, it formed the first move towards forming the bridge that was to
convey Chinese culture to Japan. The annotated translations of the Chinese
literary sources for the growth of relations between China and Ja pan form an
early example of co-operative work shared between an American and a
Japanese scholar.
Despite the appeal evinced by the Ch'u tz'u, it was long part of the Chinese
tradition to despise the lands below the Yangtse River as being uncivilised and
their inhabitants as being uncouth. Archaeological discoveries, for.both the
pre-imperial and the early imperial periods, have served to redress this
imbalance by proving the existence of the flourishing culture of Ch'u, with its
own characteristics that are seen in religion, folklore and art. The material
remains of this culture, and of the presence of colonists and officials from
97 Yii Ying-shih (1967); Hulsewe (1974); Rashke (1978); Daffina (1982).
9s Rudenko (1969); for reports on various sites and finds, see WW 1979.4, 49; 1980.7, I, 11, 13;
1983.8, 67; KKl980.4, 333; 1984.4, 367; 1987.1, 33; KGYWWl988.3, 17; 1988.3, 111; 1990.2, 5.
99 Tsunoda and Goodrich (1951).
History of the early empires 31
other parts of the Han empire, fill the pages of the special journal that treats
that region (Jianghan kaogu); Barnard (1972 and 1974) and others have
studied the anthropological and ethnic implications. Further west, sites such
as Shih-chai shan, in Yun-nan, have yielded evidence of a way oflife that was
very different from that of their neighbours of the Lower Yangtse valley.
This was the land of the cattle breeder, as well as the tiller of the soil; ox-heads
and snakes took their place as totems in the grim religious rites; and the
practice of agriculture lay some stages behind the advanced methods of
China's heartland, resting as they did on centuries of experience. Art
historians have identified the characteristics unique to the area and related
others to the culture of Dong-son. Han governments were ready to recognise
the existing powers of the local leaders, engaged as they might have been in
fighting one another. The histories record that Han Wu ti presented a gold seal
to the king of Tien in 109 BC; such an object, duly inscribed, appears in the
photographs of the artifacts found there.
Science and technology
Many aspects of scientific and technological development have been examined
by Needham and his collaborators, and by other scholars from China, Japan
and elsewhere. Research has proceeded by raising theoretical and practical
questions of a type that have not been put to the evidence previously. It has
been asked how far the principles of physics, biology, chemistry and medical
science were grasped in the Ch'in and Han periods, and what techniques
China's farmers, craftsmen, engineers and artisans adopted to speed and ease
their daily work. Much of this research work has concentrated on a
re-examination of the literary evidence, and comparison with the recently
found documents and the great wealth of artifacts from Ch 'in and Han tombs.
Maspero's early work on astronomical instruments; Dubs' attempts to
corroborate astronomical records; and reconstructions of the Han calendar
have been confirmed, corrected or supplemented by manuscripts which carry
charts of planetary movements or illustrate the varieties of comets.
study (1969) of early computation and calendrical systems and the discovery
of a number of calendars, or their fragments, that date from Han times have
served to correct the tables drawn up by Pere Hoang (1910), Ch'en Yiian
(1925) and Tung Tso-pin (1960). Among other work, Needham discriminated
between three theories, all current in Han times, that sought to explain the
relation of the earth to the heavens;
his work with Wang Ling clarified the
achievements of early Chinese mathematicians, including the refined calcula-
For reports on Shih-chai-shan, see note 60 above; see also Wang Ning-sheng (1980).
Yiin-nan Chin-ning Shih-chai-shan ku-mu chiinfa-chiieh pao-kao, p.113 and plate 107.3.
Maspero (1939) and (1950); for records of eclipses see Dubs (1938-55), vol. I, pp. 165-6, 188--9
etc.; for planetary movements, see Loewe (1977), 122-3.
Needham (1954-), vol. III, pp. 210f.
32 Divination, mythology and monarchy
tion of n as 3.14159 by the third century. Cullen's annotated edition of a Han
textbook of mathematics will provide an insight into pre-Han and Han
attitudes to the Queen of the sciences.
The history of medical science has been advanced considerably, and will
doubtless reach higher levels once the texts of the newly found manuscripts
have been published in full. Preliminary reports, that are often and regrettably
marred by the use of simplified characters, indicate the wealth of this material
and allow some conclusions to be drawn. Needham and Lu's monograph
(1980) on acupuncture may now be supplemented by reports of markedly
refined needles that were used for the purpose; Harper's preliminary con-
clusions from the medical manuscripts of Ma-wang-tui, Han-t'an-p'o
(Kansu) and elsewhere suggest connections with the theory and practice of
hygienic disciplines. The set of illustrations of exercises and postures recom-
mended for this purpose that was among the silk manuscripts ofMa-wang-tui
is itself now matched by texts from Chang-chia shan (Hu-pei) that prescribe
these procedures.
From a different angle, Sivin's analysis of the various
component parts of the Huang-ti nei ehing serves to distinguish the origins and
dates of some medical concepts;
other work on medical science has been
published by Unschuld (1985 and 1986) and Porkert (1977).
Donald Wagner's work on the history of metallurgy will draw on the
evidence of artifacts and traces of foundries. Miniature models of farmhouses,
wells or other equipment found in tombs combine with examples of agricul-
tural tools, a re-examination of textual evidence and information in legal
documents to provide more exact details of the occupations in which the great
majority of the Han people were engaged. The skills of those who nurtured
Bombyx mori, spun his threads and wove silken cloth with its intricate
polychrome patterns have been richly exemplified at Ma-wang-tui. Their
working methods and the equipment such as the looms which they used
appear on some of the reliefs that decorated Han tombs, and a few parts or
fragments, for example of spinning whorls, have been identified. One of the
drum heads from Yiin-nan illustrates graphically the Rroduction of textiles by
a people as yet unassimilated to a Han way of life or influenced by Han
culture. Kuhn's volume puts these achievemi;:nts and discoveries in their
historical context.
Archaeology and art history
Identified as some of them are, the tombs of the Ch'in, Fornier Han and Later
Han emperors still await excavation. A few stones (known as huang ch'ang),
104 For the manuscripts fromHan-t'an-p'o see WW1973.12, 18; for those from Chang-chia shan,
see WW 1990.10, 82f, 87f; for the bodily exercises, see Tao yin t'u. See also a paper 'Hygeia in
Han medicine: acupuncture's forgotten ancestress', presented by Donald J. Harper at a
workshop held in the Department of East Asian Languages and Civilizations, the University
of Chicago, in November 1991, under the title Moments and momentums in Han life.
10 In Loewe (1993). 10 Needham (1954-), vol. V, part 9.
History of the early empires
inscribed with the dates on which they were hewn and inspected and the num-
bers which indicated the positions that they were intended to take in the sur-
rounding walls have long remained the sole surviving parts of the structures
built for the Later Han emperors. A number of reports, which concern the
sites of burial for those of Farmer Han and their consorts, disclose details of
some of the artifacts found therein; explanations have been offered for the
reasons that governed the choice of locality.
The style of burial and selection of funerary furnishings, supported
occasionally by an inscription or a seal, has made it possible to identify the
tombs of a munber of the subordinate kings of the empire, as at the sites of
Man-ch'eng (Ho-pei: the King and Queen of Chung-shan, died 113 and before
104 BC), Ta-pao-t'ai (Peking: either a king of Yen, who died in 80 BC ora king
of Kuang-yang, who died in 45 BC) and Hsiang-kang shan (Kuang-tung: one
of the kings of Nan Yiieh, who died 128-117 BC). 107 Such tombs were
constructed for those who enjoyed some of the highest positions in the
dynastic and social hierarchy, and appropriate distinctions marked the
manner of their burial. An outstanding example ofa special type of structure,
which may be termed 'barricade', may be seen at the re-constructed site of
Ta-pao-t'ai; over 15,000 timbers, each a metre long, were used to build the
imposing multi-chamber residence, capped with its tumulus.
A few monographs, for example on the sites ofMa-wang-tui, Man-ch'eng,
Shih-chai shan, Ho-lin-ko-erh (Holingol) and I-nan provide fully illustrated
accounts of these discoveries.
Some of the material conditions of life stand
revealed Hglin_ggl._ Qfilfil!I.uuul.1hcir
scenes and the work of o( ___,
these actlv1y_..fooned tbe subjects chosen for.the baked or
flie provides views of agricul-
mining and hunting; Shan-tung shows the interior and exterior of
gentlemen's residences.
Remains of grains, vegetable fibres and bones from
Ma-wang-tui, combined with the reconstruction of how certain tombs were
set out and furnished in the form of a banquet enable conclusions to be drawn
about the dietary habits of the Han period.
Many of the artifacts discovered in these sites may be found in Pirazzoli-
t'Serstevens' beautifully produced account of these subjects (1982). Other
finds include a few maps, either on silk or wood, from Ma-wang-tui and Fang-
ma-t'an (Kansu); one of the tombs is dated at 239-238 BC, and the maps
found there are the earliest examples known in China. 110 Studies and
For a list of tombs of nineteen kings of Former Han, see WW 1992.2, 38; individual reports
may be seen, for example, in KK 1984.3, 222f (for the king ofYiieh-nan), and in Man-ch'eng
Han mu fa-chiieh pao-kao (for the king and queen of Chung-shan); for the royal tomb at
Ta-pao-t'ai, now open for inspection, see WW 1977.6, 23f, and Hsin Chung-kuo ti k'ao-ku
fa-hsien ho yen-chiu p. 445 and colour plate XII. 108 See note 24 above.
Finsterbusch (1966-71); Shan-tung Han hua hsiang shih hsiian-chi; and Ssu-ch'uan Han hua
hsiang chuan hsiian-chi.
For the maps from Fang-ma-t'an, see WW 1989.2, 12-22, plates III-IV and figures 8, 9 and 11;
34 Divination, mythology and monarchy
reconstructio.ns of the two capital cities of the Han empire have been based on
the meagre material finds, information in later literature which is sometimes of
a suspect nature and calculations of the dimensions of the walls.
inscriptions, mainly of Later Han, have long drawn the interest of China's
traditional scholars. More recently several series of facsimile publications
have made these texts more readily available, usually with transcriptions and
notes. A particularly valuable example is seen in Ma Heng's posthumous
volume of the texts of the classical works engraved on stone in AD 175.
Chavannes' early work on stone reliefs from Shan-tung (1893) drew
attention to the value and varieties of evidence from the artistic work of the
period. This was followed by Wilma Fair bank's reconstruction (l 941) of the
Wu Liang shrines of that area (AD 151) with their wealth of iconographic
imagery. Later publications which illustrate the distinctions that may be
drawn between the regional styles of Shan-tung, Ssu-ch'uan, Nan-yang and
the old land of Ch'u include works by Rudolph and Wen (1951), and
Finsterbusch. From China there have come a number of albums, including
photographs, rubbings or line-drawings that illustrate the themes and features
chosen by artists of, for example, Shan-tung and Ssu-ch'uan.
It may also be
possible to isolate the output of a particular region, in the pre-Han kingdom of
Chung-shan, (south-west of Peking) with its exquisite metallurgy that was
produced both before and during the Han dynasty. Art historians such as
Powers (1992) have concentrated on interpreting themes of Later Han art and
explaining the political and social implications; in a number of cases it is
possible to discriminate between work ordered for the palace, and executed by
the Palace workshops (Shangfang), and that which was designed to satisfy the
ambitions of other groups such as merchants or perhaps eunuchs. Jn a later
study of one particular and perhaps exceptional example, that of the Wu
Liang shrines, Wu Hung (1989) has interpreted the rows of mythical and
historical figures and the numerous representations of portents as a piece of
. social and political criticism, in the form of a plea for a return to the ideals of
the Confucian ethic.
The newly found wealth of material has shown how wares of bronze,
pottery or lacquered wood served the needs of .both the sacred and the
profane. Jn some cases the designs used in objects of one medium could be
borrowed or imposed on those of another; such habits produced, for example,
the simulation on pottery wares of the rings and handles that had had a
practical function on vessels made in bronze; and shapes that had been
. conceived for vessels in one medium, for example bronze, were adopted for
those made in another, such as lacquered wood. Artists used the surfaces of
the jars and wine holders for various themes or styles, whether geometrical or
for those from Ma-wang-tui, see KK 1975.1, 53; WW 1975.2, 35; WW 1975.2, 43; and WW
1976.1, 18. 111 Bielenstein (1976) and Hotaling (1978).
Ma Heng (1957); the Japanese series of facsimiles include Shodo meihin taikei (ed. Iijima
Inataro; Tokyo: Shogei bunkain) and Shoseki meihin sokan (Tokyo: Watanabe Takao).
See note 109 above.
History of the early empires
animal, treating their subjects in separate horizontal registers. The frequent
recurrence of the same themes in the stone reliefs and ornamental bricks of
Han tombs may perhaps give an impression of a type of shop work that was
devoid of initiative or inspiration; but subtle differences in the manner in
which sculptors or clay moulders fashioned their medium dispel so harsh a
judgement; so also does the persistence of the exuberant motifs of artists
working in the old tradition of Ch'u, as may be seen in the characteristic black /
and scarlet designs of lacquer wares of the Y angtse valley. /
Han art drew on mythology and folklore and served to express some of the
religious beliefs that were grounded on such traditions. These influences are
seen in the choice of the situation and the orientation of tombs; in the inclusion
of material talismans to assist the dead and guide them into the next world; and
in the profusion of symbols that appear on the structure of the tombs or their
funerary furnishings. Some of this evidence has been shown to relate to
assumptions and tales current during the pre-imperial period, as may be seen in
the Shan-hai ching; some of it developed in a markedly new manner during the
Han period.
Such considerations, and the interest that the new materials have stimu-
lated, have given rise to a number of specialist studies. Cheng Te-k'un
explored the influence of Yin---Y ang and wu hsing as seen in certain bronze
mirrors (1957); Bulling's monograph (1960) traces the sequence of different
forms of mirror decoration throughout the period. Dubs' article (1959) on the
mythological features seen in bronze 'hill' censers prepared the way for the
interpretation of the splendid examples of these objects that were found at
Man-ch'eng and the representation of the same themes on wine containers. In
several studies, Schuyler Cammann ( 1987) sought to relate some of the Han
artists' designs to the ideas of the magic squares of the day, studied in turn by
Major (1984). The detailed work on artistic expressions of mythological
themes by Hayashi (1974) and Kominami (1974) complements Yuan K'o's
assembly of the basic source material for mythology and folklore .
Reference has been made above to the continual developments that may be, .
traced during the four centuries of Ch'in and Han rule. The changes in,,
ideology, religious belief and intellectual outlook mark wide divergences
between the character of Former and Later Han, and such cultural changes are
seen alike in the art of the times. Some of those changes showed the influence of
new ideas, and before long, Buddhism was to leave its mark on many
expressions of Han culture. But other changes may reveal how a set of beliefs
may in time lose its force. The work of some of the earlier artists and craftsmen
of the Han period show their careful attention to symbols of the sacred, or their
need to express religious hope. But their successors did not necessarily share the
same motivation or the same understanding of these ideas. Later attempts to
exhibit the same symbols, for example on bronze mirrors, are often incomplete
or irregular, thereby disclosing that the artist was ignorant of their purpose. 114
For example, certain bronze mirrors described in Loewe (1979), pp.186-9 (type X).
36 Divination, mythology and monarchy
Work on the history of Ch'in and Han China has developed that of a
number of other periods and disciplines, and it has been by
willingness of scholars to take due account of their colleagues' achievements m
other fields of study. The results have been cumulative, resting on the work of
the philologist, the historian and the archaeologist, and calling the
contributions of specialists in religion, philosophy, anthropology, art hist?ry
and social studies. The process of re-assessing the received texts, evaluatmg
new manuscripts and appreciating recent discoveries has been on
the corporate results of work that has started from a number of
approaches. At the same time, today's of the West_ owe an
lable debt to their predecessors; to the multitude of. .
commentators and editors, with their unparalleled famihanty with Chinese
literature and history; to the pioneer giants who first applied the methods of
western analysis to a study ofCh'in and Han times; and to their
colleagues of China, Taiwan, Hong Kong and Japan, whose of
publications reveals their constant application to the love of learrung.
Two results of a general implication follow from the .of the _de-
tailed work that has been accomplished in recent decades. Frrst, It is possible
to appreciate with greater clarity the influence by peoples of
Ch'in and Han on their successors. In so far as the hentage which they
moulded many of the ideals, institutions and practices of later ages, attent10n
to the early empires can in no sense be isolated or relegated to of
the ivory tower. An understanding of China's of !esmt or
marxist doctrine can hardly be complete without cons1derat1on of the
made by Buddhism from the time of its entry in Later Han. The ventures. of the
Sui and T'ang emperors in the north require comparison and with the
experiences of Han forces in similar undertakings. K'ang Yu-we1's_ mtellectual
arguments of the late nineteenth century can hardly be without some
knowledge of the academic controversies said to have the Han era.
Secondly, it is perhaps not too much to suggestithat md!v1dual develop-
ments or incidents of Ch'in and Han times must now be seen m a new context,
to which a new dimension has been added. The protests of a radical critic, such
as Tung Chung-shu or Wang Ch'ung, the emergence of new of
literature and the discussions of religious or institutional issues begm to take
their place in the intellectual and political circumstances of the time. The later
poems of the Ch'u tz'u are now seen against a background and
popular belief that formed part of Han life, unvoiced and concealed as it
been in official writings. The establishment of the post of Protector General m
Central Asia must be judged against China's changing foreign policies and the
success or failure of colonial expansion. The final act of abdication by the last
of the Han emperors may now be seen within the context of those concepts of
imperial sovereignty that had been four centuries in the making.
History of the early empires
Much awaits specialists in this field oflearning. A full examination
of the manuscnpts, with their religious, mantic, legal or medical content
cannot undertaken the texts have been published in facsimile.
Although 1t cannot be cla_imed that the archaeologists' work is spread evenly
over the area, sufficient results are probably now available to justify and
make more refined oflocal cults and patterns, and to
assess the survival of ways of life of peoples that were not assimilated to H
culture. Textual and linguistic analysis is urgently needed to detennine
exte?t of the authenticity. A further problem is that of
tracmg ?ngms _of tables now included in the Shih-chi and the Han shu
and examirung therr differences. career and contributions of Liu Hsianga
(79-78 BC) could well the subject of a monograph, and a new study of
textual and academic controversies of his time and the succeeding decades
overdue. For Later Han, an attempt to relate leading personalities to
change and political decision might well be rewarding. Economic
benefit from an examination of the problems of waterways, and
studies from acounts of some of the major families of Later Han.
time may also be npe for co!11pilation of a biographical dictionary for
Ch m Han, to rest on a library s shelves alongside the works of Franke,
Goo?nch_and for the Sung, Ming and Ch'ing periods.
It is agamst this background that the articles which are collected below took
shape. They derive from a conviction that it is not possible to treat individual
aspects history in_isolation, and that developments such as dynastic
sequence, changes, mtellectual growth or religious influences must be
seen_ as affectmg one _another. Nor is it possible to fasten exclusively on one
particular type of evidence, such as historical record, without reference to
other sources such as those of literature, artistic symbolism or the material
finds Similarly, research in the subject demands more than
work_within the confines of historical method, and must take due account of
theones evolved in other disciplines such as anthropology. The articles
attempt to the_ influences or that were gathering or losing
strength at a given time; to examme the controversies that gave rise to
argument; or to place the practices of the time in their religious and intellectual
context. At the same time attention is paid to evaluating the motives and part
played by some leading figures of the day, and to observe the perpetual
mterplay between faith and reason, and between official duty and popular
Man and beast: The hybrid in early
Chinese art and literature
Our knowledge of Chinese religion and mythology rests on the evidence of art,
archaeology and literature, which may be considered very generally in two
types. There is the evidence of a natural, romantic and free tradition,
sometimes associated with the south, and that of a formal, classical and
inhibited tradition, sometimes associated with the north. Of these two major
traditions, that of the north came in time to predominate over that of the
south. For it was in the north that China's political and dynastic authorities
emerged, and from thence that they extended their sphere of influence to the
east, and then to the centre and the south. While this extension may be seen
most clearly in political terms, it also affected cultural developments. The
regimes of the north required intellectual conformity and support; there set in
a tendency whereby the temporal masters and officials of the north were wont
to mould and exploit the independent arts and mythology of the south so as to
satisfy their own immediate political needs; and in the course of such
treatment some elements of the southern tradition became subject to scorn
and even suppression.
For students of Chinese mythology, this tendency has had the unfortunate
result of overlaying some of the evidence of the natural urges of man with the
purposeful creations of his intellect. From about the beginning of the Chris-
tian era, standardisation was affecting Chinese literature, both in the choice of
the material that was sponsored for preservation and in the interpretations
that were put on early writings in order to propagate orthodox beliefs. In
studying early mythology, then, we must fasten on such evidence as preceded
the move towards uniformity, and on that which survives from the live
cultures of the south. We may consider the paintings made on neolithic
potteries of perhaps the fourth millennium BC, and the far more elaborate
patterns that adorn the bronzes of c. 1500 BC and later. The inscriptions made
' Of the many scholars who have written on this subject and associated topics, I am glad to single
out three to whose writings I owe a special debt: Noel Barnard, of the Australian National
University, Canberra; Chang Kwang-chih, of Cambridge Massachusetts; and Hayashi Minao,
of the Jimbun Kagaku kenkyiijo, Kyoto. The account of hybrid forms which is presented here is
intended only as a summary, from which all but the most important references have been
Man and beast 39
on bones and shells tell us something of the aspirations of early Chinese
monarchs of the ShangYin period (c. 1700-c. 1045 BC), and of the processes
for consulting divine powers; but they carry little information that bears on
the specific nature of those beliefs. Although some of China's literature may
date back to c. 1 OOO BC, the versions which we possess today must be carefully
examined; for we must sift the grain from the chaff, rejecting the results of the
subsequent editing that suited the needs of the imperial dynasties founded
from 221 BC onwards.
Luckily, evidence of a less orthodox frame of mind survives elsewhere.
despite the efforts of the officials of the north to deprecate its importance. It
derives from the once-thriving cultures of the south, and is seen in the art
motifs and literature that emanated from the valley of the Yangtse River and
beyond. These areas encompassed a terrain that was very different from that
of the north, giving rise to the characteristic rice cultivation, and including
large regions of swamp, forest and mountain. Such lairs lay beyond the reach
of the Chinese official, who tended to regard them as the home of the
untutored barbarian. It is from the artistic creations of such peoples, who were
free of the northern mandarins' discipline, that we may learn something of
China's early mythologies.
The Chinese believed in the existence and powers of a number of deities. Tia,
or Shang ti, or God on high, was venerated by the kings and possibly the
peoples ofShang; he was conceived as a unity, probably in anthropomorphic
terms; and he was thought to possess supreme powers over man and nature.
The kings of Chou, who supplanted those of Shang from perhaps 1045 BC,
believed in a different supreme deity, known as Tien, or Heaven. T'ien may
also have been conceived in human terms; and along with the institutions and
moral examples ascribed to the kings of Chou, T'ien was adopted as an object
of veneration by the imperial dynasties, who worshipped him right up to 1910.
Both tia and T'ien, it seems, existed on a higher plane than the shen", or holy
spirits. These were conceived in multiplicity, often being attached to specific
sites on earth. The holy spirits would respond to prayer, invocation or, if the
occasion demanded, to exorcism. For they were capable of actions which
could help or harm man; and they were conceived in animal, hybrid,
semi-animal or semi-human form. Finally, account must be taken of the kueia,
the demons who originated as manifestations of deceased human beings. They
too were capable ofbenefitting or injuring man; they responded to prophylac-
tic observances by man, and their presence could be invoked by specialist
intermediaries. That the holy spirits and demons existed on an inferior plane
, to that of tia or T'ien is shown by references in literature in which they follow
, tia in order of precedence or carry out behests at his command.
A number of changes may be discerned in the concepts of these deities. In
the earlier stages, of the Shang-Yin period, it was thought that the souls of the
LSCC 9.3b; HNT 4.4b, 4.12b.
40 Divination, mythology and monarchy
deceased ancestors of the kings shared the abode oftia, and that they acted as
intermediaries between tia and the world of mankind. This office was also
partly filled by mythological animals who served to link the two worlds. In
later stages, i.e. during the early centuries of the first millenium BC, the
importance of tia had declined; his replacement by T'ien bore a number of
social implications; and the veneration for the holy spirits may have been
growing stronger. In addition, whereas hitherto it had been the souls of the
deceased ancestors who had acted as intermediaries, from now on contact
with sacred powers was affected by specialists drawn from the human, living
world, who may variously be described as priests or shamans. ,_
V This development was accompanied by a change of treatment in Chinese
art, as may be seen principally in the decoration of bronze vessels. In the initial
stages, from perhaps 1600 to 950 BC animals are shown in full vigour, whom
man treats with affection, reverence or awe. They are creatures whose powers
are all too manifest, and there is little room for human beings beside them save
in a minor, subordinate capacity. However, in the bronzes which may be dated
from c. 900 BC, man is shown wielding strong powers with which he is capable
of challenging, fighting and even conquering the animal world; for by now the
animals are monsters which can harm man.
Similarly, in Chinese mythology,
the all-powerful animals of the early stages yield place to the human hero, who
is depicted possessing strength, courage and nobility with which to defeat the
monsters who withstand him.
Evidence for hybrid forms in Chinese art and mythology appears in objects
and literature that date from the fifth or fourth centuries BC and later. Such 1
evidence must be considered in full recognition that it had been preceded by
earlier, formative stages of cultural growth, and that those stages had lasted
for a millenium and longer. Two contradictory principles may possibly be
discerned. The first was that of identification of man with the animal world.
Tribal ancestries were traced to an animal; divination was conducted through
the medium of animal bones and shells; and attempts were made to make a
contact with the animal spirits of another worl<l by means of physical
assimilation (for example, the consumption of an animal's flesh, or wearing
part of an animal's fur or skin). The second ptjnciple which may be seen
operating in the Chinese treatment of the animal world is that of euhemerisa
tion, whereby the myths and gods of an earlier origin were transformed into
beings of authentic history, and animal figures were portrayed in anthropo-
morphic terms.
'-' The Shan-hai ching, or Classic of the Mountains and the Lakes, is a text which
reflects the southern tradition. In those parts of the book that date from c. 400
BC, we find a didactic, descriptive guide to the holy places and sacred
mountains of China. As would-be pilgrims or travellers we may learn here of
For the different stages in the treatment of animals, see Chang (1976), chs. 8 and 9.
I use the term euhemerisation in this sense, in common with writers on Chinese mythology; see
Derk Bodde, 'Myths of Ancient China', rpt. in Le Blanc and Borei (1981), pp.48f.
Man and beast 41
the location of those sites and of the rivers which lead us thither; of the
abundance of flora and the mineral wealth to be found there; and of the
animals known to populate such places, be they normal, freak or hybrid.
Probably the .text had in the instance to explain the features
of a very ancient set of pamtmgs or drawmgs; and from this explanation we
may learn not only of the bodily characteristics of these animals but also of
their cries, squeals or roars, and the consequences of consuming part of the
animal's flesh or donning part of its skin or fur. Thus:
370 leagues further east lie the mountains of Hsi-yang, with considerable supplies of
copper on the south and silver on the north side. There is an animal there whose body is
like that of a horse, with a white head, stripes like those of a tiger, a red tail and a cry
like that of a human singer. The beast is named the Lu-shu; and wearing a piece of its
skin will result in the birth of children and grandchildren (see Shan-hai ching, SPPY
l.2a; Yuan, p. 3; Mathieu, p. 6)
Various results follow the consumption of flesh or wearing the fur of these
and other animals who are to be encountered in the holy places. Fortunate
consequences included a cure from disease, relief from fear or bewilderment;
and in addition to personal blessings of the type just cited, the result could be
no less than a general blessing of peace and stability. However, the results
could sometimes be disastrous, such as the onset of floods or drought, mighty
enough to strike down a whole province. Or else, incidents of state could
ensue, such as would require a general call for military service, to the hardship
of the population. On a few occasions such calamities could follow after no
more than the sight of one of these strange creatures.
These then are the beasts who may be encountered in the flesh on the holy
mountains, and such are the results of assimilating to their persons. But our
text also decribes the .. or holy_spirits, known to reside in these sacred
hills, and informs us how they may best be served and worshipped. Altogether
there are some 400 spirits who are named and identified with particular sites,
and but some 80 are described in detail. In all cases they are hybrid form, ,
combming for example the features of bird and dragon, or horse and dragon,
or swine and snake. In a very large number of cases the holy spirits are
endowed with a human face, surmounting an animal's body, be it dragon,
horse or ox, sheep, snake, bird or pig. We may read in our guide book:
In the third stage, south, there is a total of 14 Mountains, stretching for 6530 leagues,
T'ien-yii to Mount Nan-yii. The spirits of those hills all have dragon's
bodies with human faces; they may be worshipped with the sacrifice of a white dog and
with prayer, and with rice used for the offering in grain (see Shan-hai ching, SPPY
1.1 la; Yiian, p. 15; Mathieu, p. 39)
or else:
In the third stage, east, there is a total of9 mountains, stretching for 6900 leagues, from
Mount Shih-hu to Mount Wu-kao. The holy spirits of those hills all have human
bodies with ram's horns; they may be worshipped with the sacrifice of a ram and with
42 Divination, mythology and monarchy
millet. Disaster brought about by wind, rain or flood follows the sight of this holy spirit
(see Shan-hai ching, SPPY 4.7b; Yuan, p. 113; Mathieu, p. 234)
Jn evidence of this type it would seem that the Chinese were clothing in
human garb those gods whom they first conceived in animal form. This
process is paralleled by the euhemerisation of Chinese myth, and may perhaps
be seen in the next subject to be considered.
In 1934 grave-robbers who were active near Ch'ang-sha, south of the Yangtse
River, lighted on the earliest piece of consecutive writing known in China
other than inscriptions made on oracular bones, sacred bronze vessels or a few
stone stele. The find was that of the famous silk manuscript of Ch'u, which
has suffered a somewhat chequered history since its discovery (see figure 3). As
a result it is only in the last decade or so that the results of professional
examination have become available.
The silk manuscript, which may be dated at c. 400 BC, is written and
illustrated in polychrome, with two major blocks of writing that run in
opposite directions. The text is partly defective; it is subject to grave difficulties
of palaeography and interpretation, and hardly any sentence may be read in
entirety. However, thanks to the efforts of a number of scholars, it may be
concluded that the manuscript gives an account of the creation of the orders of
heaven and earth and of the emergence of natural processes such as the
sequence of the seasons. It mentions the participation of the holy spirits and
other deities in these processes; at one point it reflects the influence of the
theory of the Five Phases that governed the cycle of creation, decay and
The manuscript text is surrounded by twelve peripheral figures, for whose
reconstruction we are largely indebted to Professor Barnard (see figure 4).
Each one of the twelve figures is accompanied by a short caption and a
descriptive notice ofits character, powers and activities. No certain interpreta-
tion of the figures can yet be given. But from the considerations which follow it
seems likely that they may represent twelve guardian gods or holy spirits,
' severally invested with powers of action for each of the twelve months.
Alternatively they may represent twelve shamans or intermediaries, wearing
masks and capable of communicating with such deities. Whatever the correct
interpretation may be, it seems likely that the twelve figures form important
evidence in the history of worship and exorcism in China.
The descriptive notices of four of the figures (nos. 2, 5, 8 and 11) associate
them specifically with the months or the seasons; and the forms of these
strange creatures is such that they invite comparison with some of the hybrids
who are mentioned in the Classic of the Mountains and the Lakes. Thus no. 7,
The manuscript is at present in the custody of the Metropolitan Museum, New York; the most
detailed study published to date is that of Noel Barnard (1972-3). The term 'Ch'u' is the name of
a state which controlled parts of the Yangtse River valley in the centuries before 221 BC.
I.e., the theory of the wu hsing, sometimes rendered 'Five Elements'; see SCC vol. II, pp. 232f.

Man and beast
[] 0 [) [ 0 0 c r; [, c '
n D o o o <'if, :f-' :r; D n o ,-,:, ! ' r: u Dr- i
$1 u tJ 1J L'. u .:) o it=:$ ...o .;:o;"':; l:. o c c n + _, 11 1

l\ 0 1f Ii D 0 1K:iFl'l Cl ill 0 0 LI 0 1k*IEE-t'rn
,(/ )\"
0 }Ja._-;( Ii G 0 C ,,,, C fJ 0 iJ 11: 3 0 lJ r; iJ
:::,$\\l: 0 0 1
Figure 3 The silk manuscript from Ch'u.
with its human body and three heads, and no. 10, with its dragon head and
bird's body, might easily fit the descriptions of spirits who are mentioned in
that text. .No: s.with its square, human head surmounting a birdlike body, and
no. 6, which is mterpreted as a snake consisting of two entwined bodies and a
single may be compared with the book's description of strange animals.
In add1t10n there are several possible, but not exact, comparisons which may
be drawn between passages in the book and no. 9, with its bird's body, human
face and antlers. 7
For no. 5, see SHC, ..sppy 5.26a (Y_iian, p.153; Mathieu (1983), vol. I, p. 312); for no. 6, see
SHC,.SPPY3.7a(Yuan,p. 78; Mathieu, p. 153); forno. 7, see SHC, SPPY5.23b (Yiian, p.150;
p. 305); forno. 9, see SHC, SPPY l. !Ob, 2.5a, 5.6a, 5.27b (Yiian, pp. 18, 27, 124, 156;
Mathieu, pp. 36, 53, 257, 317); for no. 10, see SHC, SP PY 1.4b, 5.44a (Y iian, pp. 8, 179
Mathieu, pp. 14, 369). '
44 Divination, mythology and monarchy
\ ! I I
\ 1 \;
~ ~ ~ ~
f '
Figure 4 The twelve peripheral figures of the Ch'u silk manuscript.
Man and beast 45
Both from the text that is under consideration and the unique silk manuscript
it is possible to trace features that are generally and almost universally
associated with shamanistic practice; for example, assimilation with an
animal's garb or guise; utterances made in the tongues of animals or birds; and
dependence on trees for ascent to or descent from another world. This last
feature may be seen clearly, and somewhat regularly, at the corners of the silk
manuscript. In other literary passages we learn of the prevalence of shamanistic
practice in south and central China, both for the period that is in question and
for several centuries later. Of a number of shamans who are named in the
Classic of the Mountains and the Lakes, one, called 'Hsienb', 'grasps a green
snake in his right hand and a red snake in his left hand'. This is at the summit of
Mount Teng-pao, where the host of shamans go up and down (see Shan-hai
ching, SPPY7.3a; Yuan, p. 219; Mathieu, p. 402). In another text that derived
from the south we have accounts of the shaman's work in restraining the souls
of the deceased from leaving earth for ever, and in inducing them to return so as
to enliven what is apparently a corpse;
and elsewhere we may read of the
shaman's use of grain and straw to effect a cure or to expel evil.
The twelve figures of the silk manuscript, including some hybrids, may thus
perhaps be taken to represent twelve spirits, or twelve shamans able to contact
them and to drive away evil influences. The suggestion compels us to take a
brief look at what may be said of the practice of exorcism in China. Probably
the most clear evidence, albeit for some 500 years later than the time of the silk
manuscript, is that of a description of a ceremony held at the imperial court;
this was the Great Exorcism, practised for the emperors of the Later Han
dynasty (AD 25-220), and probably stretching back to considerably earlier
At this ceremony the chief exorcist of state performed the main
rites. He was clothed in a bearskin which was furnished with four eyes,
presumably to ensure that it could command all-round vision. The object of
the ceremony is defined as being the expulsion of pestilence and evil demons
from home, court and palace; and the climax of the ceremony was reached in
an invocation to twelve named spirits, who were summoned to devour the 'ten
baleful influences'.
We therefore ask whether the twelve figures of the silk manuscript from
Ch'ua may be symbols of those twelve spirits who are defined by name for the
ceremony of the Later Han court; or whether they may be the intermediaries
sent to summon them. It may also be asked whether the twelve figures are
prototypes of other series or creatures which appear at other stages of Chinese
thought; for example, the twelve divisions of the cosmos, the heavens or the
day, who were later to be symbolised by twelve special animals;
or the twelve
See the Chao hun and Ta chao poems in Ch'u tz'u 9 and 10; Hawkes (1985), pp. 219f, 232f.
HNT 16.19a. 1For this ceremony, see Bodde (1975), pp. 75f.
11 For the duodenary series, see SCC vol. III, pp. 402f. The animals are used in their correct
sequence to designate years as, for example, 'the year of the dragon'. [For an early reference to
the symbolic use of these animals, see LH p. 990, Forke (1907-11), vol. II, p. 398.]
46 Divination, mythology and monarchy
guardian spirits of heaven, who may possibly be seen on some early Chinese
diviner's boards; or the twelve protective spirits of the household, invoked to
procure domestic safety.
These are open, but, we hope, not idle questions. The possibility that the
twelve peripheral creatures of the silk manuscript of Ch'ua, including some
that are hybrid, represent twelve protector figures gains support from other
considerations. In a number of graves that are situated in the same part of
central China there have been found a whole series ofsingle hybrid figures
carved in wood, and characterised by the corn bination of prominent antlers
and a long, protruding tor1,gue (see figure 5). In the great variety of figures of
this type, the one discovered at Ch'ang-t'ai-kuan is perhaps the most
dramatic, being finished in a variety of colours, goggle-eyed, and with hands
grasping a snake that is in process;0f being devoured. Further examples of
these features are seen in the figures painted on the coffin lids at the famous
grave no. 1, of Ma-wang-tui.
Among the multiplicity of elflike creatures,
animals, birds and hybrids there are a number of examples of antlered human
heads and snake-devouring monsters (see figure 6).
These are symbols which recur both within the Chinese cultural area and
elsewhere. Thus in the cave ofTrois Freres, the graves of Pazyryk and on the
Gundestrup bowl we may see an antlered head combined with the features of
animals other than deer, in portrayals of religious practice, shamanistic rite or
funeral service. The protruding tongue is seen in Egypt, in figures of the god
Bes, in Etruscan art and in India.
Of these features, the antler has been
interpreted as a symbol of superhuman authority, capable of warding off evil
spirits and guaranteeing everlasting life; and the tongue, which appears
sometimes with spots or drops upon it, is explained as being protruded as an
expression of hope and prayer for rain.
The figure of the snake-devourer is seen at Ch'ang-t'ai-kuan and Ma-
wang-tui, and it also appears in at least one literary context. This is in the
poem 'The summons of the Soul', which forms part of the collection of the
Songs of the South, and dates from perhaps the third century BC. The poet
describes T'u po, 'Lord of the Earth' as being 'nine-coiled, with dreadful horns
upon his forehead'. T'u po was known as the lord of the underworld, one of
whose functions was to expel demons and to devour snakes, before they had a
chance of consuming the body of a deceased person.
The suggestion that a connection may be traced between the practice of
exorcism, the symbols of antler and tongue and the figure of the snake-
devourer must remain speculative; and it may be considered in the light of no.
For this series, see Bodde (1975), pp. 90f.
13 This grave, which retained the incorrupt body of the countess of Tai, may be dated shortly after
168 BC [see Loewe (1979), eh. 2].
14 For a study of these symbols, see Salmony (1954).
1' See Ch u tz'u 9.5a, Hawkes (1985), p. 225.
Man and beast 47
Figure 5 Examples of Tongue and Antler figures; the figure from Ch'ang-t'ai-kuan is
on the right-hand side of the lower register.
2 of the peripheral figures of the silk manuscript from Ch'ua. The reconstruc-
tion and interpretation of this figure is subject to doubt; but it appears to carry
a double crest or a pair of antlers; and it is described by some scholars as
having a mouth with a divided tongue, by some as having a mouth that is
engaged in devouring a serpent.
Three other considerations or principles should be borne in mind in regard to
the hybrid forms of early Chinese imagery: the belief in bodily transformation;,,
the cults of immortality; and the composite figures of Chinese artists.
48 Divination, mythology and monarchy
Figure 6 Decorative figures from the coffins of tomb no. l, Ma-wang-tui.
Man and beast 49
The belief that it is possible for one living creature to be transformed into V
the shape of another is seen both in a grand way and in rationalist terms. In
mythology we hear that Kun, father of Yii.
the Great who saved mankind
from flood, was transformed into a turtle; in a scientific context, Wang
Ch'ung, who was writing during the first century AD, seriously cites examples
of frogs being transformed into quails, or sparrows into clams.
While there is
no knowing how prevalent such a belief may have been, it is possible that the
basic idea of transformation may have been present in the minds of some of
the artists .who fashioned hybrid creatures.
The Chinese have entertained a number of notions of paradise and
immortality. Many believed in the existence of the world of the hsiena, or
immortal beings, who could fly and roam at will throughout the unfverse, and
it was thought desirable to p_rovide deceased persons with the means of access
to such a mode of existence. To undertake a journey through the empyrean
and to achieve a life of eternity, a diet of jujubes or the juices distilled from jade
was often prescribed; and talismans with particular features were usually
buried with the dead. It is in such talismans that we may find a clue. Both the
literary sources and the example of early art give us examples of hybrids who
are equipped to escort the dead on their journey. The Cherubim or Seraphim,
or the Icarus, of the Chinese world are seen in two forms; either as
combinations of a bird's body and a human face; or as a human figure at
whose back wings have started to sprout.
The concept of transformation from one animal form to another and the v
desireto acquire immortality are possibly seen together in. one particular type\
of hybrid whose appearance is quite rare. is the figure of a human heaV'
and body, with a serpentine coil in place oflegs;
This appears at the apex of ,
the Ma-wang-tui, possibly.representing the arTival of \/
tfi! image is also seen in a stone
rehef ofa later period; here there is a pai'r of suppliants who have apparent!
received their draft of the elixir from the Queen Mother of the West and are
l!!cewise being changed into immortal beings (see figure 7). So far as is known,
the image of transformation into seri}entine forms as a means of achieving
liiiiiiortality does not appear in other cultures.
---The third principle to be borne in mind is that of the composite figure. In the v
earliest stages of Chinese art that are examplified in the bronzes of the
Shang-Yin period (c. 1700-c. 1045 BC) there are a number of instances where
two or more animals are shown in close association, in their entirety. In the art
of south or central China of c. 500 BC there are a number of beautiful
combinations of pairs of birds and animals, often snakes; these are sometimes
fashioned so as to form a drum-stand (see figure 8). But perhaps the most
conspicuous example of all composites of this period is the lacquer screen
16 LH, pp. 54f; Forke (1907-11), vol.I, pp. 325f.
17 This is to be distinguished from the pair of figures, each with a serpentine body surmounted by
a human head, which represent Fu Hsi and Nii Kua.
50 Divination, mythology and monarchy
Figure 7 The Queen Mother of the West, with suppliants; from a stone relief'.
from Wang shan, of c. 400 BC; the 51 animals on the screen include three paf
of birds who are engaged in devouring snakes (see figure 9, 1).
Composites of this type call to mind no. 3 of the peripheral figures of the
manuscript from Ch'ua. Difficult as this flgure is to interpret for certain, it.
been suggested that it is formed of a bird's body with a snake's head; an
may be asked whether a hybrid of this type may have developed fr
composite figures in which bird and snake are shown complete. I.t is
possible to look forward in time to the composite snake-cum-tortoise w
makes its appearance in Chinese art from perhaps 50 BC or so. This i
bears a cosmological significance, being one of the four animals that sig
four of the five directions, or four of the five phases of creation.
snake-cum-tortoise symbolises the north, the extreme point of Yin, the c
of the wintry season.
The heyday of the hybrid in Chinese art and literature may be placed int
fifth or fourth centuries BC; its home region seems to have been centred ou
great kingdom of Ch'ua, that bestrode the valley of the Yangtse River.
before long China became unified under Ch'in, the first of the imperiaf,c
dynasties, which was founded in 221 BC. Uniformity and
An important clue to the origin of this symbol which has yet to be explained may perhaps
found in a series of paintings of the recently excavated tomb of Pu Ch'ien-ch'iu, near Lo-y,.
which may be dated between 86 and 49 BC. Three of the four animals appear in the fi
which are well known in the following century and later, i.e., the dragon (for the east), the
(for the south) and the tiger (for the west). The fourth figure, for the north, is a composite
sheep's head, '!tiger's tail and wings, and this is accompanied by a further hybrid of a ?c.at's
and body, wings and a single horn. See WW 1977, 6, pp. l!f-11 and plates 2 and 3; [and
(1979), pp. 59, 134 note 13, 140 note 95, and figure 16].
Man and beast 51
Pairs of birds and other animals in the art of Ch'u.
52 Divination, mythology and monarchy
were promoted in art, literature and mythology. If the lively, vivid styles and
the strange tales of the south were not entirely suppressed, there was a
sufficiently strong impetus from the north to propagate other art forms; with
the propagation of the 'Confucian' cosmology, shortly after 100 BC, the
attention of Chinese artists was directed to other symbols, as befitted the new
and orthodox modes of thought. The snake-cum-tortoise, to which reference
has been briefly made above, derived from just such developments.
The hybrid creature fell out of fashion. When we meet him again, he does
not spring live from an artist's intuitive imagination; he is a creature of a
secondary order. It has been observed above that the extant text of the Classic
of the Mountains and the Lakes probably originated as explanations which
were intended to accompany a series of ancient paintings. By the Later Han
period (AD 25-220), those paintings had long since perished, although the
explanatory text survived. We find that artists of the day were portraying
creatures of fancy which may have been inspired by that text or which were
intended to clarify it. Hybrids appeared once more, by now in stone relief's,
carved deliberately to illustrate concepts which were known second-hand,
from literary sources; they perhaps lack the immediate appeal to the
sub-conscious that is carried in some of the hybrids of an earlier age.
These secondary versions of hybrids are seen most conspicuously in the
sculptures and reliefs ofl-nan, which are dated variously in the third century
and later (see figure 9, 2). The ideas persist, in the human-faced bird, who is
seen in the company of two-headed birds or two-headed deer; or in the
nine-headed monster K'ai-ming, who guarded the holy axis mundi of
K'un-lun. These figures are not identified by explanatory labels which relate
them specifically to the animals or animal-spirits of the Classic of the,
Mountains and the Lakes; but it is to such origins that they may be traced; and
all credit is due to the contemporary Chinese artists who drew on such
material at a time ~ n official efforts were being made to counteract its spirit.
Some fifteen hundred years later, at a time of intensive literary and
bibliographical activity, new editions were being prepared of that famous text.
Artists provided a new set of illustrations; craftsmen cut the blocks, so that
these could be printed for insertion with the text; and lest an ignorant reader
should be perplexed, each one of the new illustrations bore its own descriptive
caption, relating the illustration to the specific passage in the book. These were
the vignettes of strange creatures and the groups of animals set within a
landscape which were first cut in the seventeenth century and have been
adopted by publishers ever since (see figure 10). For the wheel has come full
circle; the original depictions of the holy spirits believed to be attached to some
of China's mountains had perforce been replaced by literary descriptions.
With the passage of time, Chinese artists sought to recapture the concept of
those gods. Separated as they were by centuries from the original beliefs, they
had little first-hand devotional experience on which they could call. They have
none the less succeeded in presenting posterity with woodcuts that are
graceful, imaginative and vigorous.


-<.!1!1!<lli' ; +
' "j;fl, 1 t1t1i:.;:;"ii';a .



1 I
Water, earth and fire: the symbols of the
Han dynasty
Between the inception of the Ch'inb empire in 221 BC and the restoration of
the Han dynasty in AD 25, the concept of imperial sovereignty underwent
considerable change; religious issues had entered into questions that had
hitherto been largely subject to material considerations; and claims to rule
with legitimacy had become dependent on establishing links with spiritual
powers. In the initial stages, the right to govern a Chinese empire was claimed
by virtue of practical success, which had been witnessed in the elimination of
rivals and the establishment of an authority that was acknowledged through-
out the land. By .the time of Wang Mang and the emperors of Later Han, the
claini heen linked. diEectly with the superhull1an
ro"".er of Heaven and the bestowal of its order or Mandate; the theory that
was to be invoked throughout China's imperial history had become accepted
as orthodox.
This change of attitude was fully consistent with other religious and
intellectual developments that affected policies of state and decisions of
imperial governments. Simultaneously, philosophers and statesmen were
paying considerable attention to the all-important question of the choice of
symbol, or cosmic element, with which the dynasty's future was linked and to
which it looked for protection.
Different elements were adopted by successive
governments in Ch'inb and Han times; and as some confusion is evident in the
minds of early Chinese writers, it is desirable to establish the sequence of
symbols that were actually chosen. From the evidence which is summarised
below, it appears that the symbol of Water was adopted by Ch'in and Former
Han, until its replacement by Earth in 104 BC; that Wang Mang re-adopted
Earth; and that the change to Fire took place after the accession of
Kuang-wu-ti, first of the Later Han emperors. This change had already been
suggested by some of the leading thinkers of the last decades of Former Han.
At the same time a new principle had been recognised whereby the succession
from one element to the next was governed.
For the stages whereby these changes came about, see chapter 4 below.
I.e., one of the Five Elements, wu hsing, more correctly translated as Five Phases; see Major
56 Divination, mythology and monarchy
One of the earliest statements which links these symbols with a particular
ruler or dispensation is seen in the Lii shih eh 'un-ch 'iu.
The passage assigns the
protective powers of Earth, Wood, Metal and Fire to the Yell ow Emperor, the
Hsiaa dynasty ofYiia, the Shang dynasty of T'ang and the Chou dynasty of
Wen wang respectively. The text observes that Water will be the apportioned
lot of the successor to Chou. It may be noted that the order in which the
elements figure here is the one whereby each one overcomes or conquers its
predecessor (hsiang shengr.
This order remained unquestioned in dynastic
practice until the end of the Western Han dynasty.
No less than four separate passages of the Shih-chi refer to the deliberate
adoption of the element Water, shortly after the establishment of the Ch'in
Although some doubts have been cast on the authenticity of the
account the occurrence of four references, without inconsistency, argues
against theory that the incident was interpolated after the completion of the
Shih-chi. The principal passage explains the choice of Water as following
Chou's protection by Fire, and Chou's replacement by Ch'in.
\J There is no direct statement in the histories to the effect that the Han
dynasty deliberately adopted Water, but there are several reasons why this
may be inferred. In 205 BC, before his establishment as emperor, the king of
Han ordered the inauguration of worshi to a fifth a , to
supplement the devotions that were already being paid to the other powers of
White and Red_The action be taken as the
recognition of the Power of Water, although it is not stated that this was given
preferential treatment over the other four elements,
either at this juncture or
after the king of Han had accepted the title of emperor in 202 BC. However, it
is clc::ar that Water had been adopted by the Han court, early in the dynasty,
from the account of two suggestions that it should be replaced by Earth. One
of these originated from Chia I, shortly after the accession ofWen-ti in 180
he is said to have rested his case on the plea that over twenty years had
since the foundation of the dynasty, whose state of harmony
warranted a change of protocol. Presumably he felt that sufficient time had
passed to show that the dynastic change was and that it would be
right to show a symbolical recognition of that l\appy state of affairs.
An identical suggestion which was made in 166 BC came from Kung-sun
Ch'en a man of Lu who is not known to have held any office. Like the
of Chia I it was rejected, but the appearance of Golden Dragons in
3 13.4a.
I.e., Wood, Metal, Fire, Water, Earth. For the various orders in which the elements were
arranged, see SCC vol. II, pp. 253f.
5 SC6, p.23; SC 15, p.122; SC26,p. 9; andSC28, p.19 (MHvol.II, p.129, and vol. Ill, pp. 328
and 430). For doubts regarding the authenticity of the incident, see Kurihara (1960), pp. 45-91,
and Kamada (1962), pp. 42f; [for acceptance of its authenticity, see Bodde (1986), p. 97] ..
SC 28, p. 36 (MH vol. III, p. 449); HSPC 25A.l 7a; Fujikawa (1968), p. 56 expresses the view
that in the prevailing pre-occupation with settling the empire, the Han government was ready to
accept existing protocol and practice.
1 SC 84, p.21; HSPC 48.lb.
Water, earth and fire
the following year lent some support to Kung-sun Ch'en's proposal. For, gold
m: yellow. . and the emperor relented to thepoint of
ordering a set of new regulaifoiis for protocol. In recognition of his services
Kung-sun Ch'en was appointed to be an Academician (po-shih), but his
opponents succeeded in preventing any change of patron element from taking
It was not until forty years later, in 104 BC, that the change was
actually brought about, probably with the support ofSsu-ma Ch'ien. 9 At this
time the strength of the Han empire had reached its highest point, before the
decline in Han arms and the need for retrenchment had become apparent. The
change of element, to Earth, was one of several measures which were designed
to match material achievement with symbolical .recognition of cosmic
blessings. Han had conquered its enemies, who included not only its
predecessors, Ch'in, but also those along its borders; so too must the dynasty
recognise that Earth had conquered Water, the element of Ch'in.
Almost a century elapsed before the question of the appropriate dynastic V
element next arose. In the meantime a change had overcome the predominant
attitude towards the universe, man and the state.
.... .. and derived partly froni'tlle
philosophy of Tung Chung-shu (c. 179 to c. 104). The practical or realistic
view of the state and its purposes that had been modelled on the Ch'in empire
had given place to a respect and longing for a state that was based on the
ethical ideals and the less harsh dispensation that was ascribed to the kings of
and matters,lhe won;!!!l'_Q[U1e Powers (ti2 and other
deities was givmg place to thatorrreil or Heaven. Quite consistently the
attitude towards the Five Elements, or Phases, also changed. A new view of
the principle whereby the Five succeeded one another was witnessed in the
opinions of philosophers, and put to practical effect in the hands of politicians.
The change affected the choice of element by Wang Mang, Kung-sun Shu and
Kuang-wu-ti; and it engendered the retrospective view that the appropriate
element for Western Han had been neither Water nor Earth, as had been
maintained, but Fire.
Hitherto it had been held that the elements succeeded one another by virtue
of conquest. It was now put about, on the basis of earlier thinking, that an
element rose to a position of dominance by natural growth from its
predecessor (hsiang shengh). The protagonists of the new opinion included
formative personalities such as Liu Hsiang (79-8 BC) and Liu Hsinb (c. 46 BC
to AD 23), who observed that the true sequence should proceed from Wood to
Fire, without any interloper.
They also cited an anecdote that concerned Liu
Pang, before the establishment of the Han dynasty. According to the full
SC 10, p. 32 (MH vol. II, p. 479); HSPC 25A.i9a.
SC 12, pp.j8-9 (MH vol. 111, p. 515); HSPC 25B.23b. For the accompanying changes, see
HSPC 6.3la,b (HFHD vol. II, p. 99) and CC, p. 31.
See Pan Ku's appreciation, HSPC 25B.23b. The order of the elements by natural growth is
Wood, Fire, Earth, Metal, Water.
58 Divination, mythology and monarchy
account of this story,
Liu Pang once put a large serpent to death; at the time
it was said that the ~ r n n t was the incarnation of the powe.r,ofWbite,, and that
it was as an incarnation of the power of Red that Liu Pang had succeeded in
killing it; and it is further related that when, some time after the incident, Liu
Pang had risen to be king of Hanh, he made the point of according precedence
.Jo.Red .\l,mo11g the colours.. . . . .. ..... ... . .. . ........ ""
This association of Liu Pang and the protection of Red, the colour of Fire, is
related in identical terms in the Shih-chi and "the Han shu. The s ~ l
corroborative statement to suggest that Western Han paid any special
attention to Red is seen, rather curiously, in the record of Kung-sun Ch'en's
proposal of 166 BC.
Following the rejection ofhi.s proposal, the emperor
made a progress to pay his respects to th'efiyeJ'owers (ti) at Yungc, and the
colour Red was given precedence in the robes of the officiants at the ceremony.
The absence of further corroboration has led some scholars to the conclusion
that the story of Liu Pang's slaughter of the serpent was an invention of the
first century BC.
In a somewhat strange incident of 5 BC the view was seriously put forward
that the Han dynasty had reached the end of its allotted span and that its
authority required renewal. The idea had indeed been propounded during the
previous reign, of Ch'eng-ti (33-7 BC), when the lack of an imperial heir had
given rise to dynastic problems, political intrigues and religious controversy.
The opinion which was voiced in 5 BC won acceptance to the point of
persuading the emperor and his government of the need to effect certain
changes, as a symbolical means of seeking a renewal of authority. A new
regnal title was adopted, together with a new title with which the emperor was
styled; and a formal change was made in the divisions of the day, for purposes
of calculating time. It is evident that those who believed that they were
witnessing the end of a dynastic cycle saw that cycle in cosmic terms, and it is
perhaps surprising that no direct suggestion was made for adopting a new
element as patron of the dynasty. There was, however, one allusion to the
growing power of Fire; the revelation that the end of the cycle was
approaching was ascribed to Ch'ih ching tzu, who is described as a mystic of a
very advanced degree;
the term Ch'ih ching tzu may be rendered as 'The
essential spirit of Red'.
The documents and procedures that attended the accession of Wang Mang
as emperor of the Hsin dynasty apparently assume that the elements succeed
each other naturally rather than by conquest; and they accept that Han had
existed under the patronage of Fire, which was due for displacement by Earth.
One of Wang Mang's own statements (6 January AD 9) refers specifically to
the incident of 5 BC and its prophecy of the need for dynastic renewal. B There
SC 8, p.12 (MH vol. II, p. 331); SC 26, p. 35 (MH vol. III, p.448); and HSPC 25A.17a.
SC 10, p. 33 (MH vol. II, p. 480) and HSPC 25A.l 7b.
See HFHD vol. III, p. 453, note 24. l, citing Ku Chieh-kang.
HSPC 75.3lb and CC pp. 278f. The term chen )en is sometimes rendered 'perfected adept'.
Water, earth and.fire 59
is also a definite statement in the proclamation that was circulated throughout
the empire in the autumn of AD 9, '
seeking to prove how earth had already
taken the place of Fire, and how the Mandate of the Han dynasty had thereby
become exhausted.
Wang Mang's choice of Earth as his patron is specified in
the proclamation issued immediately after his accession, on 10 January AD
The duty of adopting a patron element next fell on those who sought to
found imperial regimes after Wang Mang's death, i.e., Kung-sun Shu and Liu
Hsiu, the future Kuang-wu-ti. In both cases the theory of the natural
succession of the elements was accepted without demur. Kung-sun Shu
declared himself emperor on the strength of the possession of territories in
west China, in AD 25; as he regarded himself as the natural successor to Wang
Mang and his element Earth, he gave out that his dynasty would thrive under
Metal, the element of the west.
Liu Hsiu, however, who chose his element in
the year after his accession (i.e., in AD 26) based his decision on a somewhat
different set of assumptions. By choosing Metal, Kung-sun Shu had accorded
Wang Mang a rightful place in the sequence of dynastic authorities. When Liu
Hsiu chose Fire,
he was resuming what he believed to be the appropriate
element for the Han dynasty. In doing so he not only sought to unite his
regime, in cosmic terms, with that of Former Han; he was also branding Wang
Mang as an usurper who had never possessed a legitimate right to rule.
In two key passages of the Han shu it is asserted that Han, i.e., Former Han,
had served the tutelary element of Fire. One features in the historian's appreci-
ation of Kao ti; here the statement is linked with the legitimate succession of
Han from Yao, who had also been blessed by Fire, and with the omen of Liu
Pang's success, as seen in the story of his encounter with the serpent.
second passage occurs in Pan Piao's all-important essay on the nature of
where the same points are made. Han is also assigned to the protec-
tion of Fire in another chapter of the Han shu, which is based on the writings of
Liu Hsin. This is the treatise on measurements and astro-calendrical science;
in accordance with the order of the natural succession of the elements, Fire is
denoted as the element of Yen ti and of Yao, before the cycle had brought it
round to Han.
An interesting statement is recorded for about AD 76 from Chia K'uei, who
was a descendant of Chia I and an enthusiastic exponent of the Tso chuan. He
pointed out that the Liu family's claim to be descended from Yao, and the
Han dynasty's consequent devotion to Red, depended solely on passages in
the Tso chuan, there being no support for such claims from the Five
'' See Wang Mang's memorial to the Empress Dowager, of 6 January AD 9 (HSPC 99A.34b;
HFHD vol. III, p. 251).
HSPC 99B.9aff, HFHD vol. III, pp. 288f.
HSPC 99B.9b, lOb, HFHD vol. III, pp. 290, 293.
HSPC 99A.36a, b, HFHD vol. III, pp. 258-9. 19 HHSCC 13 (biog. 3), 16b.
HHSCC 1A.18b.
HSPC IB.26a, HFHD vol. l, p. 150.
HSPC !OOA.lOb; de Bary (1960), vol.I, p. 177.
' HSPC 21B.46a, 47a, band 72b.
60 Divination, mythology and monarchy
In a somewhat exceptional passage of the Han shu, where the
elements are arrayed in the order of conquest, the text is simply reiterating the
view set out in the Lii shih ch'un-ch'iu, according to which Fire had been the
tutelary element ofChou.
A belief that Fire had been the element for Former
Han was accepted by the commentator Tsan, who is probably to be identified
as Hsiieh Tsan (c. 350-90).
The evidence which is cited above indicates that it was a matter of no small
importance to the emperors, philosophers and statesmen of Later Han to
demonstrate that the element of Fire, thought to have watched over the
fortunes of the Liu Pang and his dynasty, was likewise the element to which
they themselves should look for protection; and it was equally important to
establish a link with the blessed Yao. These questions are of some concern to
the history of state cults and religions of China, to which Professor Eichhorn
has made valuable contributions. I am grateful for the opportunity to include
a small note in a volume which will permanently record the debt which friends
and students owe to their colleague.
HHSCC 36 (biog. 26).14b, 15a.
25 HSPC 25A.9a; for the passage from LSCC, see note 3 above.
2 See note to HSPC 1B.26a; Loewe (1960),p. 134 note2, which cites the view of Hu Shih. For the
views of an earlier commentator (Ying Shao: c. 140 to before 204), see the note to SC 8,
pp. 12-13 (MH vol. II, p. 331).
The Han view of comets
The three Han tombs of Ma-wang-tui will long rank among the most
important archaeological discoveries of China in recent years. Quite apart
from their value as examples of pit graves constructed in central China, their
contents included source material of the first order of importance for students
of Chinese religion and art, literature and science. The fortunate occurrence of
an inscribed strip in tomb no. 3, of the year 168 BC, provides a precise
indication of dating; tomb no. 1, which contained the incorrupt body of the
Countess of Tai and the famous painting that was interred with her as a
talisman, was constructed shortly afterwards.
For students of Chinese palaeography, textual history and literature, the
fifty-one items of manuscript text found in tomb no. 3 are of especial value.
They include the first finds of texts written on silk, other than the single piece
that derived from Ch'ua during the Chan-kuo period. Some of the texts from
Ma-wang-tui are unique; others correspond partially with the received
versions of works such as the Chan-kuo ts'e. Some are copies of well-known
texts which were subsequently subjected to editing and standardisation, and
thus suggest how the work of scholars and commentators such as Liu Hsianga
(79-8 BC) or Wang Pi (226-49) may have affected the transmission of earlier
Above all the content of the 51 items is extremely rich and varied, ranging
from philosophy and political theory to historical record. There are also
manuals on mantic practices, handbooks on medical subjects and astronomy,
and a few maps. In presenting a preliminary account of part of one
manuscript, which concerns both science and divination, I am glad to
acknowledge the encouragement and help of Professor Nathan Sivin.
In the following pages I shall attempt to describe certain features of the
manuscript and the terms which it uses to denote comets. These will be
compared with literary usage, and particular attention will be paid to a comet
known as the 'Banner ofCh'ih-yu'. After considering the Chinese view of the
origin of comets, there will follow a general assessment of the value of the
62 Divination, mythology and monarchy
The silken manuscript now entitled T'ien-wen ch'i-hsiang tsa chan
The manuscripts found at tomb no. 3 Ma-wang-tui included a large number of
fragments of a silken document that had originally measured 150 by 48
centimetres.' Despite the destruction of some parts and the obliteration of all
signs of writing in others, it has been possible to reconstruct the original form
of the document, whose main part consisted of six horizontal bands or
registers. Each one of these bands was subdivided vertically, in columns, with
between 20 and 50 entries each, and the total number of entries on the
manuscript, either complete or fragmentary, amounted to about 300. The
entries comprised diagrams and text, inscribed in black or red, or in a
combination of both colours. Below these six registers there was a further set
of entries on the manuscript. These were arrayed in three bands, horizontally,
each of 13 to 26 entries with text, but without diagrams. 57 entries survive
from this part of the manuscript (see figures 11, 12).
The text and diagrams are concerned with the identification of certain
climatic and astronomical features and the prognostications that are appro-
priate to each one. The features under examination fall into four types which
are generally, but not precisely, grouped together in different parts of the
document. From the upper to the lower registers, the entries concern:
(a) clouds; various shapes of cloud are illustrated as animals, plants or
(b) vapours or emanations of energy;
these include mirages,
shown by
diagrams of trees and some indefinable objects, and mists, which are
shown in various configurations of circle or line without explanatory
text; some of the entries are for rainbows.
(c) stars and constellations
(d) comets; there are altogether 29 entries (now designated as nos. 612-40),
which all except one are complete (i.e., no. 632, which has a diagram but
no text). This section has been described as the most complete and
valuable part of the manuscript; it is the only part for which photographs
and transcriptions have been published in full so far. [For photographs
of a few other parts of the manuscripts, see chapter 9 below note 2]
The terminus ad quern for the manuscript can be taken at 168 BC, the date
assigned to tomb no. 3 Ma-wang-tui.
Internal references in the text to the
states of Hane Weib and Chao lead to the inference that it cannot have
originated before 403 BC, when those states emerged at the dissolution of
1 [For a list of the manuscript textual and illustrative items found at tomb no. 3 Ma-wang-tui, see
Loewe (1981).] Chinese scholars now refer to the manuscript that is under study by the title
T'ien-wen ch'i-hsiang tsa chan. See Ku T'ieh-fu (1978), translated by Donald J. Harper (1979).
The scientific evidence is considered by Hsi Tse-tsung (1978). In terms of Han measurements,
the dimensions of the manuscript are 6.6 by 2 feet. For facsimiles, see WW 1978.2, plates 2,3
(reproduced here on pp.63-4). 2 I.e., ch'i'. 'Shend; see Ku T'ieh-fu (1978).
For the dating of the three Han tombs of Ma-wang-tui, see Loewe (1979), pp. 27f.
1lf- ,....._
7r !
5 '
0 *

' w


7rJ '
0 *
Figure 11
616 615 614 613 612
t: !It El

a If
!!J 't 'l' I
1j' 1.f
111 11
;f. jf
' fii! A. '
A. tll
;fr Ji!.

ff. '
.'.f- *
" t
M. 1'

624 623 622 621 620 619 618
,,..... l*l
Jf 'f 'i!X 11. :.f 11

! * }!
!' liiJ
0 '}! 'llt
0 '}!
' J!
}! 0 6

7cflf Ji!, If


1i 1i. ,...._

* tt

iW t: ii
. ........
t" 8.......,
:f4. 1f

:ff ff 111 "F w a
Illustrations of comets, from the silk manuscript from Ma-wang-tui.
........ f
* .k
........ }!

' f
t '

, A.0tk_'-'v .
CHJ!<l:J. !ll
:it-* a
Figure 12
632 631 630 629 628 627 626
....... *
i5" A
;i. r J! t!!l
1t :
!}3 * }!
....., ';!!
'E '

t '
}f. 1t

:it7t 'f
il- 0 1t. .a ' JiiJ '
639 638
637 636 635 634
0 1i
' 4; 'M '

* .x.
t .. f
* .I.
A 'J.
' '
-it ';t
,....... '
0 *
-{:; ;j'
f" D
;i. .._,
Illustrations of comets, from the silk manuscript from Ma-wang-tui.
The Han view of comets 65
Chin. The script of the manuscript is early, or proto, li-shu type, being similar
to that of other manuscripts from the same tomb, and retaining some
characteristics of seal writing. The appearance of the character pang in the text
cannot necessarily be accepted as a reason for dating the manuscript before
the reign of the first Han emperor (Liu Pang reigned 202--195 BC), as it
appears in other divinatory texts that were certainly written during the Han
Ma-wang-tui itselflies well within the territory of the pre-imperial kingdom
of Ch'ua and the kingdom of Ch'ang-sha which was established at the outset
of the Han empire.
Internal references treat the name Ch'u with a certain
degree of prominence and many of the artifacts and the artistic motifs of the
tomb's furnishings bear the characteristic imprint of the culture and mythol-
ogy that is associated with that state and the Yangtse River valley. It may be
suggested, tentatively, that the manuscript derived from that part of China,
perhaps within a century or so before the foundation of the Ch'in empire in
221 BC.
The place of the manuscript in Chinese astronomy and divination
Apart from the forty diagrams of physical exercises or callisthenic postures of
another manuscript,7 the document under study is the only record of the
period found so far which includes illustrations of a nature that would now be
regarded as scientific. That such diagrams not infrequently formed parts of
early Chinese writings may be seen from the entries in the bibliographical list
of the Hanshu, and in the references of much later catalogues to the survival of
a text but the loss of its accompanying illustrations.
Whereas the diagrams of
callisthenics include no more than a title or caption, the texts that accompany
the illustrations of meteorological and related phenomena, of stars and of
comets are considerably longer and more informative.
The manuscript may be considered together with a further text found at
Ma-wang-tui, also on silk, which reports the times and locations of the rising
and setting of the planets over the years 246-177 BC.
The two documents
constitute the earliest surviving original Chinese writings on astronomical
matters; for the works that are ascribed to the two famous astronomers Kana
and Shih
, of the Chan-kuo period, have long since disappeared, except for the
citations preserved in later writings. The basic evidence of the observations
and calculations made by Han, or earlier astronomers may otherwise be found
' I.e., the mantic text from Mo-tsui-tzu, which is dated in the Ho-p'ing period (28-25 BC).
For maps, see Loewe (1979), p. 10, and Twitchett and Loewe (1986), pp. 39, 125 .
' See WW 1975.6, 6f and figures 1-2 (pp. 8-9), and KK 1975.1 plate 9.1.
' For example, see HSPC 30.64a for a summary of the entries on military matters which included
43 scrolls of t'u"; and HSPC 30.68a for a single entry for a work on astronomy in 232 scrolls of
silken diagrams (po t'u); for references in the T'ung chih etc., see WW 1978.2, 3.
9 This manuscript was entitled Wu hsing chan; see WW 1974.11, 28f and 37f, and plate 4; and KK
1975.3, plates 2, 3.
66 Divination, mythology and monarchy
in the lengthy treatises of ihe Standard Histories, and in a few diagrams
painted on the walls or ceilings of tombs.
Both this document and the record of planetary movements must surely
result from cumulative, steady and sustained observation and study; for it is
not possible that a single observer could have seen and noted all the
phenomena that are recorded so meticulously. The manuscript forms a
powerful testimony to the capacity of Chinese observers in the centuries that
preceded the imperial age and to the sophisticated nature of their records. It is
difficult to believe that the manuscript under consideration was the first of its
kind, compiled without the support of earlier diagrams or descriptive
material. The suggestion that earlier documents of this type had existed
previously is partly sustained by a few citations from lost texts that are cited in
the Shih-chi or Han shu.
Records such as the Ch 'un-ch 'iu report the
observation of comets from as early as 613 BC.
In addition to recording the results of systematic observation, the manu-
script presents the prognostications appropriate to the phenomena that are
described. In doing so it introduces a mantic element into the subject. Now,
whereas a clear distinction may be drawn between science and divination at
their extreme points, there remains a wide middle ground between the two
wherein both activities merge and affect one another. This principle may be
seen in the study of the major methods of divination practised in China,
whether with bones or shells and fire, or by the manipulation of stalks and the
I ching. In each case the intuitive presentiments of a seer become subjected to
the intellectual processes of standardisation, regularisation and explanation.
Similarly geomancy (jeng-shui) seems in its origin to have been based on two
approaches; that of the seer who unconsciously divines the existence of the
unperceived properties that inform a site, and that of the observer whose
graticulated compass enables him to relate a site and its qualities to the
measured rhythms and changes of the universe. The manuscript from
Ma-wang-tui likewise takes account of both the intuitive and the intellectual
approaches; for it presents the results of observation together with guidance
with which to interpret the inherent meaning therein on grounds which we do
not yet comprehend. It may also be remarked that the treatment of stars and
10 For the Standard Histories, see SC 27 ('T'ien kuan shu'), MH vol. III, pp. 339f; HSPC 26 and
27 ('T'ien-wen chih' and 'Wu hsing chih'); HHSCC (treatise) 10-12 ('T'ien-wen'); CS 11-13
('T'ien-wen'), Ho Peng Yoke (1966). For secondary writings, see Eberhard (1933); Dubs
(1958); SCCvol. III; Sivin (1969), 52f; Maeyama (1975). For the Hsing ching and its. see
Maeyama (1977). For recently found manuscripts, see Loewe (1977), p. 123. To 1t may he
added that (i) constellations were included in the decoration of the t?mh o_f a Chmese migrant
at Takamatsuzuka, Japan, dated in the seventh or eighth centunes; (n) the Twenty-eight
Mansions were named on the cover of a lacquered box, found in a Chinese tomb dated shortly
after 433 BC (see WW 1979.7, 40f).
See SC 27, pp. 87, 93 (MH vol. III, pp. 404-9) and HSPC 26.l 9bff fo'. citations from
astronomers Kan and Shih; see HSPC 27C (2).19b, 20a, 20b for c1tat10ns from the Hsing
chuan; and HSPC 27C(2).2la for citations from Shen Hsii [or Ju], also Ch'ien-fu fun 26 ('Wu
lieh'), p. 304.
SSC 19B.13b.
The Han view of comets 67
comets together with features of climate and atmosphere in the same text
forms a valuable link in tracing the antecedents of feng-shui during the Han
Terms used to denote comets in (i) the manuscript and (ii) the Standard
( i) In the manuscript
The text that accompanied 28 of the 29 diagrams of the comets14 is in general
of the same form, comprising the name and title of the particular type of comet
that is displayed, short remarks about the duration of its appearance, and a
general prognostication of the events likely to ensue. A number of the names
are botanical terms. Perhaps somewhat surprisingly some of the names are
used to define two or even three different types of comet; in these cases the
prognostications, though phrased somewhat differently, are largely identical
in purport. At least eight of the total number of 20 different names appear in
A large proportion of the prognostications concern military
fortunes, as do those for comets whose observation is recorded in Han
It may also be remarked that, certainly at a later date, divination
which depended on another type of natural feature (i.e., the behaviour of the
winds:feng-chiao) was also largely linked with military matters. 17 Some of the
prognostications of the manuscript are attributed to named masters or seers,
i.e., Pei-kung or Yao; these are otherwise unknown.
The names of the comets that are listed on the manuscript follow, overleaf
(for a full transcription of the texts, see WW 1978.2, 5-6).
(ii) In the Standard Histories
The compilers of the Standard Histories for the Han period noted the
appearance of comets in two different ways. They feature in the chapters of
Imperial Annals, along with accounts of other phenomena, in so far as they
were thought to have a bearing on imperial destinies and dynastic con-
tinuity. The references in these chapters are terse, and they are not usually
See pp. 19lff, and Ngo van Xuyet.
The text for no. 632 is now unreadable.
For the appearance of eight of these terms together in a work attributed to Ching Fang, see
pp. 75fbelow; see also LSCC 6.IOa for the mention of five named comets, including some of
those given in the manuscript.
For example, HSPC 26.52a, for 134 BC; HSPC 26.53a for the Yiian-feng period 110 105 BC;
HSPC 26.54b, for 77 BC; HHSCC (tr.) I l.2h, for AD 76.
See chapter 9 below forfeng-chiao, a method of divination according to the direction and force
of the winds and the times at which they arose. There is considerable evidence for its practice in
both Former and Later Han, and one passage refers to this as part of the la ceremonies at the
start of the year (see SC 27, pp. 78f, Ml!, vol. III, pp. 397f). The Hou Han shu mentions a
number of named specialists at this type of divination; for its association with military matters,
see Li Ch'iian, Shen chi chih ti T'ai-po yin ching (TSCC ed., pp. 235-6).
621 B
631 B
Divination, mythology and monarchy
Name of comet (B
signifies 'botanical')
Ch'ih kuan, red drops
Po kuan, white drops
Tien hsiao (shuo), the
flute of Heaven
Tien hsiao (shuo), the
flute of Heaven
Ch'ana, (magnolia?)
Sui hsing, broom star
Po kuan, white drops
Ch'ih kuan, red drops
P'u sui, reed broom
P'u sui hsing, reed broom
Kan sui, straw broom
Kan sui, straw broom
Chou sui, sweeper broom
Li sui, whetstone-broom
Chu sui, bamboo broom
Chu sui, bamboo broom
Hao sui, artemisia broom
Hao sui, artemisia broom
Chan sui, thatch broom
Chan sui, thatch broom
Chan (? sui), thatch
Shen hsing
(Ch'iang) hsing
(Na) hsing
Kan sui, shield broom
Chan sui hsing, thatch
broom star
Ch'ih-yu ch'i, the Banner
of Ch'ih-yu
Tid, pheasant
l l 49e'
l 129q
l 124a
Notes (M signifies a
military prognostication)
attributed to Pei-kung;M
attributed to Pei-kung
attributed to Yaoh
attributed to Pei-kung
see GSR 612h
M (for the reading suia
rather than hui, see GSR)
attributed to Pei-kung
attributed to Pei-kung; M
attributed to Pei-kung
attributed to Pei-kung
attributed to Pei-kung
M; reading doubtful
M. See GSR 658f, and
p. 74 below; WW 1978.2,
26 identifies as mulberry
pips; Morohashi 14082 as
mushroom (in the Po-wu
M;see nos. 622, 623
attributed to Pei-kung
M; see pp. 77f below
The Han view of comets 69
accompanied by comment or prognostication. Comets are however treated
more fully in the special chapters that concern the movements of the heavenly
bodies or the strange phenomena of the universe. In those chapters the reports
of comets' appearances are frequently followed by interpretation or com-
ments submitted by notable or distinguished men of letters. Examples of
reports are given in the appendix below.
The observations that are reported for the Former Han period are not
always dated precisely. Sometimes two or more references may pertain to
one and the same incident. In some cases it is not always possible to
determine whether observations which were separated by a short period
were in fact concerned with the same event; and it is possible that some
reports may have derived from purposeful fabrication that was undertaken
for political motives.
For these reasons it is difficult to count the number of
different appearances that were recorded, but it may he estimated that
possibly as many as thirty separate incidents featured and were observed
between 204 BC and AD 22.
One of the observations (for 12 BC) can be
identified with confidence as Halley's comet, and the sighting of 87 BC was
probably concerned with its immediately preceding occurrence. Two other
cases (for 135 BC and 44 BC) may perhaps be linked with observations of
comets recorded in other cultures, i.e., in one case for the comet said to have
attended the birth of Mithridates, and in the other for the one that appeared
close to the murder of Julius Caesar. A sighting of AD 13 may correspond to
a report mentioned by Dio Cassius; one for AD 185 has been identified as a
A variety of terms, which will be considered immediately below, were used
in the Standard Histories to denote these events. Whichever term is used, such
18 For the Former Han period, the Han shu has a fuller set of records than the Shih-chi. In his
notes to references to comets in the Imperial Annals, Dubs draws attention to the entries in
Williams (1871) and to observations recorded in western sources. In addition to Williams'
pioneer work, lists of comets recorded in Chinese sources for the period also appear in Ho Peng
Yoke (1962). For consideration of the possible motives and ways of fabricating reports of
strange phenomena, see Bielenstein (1950); HFHD vol. I, pp.165-6, 212, 287--8, and vol. III,
p. 555; Eberhard (1957); and Sivin (1969), pp. 52f. [For a complete list of observations of
astronomical ph,enomena of all types as reported in a variety of sources, see Chuang Wei-feng
(1988). This work includes the categories of sui hsing, from the eleventh century BC to 1901
(pp. 381-574); /iu hsing yii, from the seventeenth or sixteenth century BC to 1911(pp.575--616);
and liu hsingb from 651 BC to 1911 (pp. 617-1082).]
19 Williams lists forty-one incidents (nos. 15-55) and Ho lists forty-four incidents (nos. 23-66)
including suspect novae. [For the years between 204 BC and AD 23, Chuang Wei-feng (1988)
lists forty-two incidents under sui hsing (pp. 384--8); two under /iu hsing yii (p. 577); and fifteen
under liu hsing (pp. 619-20).]
For correspondences with records from other sources, see the references in H1'1JD vol. II,
pp. 34 (note4. l), 313 (note 5. 7) and 410 (note 13.6), and HFHD vol. III, p. 333 (note 22.2). For
the incident of AD 185, see Clark and Stephenson (1977), eh. 5. For the comet sighted in 87 BC,
see HSPC 7.lb (HFHD vol. II, p. 152) and Ho Peng Yoke (1962)), p.145. Other references to
these incidents will be found in HSPC 44.9a, b and HSPC 27C(2).22b (for 135 BC); HSPC
26.56b (for 44 BC); HSPC 27C(2).23a and HSPC 36.30a (for 12 BC). Other sightings which
have been tentatively identified as Halley's comet are for AD 66, 141, 218 (see Ho Peng Yoke
(1962), pp. 150, 152 and 154). See also Kiang (1972), and Y. C. Chang (1979).
70 Divination, mythology and monarchy
details as are given in the record are of the same type, i.e., they concern the
constellation in or near which the phenomenon was seen; its subsequent
movements; its colour, and its size. This last detail is given either by
measurement in feet, however that may be interpreted, or by comparison with
material objects. How far the Chinese were able tp distinguish at this stage
between comets, novae and supernovae must remain open to question; but the
basic identification of the phenomena as comets can be accepted, by reason of
the attendant details that are reported, the correspondence between some of
the Chinese reports with those from elsewhere and the evidence of the
manuscript from Ma-wang-tui.
Six expressions are used in the Shih-chi, lI an shu and lI ou Han shu to denote
comets. As it is by no means certain, and indeed unlikely, that the Chinese
authors of these works discriminated between the terms in direct correllation
to differences of observed phenomena, it would be anachronistic to do so in
order to satisfy the conclusions of modern astronomy. It would seem likely
that the compilers or authors of the different chapters of the histories chose
terms which suited contemporary usage, which may well have changed in the
course of centuries. More expressions appear in the manuscript from
Ma-wang-tui than in the Standard Histories, where some terms are used
interchangeably and with a lesser degree of refinement.
The terms used in the Shih-chi and Han shu are hsing po, ch'ang hsing p'eng
hsing, sui"-, k'o hsing and liu hsingb; from the considerations which follow it
may be concluded that the first five of these were used synonymously. Hsing po
differs from the four other terms which include the character hsing. For in
those expressions, hsingc is qualified by the modifiers eh 'ang, p'eng, k'o and liu.
In hsing po, where the order is reversed, the character po fulfils a verbal
function, and it has been suggested that the word designated the 'burst' in the
heavens, from which comets were believed to have emerged.
Liu hsing is
perhaps to be distinguished from the other terms, with which it is not
The indiscriminate use of some of these terms may be illustrated as follows:
(1) The Ch'un-ch'iu records occasions of hsing po for the years 613, 525 and
482 BC.
This information is repeated in the Shih-chi, where, however,
the term sui or sui hsing replaces hsing po. In the corresponding chapter of
the Han shu the same incidents are described, and the comments that are
cited thereto are couched in terms of both hsing po and sui.
(2) The Kung yang commentary to the Ch'un-ch'iu defines the term hsing po
as sui (for 613 BC).
21 For po, see GSR 49la, and Schafer (1977), p.107. For Han beliefs regarding the origin of
comets, see pp. 74f, 79f below.
22 SSC 19B.13b, 48. la and 59.6b There is a further reference in the supplementary part of the Tso
chuan for 516 BC, where the expression sui is used (see Harvard-Y enching Index text p. 422).
2' SC 27, p. 86 (MH vol. III, p.403); HSPC 27C(2).19b.20b.22a.
SSC 14.8b. .
The Han view of comets 71
Terms used to denote comet: examples of interchange
Date of Ch'un- yang Hou Han Hsli llan
incident ch'iu chuan Shih-chi Han shu shu chih
Annals Treatise Biography Annals Treatise
613 BC hsing po sui sui hsing po
155 hsing po sui
154 sui, ch'ang
147 SUi hsing po sui
135 hsing po ch'ang SUi
49 hsing po k'o hsing
AD 39 hsing po sui
(3) The Shih-chi does not apparently use the term po. In contexts wherein it
appears in the Han shu, the corresponding passage of the Shih-chi writes
(4) There are occasions when the Shih-chi uses the terms sui and eh 'ang hsing
in respect of the same incident.
( 5) Both the Han shu and the Hou Han shu use hsing po and sui in respect of
the same incident.
(6) The observation for 135 BC is recorded in three separate passages in the
Han shu as hsing po, sui and ch'ang hsing respectively.
(7) K' o hsing does not seem to appear in the Ch 'un-ch 'iu or the Shih-chi,
25 The Harvard-Yenching Index to the Shih-chi carries no reference to hsing po. The expression
hsingfu occurs in SC 12, p. 39 and 28, p. 78 (MHvol. III, p. 504) for an observation of 110 BC,
which is duly reported in HSPC 6.6b as hsing po. Similarly, Ssu-ma Ch'ien's comments to the
chapter on astrology include the statement that comets (hsingfu) appeared at the destruction of
Ch'ao-hsien in 108 BC, and when the Han armies attacked Ta Yuan (SC 27, p. 92; MHvol. lll,
p. 408); HSPC 26.53a, 54a reports both these incidents as hsing po. For the use of Ju in other
texts, see Yen tzu ch'un-ch'iu 7, p. 436 and Yang Hsiung, Ch'a Ch'in mei Hsin (Wen hsiian,
SPTK ed., 48.12b).
2 SC 27, p. 91 (MHvol. III, p.407) reports the incident of 154 BC assui; SC 11, p. 4 (MHvol. II,
p. 498) as ch'ang hsing. The event is not reported in HSPC 5.
27 For example, (a) HSPC 5.3b (HFHD vol. I, p. 312) and HSPC 26.50b report the same incident
in 155 BC, once as hsing po and once as sui; (b) an incident of the ninth month of 147 BC is
reported as hsingpo in HSPC 5.6b (HFHD vol. I, p. 321) and as suiinHSPC 26.5la and SC 11,
p. lO(MHvol. Il,p. 504); and(c)HHSCC lB.llareportsahsingpofor AD 39; inHHSCC(tr.)
10.7a this is mentioned as sui.
The observation for 135 BC is given as hsing po (HSPC 6.4a; HFHD vol. II, p. 34), as sui
(HSPC 44.9a, b) and as ch'ang hsing (HSPC 27C(2).22b). The last passage includes in its
prognostication an identification of the phenomenon as the Banner of Ch'ih-yu. The term sui
hsing (seen in entry no. 617 of the manuscript) also appears in one of the medical texts from
Ma-wang-tui, in a formula of exorcism designed to eliminate infantile convulsions (see WW
1975.9, p. 37, column 53 of the manuscript). The term is also used as a general expression in
Huai-nan-tzu 3.3b.
2 There are no references to the observation of a k'o hsing; the term appears in SC27, p. 96 (MH
vol. II, p. 412 'etoile etrangere').
72 Divination, mythology and monarchy
it is used in the Han shu interchangeably with hsing po,
or, sometimes.
to describe certain sightings uniquely.
(8) Perhaps the least common of all the terms is p'eng hsing, which does not
appear to feature in the Shih-chi. Its two appearances in the Han shu
probably refer to incidents that are decribed elsewhere as hsing po.n
While it may be accepted that k'o hsing is used in the Han shu to denote q,
comet, the evidence for the use of liu hsing is far less certain. This expression
does not appear in the Ch 'un-ch 'iu; in its three occurrences in the Shih-chi it 1
used as a means of general description rather than as a direct identification ofa
particular phenomenon that can be defined as a comet.
Of the seven
references in the Han shu, three
refer to an event which was observed in the
ninth month of 32 BC. It is possible that this is to be identified with the sighting
of the hsing po for the first month of that year, and the terms whereby the liU
hsing is described are comparable with those that are used to describe a hsing
po. Two of the other references could possibly concern comets;
one other is. '
and in one context, where the liu hsing is described as being the : 2\
size of the moon and accompanied by a number of other heavenly bodies, flie:'
term has been translated as 'meteor' and explained as a fireball.
The bibliographical list which is incorporated in the Han shu includes three
items whose titles mention the expressions that are under study.
In one, k'q
[hsing] and liu [hsing] are both specified; another mentions sui and k'o [hsing]r
andin the third /iuhsingareisolated. The titlesofthefirsttwo of these works alsq
include references to the five planets, and it is to be noted that the treatise on
astrology in the Hsu Han chih lists planetary movements together ..
(a) HSPC 26.54a reports hsing po at the start of the T'ai ch'u period (104-101 BC); the textcifos
an appropriate comment from the lost ( hsing) chuan on the subject, not of hsing po but of k]J"
hsing. (b) HSPC 26.56a reports a k'o hsing for the third month of 49 BC; in HSPC 8.24b,
(HFHD vol. II, p. 263) this is given as hsing po. (c) For the sixth month of69 BC, HSPC26.56a :
writes k'o hsing, while HSPC 8.7a (HFHD vol. II, p. 215) and HSPC 27C(2).23a have hsirigpq ..
for the first month. The implications are not clear in a passage of HHSCC (tr.) 12.2a whic.h .
relates that a k'o hsing was transformed into a sui.
31 See HSPC 26.52b, 54b and 56b for incidents in 134, 77, 48 and 47 BC.
See HSPC 26.Slb, 54a and HSPC 5.6a (HFHD vol. I, p. 320) and HSPC 7.3b (HFHD vol:II; ;
p. 157); the months mentioned in these corresponding passages for the years 148 and 84 BG
not identical.
SC 24, p. 6 (MH vol. III, p. 236) reports the regular presence of liu hsing above the .
an emperor was worshipping T'ai i (Grand Unity; [in origin probably a divinity associated With
one of the constellations]); in SC 27, (MHvol. III, p. 392) and SC28, p. 10 (MHvol. III, p.422}
the term is used to describe phenomena such as lights or effulgences due to occult powers.
34 HSPC 10.3a (HFHD vol. II, p. 378); HSPC 26.56b; and HSPC 97B.4b.
HSPC 26.55a and 59a for 73 (second month) and 8 BC (first month).
36 HSPC 26.54b.
HSPC 7.lOa (HFHD vol. II, p.174), for the second month of 74 BC.
38 HSPC 30.65b, 66a. The titles are: (a) Chin tu yii-heng Han wu-hsing k'o liu ch'u-ju; (b) Han
wu-hsing sui k'o hsing-shih chan-yen; and (c) Han liu-hsing hsing-shih chan-yen. In a T'ang text
which concerns military matters (Shen chi chih ti T'ai-po yin ching, ascribed to Li Ch'u:;u,1,
preface dated 768) three chapters (nos. 7-9) concern various methods of divination for tactical
purposes. One of these includes sections on liu hsing, k'o hsing andyao hsing, but not onsui ot_,,
hsing po.
The Han view of comets
nomena that are denoted as sui, k 'o hsing or liuhsing. In general the inference
these titles is inconclusive; it could possibly be argued that they support
suggestion that for the Former Han period Chinese writers regarded k'o
gas being in the same category as hsing po, eh 'ang hsing, p'eng hsing and sui,
ileretaining liu hsing for something that may have been of a different nature .
. The foregoing considerations tend to show that the writers and observers of
ormer Han were not bound by the distinctions suggested by Wen Ying, who
as writing towards the beginning of the third century AD. In commenting on
"rt Han shu 's use of the expression eh 'ang hsing
for an observation of 172
:, Wen Ying discriminated between hsing po, sui and ch'ang hsing as being
shy', 'broomlike' and 'long', according to the shape of the light-rays of the
cts observed; but, as has been seen, it would be difficult to sustain this
nction in the references that are made to comets for the Former Han
or observations of the Later Han period, we find that p'eng hsing is
arently not used, and there is only one reference to ch'ang hsing, which is
in the treatise on astrology.
The chapters of imperial annals of the Hou
shu use the term hsing po, but not sui; and there are twelve occasions of
nts reported between AD 39 and 188 which the treatise describe as sui and
imperial annals as hsing po.
The treatise uses the term hsing po for an
ervation of AD 22, and for those that are dated between 193 and 218; for
se final years it does not use sui. There are three cases wherein both the
erial annals and the treatise describe the same event as k'o hsing:42 and
ilty-seven reports of liu hsing in the treatise are not mentioned in the
erial annals. In two passages the biographies of the Hou Han shu include
comments submitted by statesmen on the appearance of comets; they use
. somewhat loose expression sui pei.
HSPC 4.13b (HFHD vol.I, p. 251-2).
HHSCC (tr.) 11.lb for AD 65.
HHSCC Treatise Date
I0.7a 39
I I.la 60
l l.2b 75
l l.2b 76
11.3a 77
l l.7b 109
ll.12a 141
12.la 149
12.4a 178
12.4a 180
12.4b 182
12.5a 188
.. l!HSCC 3.14a and HHSCC (tr.) l l.3b for 85; HHSCC 6.6a and HHSCC (tr.) 11.IOb for 131;
l;lnd HHSCC 6.7b and HHSCC (tr.) 11.1 la for 132.
,(i) _HHSCC. 64.12b, in a memorial of Lu Chih dated 178; and (ii) HHSCC 66.12a, where
Shih-sun Jm argues in 192 that the appearance of sui pei indicates the need for speedy action.
74 Divination, mythology and monarchy
Professor Schafer suggests
that a basic and early distinction between
tailless comets (aphelial) and tailed comets (perihelia!) was reflected in the
terms po and sui; how far this distinction can be verified in Han usage may
perhaps remain open to question. At a much later stage of Chinese intellectual
history, Ma Tuan-lin (1254-1325) arranged his chapters on astronomical
phenomena in such a way that hsing po, sui hsing and ch'ang hsing are taken
together, while liu hsing and k'o hsing are treated in a separate section.
In the
manuscript from Ma-wang-tui, the term sui features not only in no. 617 (in the
expression sui hsing), but also in a number of other entries, where it is modified
by words that denote plants or shrubs (for example, entries nos. 620---31, 633,
637 and 638). It may be suggested that originally the term sui, with its vivid
imagery, was used to denote a particular and recognisable type of comet; that
it was subsequently modified, to allow for finer distinctions; and that it finally
came to be used as an expression for comets in general.
A number of the terms which designate comets on the manuscript (and, as
will be seen, in literary passages) are botanical. This need occasion no surprise,
in view of the obvious comparison between the diagrams of the manuscript
and the shapes of the plants. Of the terms which have been encountered so far
in literary passages.Ju is interpreted as meaning 'bushy', and p 'eng as the name
of a plant.
The standard commentary to the Hsu Han chih, of Liu Chao
(fl.510), carries an interesting citation that is ascribed to Han Yang, to the
effect that 'the shapes of comets are like those of bamboo brooms, or the
branches of trees, and there is no regular constancy in their size. Long ones last
a long time and their damage is severe; short ones last a short time and their
damage is more limited. '
It will be seen above that a number of the terms of
the manuscript have been identified with particular plants, i.e., ch'an as
?magnolia, p 'u as rush, reed or willow, kan as straw, haoc as artemisia, chan as
thatch and chena (or shen) as mushroom. Similarly p'ou which, as will be seen
immediately, appears in literary passages, has been identified as mistletoe.
The treatises on astrology in the Shih-chi and Han shu name several objects
which came into being as a result of the aberrations of some of the planets.
Thus, from the strange behaviour of Jupiter there emerge T'ien-p'ou,
44 Schafer (1977), p. 107.
' Liu hsing and k'o hsing are treated in chapter 281 of the Wen-hsien t'ung-k'ao together with
other phenomena such as jui huang and yao hsing. Chapter 286, which is sub-titled po sui, is
concerned only with incidents that are described by one of these terms.
See GSR 500k *p'imt/p'iu;Jt/fu shrubby, dense (Kuoyii); pimt/piu;Jt//fu clear away dense
vegetation (Shih); and GSR l 197y b'ung/b;Jung p'eng name of a plant of uncertain species
(Artemisia? Chrysanthemum coronarium? Conyza?) (Shih); luxuriant foliage (Shih).
HHSCC (tr.) 10.4a, b. Han Yang is untraced.
48 For these identifications see p. 68 above. For p'ou, see GSR 999x p';Jg/p';Ju/p'ou and
b';,g/b';Ju/p'ou raised wooden platform, a look-out built of planks (Kungyang).
49 SC 27, pp. 39, 57 and 59 (MHvol. III, pp. 362f, 378 and 380); HSPC 26.19b, 20b, 23a and 25b.
These passages may be compared with parts of the text of the other manuscript from
Ma-wang-tui which concerns astronomy (see note 9 above); for transcription of that text, see
Chung-kuo t'ien-wen-hsiieh shih wen chi, 'Ma-wang-tui Han po-shu "Wu hsing chan" shih wen',
pp. 2, 3 and 7 for columns 10, 11, 13, 17, 19 and 55 of the manuscript.
The Han view of comets
sui-hsing, T'ien-ch'an and T'ien-chiang; sui-hsing likewise emerge from the
unaccountable behaviour of both Venus and Mercury. The manuscript from
Ma-wang-tui confirms the conclusions reached by commentators that some of
these objects are to be identified as comets. Thus, T'ien-ch'an can be identified
with entry number 616, which is entitled ch'an; and when the Han shurecords,
in a totally different passage,
that T'ien-ch'an appeared in 162 BC, we may
suspect that the term denoted Halley's comet. In addition there are several
references to Ch'an-chiang which apparently describes a single phenomenon.
The expression is seen in the 'Ta jen fu' of Ssu-ma Hsiang-ju (c. 179-117 BC),
where the seeker is envisaged as 'brandishing the comet eh 'an-chiang as a
It is also seen in the 'Kan-ch'iian fu' of Yang Hsiung (53 BC-AD
18), where it denotes one of the heavenly bodies. 52 In addition the term
denotes a comet in the Hou Han shu;
and it duly appears in the Shuo yiian of
Liu Hsiang (79-8 BC) as one of the objects produced by the aberration of the
Elsewhere the expression T'ien-p'ou is used in an entirely different
sense, to denote an area of the heavens. 55
The treatment of comets in the Standard Histories56
When the chapters of the Shih-chi and the Han shu report the appearance of a
comet, they note its position in the heavens and the direction in which it was
moving; the length of time for which it was visible; its colour and its size, either
in the general terms of 'extending over all or half the heavens', or, more
precisely in terms of feet; such terms of measurement have yet to be fully
explained. The treatises of the Han shu which concern astrology and strange
phenomena include statements that are ascribed to famous figures such as
Tung Chung-shu (c. 179 to c. 104 BC) Liu Hsiang, Liu Hsin (46 BC to AD 23)
or Ku Yung (fl. c. 9 BC), as comments made on the appearance of a comet.
Curiously enough no comments are included from Ching Fang in this
connection. Elsewhere in the Han shu his views are frequently quoted in
relation to other strange occurrences; and citations from his works in the Chin
shu show that he had certainly not excluded comets from his investigation of
natural phenomena.
Sometimes the Han shu simply relates a subsequent
HSPC 26.49b; see Ho Peng Yoke (1962), p. 143.
" SC 117, p.81, Hervouet (1972), p.187; see GSR 703i. For the problems regarding the
composition of this Ju, see Hervouet (1964), pp. 288f.
" HSPC 87A.14a; Knechtges (1976), p.48.
" HHSCC 52.4a.
For the reference in the Shuo-yiian, and the question of the emergence of comets from planets,
see note 58 and p. 79 below.
" HHSCC (tr.) l l.3b.
For examples of passages, see the appendix.
Ching Fang's explanations of strange phenomena are cited in HSPC 27B(l ), HSPC 27B(2),
HSPC 27C(l) and HSPC 27C(2), in connection with incidents such as the odd behaviour of
rats, or unaccounted sprouting ofleaves in dead wood. For his remarks on comets, see Chin shu
chi?o chu 12.12b. [For the distinction between Ching Fang the Elder (?c. 140 to c. 80 BC) and
Ching Fang the Younger (79-37 BC), see A.F.P. Hulsewe, 'The two early Han I ching
76 Divination, mythology and monarchy
historical event, leaving it to the reader to draw the obvious inference that the
comet's appearance foreshadowed the incident in question, which was usually
of a dynastic or political nature.
In addition the Han shu includes on at least
three occasions a general statement to the effect that 'comets eliminate the old
and inaugurate a new order'.
There are also a few tantalising citations from
works that are now lost, such as the Hsing chuan, or the opinions of Shen Hsu
[or Ju]. The question may naturally be raised whether such writings included
material of the same form as that of the manuscript under study.
The treatise on astrology in the Hsii Han chih reports planetary movements
and observations of phenomena such as the appearance of comets. The text
notes the size of the comet, sometimes in graphic terms (for example, as large
as a peach, a melon or an egg), or sometimes in terms of the measurement of
the length or breadth of the tail (in feet). Occasionally it is stated that the
comet's appearance was accompanied by a noise like that of thunder, and the
comet is sometimes described in terms of colour (i.e., white, blue and yellow,
red, green and white, red and yellow, blue and white, or yellow white).
Sometimes it is said that a comet broke into several fragments. The text of the
treatise states in which part of the heavens the comet was sighted and reports
its movements; it relates the area of the heavens in question to the correspond-
ing and appropriate region on earth. Such a relationship is sometimes
restricted to the terms of a prognostication that is given, usually from
unnamed sources, or to the subsequent verification in terms of political and
dynastic change.
The treatise includes the following general explanatory statement, follow-
ing its report of the appearance of a comet in the eleventh month of AD 22:
Disrupting stars [po hsing] are the product of evil exhalations and give rise to disorder
and violence, by which they disrupt [po] natural qualities of good [te]; a disruption of
specialists called Ching Fang', TP 72 (1986), 161-2. See also note 80 below.]
58 For example, see HSPC 27B(2).22b, 23a; after reporting the comet of 110 BC, the text adds
'Thereafter Chiang Ch'ung staged his revolt and the capital city was thrown into confusion'
(for this incident, of 91-90 BC, see CC, eh. 2). See also Shuo-yiian 18.36 for credence in the
connection between a comet and an historical event such as the accession of the First Ch'in
HSPC 26.59b and HSPC 27C(2).20b, 2la. In Wen Ying's (ft. 196-220) note to HSPC 4.13b
(HFHD vol. I, p. 251), this statement is ascribed to Ta fa.
For citations from the Hsing chuan, see HSPC 27C(2).20a, b. No work with this single title
appears in the large number of books on astronomy and astrology that are listed in HSPC
30.65aff. Shen Hsii, or Ju, was a prominent man in Lu, who features in incidents that are
recorded in the Tvo chuan for 706, 694 and 680 BC (SSC 6.22b, 7.25b and 9.86); these do not
correspond with the reference in HSPC 27C(2).2la. The incident recorded in the Tso chuan for
694 BC is also reported in the Kuan-tzu ('Ta k'uang') 18.2a, where Shen Hsii appears as Shen
Yii (Rickett (1965), p. 48 and (1985), p. 287). It is also recorded in one of the historical
documents found in Ma-wang-tui which is entitled now Ch 'un-ch 'iu shih-yii, but in this account
neither Shen Hsii nor Shen Yii are mentioned; see WW 1977.1, 35 (incident no. 16).
" HHSCC (tr.) 10.4a; for the observation, see HSPC 99C.19b (HFHD vol. III, p. 435), HHSCC
1A.2b; and Williams (1871) no. 55. The bracketed passage reads: ts'anjan pei yen ping chih lei
The Han view of comets
such qualities is a sign of violence, a manifestation of darkness [pu ming]. In addition
[eight characters not understood] ... hence they are termed 'disrupters', an expression
which implies that something has been injured and that something has been
obstructed. Sometimes they are called 'broom-stars' [sui hsing], the means of
eliminating corruption and inaugurating a new order.
Liu Chao's comment to this passage includes two interesting citations, of
which one has been reported above.
The other is a note written by Sung
Chun to a document entitled Kou ming chiieh;
he writes of the existence of
five types of 'broom', distinguished according to the five symbolical colours of
the wu hsing, i.e., green, red, yellow, white and black.
The Banner of Ch'ih-yu
The diagram that forms part of entry number 639 of the manuscript is
identified there as 'the Banner ofCh'ih-yu', and the prognostication that was
appropriate to this particular type of comet reads: 'armies are without; they
will return'. There is further information about this comet in literary sources.
Ch'ih-yu himselffeatures as a hero, or a villain, of Chinese mythology who
was involved in a number of escapades. Sometimes he is cited as a byword for
the outbreak of violent conflict; sometimes he is described as one of the
ministers who served Huang-ti, the Yellow Emperor; at a later stage he
became accepted as the God of War. 64
The term 'Banner of Ch'ih-yu' is used to denote certain types of exhalation
(ch'i) as well as a comet of a particular definable type. In the Lii shih
eh 'un-ch 'iu
the term appears among a number of others that denote clouds or
exhalations; but although the same passage refers to comets, the Banner of
Ch'ih-yu does not feature in that connection. The T'ai-p-ing yii-lan66 cites a
passage from a work which is entitled Huang Ian; this recounts some of the
consequences that followed the death ofCh'ih-yu at the hands of Huang ti and
his subsequent burial. A form of popular worship grew up at his tomb, taking
place in the tenth month of the year. In addition a red emanation which
See note 47 above.
Re-collected fragments of the Hsiao-ching wei kou ming chiieh, with annotation by Sung Chiin,
are included in the Yii-han-shan fang chi-i-shu eh. 58; for the full citation from which this
passage is drawn, see f.8a. In that collection Sung Chiin is described as being of the Wei period,
presumably third century AD. He is to be distinguished from the Sung Chiin of HHSCC 2.19a,
whose pupils are stated to have offered advice regularly to Ming-ti (reigned 57-75), after
inspecting comets. See also HHSCC 41.13b for a biography of Sung Chun, who died in AD 76.
There is nothing in the biography to confirm the statement of HHSCC 2. l 9a, or to show that its
subject possessed skills or specialised knowledge of comets. See also Wang Hsien-ch'ien's note
to HHSCC 41.13b for the view that Sung Chiin is an error for Tsung Chiin.
See SC 1, pp. 6, 8 (MH vol.I, pp. 27, 29) and SC 27, p. 60 (MH vol. III, p. 107), and Bodde
(1975), p.120. A shrine to Ch'ih-yu was included among those places of worship which were
abolished in 31 BC (HSPC25B.15a). In Yen-t'ieh lun 52, p. 328, Ch'in is criticised for believing
itself to be impervious to damage, even by Ch'ih-yu. [For Ch'ih-yu's place in mythology, see
chapter 11 below.]
LSCC 6.9b.
TPYL 875.!0a; see also the note in SC 1, p. 8 by P'ei Yin.
Divination, mythology and monarchy
appeared at the tomb and resembled a bolt of deep red silk was popularly
called the 'Banner of Ch'ih-yu'. Elsewhere, in an incident which is reported for
AD 254 an emanation that was white was identified by Wang Su (195-256) as
the he added the prediction that its appearance meant trouble in the
At least three passages refer to the appearance of a comet in 135 BC. The
shortest of these, in the imperial annals of the Han shu,
reports that in the
eighth month 'a comet appeared [hsing po] in the eastern quarter; it was long,
extending throughout the entire sky'. In the chapter on the wu hsing,
information is somewhat fuller; there was a comet (hsing po) in the northern
quarter in the sixth month, and another (eh' ang hsing) in the eastern quarter in
the eighth month. The text includes a prognostication that was given for the
second appearance; this identified the comet as the 'Banner of Ch'ih-yu',
which would be followed by royal military expeditions in all directions. The
third reference is by no means as precise as those that have been mentioned,
but it may well concern the same incident. The Shih-chi
alludes to two
observations of Ch'ih-yu's Banner during the Yiian-kuang and Yiian-shou
periods (i.e., 134-129 and 122-117 BC). The text adds that the length of the
comet was such that it extended for half the sky.
The next appearance of a comet which was identified as the Banner is dated
for AD 191, and again there are two separate references. The short entry in
imperial annals71 simply states that in the ninth month the comet appeared m
the lunar mansions Chiaob and K'ang. The treatise, however, is more
informative; 72 the comet is dated to the ninth month; it is described as being
over ten foot long, and white, emerging from the south of Chiao and K'ang.
The chapter includes a similar prognostication to the one seen in the treatise of
the Han shu, i.e., that its appearance would be followed by military activity in
all quarters. Finally, the Hou Han shu carries a passage in which a Liu hsing was
described by Su Ching (ft. AD 10-20) as being like Ch'ih-yu's Banner.
For attempts to describe the phenomenon we must turn in the first instance
to an earlier text. According to the passage of the Lii shih ch'un-ch'iu that has
been cited above, 74 the Banner was yellow above and white below. Here the
text concerns the appearance of an emanation rath,er than a comet. Neverthe-
less it was cited as an authority by Chin Cho (ft.c. AD 208), in a comment to
the Han shu. 15 The text on which the comment is made carries part of a
prognostication with which we are already familiar; 'the Banner ofCh'ih-yu,
while being like a comet [sui1, is curled at the rear, in the shape of a flag; when it
is seen, those who are kings will undertake military expeditions in all
67 San kuo chih 13 (Wei), p. 418. For a further observation of the banner ofCh'ih-yu, in AD 501,
see Wen-hsien t'ung-k'ao 286, p. 2269B.
HSPC 6.4a (HFHD vol. II, p. 34).
69 HSPC 27C(2).22b. 10 SC 27, p. 91 (MH vol. III, p. 408).
HHSCC 9.3a.
12 HHSCC (tr.).12.5b. ,. . .
13 HHSCC 30A.3b. The object was compared variously with the banner of Ch 1h-yu, With Ymg
t'ou or T'ien chiang (for T'ien chiang, see p. 75).
LSCC 6.9b.
" HSPC 26.37b; the passage occurs also in SC 27, p. 72 (MH vol. III, p. 392).
The Han view of comets 79
directions'. The same passage recurs in the treatise on astrology in the Chin
shu, with some further elaboration regarding the shape and colour of the
Chinese views of the origin of comets
Reference has been made above to the statements of the Shih-chi and Han shu
to the effect that certain named comets were produced thanks to the
aberrations of the planets.
A similar concept is voiced by Liu Hsiang (79-8
BC), whose interpretations of comets and their significance in dynastic terms
are included in the Han shu. In a separate work, the Shuo yiian, he describes
some of the features of the heavens in connection with prognostication. He
names five comets which he regards as being the product of the waxing and the
waning of the five planets, but he does not relate individual comets to
particular planets. We have already encountered two of the comets which he
names, i.e., the Banner of Ch'ih-yu and Ch'an-chiang.
The treatise on astrology of the Chin shu
carries a long citation that is
ascribed to Ching Fang, a famous contemporary of Liu Hsiang who has
good cause to be named among the foremost of Han scientific observers.
The passage refers to a number of comets by name, eight of which appear in
the manuscript from Ma-wang-tui.
Ching Fang attributes their origins to
the planets, and lists the Banner of Ch'ih-yu among the products of Mars.
The same chapter of the Chin shu also carries a citation from a source which
is named as Ho t'u.
Here the comets are said to originate from the
dissipation (san) of the planets; it is also suggested that the essence (ching) of
the planets may become comets; and the Banner of Ch'ih-yu is variously
ascribed to Mars and Saturn. In a comment to the Han shu, Meng K'ang
(ft.c. 180-260) likewise identified the Banner of Ch'ih-yu as the essence of
The same concept is repeated in two short citations in the T'ai-p'ing
yii-lan, from a work which is entitled Ho t'u ch'i yao kou.
According to one
of these passages the five planets produce the five comets by a process of
dissipation; the other passage states that the Banner of Ch'ih-yu originated in
this way from Mars.
76 Chin shu chiao chu 12.9b; Ho Peng Yoke (1966), p.130.
11 See p.000 above, and SCC vol.111, pp.430f. 78 Shuo yiian 18.26.
" Chin shu chiao chu 12.12b; Ho Peng Yoke (1966), p.134.
0 Of the two scholars named Ching Fang who were each specialists in the interpretation of the
Chou i, the first had probably died before the reign ofHsiian-ti (73-49 BC). The second (79-37
BC), whose sayings are cited in the Chin shu, was the founder of one of the four chin wen schools
of the I ching. Part of his intellectual effort was directed to reconciling the cycle of the sixty-four
hexagrams with Yin-Yang and the Five Phases' cycle of change in nature. For his biography,
see HSPC 75.5b. The most reliable fragments of his writings are probably to be found in
citations in works such as the Han shu or Chin shu.
81 I.e., in entries nos. 613, 616, 617, 626, 628, 635; 639 and 640.
12 Chin shu chiao chu 12.12a.
83 See Meng K'ang's comment to HSPC 26.37b and SC 27, p. 72 (MH vol. III, p. 392).
84 TPYL 875.4a and lOa.
80 Divination, mythology and monarchy
The value and purpose of the manuscript
The manuscript that is under study reflects both the progress of scientific
observation and the Chinese concern with divination. Its implications have a
bearing on three topics: the importance of linear patterns in divination; the
Chinese attitude towards the ordinary and the extraordinary movements in
the world of nature; and the form of the earliest parts of the Book of Changes.
Finally it may be asked what the motives were for the compilation of the
Comets are classified in the manuscript according to their linear shapes, and
the prognostications follow suitably. An analogy is immediately suggested
with the importance of the lines or patterns that were induced on turtles' shells
or animals' bones, or the creation of a pattern of six lines by the cast of yarrow
stalks. However, there is one difference between such patterns and those of the
manuscript's diagrams. For whereas in the first two cases the patterns emerge
only after purposeful human manipulation, the shapes of comet are provided
by nature, for all to gaze at and admire. Other cases wherein Chinese
divination takes account of natural rather than artificial lines may be seen in
feng-shui, where the inherent patterns of the earth are regarded as the
conductors of good or evil influences.
Divination in China fastened on the features and phenomena of nature in
two ways. Seers discerned a message in the normal patterns of regular
occurrence, such as the direction of the winds or the shapes of clouds; they also
saw abnormal, irregular events as the harbingers of the future. In both cases
they relied on a philosophy that saw the universe as a whole, whose integral
parts were subject to the same overriding rhythms. Linear patterns could show
how those rhythms were working out in their normal, regular and expected
course; they could also show the disturbances that had been wrought in the
operation of those rhythms. Extreme examples may be quoted in two
instances. The practice of feng-shui fastens on the permanent, static and
regular features of a landscape; the attention paid to an eclipse or the
extra-orbital movements of the heavenly bodies seeks to interpret violent and
abnormal incidents. It must, however, remain open to question how far the
Chinese consciously drew a distinction between these two types of inference. It
is of some interest that the single document from Ma-wang-tui includes entries
for both types of phenomenon, for example, the shapes of clouds, that are of
regular occurrence, and the patterns of rarities such as comets.
The text that accompanies the entries for the comets includes the name
allocated to each pattern and the prognostication thought appropriate
thereto. Comparison is immediately suggested with parts of the Chou i. In
both cases the names of the linear patterns include material objects, as if they
were chosen as mnemonics. Whereas in the Chou i the prognostications of the
t'uan and the yaoc are couched in formulaic terms and give general indications,
in the manuscript from Ma-wang-tui, which is dated several centuries later,
The Han view of comets 81
the predictions are more specific. It may be asked whether, in their earliest
forms, some of the writings that we now know as the Chou i were not of a
similar layout to that of the manuscript, i.e., a series of linear patterns,
accompanied by the name whereby each one could be identified and
remembered, and a suggested prognostication.
These considerations prompt the question of the prime motive for compil-
ing the manuscript, and perhaps that of recording part of the Chou i in writing.
It can hardly have derived from a seer's own needs; for a master whose
pronouncements depend on intuitive processes would hardly require a record
to stir his memory, or for its own sake. But possibly a less skilled disciple,
whose claims to mantic powers were pretentious rather than genuine, would
be anxious to possess precisely such a document as the one under study; for it
could be used both as an authority from which his own statements could draw
support and as an aide-memoire in times of difficulty. The inclusion on the
manuscript of the names of certain masters as the sources may be relevant in
this respect. The document may have emerged at a stage when divination in its
real form, that depended on unconscious powers of vision, had given way to
the next stage, when intellectual powers were being invoked, memories were
being consulted or some form of instruction sought. T'ien-wen ch'i-hsiang tsa
chan remains in the middle ground between science and divination, allowing
scope both for the fruits of systematic observation and the intuitive messages
of a seer.
The following passages from the Standard Histories may serve as examples of
reports made in those works for the observation of comets or similar
(i) HSPC 10.13b (HFHD vol.II,p.410)
In the autumn, seventh month [i.e., August to September, 12 BC] there was a
comet [ hsing po j in the constellation Tung-ching.
This incident has been identified with the appearance of Halley's comet. It
followed a solar eclipse which was reported for the first month, and
unexplained claps of thunder and flashes oflight which appeared out of a clear
sky in the fourth month. The events prompted the promulgation of an
imperial edict, in which the emperor expressed his concern and anxiety at the
state of the universe, and asked his principal counsellors to tender their advice
regarding the prevailing situation.
The constellation Tung-ching was one of the twenty-eight mansions (for
identification, see SCC, vol. III, p. 237).
For the diversity of material that came to be included in the Chou i and its different origins, see
Waley (1933), [and the entry for I ching by E. L. Shaughnessy, in Loewe (1994)].
82 Divination, mythology and monarchy
The incident is also reported in HS 27C (2).23a, for a date corresponding
with 26 August. The passage is translated by Dubs (HFHD vol. II, p. 410 note
13.6) as follows: .
'In Yuan-yen, I, vii, on hsin-wei, [Aug. 26], a comet appeared m [the
constellation] Tung-shing [, v, y, C 36, e Gem], and marched over the
Five Nobles [8, z, -r, u, K Gem]. It rose north of the [two] [the same as
the Nan and Po-ho, p, a, f3 Gem and e, {3, a C Min] and directed itself towards
and traversed Hsien-yuan [35 Lyn; 10 U Ma; 38, a Lyn; 59, z, Cnc; A, e, , ,, y,
17, a, o, 31 Leo] and T'ai-wei [6, 8, 1, a Leo; (3, 17, y, b, e Vir; a It .daily
progressed six degrees [of equatorial longitude] or more._ At 1t rose the
eastern quarter. On the thirteenth day [Sept. 7], at evenmg, it appeared m the
western quarter. It invaded the Second Consort[' Ser (cf. SC27: .14 the
Harem [another name for Wei3 (c, , C 17, fJ, 1, K, A., v Ser) and Chi (y, i5, e, 11 Sgr;
cf. HS 26: 8b, 9a)] the Bushel [C r, rr, <p, ll, Sgr] and Saturn. [On 7 Sept.,
Saturn was in R. A. 283.5.]
'The point of its flame [its tail] twice Tzu-kung
circle of stars about the north polar reg10ns: 6, ll, Ora; Ptazz110 126, 27 U Ma,
Piazzi 7h 187 48 H Cep; 19 H Cam; ex, z, 17, C (), e Ora; (3, y Cep]. Its great fire
[head?] later ;eached to the Milky Way and swept [away evils] in the of
the Consorts and Empress [U Mi?], went south, moved on, and mvaded
Ta-chio [Arcturus], the [two] She-t'i[17, -r, v and o, n, 'Boo], and to the
Heavenly Market-place [four stars of the six in Ch'i (cf. HS 26: 7b), 1.e., v,
43, n, o, v Sgr], where it stopped for a lunation, travelling slowly. Its flame [tail]
entered into the Market-place for ten days and later went west and left. On the
fifty-sixth day [Oct. 20], it hid itself together with the Azure
as the Eastern Palace (cf. 26: 7a), one-quarter of the zodiac, mcludmg Ch10,
K'ang, Ti, Fang, Hsin, Wei
, and Chi (Vir, Lib, Ser, and part of Sgr)].'
The text of the Treatise continues:
Ku Yung's answer [to the edict] was as follows: "Since the remote past,
occasions of extreme disturbance have occurred only rarely. When we observe
the fast rate [of the comet's-] speed, the varying length of its the
infringments that it perpetrated in its course, it is apparent that 1t s1?mfies,
within the violence wrought by the womenfolk of .the palace, and, without,
the di;astrous outbreak of rebellions throughout China."
Liu Hsiang, for his part, said: "At the ruin of the Three Dynasties the [two]
She-t'i changed places, and at the destruction of Ch'in and Hsiang [Yu] there
were comets in Ta-chiao."
It was in this year that the Chao-i consort Chao treated the tw? sons
emperor with violence. Five years later, at the death of Ch'eng-t1, the Chao-1
consort took her own life; with the accession of Ai-ti all members of the Chao
family were deprived of their offices and orders of rank, and banished to
Liao-hsi. In the absence of a direct heir to Ai-ti, P'ing-ti acceded to the throne,
and Wang Mang controlled affairs of state; he had the Chao empress.
Ch'eng-ti retrospectively demoted, and both she and the Fua empress of A1-tl
The Han view of comets
committed suicide. All members of the Ting and Fu families who had been
allied by marriage to the imperial house were deprived of their offices and
orders of rank and banished to Ho-p'u, or sent back to the commanderies of
their origin. When P'ing-ti died without a direct heir, [Wang] Mang forthwith
assumed unlawful control of the state.
(ii) Hsu Han shu(tr.) 10.6a
On the day chi-hai of the twelfth month [of the IOth year of Chien-wu, i.e., 25
January AD 35] a large roaming star [liu hsingj such as a broadly shaped jar
emerged from the south-west of the constellation Liub86 and proceeded to
enter Chen b. At the time when it was about to be destroyed, it split into ten or
more fragments, with the appearance of embers; shortly there were sounds as
loud as a clap of thunder.
Liu corresponds with Chou, and Chen with Ch'in and Shu. That a large
roaming star emerged from Liu and entered Chen corresponds with the
penetration of a large expedition from Chou into Shu.
At this time Kuang-Wu ti had sent Marshal Wu Han to mobilise 300 OOO
conscript troops from Nan-yang, they were to be embarked on boats,
orders to proceed upstream along the [Yangste] River, to take the offensive
against Kung-sun Shu, Emperor of the White, of Shu. In addition he had
ordered Generals Ma Wu, Liu Shang, Kuo Pa, Ts'en P'eng and Feng Chun to
bring order to bear in Wu tu and Pa commandery. In the tenth month of the
twelfth year, advanced troops of Han attacked Yung, Superintendent of the
Guards and cousin of[Kung-sun] Shu; they advanced as far as Kuang-tu and
put to death Shih Hsing, son-in-law of [Kung-sun] Shu. Feng Chun, the
Wei-Ju general, took over Chiang chou and put to death [Kung-sun] Shu's
general T'ien Jung. In addition, Wu Han attacked [Kung-sun] Shu's Marshal
Hsieh Feng, putting over 5,000 men to death; Ts'ang Kung conquered Fu,
killing [Kung-sun] Shu's brother Hui, who was Ta ssu-k'ung.
On the day ting-eh 'ou of the eleventh month [23 December AD 36] Kao Wu,
General of the Han Reserve Army, stabbed [Kung-sun] Shu, piercing his
breast. He died that night, and the next day Han [troops] entered the city of
Shu; they butchered its inhabitants, putting to death [Kung-sun] Shu's
generals of prime rank Kung-sun Huang, Yen Ts'en and others. Myriads of
individuals were killed and over 10,000 members of the families related to
Kung-sun Shu by marriage were exterminated.
This was the response [to the comet], in the form of attack and slaughter by
the generals of prime rank. The shooting movements of the smaller stars and
the fragmentation of the comet into ten pieces or more, like embers, were signs
that the generals of secondary rank would follow and join in the fight. The
sounds like thunder were indications of armed conflict.
For the identification of the constellations Liu and Chen, see S'CC, vol. III, p. 237.
84 Divination, mythology and monarchy
(iii) Hsu Han chih (tr.) 11.la
On the day ting-mao, in the sixth month of the third year 9
August AD 60], a comet [sui hsing] from the 1en-ch, uan for
a length of two feet, moving gradua_lly m a northe:ly direction to the south of
K'ang; after being observed for thirty-five days 1t departed.
T'ien-ch'uan signifies water, and that the comet emerged therefr?m
indicated a great flood. In that year the rivers Iand Lo
the Chin-ch'eng gate, destroying the bridge over the I River and mundatmg
thirty-two prefectures in seven commanderies. . .
(The translation follows the text as emended in the ed1t1on of
Chung huashu-chii, Peking 1965, p. 3229, as suggested ?Y Chien
constellation Tien ch'uan is identified by Morohash1 (5833.932) with T ten
Huang. For K'ang, see SCC, vol. III, p. 235).
The authority of the emperors of Ch'in
and Han
Introductory remarks
A deep contrast may be drawn between the concepts of imperial sovereignty
that were accepted in 221 BC and AD 220. At the outset of the Ch'in empire
(221-207) the right to rule needed no greater defence or explanation than that
of successful conquest, achieved by force of arms; by the end of the Han
empire (202 BC -AD 220) a new dynasty was obliged to demonstrate that it
had received Heaven's command to rule and Heaven's blessing on its
undertakings. The first Ch'in emperor formulated and assumed his title by
arbitrary action;' in AD 220 Ts'ao P'i (187-226) went through a form of
reluctantly accepting nomination after receiving an instrument of abdication
from his predecessor, the last of the Han emperors.
While the governors of
Ch'in had been content to take material wealth and strength as the objective of
their rule, by the third century AD it had become firmly established that a new
dynasty could only claim support if it could show that its purpose lay in the
unfolding of cosmic destiny.
In these four centuries, political theory and constitutional practice had
developed alongside a problem that was destined to recur throughout imperial
times. This was the compelling need of a new dynastic house to justify its
displacement of a predecessor (often by means of force) and yet to show good
reason why it would be unjust for a rival group to seize power in its turn. It
became necessary to conceive or formulate ideal qualifications for sovereignty
that could be claimed without unduly manifest hypocrisy and which could be
shown to have eluded a dispossessed or conquered rival. The problem grew in
intensity, both as more and more contenders arose for supreme power, and as
statesmen and historians were more frequently obliged to make a definite
choice of those whom they would support in action or justify in writing. It was
solved only by incorporating temporal authority within a transcendent system
that embraced all activities and values of the universe.
1 SC 6, pp.19f; MH vol. II, pp.122f.
2 This incident is summarised in San kuo chih 2 (Wei), p. 62, where the notes cite a number of
accounts and documents. In particular, seep. 75 note 3, for a citation from the Hsien-ti chuan;
for critical comments on these sources, sec de Crespigny (1970); see also Hou Han chi 30.l 7a and
Liang Han chin shih chi 18.4b and 18.lOa, and Leban (1978).
86 Divination, mythology and monarchy
The new view of imperial sovereignty should be considered within the
context of the major change that overcame religious, intellectual, and political
attitudes in these centuries, and which is sometimes described as the victory of
In addition the change in the concept of imperial sovereignty
should be considered in relation to the role actually played by the emperor in
affairs of state. For while the four centuries in question witnessed a growing
insistence on the religious values and ethical considerations of empire, they
also witnessed a number of occasions when the personal powers of the
emperor were reduced to a point of disappearance, in favour of the control of
government by other parties. Thus, from almost the outset of the Han dynasty
there were occasions when infants or minors were installed as emperors, while
an empress dowager, an imperial concubine, or her relatives moulded the
decisions of state. There is also a further hint that the emperor's personal
powers were not of a practical, immediate nature. Of all the emperors of
Former Han (202 BC --- 8 AD), Wu-ti (141-87) is credited with the greatest
vigour, personality and achievement; but on inspection it appears that the
majority of actions in which he took a personal part were neither political nor
military; they were concerned with religious cults.
On the one hand, greater emphasis was being placed on the divine nature of
imperial sovereignty and Heaven's part in ensuring dynastic continuity; on the
other, those who exercised supreme power of government stood beyond an
emperor's control rather than beneath his supervision. It appears that
statesmen, however strong, who wished to fulfil their ambitions, could not
dispense with the existence of a formally installed emperor; but only a
relatively small number of emperors can be shown to have taken an active,
personal part in controlling China's destinies.
Whatever the later protestations may have been, in at least two instances
Han statesmen recognised with some embarrassment that their dynasty had in
fact been founded by force majeure, in the same manner as Ch'in. The first
occasion is that of a discussion held between two men of learning before
Ching-ti (157--141). The question had arisen of the morality of the action
taken by T'ang and Wud in putting to death the last kings of the Hsia (trad.
dates 2205-1767) and the Shang (c. 1700-c. 1045 BC) dynasties. Yuan Ku
(third/second century BC) insisted that the action was defensible and even
commendable, and that the extent of the popular support they had enjoyed
showed that they had received a mandate for their actions. Mr Huang
(Huang-sheng) replied that, in so far as they were of an inferior status to that
of men who had been born kings, they had had no right to murder their
superiors; to which Yuan Ku raised the question of the legality of Kao-ti's
(202-195) action in replacing Ch'in, ifHuang's principles were to be accepted.
It was at this point that Ching-ti saw fit to bring the discussion to an abrupt
close; and according to one account of the incident, academics did not
' For this interpretation, see HFHD vol. II. pp. 34lf, and Hu Shih (1929). For the relation of the
change to political manoeuvres, see CC.
The authority of the emperors of Ch 'in and Han 87
subsequently presume to discuss such questions or to clarify these issues.
The second instance occurred about a century later, and the implication
which had alarmed Ching-ti was voiced explicitly. In 46 BC, I Feng (jl. second
half of the first century BC) was pleading for the removal of the seat of
imperial government from Ch'ang-an to Lo-yang. He based his case on the
need to return to the ideals of economy or parsimony that had been ascribed to
Wen-ti (180---157) and to strive for the qualities postulated of the old kingdoms
of Chou (c. 1045-256). He drew a blunt contrast between the ideals of those
kings and contemporary Han practice, and even went so far as to state that
Han had been founded on the basis of military strength and low cultural
standards. This view was completely contrary to that of Pan Piao (3-54), who
shortly afterwards stated his belief; he regarded it as a vulgar misapprehension
to think that Kao-ti had established his empire by means of force alone.
From these passages it may be inferred that well after the foundation of the
Han dynasty there were still some statesmen who could not accept that it had
been fully backed by moral considerations from its time of origin. The
development of the belief that sovereignty must rest on right as well as might
and the steps that were taken to propagate that view may be traced in evidence
of three types, documentary, procedural, and philosophical. The documents
are the summaries of memorials and edicts that attended major occasions of
dynastic change, such as the accession, deposal, or abdication of an emperor.
The procedures and formalities of these occasions, as described in the
histories, reveal the importance of certain symbolical acts; and both the
histories and other sources preserve contemporary tracts which bear directly
on the theoretical issues under discussion. In no case is the evidence complete,
as the habits of Chinese historians and the accident of literary survival have
denied posterity a full archive. Nevertheless the value of the evidence,
particularly that of the documents, is strong. We may possess no more than a
selection and summary of many of such pieces, but their very retention implies
that they promote or defend the preferred opinion of those who practised
imperial sovereignty, and that they served to support their claim for
The principal occasions of dynastic change or controversy which gave rise
to significant edicts, memorials, decisions, and procedures were the founda-
tion of the Ch'in empire (221 BC); the accession of Liu Pang (248-195) as
emperor of Han (202 BC); the effective control of state by the Empress
Dowager Lua, at a time when two infants successively held the title of emperor
(187-180 BC);
the elimination of the Lu family and the accession of Liu Heng
(202-157), known as Wen-ti (from 180 BC). The death ofChao-ti (87-74) in
74 BC gave rise to a crisis in which Liu Ho's (d. 59 BC) accession was abruptly
followed by his deposal twenty-seven days later
and the accession of Liu
SC 121, pp. 16f; HSPC 88.18b.
' HSPC 75.20aff; for Pan Piao's views, see p. 109.
These are referred to as Shao-ti Kung (188-184) and Shao-ti Hung (184-180).
Divination, mythology and monarchy
Ping-i (91-49) as Hsiian-ti (74-49). The basis of sovereignty was next brought
into question in 5 BC by some who thought the survival of the dynasty to be in
jeopardy and sought a means of dynastic renewal.
There followed the
accession of Wang Mang ( 45 BC - 23 AD) in AD 9 as first and only emperor of
the Hsin dynasty (9-23), and in AD 23 that of Liu Hsiu (6 BC- 57 AD) as first
emperor of the restored Han dynasty (Hou Han, 25-220). By now the main
changes in the view of imperial sovereignty had been effected; the documents
and procedures which attended the abdication of the last of the Han emperors
(Hsien-ti, reigned from 189 onwards) in AD 220, and the accession of Ts 'ao P'i
as emperor (W en-ti, 220-6) of the new dynasty of Weib (220-65) elaborate and
underline the principles that had by then become accepted. In addition there
were a number of occasions marked by dynastic plot, or when there was no
recognised and nominated heir apparent, which led to a consideration of rival
claims to the title of emperor, and a definition of some of the principles of
hereditary accession (i.e., in 80, 45, and 8 BC and AD 5).
The issues at stake will be considered below under the three headings of (I)
the links forged with a superhuman world; (II) the properties required of an
individual emperor; and (III) the symbolical procedures which were deemed
necessary at an emperor's succession. We will then briefly examine (IV) some
of the theories put forward by writers such as Tung Chung-shu (c. 179-c. 104
BC), Pan Piao, Wang Ch'ung (27 - c. 100) and Wang (c. 90-165).
with a superhuman world
'Pfte Mandate of Heaven
The expression Mandate of Heaven (t'ien ming) appears some seven times in
the Book of Songs (Shih ching), in the sense of the bidding given by Heaven to
certain kings and their obedience thereto. In one passage
we read of Heaven's
orders to the Black Bird (Hsiian niao) to descend and give birth of Shang, and
the poem continues with references to the continuity of the line of the kings of
Shang. Other poems refer to Heaven's orders to the king of Shang and to King
Wen of the next dynasty. 10 One of the most telling references to the Mandate
of Heaven, part of which is cited in the Meng-tzu, is in praise of King Wen;
the poem mentions his debt to the Mandate of Heaven and the possibility that
Heaven will bless his descendants. Above all the poem asserts the highly
important thesis that the Mandate does not remain constantly with one
incumbent, as may be shown in the transfer from the house of Yin ( = Shang)
7 For details, see CC pp. 75f. ' See CC pp.278f.
Shih ching no. 303, SSC 20(3).l2a; Waley (1937), p. 275; Karlgen (1950a), pp. 262f.
10 Shih ching nos. 244, 305, SSC 16(5).lOb and 20(4).9b; Waley (1937), pp. 263, 279; Karlgren
(1950a), pp. 198, 265-6. [Forother aspects of the Mandate at this early stage, Allan
for example, p. 531, for the suggestion that the seeds of the theory lay m the of
Shang to maintain its monopoly for divination and Chou's assertion that, thanks to its own
exercises in divination, their own cause was favoured by the supreme powers.]
11 Shih ching no. 235, SSC 16(1).la; Waley (1937), p. 250; Karlgren (1950a), pp. 185f.
The authority of the emperors of Ch'in and Han 89
to that of Chou. It was this
key expression, t'ien ming mi ch'ang, that was
selected for citation in the Meng-tzu, where there are surprisingly few
occurrences of the term t'ien ming.
Four crucial passages in the Book of Documents (Shu-ching) allude to the
belief that Heaven had transferred its mandate from Yin to Chou, or ordered a
king to eliminate an unjust predecessor.
However this view does not appear
to be re-iterated in the Tso chuan, whose references to the Mandate of Heaven
are of a somewhat different type;
for here they involve the fate or behaviour
of individuals only, without the majestic concern with the destiny of royal
houses. Nor does such a concern form a central point of Confucius' teaching
in the Analects (Lun-yii).
[In what has been described as one of the lost texts
of Huang-Lao thought ('Liming', in Shih liu ching; see Ma-wang-tui Han mu
po shu vol. I, p. 61) Huang-ti, the Yellow Lord, states that he received his
charge (mingb) from Heaven, established his position on earth and created his
reputation among mankind.]
It has been suggested by one scholar
that, while it cannot be known when
or how the doctrine of Heaven's mandate to kings originated, its formulation
or propagation may possibly be traced to the Duke of Chou. The doctrine was
doubtless of value to those who wished to exculpate the house of Chou from a
charge that it had eliminated its predecessor unjustly, by force. In due course
the doctrine became an integral part of the imperial creed of life; but it is of
considerable importance to observe the absence of a direct continuity between
the statements attributed to the Duke of Chou and the claims made by the
empires. r;,
In the first instance there was a long interval of time between the early Chou 0
kings and the full invocation of the Mandate of Heaven in support of imperial
government. The political circumstances of the Ch'un-ch'iu (722-481) and
Chan-kuo (403-221) periods were hardly such that the leaders, protagonists or
monarchs of the day could expect to call on the blessing of Heaven; and as will
be shown below, after the establishment of imperial government over two
centuries had still to elapse before an emperor would claim to be the recipient
of the Mandate.
Secondly, there is a considerable difference in the principles whereby the._
Karlgren (1950a), p. 186: 'Heaven's appointment is not for ever'; Legge, vol. II, p. 297;
Meng-tzu, SSC 7a.10b; Lau (1970), p. 120: 'Because the Mandate of Heaven is not immutable'.
SSC ('T'ang shih') 8.2aff; ('Ta kao') 13.15bff; ('K'ang kao,) 14.2bff; ('Chiin shih') 16.18aff;
Karlgren (1950b), pp. 20, 37, 39 and 59.
Tso chuan (SSC ed.) 39.22a, Couvreur (1914), vol. II, p. 541 for 544 BC; 41.25b, Couvreur,
vol. III, p. 36 for 541 BC; 52.l 7b and 20b, Couvreur vol. III, pp. 424, 430 for 515 BC; 54.25b,
Couvreur vol. III, p. 514 for 506 BC; and 59.2la, Couvreur vol. III, p. 707 (for 480 BC).
The best-known reference to the Mandate of Heaven in that work is the famous statement that
on the age of fifty, the Master 'understood the ordinances of Heaven'; see Lun-yu 2
('Wei cheng') SSC 2.2a; see also Lun-yu 16 ('Chi shih') SSC 16.7b (Legge vol.I, pp. 146, 313).
Creel (1970), pp. 82f; [for other views of the Mandate as seen in Western Chou see Hsu and
Linduff (1988), pp. 101-6, 382-3. For King Wen's receipt of the Mandate, see' Shaughnessy
(1991), pp.233, 227-8 and 246-7.]
90 Divination, mythology and monarchy
doctrine was applied. For the kings of Chou it served to explain the right of
succession which followed the conquest of a predecessor.
By the time of
Wang Mang it was linked with a theory of cosmic and dynastic change that
derived not from conquest but from the natural processes of birth, death, and
rebirth; the view that the doctrine survived without change from the time of
the Duke of Chou throughout the imperial period perhaps needs some
Some two centuries passed in the imperial age before it was recognised that
the existence and successful survival of a dynasty was bound up with the
bestowal of a specific trust or order. In the account of the first Ch'in emperor's
adoption of his title, we read solely of his successful conquest of his enemies,
but by the time of Han Kao-ti there is a perceptible change. For it is
acknowledged that such achievements depend on something more than
human proficiency, and that the strength of the victor derived from Heaven.
1 However, gratitude to Heaven for such gifts is still far removed from a belief
that Heaven is the ultimate source of temporal authority; such a concept does
not appear in the documents of accession until the time of Wang Mang.
Several of Kao-ti's supporters argued that he owed his personal success to
the strength drawn from Heaven.
On one occasion, the emperor himself
acknowledged that he had achieved the foundation of the empire thanks to
Heaven's spiritual power (lingb);
and as his life was drawing to a close he
likewise recognised that the decree of Heaven, which had vouchsafed him
strength earlier in his career, was inexorably bringing about his death.
In the
same way the Empress Dowager Lu recognised that Heaven had decreed her
own end;23 and the accession of Liu Heng was accompanied by the assurance
that the transfer of popular support from the family of Lu to the house of Liu
was due to superhuman strength.
The only reference in these early days to
Heaven's bestowal of a Mandate was not made in support of a Han emperor.
It occurs in the arguments put forward by Lou Ching (jl. third/second century
BC) for the establishment of the new imperial capital at Ch'ang-an rather than
Lo-yang; Lou Ching took the opportunity to contrast the Chou dynasty,
which had ruled by virtue of moral qualities and the Mandate of Heaven, with
Han which had not been thus favoured.
17 Creel ( 1970), pp. 84f.
" Creel (1970), p. 93, 'A new concept of the state, based upon the idea of the Mandate of Heaven,
came into being. No alteration of this basic concept, of anything approaching comparable
scope and depth, would again occur in China before the twentieth century. The doctrine of the
Mandate of Heaven became the cornerstone of the Chinese Empire.'
19 SC 6, pp. l 9f; MH vol. JI, pp. 122f.
20 Such supporters included Chang Liang(d. 168 BC), Han Hsin (d. 196 BC), Li I-ch'i (d. 204 BC)
and Lu Chia (c. 228 to c. 140 BC). For the former two, see SC 55, p. 7, and 92, p. 37, and LH 3
(Ming lu), p. 23; Forke (1907--11), part I, p. 148; for the other two, see SC 97, pp. 10, 13.
21 HSPC 1B.17b; HFHD, vol.I, p.131.
22 HSPC IB.22b; HH!Dvol. I, pp. 142-3; SC 8, p. 84; MHvol. II, p. 400; sec also LHas cited in
note 20 above. 23 For her interpretation of portents. seep. 95 below.
24 HSPC 4.2a; HFHD vol. I, p. 224.
2' SC 55, p. 20 and SC 99, p. 2; HSPC 1B.7a; HFHD vol. I, p. 108; and HSPC 43, p. !Ob.
The authority of the emperors ()f Ch'in and Han 91
Reference will be made below to theories formulated by Tung Chung-shu
and the place that he found for imperial sovereignty within his concept of the
universe. In the record of political events and discussions, it seems that we
must wait until c. 45 BC before we find a significant new departure. A new
attitude is discernible in a memorial submitted by K'uang Heng (first century
BC), a statesman who dissociated himself sharply from the realist and
expansionist policies ofWu-ti and his predecessors, and strove to promote the
ethical ideals of government that were ascribed to the kings of Ch(m. In
writing about dynastic destinies,
he observed that the duty of a 'king who
had received the mandate (shou ming chih wang) lay in transmitting his
inheritance as a possession for ever'. In expanding his theme, K'uang Heng
cited precedents from the kingdom of Chou and a passage from the Book of
Songs to show that a sovereign's attention to correct principles merited the
blessing of various types of deity, at whose head stood Heaven above.
Shortly afterwards, in the reign of Ch'eng-ti (33-7) the somewhat revol-
utionary view was put forward by Kan Chung-k'o, a specialist in calendrical
science, that the Han dynasty had run its course and stood in need of
This assertion included a crucial statement which referred to
the Han dynasty's receipt of a mandate from Heaven: 'The Han dynasty has
come to the final end of the cycle of Heaven and Earth, and stands ready once
more to receive the Mandate from Heaven'. The view was repeated in 5 BC by
a pupil of Kan Chung-k'o named Hsia Ho-liang; and he even went so far as to
suggest that the failure of the previous emperor, i.e. Ch'eng-ti, to produce an
heir had been due to his inability to respond to the Mandate of Heaven.
Attention has been paid above to the claims made on behalf of the. emperors
of two 11ewly arisen as to assert their exercise of authoritv. For the
first emperor of Ch'in, the claim rested the of the
unification of China's rival kingdoms; on behalf of Liu Pang, it was alleged
that his strength and success had been due to gifts bestowed by Heaven
personally. As compared with these two cases, the accession of the first
emperor of the next house, Hsin, was marked by a studied attempt to trace the
claim and title to the transcendent will oTsuperhuman powers.-tl:le
M<i11date of Heaven ..vas closely linked with the rhythmical cyde of the Five
.. 2,f_}>QaSeS (WU the operation of that cycle was plain to see.in
the preliminary portents which were reported from many quarters of the
t!mpire. The new dependence on the Mandate of Heaven as a meaiis of
conferring legitimate power is ifhistrated in the contrivances devised immedi-
ately before Wang Mang's assumption of the imperialtitle in AD 9.
A bronze casket, which was found fortuitously and most opportunely in the
memorial shrine dedicated to the founder of the Han dynasty had been
secured and sealed in the usual manner, but it was the terms of the two
inscriptions with which the fastening was made that were of prime interest.
2 HSPC 81.6a. 21 See CC p. 279. 2
' HSPC 75.3lb.
2 HSPC 99A.35b; HFHD vol. III, p. 254.
. f
92 Divination, mythology and monarchy
For they permitted Wang Mang to claim that he was accepting the abdication
of the Han house, and that his own descent could be traced to Huang-tih, the
Yellow Emperor and Yu ti (Shun). In the rescript that he issued, Wang Mang
wrote that Almighty Heaven and God on high had vouchsafed the clearest
indications of their support, and enjoined upon him the responsibility for the
welfare of all peoples of the world; he could not but accept the Mandate, in the
deepest sense of reverence. By way of advertising the strength of his claims,
Wang Mang had the banners and the devices of the new dynasty's officials
inscribed in such a way that they would make clear that imperial authority
rested on the awesome mandate of Almighty Heaven and God on high.
In the civil wars which attended the downfall of Wang Mang and the
restoration of the Han dynasty, a number of claims were made by rival
contenders to show that they stood possessed of the Mandate of Heaven. One
supporter of Liu Hsiu (Kuang-wu-ti 25-57) is reported as claiming that Wang
Mang's usurpation constituted an offence against Heaven, whose Mandate
really served to uphold Han.
In a despatch which was generally circulated
throughout the empire, Wei Ao (d. AD 33) likewise accused Wang Mang of
lodging an unsubstantiated claim to possess the Mandate,
and elsewhere
claimed that he was the true recipient; Kung-sun Shu (d. 36), another
pretender, acted in like fashion.
The claims of Liu Hsiu (Kuang-wu-ti) were made no less clearly in the
documents which attended his accession,
where a new principle may be
observed that was of no small importance. This was to the effect that, first, the
Mandate of Heaven cannot be gainsaid; and secondly that it must have a
repository. As has been seen, in its earliest forms the doctrine provided for the
removal of the Mandate from one house to another. In the intervening
centuries since the Book of Documents and the Book of Songs, the place of
imperial sovereignty had become enshrined as an integral part of the cosmic
system, thanks partly to the teachings of Tung Chung-shu. The new principle,
that Heaven cannot brook an interruption of the Mandate and must ensure
that it is vested in a recognised incumbent, follows as a logical necessity from
those beliefs. It was a principle that was destined to be of profound importance
in the subsequent stages of China's dynastic history and historiography.
There were other occasions in the early part of the Later Han dynasty when
statesmen or emperors referred to the Mandate. Feng I (d. 34) once re-assured
Kuang-wu-ti that one of his dreams signified the manifestation of the
Mandate on his behalf.
In the account of his final acceptance of the title,
Kuang-wu-ti referred to its bestowal as a gift from the powers of Heaven and
Earth; and he avowed his faith that his successful conquest of his enemies
: proved his conformity with the will of Heaven and his command of popular
30 This supporter was Wang Lang; HHSCC 12. lb. [For a reconsideration of the view of Wang
Mang as an usurper, see Bielenstein (1954--79), vol.I, pp. 145f, and Bielenstein (1986),
pp. 223f.] 31 HHSCC 13.2a and 4b.
HHSCC 13.14aff.
33 HHSCC !A.14b. 34 HHSCC !7.4b; HHSCC 22.lb. " HHSCC (treatise) 7.lb.
The authority of the emperors of Ch'in and Han 93
support. In a solemn declaration made towards the end of his life, in AD 56, in
the shrine dedicated to Kao-ti, the same emperor referred to the seizure of
power by the Empress Dowager Li.i and the elimination of the Li.i family after
her death;
he epitomised the process as the fall of the Mandate, followed by
the restoration of peace to a dynasty that had been in danger. The first edict of
Kuang-wu-ti's successor, Ming-ti (57-75), referred to his predecessor's receipt
of the Mandate.
Most significantly, on the next occasion when a new
dynasty was established in place of a defunct regime, the king of Wei could
write: 'It is not right to refuse or withstand the Mandate of Heaven; it is not
right that the Sacred Instrument of power should long be left abandoned; it is
not right that the whole body of servants of state should be without a master; it
is not right that the manifold problems of state should be without control.'
The Five Phases
A well-known passage in the Lu shih ch'un-ch'iu
declares that when a
sovereign or king is about to arise, Heaven invariably takes the preliminary
step of displaying tokens of his good fortune to mankind. The theme is
illustrated in the case of four rulers, each of whom arose in his due season,
preceded by the appearance of appropriate material signs. Thus, at the time of
the Yellow Emperor, Heaven produced creatures of the earth, such as worms
or crickets, whose presence suggested his association with the energy of Earth;
his successors were likewise accompanied with symbols of their own particular
phase, i.e. Wood (for Yi.ia, founder of Hsia), Metal (for T'ang, founder of
Shang) and Fire (for King Wen, founder of Chou).
The passage shows how, in pre-imperial days, a link had been forged
between the power of a particular ruler or house and one of the Five Phases
whereby the processes of nature unfold. There are two important points in the
theory as first expressed. First, the scheme or rhythm was based on the belief
that each phase was initiated after the conquest of its predecessor, i.e. by the,-,
process of successive conquest (hsiang shenga); and secondly, the fifth phase f . - ~
the process, which was symbolised by Water, was still unengaged, awaiting
association with a temporal dispensation yet to come.
It cannot be said how widely the belief in this theory was held at the time of
Ch'in's unification. Although it would seem to have had little appeal to the
highly realist statesmen of that regime, no less than four passages of the
Shih-chi refer or allude to Ch'in's deliberate adoption of water as its patron
symbol; and although some doubts have been cast on the authenticity of such
" HHSCC IB.21b.
HHSCC 2.lb.
38 See the passage from the Hsien-ti-chuan, as cited in SKC 2 (Wei) p. 75, note 3.
" For the adoption of patron symbols in the Ch'in and Han empires, see chapter 2 above. For the
term 'Five Phases', see Major (1976).
40 LSCC 13.4a. Being an eclectic work, this book includes passages whose political thought is of
varied types; see for example, 20.la, which may be compared with Hsun-tzu 9 ('Wang chih'),
94 Divination, mythology and monarchy
an act, the statements can hardly be gainsaid.
It would appear that the early
rulers of Han paid similar attention to the belief, by retaining Water as their
symbol, although there is no specific statement of such a decision. However
the importance attached to associating the dynastic rule of Han with one of
the Five Phases became apparent quite soon. Shortly after the accession of
Wen-ti, in 180 BC, Chia I (201--169 BC) suggested that the element should be
changed to Earth, and the proposal was repeated by Kung-sun Ch'en (fl. first
half of second century BC) in 166 BC. On both occasions the idea was rejected,
and it was only in I 04 BC that it was adopted. The change to Earth in that year
accompanied a number of new regulations for protocol which were designed
to display the strength of the Han dynasty and the success of its expansionist
policies. The adoption of Earth, the known conqueror of Water, symbolised
Han's victory, not only over the enemies ofits own choosing, but also over its
predecessor, Ch'in.
The link between Han and Earth remained unquestioned until the end of
the Former Han dynasty, when new ideas were beginning to circulate. It was
now becoming accepted that the Five Phases succeeded each other not as a
result of conquest but by way of natural production or creation (hsiang
shengb), and this view was incorporated in some of the documents which
accompanied Wang Mang's accession.
In addition those documents insisted
that the Han dynasty had enjoyed the protection not, as had been maintained
hitherto, of Earth, but of Fire; and as the natural successor to Fire was Earth,
Wang Mang declared that his own dispensation was blessed by Earth. It may
be noted that when the time came for the restored Han dynasty to choose or
identify its patron, it accepted precisely the same premises. Kuang-wu-ti chose
Fire, which he believed to have been the patron of his ancestors of Former
Han and to be in the ascendant by virtue of natural succession rather than
conquest. Kung-sun Shu, however, in his brief bid for power as emperor chose
White or Metal.
He, too, followed the theory that the Phases followed one
another naturally and not by conquest; but while Kuang-wu-ti ignored the
claims of Wang Mang as being the acts of an illegitimate usurper, Kung-sun
Shu accorded him recognition; for he chose the symbol of that phase which
followed Wang Mang's symbol of Earth.
At the time of Wang Mang's accession a new measure of emphasis had been
imparted to two concepts, that of the Mandate of Heaven and that of the link
between a temporal dispensation and the predominance of one of the Five
41 SC 6, p. 23; SC 15, p. 122; SC 26, p. 9; and SC 28, p. 19; MHvol. II, p. 129 and vol. III, pp. 328
and 430. For doubts regarding the authenticity of Ch'in's adoption of a patron symbol, see
Kurihara (1960), pp. 45f, Kamada (1962), pp. 42f, [and Bodde(l986), p. 97 for a review of such
42 HSPC 99B.9b, lOb; HFHD vol. III, pp. 290, 293; see also HFHD vol. III, p. 259, note 36.5.
43 HHSCC 13.15a; Biclenstein (1954-79) vol. II, p. 234.
HHSCC 13.16b.
The authority of the emperors of Ch'in and Han 95
Phases of creation, death, and rebirth. In addition the treatment and
interpretation of portents had undergone a change which was of no less
importance, as far as the sources permit conclusions to be drawn. It is at times
not possible to determine whether the histories include contemporary or
retrospective interpretations of these events.
The change was from a negative to a positive attitude. The passage from the
Lu shih ch'un-ch'iu which is cited above asserts the principle that natural
phenomena foretell the rise of certain individuals to their rightful position as
leaders of the world. This positive principle seems to have been conspicuously
absent at the beginning of the imperial age, but it appears from the time of
Hsiian-ti (r. 74-49 BC). In the meantime a negative aspect of the principle had
been formulated and propagated by thinkers such as Tung Chung-shu. This
was to the effect that Heaven brings about strange or even unnatural portents
as a warning to a badly disposed monarch in the hope that he will mend his
ways; and in the last resort such phenomena, or miracles, may be interpreted
as a prediction of the end of a monarch's period of rule. It was this view of
portents which later attracted the sharp criticism of Wang Ch'ung.
On two occasions the Empress Lii is reported as interpreting events as
portents of her own demise, i.e. an eclipse which took place in 181 BC, and the
sight of a peculiar dog, who reared up and bit her in the side, in the following
However, there is a conspicuous absence of references to favourable
portents in the pleas put before Liu Heng to assume the imperial title.
0n this
occasion (180 BC) there had been an interruption in the imperial succession,
and the counsellors of state were anxious to draw on all possible reasons to
persuade Liu Heng to accede to their request. For this reason, perhaps more
weight should be attached to the absence of such references than is usually
warranted for an argumentum ex silentio. Similarly there is no reference to
favourable portents in the documents that attended the accession ofHsiian-ti
in 74 BC, again after a dynastic crisis. Portents of failure, or disaster, are
reported for two incidents of an unsuccessful bid for imperial power. These
were the strange events said to have presaged the king ofYen's
abortive plot
to seize the throne in 80 BC, and the failure of Liu Ho to rule as emperor for
more than twenty-seven days in 74 BC.
The development towards a positive attitude took place during the reign of
Hsiian-ti. The first report of a superhuman activity to be associated with the
rise of a monarch lies in a slightly different category from that of specific
45 LH 42 ('Ch'ien kao'), p.634; Forke (1907-11), vol.I, pp.119f.
46 For the eclipseof4March181 BC, see SC9, pp. 18--19, MHvol. II, p.422; HSPC3.4b,HFHD
vol. I, pp. 199, 211; for the sight of the dog, see SC 9, p. 21, MH vol. II, p. 425.
47 Portents are to be distinguished from divination, for which seep. 106 below.
Portents reported for 80 BC included (i) an eclipse (20 September): HSPC 7.6a; HFHD vol. II,
p. 164; HSPC 27C(2).14b; (ii) the strange behaviour of rats, birds and swine: HSPC
27B(l).14b; HSPC 27C(l).5b; HSPC 27B(2).8a and 20b; and (iii) the outbreak of fire: HSPC
27 A.13b. For 74 BC, darkness was prevalent by day and night, in the absence of the sun and the
moon: HSPC 27C(l).11a; and a meteor was reported: HSPC 7. lOa; HFHD vol. II, p. 174;
HSPC 26.55a.
Divination, mythology and monarchy
incidents witnessed in material form. Towards the end of Wu-ti's reign it was
asserted that the 'invisible presence of the Son of Heaven' (t'ien tzu eh 'i) lay
within the prisons of Ch'ang-an; as a precaution against treason or dissidence
might arise therefrom, the government ordered the death of all those
mcarcerated there, lest one of them should attempt to become emperor. One
?f was Liu Ping-i, later to become emperor as Hsiian-ti. The
mc1dent is related not as part of the significant actions ofWu-ti's reign, but as
indication that the imperial destiny of the future monarch, still in his
mfancy, had been revealed by superhuman means.
It was during Hsuan-ti's reign that a change is noticeable in the official
treatment of phenomena. Between 65 and 51 BC the imperial edicts which
referred to strange phenomena fastened almost exclusively on the beneficent
nature of certain portents.50 The selection of terms such as Holy Bird
(Shen-chiieh), Five Phoenixes (Wu-feng), Honeydew (Kan-lu), and Golden
(Huang-lung) as the regnal titles for the period 61 to 49 BC indicates
the importance attached to omens of felicity at this time. As yet, however,
there was no occasion of dynastic crisis wherein the concept of favourable
omens could be invoked to resolve doubts regarding the imperial succession or
to support the claims of a particular candidate. The principle appears,
somew.hat obliquely, in 5 BC, in regard to the suggestions of Hsia Ho-liang for
dynastic renewal or rededication. 51 Reference was duly made on this occasion
to the warnings that Ch'eng-ti had received from Heaven for his failure to
respond to his proper calling. But one of the reasons why Hsia Ho-liang's
proposals were rejected is stated to be the absence of felicitous portents which
would corroborate the advent of a new imperial era. It would seem that in 5
BC there were some circles in government who looked to a positive sign of
Heaven's blessing before dynastic change could be contemplated.
The case of Wang Mang marks the first definite instance in which attention
P.aid to the value of favourable portents in establishing a claim for
imperial sovereignty.
the of his from of f ovnda,tion
.cl.ane.w __dyrillsty. Reference has been made above to the inscriptions attached
to the casket, whose discovery took place so opportunely before Wang
Mang'.s accession; but this was only the final example of many incidents
wherem Wang Mang showed his faith in felicitous omens. When his
supporters bestowed upon him the honour of the Nine Distinctions in AD 5,
HSPC 8.2a, HFHD vol. II, p. 201; for a comparable example later, see HHSCC 5. la.
Of _edicts which concerned phenomena between 178 and 66 BC, eleven fastened on
cal_am1ties as.n:a.tters of warning; seven, which reported happy events, were linked with Wu-ti's
act1v1ties. Between 65 and 51 BC, there is a different pattern, with three edicts
concernmg warnings and nine concerning happy phenomena. It may be noted that thereafter
no edicts of Former Han concerned phenomena of happy augury.
HSPC 11.Sa, HFHD vol. III, pp. 29f; HSPC 7 5 .31 b.
HSPC HFHD vol. III, pp. 204f; HSPC 99B.12a, HFHD vol. Ill, p. 288. The Nine
Disti_nct10ns consisted of privileges and insignia which displayed and confirmed Wang Mang's
dommant and favoured position in the empire.
The authority of the emperors of Ch'in and Han 97
they referred to the appearance of something . more than 700 auspicious
omens. Very soon after the death of P'ing-ti (1 BC--6 AD) Wang Marig's
future entitlement was proclaimed by the discovery of a stone, inscribed with
the message that Wang Mang should become emperor.
After Wang Mang's time, felicitous omens featured in the accessions of
and the first of the Wei emperors in 220. On that occasion, the
rescript in which Hsien-ti announced his abdication justified his action in part
on the appearance of 'auspicious omens sent down by the Almighty'
(Huang-ling chiang jui).
Tli.e. foregoing consider11ti()nS_l1lay 1Je,. S1Jl1ll!lar:is,eci by o 1Js.erv..i11g .. th::It .. the V
Q[igins qf thJee.Pdncipks or .concepts .. may .be.Jraced.to .. the pre-imperial
p()rioci; these are.those of the Mandate of Heaven, the Five Phases and the
of po.rte11ts_;is harbi11gei:s.of 3: fe,licit.011s.n::ign. In the early days of
imperial government they did not take a predominant place in political
thought; but by the time of Wang Mang the three concepts had come to
feature as cardinal points in an assertion of the right to sovereignty. In the
initial stages of empire, it was sufficient to claim the right of replacing a
predecessor by conquest; portents were interpreted as predicting the end of
an unsuccessful rule; and Heaven was simply cited as a bestower of strength
on a conquering hero. In the later stage, it was asserted that the replacement
of an imperial predecessor followed as naturally as the growth of one phase
of creation from the last; it had become essential to demonstrate that
sovereignty was exercised thanks to the direct Mandate of Heaven; and
portents were forthcoming to prove the valid authority of a newly arisen
(/'ifl;'he properties required of an emperor
rise and fall of the houses of Ch'in, Former Han, Hsin, Later Han and
Wei, and the circumstances in which some of the emperors were chosen to
to throne, witnessed 1.he .. c.Q11.flict 9.U.WQ !b;:it_9ja_n_-;f;'
__i:ig}itJQ .. e111pe,ror.. a11d ... t!ie .. 9.1la:!!fis;aJiC>t18-. i:!!.etiteq_ by
P.ernQl!!!l .11lOJ.l!l The conflict was by no
means new to Chinese thought and practice, and recurs on a number of
occasions and in a number of key statements of later imperial times.
In the early monarchies of China the principle of patrilinear succession was
by no means clear cut. In the kingdom of Shang, and in the system of kinship
of Chou, the transmission of authority and seniority had been regulated by
highly complex schemes that often superceded the direct right of a son to
follow his father.
By the time of the Warring States, however, the principle
" HSPC 99a.25a, HFHD vol. III, p. 218. For Wang Mang's attention to these matters after his
assumption of the imperial title, see HFHD vol. III, pp. 288f.
54 HHSCC 1A.15a. 55 SKC 2 (Wei), p. 62.
56 See K. C. Chang (1976), pp. 72f.
98 Divination, mythology and monarchy
was generally accepted, particularly in the kingdom of Ch'in. Nevertheless it
was subject to criticism by some writers such as the author of the Hsun-tzu,
who believed that man is greater than his institutions, and that merit is of more
significance than heredity in choosing a monarch.
It may be observed that in practice only a half or less of the titular emperors
of Han acceded in a regular undisputed manner as sons of an immediately
deceased predecessor.
However, the imperial houses theoretically accepted
the principle of hereditary succession on a patrilinear basis, so long as certain
conditions were fulfilled. The heir to the throne must be the son of the duly
nominated empress, and not of one of the minor consorts. The emperor could
choose which one of his sons he desired to nominate; but if he insisted on
selecting a son by another woman, it would first be necessary to demote the
existing empress and elevate that other woman in her place. In addition, from
early on in Han a show was made of ensuring that an incumbent to the
imperial throne stood possessed of the requisite moral qualities; by 74 BC this
principle could be invoked to the point of demoting a monarch who
demonstrably failed to reach the required standard.
The empire of Ch'in had arisen by the extension of the domains and power
of the kingdom of Ch 'in. Prince Chenga (259-210) had succeeded his father as
king of Ch'in in 246 BC, and it was from that position that he assumed the new
title of Huang-ti in 221, so as to correspond with his new dignity and strength.
In doing so he made it clear that the future succession would be hereditary,
passing from the first to the second and eventually to the ten-thousandth
emperor. Very soon it was shown that the right of nominating a successor lay
with the emperor of the day, as may be shown from the manner in which the
accession of the second emperor was presented to the public.
There is also
one further circumstance which is worthy of consideration. The Shih-chi
carries the allegation that, so far from being a true son of his royal father,
prince Cheng had in fact been sired by Lii Pu-wei (d. 235 BC), but the
statement is by no means proven.
The importance of the allegation, whether
true or not, presumably lay in its implication. If he was in fact no true son of
the king he had no right to succeed him, and thence become emperor.
There could be no question of claiming an hereditary right for Han Kao-ti,
but there is a somewhat uncertain allusion to the moral worth and wisdom
Hsiin-tzu 12 ('Chiin tao'), pp. 158f, and 17 ('Cheng lun'), pp. 234f.
" I.e., for Former Han, Hui-ti (195-188), Ching-ti, Wu-ti, Yiian-ti (49--33) and Ch'eng-ti only, as
against Kao-ti, the two infants enthroned under the Empress Lii, Liu Ho, Hsiian-ti, Ai-ti (7-1),
P'ing-ti and Liu Ying (b. AD 4). While Chao-ti was indeed the son of his immediate
predecessor, the circumstances of his nomination and accession were anything but regular. For
Later Han, Ming-ti, Chang-ti (7588), Ho-ti (88-106), Shun-ti (125-44), Ch'ung-ti (144--5),
and Hsien-ti, as against Kuang-wu-ti, Shang-ti (106), An-ti (106-25), Chih-ti (145--6), Huan-ti
(146-68) and Ling-ti (168-89).
For the manipulation practised by Chao Kao (d. 207 BC) and others, and the need to forge
imperial documents, see SC 6, pp. 66f, MH vol. II, pp.19lf and SC 87, pp.14f; and Bodde
(1938), pp. 25f.
SC 6, p. 2, MH vol. II, p.100, and SC 85, p. 7.
The authority of the emperors of Ch'in and Han
(hsiend) that an emperor was expected to possess. According to the sources,
when he was invited to assume the title of emperor, the king of Han protested
that he could not claim such qualities. His supporters immediately sought to
re-assure him; but the detailed report of the encouragement that they voiced
emphasises Liu Pang's achievements in the field and the measure of his
popularity rather than his claim to righteousness or wisdom. 61
Liu Ying (Hui-ti, 195-188) duly succeeded Kao-ti, being his son by the
Empress Lii. On his death in 188 BC, no son had been born to his own
empress, and his mother retained her position of dominance by contriving the
succession to the throne of two infants. The legality of this procedure could be
brought into question on two counts; first, that of enthroning a child who had
not been born of an acknowledged empress; and second, the possibility that
they had not been fathered by Hui-ti. 62
In the constitutional crisis which followed the elimination of the Lii family,
the statesmen and generals of the Han empire were faced with the problem of
enthroning an emperor who had not been nominated as heir apparent by his
father. The choice of Liu Heng was supported by considerations of hereditary
and personal qualities. As the older ofKao-ti's surviving sons, he was judged
to have a prior claim both over his younger brother and a nephew, although
the latter's father had in fact been a yet older son ofKao-ti. 63 In addition it was
held that Liu Heng possessed a reputation for those personal qualities that
had been sought by Kao-ti, i.e., moral worth and wisdom, together with a
sense of duty and obligation towards mankind and to his parents.
It was not until 87 BC that the next crisis arose concerning the imperial
succession. The dramatic events of the last few years ofWu-ti's reign had left
the empire without a nominated heir apparent, and with the deceased empress'
family virtually extinct. In Wu-ti's final illness the formalities of state were
duly observed,
when his infant son Fu-ling (94-74) was nominated as heir.
At the time there was no established empress, and as Fu-ling's mother had
already died, there could be no question of establishing her with that title. 65
Wu-ti died two days after Liu Fu-ling's nomination. His succession was
challenged by Liu Tan, king of Yen and one of Wu-ti's sons by another
consort. He staged two abortive attempts to seize the throne, 66 claiming that
he held a prior right over Liu Fu-ling (Chao-ti), as the eldest surviving son of
the late Wu-ti; and he even alleged that Chao-ti was no true son of his father. 67
The death of Chao-ti without a successor in 74 BC caused a further
constitutional crisis, in which his immediate successor, Liu Ho, was deposed
after twenty-seven days and replaced by Liu Ping-i (Hsiian-ti). While the
political motives and implications of these changes are not of immediate
SC 8, p. 63, MH vol. II, p. 380, and HSPC 1B.3a, HFHD vol. I, pp. 99f.
SC 10, p. 7, MH vol. II, p. II, p. 449; HSPC 3.8a and 4.3b, HFHD vol. I, pp. 209-10 and 227.
I.e., Liu Hsiangb, whose father Fei had been king ofCh'i until his death in 188; see HSPC 14.6a.
Sec CC p. 67.
The title was in fact granted posthumously at a later date; see HSPC 97 A.17a, CC p. 54.
CC pp. 57f, 74f. 67 HSPC 63.!0a.
100 Divination, mythology and monarchy
concern, 68 they necessarily involved discussion of suitable qualifications for
the position of emperor.
Strictly speaking, the candidate with the highest claim to succeed was Liu
Hsu, only survivor ofWu-ti's sons, and his cause was espoused by a number of
counsellors of state. 69 He was rejected on the grounds that his moral conduct
did not measure up to the position of emperor and that he would thus be
unsuited to receive charge of the imperial shrines. Historical precedent was
cited from the days of Chou to show that in certain circumstances an older son
could be rejected in favour of his younger brother. The choice actually fell on
Liu Ho, grandson of Wu-ti by one of the minor consorts. He was summoned
to Ch'ang-an to receive his charge and actually acceded to the throne. But
before long it was being put about that his improper conduct, extravagance
and lack of scruple made him totally unfit for the position. Detailed charges
were levelled against Liu Ho; the need for a ruler to possess moral
qualifications was shown by reference to the Book of Songs and the Spring and
Autumn Annals ( Ch 'un-ch 'iu );
and a precedent was found for the deposal of a
monarch who proveq to be unsuitable.
The choice of Liu Ping-i as emperor was justified on the grounds that he was
a great-grandson of Wu-ti, and that it was permissible to choose a candidate
from a collateral line if a monarch died without a successor.
At the same time
it was asserted that Liu Ping-i possessed the requisite qualities; he had been
educated in the tradition of the Book of Songs, the Analects and the Book of
Filial Piety (Hsiao ching); he practised thrift in his own behaviour; and he was
of a kind and philanthropic disposition.
The importance of choosing an imperial heir who was the son of the
established empress features in the advice tendered to the throne by K'uang
Heng, whose reference to the Mandate of Heaven has featured above (see
p. 91). This occurred in c. 45 BC at a time when the reigning emperor (Yiian-ti)
was showing marked favouritism towards a secondary consort and was
contemplating naming her son as heir apparent. K'uang Heng insisted on the
need to distinguish between the claims of the official empress and other
consorts; he believed this to be an inherent provision of the conventions for
The problem of the succession arose shortly before the death of Ch'eng-ti,
in 7 BC, with no son born to his name and still alive. The controversy centred
around the rival claims made for a son and a grandson of Ch'eng-ti's father,
neither of whom were descended from Ch'eng-ti's own mother.
The case for
Liu Hsinga, Yuan-ti's son, rested on the grounds that the son of a previous
" See CC pp. 75f and 119f. 69 HSPC 68.4b.
HSPC 68.9b.
71 HSPC 68. lOb. Liu Ping-i was a great-grandson ofWu-ti by descent from his Empress Wei and
the heir apparent Liu Chii, who had been forced to commit suicide in 91 BC; he thus came from
a line which was collateral with that ofChao-ti and Liu Ho; see CC table 2 (facing p. 64), [and
Twitchett and Loewe (1986), table 6, pp. 174--5].
72 HSPC 81.6b, and CC pp. 155.160.
73 HSPC 81.16a; CC table 4 (p. 156), [fwitchett and Loewe (1986), table 8, pp. 216-17].
The authority of the emperors of Ch'in and Han 101
emperor possessed a higher claim than a grandson; the supporters of Liu
Hsin, Yuan ti's grandson, argued that, as a nephew, he could claim a right
that was comparable with that of a son of Ch'eng-ti.
In the event it was Liu Hsin, grandson of Yuan ti and half-nephew of
Ch'eng-ti, who was chosen to be the new emperor, in preference for Liu Hsing,
Ch'eng ti's half-brother. Liu Hsin, or Ai-ti, died in his turn without a
nominated successor in 1 BC, and he was followed by the son of Liu Hsing,
i.e., Liu Chi-tzu (9 BC - 6 AD), or P'ing-ti. The problem arose once again
when P'ing-ti died in AD 6 at the age of fourteen. By then all the descendants
ofYiian-ti had died out, and the choice lay among the surviving descendants
of his father, Hsiian-ti. Ifwe are to believe our sources, Wang Mang was able
to ensure that an infant, whom he could manage, would succeed, and to
eliminate the claims of any surviving adults; he did so by citing the principle
that cousins should not succeed cousins of their own generation.
Wang Mang's assumption of the throne is perhaps the best documented act
of accession for the period under review, apart from the welter of reports that
refer to AD 220.
Wang Mang achieved the replacement of one dynastic
house by another without bloodshed, and by a series of cumulative steps
whereby he held the titles of Regent, Acting Emperor, and, finally, Emperor in
his own right.
But despite the essential rupture in the line of hereditary
succession Wang Mang was at pains to maintain some measure of continuity
with the house of Liu, displaced as this had been from the supreme position of
empire. Thus, the initial stages of the ceremonies which heralded Wang
Mang's accession took place in the shrine dedicated to Han Kao-ti. In his final
declaration to Liu Ying, the infant who had been named as heir apparent in
AD 6,
Wang Mang referred to the Han dynasty in terms of respect and made
provision for the continuation of services to its ancestral founders. In place of
descent from Liu Pang, Wang Mang claimed that his line originated from the
Yell ow Emperor and Shun.
Naturally enough Liu Hsiu could assert his valid claim to be a scion of the
house of Liu which he was restoring to the throne.
At the same time his
supporters genuinely allude to his personal achievements in bringing this
about, in the same way as the personal successes of forging the unification,
overcoming rivals and winning popular support had featured in the claims of
the first emperor of Ch'in, Han Kao-ti and Han Wen-ti.
On a number of occasions in Later Han an emperor died without a
nominated successor, often being no more than a child himself. The candida-
ture of An-ti, who succeeded at the age of thirteen in AD 106, was supported
on the grounds that he was the grandson of the previous emperor (Chang-ti),
74 HSPC 99A.24b, HFHD vol. III, pp. 217-18. 75 See note 2 above.
76 See HFHD vol. III, pp. 49f and 103f. 77 HSPC 99A.35b, HH!D vol. III, p. 254.
" HSPC 99A .35b and 99B.lb, HFHD vol. III, pp. 255, 261.
79 Liu Hsiu (Kuang-wu-ti) was actually descended from Ching-ti; HHSCC !A.la; [Bielenstein
(1986), p.245.] so HHSCC 1A.13b.
102 Divination, mythology and monarchy
there being no son available. It was added that he had been suitably educated
and that he possessed the right qualities of character. Other cases in which a
candidate was chosen on the grounds that he was descended from Chang-ti
included Chih-ti, Huan-ti and Ling-ti. Shun-ti, whose accession had been
thwarted by intrigue, was in fact a son of the previous emperor, An-ti. The
memorial which preceded his enthronement used an important expression for
later dynastic history. It observed that he was in the direct and correct line of
succession (cheng-t'ung).
In the final case, when the first emperor of Wei acceded, there could be no
question of asserting an hereditary right to the imperial throne, which had
been won by personal exertions. The deed of abdication ofHsien-ti, the last of
the Han emperors, draws attention to the moral qualities possessed by the new
incumbent, and credits him with regaining the inheritance of the kings of
III. ,The formalities of accession, deposal and abdication
\ ;
.. TM habits and accidents of historiography are such that the treatment of the
various state occasions of the Ch'in, Han and Wei empires, and the extent of
detail that is provided, vary considerably. While it is possible to observe the
appearance of certain significant features, their development cannot necessar-
ily be related to other changes of a religious or intellectual nature. We may
consider (i) the part played by counsellors of state in the validation of
accession or deposal; (ii) the reluctance displayed by a candidate before
accepting the title of emperor; and (iii) the insistence on an act of abdication
by a predecessor. Attention should also be paid to (iv) the importance of an
Empress Dowager on these occasions, (v) the value of the imperial seal, and
(vi) the religious ceremonies that accompanied an accession.
(i) The counsellors of state
Occasions of dynastic change were frequently marked by a formal request
submitted by senior counsellors to a dominant leader or chosen candidate,
inviting him to accede to the imperial throne. In 221 BC the king of Ch'in
sought the advice of senior officials concerning the title that he should adopt
and other matters of symbolical procedure that would demonstrate his
exercise of imperial power;
and the king himself only intervened when they
failed to respond adequately to his ambitions. More certainly and effectively,
the formalities of state were subsequently so ordered that it could be shown
how the dynastic succession, when not hereditary, depended on the expressed
" HHSCC 6.lb. 82 SKC 2 (Wei), p. 62.
" SC 6, pp. 2lf, MH vol. II, pp. 124f. Other matters which were concerned included the correct
terms for imperial commands and the expression used by the emperor when referring to
The authority of the emperors of Ch'in and Han 103
will of noblemen, generals, and ministers of state. Of particular note, this
procedure was followed in the dramatic events of74 BC; similarly the disputed
accession of Shun-ti, in AD 125, followed the presentation of a memorial by a
number of statesmen.
Finally, there was one circumstance of note in the
ceremony wherein the king of Wei received the instrument of abdication in his
favour from the last of the Han emperors.
This was delivered in the presence,
and with the assent, of senior statesmen, nobles, and general officers; in
addition the Shan-yii of the Hsiung-nu was present, together with a large
number of visitors to court from foreign parts.
(ii) The show of reluctance
In contrast with the eagerness whereby the first emperor ofCh'in assumed his
title, some of the later emperors went through a form of expressing reluctance
to accept the onerous charge that had been thrust upon them. This is first
reported for Han Kao-ti; but curiously enough it is mentioned only in the
Shih-chi, and not in the lengthier, more detailed account of the Han shu.
According to the Shih-chi, the king of Han thrice demurred from accepting the
title, and only did so when it became clear that his refusal was not to be
brooked. Precisely the same procedure was adopted in later cases, such as
those of Wen-ti and Kuang-wu-ti.
(iii) The act of abdication
Kao-ti's foundation of the Han dynasty in 206 BC, the elimination of the Lii
family in 180 and the deposal of Liu Ho in 74 were contrived without recourse
to a form of abdication by a displaced occupant of the throne. But the changes
of the latter part of the Han dynasty were subject to significant influences of
history and learning, owing partly to the propagation ofliterature that harked
back to the pre-imperial age. Wang Mang in particular was anxious to pose as
an exponent of traditional morality, and it is not surprising that he sought to
draw a comparison with earlier examples, such as the resignation of Yao and
Shun in favour of chosen successors. Wang Mang made a formal progress to
the shrine that was dedicated to Kao-ti, and there accepted the 'act of
abdication conveyed by the bronze casket'.
The importance of abdication also features in Pan Piao's treatise on the
See SC8, p. 63, MHvol. II, p. 380; HSPC 1B.3a, HE'HDvol. I, p. 99 for Kao-ti; HSPC4.lbff,
HFHD vol. I, p. 222 and HSPC 4.3aff, HFHD vol. I, p.226 for Wen-ti; HSPC 68.6bfffor the
events of74 BC; HSPC 99A.32a, HE'HD vol. III, p. 243 for Wang Mang; HHSCC 1A.15a for
Kuang-wu-ti; and HHSCC 6. la for Shun-ti.
s See SKC 2 (.yvei), p. 75, note 3 for a passage from the Hsien-ti chuan and de Crespigny (1970) for
the value of that document.
SC 8, p. 63, MH vol. II, p. 380.
For Wen-ti, see HSPC 4.4a, HFHD vol. I, p. 229; for Kuang-wu-ti, see HHSCC (treatise) 7.3a.
For the metal casket and its attached inscription, see pp. 91f above; for the interpretation of the
text, see Yen Shih-ku's (581-645) note to HSPC 99A.35b.
104 Divination, mythology and monarchy
destiny of kings, which will be considered below. But it did not enter into the
formalities of Kuang-wu-ti's accession, for the very good reason that in the
official view Wang Mang was an interloper in the appointed cycle of rulers,
and had no right either to rule or to abdicate. Perhaps the most conspicuous
example of the procedure for abdication was that of the last Han emperor,
whose instrument began by referring to the example of Yao and Shun in the
remote past.
(iv) The part played by an Empress Dowager
We are not concerned here with the general manner in which an empress or
empress dowager could predominate in matters of state,
but solely with the
part that she played in determining the imperial succession, whether as a direct
act of will on her part, as a formality wished upon her by senior statesmen, or
as an essential part of constitutional procedure. The first instance is seen in the
deliberate action of the Empress Dowager Lu, in enthroning two infants of her
own choice, successively.
It was on the initiative of the Empress Dowager Lu that, in the absence of a
son born to Hui-ti and his empress, the child of another woman of the palace
was nominated heir apparent;
on the death of Hui-ti he duly became
emperor, young as he was. This arrangement enabled the Empress Dowager
to take a leading part in affairs of state, and her position only came under
threat some four years later, when the young emperor realised the circumstan-
ces of his birth and conceived a hatred for her. It was in response to her
command that senior statesmen met to consider replacing him, and they
obliged her by accepting the decision that this step was necessary. In such
circumstances a second infant emperor was enthroned, in 184 BC;
it was in
the belief that neither he nor his brothers were true sons of Hui-ti that senior
officials of state felt entitled to take up arms to expel him and establish Liu
Heng (Wen-ti).
One of the measures taken by Liu Heng before finally agreeing to accept the
title of emperor was to inform his mother of the invitation proferred to him.
In the crisis of 74 BC it was through the agency of the Empress Dowager that
crucial decisions were taken, and it can only be concluded that by then her
position had acquired a recognised authority. Chao-ti's empress was a mere
fifteen years old at the time of the emperor's death; by a command issued in
her name senior officials were sent to summon Liu Ho to succeed to the
imperial throne; and it was to the Empress Dowager that Huo Kuang (d. 68
BC) and his colleagues brought their plea that he should be deposed, in the
89 SKC 2 (Wei), p.62. 9For this aspect of the subject, see Yang (1960).
"' SC 9, p.15, MH vol. II, p. 418; HSPC 3.lb, IJFHD vol. I, p. 191.
92 HSPC 3.3b, HFHDvol. I, p. 197; for the murder of this boy after Wen-ti's accession, see SC9,
p. 37, MH vol. II, p. 441. 93 HSPC 3.8a and 4.3b, HFHD vol.I, pp. 209, 227.
94 HSPC 4.2a, HFHD vol. I, p. 225.
The authority of the emperors of Ch'in and Han 105
interests of the dynasty. She had even been primed, as it would seem, to
interject an expression of horror at one point when the charges were being laid
against Liu Ho; it was by her command that both the final steps of deposal
were taken and a further commission was sent out to summon Liu Ping-i to
accede to the throne in his turn.
On the death of Ai-ti without a successor (1 BC), the Grand Empress
Dowager issued the edicts which prepared the way for the accession of Liu
Chi-tzu (P'ing ti).
Again, in AD 5, the edicts which entitled Wang Mang to
style himself regent and then acting emperor originated from the same
A number of the Later Han emperors died without a successor sometimes
while still in their infancy. In such circumstances the Empress Dowager was
able to determine who would follow as emperor; it was her edicts which
established him in his rightful position, while she herself retained a domina-
ting influence at court. The accessions of An-ti (AD 106), Chih-ti (145),
Huan-ti (146) and Ling-ti (167) were brought about in this way. In addition
Ch'ung-ti, the infant son of Shun-ti, owed his accession and his short reign
(144--5) to the intervention of an Empress Dowager.
( v) The imperial seal
From early times, the imperial seal came to constitute a material symbol of
authority which was lodged in the person of a rightly acclaimed emperor. To
distinguish it from the seals used by nobles and officials in the conduct of their
duties, the imperial seal was described as hsib rather than changb or yina. Its
special character is revealed in the term sacred instrument (shen eh 'i), which is
sometimes used in place of hsi.
The imperial seal featured in 180 BC in part of the ritual whereby the senior
counsellors of state persuaded Liu Heng to accept the title of emperor.
all due deference they presented him with the Son of Heaven's seal, and he
finally agreed to accept the burden of empire. One of the charges brought
against Liu Ho was misuse of this precious jewel.
When, in 74 BC, the
Empress Dowager had approved the suggestion that Liu Ho should be
9' HSPC 68.9b, !Ob. 96 HSPC 12.la, HFHD vol. III, p. 61; HSPC 97B.2lb.
97 HSPC 99A.25b and 29b, HFHD vol. III, pp. 219f and 233f.
98 See HHSCC 5.lb (for An-ti); 6.14a (for Ch'ung-ti); 6.15a (for Chih-ti); 7.la (for Huan-ti) and
8. la (for Ling-ti).
99 Thetermhsiis used for kings of the pre-imperial period (SC6, p. 20, MHvol. II, p. 123) and for
kings of the Han period (HSPC 19A.26a). For institutional usage, see Tu tuan, A.3b, and Han
chiu i (SPPYed.) A.la. For an example in which the term hsi features, see WW 1973.5, 26fand
1976.11, 8f, for ajade seal inscribed Huang-hou chih hsi and believed to have been the seal of the
Empress Lii. [Exceptionally the privilege of referring to his seal as a hsiwas conferred by Wang
Mang on a leader of the Hsiung-nu, only to be withdrawn at a later date; see H SPC 99B.1 lb
and 13a; HFHD vol. III, pp. 295, 301. For an early reference to the term shen ch'i, see Tao-te
ching 29.]
100 HSPC 4.3a-4a, HFHD vol. I, pp. 226-9.
HSPC 68.7b.
106 Divination, mythology and monarchy
deposed, Huo Kuang loosed the cord attached to his person, detached the seal
and delivered it to the Empress Dowager; as a final act in the drama it was
presented to Liu Ping-i.
As the first action to be taken after the proclama-
tion of the Hsin dynasty in AD 9, Wang Mang, together with all senior
officials of state, solemnly presented the Empress Dowager with a new seal, as
a symbolical means of expunging the dynastic title Han.
The seal duly played its part when An-ti came to the throne in AD 106, and
at the accession ofShun-ti (125) it was necessary to take possession of the seal
by force.
Finally, in the declaration made before the great congregation of
officials of state in 220, the king of Wei claimed that 'The Lord of Han has seen
fit to bestow upon me the Sacred Instrument (shen ch'z)'; and at what was
perhaps the climax of the ceremony of accession, he solemnly took delivery of
the seal.
From one account of the dynastic upheavals of this time,
learn that this had not been achieved without difficulty. When the king of
Wei's envoys called to request the seal, the Han empress angrily refused to
yield it. Finally she had the envoys admitted to her presence, roundly cursed
them,, and flung the seal to the ground whence they could retrieve it.
--- 1
09eligious Ceremonies
Newly acceded emperors paid attention to unseen powers in at least four
ways, by (a) divination, (b) purification, (c) notification to deceased ancestors,
and (d) the worship of Heaven and Earth.
(a) On two occasions we hear of the practice of divination before a new
incumbent agreed to accept the imperial charge. Still undecided whether he
should do so, Liu Heng had specialists use the age-old method of heating
bones and interpreting the resulting cracks;
and their answer proved to be
satisfactory enough. When the king of Wei consulted occult sources in AD
220, he may have employed other methods as well, including that of the
diviner's board;
once again the answer proved to be favourable.
(b) Immediately after Wen-ti's accession, senior officials were ordered to
purify the palace (ch'ing kung).
This action was explained by one early
commentator, viz. Ying Shao (c. 140 to before 204), as a normal statl,ltory
precaution against unexpected untoward incidents; but it is not mentioned
again as part of the formalities of accession. Possibly it is to be interpreted as
an act of ritual purification following the violence whereby the Lii family or
the last puppet emperor had been eliminated.
( c) The most noticeable and regular of the religious aspects of an accession
HSPC 68.lOb. 10' HSPC 99B.la, HFHD vol. III, p. 260.
104 HHSCC 5.2a and 6.2a. 10 SKC 2 (Wei), p. 75, note 3, citation from Hsien-ti chuan.
HHSCC 10B.l3a.
HSPC 4.2a, HFHD vol. I, p. 225.
108 SKC2 (Wei), p. 75, citation from the Hsien-ti chuan. In the process of divining, the king of Wei
is said to have 'retained the turtle' (shou kue1). This may refer to the indication of fortune on a
diviner's board, where a man's destiny was foretold by the appearance of a particular animal.
HSPC 4.4a, HFHD vol. I, p. 230.
The authority of the emperors of Ch'in and Han 107
were the services paid to the ancestors of the imperial house. This feature is of
obvious importance, in view of the need to maintain and prove continuity on
an hereditary basis.
J.here .. the f!rst9(the. took any
t,he f c)r due

It was not possible to include these features in the
-ceremonies which marked Kao-ti's accession. One of the first acts ofHui-ti's
reign was to order the erection of shrines in honour of Kao-ti in the
commanderies and kingdoms of the empire.
In urging Liu Heng to accept
nomination as emperor, the counsellors of state stressed the importance of
taking responsibility for the upkeep of Kao-ti's ancestral shrines; and in the
first year of his reign, Wen-ti duly presented himself there.
Chao-ti likewise
did so immediately after his accession, and the significance of maintaining
continuity in the services to these shrines is seen in the deposal of Liu Ho and
the accession of his successor.
'The Son of Heaven is the instrument for the
preservation of the ancestral shrines and for imposing unity on all within
the four seas', the ministers ofstate urged; and Liu Ho's failure to pay his
respects to Kao-ti's shrine showed his unfitness to rule as emperor. The
decision to depose him was to be formally announced to Kao-ti's shrine, 114
and the accession of Hsiian-ti concluded with his presentation there. Similar
acts of notification and acknowledgement were likewise performed by
Yiian-ti, Ch'eng-ti, Ai-ti, P'ing-ti and Kuang-wu-ti, immediately after their
The brazen casket, whose inscriptions formed the final piece of evidence
that Wang Mang was to rule as emperor, had in fact been 'discovered' in the
shrine dedicated to Kao-ti; ironically enough it was on that sacred site that
Wang Mang received indication of the abdication of the last of the Former
Han rulers.
Of the emperors of Later Han, Shun-ti, Huan-ti, and Ling-ti are
recorded as visiting the shrines of Kao-ti and Kuang Wu-ti.
abdication was made as part of an act of worship in the shrine of Kao-ti. 118
(d) More elaborate religious ceremonies are recorded for the accession of
Kuang-wu-ti and the first of the Wei emperors than for their predecessors.
That some of these rites were directed towards Heaven is consistent with the
major development that had affected the imperial cults of the Han dynasty. 119
SC 6, pp. 69f, MH vol. II, p. 195.
HSPC 2.3b, HFHD vol. I, p. 178; for the subsequent erection of shrines to Wen-ti, see HSPC
5.2b, HFHD vol. I, p. 308; for the proliferation of imperial shrines, see CC pp. l 79f, [and
chapter 13 below for the subsequent dynastic problems and economic considerations that led
to the abolition of some].
HSPC 4.5a, HFHD vol. I, p.231; and HSPC 7.la, b, HFHD vol.11, pp.151-2.
HSPC 68.7a. 114 HSPC 68.9b.
"'HSPC 9.2a, 10.2a, ll.2a and 12.la; HFHD vol.II, pp.302, 375 and vol.III, pp.18, 62;
HHSCC 1A.16b.
HSPC 99A.35b, HFHD vol. III, pp. 254--5.
HHSCC 6.2a, 7.lb and 8.lb. 118 SKC 2 (Wei), p. 62.
I.e., the final adoption of rites which were directed to Heaven rather than to the Five Powers;
see CC eh. 5.
108 Divination, mythology and monarchy
According to one account of Liu Hsiu's accession, an earthen altar was
constructed and a pyre laid.
The wood-smoke which in time ascended
therefrom conveyed the notification to Heaven that Liu Hsiu had become
emperor; and at the same time the message was sent to the six powers of the
universe (liu tsung)
and the whole host of holy spirits (shenc).
Little is recorded for the ceremonies which marked the accession of the
subsequent emperors of Later Han, and it is possible that this elaborate
procedure was performed solely in case of dynastic innovation. We next hear
of it in AD 220. According to one account,
'an altar was made after
Hsien-ti's declaration of abdication. The king of Wei mounted the altar and
proceeded to his due place, with the officials of state drawn up in attendance
on either side. When the ceremony was completed he descended from the
altar; and when he had observed that the conflagration had completed the
formalities he made his way back.' From a further record, in the lost
biography of Hsien-ti, we may conclude that the ceremony in question had
been that of accepting the deed of abdication, and that the fire was lit as an act
of worship to Heaven and Earth, the Five Sacred Mountains and the Four
Mighty Rivers.
IV. Statements of political theory
In addition to the evidence of the documents which were presented at the time
of an imperial accession and that of the formalities recorded for those
occasions, there are a few statements of a theoretical nature which refer to the
concept of imperial sovereignty. The writings of Tung Chung-shu, Pan Piao,
Wang Ch'ung and Wang Fu fortunately span the four centuries in question
and may be judged in the light of the changes already observed in the concept
and practice of the institution.
For Tung Chung-shu it seems advisable to restrict consideration to those
writings which the Han shu incorporated and ascribed to his brush.
views are expressed in an essay which is dated in the early part ofWu-ti's reign,
at a time when the Han dynasty saw itselfas the successorofCh'in and did not
seek its models in the kingdom of Chou. The monarch's duties were seen to
comprise three basic principles. He must ensure a conformity with the
ordinances of Heaven, so as to attain harmony between the three estates of
Heaven, Earth, and Man; he must so order the ways of humanity that man's
better nature is fulfilled; and he must establish the requisite social norms and
distinctions that will restrain human greed. Tung wrote as follows:
HHSCC 1A.15b and HHSCC (tr.) 7.lb.
Identified, according to the commentary to HHSCC (tr.) 7.lb, as sun, moon, star, mountain,
river and sea. For this question, see HSPC 25A.2a (notes), HSPC 25B.21a; HHSCC 1A.15b
(notes); MH vol. I, p. 61 note l; Laufer (1912), p. 120 note 1.
SKC 2 (Wei), p.62.
' Cited in note 3 to SKC 2 (Wei), p.62.
This decision is taken in view of the doubts which exist regarding the authenticity of parts of
the Ch'un-ch 'iu fan-lu.
The authority of the emperors of Ch 'in and Han 109
The ordinances (ling") of Heaven are termed destiny (mingb) which cannot be put into
operation except by a man of holy qualities. The fundamental substance of man is
termed human nature (hsingb), which cannot be brought to completion save by cultural
example and precept. Human desires are termed emotion (ch'ing) which cannot be
restrained save by regulations. This is why a man who is a true king pays great attention
on the one hand to receiving the intentions of Heaven so that he may conform with
destiny; and on the other hand he strives to educate his people intelligently, so that their
natures may be fulfilled; and he establishes the correct norms for their institutions,
distinguishing between the upper and the lower orders of humanity, so as to preclude
desire ... Man receives his destiny from Heaven and is thereby pre-eminently different
from other creatures.
The same theme appears elsewhere in Tung's writings,
but there is no
implication that the term t'ien ming refers to the appointment of a particular
dynasty or person to rule over mankind. Similarly there is no specific reference
to a Mandate of Heaven in the charges brought against Liu Ho in 74 BC. The
closest reference there is to blame Liu Ho for failing to attend at the shrine of
Kao-ti, thereby showing that he was unfit to 'receive the dispensation of
Heaven' (eh 'eng t 'ien hsu).
For the first full theoretical statement of the doctrine of the Mandate of
Heaven, we must turn to the famous essay of Pan Piao entitled On the
Mandate [or Destiny] of Kings (Wang ming lun). As presented in the Han
Pan Piao's essay is the work of a man of some twenty years of age who
had witnessed the rise and fall of the Hsin dynasty and the recent restoration
of Han, amid the claims of rival pretenders to power. The essay was allegedly
compiled in the hope of disabusing those who, like the pretenders Wei Ao or
Kung-sun Shu, could not see that ultimate dynastic success lay with the house
of Liu.
The essay sets out to re-affirm the cause of imperial unity; such unity had
last been forged by Wang Mang, under the plea that he had been entrusted
with the task by Heaven. It was therefore of no less importance to
Kuang-wu-ti's government to show that he too could command universal
loyalty throughout the empire in response to that divine cause.
Pan Piao argues that true sovereignty is the portion of the man appointed
thereto by Heaven, in the same way as other opportunities or functions are
presented to other mortals. The rule of the true sovereign is marked by the
blessing of suitable omens; it conforms with the cycle of the Five Phases, of
which Fire was then dominant; and it depends on communion with divine
powers. This principle remained identical, whatever the circumstances in
which temporal power had been attained in the past, be they peaceful or
violent; and continuity could be traced from Yao to members of the Liu
family. Just because true sovereignty is acquired only by divine help, it was a
vulgar misapprehension to believe that Kao-ti had seized it by force alone.
125 HSPC 56.15a. 12 HSPC 56.4b.
' HSPC 68.9b.
12 HSPC 100A.6aff, de Bary (1960), vol. I, pp.176f.
110 Divination, rnythology and monarchy
Only those who are capable of discharging the office appropriately can
claim the true honour of sovereignty; for the right vessel must be used for the
right task. While Pan Piao implies, but does not state, that Wang Mang was
not a suitable vessel, he names five signs whereby it could be shown that Kao-ti
had possessed the qualities that marked him out as being fit for the Mandate.
Other interesting points in the essay are seen in the allusion to the
abdication of Yao and Shun, and the place of the Sacred Instrument, or seal,
as a symbol of imperial rule. While the essay draws on concepts which, as have
been shown, originated long before the days of Pan Piao, its importance lies in
its systematic formulation of a theory of imperial sovereignty and its specific
dependence on the gift of Heaven. The tone and emphasis of the document
complement the practical procedures and theoretical claims of Wang Mang.
WangCh'unggrewup at a time when the dignity of the Han house had been
restored and the imperial system was being vindicated. There was no conflict
here with his belief in the natural force of reason, although his idea of fate, or
destiny, is somewhat difficult to reconcile with some of his cardinal principles
of spontaneous growth. On a number of occasions Wang Ch'ung alludes to
the Mandate of Heaven, citing the views of Confucius and Mencius. He also
refers to the concept in the two chapters
where he seeks to demonstrate that
the achievements of the Han period had been in no way inferior to those of
Chou. He argues that Han's superiority is clear on all counts, whether by
virtue of the appearance of good omens, the growth of cultural standards, or
the extension of territorial dominion. In describing the virtues and achieve-
ments of Han, Wang Ch'ung was ready to accept the statements of the
scholastics Uua) that kings, and in particular Kao-ti and Kuang-wu-ti, had
been in receipt of the Mandate. From other chapters
it seems that he
understood the term mingb in a sense not wholly different from Pan Piao. He
wrote that individuals are born with their own characteristics, or destiny, be it
for glory or shame; but Wang Ch'ung could not accept that a mandate to rule
was imparted to individuals after birth, by a deliberate act of Heavcn.
Finally attention should be paid to the views of Wang Fu, which may have
been coloured by his own personal circumstances and contemporary political
conditions. Coming from humble origins, Wang Fu wrote as a critic of the
world he saw around him, at a time when sovereigns had been reduced to
ineffectual positions and the Han government was under the domination of
powerful families at court. Wang Fu' s criticisms were directed equally against
the imperial house and other parts of society, and he protested that true
leadership derives not from heredity but from individual qualities and
conduct. He cites historical examples of kings, statesmen and generals to show
129 Lll 57 ('Hsiian Han') and 58 ('Hui kuo'), pp. 817f, 826f; Forke (1907-11), vol. II, pp. 192f,
" 0 Lll 12 (Ch'u ping'), pp.115f; Forke (1907--11), voL I, pp. 130f.
For the idea that certain men are born with a natural ability to be monarchs and that it is their
destiny to rule as such, see Lll 3 ('Ming lu'), pp. 21f, Forke, vol. I, pp. 146f; and 10 ('Ou hui'),
p. 99, Forke, vol. II, pp. 7f.
The authority of the emperors of Ch'in and Han 111
either the successes achieved by those who did not have the benefit of
hereditary succession, or the failure of those who had inherited such
At the times when Wang Fu and his predecessors were writing, the dynastic
history of the Chinese empires was still in its infancy; the major battles for the
continuity, interruption and restoration of unity had yet to be fought. From
Wang Fu's distrust of the infallibility of hereditary succession, we may look
forward to the more mature considerations of Chinese thinkers, grounded on
the experience of several centuries; to Han Yu's (768-824) vindication of
hereditary succession as against a system of choice; to the rejection by Ssu-ma
Kuang (1019--86), Ou-yang Hsiu (1007-72) and others of the attempts to force
the pattern of the correct line of succession (cheng t'ung) on historical
incidents; and to the compromise reached in Chu Hsi's (1130.-200) re-
assessment of the place of dynasties and emperors. Professor Biinger's
life-long studies and contributions have been directed to the root of these
discussions, concerning as they do the basis of the state and its authority. It is
with profound admiration of his work that the foregoing pages are offered to
an esteemed colleague, with hearty congratulations on the attainment of his
seventy-fifth birthday.
CFL 4 ('Lun jung'), pp. 32f, [Kamenarovic (1992), pp. 391].
The term K' an-yii and the choice of the
The term k'an-yu, which is seen in various texts from the Huai-nan-tzu
(completed by 139 BC) onwards, has been subject to various attempts at
explanation, and it is possible that its original meaning may have become
forgotten at a comparatively early stage. It will be noted below that from the
third or fourth century commentators were explaining k'an-yu as referring to
heaven and earth, but apparently without a clear comprehension of the term.
In much more recent times, k'an-yu-chia has been used as a synonym for
feng-shui-chia, meaning, specifically, experts in geomancy whose aim lay in
ensuring that a site on earth would be auspicious, either for occupation or for
burial. 1 Consideration of the evidence that is set out below, however, leads to
the conclusion that in its first occurrences k'an-yu was concerned with the
choice of an auspicious time rather than an auspicious site, and that decisions
taken by the k'an-yu-chia for this purpose involved the use of an instrument.
In this connection, attention should be paid to the valuable suggestion made
by Donald Harper that the term k'an-yu referred to two parts, 'canopy and
carriage', of an instrument that was comparable with the cosmic board (shihc).
From fragments and examples of these which have been studied, it may be
seen that they were in all probability used to regulate human behaviour so that
it would accord with the major situation prevailing in the cosmos, as measured
in terms of the movements of the sun and the earth.2
There is no shortage of evidence to show that the importance attached in
Shang-Yin times to the choice of a propitious time for embarking on major
religious observances or initiating large-scale projects persisted well into
imperial times. 3 It may be suggested now that, from denoting the two parts of
1 According to informants, in recent colloquial usage k'an-yii tends to be restricted to o.r
literary speakers. See also the preface (dated 1969) of Nan Huai-chin to a reprmt (Ta1pe1:
Chen-shan-mei 1970) of Ti-Ii t'ien-chi hui-yiian (original preface 1614).
2 For these see 'Fu-yang Shuang-ku-tui hsi Han Ju-yin hou mu fa-chiieh chien-pao'
15f, 19, figure 8; 25, figures 9-10; and plate III; Yen Tun-chieh (1978); Yin Ti-fei
(1978--9) and (1980--1); Cullen (1981); and Loewe (1979), pp. 75-80, 204-8 [and Kalmowskt
(1983)]. For the association of the shih with hsiian chi yii heng, see Cullen (1980-1 ), p. 39 and
Cullen and Farrer (1983).
3 The importance of choosing auspicious occasions may be seen in documents such as the 'Yiieh
ling' (as in the initial p'ien of the Lii shih ch'un-ch 'iu, chs. 1-12; Li-chi eh. 5; Huai-nan-tzu eh: 5);
and in the later and more practical Ssu min yiieh ling. For criticism that was kvelled. agamst
undue attention to the choice of a propitious time and against other mantle practices, see
The term K'an-yii and the choice of the moment 113
an instrument of a particular type that was used for this purpose, k 'an-yu came
to signify the instruments themselves, of various types. In an early occurrence,
in the Shih-chi (see passage 2 below), k'an-yu-chiais believed to refer to experts
in the use of these instruments, as distinct from specialists in other methods
(for example, calendrical science, consultation of almanacs, or wu hsing
theory), whose services were likewise invoked in order to determine the
auspicious nature of a particular moment in time.
The following notes will attempt to review the early evidence for the use and
meaning of the term k'an-yii.
I Huai-nan-tzu 3.29a
The passage is concerned with the movements of the heavenly bodies and their
relation to the sequences of the months. Certain categories of days are
specified as being suitable or unsuitable for activities, and the relevant part of
the text reads: 00'. El iiJ t.J :f s h11H-J 1f -Jio J1$ .
In view of the questions that have been raised regarding the validity of the
text in the immediately preceding part of the work (see Wang Yin-chih's
(1766--1834) note, cited by Liu Wen-tien), we may well be justified, as Harper
suggests, in following a reading given in the commentary to the Wen-hsiian,
which omits the character yind. Harper's rendering of the passage reads: 'On
the Canopy and the Chassis the masculine is slowly moved, thereby perceiving
the feminine.' He explains the passage as meaning that the circular part of the
k'an-yu instrument was being rotated, until it was properly aligned with the
square board (i.e., the feminine part of the board). An alternative rendering,
following the punctuation of Kusuyama reads: 'On yen days it is not suitable
to undertake general activities. The k'an-yu moves slowly, the male thereby
comprehending the female.'
The 'male' and the 'female' may well refer to
parts of the instrument, as Harper suggests, but it is perhaps possible to
interpret the action as that of a male piece or pin (hsiung) fitting into a female
socket (tz'ub). Whatever the precise meaning and connotation of the passage
may be, there is no doubt that its context is that of regulating activities in order
to conform with the appropriate sequence of time.
Lun-heng as cited under passage VIII below, and in the followingp'ien of that work: 68 ('Ssu
hui'), pp. 964f, Forke (1907-11), vol. II pp. 376f; 70 ('Chijih'), pp. 985f, Forke, vol. II pp. 393f;
71 ('Pu shih'), pp. 984f, Forke, vol. I pp. 182f; and 72 ('Pien sui') pp. 1004f, Forke, vol. I
pp. 525f. 4 See Harper (1978-9), p. 9 note 53; Wen-hsiian 7.2b.
' Kusuyama, vol. I, pp. 190--1. In his note, Kusuyama suggests that the k'an-yii chia pronounced
on questions of fortune after taking into account the movements of the heavenly bodies and the
situation on earth, on the basis of the sexagenary cycle. See also the notes to the passage by
Ch'ien T'ang (1735-90) (rpt. HNT, ed. Liu Wen-tien, 98aff). [No precise interpretation is
offered here for the termyenjih (readyehjih according to Morohashi 3025.35 and 14330.15; see
GSR 616c). In his note to the immediately preceding passage in HNT, Wang Yin-chih
(1766--834), who believes the text to be corrupt, cites Chia Kung-yen's (jl. 650) sub-commentary
to Cheng Hsiian's (127-200) note to Chou Ii, SSC 25. la. The term possibly refers to days when
the moon is apparently engaged in retrograde movement; see SCC vol. III, pp. 214, 219.]
114 Divination, mythology and monarchy
II Shih-chi 127.14
In this addendum to the chapter, Ch'u Shao-sun describes an incident in
which Wu-ti consulted experts of a number of types in order to determine
whether or not a particular day would be suitable for taking a wife. The
specialists included masters of wu hsing, (i._e., the use of
almanacs), and ts'ung-ch'en, as well as experts m
omy, and T'ai-i.
There can be no question in the is
recounted in this passage, the purpose of consulting experts m k an-yu was to
determine a suitable time, rather than a suitable place.
ill Han shu 87A.9b
The term k'an-yii occurs in a somewhat difficult passage of Yang Hsiung's
'Rhapsody of the Sweet Springs' ('Kan ch'iian fu').
This[u was presented by
way ofcriticism to Ch'eng-ti (reigned33-7
to perform religious services in hope of birth of an hetr'.
were to be carried out at Kan-ch'iian and Fen-ym, mhonour of the de1t1es T al-1
and Hou-t'u, as had been the customary practice until 31 BC. In that year,
however, they were discontinued in favourof the state cult.s that
to Heaven;
a reversion to the traditional objects of worship and their sites took
place in 14 BC. It is to an occasion in these years, between 14 an? 7 BC that Yang
Hsiung'sfu refers. The introductory passage relates how officials were ordered
to determine a fortunate day for the ceremonies and to ensure that the
conjunction of the heavenly bodies would be appropriate. The text continues:
'He summoned Chao-yao and T'ai-yin, he secretly stationed Kou-ch'en to
oversee the arms [or armed men]; he charged k'an-yii with pi-lei.'
T'ai-yin and Kou-ch'en may all be identified as stars or the spmts who
them and some of these names recur below (see under passage VIII).
While pi-lei ma; well be interpreted in a military sense, as it is by it is
also possible that it may denote a con.stellation.
.As in the .
cited, k'an-yii is again concerned with the choice of a ttme.
IV Han shu 30.70
The bibliographical list of the Han shu includes a class of writings as
shu shu or studies of the occult, which is divided into six sub-categones. Each
sub-ca{egory bears its own title, and there are 31 entries under wu hsing,
For Ch'u Shao-sun (second half of the first century BC), see MH vol. I, pp. and Pokora
(1981) who suggests the dates 104 to 30 BC for Ch'u's life. For the chien-ch'u ch1a, see chapter
10 Ts'ung-ch'en and T'ai-i form part of the titles of works included in HSPC 30.65a, 70b
and 71a. For the state cults addressed to T'ai-i, see CC p. 169.
Knechtges (1976), pp. 46, 128 note 21. CC p. 178. . .
Knechtges 1976), p. 46 renders: 'He assigns K'an and Yii to. the P11s fourteenth
the 28 Lunar Lodges, sometimes known as Tung pi. Both m this m Chang Heng s
(78-139) 'Ssu hsiian fu' the term pi-lei seems to be used as an extenSion of pz (HHSCC 59.23a).
The term K'an-yii and the choice of the moment
amounting to a total of 652 chiian.
Some of these entries probably denote
manuals for the use of instruments of various types, for example, Hsien men
other entries are for works which were concerned with the regulation,
apportionment, or calculation of time and its cosmic considerations. The list
also includes an entry for K'an-yii chin k'uei in 14 chiian.
In his comment to this sub-category ofwritings
the author of the treatise
explains that they derive from calendrical science and considerations of the wu
hsing cycle. He adds that there are those who make use of these works or their
principles in order to determine the auspicious and the inauspicious, thereby
giving rise to confusion. It may also be noted that the title of a book which
probably concerned the choice of a propitious situation rather than an
auspicious time (Kung-che ti-hsing in 20 chiian) is entered in a different
sub-category of shu shu, i.e., hsing fa. 13
V K'an-yii shu
A fragment of the Feng-su preserved in the T'ai-p'ing yii-lan, 14 cites a
short passage from a work that is entitled K'an-yii shu, in the context of
entertainments and food. This reads: (A quarrel will
certainly break out if guests are assembled on the first day of [the earlier
months, or the months specified above?]).
VI Chou Ii (SSC) 26.20aft"1
The duties of the Pao chang shih are specified as being concerned with noting
the movements of the heavenly bodies, with a view to observing the changes
that occur below the heavens and distinguishing that which is auspicious and
inauspicious. This official was also responsible for drawing up territorial
divisions that corresponded with the apportionment of the stars, thereby
observing the incidence of calamity and good fortune. In commenting on the
passage (26.21a), Cheng Hsiian (127-200) wrote that 'although the k'an-yii
possesses measured provisions whereby commanderies and kingdoms be
entered, these are not ancient calculations' ( 1J:
This comment can be interpreted as meaning that the k'an-yii instruments
of Cheng Hsiian's time made provisions for adjustment, so as to fit the
contemporary territorial administrative divisions of the Han empire. If this
interpretation is correct, it implies that the instruments were designed so as to
link astronomical phenomena directly with incidents that occurred in different
regions on earth, and the association of time and place is of considerable
This figure is given in HSPC 30. 72b; the actual total of chilan listed is 654.
HSPC 30.72a.
HSPC 30.73a. " HSPC 30.77b.
TPYL 849.5B; Centre Franco-chinois d'etudes sinologiques, Index du Fong su t'ong yi (Peking,
1943), p. 85. Avoidance of holding assemblies on shang-shuo days is also mentioned in LH 72
('Pien-sui'), p. 1010, Forke (1907-11), vol. I, p. 530. The expression is also seen on contempor-
ary calendars from Hong Kong. " Biot (1851), vol. II, p. 114.
116 Divination, mythology and monarchy
importance. Cheng Hsiian's note of the difference with ancient practice may
mean either that older instruments carried no such device; or that while they
did make provision for territorial divisions, these were different from those
that were valid for the days of the Han empire. In any case, the passage is
perhaps the earliest reference to a connection that can be drawn between the
term k'an-yu and considerations of place.
VII Hou Han shu 76.6b
Wang Ching is best known as an engineer who had specialised in problems of
water control. His training had included a study of the Changes (we are not
told which school), and he had acquired an interest in astronomy and occult
matters (shu shu). His last appointment as governor of Lu-chiang com-
mandery was dated in AD 83, and he died while holding that office.
Wang Ching had noticed examples of confusion or discrepancy in the
records of questions put to the turtle shells and the yarrow stalks found in the
classical texts. He therefore set about consulting a wide range of documents
that concerned the occult; or prohibitions imposed on the choice of the site of
a residence or tomb; or works that belonged to k'an-yu and considerations of
time (Jih hsiang). He collected those parts that applied to practical affairs in a
book that was entitled Ta yen hsuan chi. It may be noted that a clear
distinction is drawn in this passage between works that concerned a choice of
place and those which, including k'an-yu, concerned time.
VIII Lun-heng 70 ('Chi jib'), pp. 985f1
In his usual rationalist, scientific frame of mind, Wang Ch'ung addresses
himself in several passages to exposing the fallacies inherent in the belief that
certain days are fortunate and others unfortunate for certain activities.
argues that such a belief and its subsequent implementation in practice rests
on inconsistencies, for there is no reason why some activities only should be
subject to such qualifications and others should be exempt. He observes that
no causal connection can be traced in the choice of avoidance on certain days.
He points out the absurdity of believing that, although human beings may
have no reason to object to certain types of activity on particular days, they
should refrain from such practices because superior beings or spirits object to
It is within this context that Wang Ch'ung refers to K'an-yu Ii. He writes:
As for the K'an-yii, the various holy spirits (shenc) that are upon this cycle are not
uniform. The Sages do not mention them; our own body of teachers do not transmit
accounts of them and they hardly possess any reality. The order of heaven is difficult to
1 Forke (1907--11 ), vol. II, pp. 393f.
17 In addition to the passage under reference, see LH 68 ('Ssu hui') and 72 ('Pien-sui'), Forke
(1907---11), vol. II, pp. 376f and vol.I, pp. 525f.
The term K'an-yii and the choice of the moment 117
understand; but were it to be granted that it does possess these spirits, what happiness
would follow from avoiding actions on days which they use for activities, and what
misfortune would follow from not avoiding such days? Those who are kings initiate
actions on chia tzu days, and members of the civil population likewise make use of such
days; but when those who are kings hear of this they do not punish them with the laws.
Now, if those who are kings are not angry, if the civil population does not avoid days
when they are themselves active, why should the holy spirits of heaven alone be ready
to blame them for doing so?"
The passage suggests (p. 993) that the K'an-yu Ii, whether in the form of an
instrument or a written calendar, marked days on which certain actions were
to be avoided, with the appropriate deities being shown. Almost the same
reasoning is put forward by Wang Fu (c. 90-165), again with critical intent, in
a passage where some of the holy spirits of the stars are specified.
Two of
these, T'ai yin and Kou ch'en also appear in the passage that is cited above
from Yang Hsiung's 'Kan-ch'iian fu'. Wang Ch'ung's point that the spirits
shown on the K'an-yii Ii are not mentioned in literature perhaps suggests that
they were featured more on a popular than on an official level.
IX Wei shu 91 (biog. 79) pp. 1955f.
Yin Shao, who enjoyed the study of Yin-Yang and the occult sciences (shu
shu), rose to be an academician during the first half of the fifth century. In 458
he presented to the throne a work entitled Ssu hsii k'an-yii together with a long
memorial. He recounted how he had met the great scholar Ch'eng-kung
Hsing, who lived the life of a recluse, but later introduced him to the Sramana
Shih-t'an Ying and a master named Fa Mu. These last two had explained the
mysteries of mathematics to Yin Shao together with a number of other
subjects. In addition, Fa Mu taught Yin Shao the fundamental principles of
the universe and the rule ofYin--Yang, on the basis of a text entitled Huang-ti
ssu hsu ching. This book had been annotated by Fa Mu's teacher Ho Kung. It
comprised 324 sections ( changb) in 36 chiian which were arranged in four parts
of equal length, concerning respectively: (i) the pairing of Yin and Yang; (ii)
seasonal energies, with their growth and decay, their fortune and misfortune;
(iii) movements of the sun, moon, and heavenly bodies, and their conjunctions
and relationships; and (iv) the rise and fall of fortune and misfortune, of
reward and punishment in the cycles of time.
Fa Mu gave this book to Yin Shao, who lived in seclusion for forty-five
years, observing in turn the 'eight meetings of the contemporary popular
k'an-yu' ( ). In so long a period of time he could not
possibly record in entirety all the errors of transmission or the bans imposed
for considerations of good or bad fortune. Sometimes he had found that on
" LH 70 ('Chi-jih'), pp. 993f, Forke (1907-11), vol. II, p. 401.
1 CFL 25 ('Pu lieh'), p. 299.
For a shorter account, see Pei shih 91 (biog. 79), pp. 1955f.
118 Divination, mythology and monarchy
days regarded as good an evil occurrence had taken place; or that while good
fortune had been expected, a calamity had occurred. Finally Yin Shao had
selected the main points of the Ssu hsii ching together with actual experience of
good and bad fortune, and he had assembled this information in one volume
(chiian). This book was intended to apply to all levels of society, from emperor
to commoner, irrespective of rank or status; and in due course it was presented
to the throne. The Wei shu concludes the biography of Yin Shao by observing
that thereafter his Ssu hsii k'an-yii circulated widely.
X Sui shu 34 pp. 1035-6f
A long section of the bibliographic treatise in the Sui shu is described under the
general term wu hsing, with a total of 272 entries amounting to 1022 chiian.
The entries include items that derive from a number of methods used in
divination and the consultation of oracles, and some of these may be identified
as manuals for the use of certain types of instrument or shih.
The work that is
mentioned immediately above (i.e., the Ssu hsii k'an-yu by Yin Shao) takes its
place among ten titles which include the term k'an-yii, written, however, with
yiie in place of yii. Possibly this change may be taken as an indication that by
the time when the list was being compiled the significance of the term k'an-yii
had been forgotten, with yiic being replaced by a homophone.
It has been suggested that some of the fragments of inscribed wood that were
found a few years ago at Mo-tsui-tzu, Wu-wei (Kansu) derived from k'an-yii
and its use, but this identification should now be revised.
Following the
discovery of the fragments at Mo-tsui-tzu, a far larger cache of mantic
documents has been found at Shui-hu-ti, Yiin-meng (Hupei), and some of
these have shed considerable light on the chien-ch 'u system of consulting the
oracle. In view of the similarity of certain features of the documents, it may
now be suggested that the fragments from Mo-tsui-tzu formed part of a
table or almanac that derived from chien-ch'u rather than from k'an-yii
In the notes to at least two passages of the Han shu and Hou Han shu, Hsii
Shen (c. 55-c. 149) is cited as explaining k'an as the order of heaven (t'ien tao)
andyii as the order of earth (ti tao).
In commenting on the termk'an-yii at its
appearance in Yang Hsiung's 'Kan-chiian fu', Chang Yen (third or fourth
century) wrote that it was a general designation for heaven and earth. Yen
Shih-ku (581-645) expressed his agreement with this view. Meng K'ang (jl.
180--260), however, had departed from this type of explanation, identifying
k'an-yii as the name of a holy spirit who had compiled maps and writings on
This work is no longer extant.
See entries in SS 34 (tr. 29), pp. 1028, 1032.
See GSR pp. 40, 42 (821 and 89j), for the ancient reading two.
Wu-wei Han chien, p. 139.
' HSPC 30.70a and HHSCC 76.7a; Shuo-wen 13B.22a.
The term K'an-yii and the choice of the moment 119
In a recent statement, Nan Huai-chin has traced the attention
to k'an-yii back to the Spring and Autumn period, and he observes the
contemporary use of the expression to mean geomancy.
The conclusion that may be tentatively drawn from the evidence of the
passages cited above and from a consideration of a few discoveries from Han
tombs may be summarised as follows.
By the time of the Huai-nan-tzu, the tenn k'an-yii was being used as a
general term to denote a variety of instruments used to show the relationship
of the earth and the heavenly bodies, and to prognosticate the good or bad
results that would follow actions undertaken at a specified time. The term had
started as a description of the two parts of a particular instrument, of which
one represented the heavens (k'an: 'the canopy') and one the earth (yii: 'the
chassis'). Of the instruments that were included under the general description
of the term k'an-yii, the cosmic board (shihc) has already attracted consider-
able attention, and reference should now be made to some other types, of
which examples have been found in tomb Ml at Shuang-ku-tui, Fu-yang
This has been identified as the tomb of the second marquis of
Ju-yin, Hsia-hou Tsao, who died in 165 BC.
One of these instruments, which has been entitled Erh-shih-pa hsiu yiian
p'an, is described as a circular board showing the 28 Lodges, with a series of
apparently 365 sockets or grooves cut into the edge of the board, but without
penetrating it completely. It may be asked whether those grooves had been
prepared for the insertion of a pin or peg at successive positions round the
circle; and whether it was such an action that was described as 'the k'an-yii
moving slowly, the male thereby comprehending the female'.
A second instrument, which has been entitled T'ai i chiu kung chan p'an
consists, like the shih, of a square plate within which a second plate, that was
circular and could rotate, had been fitted. The inscriptions on the two plates
refer to the sequence of the seasons, the passage of days, and the likely
outcome that would occur in certain circumstances. There is also a clear
reference to the heavenly bodies, or their guiding spirits, as may be seen in the
prominent position given to the characters Chao-yao. It is possible that Yang
Hsiung's reference to the constellations, and to 'charging k'an-yu with pi-lei'
may be understood with reference to an instrument that possessed these or
similar features.
In this way it seems that the k'an-yii, i.e., instruments that represented
heaven and earth, and whose moving parts could be described as 'canopy and
chassis', enabled an operator to consult the stars regarding the choice of a time
for activity, and to prognosticate its probable outcome. In so far as the choice
of time was intimately linked with considerations of space, it became
necessary to take territorial arrangements into account; for this reason some
HSPC 87A.9b.
As cited in note 1 above.
See 'Fu-yang Shuang-ku-tui hsi Han Ju-yin hou mu fa-chiieh chien-pao', as cited in note 2
above, and Chao T'ieh-hua (1980), plate 25. 2 HSPC 16.6a.
120 Divination, mythology and monarchy
of the later instruments, which were known to Cheng Hsiian (127-200), were
fitted with markings that showed the contemporary administrative divisions
of the empire. .
In the course of the centuries, these early instruments underwent consider-
able development, not least the incorporation of the magnetic needle, from
perhaps the eleventh century on.
Possibly this development may have
exercised a radical influence on the whole process of using the instruments for
consulting the oracle. For the new device would tend to_
attention of both the inquirer and the operator on quest10ns of s1tuat10n
rather than time, as the needle could be seen to be pointing in a recognisable
direction on earth. But while the emphasis of feng-shui lay in the choice of
place rather than time, it made use of an instrument tha_t was the _direct
descendant of the k'an-yii of Han times, and that term persisted, despite the
change of design, intention, and usage. For these reasons, the term k'an-yil
chia continued to be used, and became synonymous with the term feng-shui
Attention may perhaps be drawn to the vigorous criticism voiced by Ssu-ma
in 1084 against the funerary practices of his day, which included, to his_mmd, excessive
consideration of geomancy. It is tempting to speculate that the recent mtroduct10n of the
magnetic needle into the instruments and their subsequent use to excess may have been one
reason for Ssu-ma Kuang's outspoken views; see Ssu-ma Kuang 'Tsang lun'; also de Groot
(1892-910), vol. III, pp. I 02 lf, and SCC vol. III, pp. 3 !0f. At a slightly date, Chang Tsai
(1020-77) had, in a single sentence, dismissedfeng-shui as bemg meamngless; see Chang tzu
ch'iian shu 8.6a ('Sang chi').
Imperial sovereignty: Tung Chung-shu's
contribution and his predecessors
Over four centuries separated the accession of the first Ch'in emperor, in 221
BC, and the abdication of Han Hsien-ti in favour of the king of Wei in AD
220. In the the concept of imperial sovereignty had changed
fundamenta.lly Ill religious, intellectual and practical terms. Imperial rule
based on might had given way to the need to support a claim to rule with
intellectual sanctions. With the renewed worship ofT'ien, the old idea of the
t'ien-ming, or Mandate of Heaven, had been revived with some force. While
the early emperors of Han had been ready to adopt Ch'in's institutions, from
-- the start of the Later Han period (AD 25) at least emperors and statesmen
lOoked to the principles and practices of Chou as ideals that they should
emulate. In the procedures whereby the succession of emperors was accom-
plished, a new stress was being laid on correct ceremony, on the formulation of
documents and on the material symbols of majesty. 1
At the same time the importance of the emperor had also been subject to
It was apparent that the choice of incumbent for this position was
hkely to be ofless significance than the compelling need to see that it was filled.
powers of the emperor had been reduced, his significance
as a rehg10us furict!oiiarynad beep While his existence as the
"fouiitain:nead from wh-!Ch- ail devolved was essential to the
operation of the empire, the powers that he had once exercised had fallen into
other hands. In political terms, emperors could be reduced to nonentities; in

The process of change was due both to dynastic and political circumstance
and to intellectual development. QEJ)'_ __ w.ere_s1rnng
... .. ..
ts: of imperial consorts had risen to eminence and
had tlieir powers iii-1n.e .. scant

qt\ en
fl1emselves c;op.sqri, a.nd.on. the .. settlement.Oisuc.h
. !.h.e ha9 ....s.ometimes In such
For these developments, see chapter 4 above.
122 Divination, mythology and monarchy
circumstances it would take an exceptionally strong emperor to assert his
personality and influence events as forcibly as, for example, the first Ch'in
emperor. All too often an infant or a weakling would sit enthroned at the head
of the empire while major decisions were being taken elsewhere.
The changes that took place in religious and intellectual terms are often
attributed to the influence of Tung Chung-shu, and while it may in general be
agreed that much of the achievement was his, it is proposed here to examine
some of the antecedents to his work. In some respects Tung was an innovator;
in others he may have been drawing on the work of his predecessors and
lending it renewed force.
But wherever the precise credit for originality is due,
the measure of those achievements is clear. .. !<Kercise . ...Qf
imperial was regarded asan integral past of a
the powers of T'ien were recondled with the rhythm of the Five Phases (wu
hsing) and both were seen to affect the destiny of kings. The imperial regime of
Han was shown to be a descendant of the regimes of the remote past; a new
force was given to the teachings of Confucius as a paragon master of ethical
and political values. In addition, Tung Chung-shu's writings and his reputa-
tion carried two further results. In the :fi.!tP.la.cJ< .
mit1ct: a.pr3:c;ti.9e. ()f fa,ste11ii,ig .o:ri .f uen t of 11at.ll:r,e, qg
untowa.rd di11asters..as a .. oLcriticising the .secondly, Tung's
to literature, history and education exerted a formative influence on
the training of Chinese officials, for better or for worse.
Tung Chung-shu, whose dates are usually given as c. 179 to c. 104, did not
attain a high position as a statesman, nor did he enjoy a particularly successful
career as an official. Shortly after the accession of Wu-ti (141 BC), he was
appointed chancellor (Hsiang) of the small kingdom of Chiang-tu (Chiang-su
province), but he was later reduced to the position of counsellor of the palace
(Chung ta-ju). In 135, fires broke out in one of the provincial shrines dedicated
to the memory ofKao-ti, first Han emperor; Tung's ..
strange phenomena and. omens c9nsiderable-cri ticism.i 11-l)Q IJi< .. waa.
... brough:t"up oil a .. charge;whose nahire is iioi
death penalty.An l!nperial pardon saved'fiimTrom tliis"t>uriishment. Later he
focuried the enmity of Kung-sun Hung, who came to hold the highest post in
the imperial government (i.e., Ch'eng-hsiang) from 124 to 121. Tung Chung-
shu was relegated to be chancellor in the distant kingdom of Chiao-hsi
(Shan-tung province) and died after retirement due to illness.
By contrast with the short biographical account that is included in the
Shih-chi, the Han shu incorporates the text of three memorials which Tung
submitted to the throne in response to imperial edicts.
These specifically
For general assessments of Tung Chung-shu, see Fung Yu-Ian (1937-53), vol. IJ, pp. 7f; Chan
Wing-tsit (1963), pp. 27lf; and Hsiao Kung-chuan (1979), pp. 484f. Specialist studies include
Seufert (1922), Tain (1974) and Helliwell (1981).
' For the bare facts of Tung's life, see SC 121, p. 26, HSPC 56.la and 19a; for an attempted
chronology, see Su Yii's edition of the Ch'un-ch'iufanlu; for the record of the Fires in Kao-ti's
shrines, see HSPC 6.3b, HFHD vol. II, p. 33.
Imperial sovereignty 123
requested advice on the nature of dynastic government and the means of
ensuring a successful and prosperous dispensation on earth. The three
memorials form the basic evidence for Tung Chung-shu's political philos-
ophy, and they may be accepted as possessing a higher degree of authenticity
than the more voluminous material that is collected in the Ch'un-ch'iufan-lu. 5
!9Jf!Jrn in
of reducing the between rich and poor; audit.refers
tO q( rites to induce rain to fall at ti111es of drought, or
briiig a downpour to an en din till1es of excess. Irutddition, the dtcs a
considerable number of the interpretations which
of strange !he biographies,
_Tung after hi ..

In submitting his answers to the questions raised by Wu-ti, Tung Chung-
shu was able to look back on over half a century in which the Han emperors
had presided over the destinies of China; and his answers were of such a nature
that they affected the ideas of imperial sovereignty for some centuries. In
general it is difficult to answer the question of how far Tung's ideas were
original. The nature and extent of the sources is such that they are more
informative for the first than for the second century BC and in addition it is
likely that political ideas were relatively unsophisticated before Tung's time.
Certainly there were moments of dynastic history that could have been
expected to raise the whole question of the appropriate form of government
for China (for example, the crisis that accompanied the rise and fall of the
The texts of the three edicts, requesting advice from officials, and ofTung's replies, are given in
HSPC 56. lb, Sb and 13b. These are dated at 134 BC in Ch'ien Han chi 11.1 band at 140 BC in
Tzu-chih t'ung-chien 17, pp. 549f; they may have been related to the orders given in 141and136
BC that concerned the qualifications required by candidates for office and the means of
education. However, there may be reason to doubt whether the three edicts and memorials were
composed at so early a date. The tone of the edicts hardly suits the sixteen-year-old newly
acceded emperor and it does not reflect the opinion or optimism of the leading statesmen of the
day. It is also possible that the criticism implied in the documents was more appropriate for the
later days ofWu-ti's reign than for these early stages. There may be grounds for believing that
the death of the Empress Dowager Tou, in the fifth month of 135 BC, allowed some
consideration to be given to a change of policies and a re-assessment of political motives, as is
invited in the edicts.
In the second of the memorials (HSPC 56.l lb), there is a reference to the loyal submission of
K'ang-chii and Yeh-lang; it was only in the reign of Ch'engti (33-7 BC), or possibly Yiian-ti
(4933 BC) that K'ang-chii first took formal steps to adhere to Han; see Hulsewe (1979a)
pp. 123f, note 298, for consideration of the evidence and the reference to K'ang-chii hy Ssu-ma
Hsiang-ju (c. 179 ll 7 BC). Yeh-lang made over to Han in 111 BC (HSPC 95.4b).
' The authenticity of all or parts of this work has been brought into question, on philological
grounds. In addition, the content of some of the chapters is not entirely consistent with what is
known of Tung from other references or with contemporary intellectual development; [see the
entry for Ch'un-ch'iufan-lu in Loewe (1993)].
See HSPC24aJ6a, Swann (1950), pp. 177ffor Tung's stand on the question ofland tenure. His
performance of rites to induce or to prevent rainfall is mentioned in HSPC 56. l 9b; see also
Ch'un-ch'iu fan-lu 74 ('Ch'iu yii') and 75 ('Chih yii'), and chapter 7 below. For Tung's
interpretation of omens, see, for example, HSPC 27A.2a, 6a; for his refraining from such
interpretation, see HSPC 56.20a.
124 Divination, mythology and monarchy
Empress Lu), but documents which reflect arguments raised on this subject at
that time do not survive.
In the early decades of the Han dynasty, the idea and practice
the kings of Chou and of the moral leacfcrship that they were expected to exert.

wlien compareci with those qUhe ,lcings of the Warnng
demonstrably short-lived anci it had beenbroughtdown. by .. nf

manipuiation ofill1perl'ai in Qf the ggreedJgiaUies
to the.house of Liu, and the accession of Liu Heng as Wen-ti in 180 h<1,sf.11Qt,been,
controversy and challenge; A
fol rebellion against the central government ,t_ook l
.. In such circumstances if could well be asked whether the house of Han could
expect to survive with greater success than Ch'in .. If
to ensure th,e continuity.Qf their line, they would need to

need to show that there were greater mornt.andjntellec.tual
grounds, why Han had a better.charic;e t_hanjts predeq;:ss9r. l:faw.eY.er,.ih.e.
establishment of such a claim involved some difficulty .or even contradiction.
f(Tt was10 fie siiown despite th.e recent' shortcomings
an .imperial form ofgoveinment could be established for al.l
also becle11rly explainec.fwh.y Ch'in had failed to acliieve this
Han could. hope to do so. . ..
were all too obvTuus. Han, no less than Ch'in,
of milihfryTare:e.- 13:'L
sanC11oi1S.asct115eato'ilieCiispet1sation gfJht;.ki11gs.Q(g.h911:. Han had adopted,
Haii'ifp'fafa]'fJieasurestopreserveTaw'andorder were to
markedly,less of
emperors.had l.nhei-iied. and elaborated observances to the same gods that had
1Iieserv1ces 0rcli'111.-r ------ ----------------
to lend suppoiito. the .. Jhe....
Shih:chTanailaiishu sfiOw some recognition of the problem, there is little
.evidenceofal.l"obiectiveanalysis ohhe issues, such as had already i!ven
- "" ,- - ' -" -., >- _,__., J_,,,...,.. & ,,. , = ,- """ - ''"' - '"'-"" '' '"" '' -
political in the .. .. -
7 The principal changes in the administration concerned provincial government, with the
institution of hereditary nobilities and kingdoms. For the provisions of Han law, see Hulsewe
(1955). For the state cults of Ch'in and Han, see CC eh. 5 and Loewe (1982), eh. 12.
Imperial sovereignty 125
recorded in the histories derived more from the. practical needs of the
in- a.
evolve politicaLg1jp,g:Ples. ... , -- ---
. - iw a.
imate means of in.a.m.anver ..
His success drew on the application of wu hsing
theory-to-human institutions, and on the acceptance of ethical norms praised
by Shu-sun T'ung (fi. 200 BC) and Lu Chia (c. 228 to c. 140 BC). Before his
time several voices had drawn attention to the practical reasons for Ch'in's
failure; Tung Chung-shu emphasised the moral shortcomings of that regime.

had drawn on,..r]1.gi!].yIJ-!a[y,llletaphysics and on mythology to explain the
growtlioThuman culture and the place of government within the cosmic
. estabiisiledhis own cosmic the
intellectual authority and ethical precepts of a highly acclaimed teacher,
Confucius, with a new degree of emphasis.
In the following pages attention will first be paid to some of the intellectual
characteristics of the age immediately preceding Tung Chung-shu; thereafter
we shall proceed to examine some aspects of his own contributions.
The Five Phases
In perhaps the fourth or the third century BC an important advance had been
made in Chinese thought by the combination of the theory of the Five Phases
(wu hsing) with the concept of Yin-Yang. This theory was basically an attempt
to explain the operation of the observed world of nature according to a
cyclical scheme of stages.
As far as may be told, the first application of the
theory to man-made institutions may be traced to a famous passage in the
Spring and Autumn Annals of Mr Lu (Lu shih eh 'un-ch'iu) which was compiled
some two to three decades before the establishment of the Ch'in empire in 221
BC. The passage is of the utmost significance to the growth of concepts of
imperial sovereignty. By linking the theory of the Five Phases with the active
participation of Heaven in human destinies and with the symbolic importance
of omens, the writer was expressing ideas which do not seem to have been
taken up again in earnest until the time of Tung Chung-shu. In accepting that
human institutions are subject to universal laws of nature, the contributor to
the Lu shih ch'un-ch'iu was suggesting a new context within which temporal
power should be placed, and he was separating its rise and decline from the
sole arbitration of human will-power and force. The passage runs as follows:
Whenever a sovereign or king is about to rise to power, Heaven will certainly manifest
a favourable sign to mankind in advance. At the time of the Yellow Emperor, Heaven
had displayed creatures of the earth, such as worms, beforehand. The Yellow Emperor
8 For WU hsing and Yang, see sec vol. II, pp. 232f.
126 Divination, mythology and monarchy
said that the energy of earth was in the ascendant; and in those circumstances he
singled out yellow for prominence among the colours and modelled his actions on
earth. In the time of Yu, Heaven had displayed grasses and trees that were not killed
off in autumn or winter. Yii said that the energy of wood was in the ascendant; and in
those circumstances he singled out green for prominence among the colours and
modelled his actions on wood. In the time of T'ang, Heaven had first shown how metal
blades were produced from liquid. T'ang said that the energy of metal was in the
ascendant; and in those circumstances he singled out white for prominence among the
colours and modelled his actions on metal. In the time of king Wen, Heaven had
displayed fire, with scarlet birds holding texts inscribed in red in their beaks, and
assembling at the altars of Chou. King Wen said that the energy of fire was in the
ascendant; and in these circumstances he singled out red for prominence among the
colours and modelled his actions on fire.
It will of course be the energy of water that must displace that of fire, and Heaven will
make a display of water in advance, so that the energy of water will come into the
ascendant. When that occurs, the ruler will single out black for prominence among the
colours and model his action on water.9
Tung Chung-shu fastened especially on the relationship between Heaven
and the cycle ofYinYang and its phases, as may be seen below; but the Han
shu does not include any allusion to his deliberate association of either the
Ch'in or the Han dynasty with the predominance of a particular phase in the
cycle. In the meantime there is reason to believe that even the highly practical
rulers and officials of Ch'in had paid some attention to the theory, by claiming
that their power existed under the aegis of water.
During the Han period a
more pronounced interest is evident, with definite attempts to claim that the
emperors were exercising their authority in accordance with the pre-destined
sequence of the Five Powers. During the second century BC there were several
abortive attempts to persuade the emperors that their power was blessed no
longer by the symbol of water, as had been inherited by Han, but by the next
symbol or phase in the sequence, i.e., that of earth; in 104 BC the change was
duly introduced. Considerable importance was laid on this aspect of imperial .
authority at subsequent moments of dynastic change, for example by Wang
Mang, who re-adopted earth in AD 9, and by Kuang-wu-ti who chose fire
(AD 26).
The implications of this determined act of symbolism were far reaching. In
declaring which element was its patron, a dynasty both claimed to exist as part
of the pre-determined order of nature and defined its relationship to its
9 LSCC 13.4a. Yii and T'ang were the founders of the Hsia and Shang dynasties and king Wen
was the effective founder of Chou. It may be noted that the order of succession of the phases is
that of conquest and not of natural birth. The 'texts inscribed in red' were books thought to
describe the ways of antiquity. King Wu of Chou had been informed that they included
information regarding the Yellow Emperor and other mythological rulers and wished to
consult them. Their description as 'red' implies that they were written in imperishable materials.
For references, see chapter 2 above note 5.
11 SC 12, p. 48, Mll vol. III, p. 515; HSPC 25B.23b; HSPC 99a.36a, b and 99B.9a, HFHD
vol. III, pp. 258-9, 290; HHSCC 1A.18b.
Imperial sovereignty 127
,;predecessor. changing from water to earth in .1 ?4, Han was
its claim to exist as the conqueror who had legitimately displaced Ch'm. By
te-adopting earth, Wang Mang was deliberately claiming that his dispensa-
tion had come into being not by way of conquest but by way of natural
processes. In addition, both Wang Mang and the first of the Later Han
emperors, who declared that his dynasty existed under the protection of fire,
were exhibiting their belief that they were the normal and legitimate successors
rtot of the emperors of Ch'in but of the kings of Chou. Kuang-wu-ti was also
Wang Mang as an illegitimate usurper.
'Urttil recently the expression Huang-Lao was little more than a term seen in a
f'W passages of the Standard Histories. That it was a mode of thought which
to some of the most highly placed persons in the land was clear, but its
implications were unstated.
The position has been changed by the discovery
copies of hitherto unknown texts at Ma-wang-tui, and the
. 1dentification of some of these as expressions of Huang-Lao thought.
' One of these documents (Ching fa) appears to be a handbook of guidance
for a ruler of mankind, with a description of some of the techniques required
jfor his task. The government of man is justified on the grounds that it is part of ,
a universal order of being (tao). The book refers to Yin Yang and to the belief
that correspondences (kan-ying) exist between phenomena manifested in
tiifferent parts of the universe. But there are apparently no references to wu
:ihsing. The essential point is that in this hitherto lost text temporal rule is seen
:i\'spart of the cosmic order, and the concept of that order is very different from
that which was adopted by Tung Chung-shu. The same principle is brought
;.'Dirt in two of the other documents found at the site.
, . Huang- Lao thought saw tao, which one is tempted to translate as 'ultimate
.. as the origin of all things and the home to which all things return. The '
.authority of a sovereign derived from his responsibility for seeing that his
Ts'ao Shen, who became chancellor of state in 193 BC, had been subject to the influence of an
exponent ofHuang-Lao thought named Kai Kung HSPC 39. l la (SC 54, p. 13, SC 130, p. 59).
.., Others recorded as having favoured or studied Huang Lao thought included T'ien Shu,
governor of Yiin-chung commander during Wen-ti's reign (HSPC 37.4b) and the Empress
Dowager Tou (consort ofWen-ti, mother of Ching-ti); after her death in 135 BC, T'ien Fen
took steps to remove its influence from the court (HSPC 52.4b; HSPC 88.3b; HSPC 97 A.Sa).
Ch'en P'ing, who took a leading part in ousting the Lii family and enthroning Wen-ti, had also
favoured HuangLao thought (HSPC 40.12a). Subsequently, others included Chi Yen, who
attained ministerial rank in Wu-ti's reign (HSPC 50.9a); Cheng Tang-shih, appointed
Ta-nung-ling (superintendent of agriculture) in 130 BC (HSPC 50. l 5a); and Liu Te, who held
the office of Tsung-cheng (superintendent of the imperial family) for a few months in 80 BC.
[For others who are known in this connection, see Vervoorn (1990), p. 268 note 31.)
' "' .See the four texts which preceded copy B of the Lao-tzu, entitled 'Ching fa', 'Shih liu ching',
' and 'Tao yiian'. These are transcribed in Ma-wang-tui Han mu po shu vol. I. See also
. . Loewe (1981) and (1982), p. 209.
} 'Ch'eng' and 'Tao yiian'.
128 Divination, mythology and monarchy
subjects' lives conformed with tao. For this purpose he must, by his issue of
orders, establish suitable norms and patterns of behaviour. He must also
practise the technique of observation (kuan) so as to acquire an insight into the
inner workings of the universe and the relationship between man, heaven and
earth. In so far as tao comprehends models that are to be imitated and their
prescriptions, Huang-Lao thought provided a framework for laws that was
utterly different from that of the tradition of Ch'in. The purpose of a
sovereign's commands thus varies fundamentally from that of the rulers of
Ch'in, whose orders were intended to achieve obedience to a human will and
the attainment of man-made ambitions. Huang-Lao thought also varied from
\ the view of man of the Chuang-tzu that saw no place for the organisation, let
! alone the compulsion, of mankind.
These texts also include a view of human evolution and of the growth of
civilisation that varies from the one traced to Confucius and his disciples and
taken over by Tung Chung-shu. One document sets out a metaphysical
framework that included the categories of wu (matter), hsinga (form) and
m i n ~ (name). According to another text, the four orders of chic (discipline),
hsinga (form), shihd (seasonal distinction) and minga (name) had been created
' deliberately by the servants of Huang-tib, the Yellow Emperor, for the benefit
and improvement of mankind. By these means man was enabled to achieve
moral progress and to raise his state of existence from chaos to order.
These concepts and ideals vary considerably from those that grew up as a
result of Confucius' teaching, and the idea that human amelioration i's
achieved by attention to Ii" ,jen and z-e. The texts also call on a mythology that
differs from that enshrined in the Confucian canon, where the paragon rulers
are named as Yao and Shun. The Huang-Lao texts invoke the splendid
victory ofHuang-ti over Ch'ih-yuat a crucial stage in the advance of mankind
from barbarism to civilisation, from a state of anarchy to one of ordered and
recognised government. Although the texts do include at least one reference to
the t 'ien ming there is a conspicuous absence of reference to Confucius' ethical
values or to his contribution as a teacher.
The ideas of Huang-Lao thought were current in\. China before 168 BC.
Although it cannot be told how widely they had ~ n acclaimed, the influence
exercised by one devotee, the Empress Dowager Tou, who died in 135, may
have been significant. But however influential this mode of thought and its
view of imperial sovereignty may have been in the early decades of the Han
period, it was. due to be eclipsed by other ideas, for reasons which may be
suggested, rather than proved, as follows.
In 139 BC the Huai-nan-tzu was presented to the throne. The work derived
from the speculations and discussions of a number of writers, and it set out to
present a highly systematic explanation of the universe. As far as may be said,
in the absence of full source material, the scheme of the Huai-nan-tzu was of a
. " For example, see the transcriptions under reference in note 13, pp. 43, 53, 62, 66, 72, 81and87.
Imperial sovereignty 129
more comprehensive and persuasive nature than the ideas of Huang-Lao
thought, .and it is hardly surprising that subsequent interest in tao fastened on
the Huai-nan-tzu rather than on other writings. When, several centuries later,
a new thrust was given to 'Taoism' by religious motivation, it was Lao Tzu
rather than Huang-ti who formed the ancestor to whom the new practices
were ascribed. Thereafter there was little impetus to revive the texts of
Huang-Lao thought.
A further reason for the eclipse of Huang-Lao thought may perhaps be seen
in the change that was affecting the image of Huang-ti. Huang-Lao thought
saw him as an ancestral figure from whom the government and organisation of
man, among other things, derived. However, he features in a different capacity
in the state cults of both Ch'in and Han, as one of four, and then five, powers
(tz) to whom worship was seasonally due. A further change of emphasis may
have taken place in 104 BC, when Huang-ti was invoked as an intermediary
who could procure immoriality.
This cult probably did not last for long and
it does not seem to have made much of an impact; but it may well have
detracted from the image of Huang-ti as a forerunner of human sovereignty,
at a time when other ideals were being propagated for this purpose.
It is perhaps more likely that the replacement of Huang-Lao thought by
other views was partly due to the growing ascendancy of ethical ideas, as
espoused by Shu-sun T'ung, Lu Chia and Tung Chung-shu, and the place
taken by these ideas in the training of officials. The principal steps to develop
that training were taken in 141, 136 and 124 BC.
Shu-sun T'ung and Lu Chia
Early in the Han period there were some who realised that a considerable
difference lay between the successful seizure of imperial power by forceful
means and the maintenance of effective leadership over an empire at a time of
peace and stability. The Standard Histories record two instances in which the
contrast was drawn in explicit terms to Kao-ti (reigned as emperor 202 to 195
BC) and two results followed.
In c. 201 Shu-sun T'ung drew up a set of guide-lines for the correct
procedures to be followed at court, with a view to characterising the new
regime as one backed by attention to cultural refinement rather than by the
crudities of a military way oflife. These conventions were intended to enhance
the majesty and dignity of the imperial throne and to establish it as a patron of
civilised values. These included a recognition of the correct relationship
between the sovereign and his counsellors. 18
There is no separate category for books deriving from Huang-Lao thought in the bibliographi-
cal list incorporated in chapter 30 of the Han shu. Titles which include the term Huang-ti are to
be found among the Tao chia; one item is described as being comparable with the Lao-tzu
(HSPC 30.36b, 37a).
HSPC 25A.3lb, 35a; CC p.184.
HSPC 43.15a; HFHD vol. I, p. 21; see also HSPC 22.3a and Hulsewe (1955), p. 433 .
Divination, mythology and monarchy
In the second well-known instance, Lu Chia had pointed out the essential
difference between the achievements of Ch'in and the maintenance of a stable
empire. There followed the production of a set of essays in which Lu Chia
explained the principles of government. This short text, entitled Hsin yii,
attempts to demonstrate that ethical values distinguish a reputable govern-
ment from others. It is the unsuccessful regimes which are bound only by
material considerations and which take unduly oppressive measures; it is these
that are eventually brought down by the force of popular pressures.
Possibly owing to the prominence attained by Tung Chung-shu, insufficient
attention may have been paid to Lu Chia and his provision of intellectual
support for an imperial regime. Lu Chia's brief, as ordered by the emperor,
was not to expound a set of ethical concepts; it was to show the reasons for the
success or failure of a particular regime.
In the course of his writings, Lu Chia recognises the existence of a
relationship between Heaven and Earth, and he points out the importance of
phenomena and omens as a voice of warning. He stresses the value of moral
virtues, as espoused by the sages, arguing that cultivation ofjen and ileads the
way forward to ordered government (chihh). Lu Chia observes the failure of
the Ch'in empire, following the exercise of punishments and other intensive
activities. The arrogance and extravagance that had been involved forms a
strong contrast with the ideal of achieving ordered government by taking as
little an active part as possible. The Hsin yii notes that the best type of ruler
should conform with the seasons ordained by Heaven, and so control his
movements that they accord with the rhythm of Yin and Yang. Lu Chia
deprecates eremetism as a form of escapism from public duties, and he stresses
the need to recognise talent among human beings. In pleading for a consistent
single-mindedness, he asks his ruler to concentrate on matters of principle and
to avoid attention to material profit. In discussing the need for good faith and
trustworthiness, he argues that the decline of a dynasty is due not to Heaven
but to the faults and shortcomings of m.an. The book does not mention the
t'ien ming.
The Hsin yii carries citations from works such as the Book of Songs and the
Book of Documents, and from sayings attributed to Confucius. Reference is
also made to some of Confucius' actions, as examples of wise and uncompro-
mising behaviour. There are three references to the Ch'un-ch'iu's record of
historical fact, one citation from Lao Tzu and one reference to Mo Tzu.
Lu Chia's views carried with them deep implications that recur throughout
imperial history, but they also raised one awkward question. If it were
accepted that successful government rested on the practice of ethical prin-
19 HSPC 43.6a.
20 The authenticity of the Hsin yii was brought into question by the editors of the Ssu-k'u
catalogue, but it has been accepted by a number of notable scholars including Sun 1-jang
(1848--1908), Yi.i Chia-hsi (1883--1955), Hsi.i Fu-kuan (twentieth century) and Miyazaki
Ichisada (b. 1901); [see the entry for the work in Loewe (1993)].
Imperial sovereignty 131
ciples and on moral leadership rather than on force, how had Liu Pang been
justified in winning his way to imperial eminence by military means? The
question was openly voiced slightly later, in the presence of Ching-ti (reigned
157--141 BC) and the resulting discomfort led the emperor to bring the
discussion to an abrupt close. The Han shu records a further occasion on
which this question of principle was raised, in 46 BC.
The awful example of Ch'in
The question of the success and failure of imperial regimes led inevitably to
assessments of Ch'in, as may be seen in expressions of thought shortly after
the time of Shu-sun T'ung and Lu Chia. These were the opinions of statesmen
faced with the practical problems of administering the empire, but the political
circumstances in which they were writing differed from those of the previous
generation. Shu-sun T'ung and Lu Chia had been tendering their advice to a
newly victorious emperor whose position was apparently unchallenged; their
successors, of Wen-ti's reign (180-157), were writing after a period when the
imperial succession had been subject to manipulation and the emperor's own
place had been taken over by the empress Lii. When Chia I (201-169 BC) and
others were advising Wen-ti on the principles of imperial government, they
had in their minds the very recent experience of dynastic instability and the
threat posed to the fabric of empire.
Chia I and Chia Shan were two of those who addressed themselves to the
question of why the mighty Ch'in had succumbed to a minor attack of rebels
led by Ch'en She. 'Nailing the errors ofCh'in' was the title of a famous essay of
Chia I that is incorporated in the Shih-chi and the Han shu.
By asking this
question about Ch'in, Chia I and his contemporaries may, in addition to their
main purpose, have set a pattern of criticism that lasted for two thousand
years and more since their time. Questions were not, and are not put,
regarding the quality, legitimacy or wisdom of an existing regime; by voicing
them in respect of a comparable person or institution of the past it is possible
to express criticism of the present, for the reader is expected to draw the
obvious inference. In this way Chia I may well have been masking his criticism
of his own masters.
Chia I's essay should probably be interpreted as a warning directed against
the improper operation of imperial government but not as a criticism of the
imperial system as such. He described the errors of Ch'in in the hope that Han
would avoid making the same mistakes, and he blamed the failure of Ch 'in to
withstand the rebels on its excessive severities and penal sanctions. Chia I was
SC 121, p.16; HSPC 88.18b; HSPC 75.19a; see chapter 4 above, p.86.
22 Chia I's essay 'Kuo Ch'in' divides into three sections, which appear variously as follows. It may
be noted that the entire text is given in SC 6, where the order of the three sections is incorrect.
For section I, see SC 6, pp. 91-9, MHvol. II, pp. 225-31; SC48, pp. 21-5; and HSPC 31.24b;
for section 2, see SC 6, pp. 99-103; MH vol. II, pp. 231-6; for section 3, see SC 6, pp. 87-91,
MH vol. II, pp. 219-24.
132 Divination, mythology and monarchy
anxious to see the fabric of empire maintained and to avoid the fate that had
overtaken Ch'in, or the dangers caused by the empress Lii and her family. For
this reason he stressed the need for ethical values. By doing so he has earned
the traditional classification among the ju chia or 'Confucians', but that
classification is somewhat misleading. Chia I was more concerned with
defending the idea and practice of imperial government in the face of possible
danger than with preaching the virtues advocated by K'ung Tzu. He would
better be regarded as a staunch adherent of empire as against other forms of
government, and in this sense he is comparable with his contemporary Ch'ao
Ts'o (executed 154 BC), who is classified among the fa chia or 'Legalists'. It
may be of interest to recall that, during the campaign to 'Criticize Lin Piao and
Confucius' that was raging in China in 1974, Chia I was assessed and praised
as a statesman who served China with merit and without the taint of
Confucianism. 23 It will be seen below that a new emphasis was given to the
criticism of Ch'in on moral, rather than on practical, grounds, by Tung
Chia Shan, who was no relation of Chia I, was likewise drawn to explain
how the defeat of Ch'in had come about. It is clear from his essay that he was
writing byway of warning to his emperor, in the hope that he would not repeat
some of the errors that had led to Ch'in's downfall.
Like Chia I, he was
addressing Wen-ti. He blamed dynastic ruin on the excessive demands that
Ch'in had imposed on the population, or on Ch'in's material extravagances
and lack of scruple. Coupled with greed and ambition, and in the absence of
restraint, these faults constituted an abuse of imperial sovereignty, to the
extent that Heaven brought about the end of the regime.
In making this statement, Chia Shan was possibly proceeding further
forward than Chia I, and he advances even further in the next part of his
submission. He reflects on the nature of Chou's clement rule as a paragon and
in doing so he expresses a theme that is to be heard for ever afterwards. Chia
Shan also took the opportunity to insist on the importance of offering
remonstrance and criticism to the throne, and on the need of emperors to
establish officials who were charged with this duty. He also made one other
point, which was specifically directed at contemporary practice rather than
the failings of Ch'in. He urged a reduction of material extravagances and of
indulgence in pleasures such as hunting. This warning, apparently given
somewhat late in Wen-ti's reign, perhaps seems odd in view of that emperor's
reputation for parsimony and reluctance to tax the population unduly.
Possibly his reputation for thrift was earned after due heed had been taken of
Chia Shan's words, which were submitted in the interests of maintaining
imperial sovereignty securely.
At very much of the same time, or perhaps slightly later, the contributors to
23 This assessment of Chia I was displayed on placards observed in Lo-yang in 1974.
24 HSPC 51.la.
25 For Wen-ti's reputation for frugality, see HSPC 4.2la, HFHD vol. I, p. 272.
Imperial sovereignty 133
the Huai-nan-tzu were re-iterating some of these arguments, but from a
somewhat different point of view. Again, there is a recognition of Ch'in's
material achievements and military success, and a corresponding contrast is
drawn with the ease with which the empire had succumbed to the pressures of
weak and poorly armed forces. The reasons are seen in Ch'in's excessive
Concentration on military ventures, the extravagance of the regime and the
failure to give a proper hearing to critics.
The conclusion is thus similar to
that reached by Chia I and Chia Shan, but the reasons for doing so vary. Chia
I and Chia Shan believed that it was the lack of respect for ethical
considerations, in purely practical terms, that had brought about Ch'in's ruin.
The writers of the Huai-nan-tzu, however, had other principles in mind; they
criticised Ch'in's excesses not on the grounds that they were unjust to man, but
because they conflicted with man's proper relation to the world of nature.
The needs of empire
During Wu-ti's reign (141-87 BC) a new emphasis came to be placed on
ethical values as the basis for imperial government rather than on materialist
riches as ends in themselves. The change may have been due in part to the
expressions of opinion that have already been noted, but the very needs of that
government were partly responsible.
Wu-ti's reign witnessed active efforts to expand the scope of the administra-
tion, to intensify its hold on the population and to co-ordinate the use of its
material resources. To achieve these aims it was necessary to expand the civil
service. Loyal and well-trained officials were needed in considerable numbers
to staff the organs of government as they grew in size and complexity, and the
government needed to take an active part in education and recruitment. These
developments involved a paradox. For, by responding to the growing need for
training candidates for office, the government came to engender a respect for
values that ran counter to those of the prominent officials of the day and their
policies. Those policies were directed towards achieving material enrichment;
the texts with which candidates were trained looked to the moral improve-
ment of man.
The steps taken to train and recruit civil servants are known well enough.
By an edict of 141 BC, discrimination was to be shown against those
candidates whose main claim to ability lay in their familiarity with the realist
aspects of Chinese thought. The edict voiced a specific reaction against writers
such as Shen Pu-hai (b. c. 400 BC), Shang Yang (c. 385-338 BC), Han Fei (280
to c. 233 BC), Su Ch'in and Chang I (Chan kuo period), i.e., some of those
whose teaching and principles lay behind Ch'in's policies of expansion and
enrichment. Possibly the promulgation of this edict owed something to the
HNT 12.24b, 13.llb, 15.8a, 18.23a and 20.15b, 25a.
See HFHD vol. II, pp. 20f; Bielenstein (1980), pp. 132f; [and Twitchett and Loewe (1986),
pp. 463f].
134 Divination, mythology and monarchy
pleas made by Tung Chung-shu in corresponding terms. The peroration of
one of the memorials included in the Han shu ends with the suggestion that
active discrimination should be practised against those who were not versed in
the curriculum of the six approved subjects (liu i) and the methods of K'ung
According to at least one scholar (Su Yii) this memorial preceded the
issue of the edict, but the precise dating must remain a matter of controversy.
Other steps that followed shortly carried the ideas of Shu-sun T'ung and Lu
Chia further forward.
In 136 BC there was established a series of official
posts for academicians (po-shih), who were specialists in named subjects or
texts, including those that concerned protocol and approved behaviour (Ii); it
was Shu-sun T'ung who had advocated the cultivation of these habits. Other
academicians were to be specialists in documents that already possessed a high
reputation for their moral values, which had been cited by Lu Chia. One of the
academicians was to be a master of the Ch'un-ch'iu, compiled by Confucius
and redolent with moral lessons that could be drawn from the past and applied
to the present.
In the very same memorial to which allusion has just been made, Tung
Chung-shu emphasised the continuity of past and present and the value of the
Ch'un-ch'iu; that text, he argued, showed both how events of the present could
be matched with those of the past, and how human ideas relied on those of a
superior body (t'ien).
Both the edict of 136 BC and Tung's memorial show an
intention of incorporating ethical values and a respect for the past into the
regular training of officials of the empire. A further significant move in this
direction took place in 124 BC, with the deliberate increase in the number of
those admitted for training. This measure derived from a suggestion of
Kung-sun Hung, chancellor of the empire.
The many calls for recruitment, and the terms in which they were made,
reflect the acceptance of ideas voiced by Tung Chung-shu and expounded by
the academicians. Senior officials of the central government and the provinces
were ordered to recommend candidates known to possess named qualities.
These were those enshrined in the texts associated with Confucius and cited by
Lu Chia. They included moral integrity, intellectual ability, conformity with
conventional values, readiness to speak openly in criticism without fear of a
superior and a sense of family responsibilities.
Tung Chung-sbu's contribution
Tung Chung-shu is usually described as a syncretist and as the founder of Han
Confucianism. While it may well be true that he drew on a number of ideas
2 HSPC 56.19a. For the liu i(i.e., the Book of Music, Book of Songs, Book of Documents, Book of
Changes, Compendium of Ii, and Spring and Autumn Annals), see HSPC 30.26b.
2 HSPC 6.3b, HFHD vol. II, p. 32.
HSPC 56.14b.
31 HSPC 6.1 lb, HFHD vol. II, pp. 24, 54.
' 2 HSPC 6.4a and 8b, HFHD vol. II, pp. 34, 46; HSPC 65.lb; HSPC 85.lb and 19a.
Imperial sovereignty 135
that had already been expressed, and that in this sense he is less of an
innovator than has sometimes been believed, he went considerably further
than his predecessors in taking some of their ideas to a logical conclusion and
in formulating a systematic view of the universe. In doing so he played a
crucial role in reconciling ideas that were contradictory. Possibly he was aware
of a conflict between faith and reason; between belief in the power of
revelation through omens or by divination, and an explanation of the universe
in rational terms according to the principles of the Five Phases.
The three memorials that are included in the Han shu
were submitted in
direct response to questions raised in imperial edicts. Those questions were
concerned directly with the nature of imperial sovereignty. The first of the
edicts took note of the way in which ideal methods of government had been
practised in the past, to the general benefit of mankind, and proceeded to
enquire on what basis such results had been achieved, with the blessing of
Heaven. In his second edict the emperor enquired why the steps that he had
been taking to govern China responsibly and appropriately had failed to
produce adequate results. The third edict referred to the relationship between
Heaven and man; it pointed out that the style and methods of government
adopted by the ideal kings of the past were by no means identical, and sought
the reasons for such obvious discrepancies.
These questions provided ample scope for long replies. Attention will be
restricted here to four major matters of principle to which Tung referred, i.e.,
(a) the cosmic view of the universe; (b) the role of Heaven, Yin-Yang and the
Five Phases; (c) omens and their significance; and (d) the critique of Ch'in.
(a) Towards the end of his third memorial, Tung wrote:
The grand co-ordinating unity (ta i t'ung) that is mentioned in the Spring and Autumn
Annals is a thread which runs constantly through Heaven and Earth, and it forms the
principles of action that have been generally accepted in past and present. At present
our teachers propagate strange principles; our fellow human beings hold to unusual
practices; the many schools of thought have idiosyncratic methods, and the con-
clusions to which they point are not identical. It is for these reasons that the upper
reaches of society have no means with which to up-hold the co-ordinating unity; and as
the models for behaviour and institutions have been frequently changed, the lower
reaches of society do not comprehend what is being preserved.
Elsewhere Tung re-iterates the same belief, in a single organised system
which formed the constant everlasting principle of Heaven and Earth, of past
and present alike. Within this scheme man was playing his part as but one of
several elements in the system. Tung seems to have been in advance of his
contemporaries in recognising a unitary system with a place for Heaven,
Yin-Yang and the Five Phases, for the exercise of temporal government and
for the maintenance of the approved ways of a cultured existence.
For references to these memorials, sec note 4 above. 34
HSPC 56.19a.
136 Divination, mythology and monarchy
(b) The passage from the Lii shih ch'un-chi'iuwhich is cited above (pp. 125f)
explains the incidents of dynastic succession in terms that involve three agents,
i.e., Heaven, the symbols of the Five Phase and the occurrence of omens. It
would seem that the passage is of unique importance in combining these three
elements together at an early stage of Chinese thought, i.e., the middle of the
third century BC. As far as may be told, we must wait until the time of Tung
Chung-shu before this synthesis could be taken up and expanded, despite its
inherent contradictions.
It need hardly be recalled that t'ien had occupied a central place in the
religion and political thought of the kings of Chou. Tien was a god to be
worshipped and the fountain-head of authority. As the idea was propagated,
so was the king of Chou styled the Son ofHeaven,
and recognised as the sole
arbiter of human destinies. But with the emergence of the self-styled kings
(wang) in different parts of China, the unique position ascribed to the kings of
Chou became a fiction, respected in theoretical but hardly in practical terms.
As none of the kings of the Warring States could claim that their authority
derived uniquely from Heaven, the concept of the 'Son of Heaven' became less
and less meaningful and the worship of Heaven grew more and more tenuous
as a means of uni ting temporal rule with the authority ofa supreme power. By
the time that the Ch'in empire was founded, the official cults of state were
being directed to other deities, which were likewise adopted by the new
emperors of Han. Heaven took little or no place in the religious cults
patronised by the Han emperors until c. 31 BC.
In the same way, the idea of the heavenly mandate, which had made its
appearance in the Chou period
at least, could hardly be valid in the centuries
of the Warring States. These were the years when several kings were exercising
comparable powers of government simultaneously, on the basis of the
principle that might is right. No single one of the rival kings could claim that
his authority effectively drew on the blessing of Heaven.
In addition there was a fundamental conflict between this concept and the
35 For early references to the term t'ien tzu, see I ching 'Ta yu', SSC 2.30a, Wilhelm (1951), p. 61;
Shih ching 'Ts'ai shu', SSC 15(1).6a, Karlgren (1950a), p.176; Shang shu 'Hung fan', SSC
12.14b, Karlgren (1950), p.32 (item 15).
36 The first ruler to claim the titleofwang, other than the king of Chou, was Ch'u Wu wangin 740
BC (SC 14, p. 36); by 230 BC the title had been adopted in seven states including Ch'in.
37 CC eh. 5.
38 For early references to t'ien ming, see, for example, Shih ching 'Hsiao yiian' SSC 12(3).lb,
Karlgren (1950a), p.144; 'Wen wang' SSC 16(1).6a, Karlgren p.186; 'Ruan' SSC 19(4).18b,
Karlgren p. 252; 'Lai' SSC 19(4).20b, Karlgren p. 253; and 'Yin wu' SSC 20(4).12a, Karlgren
p. 265. Shang shu 'P'an keng' SSC 9.3a, Karlgren (1950b), p. 20, line 3; 'Ta kao' SSC 13.16a,
23b and 24a, Karlgren pp. 36, (line 1), 39, lines 13 and 15; 'Wu i' SSC 16.!0a, Karlgren p. 58,
line 4; 'Chiin shih' SSC 16.18a, Karlgren p. 59 line 1; and 'Wen hou chih ming' SSC 20.2a,
Karlgren p. 78 line 1. See also SC 127, p. 2 for a reference to the relationship between receipt of
the Mandate, the prosperity of a regime and divination; Creel (1970), pp. 8lf; and chapter 4
above pp. 88f.
Imperial sovereignty 137
application of the Five Phases to political destinies. The one doctrine held that
Heaven deliberately chose a person or a house for investiture with responsibil-
ity for the government of man; the choice depended on that person's or that
house's possession of qualities that were adequate for the task of bearing
temporal authority. But according to the theory of the Five Phases, there is a
fundamental and inescapable rhythm which underlies all activities and
movements and regulates change according to successive and predetermined
stages. This theory provided for changes of temporal power to be brought
about either by one regime's conquest of another, or by a process of natural
As has been seen, the writer of the Lu shih ch'un-ch'iu saw no difficulty in
reconciling the part of Heaven with the predetermination of natural se-
quences. Tung Chung-shu accepts and re-inforces this compromise in a
number of ways, and in doing so he is following earlier writings. The opening
passage of the Hsin yii of Lu Chia cites a statement that attributes to Heaven
the creation of all things. It proceeds therefrom to the regulation of Yin and
Yang and the establishment of the Five Phases in their due order.
The same
idea, that Heaven is responsible for the adjustment of Yin and Yang, appears
in a splendid passage of a writer with a different point of view, in the
Tung Chung-shu elaborates the idea as follows, in the first of
the three memorials:
A major element of the order of Heaven lies in Yin-Yang, with Yang constituting
bounties and Yin constituting punishments. While punishments control slaughter,
bounties control living. For these reasons Yang constantly takes its place in the depths
of winter, being concentrated in places that are void and of no practical application.
We may thus observe that Heaven deputes its charge to bounties and not to
Heaven commands Yang to appear without, and by spreading its benefits above, to
bring about the successful completion of the year's work. Heaven causes Yin to go
within, and by concealing itself below, to emerge at due seasons to assist Yang. For if
Yang does not receive the help of Yin it cannot by itself complete the work of the year.
Ultimately it is Yang that achieves renown by completing the work of the year. Such is
the intention of Heaven.
The passage proceeds to draw the analogy between the role of Heaven and
that of an earthly ruler. The same point, i.e., Heaven's creation of matter and
its control of Yin--Y ang, is explained at considerable length in the third of
Tung's memorials, with an elaboration in terms of the seasonal changes
whereby the Five Phases are manifested.
Later in the third memorial Tung seeks to explain why there are apparent
differences in the principles adopted by the sage kings of old. In doing so, he
insists that throughout their regimes there was no deviation from basic
Hsin yii l.la.
HNT 20. la; Loewe (1982), p. 64. 41
HSPC 56.5a, b.
HSPC 56.14b.
138 Divination, mythology and monarchy
principle; it was only their application, by way of expedient, that varied, as
required by circumstance. He stresses that it is from Heaven that the major
principles of the world's order (tao) proceed: 'The main origin of tao comes
from Heaven; Heaven does not change and tao likewise does not change.'
This view is at variance with a statement in one of the ancillary texts of the
Book of Changes, to the effect that it was the holy men of old who 'in making
the Changes sought to accord with the principles of nature and destiny, and for
this reason established the order of Heaven, namely Yin and Yang, and
established the order of earth, namely pliant and adamant'.
On several occasions Tung Chung-shu refers to the t 'ien ming, but in
somewhat vague terms as compared with later passages that allude to this
idea.45 In imperial times it is necessary to wait for the time ofK'uang Heng (d.
c. 30 BC) and Pan Piao (AD 3-j4) before the idea of t'ien ming can be seen to
be making an impact on the ideas of imperial sovereignty.
Parallel with the concept oft 'ien ming is the view of the king who forms the
essential link that binds the three estates of Heaven, Earth and Man together
and acts as a channel for communication. This idea is set forth in a famous
passage of the Ch 'un-ch'iufan-lu that explains the form of the character wang
as symbolising the monarch's role;
the Han shu does not include a reference
to this symbol. Ifit can be accepted that that part of the Ch'un-ch'iufan-lu
stems from Tung chung-shu himself, it could also be accepted that further
definition had been given to the concept of sovereignty in the second century
(c) The passage from the Lu shih ch'un-ch'iu that is cited above (pp. 125f)
draws a specific association between the appearance of certain signs, which
may be termed omens, and the succession of temporal powers. Here again,
although omens are taken as indications of personal destinies, it is necessary
to wait for some decades before they are linked with the fate of an imperial
In personal terms, omens formed a salutary warning to the empress Lu, who
recognized them as signals. An eclipse that was reported for 178 BC formed
the mainspring of an edict in which the new emperor searched his conscience
and tried to identify his errors.
Thereafter a number of edicts sought to
explain untoward events in the light of the practices or malpractices of
officials, or the material circumstances of the empire.
Perhaps the first deliberate attempt to exploit an unexplained event as a
means of arousing faith in the dynasty may be seen in the retrospective
adoption of the regnal title Yiian-ting, for enumerating years from 116 BC
onwards. In this instance the fortunate discovery of bronze tripods was
43 HSPC 56.16b. 44 I ching ('Shuo kua') SSC 9.3a, Wilhelm (1951), p. 264.
45 HSPC 56.2a, 4b and 15a.
See chapter 4 above, pp. 88f.
47 CCFL 'Wang tao t'ung san' l 1.9a; de Bary (1960), vol. I, p. 163.
48 SC 9, pp. 18-19, Mlfvol. II, p. 422; SJ 9, p. 21, Mlfvol. II, p.425; HSPC 4.9a, HFJID vol.l,
Imperial sovereignty 139
commemorated and exploited as a sign of the blessing that the house of Han
had merited.
It was precisely at this time that Tung Chung-shu had been enunciating this
view on omens, their association with dynastic destinies, and the value of the
Spring and Autumn Annals in enabling a comprehensive view to be taken of
events past and present.
His view that omens were warnings sent by Heaven
to direct an earthly ruler to reconsider his policies and his treatment of man is
certainly reflected in the edicts of the next few years that refer to such events.
But perhaps the clearest examples of a deliberate attempt to exploit omens in a
dynasty's favour, and thus to demonstrate that it was in receipt of the blessing
of Heaven, are seen in the years 65 to 51 BC, when no less than nine edicts each
singled out events such as the felicitous behaviour of birds or the fall of
honeydew, and coupled them with announcements of bounteous acts. This
series was followed by a similarly striking series of edicts which fastened on
strange events that bore implications of the opposite type. From 48 to 42 BC
nine edicts referred to disasters or events that had upset the balance of nature,
thereby highlighting the inadequacy of the emperor and his dispensation. It
would seem that from 65 BC onwards Tung Chung-shu's view of omens as a
sign of Heaven's concern with the destiny of a dynasty, either for good or for
ill, had received some measure of acceptance.
(d) In the early part of Former Han a number of statesmen had criticised the
regime of Ch'in. It has been shown above, for example, how Lu Chia, Chia I
and Chia Shan attempted to explain Ch'in's dynastic failure in practical terms,
on the grounds of excessive and self-defeating measures that had been
imposed on the population; they had also made some allusion to the principles
that were involved. The same style of criticism had been raised in the
With Tung Chung-shu a new type of criticism enters in with a far sharper
emphasis on Ch'in's moral failings and on the measures taken to destroy the
cultural basis of Chinese civilisation. Ch'in's practice of imperial sovereignty
is not being examined in the hope of ascertaining where it had been at fault, as
it was by Chia I. Tung sees Ch'in as an example of an unjust dispensation and
of the unjustifiable use of temporal authority. Ch'in is accused of trying to
eliminate the traditional virtues and values of the past.
It may be asked to what extent, if any, Tung's strictures were really being
directed against the contemporary regime which he witnessed and with whose
policies he was by no means entirely in sympathy. Ifhe did in fact choose this
method of commenting on the government of his own day, he may have
rendered a profound disservice to the study of Chinese history, by drawing
49 HSPC 6.l 7b, 19b, HFHD vol. II, pp. 71, 75. It has been suggested (HFHD vol. II, p. 121) that
the adoption of the regnal title Yiian-kuang for 134 to 129 BC had already derived from the
occurrence of an omen. However, the omen in question has been identified as the appearance of
a comet, and it is by no means certain that comets were ever interpreted as signs of blessing
worthy of commemoration. See chapter 3 above, pp. 75f.
JISPC 27A.lla; Loewe (1982), p.86.
HSPC 56.7a, !lb and 16b.
140 Divination, mythology and monarchy
attention away from the faults of Han and exaggerating those of Ch'in. But
whatever Tung Churtg-shu's intentions may have been, he set new standards
for judging the quality of dynastic achievement.
Summary and conclusions
In the foregoing pages an attempt has been made to trace how ideas of
sovereignty developed during the first century of the Former Han period.
Some ideas of earlier origins, such as those of the Five Phases and the t'ien
ming, were elaborated and formed into part of a cosmic system; some, such as
the philosophy ofHuang-Lao, failed to mature in the face of competition; and
a sense of purpose was lent to imperial rule by the claim that the emperors were
following ethical guide-lines rather than relying solely on force. Within this
framework, Tung Chung-shu both leant on some of the ideas of his
predecessors and added his own characteristic emphasis.
Comparison with Lu Chia adds force to Tung Chung-shu's contribution;
for although no more than a few decades separated the two men and their
writings, the difference in intellectual terms is striking. Tung Chung-shu
stands out with a degree of sophistication, a power of sustained argument and
a power of analysis that had hardly been seen previously. He shows a deeper
sense of Heaven's personal part and its devotion to man's interests than Lu
Chia; he expresses a clearer concept of Heaven's power of warning. He draws
a subtle discrimination between human motives and qualities, while insisting
on the pre-eminence of man over other creatures. Tung Chung-shu also
demonstrates a more subtle view of the past, by distinguishing between the
constant value of certain principles and the need to adopt expedients from
time to time in order to preserve such principles. He points out that it is up to
man to apply the correct tao to his behaviour and government; it is not tao that
glorifies man. By contrast with Lu Chia's rather simple references to Ch'in and
its demise, he argues on rational grounds why kings and emperors should
refrain from exacting undue punishments and from indulging in other
excesses. Tung Chung-shu also treats the Ch 'un-ch 'iu,\compiled by Confucius,
with a new type ofrespect; he recognises that it may be! used as a source for the
study of man and his relationship with Heaven; ailid he pays careful attention
to the wording of the book, in the belief that its formulae conceal basic truths
and moral lessons.
Attention should be paid briefly to some of the developments that followed
Tung's own time. At the close of his life he may have witnessed the symbolical
changes that added dignity to the imperial structure and proclaimed that the
-start of a new age was being envisaged. These changes were not limited to the
adoption of earth in place of water as the symbol of the Five Phases. In
addition, a new calendar was introduced as a means of ensuring that the
regulation of mundane matters would accord more accurately with the major
cycles of the universe, and the adoption of the new regnal title T'ai-ch'u (The
Imperial sovereignty 141
Grand from BC stands as a public measure that was designed
to enhance 1mpenal prestige at that juncture. 52
Some decades later there followed the introduction of the imperial cults to
Heaven, finally accepted in the Later dynasty, and the revival of the
attention due to the Mandate of Heaven. Under Wang Mang's rule and that of
the Later Han emperors, omens played a more significant part than hitherto in
demonstrating the link between Heaven and the imperial dispensation. Tung
Chung-shu' s encouragement of education as a means of training civil servants
soon achieved noteworthy results; these were particularly impressive in the
Later Han period.
In one respect, however, Tung Chung-shu's influence is not so evident. The
Yen-t'ieh lun is an account of the debate held at court in 81 BC, being compiled
perhaps some two or three decades subsequently. The dialogue records a
number of instances in which the regime of Ch'in was criticised and even
However, such strictures were by no means always acceptable. For
the dialogue also includes occasions when spokesmen sprang to the defence of
methods; or they may have praised the basic policies of Shang Yang,
as the way that led forward to the successes of the Ch'in empire. In
add1t10n, those who are shown as defending Ch'in in the debate attributed the
fall of that dynasty to the failures of the individual officials or advisers who
served the second emperor. Tung's arguments that Ch'in's ruin followed from
ideological causes evidently did not command universal agreement.
In later ages, Ch'in has been criticised for pursuing 'Legalist' theories and
Han has been characterised as the champion of 'Confucianism'. As need
hardly be stressed, the situation was in fact far more complex, with varying
degrees of compromise being achieved between the extreme rigours ascribed
to Ch'in and the idealistic ethical approach to government that is attributed to
the Han Confucians. It was one of the achievements of the Han age to have
operated government within a framework that was acceptable to
both parties and that was backed by religious and intellectual support.
CC eh. I.
YTL 7,('Fei Yang'), p. 51, Gale(1931) pp. 42-3; 12 ('Yupien'), p. 91, Gale (1931), p. 79; 16 ('Ti
),p.115, Gale (1931),p. 102; 19 ('Pao hsien'),pp. 136--7, Gale (1931),pp. 123f; 23 ('Tsun
tao ), p. 168, (Gale p. 88; 24 ('Lun fei'), p. 172, Gale (1934), p. 91; 28 ('Kuo chi'), p. 192,
Gale (1934), p. 105; 29( Sanpu tsu'),p. 208; 38 ('Peihu') p. 263- 41 ('Ch'iihsia') p 275 and43
('Chieh ho'), p. 287. ' ' ' '
The cult of the dragon and the
invocation for rain
Chinese historical sources frequently mention the occurrence of drought ah'd .
the measures that were adopted to relieve the population from such '
Some of the methods reflect the early belief in a connection between thl:l.
appearance of dragons and the downfall of rain, and it is with this subject that
the present chapter is concerned.
For a variety of reasons the connection drawn between dragons and rainfall''.
bears an intrinsic interest. First, it is an example of sympathetic magic of ;
imitative type, which seeks to bring about material results by a display of;
phenomena similar to those that are desired.
Secondly, a blend of faith ani<
reason may be observed in the practices which derived from this belief..'
Finally, both the theory and the practice demonstrate a process that is seeniti;;);
other aspects of China's cultural development: a comparatively late rational,,-;,
isation and standardisation, based on philosophical principle, becoll1i&"
imposed on an original act of faith that could well have been of a very e
mythological origin.
Fortunately, sources which spring from different intellectual attitud
provide evidence for this study. In addition to a few
statements of historical fact and records of formal institutions, there.
references in the writings of several types of philosophy. The subje6
mentioned in the Huai-nan-tzu, which was presented to the throne in 139 .
and which sets out to describe the workings of the universe as a regular procest
of nature. In the chapters of this collection the writers are highly
to a belief that the appearance of dragons can bring about a fall of rai11L1a
addition the rationalist critic Wang Ch'ung (AD 27 to c. 100) seems, despite
his own principles, to have been unable to refute the existence of unexplained:
phenomena within this context, as may be seen in the Lun-heng. Finally,&.,
detailed description of a whole ritual that was designed to bring about a fall o,'>
rain is included in the Ch'un-ch 'iu fan-lu. Traditionally this work has
ascribed to Tung Chung-shu (c. 179 to c. 104), but doubts have been raised";
regarding the authenticity of all or some parts of the text, which may possibl)",
be dated up to four centuries or so after his time.
' For the distinction between imitative and contagious magic, see Frazer (1911), vol. I, pp,
2 See Chang Hsin-ch'eng (1957), vol. I, pp. 475f, [and the entry for Ch'un-ch'iu fan-Ju in
Cult of the dragon and invocation for rain 143
origins of the legend
evidence for the invocation of dragons to procure rainfall comes from
Buddhist China and may be traced back to the early centuries ofrecorded
Nevertheless the subject may perhaps best be introduced by reference
much later passage. This occurs in a famous essay of Han Yu (768-824),
eat exponent of Confucian rectitude and antagonist of Buddhist belief.
passage shows how, by the time of Han Yu, the belief in the dragon's
ers had so far become encapsulated in the Chinese tradition that an
yist could exploit it as an allegory so as to illustrate and add force to his
''ffing out his breath with a roar, the dragon forms the clouds; and the clouds are of
not possessed of greater spiritual power than the dragon. However, it is by
ng his own breath that the dragon journeys to all corners of the empyrean. He
close to the sun and the moon and he crouches within their effulgence. He gives
thunder and to lightning; he brings about transformations of nature such that
pours down upon the earth beneath, submerging the hills and the valleys.
,Writing this essay, Han Yii was putting forward the case of a disap-
ed and disgraced official, who was anxious to point out that his talents
scured and unrecognised. In his allegory, the dragon naturally enough
for the sovereign, or the emperor, and the clouds for his servants and
ials. Just as the dragon relies on the clouds to enact his purposes, so does
. peror no less require officials to carry out his will. It would follow that a
and saintly ruler would employ officials such as the writer, in order to
his function adequately. The metaphysical, or even theological, problem
ther an active agent or creator can only operate through the medium of
n creations hardly enters into Han Yii's argument; he contents himself
'xpressing his surprise at the accomplished reality.
mthis late and highly sophisticated allusion to the myth, we may revert
ljer references, which are to be found in writings of a very different type
a:s the I ching, the Tso chuan, and the Shan-hai ching.
mgons appear in a number of contexts in the account of the lines of the
.hexagram ch'ien
, and the enigmatic expressions of the yao
have given
to .a number of interpretations, including the elaboration of Han Yii that
i>t been cited. In particular, in one of the explanations of'nine in the fifth
\ we read that clouds accompany the dragon and winds follow the tiger,
of the universal rule that like things respond to like, be they sounds,
.. ions of energy or material, or visible objects such as water. It should
et the date of the CCFL may be, it has yet to be shown that the text has been influenced
way by Buddhism. For a general study of dragons in the East, see de Visser (1913); book
and book II, eh. 3 refer specifically to the link between dragons and the rain in China and
Yu; 'Tsa shuo ssu shou', in Ma T'ung-po, p. 19. For an early reference which links the
Ii with the invisible energy of the clouds (yiin ch'i). see Kuan-tzu 'Shui ti' 39.2b.
144 Divination, mythology and monarchy
perhaps be noted that this explanation of a highly difficult passage cannot be
dated back to the earliest parts of the I ching.
The Ch'un-ch'iu records a performance of the Great Rain sacrifice (yiid) for
the year 707 BC. In that part of the Tso chuan which is associated with the
record of this incident, the writer is concerned with dating the sacrifice
precisely, and in this connection he writes that 'the ceremony is conducted
when the dragon appears' (lung chien erh yii). This expression has required
explanation. It is understood by some commentators to refer to the fourth or
the sixth month, indicating that the ceremony should be deferred until the
right season of the year had been reached, when it would be most necessary.
Another explanation, which perhaps carries greater conviction, is that the
term 'dragon' denotes a constellation or possibly a planet. The text would then
mean 'When the Dragon Star, or possibly Jupiter, rises, that is the time when
the sacrifice of rain should be performed. '
A passage in the Lun-yu shows the
early association of a dance with the ceremony.
One of the later chapters of the Shan-hai ching, which was probably
compiled in the fourth century AD, alludes to the origin of the myth. Late as
this particular passage may be, it may with some confidence be taken to be
based on considerably earlier material.
Ying lung [or Ying the winged dragon] was situated at the southern extremity of
Mount Hsiung-li-t'u-ch'iu. He put to death Ch'ih-yu [god of fighting] and K'ua fu [a
mythical animal] and had no means of climbing up the mountain again. As a result,
down below there were many occasions of drought; and at times of drought, images of
Ying the Dragon were fashioned and great showers of rain were procured.
It is to this story that Kuo P'u traces the origin of the
contemporary use of clay dragons to attract the rain. He also draws attention
to the tale that a divine dragon helped Yiia the Great to control the waters, by
marking on the ground with his tail those watercourses that required to be left
The evidence for early practices
Two reasons may be suggested why there is an .absence of archaeological
evidence for this practice, which is mentioned frequently enough in literature.
In the first instance, there is a hint, to which reference will be made below, that
' The passage occurs in the 'Wen-yen' elaboration of 'Nine in the Fifth'; I ching, SSC 1.15a;
Wilhelm (1951), p. 382.
6 Ch'un-ch'iu SSC 6.8a and Tso chuan 6.,llb; Legge, vol. V, part I, pp. 45-6; Couvreur (1914),
vol. I, p. 84. For the dating of the yii see Huang Hui's note in LH 'Ming yii' 45, p. 671, Forke
(1907--11), vol. II, pp. 335f. [See passage (i) in the addendum below.]
7 For the inclusion of a ritual dance, see Lun yu 'Hsien chin' SSC 11.lOb, Legge vol. I, p. 248;
Chou li'Nii wu' SSC 26.lOb, Biot (1851), vol. II, p. 104; Li chi 'Yiiehling' SSC 16.3a, Couvreur
vol. I, p. 361.
' SHC, SPPY 14.6a, Yiian p. 359, Mathieu (1983), vol.I, p. 544.
9 See Kuo P'u's (276--324)commentin SHC 14, p. 360; Ch'u tz'u 'T'ien wen' 3.5b, Hawkes (1985),
pp. 128, 138-9.
Cult of the dragon and invocation for rain 145
the clay dragons used for this purpose were deliberately abandoned after use
as being valueless; and it is likely that if they had been made in a somewhat
rudimentary way, without firing by artificial means, they would soon have
disintegrated, particularly if they were exposed to the climatic elements which
they had been made to induce. Secondly, the great bulk of archaeological
evidence is derived from graves; these were hardly situations that called for a
talisman that would bring about a downpour.
Observance of the great ceremony of sacrifice and prayer for rain (yii or ta
yu) is well evidenced both for the pre-imperial period and during the Han
dynasty. The Tso chuan records a number of these occasions between the
eighth and the fifth centuries BC.
In addition, the Hsiin-tzu discusses the
matter and rejects the possibility of a connection between the ceremony itself
and the incidence of rainfall.
In the Chou Ii the ceremony is coupled with
dances that were performed by shamans.
For the Former Han period, there
is a record of this specific ceremony for 81 BC. For two other occasions of
drought (109 and 108 BC), prayers of a different type appear to have been
but the yu is specified for a number of occasions in Later Han (AD
113, 132, 145, 158, 161 and 176).
As it is sometimes specified that the
ceremony took place in the capital city and it is sometimes described as ta yu, it
seems likely that it was regarded as one of the major imperial cults. [Passages
(ii) (iv) and (v) in the addendum below refer to specific orders to pray which
were given to officials, including some of high rank].
In none of these passages is the use of clay or earthenware dragons
mentioned specifically as part of the rite. However, this appears in a number of
passages of the Huai-nan-tzu, as part of the ceremony for praying for rain. A
commentator's note ascribes the habit to T'ang, founder monarch of the
Shang dynasty, who is said to have fashioned these imitative talismans in time
of drought, in order that the clouds would accompany them and bring down
the rain. Elsewhere the Huai-nan-tzu observes that the manufacture of these
models was an act of the holy saints, which was comparable with their
provision of storage tanks at the time of a downpour.
In several highly informative passages the Huai-nan-tzu links the use of clay
dragons to procure rain with the fashioning of straw figures of dogs as a means
of seeking good fortune and forfending all evil and errors.
These images -
For the records of this ceremony on a number of occasions between 707 and 480 BC, see
Harvard-Yenching Institute Sinological Index Series (d), pp. 286, 1499. A few of these occasions
attracted sufficient comment from Han writers to warrant inclusion in the Han shu's treatise on
the Five Phases (wu hsing); see HSPC 27B(l).21aff.
11 Hsun-tzu 'Tien lun', p. 228.
12 Chou Ii 'Ssu wu' and 'Nii wu', SSC 26.7b and lOb, Biot, vol. II, pp. 102, 104.
13 HSPC7.5a,HFHDvol. II,p.161 (for81 BC); HSPC25A.38b(for 109 BC); HSPC25B.2b (for
108 BC). It is possible that this ceremony together with other methods featured in a lost work of
26 chuan whose title is entered in HSPC 30. 76a as Ch 'ing yu chih yu; see also CCFL 'Ching hua'
HHSCC 5.9b; 6.6b; 6.15b; 7.8b; 7.lOb; and 8.6a.
" HNT 4.8a, with note by Kao Yu (c. 168--212) and l 7.19b.
HNT ll.llb; see also HNT 16.13b and 17.la.
146 Divination, mythology and monarchy
and it seems that the text is referring both to the clay dragons and to the straw
dogs - were decorated in green and yellow; they were bound in figured or
embroidered silk and clothed in scarlet silk. Officials and prayer readers were
dressed in black garments, and counsellors of state wearing their official
headdress were in attendance as an escort for the figures. According to one
version of the text, once used the figures were discarded as being valueless;
possibly they were destroyed.
Philosophical or scientific considerations intrude into these matters in an
historical reference to Tung Chung-shu; but it is noticeable that as yet he is not
said to have made use of clay dragons. According to a passage in the Shih-chi
and the Han shu,1
in praying for rain Tung Chung-shu 'closed up the
openings which would admit the influence of Yang, the energy of the sun, and
opened up those orifices which would allow the entry of Yin, the energy of the
waters. To stop rainfall he took precisely the opposite steps.' The same
principle is inherent in a well-known chapter of the Ch'un-ch'iufan-lu that is
concerned with the subject of mutual attraction and thus involves the question
of sympathetic magic. The argument seeks to show how like attracts like and
repels unlike, in such a way that the Yin and the Yang forces that operate in
heaven and on earth correspond with the activation of those Yin and Yang
forces that are inherent in man:
Anyone who understands this principle and hopes to bring about a fall of rain will
therefore put Yin energies into action so as to stimulate Yin; and anyone who wishes to
stop rain falling will put Yang energies into action so as to stimulate Yang; and we find
that inducing a fall of rain is not due to the holy spirits (shen).
It will be shown below how this principle was elaborated. Two further
references to clay dragons require attention first. From the Hsu Han chih we
learn the following details of the ceremonial procedures of the court that
accompanied religious festivals and seasonal changes:
In times of drought the senior officials, drawn up in order of seniority, perform the Yii
ceremony as a prayer for rain. The Yang openings are closed; the officiants don black
silk and set up clay dragons. They erect two rows of earthep.ware figures of dancing
youths, which are changed once every seven days in accordance with precedent. The
altar of the soil is encircled with bands of scarlet rope and drums are beaten. After
prayers, they give thanks and sacrifice animals, in accordance with the prescribed rite.
From a fragment of the lost works of rationalist philosopher, Huan T'an
(43 BC to AD 28), we learn that no less a person than Liu Hsinh (46 BC to AD
23) made use of clay dragons to attract the rain, together with a number of
other devices.
Liu Hsin was one of the most highly advanced men ofletters of
See the notes to HNT l l .12a SC 121, p. 26, HSPC 56.19b.
CCFL 'T'ung lei' 57 .6a. [For a study of the idea of attraction or resonance in the Huai-nan-tzu,
see Le Blanc (1985).] 20 HHSCC (tr.) 5.laff.
21 See Pokora (1975), p. 121item129. Other devices mentioned include the use of pitch-pipes, for
which see Bodde, 'The Chinese magic known as watching for the ethers', first published Egerod
and Glahn (1959), pp. 14-35, 1959, rpt. Le Blanc and Borei (1981), pp. 351-72.
Cult of the dragon and invocation for rain 147
the day, dying in AD 23, and his precepts and practices were of no small
significance in the Chinese tradition. 'When dragons make their appearance',
Liu Hsiu is said to have explained, 'winds and rain rise up to welcome and
escort them; so clay dragons are made to simulate the phenomenon'.
Contemporary opinion and criticism
We may now consider two expressions of opinion regarding the use of clay
dragons and other means of attracting the rain, first from Wang Ch'ung and
secondly from Lang I (ff. AD 132). Usually in his discussions Wang Ch'ung
presents a rationalist exposition of a systematic order of nature which is
unaffected by personal decisions or the whims of a higher authority, and
which is in no way subject to human manipulation. In this instance his
reaction is somewhat unexpected. He mentions Tung Chung-shu's attempts to
invoke rain on no fewer than seven occasions; in six of these he refers to Tung's
use of clay dragons for the purpose, and it would seem that these are the
earliest ascriptions of this particular practice to Tung, apart from the
possibility that the Ch'un-ch'iufan-lu is earlier.
What is remarkable is that
Wang Ch'ung on several occasions expresses the view that Tung Chung-shu
was perfectly sincere in his belief, and he adds that there were distinct grounds
for that belief.
In one chapter, Wang Ch'ung adduces fifteen and a further four reasons to
support Tung's thesis. The force of the arguments and the tone of the chapter
are such that some scholars have called its authenticity into question, in view
of the flagrant contradiction with the general principles that Wang Ch'ung
enunciates so forcefully elsewhere in the book. However, the authenticity of
this particular chapter has been defended by Huang Hui (1935). He suggests
that it does not necessarily represent views which Wang Ch'ung held himself,
and he credits Wang with including them in the interests of setting out the case
Possibly Wang Ch'ung had himself witnessed an occasion when clay
dragons had been displayed and a shower of rain had duly followed, and he
felt obliged to record reasons for something which, in accordance with his own
methods, could only be regarded as a successful experiment. It should perhaps
be noted that Wang Ch'ung's account of Tung Chung-shu's practices does not
include the details given in the Ch 'un-ch 'iu fan-lu and considered below.
Somewhat later than Wang Ch'ung's time, Lang I is reported to have
LH 'Lung hsu' 22, p. 283, Forke (1907-11), vol.I, pp. 356f; 'Luan Jung' 47, p. 691, Forke
vol. II, p. 349; 'Kan lei' 55, pp. 787, 790, Forke vol. II, pp.17, 19; 'Ssu wei' 63, p. 889, Forke
vol. I, p. 206; 'Ting hsien' 80, p. 1100, Forke vol. II, p. 132; 'An shu' 83 p. 1162, Forke vol. I,
p. 465; 'Ming yu' 45, pp. 666, 676, 678, Forke vol. II, pp. 330, 336, 338.
See LH 'Luan lung' 47, p. 691, with Huang Hui's note to the title; see also Forke's comment,
vol. II, p. 349 note I.
Wang Ch'ung does not refer to the elaborate wu hsing associations that are incorporated in
CCFL 74; where he does refer to some of the practices that are mentioned here (for example,
the burning of shamans: LH 'Ting kuei' 65, p. 942; Forke vol. I, p. 246), he does not ascribe
them to Tung Chung-shu.
148 Divination, mythology and monarchy
mentioned the display of dragons that was practised at court along with other
measures that were taken at times of seasonal imbalance.
Lang I was a
well-known expert in making prognostications, and for noting oracular
messages that were conveyed by climatic phenomena (i.e., by the processes
known as feng-chiao ).
He rejects the measures that he mentions as being
quite ineffective and powerless to bring about climatic change such as a
shower of rain; forit is August Heaven (Huang t'ien) that controls the balance
of the seasons and the changes of climate, which remain unaffected by human
deceits, pretensions, or manipulations.
The major context of the practice
The use of earthenware dragons to induce rain to fall may be set in context in
several ways. It must be considered along with other measures thought to
procure this result, and with other ways in which dragons took part in prayers
for rain. In addition the manufacture of clay dragons should be compared
with that of other images used for other purposes by way of sympathetic
The use of clay dragons is mentioned in a small treatise entitled Tao-yii
tsa-chi, compiled by Ch'ien Ch'i, with a prefatory note dated 1545.
The work
includes an account of a number of incidents in which invocations were
offered together with a whole host of other methods. The use of clay dragons is
recorded here for as late as the eleventh century. Other devices included the
offering of supplications to deities, often of a local type, and the performance
of the religious or magical dance known as the 'Steps ofYii' ( Yii pu), as will be
observed below.
From some of the manuscripts found at Ma-wang-tui we know that the Yii
pu took its part among incantations designed to exorcise evil influences that
had caused illness.
As a means of inducing rainfall, it features in other texts
as follows:
Specialists from the western regions who are capable of uttering spells and incantations
stand at the side of a deep pool and perform the Steps of Yu. As soon as they breathe
out, a dragon emerges, floating on the water, measuring'several tens offeet in length.
When the specialist breathes out again, the dragon promptly shrinks to a few inches,
and it is then collected and placed inside a vessel. There may be as many as four or five
and they are fed with water, sparingly. When there is news of an area that is afflicted by
drought, the dragons are taken there to be sold, and a single one may fetch some tens of
units of gold. When the vessel is opened, a dragon is let loose into a pool. The specialist
HHSCC 30B.14b. For Lang I, see de Crespigny (1976), pp.
HHSCC JOB. la. For Jeng chiao, see Ngo van Xuyet (1976), pp. 186f and chapter 9 below.
The Tao-yii tsa-chi is most easily available in the TSCC reprint from Pai ling hsiieh shan, to
which reference is made here. For an account of this evidence, with other means of inducing
rainfall, see Cohen (1978). The author of the work, Ch'ien Ch'i, may possibly be identified as a
person who achieved his chin-shih degree in 1508. 28 See Loewe (1981), p. 193.
Cult of the dragon and invocation for rain
performs the Steps of Yu once more; as he breathes out, a dragon measuring several
tens of feet in length emerges, and shortly the rainclouds rise up from all directions. 29
A further means of inducing the rain to fall was that of exposing a suppliant
or shaman to the full heat of the sun, or to man-made fire. A reference to the
practice is recorded in China for 639 BC.
At a time of severe drought the
Duke of Lu proposed to burn a shaman and an emaciated person, but was
dissuaded from doing so. Similar practices are cited by Frazer, who writes of
the exposure of a deity named Lung Wang near Canton, in 1888, and of the
Japanese custom of casting a guardian dog into an arid rice-field, with the
exhortation that he should suffer the heat himself. 31
A number of explanations are offered for this practice. Yoshinami32
suggests that it derived from the self-sacrifice that a ruler was ready to make,
by way of propitiation, and that responsiblity for self-sacrifice in times of
drought was transferred to a shaman. Elsewhere there is an instance of the
same principle, where a local magistrate, being responsible for the welfare of
his flock, is alleged to have been ready to make the supreme sacrifice by
throwing himself upon a flaming pyre; in this way he would bring moral
pressure to bear on a local god so that he would relent and provide rain. 33
The reference to this practice is in a text that is dated after the Han period,
but it cannot be said how old the custom may have been or how prevalent it
was in different parts of China. Possibly intermediaries were exposed to the
heat of the sun, or of a man-made fire, in order to induce them to redouble
their efforts at intercession, as a result of personal severe suffering. We shall
revert to this practice later.
Other methods of bringing pressure to bear on the gods are also recorded.
As a first step, their titles were revoked as a means of showing them that they
had forfeited the right to such symbols of power and dignity, and that they
must take positive action so as to regain them. As a further step, threats could
be uttered to damage the precincts of their shrines. If the local gods were still
obdurate, their very images could be exposed; and, as a final and desperate
step, these images could be smashed. 34
The translation follows the text given in Tao-yii tsa-chip. 8, with some corrections in view of the
readings.given elsewhere, for example, TPYL l l.6a, 736.Sa and 929.7b. Possibly the earliest
account 1s to be found m the Chin-lou tzu of Hsiao I (Liang Yiian-ti; born 508, reigned
TSCC ed. 5(10).93. I am mdebted to Mr Matthew Henderson for drawing these references to
my attention.
See Tso chuan SSC 14.26b, Legge, p. 179, Couvreur (1914), vol. I, p. 327; see also Li chi, 'T'an
kung' SSC 10.32a, Couvreur (1913), vol. I, p. 261. For practices during the Shang period see
Ch'en Meng-chia (1936), Qin Xigui (1983) and Allan (l 984a), pp. 528f. For a study of
to heat as a means of invoking a fall of rain, see Schafer (1951). [For literary references to
T'ang's willingness to present himself as a sacrifice in order to assuage a drought, see LSCC
9.3b, 4a and HNT 9.Sb, Ames (1983), p. 173.]
" See Frazer (1911), vol.I, p. 299 (citing from E. Z. Simmons, 'Idols and Spirits'; Chinese
Recorder and Missionary Journal 19, 1888, p.502); see also p.300 for reference to a similar
practice in Sicily.
Yoshinami (1978), with reference to LSCC 9.3b.
" TPYL 529.?a, b, citing from Ch'ang Ch'ii, Hua-yang kuo-chih.
Cohen (1978), pp. 247f.
150 Divination, mythology and monarchy
If dragons were not used in the form of clay models, they sometimes
featured as paintings. Wang Ch'ung refers to an early instance of this in the
state of Ch'u. Much later, at a time of drought in the K'ai-yiian period
(713-41), we hear of an official who had a single white dragon painted on the
walls of his office; immediately a dragon arose from the lake and mounted the
clouds, and the winds and the rain duly followed.
In addition, Frazer gives
the following account of the use of an artificial dragon in Japan:
In Okunomura, a Japanese village not far from Tokio, when rain is wanted, an
artificial dragon is made of straw, reeds and bamboo, and magnolia leaves. Preceded
by a Shinto priest, attended by men carrying paper flags, and followed by others
beating a big drum, the dragon is carried in procession from the Buddhist temple and
finally thrown into a waterfall.
There is also a reference to the performance of a dance with dragons that is
of some interest in view of the occurrence of this rite in the Ch 'un-ch 'iu fan-lu,
as will be described below. There can be little means of assessing the antiquity
of the allusion, and in so far as it ascribes a custom to the remote past, the
reference may be suspect as being an anachronism. We read that no less a
person than Shen Nung resorted to this device once, when he had suffered a
drought that had lasted for nineteen days. In addition to ordering the
manufacture of Yellow Dragons, he had an outsize dragon made with which
dances were performed by fully grown men.
The use of clay dragons may also be compared with the use of images of
other types. Reference will follow below to the use of earthenware suns, made
as a means of preventing rainfall. We also hear of a stone bull that was used for
inducing rainfall. This image stood within a pool to the south-east of the
mountains ofYii-lin commandery, in south-west China. At times of drought,
the inhabitants would slaughter a bull in order to pray for rain; the bull's
blood would be mixed with mud, and the mixture was then daubed on to the
back of the stone bull. When the prayers had been completed, the rain would
start to fall, and then to pour down, ceasing only when the mud on the stone
bull's back had been cleaned off.
The motive for this practice was presumably to induce the god to remove
something which was both polluted and polluting. The mixture of blood and
mud could be nothing but offensive to a deity, particularly if it was placed
deliberately on a sacred image; and it could be hoped, or even expected, that
35 LH 'Luan lung', p. 692, Forke (1907-11), vol. II, p. 349; Tao-yii Isa-chi pp. 7 and 12.
36 Frazer (1911), vol. I, p. 297, on the basis ofR. Lange, 'Bitten um Regen in Japan' (Zeitschrift
des Vereinsfiir Volkskunde, iii, 1893, 334f), and W. G. Aston, Shinto (The Way of the Gods),
(London: Longmans, Green & Co., 1905), p. 153.
37 Cited by Hui Tung (1697--1758), in a note to HHSCC 5.2b.
3s Tao-yii tsa-chi p. 5; it is not possible to date this reference precisely; a commandery of Yii-lin
was founded in Former Han and existed in Later Han.
Cult of the dragon and invocation for rain 151
the god would take immediate steps to purify the spot. Once more there is a
comparable rite in Japan, as quoted by Frazer:
Among the high mountains of Japan there is a district in which, ifrain has not fallen for
a long time, a party of villagers goes in procession to the bed of a mountain torrent,
headed by a priest, who leads a black dog. At the chosen spot they tether the beast to a
stone, and make it a target for their bullets and arrows. When its life-blood bespatters
the rocks, peasants throw down their weapons and lift up their voices in supplication to
the dragon divinity of the stream, exhorting him to send down forthwith a shower to
cleanse the spot from defilement.
Finally, the use of clay dragons may be compared with the use of clay bulls,
which appear in Han times as a means of welcoming the spring and
encouraging the new season of growth. According to a passage of the
at the annual ceremony of the Great Exorcism, dogs and
sheep were sacrificed in order to eliminate evil influence, and clay bulls were
set up in order to inaugurate the season of ploughing. In one version of the
text, the bulls were accompanied by clay images of ploughmen. Professor
Bodde has suggested that this whole rite was a surviving element of a sacrifice
of bulls rendered as a means of fructifying the earth.
The elaboration of the cult
We may now consider the description of the rite for inducing rain that appears
in p'ien 74 of the Ch'un-ch'iufan-lu, entitled 'Ch'iu yii'. It will be seen that the
rite is set out in a schematic form, with signs of a considerable debt to
philosophical principles hardly formalised in Tung Chung-shu's own life-time
(c. 179 to c. 104). For this reason it may be supposed that, although the
description may call on actual practice, it may have been composed anach-
ronistically (some centuries later), at a time when considerable standardisa-
tion had set in. [An assumption that Tung Chung-shu had personally
performed the rite is carried in passage (iv) of the addendum below.]
The chapter describes the steps that should be taken to procure rain for each
of the four seasons of the year and also for the fifth point of chi-hsia, the final
part of summer.
For each of these five occasions the chapter sets out a list of
items of procedure. While the purpose and type of activity is identical (for
example, prayer or sacrifice), the details that are prescribed in each of the five
cases vary according to the characteristics of the particular season for which
39 Frazer (1911), vol. I, p. 291, citing from W. Weston, Mountaineering and Exploration in the
Japanese Alps, (London; John Murray, 1896), pp. 162f.
40 HNT 5.16b, with Kao Yu's note for the specification of dogs and sheep.
41 See Bodde (1975), pp. 20lf. For the manufacture of clay oxen in Later Han, see Cheng
Chieh-hsiang (1963).
For the need to find a fifth season, so as to accord with the wu hsing, and the solution of this
problem, see Bodde (1975), p.192.
152 Divination, mythology and monarchy
provision is being made. In this way the prescriptions are made to accommo-
date the major view of the five stages or phases of the major process of birth,
death, and rebirth. Thus, the colours of the robes that are worn, the number of
the participants in the rite and their age, and the measurements of the dragons
accord with the characteristic colours, numbers, and other details associated
with each particular phase of the wu hsing. Similarly, certain activities are
banned, so as to avoid running counter to the natural features of a particular
phase; and the rites are carried out in the appropriate position or quarter of
the site.43
The ceremony for spring is the fullest to be treated, including more items
than those recorded for the other seasons. It is therefore taken as a basis for
the following account of the principal features of the invocation for rain that is
described in the Ch 'un-ch 'iu f an-lu.
Prayers are offered by the inhabitants of a locality to its deities, which may
include the gods of the soil and the grain, and the lords of the mountains and
the rivers. A ban is imposed on felling specially named or well-known trees,
and on chopping timber from the mountains; this is matched in other seasons
by other prohibitions, such as that on raising fire in the autumn. Next, a
shaman (or shamaness) is exposed to the heat of the sun. Attention has been
paid above to the possible motives for this practice, which attracted the
attention of Wang Ch'ung. In one passage he suggests that it was due to the
way in which shamans were imbued with Yang and the need to eliminate
excess Yang at a time of drought. But Wang Ch'ung seems to question the
efficacy of the action, and the explanation may be no more than an
anachronistic rationalisation.
As the next step, altars were erected at an appropriate part of the site and
prayers were offered to other deities of a more universal nature, such as Kung
Kung in the spring (outside the east gate) and Ch'ih-yu in the summer (outside
the south gate). On these occasions, offerings were made of fish, alcoholic
spirits, or dried meat; a specially gifted shaman, distinguished for his or her
purity, ability to speak, and powerful delivery, was chosen to recite the prayer
of intercession. For three days the shaman would keep vigil and fast by way of
preparation. Thereupon, clothed in the appropriately coloured robes for the
season (for example, green for spring), he or she would make obeisances and
kneel, praying as follows:
Almighty Heaven that hath given growth to the five crops in order to sustain mankind,
the withering of these crops by drought that we now suffer is such that they are unlikely
to grow to their due fulfilment. W c reverently bring out offerings of pure wine and dried
meats and twice prostrate ourselves in entreaty for the rain, praying that it may fall in
abundance in its due season.
For the correspondence of various categories so as to suit the wu hsing, see SCC vol. II,
pp. 262f.
CCFL 74.5bff .
., LH 'Ming yu' 45, p. 664, Forke (1907--11) vol. II, p. 329, and LH 'Ting kuei' 65, p. 943, Forke
vol.I, p. 247.
Cult of the dragon and invocation for rain
The next part of the ceremony took the form of the dance of the dragons.
These were fashioned in varying sizes, with one major dragon set up at the
centre of the site and several minor ones being arrayed in one of the four
quarters. The size, number, and colour of the dragons, and other details, were
specified as follows for the five occasions of the year:
Spring Summer Midsummer Autumn Winter
Length of major dragon
(feet) 80 70 50 90 60
Number (length) of minor
dragons 7 6 4 8 5
(40ft) (35 ft) (25 ft) (45 ft) (30 ft)
Colour of major dragon green red yellow white black
Position of minor dragons east south south west north
Direction faced by major
and minor dragons east south south west north
Interval between
minor dragons 8 ft 7ft 5 ft 9ft 6 ft
It will be noted that for midsummer the position specified for the four
acolytes of the yellow dragon was at the south of the site, and not, as might
have been expected, at the centre. Possibly this detail may imply that the
central part of the site was too confined to accommodate the dance, which was
regulated by precise prescriptions. It was performed by eight youths in the
spring, seven able-bodied men in the summer, five adults in midsummer, nine
widowers in autumn and six elderly men in winter. Prior to the dance, all the
performers would fast for three days, after which they would don robes of
the appropriate colour for the season. Specially designated officials, such as
the overseer of the fields (t'ien ssu-fu) were commissioned to set up the dragons
for the purpose.
As the dragons were presumably designed to play a mobile part in the
dance, it is to be assumed that they were made ofa flexible frame, covered with
a coloured fabric. One scholar, however, assumes that these large models were
of earthenware, but this is unlikely.
In addition to the difficulty of
manipulating such models in an active dance, the specific use of clay dragons is
mentioned separately, later in the text of the Ch 'un-ch 'iu fan-tu. Moreover, as
far as may be known, no surviving remnants have been found of models of
dragons that could have been made to such a size for this purpose. [Very long
dragons made of bamboo frames dressed in textiles and illuminated internally
are skilfully and speedily manipulated in accompaniment with the dances of
certain exorcist rites even now.]
At the next stage of the rite holes are dug in the shrine dedicated to the soil
so as to connect with watercourses that lie outside the gates of the village.
De Visser (1913), p. 115.
154 Divination, mythology and monarchy
Thereafter frogs are collected and deposited within the shrine dedicated to the
spirit of the soil, being introduced into tanks that were prepared carefully,
with due account of the numbers appropriate to the season. For spring,
summer, and possibly winter, five frogs were collected. They were set in tanks
made to measure for each season, being respectively eight, seven, five, and nine
feet square for spring, summer, midsummer, and autumn; for the winter the
text merely writes 'as in spring'. Each tank was to be one foot deep.
Frazer draws attention to a widespread connection established between
frogs or toads, and rainfall.
He suggests a number of explanations, to which
may be added the generally held belief that croaking frogs foretell a shower of
rain. This is found in a number of sources and cultures. For the Roman world,
the belief is mentioned by Cicero, who writes to Atticus (44 BC) 'Besides, I am
afraid that it is going to rain, if there is any truth in prognostication; for the
frogs have been talking like orators.' He also alludes to the subject in a passage
on divination, which was taken up by Edward Topsell in 1608:
When Frogs do croak about their usuall custome, either more often, or more shrill than
they were wont to do: they do foreshew raine and tempestuous weather.
Wherefore Tully saith in his first book of Divination, who is it that can suspect, or
once thinke that the little Frogge should know thus much, but there is in them an
admirable understanding nature, constant and open to it selfe, but more secrets
obscure to the knowledge of men; and therefore speaking to the Frogs he citeth these
Vos quoque signa videtis aquai dulcis alumnae,
Cum clamore paratis inanes fundere voces;
Absurdoque sono fontes et stagna cietis.
In English thus:
And you 0 water-birds which dwell in streames so sweet,
Do see the signes whereby the weather is foretold,
Your crying voyces wherewith the waters are repleate,
Vaine sounds, absurdly moving pooles and fountaines cold.
The same belief may be noted in other parts of the world, such as Africa;
and the connection between frogs and wet weather appears in a charming
folk-tale from Korea, whose text is given in the appendix below.
The rite proceeds with offerings, fasting, and prayers. For the spring
ceremony, three year old cocks and three year old pigs are chosen and burnt at
Frazer (1911), vol. I, pp. 292f.
Cicero, ad. Att. 15.16b 'equidem etiam pluvias metuo, si prognostica nostra vera sunt; ranae
enim p11wpevo1Ju1v'; see Shackleton Bailey (1967), pp.104-5; Topsell (1608), p. 183; Cicero, de
divinatione I. ix (translation by William Armisted Falconer, in the Loeb Classical library, XX,
Cicero, 238-9). Virgil mentions the croaking offrogs as a sign of impending rain, along with the
behaviour of cranes, who fly before it, heifers who gaze up to the skies in hopeful expectation,
and swallows who flit round the pools of water (Virgil, Georgics I, 378 'et veterem in limo ranae
cecinere querellam'); see also Pliny, Natural History XVIII, 87.
For the belief in the frog's capacity to affect rainfall, as held in certain parts of Africa, see the
entry 'Rain frog' in Burton and Burton (1969), vol. XIV, pp. 1905-<i.
Cult of the dragon and invocation for rain
the shrines of the four quarters. Orders are given to close the southern gate of
the settlement or village and water is placed outside; the northern gate is
opened. One of the pigs is placed outside the northern gate and another in the
market-place, and at the sound of the drum's beat their tails are set alight. The
ceremony for summer was similar, but no prescriptions are given for
midsummer, autumn, or winter.
In the spring, unburied human bones are collected and buried. The belief
here would seem to be that unburied bones suffer unduly from rainfall and are
wont to complain so vociferously that the souls which once inhabited them
intercede with higher authorities so as to prevent further showers. Burial of
the bones would preclude such suffering and the need for intercession, and a
possible cause of drought would be eliminated.50
The final steps for the spring include opening the springs on the mountains
and collecting firewood, which was then set alight. All obstructions to free
passage at bridges are removed and blocked watercourses are freed. Ifby then
the rain has happily begun to fall, further offerings are made of a pig, wine, salt
and grain; and finally mats are woven of thatch: on no account should this be
The chapter of the Ch 'un-ch 'iu fan-lu ends by specifying two observances
which were to be kept for each of the four seasons. On appropriate days,
dragons were to be made from pure clay and exposed; and married couples,
whether officials or commoners, were to mate together in sexual intercourse.
This final injunction can readily be explained as a means of inducing natural
harmony between Yin and Yang forces, and thus procuring the seasonal fall
of rain.
It may be contrasted with injunctions found elsewhere to abstain
from sexual intercourse at the vernal equinox and at the summer and winter
solstices; these last two occasions were characterised as times when Yin and
Yang were in active contention to achieve dominance. 52
Methods of preventing rainfall
Following the section whose content is summarised above, the Ch'un-ch'iu
fan-lu includes a section which is entitled 'Chih yii' (no. 75), 'Stopping the
rain'. In some editions of the book the text is in a defective state, and different
passages have been inserted by various editors by way of reconstruction. The
following account is based on the version which is printed in Su Yii's edition.
See de Groot (1892--1910), vol. III, pp. 918f.
See Derk Bodde, 'Sexual Sympathetic Magic in Han China', first published 1964, rpt. Le Blanc
and Borei (1981), pp. 373--80.
The Ssu min yiieh ling provides that this should be avoided for five days before and five days
after the vernal equmox, summer solstice and winter solstice; no proscription is included for the
autumnal equinox. A n?te to the prohibition for the vernal equinox suggests that it was
mt_ended to avmd the buth of deformed children. See Shih Sheng-han, Ssu min yiieh ling
chzao-chu2: 5, 5:6and 11: 1(pp.220, 44and 71); translated in Hsu Cho-yun (1980), pp. 218, 222
and 226.
156 Divination, mythology and monarchy
In cases when the rainfall is excessive, officials of the county or locality make
earthenware suns with which they stop up the channels; they cut off the
watercourses and cover over the wells. Women are forbidden to travel or to
make their way to the markets, and the shrines dedicated to the soil are all
cleansed. Officials of a rank and dignity suitable to each locality then fast for
three days. They all don robes coloured appropriately for the season and make
offerings of pigs, grain, salt, and wine to the shrine. The drums are beaten for
three days, and prayers are then offered with suitable obeisances, ending with
the following invocation:
Heaven that has created the five crops to sustain mankind is now pouring down a
surfeit of rain, such that the five crops are not in harmony. We reverently proffer our
fattened animals and pure wine and thereby ask the divine powers of the shrine
favourably to cause the rain to cease, that the people may be relieved of suffering. We
ask that Yin be not allowed to destroy Yang, for if Yin destroys Yang, there remains no
accord with Heaven; and it is the constant will of Heaven to bring benefit to man.
Intimation to the shrine of man's desire to bring the rain to a stop is
accompanied by the drum but not by human song, and it is only when the
procedures have been carried through to the end that they are brought to a
The Yang openings, i.e., the gates on the south side, are opened and
the Yin openings are closed, so as to shut off the waters and admit the heat or
fire. The shrine is surrounded with ten circlets of red silk, and [the officiant]
wears red clothes and vermilion headgear.
There may possibly be records of the use of red silk for this purpose by
officials who were ordered to stop the rain by this means, during the reigns of
Wu-ti (141--87 BC) and Ch'eng-ti (33--7 BC).
In addition, the chapter of the
Ch 'un-ch 'iu fan-lu next includes a factual account of an attempt to stop the
rain by the methods that have been described. This is dated precisely to a day
corresponding with 10 November 134 BC, when we are told that [Tung]
Chung-shu, chancellor of the kingdom of Chiang-tu,
notified some of his
subordinates that rain had been falling to excess, with the consequent fear that
the crops would be damaged. He therefore enjoined them to prevent the rain,
by carrying out the ceremony of preventing Yin and arousing Yang. Written
notification was duly sent to seventeen counties and eighty detached districts,
and orders were given to forbid womenfolk from going to market. In the
market, people were not to proceed to the wells, which were to be covered so as
to prevent overflow. The procedure of beating drums and sacrifice was to
follow, and the prayer whose text has been quoted above was to be recited.
" Alternative rendering: 'Only when the procedures have been carried through to the end does
the rain stop'.
See passages (iv) and (v) that are cited in the addendum below, and Otto Franke (1920),
Zweiter Tei/, pp. 268f.
" See HSPC 14.14b for the succession of kings of Chiang-tu, which existed from 153 to 121 BC,
and again from AD 2 for five years, under a different title. For Tung Chung-shu's appointment
as chancellor, see HSPC 56.19a.
Cult of the dragon and invocation for rain 157
Within five days the notification had been duly received at all the
subordinate offices and immediate steps were taken to order the local officials
to proceed to the shrine. They were to carry out these procedures ending only
at the hour of hsia pu,
and after three days they were brought to a close.
Before three days had passed the skies had cleared and the rain had stopped.
The procedures laid down for stopping the rain were far less elaborate than
those for inducing it to fall, and far less subject to the regularisation imposed
by deference to the wu hsing frame of mind. In addition, the account of an
incident, seemingly dated correctly, in which Tung Chung-shu was personally
concerned, adds some degree of veracity to the authenticity of the text, whose
defective state may be a further reason why it can be accepted as being free
from later correction or amplification. Possibly it may be concluded that while
this section has been free of later elaboration, the previous section, on
inducing rainfall, includes material of a time later than Tung Chung-shu.
The evidence that has been examined includes elements of early Chinese belief
and mythology. In areas where the gift of seasonal rain, the danger of drought,
and the calamity of flood were of paramount importance, it is hardly
surprising that the subject features with such prominence, and that the dragon
cult formed but one element in a variety of practices. It may also be seen how
the elaborate procedures bear the imprint of a sophisticated mode of thought
and a desire to conform with its principles. A similar process wherein the
intellect has imposed its own categories on acts of faith or instinct may be
observed in connection with divination.
Nevertheless, the evidence remains
as a remarkable example of the practice of sympathetic magic in early China,
whose continuity may be witnessed today. It may be seen, for example, in the
dragon dances performed for theatrical purposes by actors from Ssu-chuan; in
the annual ceremony that takes place in Nagasaki, as a survival of Chinese
influence; and above all the Dragon Boat festivals that have been regularly
enacted in south China. 58
Once upon a time there lived a green frog who would never do what his mother
told him. If she told him to go to the east he would go to the west. If she asked
This was the ninth of the twelve hours into which the period of day and night was divided. See
Loewe (1967), vol. II, p. 20.
See Loewe and Blacker (1981), pp.40f; and Loewe (1982), pp.91f.
" See Schneider (1980), p. 148. In recent times actors from Ssu-ch'uan have performed the
dragon dance as a finale to a theatrical display (for example, at the Dominion Theatre,
London, in July 1981).
I am grateful to Routledge and Kegan Paul Ltd for permission to reprint this tale from Zong
In-sob (1952), pp. 34-5.
Divination, mythology and monarchy
him to go up the mountain he would run down to the river. Never, never
would he obey his mother in anything.
At last his mother grew very old, still worrying about her son's future. At
last she fell ill, and realised that she was about to die. So she called her son to
her bedside and said to him, 'My dear son, I shallnotlivemuchlonger. When I
die do not bury me on the mountain. Do you hear? I want to be buried by the
river.' She meant of course that she wanted to be buried on the mountain, for
she well knew her son's perverse ways.
Very soon afterwards she died. Then the green frog was very sad and wept
bitterly. He repented of all his misdeeds in the past, and made up his mind that
now at least he would do as his mother had asked. So he buried her by the
riverside. And whenever it rained he worried lest her grave should be washed
away. He used to sit and lament in a mournful voice. And to this day the green
frog croaks whenever the weather is wet.
A few fragments of the Han chiu i, which is ascribed to Wei Hung (first century
AD), lend corroboration to some of the suggestions that have been made
above. The mention therein of some of the details that are included in the
essays of the Ch 'un-ch 'iu fan-lu is of particular interest in view of the doubts
regarding the authenticity of that text. The passages are assembled together in
the short collection Han chiu i pu i B, in Han kuan liu chung (B 4a--5b) and Han
kuan ch'i chung (B 6a-7b).
(i) Citedfrom T'ai-p'ing yii-lan but not found therein
It is in the summer, when the Dragon star (Lung hsing) is seen, that the yii
ceremony is first performed.
The text continues with details of the Lung hsing, also termed Ling hsing,
some of which are recorded in (i) The So-yin note ofSsu-ma Chen (early eighth
century) to Shih-chi 11, p. 16; and (b) the Cheng-i note of Chang Shou-chieh
(Ji. 737) to Shih-chi 12, p.43.
(ii) Hou Han shu chi-chieh (tr.) 5.lb, note by Liu Chao (Ji. 502--20)
In the prayers for rain, the superintendent of ceremonial (T'ai-ch 'ang) prays to
Heaven Earth the ancestral shrines, the shrines dedicated to the soil and the
crops, the and the rivers to beseech blessings, in each case in a
manner like the rite of the regular sacrifices.
The text continues with prescriptions for the timing or repetition of the rite,
from the fourth month onwards.
Cult of the dragon and invocation for rain 159
(iii) Hou Han shu chi-chieh (tr.) 5.3a, note by Liu Chao
This passage, which appears to be defective, seems to refer to a drought or
droughts which occurred during the Yiian-feng period (110-105 BC). These
are recorded for the summer of the fourth year ( l 07 BC) and the autumn of the
sixth year (105 BC); see HSPC 6.28b and 31a, HFHD vol. II, pp. 94, 98. The
meaning is far from clear, except for the statement that prayers were not
offered at certain seasons.
(iv) T'ai-p'ing yii-lan 526.3a, with a slightly different text in Su Yii's note
to CCFL 'Chih yii' 75.13a
The first two characters of the passage, for which no translation is offered, are
regarded as being corrupt.
... first year, the scholars and professionals (ju shu) submitted a suggestion for the
performance of Tung Chung-shu's request for rain. For the first time the chancellor
and subordinate officials were ordered to pray for rain. There was a violent snowstorm
south of the city wall. Dancing youths and girls prayed to Heaven, the holy spirits and
the Five Powers. In the fifth year
orders were first given to all offices to [pray] for the
rain to cease. Scarlet ropes were bound to encircle the shrine dedicated to the soil, the
drums were beaten and it was struck.
( v) Hou Han shu chi-chieh 5 ( tr.) 3a note by Liu Chao
In the sixth month of the second year
of Ch'eng-ti, for the first time the
various offices were charged with the duty of praying for rain to cease. Scarlet
ropes were bound so as to encircle the shrine dedicated to the soil, the drums
were beaten and it was struck. Thereafter water and drought were constantly
out of balance.
60 Su Yii suggests reading ch'eng nien rather than wu nien.
61 Su Yii reads chub in place of kung.
This is given as the third year in T'ung tien 43 (Li 3), p. 249B, and in the Chung-hua shu chii
punctuated text of the Hou Han shu, p. 3120. In his note to Liu Chao's commentary, Hui Tung
(1697--1758) cites a north Sung version which reads fifth year. The order that is mentioned is
not recorded in Han shu eh. 10. If the date is to be taken to refer to the years of Ch'eng-ti's reign,
as enumerated from his accession and without specification of a nien hao, the year in question
would have been 31, 30 or 28 BC.
None of the nien hao that were adopted under Ch'eng-ti remained in force for more than four
years, and the reading of the fifth year may perhaps be questioned. However, there is reason to
show that on some occasions obsolete terms were still used, possibly because orders to adopt
the new one had not been fully circulated; see Loewe (1959), 31 Sf, and ( 1967), vol. I, p. 136 note
27, and vol. II, p. 73 note 4.
Divination by shells, bones and stalks
during the Han period
Sources of information
Our knowledge of divination by means of shells and bones (pu) during the Han
period is bound by completely different considerations from those that affect
earlier times. For the Shang-Yin and earlier ages, research depends almost
exclusively on the material remains of the process .itself;. there is no
from contemporary documentary evidence compiled to. its
mysteries and forms, or to record in chronicle form m which the
process played a part. For the Han period, while archaeologists have yet to
identify examples of shells or bones actually use.d for the and
inscribed accordingly, it is possible to call on a considerable body of hte.rary
evidence, dating from both pre-imperial and imperial times, and at times
scattered among different texts. . .
For sources that were compiled in their original form before the impenal
period, attention focusses on the various compendia Ii, a.re concerned
not only with the use of bones and shells, but also with d.iv!nat10n by stalks
(shW); there arises the immediate difficulty of determmmg how far . the
references in such books may be taken as having a bearing on Han
For while these texts purport to describe the practices o_fthe kings of it
is only too likely that at times they may referrmg to
continuing even into the imperial age. In add1t10n, the respect m which such
books were held during Han times was such that. they v.:en have
constituted a framework within which much of official Han thmkmg took
place. . fd' b
When the compendia on li were being written,
the practice o 1vmat10n Y
shells or bones was already at least a thousand years old, and the use
stalks was not much younger. Some measure of stylisation had long smce set
in, with a rigorous set of procedures taking the place of the sp?nt_aneous
actions and reactions of a seer. It may even be suggested that by this time the
original motives for the practice had become outmoded, and that much of the
1 For authorship and history of the four principal texts, i.e., Chou Ii, I Ii, Li chi and Ta Tai Ii chi,
see the entries in Loewe (1993); references given below for the first three of these texts are to
Divination by shells, bones and stalks 161
procedure was being conducted without an understanding of its purposes. In
such circumstances the force of a written set of rules for the ritual may well
have acquired an overwhelming and disproportionately high influence,
whatever the motive may have been for the compilation of such documents.
Possibly the rules and procedures were set down in writing as a means of
asserting the permanent nature of certain considerations that transcend the
ephemeral lives and rule of individual men. Possibly the books on Ii were
intended to form a defence of certain aspects of human behaviour, on
intellectual grounds; possibly they were compiled as handbooks for consulta-
tion, or to ensure the maintenance of orthodox procedures and ceremonies.
Whatever the dates may have been when the compendia on Ii were
originally written or reached their present form, the terms of reference and the
technicalities that they describe partake of the pre-imperial age, and it must
remain open to question how far this may have been due to deliberate
anachronism. We cannot tell how far these books were describing practices
that had never been operated; or procedures that had long become obsolete;
or rites that survived from the Chan-kuo period until Han times. It may
however be concluded that they formed an integral part of the background to
Han thought, perpetuating a tradition that affected the training and intellec-
tual outlook of Han officials. Recognition of the importance of these texts is
seen in their inclusion among the Five Classics, and the appointment of
Academicians to specialise in their exposition (136 BC), and in the citations
that appear from time to time in statements attributed to officials. 2 If an
analogy may be risked, it may be asked whether the influence of these classical
or scriptural texts was in any way parallel with that of the ritualistic portions
of the Old Testament in Victorian England. Attitudes may well have varied
from fundamentalist acceptance to criticial scepticism or rejection; but at each
extreme, the texts bore great intellectual significance.
Sources which date from the imperial period include two chapters of the
Shih-chi, each with an appended supplement by Ch'u Shao-sun (?104-?30
Chapter 127 sets out to vindicate the profession of diviners and their
standards of honesty. Chapter 128 gives a short historical note on the practice,
followed by a considerable body of technical information concerning the
qualities and properties of turtles, and a few anecdotes. The chapter also
includes catalogues of the types of crack that appear on the shells during
divination and the types of question that may be put to this source of occult
Records of actual incidents of divination, or of regular occasions for its
performance, occur in the Shih-chi, Han shu and Hou Han shu. References to
the officials whose duties were concerned with these practices and with the use
of stalks are seen both in these works and in fragmentary texts such as the Han
For the relevance of these books to Han times, see Bodde (1975), pp. 7, 15, Tjan Tjoe Som
(1949-52), vol. I, pp. 82f and Fujikawa (1968).
For Ch'u Shao-sun, see Pokora (1981).
Divination, mythology and monarchy
kuan. 4 In addition, the extent of criticism, usually adverse, to which the
practices gave rise suggests that they were matters of regular occurrence which
thinking minds could not ignore. Such criticism is seen, for example, in the
Huai-nan-tzu, Yen-t'ieh fun, Han shu (Chapter 30), Lun-heng and Ch 'ien-fu fun,
and in the writings of Chung-ch'ang Tung (c. 180-220); they are thus spread
in time from perhaps 150 BC to AD 200.
The powers of the turtle and the yarrow stalks: faith in the practice
Some of the available evidence suggests the existence of a deep-rooted belief in
Han times that divination by turtle shells
or stalks was an ancient traditional
practice, and that it was the continuation of an age-old hallowed rite. In
discussing the stress placed on various forms of religious observance, the
Li-chi6 refers to the reliance that the kings of old had placed on
these means, and there are similar references in the Shih-chi. In one passage tli,e
latter text mentions that the holy kings of the past had always performed this.
precautionary rite before establishing their regime, accepting their charge 1Q
rule or initiating amajorproject.
At a later stage we hear of a senior official of
government citing such examples by way of admonition. This was when Fu
Chan, appointed minister of finance (Ta ssu-t'u) shortly after Kuang-wu-ti's
accession (AD 25), was advising the new emperor of the dangers of personally
leading a campaign against a rebel. Fu Chan drew specific attention to King
Wen's example of divination by both methods in comparable circumstances.
In an important passage, the Li-chi refers to the powers of the shells imd the
stalks as instruments which may communicate the will of the holy spirits of
heaven and earth.
Elsewhere the same book refers specifically to the way in
which the rise or fall of a state may stand revealed in these objects' signs;
this passage occurs in the 'Chung-yung' chapter, its message found a
prominent place in the regular curriculum of Chinese education during the
imperial era. A reference in the Ta Tai li-chi
points out that the powers of
prognostication that the turtle, the most refined of all creatures, possesses
depend on the application of fire. In regulations for setting out items of
equipment or other goods for banquets, or those brought as items of tribute,
the turtle was sometimes given priority, owing, it 'was claimed, to its gift of
prior knowledge.
In a completely different type of writing that may well date
from pre .. imperial times, turtles are cited as possessing prophylactic powers
against deafness or curses.
See the collections in Han kuan 'ch'i chung and Han kuan /iu chung.
' For the identification of kueib as turtle, see Keightley (1978), pp. 8f, 157f.
Li chi 22.16b, Couvreur (1913), vol.I, p. 525.
1 SC 128, p. 2; see also SC 127, p. 8. ' HHSCC 26.2a.
Li chi 54.26a; Couvreur (1913), vol. II, p.510.
10 Li chi 53.4a; Couvreur (1913), vol. II, p.462.
" Ta Tai Ii chi 58 'Tseng tzu t'ien yiian', SPTK 5.8b.
12 Li chi 24.14a and 25.12a; Couvreur (1913), vol. I, pp. 568, 578.
" SHC, SPPY 1.2b and 5.36b (Yuan, pp. 3, 168); Mathieu (1983), vol. I, pp. 7, 345.
Divination by shells, bones and stalks
A number of references may be found in Han writings to the belief that the
turtles' ability to prophesy depends on their great age and their accumulated
store of wisdom. A general statement in the Huai-nan-tzu explains that turtles
are used for prognostication and that other types of bone are not used for this
purpose, owing to the turtles' longevity; elsewhere in this book this is stated to
extend to 3000 years.
In a note to the Li-chi, Liu Hsiang (79-8 BC) is
recorded as remarking on the age of turtles' utterances, and adding that thev
acquire numinous powers (ling) after living for 1 OOO years. 15 The same reaso;
is cited in the Lun-heng, in an answer given by K'ung-tzu to Tzu lu, and again
in the Po-hu t'ung. '
The Shih-chi writes that turtles with divine (shen) powers are to be found in
the of the Yangtse River. They are taken alive regularly each year in
Lu-chiang commandery; as many as twenty specimens measure one foot and
two inches, such a length being attained only after 1 OOO years; while those of a
seven or eight inches, when caught by the local inhabitants, are prized
Another passage adds that the inhabitants of the Yangtse region
regularly breed turtles and consume them in the belief that they are able to
induce the onset of vital energy (ch'i), and that they are beneficial for
problems of decrepitude and old age.
That considerable importance was
attached to the acquisition of turtles may be seen in the injunction to
fishermen to catch them, as specified for the last month of summer in the
Ordinances of the Months ('Yiieh ling'). 19
Turtles, or creatures that resemble them, are credited with somewhat
different forms of power in an early source which draws on folklore. Such
references be found in the first and fifth chapters of the Shan-hai ching,
of which have been ascribed to the end of the fourth century BC, but
which perhaps reached their present form towards the end of the Former Han
period. Two passages mention a particular type of animal like a turtle which
may be worn as a prophylactic against deafness and calluses; another creature
may be used as a preventative against fire.
Consumption of parts of a special
type of three-legged turtle will act as an antidote for major diseases and
tumours; consumption of parts of another three-legged turtle, described as
pieh, will provide protection against imprecation or disease. 21
In some of the passages that are cited above the divine powers of the yarrow
plant are explained by the same reason as that given for the turtle, i.e., its age.
HNT 14.8b and 17.Sb.
" Li chi 3.15b; Couvreur (1913), vol.I, p. 61; for the value of those turtles imbued with ling as
against others, see HNT 16.12a; see also CCFL 'Feng pen' 9.lla.
L!f 'Pu shih', 71, p. 995; Forke (1907-11), vol.I, p. 182; PHTB ('Shih kuei') 16b (6.4a, b); Tjan
TJ.e .Som (1949-52), vol. II, p. 523. 17 SC 128, pp. 9, 10. 1s SC 128, p. 6.
L1chi16.9a; Couvreur (1913), voLI, p. 367; LSCC 6.lb; HNT 5.Sb; see also Chou /i4.20a; Biot
(1851 ), vol.I, p. 90.
SHC, SPPY l.2b, 5.15a and 5.43b (Yuan, pp. 3, 138 and 177) Mathieu (1983) vol I pp 7
282, 367. ' ' . , . '
SHC: SPPY 5.19b and 36b (Yiian, pp. 144, 168); Mathieu (1983), vol.l, pp. 295, 345; for the
practice of ku, see CC eh. 2.
164 Divination, mythology and monarchy
However, the multiplicity of stalks stemming from the single root together
with the length of the stalks forms another reason why the plant is credited
with special characteristics. Both the turtle and the yarrow feature in a passage
which was added to the Shih-chi by Ch'u Shao-sun and wherein a considerable
degree of folklore and tradition surround their mysteries.
Ch'u Shao-sun reports
that among the records that he found in the office of
the director of divination (T'ai pu) he came across the statement that 'below
there is the pine-root (ju ling),
above there is the dodder (t 'u ssu);
above there
are the massed yarrow stalks, below there is the holy turtle'. He explains the
term fu ling as something lying below the dodder plant, being shaped in
appearance like a bird in flight. When the new rain has ceased falling and the
skies are clear and calm without winds, the dodder is cropped by night and
removed. The area is forthwith illuminated with the use of a fire-basket, and at
the moment when the flames of the fire-basket are extinguished the spot is
marked and surrounded with forty feet of new cloth. At dawn the spot is
immediately excavated to collect the/u ling. This is found after penetrating to a
depth of four to seven feet, but it will not be found at a depth ofoverseven feet.
Ch'u Shao-sun continues:
The ju ling is the root of the thousand-year-old pine tree. Eat it and you will not die. I
am informed that below those yarrow plants which produce their full hundred stalks
there will inevitably be a holy turtle on guard. On top there will invariably be blue
clouds as a cover. Traditionally it was said that when the world is at peace and the way
of true sovereignty is being practised, the yarrow stalks grow to a length of ten feet, and
the luxuriant plant will produce its full hundred stalks. But those who collect the plant
nowadays are not able to attain the standard of the past. Those who cannot find plants
with a hundred stalks ten foot long even find it difficult to get them with eighty stalks
eight foot long. Members of the public who like to have recourse to the lines of the
hexagrams collect plants with sixty stalks or more measuring six feet and find that they
are fit to be used.
Ch'u Shao-sun quotes the saying that material wealth will fall to those who
procure the named (or famous) turtles; their families will invariably grow to be
very rich, to the extent of a fortune up to ten million. The names of the eight
famous turtles are then given, following major elements of the universe that
are usually seen in the heavens, i.e., the turtle of the north dipper, the south
pole, the five stars, the eight winds, the twenty-eight lodges, the sun and moon,
the nine regions, and finally the turtle of jade.
'In each case the turtle's
pattern comprises written signs beneath the plastron; the message that the
written signs convey gives the turtle its designation and I have simply
summarised what is indicated without copying the pattern.' After citing
examples of other objects of great value, the text proceeds: 'Whoever succeeds
in finding the hundred stalk yarrow plant, and of acquiring it together with the
turtle below, and uses these objects for purposes of divination will be correct in
SC 128, pp. 7f. 23 Literally hidden or latent magic. 24 Literally hare's silk.
SC 128, p. 9; in some texts this is given as king's jade.
Divination by shells, bones and stalks
all his pronouncements such that he can determine good or bad fortune.'
From these considerations it may be suggested that the value of the turtles
~ n d their particular qualities was linked with the idea of permanency, as seen
m the turtle's own age and the connection with the heavenly bodies, whose
lives outlast those of generations of human beings. The source of truth, the
vehicle for divination must be seen to transcend the brevity of human life, and
it will be found in creatures believed to live longer than any others, named
after the most permanent objects that man encounters. The turtle both
_contains an accumulated store of knowledge and wisdom, and may be viewed
as a timeless source of truth that stands above human transience.
While there are hints of a hierarchy whereby divination by turtle-shells was
regarded as being superior to that by yarrow-stalks, the evidence is somewhat
conflicting. According to one passage,
while the Son of Heaven does not
divine with stalks, the leaders of the states (Chu hou) do so; but they do not use
the stalks when they are outside their own states, except when they are
determining a place of residence. The statement that the Son of Heaven does
n,g_tuse turtles to determine the situation of the ancestral shrine is explained on
the grounds that he has already carried out divination by this means in order
to establish the seat of his kingdom.
From the foregoing references it would appear that divination by turtle had
acquired a higher place than that of yarrow stalks. In the Chou Ii however we
may read that 'for major decisions of state, divination should be done first by
stalks and then by turtle shells'.
The I-Ii includes detailed prescriptions for
divination by turtle to determine a suitable time for burial; there follow the rules
laid down for the use of stalks to determine the appropriate place for burial. 28
Officials concerned with divination
In describing the needs and purposes of the religious observances of the kings
of old, the Li-chi notes how those sovereigns would be preceded by the
shamans, with officials responsible for astrology following behind; and they
would be accompanied on all sides by diviners who operated with both shells
and stalks, and by musicians and their supporters. In another passage,
specialists with the use of shells are mentioned in the same category as
specialists responsible for prayer, archery, riding and medical care. 29
The Chou li includes a regular establishment of officials or dignitaries whose
duties were concerned with divination. They are set out in hierarchical manner
l o ~ ~ with other officials who \\'ere responsible for other types of religious
act1v1ty (for example, music, sacrifices, funeral ceremonies), under the general
Li chi 54.27b; Couvreur (1913), vol. II, p. 512.
Chou Ii 24.24a; Biot (1851), vol. II, p. 81. For the relative trust to be placed in the shells and the
stalks, see Shang shu ('Hung fan') SSC 12.17aff., Karlgren (1950b), p. 33.
I Ii 37.lSb, 17a; Steele (1917), vol. II, pp. 73, 75; Couvreur (1928), pp.474, 477.
Li chi 13.7a, band 22.17a; Couvreur (1913), vol.I, pp.304, 526.
Divination, mythology and monarchy
title of 'Offices of the Spring' .30 The director of divination (T'ai-pu) was
supported by two counsellors (Ta-Ju), four masters of divination (Pu shih);
eight diviners (Pujen) of middle rank and sixteen diviners of lowest rank; two
store keepers (fub); two scribes (shiha); four assistants (lzsub) and forty
attendants (t'ub). In addition there were two officials designated 'turtle men'
(Kuei jen), supported by two store keepers, two scribes, four craftsmen
(kungb), four assistants and forty attendants. There was also a of
two Chui Jen, with one scribe and eight attendants, who were responsible
burning the necessary materials, and eight prognosticators (Chan }en), with
their supporting staff, who were concerned with divination and its results both
by means of turtle shells and yarrow stalks. For the specific use of yarrow
stalks, the text lists two specialists (Shihjen) supported by one store keeper,
two scribes and eight attendants.
There is no immediate or certain means of relating this highly idealised
establishment to the actual practice of the centuries before the imperial era,
but it can hardly be expected that the account is entirely realistic. While the
passage may owe something to the Chinese addiction for hierarchical schemes
for the organisation of man and his affairs, it may possibly also refl\:<;t_a rs:al
characteristic of the arrangements that were sometimes made, i.e., a duplica-
tion of participants in the ceremonies, in order to prevent abuse or the
exploitation of mantic practices for ignoble ends. . .
A second passage in the Chou li
is concerned with the duties of these
officials. The T'ai-pu was responsible for the three methods of augury
chao ), depending on the resemblance of the cracks in the shells to the patterns
of jade, pottery or the land. 32 He was also responsible for the three methods of
the 'Changes' (i), i.e., Lien-shan, Kuei-tsang and Chou-i. The masters of
divination were responsible for 'opening the four auguries of the turtles', i.e.,
fang chao, kung chao, i chao and kung chaob. The turtle men were
for the six varieties of turtle, each with its own distinct name. The specialists m
yarrow were responsible for the three methods of the Changes, thereby
discriminating between the names of the nine types of yarrow. . . .
Information regarding the post of T'ai-pu or T'ai-pu ling durmg irnpenal
times is not entirely clear, and the evidence is partly contradictory. In the list of
officials that is given for Former Han, the T'ai-pu ling and his assistant (ch'eng)
duly feature as subordinates of the T'ai-ch 'ang of
nial), who had been known under the title of Feng-eh 'ang until 144 BC. In this
list the director of divination is treated in precisely the same way as five other
senior officials who were responsible for specialist tasks, i.e., the directors of
music (T'aiyiieh ling), prayer(T'ai chuting), butchery (T'ai tsai ling), astrology
(T'ai shih ling) and medical care (T'ai i ling). These officials were usually of the
grade of 600 bushels (shW).
3Chou Ii 17.12b; Biot (1851), vol.I, p. 409.
31 Chou /i24.10a; Biot (1851), vol. II, pp.69f.
33 HSPC 19A.6b; Bielenstein (1980), p. 19.
32 I.e., yii chao, wa chao and yiian chao.
Divination by shells, bones and stalks 167
to the descriptive notice in this list the T'ai-pu was first
established m 104 BC.
A passage in the Shih-chi
which refers to the
establishment of the T'ai-pu from the beginning of Han was misinterpreted by
some commentators to signify that the post had been established at the time of
Wen-ti's accession in 180 BC. However, another passage of the Shih-chi refers
to Han's inheritance of the office of T'ai-pu from Ch'in;
elsewhere we are told
that the second Ch'in emperor once (207 BC) summoned the T'ai-pu to
divination wi_th the use of the hexagrams.
An unnamed Chiang-chun
t az-pu was captured m one of the incidents whereby Kao-ti consolidated his
of the empire in c. 197 BC, but this man presumably did not hold
title fro_m the central imperial government.
The advice of the T'ai-pu and
his subordmates was sought in 90 BC on the question of whether an attack on
the Hsiung-nu would be auspicious, and in AD 3 in connection with the
position .r status that should be accorded to Wang Mang's daughter.
Acco_rdmg to the treatise on officials in the Hsu Han chih,
the T'ai puling
(600 shzh grade) was later suppressed, at a date which is not specified, and the
post was combined with that of director of astronomy (T'ai shih ling). A
fragment of one of the lost treatises on officials of the Han empire includes
three specialists on divination with turtles (kuei pu) among the subordinates of
T'ai_ shih ling.
These posts and officials are, needless to say, to be
d1stmgmshed from the occupation of professional diviner which was open to
members of the public.
Early writings on divination
A further indication of the attention paid to these types of divination by
educated men of the day may be seen in the writings known to have existed on
the subject at the close of the Former Han period. The bibliographical list that
in the Han shu_names a total of fifteen works the category of
Shih kuei ; of these, five included the word kueib in the title. It may safely be
assumed that they were concerned with divination by turtle, and they
amounted altogether to 158 chuan.
One other item, which was entitled Shih
HSPC 19A.7b. " SC 127, p. 2.
SC 128, p. 3.
37 SC 87, p. 43; Bodde (1938), pp. 52-3. For references to diviners who were apparently not part
of the official see Hulsewe (1985), pp.85, 176-7 (A 94 and D 173).
SC P 10; ID this somewhat doubtful passage, T'ai pu is taken by one commentator
(fakigawa) as the name of a general see also HSPC 41.Sa.
39 HSPC 96B.18b; Hulsewe (1979a), p. 171; HSPC 97B.23a.
HHSCC (tr) Bielenstein (1980), pp. 19, 22 and 163 note 66.
Jf_an kuan, cited ID HHSCC (tr.) 25.lb (note); Han kuan liu chung la; Ch'en Tso-lung lb
B1elenste1D (1980), p. 22. '
See for example, HHSCC 12.la and 48.4b for examples of individuals who acted aspu hsiang
HSPC (a) Kuei shu 52 chilan; (b) Hsia kuei26 chiian; (c) Nankuei
shu 28 chuan, (d)_ Chu 36 _chuan; and (e) Tsa kuei 16 chiian. Shen Ch'in-han (1775 832)
suggests that_ the mformahon by Ch'u Shao-sun in his addendum to SC 128 is a summary
of the item listed here as Kuez shu, whose length is sometimes given as 53 clu"ian.
Divination, mythology and monarchy
Evaluation and criticism
d" the practice and validitv of divination may be
A number of 111g .. ,,,.. ;
., ..... ,............. d f the Han nenod and
found m hterature_.J!!HULr..<?m . - .1=:_- . - ............ d d t 1
deliberate_ 0 er-
'criticism is expressed ?n. the . grounds. that the
is inexpedient or ineffective, or. because it is _rational n?r
consistent. A few writers protest for moralist reasons; some cnt1c1se way m
which was or undue reliance.
The Hsi-tz 'u-ch'7t;;;; of the Book of Changes includes the that four
principles of the holy sages are included in the Changes. o.ne of
these, 'in divination by turtles or stalks, they respected the_1r prognostlcat10ns ,
and this passage is repeated in the Hou Han shu, bemg qu?ted as.
Confucius.45 As m!ght be expected, a somewgat RQ!nt of VIeWJS.
expressed in the Han-[ei-tzu._ a .. ways
b b gh t mav read that one denves from..
whereby a state can e rou .t o fJ:lln, we __ ..._------ .. , ... _ ...
. d service to the demons
th,e - .. . . . . ..,. .
'and holy .'.ill:L.!.h.L
......... ----- d t" t ra er and sacnfice
Another chapter of the
accompanymg. evo o .. ....,'.c:c. . t t
.. - 1;"'"hh d with whtarv fortunes stresses tne mcons1s en
....... - ..
nature of fiUU:.sheU& .... . . . = i.. -. -
.. of
cons1derat10ns. . r d
,.\'poemorthe Ch'u-tz'u, which be wddle of the
century BC, records an anecdote in which an opm1on is
the validity and correct use of divination. We read that Chu Yuan once
consulted a named diviner, who was master of the techniques of both
and stalks, and who set about preparing these media in order.to answer Chu
Yiian's questions. these proved to be concerned with matters of a
44 HHSCC 26.4a.
4s Hsi tz'u chuan, SSC, 7.23b; HHSCC 82A.la; Ngo Van Xuyet (1976), P 73.
4 'Wang cheng' 15, p.113; Liao (1939-59), vol. I, p. 134.
41 'Shih hsieh' 19, p. 131; Liao (1939-59), vol.I, PP 156f.
Divination by shells, bones and stalks 169
!_1!gi} nature, pri1_1c;iles; the diviner is ...
quoted as explaining that SUCh matb;!rS lay beyond the scope of the turtle and
the yarrow.
The Huai-nan-tzu refers to divination on a number of occasions, and as that
text derived from a collection of a number of authors' writings, it is hardly
surprising that the passages are not entirely consistent. The eighth chapter
refers to the primaeval state of purity in which conformity with the order of
nature assures the regular fulfilment of cosmic cycles. This was an era when
the human intellect was not being used for deceitfulPJlIQOSes; there was no
resoff fo-a'cnmceoraaysmafwoiil01Je'f [;i a
na. sclieming. The passage ... ... .
Thefeng-[huang] and the [chz]-lin make their appearance; the yarrow stalks and the
turtles bear their signs of augury; honey-dew falls; the bamboo seeds are fully formed;
liu-huang jade is found; scarlet plants grow; there is no room for contrivances or deceit
of heart and mind.
}'he passage wouJd errorJies in the deliberate
sources of wisdom, such as the iarrow stalksandtfiefortfes'
when 1eft

the Huai-nan-tzu describes the value and virtues of a world
to .... ..
competitivs: . .1! .. of harmony achieved. The passage then
traces the 'fall' from the golden age of Huang-ti, itsel(inferior to the order
imposed by Fu Hsi. When chaos set in, the balance of nature was restored by
the efforts of Nii Kua. At this time all achievements were attained uncon-
sciously and the order of nature takiggJhe ili>.filinant place
human intelligence. The neitstage"was seen.in-the and abl:tses
ill of harmony.
.. Mother of the,Wes.t her blessings
c_osmos, by .!_IJ:e deliberate rupture of her symbolic headdress. It was iii
fills period that excesSTVeusehadbeen With tlie result thaf
it was by over use .
A further passage inthe Huai-nan-tzu
refers to divination by turtles and
stalks as one of a number of professional means of enquiring into the order of
nature; the process is mentioned along with comprehension of the principles
of Yin-Yang and the Five Phases, watching for the progress made by vital
energy (ch'i) and awaiting the movements of the heavenly bodies. The subject
Ch'u tz'u 6.la; Hawkes (1985), pp.203f.
HNT 8.1 b. For the auspicious nature of the bamboo's production of seeds, see Chin shu chiao
chu 28.32a, which records such a rare event for AD 292. Liu huang is explained as jade in Kao
Yu's commentary to HNT; elsewhere it is mentioned as a species of bamboo which was of
sufficient rarity to warrant annual presentation as tribute (Hsi ching tsa chi, SPTK ed., 22.6a).
For chu ts'ao as a plant of good omen, see Ta Tai Ii chi 67 'Ming t'ang' SPTK 9.1 la; HSPC
99A.7b, HFHDvol. III, p. 151; PHTB 'Fengshan' 2b, 3b (5.2b, 3b), Tjan Tjoe Som (1949-52),
vol. I, pp. 241, 143, 335 note 330 and p. 341 note 358; and LH 'I hsii' 18, p. 209, Forke
(1907-11), vol. II, p. 165. so HNT 6.9bff. si HNT 15.2la.
Divination, mythology and monarchy
also occurs in the famous chapter of the same work which declares the glories
of heaven to be an institution of the order of nature, in contrast with the
limited power of human skills. The passage points out that the majesty and
might of heaven and earth lie beyond measurement in human terms. It alludes
to the invisible power of spiritual beings and forces, along with the practice of
prayer and worship so as to seek good fortune,
to ()f dhr.in.a.!imL
dec;ide J.llaj()r As .a 12
B..ook. efjSflll:E}_,

thej!,Ql)r,J<lngs .. of old ..
;;:son or .a!! .. .. .f9.E"!J:i.e 9.eJl1QJ1S ..
their of the laws; it was the means of resolving
la1neci the
when in dou_bt is imputed to your action;. i(
______ ...........,,,...... . - - .
is to of course
carrv it out.
. - _t;.J. -"''"""'""'""""''""''
The diviners' profession is defended at considerable length man entertam-
ing passage of the Shih-chi. 54 _!his begins by observing that from the remQte
past an appeal had always been made to divination by turtles l!,J!.d stalks on the
when monarchs rose-tOi)Ower. The practice occurred especially.
.. and also _
wlient1leKing"'orTaI''(Han Wen:ti) the city.
---n:le Sh1h-cmllien - proceeds to recount an anecdote in whTcn-Ssu-ma
Chi-chu, of Ch'u, who was practising his profession in the eastern market of
Ch'ang-an, received a visit from two senior and well-established officials,
namely Sung Chung, counsellor of the palace, and Chia I (206-169 BC) who is
described as an academician. The two men paid their visit
at the suggestion of Chia Q11.ilQ.bismlJ.eaguelhaUhere..w..a&Jl-
long of h()!Y among
specialists in or .. the .. QQ..th well
2.f!ic;ials of
.. ideafO"examine what ffiiglit be .. ampng;those. _who practised _tlie
occult arts. It was such circumstances that the twd officials made their way
"by carriage to the market place, on a day when it had been raining and there
were few persons about. They found Ssu-ma Chi-chu engaged in discussing
weighty matters with a few pupils, such as the order of nature of heaven and
earth, the cyclical motions of the sun and the moon, and the fundamental
principles of Yin-Yang, and good and bad fortune. Ssu-ma Chi-chu treated
2 HNT 20.3a; 'I', SSC 18(1).14a; Karlgren (1950a), p.218.
3 Li chi 3.18b; Couvreur (1913), voLI, pp. 61-2.
SC 127 .2a. Different views have been expressed regarding the authorship of this chapter of the
Shih-chi (see the commentaries).
ss HSPC 4.2a; HFHD vol. I, p. 225.
Divination by shells, bones and stalks 171
his new visitors with the courtesies due to their position and continued his
discourse, ranging from cosmology to astronomy, ethics and signs of
auspicious and inauspicious events. The two visitors were somewhat taken
aback to realise that they were in the presence of somebody who was worthy of
deep respect, and they expressed their surprise at the humble way oflife that he
practised. Ssu-ma Chih-chu then expostulated, asking his visitors where their
real values lay. They replied by remarking on the low esteem in which diviners
generally liela, being believed. to stoop to anYfilmg by
aeceifiiforaer fo promore111e1r-0wiiTiiferestsTn'IITs
. turned the tables. J:Ijt8;lke.d_()rlii,tegnty as a quality to men of true
worth who do valUes or lionour.for the
wtth i:ile'iillghty:--He

J!l. for honours or official stipends they
ftoII1: ro,b hers, beiii Jeiermfiied to "pursue their own
.. Pl1hlic interest:
After bitter criticisql. of the inherent deceit and byJW,crisy oL
public life, Ssu-ma Chi-chi turned to the case of the diviners. He said that of
necessity \\feiii"concerne(f"W!th considerationsc,rc-osiiiology, cyclical
and ethical __9n 5>f a,
to a, only aJter in_anipulating.
It kings.iQ
consult the. .. ..
.. bung it
.. .P.f._<?,()f9fthe.ni:actic.e:s.v.alidity cold.be seen in .thestateofru,Aer of the
!?rid Wet?-J2Qg .. QLQhgyJl.id
.t: ........ ng 0.1..Yueh, .. as his
model. . .. . . .. " ..._._,._,. ....
Chi-chu then referred to the respect that diviners had for due form.
They would deliver their an of ' .
after due attention been paid to setting their ceremQ'filiiIIlatsand befis
in which it was'
after their utterances that the lioly Bemgs woUIO acC'eQ! the sacrifices ,
iheir honour; ffia:tlofar niitiisteis"of

Contnvinghlessinus_ of
such .types could qar,<Jh' . ..
his defence of the profession citing
opposite fro?1 and Chuang Tzu, and by
of d1vmers were in material
He then pol'iiteOoUt the existence of a number .Qf apQifeiit_ ..
172 Divination, mythology and monarchy
imperfections in the world of nature which had been put right the
reactiOns ()f, for example, the stars; and he Ilotedthe
rhythm of sun and moon. However, there was no assured permanency in the
teachings of the kings of old; in such circumstances i!_}YQ_u,!c,l __
misguided to diviner.s.'. l:ltt(!Jances would. oido.uht,._
bntne-othef liandthosewho lived by their wits and their
life hesitated to cite precedents from the past as l:r1:C.3:!1:(;[
bringing pressure to bear ontbeirp:ionarchs or. of seekin&to
their road to success lay by way of exagge.i:aJi,on. as
guides to th.e J>erplexed at1d mentors fortbe foolish,
their sovereigris-by cfarifying the rn:tture of Heaven, and with !lO .. thm1ghtJ}f
honours or fame. ---
One of the spokesmen whose views are set down in the which
was completed perhaps c. 60 BC, inclined fowars!..Q.iyjna_tion.
In the chapter :whichco!Ilpares past andpresent practice,
divination aswitnesse.djnJhe firstce11turyBCfares no better at of
the critic than material aspects of daily life, which are castigated
extravagant. We read:
In ancient times virtuous conduct was practised in the search for happiness, and as a
result generosity accompanied sacrifice and prayer; ethical ideals marked the search for
good fortune, and as a result it was rare for divination to be performed by turtle or
yarrow. 8-_ll_! ill. t<i.!ht: cll!_l11QJ\.!L(ku.e,l).Jhaj;
men look. for goodh1ck; and whilebeing carelessin regardto proper
parffcl.i!ar . friW.ard s
pare11ts, put a high value ()I\ and_111 their extreme .. .. !hex 12ut
their frusfili the fCirtune -t4at the..riE!?:t day will brin&.: .. : ..
Similar tholJghis..are..expre.ss.e.d.in a comment that follows entries for books
on divinatl;;in the Han sh11,_Divination by turtle and stalks was practised by
the holy men of old, we'iead, and citations from the Book of Documents and
the Hsi-tz 'u chuan of the Book of Changes illustrate this

In the age
of decline, however, res,orted to
t11estalks sucb,auhtl< ...
vl.giLAs ar:es-uft there was no responseJrom theh()lY.pirits;
C>fte-n a question was repeated to the .. !10 message was fortli9.nl:_
.Elsewhere, however, the Han shu includes a defence of the profession that
practised with. the yarrow.sfalks in Cli'eni-fu, during Ch'eng-ti's reign (33-7
BC). He is a.s t.tiat, _a.Hh.ol:lgh. giyb:i.e.r:s.who .. use ..dt.:ur.tle..shclls.gr
stalks were mem:tJers of a somewhat despised profession,they had
bringing a: dviiis1ng influence to bear on members.of the_
YTL 'San pu tsu', 29, p. 204.
57 HSPC 30.74b; see Shang slzu, as cited in note 27 above; Hsi tz'u chuan, SSC, 7.27a.
Divination by shells, bones and stalks
questions . raised, they exploited the
slidfo and the. to recite .. tbe. . or
Wb.en.falKirig to children, their answer depended on the principles of their duty
to their parents, and they likewise spoke on a basis of fraternal duty or loyalty,
when answering men with brothers or advising servants of the empire. Jn this
way members of the profession were able to give a lead to good conduct, and it
was noticeable that over a half of those who heard their advice heeded it. 58
.. AD
85 Chang-ti offered K'ung Hsi, a descendant of Confucius who had b(!en
engaged in !he wOrk of collatiOri in the Tung kuan, an appointment as
fuagistiaie of Lin-chin. Oneof his friend.s.cons11lted the stalks to see whether
ne should accept the position that he should refuse to do so.
lt'Uhg Hsi's reaction was indignant, as he insisted that good or bad fortune
depend on and no.t gn divination
59 At much the same time
Wiillg-ch.Tng haci"been by that were apparent in
writings tlfat concerned the use of turtle shells and yarrow stalks; his interest at
least shows that he thought the subject worthy of considered treatment. 60
Perhaps the most vociferous criticism of divination was expressed by Wang
whose attitude appears at first sight 1:; be
Ambivalent. He gives due credit to the principle that material features exist
which, if examined, will give an indication of good or bad fortune to come; at
the same time he rejects outright a dogmatic belief that particular objects such
as turtle shells or yarrow stalks are imbued with numinous powers or holy
qualities. In fact, however, Wang Ch'ung is consistent with his own principles.
By recognising that events of the universe take place according to a natural
order, he also. believes that .. .. !11ay
In this sense indications of good or
bad fortune may be implicitly present in some features or phenomena of the
natural world .. This point is brought out in one passage which concerns the
..1lrnJ .. ..

It may also be seen in the
'following passage: 62
;, are kings take the whole world as their family. \Vhen the members of .a
.. aboumr 1nmatErs lliatmay bad fortune,
5?Ls,11ch beT()renan<ll:iy-huma
who un<lerstandthem
P!()gnosticate fron:ithem, orbad
.. .. .. ortune..pussess.
v.a!. pf lm!Tian beings who are
lOrtunate or unfortunate.
. ...,..ThlsTscomparabie with the signs in the turtle shells or the numerical cast of the
stalks. There are constantly elements of good or bad fortune in the signs and in the cast.
Whel1 divines by turtle or practises divination by the stalks he
HsPc n.2a. ;, ------...
LH 'Pien tung', 43, pp. 649f.; (1907-11), voL I, pp. 105f.
LH 'Chih jui', 51, p. 749; Forke (107-11), voL II, pp. 313f.
Divination, mythology and monarchy
falls in with signs that are fortunate; a man who is not fortunate faUs in
arc-unforfiinate; if !s.noftlie case that the Turne ()r the yilHQ\>( stalb_ -
or nu1ninousquafities such that they comprehend humanfonune or
an<f:ProCiuce signs ora numerical co111bination by wfty ofnotjfication,
Those who sit idly _giyi.11\ltiQn. PJ'. Jmtl.1; or .stalkf> .... w:ill:UN _premous
character may just the
s6::Fc- find that good or bad fortune exists constantly in the
au"d the. advenfofobjecfsbearirig.signs oT &ooa or bad fortune will of
with human beings ofgooaor ...
The same points are brought out in some of Wang Ch'ung's most forceful
chapters,63 where he attacks the_ commonly
yarrow . .and. earth. -.
respecffvefy:wang.Ch;ung sets out to show that neither object possesS((S
or the means of communicating with human beings. _tie
repeats his thesis that it is g_uite fortuitous for successful men to
r - --- -
A famous near-contemporary ofWangCti'ung, however, evidently allowed
that there was some value in divination.
drew a sharp .. ..
-- ---.,"'---.,--- rt t { h' b\ h. h
:?.o.th n:ieth()ds, the .1:1!.e,. , rL ...f!!L.i. ... __
he regar<ie,<i as i!l,Y':llig.J:}e to }}gJect --
against ch'an writings a_nd)ll. J:iis __ on the . ..

..... .. f[Qlll: __
tual to practical grounds. This was a
of Han. society_ anci tb_e,Jack of
impartiality.in . .life.-Ifi" aspeCial chapter on the subject of divination,
Wang Fu_ (90--1 the

However, Wang Fu stressed that 1t was essential for div111at10n tQ
be .used for its proper purposes, and not for matters that were not reaJly
subject to doubt:It was in tills respect that he castigated the contemporary
abuse of turtle shells and yarrow stalks, and their use _
for praying, as it were, to the

seen in the context of his ..
shamanl.sm,'fudgements-dependent on physiognomy,
t 63 shih', 71, pp.994f; (1907-11), vol.I, pp.182[.j
64 HHSCC 59.9b 14b.
6' See also Po hu /ung, as cited in note 16 above, for the view that the of divination lay_ in its
demonstration that decisions of state were not being taken to satisfy personal or arbitrary
66 CFL 'Pu lieh', 25, pp. 29lf; [Kamenarovic, (1992) pp.169f].
Divination by shells, bones and stalks 175

again he was stressing the moral improvement that such
practices inherenfCfangel=s of excessive l:eifunce.or ITllSUSe.
The cau_se was 1n an essaiaHiTbutecfio Cli'iing:w'"aiigT;iing (c.
180--220), If1 tha! Fu, th;it.concerns the way of
yen. Chung-ch'ang T'ung emphasised the- need to pay full to
liuma? values and achievements in dynastic matters, as against an excessive
trust m ui:seen powers. .1".ho. clailn an ..
tf.eaven w1thoqt a comprnhension of human values are comparable with
diviners or prayer J11akers,they are low grade of
folly. Gli!!ng::_ch '_i_t!_l_g t? ... )1is_ ple"1.for impartial

f oEce, to_ the io pu6ilc illierests: wliafever steps that he
Wlll not Serveto averf ruin, auspicious the omens, however
his shrines niay be with turtle shells or yarrow

Forms, procedures and types of question

Reference has ben made above to the injunction that is included in one of the
_to the effect. that
t. Iye
tEe <Jf.amysttcal whicli differs
for the Gludance of that type occurs in texts such as the
13--chi, or, mon:i specificafiy'f or the imperial period, in the Shih-chi. Attention
will first be paid to texts such as the Li-chi.
According to one passage, whose ma_}' be in doubt before
consultation of the shells or the cffwas .. the
motive for doing so derived from a right and proper reason or whetilerTt
aue-r015e-rs-0nai-mofives; omf1ntne 'first ca-se-woliid ii t - t
--- .... --
- .. .... . .. . . . '"-"" . . . . glL.L ..put..
questI()n. .. . .1h!!LQeither shells nor stalks should be
than three times in
.. to
.'Y!!h .. th,eother 1Ile!l1,<&, The Chou lz lists eight specffic types of question
that concern matters of state and which may properly be put to the turtle
shells, i.e., campaigns; the meaning of strange phenomena; the

ramfall; the hkehhood of sickness.
That the turtle shells are not to .. be

CFL 'Wu 26, pp. 301f; 'Hsiang lieh', 27, pp. 308f; and 'Meng lieh', 28, pp. 315f;
(1992), pp. 177f, 18lf and 185f].
Ch'iin shu chih yao 45.26b- 28a.
Lz c_hz 35.6a;_ Couvreur (1913), vol. II, p. 6; the passage is sometimes taken to be an injunction
agamst puttmg the same question for a second time.
10 Li chi 3J4b; Couvreur (1913), vol. I, p. 61.
11 Chou li 24J3b; Biot (1851), vol. II, p. 72.
Divination, mythology and monarchy
consulted over the timing of the major sacrifices to heaven and earth is
explained such f!!S.t::Q

winter solstices. 72 However, this may have been a late, somewhat standardised
arrangemeiif"Elsewhere in the Li-chi we are told how the
wang) of the three dynasties
to ;{heaven and earth, ii:t
prohibition of timing. 73 The sanie passage states
not consult thefortfo"shclls. foraefefmfo1ngl1ie. .the.
14 " .
( Lt-chi provides that those who are about to consult the turtle must take
\ the preparatory steps of bathing and donning their ceremonial jades at
) waist, presumably by way of purification and in order to mark the solemmty
\ of the occasion. 75 The same work quotes a statement of the men of the south to
\ the effect that those who lack constancy of character are not fit to act as
' diviners with shells or stalks. 76 In addition, the diviners' gifts
pep,a,UJ ...
()f a,nq ):i(JlX c;hoic.e
.. shll ... 2i:: ... .. . ..
,. 1n .. to ensure that all religious rites are
a .. ... sJlch ... .. .
with shells and stalks. 78 Some sovereigns treated these specialists
. agothe principles
lying behind Yin and Yang and Heaven and Earth, and with these they
formulated the Changes. The diviner bore the turtle in his hands and faced
south, and the Son of Heaven, wearing formal robes and headgear, faced
north. However inspired and intelligent his heart and mind may have been,
he would invariably draw forward to determine his decision.'
shows the great degree of
refainf orthe-dlvlners;-evento "'i:he J2 .. .. ho1!9_t:red_

'fuforl.ii f!:i.1s-ca11liecflona:re.<lefined in ...

...... -.. <nr . -- ---------" _.... -- ...... . -- ... - ..... .......... .. ........ --
--- -"
72 Li chi 5.24b; Couvreur (1913), vol.I, p. 105.
73 Li chi 54.26a; Couvreur (1913), vol. II, p.510.
74 Couvreur explains this as being due to regulations which provided that this must be placed to
the left side of the palace.
75 Li chi 10.2a; Couvreur (1913), vol. I, p. 225.
76 Li chi 55. !8a; Couvreur (1913), vol. II, p. 532.

78 Li chi 22.17a; Couvreur (1913), voLI, p. 526.
79 Li chi 48.15a; Couvreur (1913), vol. II, p. 314.
Divination by shells, bones and stalks 177
1f' The Li-chi includes detailed prescriptions for the notification of deaths and/f
varying for officials and dignitaries of different grades. When i
d1vmat10n is performed to determine the place and time of burial for a '
counsellor of state (Ta-ju), the diviners must wear prescribed items of
clothing; the officiant who pronounces the prognostication must wear a fur
hat. Similar prescriptions provide for the correct way in which yarrow stalks
are used for the purpose. The same passage provides that at the burial of a
counsellor of state, the Ta tsung jen assists the master of the house in the
, conduct o.f ceremonies; the Hsiao tsung jen puts the charge to the turtle;
and the d1vmer (Pu jen) carries out the act of divination.
When the shells or the stalks were consulted in order to determine the right
date for a specific action, a distinction was drawn between those actions which
followed from happy and unhappy events. Thus, for funerals the first
enquiries were made to settle a date in the distant rather than in the immediate
future (i.e., enquiries concerned a date beyond the current period often days)
but for actions that ensued from happy events the reverse procedure
correct, and the first steps that were taken were to fix a day within the current
ten-day period. In specifying this principle, the Li-chi cites the formulae that
were to be used on these occasions: 'To determine the day, we rely on the
constant truths invested in you, oh mighty turtle, oh mighty yarrow stalks.'
A distinction of terminology was also maintained when divining over the
details of a funeral, depending on the relationship of the chief mourner and the
deceased person.

r 1;!nsitiQ!1 'Yht:nJbi ..s.js

.......... --- ................ ........... J?Q ..... 12.-. ......... .JL ........... -................ _,................ __ Jhe.
text bee?. subject. ()f the time. of Cheng
Hsua_ii p21-290) onwards, and a number infe'rpret:itfon 81111
remam in d011bt. f!(;COrdance with of the
Opfofons that havebeen expressed. 84 " .....
AS'anrsfslep ...the for graves makes a survey of the land
that may be selected for the purpose and for which enquiry is being made. The
four corners are dug out, and the excavated earth is deposited outside the area
while earth that is excavated from the centre is placed on the south side of
When the morning ceremony of lamentation has been completed, the
mourners proceed to the south side of the site, facing north, and discard
their and. belts. The diviner who is charged with the duty of putting
the takes his place at the right hand side, that is the place of honour,
of mourners. The diviner faces east; he draws off the upper part of the
case m which the yarrow stalks have been contained and holds both parts in
Li chi 40.lla; Couvreur (1913), vol. II, pp.122f.
82 Li chi 3.14a; Couvreur (1913), vol. I, pp. 60.-1; the same formula occurs in I Ii 47.3b; Steele
(1917), vol. II, p, 159; Couvreur (1928), p. 582.
Li chi 42.lOa; Couvreur (1913), vol. II, p.167.
I li 37.15b; Steele (1917), vol. II, p. 73; Couvreur (1928), p.474 .
178 Divination, mythology and monarchy
his hands. Facing south, he receives the charge, such as 'X who is in mourning
for his father Y seeks by means of the stalks to find a suitable place of burial.
He is contemplating the use of this dark spot as the place and wishes to know
whether evil consequences will follow such a choice.'
The diviner agrees to accept the charge but he does not enunciate it. Turning
right about, he faces north and performs his act of divination pointing to the
central plot, with the assistant who draws out the lines on his left. When an act
of divination is completed, the pattern of lines is taken and shown to the
diviner charged with the question; he receives it; inspects it; and returns it. The
assistant who has drawn the lines faces east and the prognostication is made
according to a number of methods.
He moves forward and announces this to
the diviner charged with the question and to the chief mourner. If the
prognostication signifies 'proceed', the chief mourners don their headbands
and belts, and perform their act of lamentation. If the prognostication
signifies the reverse, divination is performed to choose a site for burial
according to the ceremony carried out in the first instance ..
[,i-chi, fqr ..a . second ()r . s ubseq b,urial in ... .wherein, a
aiieady been buried (for for a spo s(:) n,9 diyin,i;lt,i on is carriedol1t. RY
stalks; presumably this was bec;ause the site had already been
as favourable.

The I-li also itrcludes a detailed description of the ceremony of divination
for determining an appropriate day for burial. This was conducted with turtle
shells, and much the same attention was paid to the niceties of the occasion
and its procedures as in the case of consultation of the stalks to choose a
suitable place. Considerable stress is laid on the positions adopted by the
various participants and the way in which the turtle's plastron was handled.
Again, should the answer be unfavourable, the ceremony was to be repeated.
A ceremony whose purpose was completely different, which is described in
the opening passage of the I-li, concerned occasions when a young man was
confirmed in his majority and put on adult headdress. But despite the
differences in purpose, the ceremony had many features in common with that
undertaken to choose the right situation for a grave, as has been described
above. The yarrow stalks were first consulted to determine an auspicious day
for the event. The stalks, mat and means of figuring out the hexagrams were
laid out in the western portico; the diviner drew off the upper part of the case in
which the stalks had been contained, and held both parts in his hands. When
he had learnt what charge was to be put, he resumed his seat, facing west, with
the assistant who drew out the lines on his left. When the act of divination had
been completed, the drawn out hexagram was shown to the master of the
The commentators suggest that these included use of Lien-shan, Kuei-ts'ang and Chou i.
Li chi 33.5b; Couvreur (1913), vol. I, p. 760.
J /i 37.l 7a; Steele (1917), vol. II, pp. 75f; Couvreur (1928), pp. 477f; and Iii 41.8b; Steele (1917),
vol. II, p. 101; Couvreur (1928), p. 509; and 44.2a; Steele (1917), vol. II, p. 127; Couvreur
(1928), p. 541.
Divination by shells, bones and stalks
house. It was inspected by the diviner and his assistants, who would report if it
should be regarded as being favourable. If it showed that the day would not be
favourable, the stalks were consulted to see whether a day somewhat later
would be appropriate. 88
stalks was also practised on this occasion to choose
d1vmat10n for both these purposes was a means of showing respect to the
assumption of majority by the young man in question; demonstration of such
respect was in itself a means of stressing the importance of /i. 89
i11struments of.divination were treated
with .that wastlJeir ..m
poss1qlepol!llt\QJ1. j\longwith a number of other objectsthat 'were for
t() 6e brought

of the respect due to a ruler--P-un!Shmeiitwas
it ..
... or \\,ere.pena,fiie.s if it was
necessi;try. tg. thi;lf had been inverted, 'or
that of turt!estli<lthad bGe.n turnGd.C>11 s.i9t;:
This regulation presumably
.. of divination was
Once, when Ch'ii Yuan is reported to have consulted the diviner Chan Yin
without any success, the latter had started the procedure by setting his
straight and brushing the plastron.
We are also told that rapid movement
be when carrying the turtles or yarrow stalks. 93 When they were
bemg earned, they were to be handled in the same way as a number of other
ritualistic_ objects, i.e., with the left hand raised higher than the right hand.94
Ceremomal batons were to be discarded during acts of divination that
concerned details of burial.
In addition, the insistence on strict procedures
for religious ceremonies provided for the disposal of spoilt materials; robes
and vessels or other equipment used in sacrifices, if spoilt, were to be burnt or
bu:ied; spoilt or stalks were also to be buried, along with sacrificial
ammals who had died naturally before the sacrifice for which they were
It has been seen above that in defending his profession Ssu-ma Chi-chu
attention paid by diviners to form and.b.ehallio11r.97
Ch 'u Shao-sun re-i tera'ted-ihe pa1ni:9s ... ..... ----c. .... __,. ..
" I '.i Lia; Steele (1917), vol. I, pp. lf; Couvreur (1928), pp. If.
Lz chz 61.1 b; Couvreur (1913), vol. II, p. 637.
Li chi 4.9b; Couvreur (1913), vol.I, p. 76.
Li chi 4.9a; Couvreur (1913), vol. I, p. 76.
Ch'u tz'u 6.lb; Hawkes (1985), p.204.
Li chi 35.12b; Couvreur (1913), vol. II, p. 13.
Li chi 35.16b; Couvreur (1913), vol. II, p. 18.
Li chi 44.14b; Couvreur (1913), vol. II, p. 216.
Li chi 3.12a; Couvreur (1913), vol. I, p. 57.
SC 127, pp. 2f, as cited above.
180 Divination, mythology and monarchy
At the time when I was serving as a gentleman of the court I used to wander around
Ch'ang-an to look at the sights. J.J:ll'()..<l!.9._e_i;:
professionand ob?erved their deportment and g:ait In the m<lnrier in which they
adjusted their and hats they were just like mcnup
possessed the demeanour of I11en of quality. they a person's cl;iar<!cter
-thij\Vere naiure.' ff a C<l.llle for COUSU]t<i;!i9J!!Y.i!hJh_t; ..
turtle 'slierrs; tlieFdiviiiers c9p:egt _
behaviour; !lev'ef so much as showing a tooth by way of a smlie. . . .. . . -
In Han times there were evidently different views regarding the correct
procedure for storing or disposing of materials used in divination. The
includes the report that when it was desired to perform an act of
divinationin Hsia and Yin periods, stalks and turtles were collected for the
puip-ose; and once the activfff was over they were.disca_rcied. and
the bel1efthatturtles woula"fose ln
store arid tiiatYarrowsfaIKs;1r kept.for-long, would no 1P:oi:e i:eh1}.ri ..!Eeif
spiritual virtue .. t bf--
the time of the office of di_viners of the Chou period: both the.
the sl:iel!s'-wer'e constantly pre'sefvecras being .. Ch::\i_ :
Shao-suriWrMe of the 'turtle house'tha1 exisfocfin his own time within the
cofiecHono"t"iilii1e shelhthatV{er(;Tilought'to.possess.sp!ililiif In the
same context he tells of the use. of shells.or 'If you
collect the bones of turtles' forelegs, pierce them through and wear them at the
waist; or if you collect turtles, place them in the north-west corner of the house
and hang them up; you may then penetrate the fastnesses of the mountains
and the forests without losing your senses.'
In aspecial p'ien which isentitled Shih kuei,
thePo-hut'ung treats the
pa_i!f wax_o,{cfliect"statemeni"an(f partly in ..
a:n<l answer. The text inch.l<les .. a.more siandardISed-degree offormula tion "ihail'
Thatseen"Iiitherto; it repeats a number of principles that had been enunciated
in earlier writings; and it explains some aspects of the procedures, usually by
the simple device of citing from such texts.
by both_ metho,ds was practised, accordingto .this source, from
t&e -sO'ii -;r 'iiiose '"Cif the- rank its.
aeiiionsffatetFiai aeclsi oiis"of siafe were .. no'ffaK:en ...
materials to be useci\vere graded according to .size, and the use" of different
types was specified in accordance with the rank of the consultant. Thus the
Son of Heaven used turtle plastra that measured one foot and two inches, and
yarrow stalks that were nine foot long. Smaller shells and stalks were used by
the nobility (Chu hou), counseJlgrs (Ta-ju) and finally those who were of shih
status, who had to be con,tent with turtles of a mere six inches, and stalks of
three feet. The prescribed measurements wer,e in even figures for the turtle, as
See his addendum to SC 128, pp. 12f. 99 SC 128, p. 3. 100
SC 128, p. 10.
P HT B 16aff (6. 3bff); Tjan Tjoe Som, vol. II, pp. 522f.
Divination by shells, bones and stalks
this belonged to the category of Yin, but odd for the yarrow stalks, which
belonged to Yang. The number of diviners also varied according to the rank of
the r_anging from nine for the Son of Heaven to three for shih. Of
all vanet1es of withered stalks and dried bones, yarrow stalks and turtle shells
were chosen for use because they are the most long-lasting objects in the
world. When the stalks were consulted in order to choose a suitable day for
the ceremony of assuming adulthood, the ceremony took place outside the
ancestral all cases it was apparently essential for the hexagrams
to be_ delmeated w1thm the shrine, in so far as it was the ancestors who
the This latter principle does not appear to have
been m earher texts, and is a somewhat noteworthy, and perhaps
unusual, add1t10n to the subject.
. the prescription that the correct position for
d1vmat10n with milfoil IS on the west side facing east. The reason for burying
shells and stalks after use is to prevent pollution of something that is
honourable. For reasons which are stated but which are by no means clear, the
text also lays down that if divination by stalks fails to reveal whether an action
or proposal will be fortunate or unfortunate, divination is tried again by
means of turtle shells. 102
of manipulating forty-nine out of a total of fifty stalks is
m 'u chuan of the Book of Changes, and the passage is cited
with some vanat10ns m the Lun-heng. The stalks are divided into two lots to
symbolise heaven and earth; four lots are counted off to symbolise the
the being retained between the fingers to symbolise an
mtercalary penod or month.
A note to the Huai-nan-tzu that is ascribed to
Kao Yu (c. 168-212) alludes to the same method of using forty-nine stalks.104
Operators grasp (ts'ao) the turtle and the yarrow are described as setting
the stalks straight (tuan) so as to make enquiry of their numbers.1os
Practice, incidents and occasions: the pre-imperial period
__ <?f .. practice of divination in
pre-1mperrn!hmes. The selected references and incidents is
illawirnere"Iiaveoeen chosen from the various compendia on Ii, in the belief
that they would have been present as examples in the forefront of the minds of
those concerned wit_e._ J:fan period. 106 These incidents
102 Tjan Som rende_rs as 'Th;;:-Ui"iir oil follows the way of yang; it has many
and by its permutations [thmgs] come to completion' (vol. II, p. 525).
Hsi tz u SSC, 7.20aff; LH 'Pu shih', 71, p. 997; Forke (1907--11), vol.I, p. 184. For a
system used m connection with the 81 tetragrams of Yang Hsiung's scheme, sec HHSCC
59.lb, commentary [and Nylan and Sivin (1987), pp. 63f]. 10 HNT 8.lb, note.
HN_T_ 117.3a. See also Ch 'u tz 'u 6.1 b, Hawkes (1985), p. 204, for the diviner who 'set out his
and dusted his (Hawkes' translation).
In there are the many mc1dents reported in the Tso chuan, which merit consideration
as a special study; they wo_uld be more likely to have been brought to attention in Later Han
than m Former Han. I am mdebted to Professor Zurcher for pointing out that eight records of
182 Divination, mythology and monarchy
concern religious activities, the human life-cycle, family affairs or the selection
of individuals for particular purposes or duties. In some cases the references
may be somewhat anachronistic.

110! ..
__ to __ _l'?,wers ..
motivation, they. mtght be.
While tliere were fixed trmes for fire
major services, it was necessary'fo consult the stalks to determine a date for
the minor
turtle or the yarrow. The duties of the Ta tsaimcluded divmat10n to deternune
1Jiei1gnraayror-saerifice to the Five Powers.
Similarly the Ta tsung po led
his subordinate officials in acts of divination to choose the time for sacrifice to
the more important spirits and demons (Shen kuez).1
A further reference to
::'.di::: . J;)rereguisite to thJ<
inspection of schools.
In determining the right day for performing the sacrifices at the perimeter of
the city (chiao ), the charge to do so was announced in the ancestral shrine, and
the act of divination with the turtle shell was performed in the shrine dedicated
to oyrn fathi<r. In tbi.._way of Heawn
oemo!Jfil:.ra.t.ed.bis reverence for his and his love for his late
The consulted io deci3e wlleffierapij'.iicUiif
sacrificial animal had been so selected that good fortune would prevail.
Yarrow stalks were used in connection with minor offerings of animals made
in honour of ancestors.
Reference has been made above to the ways in which divination took place
to determine the time or place of burial; in addition, the yarrow stalks were
consulted to choose the individual who was to impersonate the corpse during
the ceremony.
Divination also featured in connection with marriage and its
possible outcome, as, at a crucial point of the preliminary arrangements, the
turtle shells were consulted to see whether marriage with a certain woman
would be fortunate.
This precaution may have had considerable signifi-
cance in cases wherein the name of the woman was not known for certain.'1
divination (pu) in the Ch 'un-ch'iu and Tso chuan, that are daled 629 and 494 BC, are
concerned exclusively with the ritual purity of the bull chosen for sacrifice at the bounds of the
city (chiao),
Li chi 54.26a; Couvreur (1913), vol. II, p. 510.
Chou Ii 2.20a, b; Biot (1851), vol.I, pp. 37-8.
Chou Ii 18.27b; Biot, vol.I, p.436; see also Chou Ii 19.18a, Biot, vol.I, p. 455.
Li chi 36.5b; Couvreur (1913), vol. II, p. 32.
Li chi 26.4a; Couvreur (1913), vol. I, p. 591; see also CCFL 'Chiao ssu' 69.6a for reference to
divination on the occasions of the sacrifices at the boundary.
Li chi 48.lb; Couvreur (1913), vol. II, p.293.
I Ii 47.la; Steele (1917), vol. II, p.158, Couvreur (1928), pp.581f.
I Ii 47.5a and 24.14b; Steele Vol. II, pp. 128, 159; Couvreur (1928), pp. 345 and 584; Li chi
33.9b; Couvreur (1913), vol. I, p. 764.
I Ii 6.9a, b; Steele, vol.I, p. 21, Couvreur (1928), pp.49-50.
Li chi 2.14a, and 51.25a; Couvreur (1913), vol.I, p.31, and vol. II, p.423.
Divination by shells, bones and stalks
When Shih Tai-chung died his principal consort had not born him an..heit..
it c!ioose been
was had to the shells for
incident is .. connectfon with the-ceremonial. that
liP. againsi<livifi,atfon on the
furth_er passage iD.'il1en:cJii'18'iI1terpreiecfas''-
referring to divination with shells when it was necessary to decide who was to
1Je. the ruler of a state (for.exampTe;-cruringfiieenfufced absence of an
.. ?..'IDg !e-ai_!!l_rrsFor
somewhat less exalted positions, it was the turtle who decided to which official
there was assigned the duty of carrying a ruler's newly born heir, and to which
women there fell the responsibility of rearing the child. 119 Prescriptions for the
buildings used by the Son of Heaven and the nobility for sericulture laid down
that those women whom the turtle declared to be fortunate were to be chosen
to supervise the work. 120
Two other references to divination may be cited from the Li-chi. The first
concerns the building of a city: 121
__..,..--.._____:__. -------
The master said: when a man ascribes what is good to others but assumes personal
responsibility for what is bad, members of the public will yield to each other the credit
for performing good deeds. The Book of Songs has it: 'It was the king who examined
the cracks; he planned this city of Haoh; it was the turtle that set it right; it was Wu
Wang who completed it. ' 122
The second reference is of a more general nature: 123
The master said: the [man tic] instruments used by great men are treated with awe and
respect. In general the Son of Heaven does not make use of yarrow stalks; the nobility
use yarrow stalks when in their own domains. When he is embarked on a journey, the
Son of Heaven makes use of yarrow stalks; unless they are within their own state the
nobility do not make use of yarrow stalks; they consult the turtle in respect of a choice
of residence or chambers; the Son of Heaven does not consult the turtle in order to
determine the site of the principal ancestral shrine.
The master said: when the man of quality is showing respect he makes use of the
sacrificial vessels. It is for this reason that [in paying visits to his superiors] he does not
infringe the regulations that prescribe the day for doing so; nor does he go against the
guidance of the shells and the stalks, for he thereby intends to show respect to his
superiors. It is in this way that those of higher rank do not behave insultingly to those
below and those below do not lack respect in relation with their superiors.
Li chi 10.1 b; Couvreur (1913), vol.I, p. 225; LH 'Pu shih' 71, p. 999; Forke, vol.I, p. 186.
Li chi 51.21a; Couvreur (1913), vol. II, p. 419.
Li chi 28.12a; Couvreur (1913), vol. I, p. 663.
Li chi 48.2a; Couvreur (1913), vol. II, p.294.
Li chi 51.1 Sa; Couvreur (1913), vol. II, p. 408.
Shih. ching 'Wen wang yu sheng', SSC 16(5) 14a; Karlgren (1950a), p. 199; for the Li chi
readmg tu, translated here as 'planned', the text of the Shih ching reads che 'resided in'.
' Li chi 54.28a; Couvreur (1913), vol. II, p. 512.
Divination, mythology and monarchy
Practice, incidents and occasions: the Cb'in and Han period
. rln describing the types of crack are formed on shells, the addendum
\ to chapter 128 of the Shih-chi
lists twenty-three topics that may form the
I subject of enquiry. These are illness and its outcome; the demonic nature of an
J, \ illness the possibility of the release of detained persons; the acquisition of
wealth; success in the sale or purchase of slaves and livestock; the advisability
c6),., of attacking robber bands; the advisability of undertaking journeys; the
;,..J l!) . likelihood of encountering robbers either in an expedition to attack or in
K one designed to find them; the reliability of reports of robber acbVIty; the
"" advisability ofleaving official service on transfer; the advisability of remaining
in office; the fortunate or unfortunate outcome of taking up a residence;
prospects for the year's harvest; the chances of an outbreak of epidemic during
the year; the likelihood of armed violence during the year; the results of an
interview with highly placed persons; success in addressing requests to others;
the chances of finding a lost person or persons; the results of fishing and
hunting; the likelihood of meeting robbers when on a journey; the prospects of
rain; the likelihood of rain ceasing.
It is evident that these were the types of question that were likely to be put to
divination on a popular level, and many of the topics recur in the subjects that
are mentioned in the newly found almanacs, and in prognostications that
derived from comets or clouds. In other literacy evidence the accounts of
are of
than poptrtarfeafs.
consultation of the shells and period,
attention may be drawn to a passage in the Ordinances of the months ('Yiieh
ling'). This text may be seen in nearly identical form in no less than three
works, i.e., the Li-chi, the Lu shih ch'un-ch'iu and the Huai-nan-tzu. As
Professor Bodde has pointed out, the 'Yiieh ling' may be accepted 'as a late
but authentically pre-Han text' .1
Its inclusion in three works of such
different types and purposes as the three that are mentioned, and its
importance in educational terms thanks to its in the Li-chileave
little doubt that these regulations carried a significance in Han
times. I
The text tells us that on the day of the winter solstice the Son of Heaven
personally leads his senior officials to greet the incoming year in the sacred site
set apart for worship on the northern outskirts of the city. On return he pays
tribute to those who had died in his service and comforts their orphans and
widows. It is in this month (Meng tung) that he orders the director of
astrology, divination, or prayer
to pray to the spirits and to prognosticate
124 SC 128, pp. 35f. 125 Bodde (1975), p.16. .
12 Li chi 17.lOb; Couvreur (1913), vol.I, pp.392f (reading t'ai-shih); LSCC 10.lb (readmg
t'ai-pu); HNT 5.14a, b (reading t'ai chu).
Divination by shells, bones and stalks 185
from the signs of the turtle shells and the yarrow stalks, so as to examine their
good or evil fortune .
Divination by turtle shells features in the famous anecdote in which the
second Ch'in emperor (reigned 210-207 BC) was worsted by Chao Kao.
Astonished at the general agreement that an animal that was palpably a deer
was a horse, the emperor sought guidance from the stalks to see if he was
suffering from delusions. The record simply reports how the director of
divination blamed the em eror's stateoTm1n<lon-liis-faililletO'Carout.his.
religious duties with sufficient attention o o mess.
-.on at least two_occaJions or
to have ..
measures pnor to his enthronement in 180 BC as emperor, Liu
Heng, king of Tai, made the appropriate gestures ofreluctance, modesty and
reverence for occult powers. _!!is final step was to consult the turtle. which duly
produced a large horizontal stroke. This was interpreted favourably with the .
llelQ_9t from a lOiig lost worlc; according to Chavannes, the outcome
was .
graa:ntcreret; Ge serai comme) K'i, de la dynastie Hia, par gJQii:ct/,)n
'o'ftheTnCICieiit,' it- is of considerable
interest to note the reference here to an established set of interpretations, not
altogether unlike those that have survived in the Chou i for divination by
means of yarrow stalks and the hexagrams.
In the second instance the
divination on these occasions may have been to make a publiccilspiay of proof
that a newly se_lected sovereign was assuming his charge Willi tiie full 61essigg ____ _
.of occult powers. In 74 BC Liu Ho, king as Han emperor
for a mere twenty-seven days before being deposed. His successor, Liu Ping-i,
was brought forward largely owing to the recommendation of Ping Chi. This
statesman reiterated the young man's virtues and urged that both the stalks
and the shells should be consulted. We are not told for certain whether these
steps were taken; in the event Liu Ping-i duly acceded to the throne; he is best
known under his posthumous title of Hsiian-ti.
Divination played an important part on a number of occasions when
imperial policy or matters were concerned. It is said to haye been
Jnvnlved Jn...a ....case . ..ilL ..
Witchcraft, the ..

Accordmg to the Huai-nan-tzu,
whenever a state faces an emergency,
the ruler issues a summons to his general or generals with the command 'The
12' SC 83, p. 43; Bodde (1938), pp. 52-3.
12 SC 10, p. 4; MH vol. II, p. 446; HSPC 4.2a; HFHD, vol. I, p. 224.
12 HSPC 74.8a; CC pp. 75-81.
SC 128, p. 4. Takigawa notes that this incident refers to the case of the empress Ch'en and her
daughter, and not, as had been suggested, to the major affair of91 BC. For the former case, see
HSPC 59.2a; for the latter, see CC eh. 2. 131 HNT 15.22b.
186 Divination, mythology and monarchy
destiny of Our hearth and home lies with you in person; Our state is facing an
emergency and it is Our desire that you should take command and respond to
the situation.' When he has received such a commission, a general gives orders
to the officials responsible for prayer and to the director of divination
to fast and keep vigil for three days; they should then proceed to the principal
shrine, there to pierce the divinely inspired turtle so as to determine by
divination a day that would be favourable for taking receipt of the drums and
An edict of 89 BC
referred to the ancient practice of divination by both
methods on the part of ministers and counsellors of state, and their refusal to
undertake a project if the guidance was unfavourable. On the occasion to
which the edict referred, the stalks and the Chou-i had been consulted to reveal
the meaning of a gesture whereby the Hsiung-nu had been tethering their
horses close to the Han lines. The sign was duly interpreted as meaning that
the Hsiung-nu would be defeated, and a whole variety of specialists who were
consulted, including experts in the stalks and shells, pronounced that they
themselves were satisfied. The turtle was then consulted for a second time,
to determine which of the generals should be put in command of the
campaign; it was in these circumstances that Li Kuang-Ii was appointed Erb
shih general.
There are also references to diviners, or divination, during the civil wars
that preceded the Han restoration, possibly on a more personal level.
Specialists in both types were among those who interpreted a lady's dreams as
an indication of an imminent armed uprising in AD 27.
In the following
year, 'J''it::'.g J111!g, of .the self-established leaders who had arisen in those
troubled times, is said to have consul.ted the turtle to decide gr11ot he
should surrender to Ts'en P'eng; as there was a split in the sign he did. J101 do
A few incidents may be cited in Han times wherein the shells or the stalks
were consulted to solve problems of choice in a family. Specialists in these
methods of divination are, it is true, conspicuously absent from mention
among those whom Wu-ti consulted to fix an appropriate day for his
but a century or so later the force of tradition was
the case of P'ing-ti's marriage toWang Mang's daughter.
This was shortly
after his accession, at the age.of years; with the example ofHuo Kuang iii
Wang Mang succeeded in having his daughter espoused fo the young
HSPC 96B.18a; Hulsewe (1979a), p. 170.
Le., with the reading pu in place of kua, as suggested by Wang Nien-sun (1744-1832).
HHSCC 12.9b.
' HHSCC 117.14a; Bielenstein (1954-79), vol. II, pp.26f.
SC 127, p.14; for this incident see Loewe (1982), p. 100.
HSPC97B.22bff. It is of considerable interest to note thatin commenting on the latter passage
Fu Ch'ien (c. 125-95) explains how the five different types of crack produced on the shells can
be related to the material elements of the wu hsing.
Divination by shells, bones and stalks
emperor, in the teeth of some opposition. Some of the most senior officials of
s-fafe-were present to take part in the formal ceremony of choosing the bride,
and among them were forty-nine subordinates of the directors of divination
and astrology. These were robed in the formal dress and headgear prescribed
by the compendia on ceremonial,
and they conducted divination by shell
and stalk according to the rules of Ii. For the remaining stages of the
procedure, Wang Mangtook equal pains to ensure that due deference was
paid to the tradit!Onai rites that these occasions demanded.
In AD 128 Gang Na, daughter of Liang Shang, was introduced into the
imperial women's quarters at the age of thirteen. 140 Liang Shang had
conceived the ambition of marrying her to the emperor, doubtless as a means
of enhancing his own position, and by way of acquiring a reward for the
services which his ancestors had rendered to the Han house. When the
specialist in physiognomy had pronounced on the exceptional qualities that
the girl possessed for the august position that was intended, the shells were
consulted, and they yielded a message that was highly encouraging. When the
stalks were consulted they produced the hexagram k'un moving into pi which
was regarded as being just as auspicious.
According to the Shih-chi it had been the practice to put the question
before deciding whether a newly born infant should be reared. 141 This was
presumably intended for cases where the health of a child was precarious, or
if it had been born with some deformity.
Towards the beginning of
Ling-ti's reign (168-189) we hear of one statesman citing the principle laid
down for the Ch'un-ch'iu period that, in the absence of an heir, the eldest son
of a collateral branch is chosen; if there are sons of equal age, the choice
depends on their relative character; and in cases where such considerations
are equal, !he bythe shells a11d the stalks. 143 As a further
example \Vas to a personal question,
there is a reference to the use of shells to fix a site for the burial of the
emperor's mother (AD 194). 144
An anecdote that may be apocryphal refers to consultation of the turtle
shells to determine the meaning of strange phenomena. This occurred after the
birth of Liu Hsiu, destined to be the first of the Later Han emperors. On the
night of the day chia tzu, in the twelfth month of the first year of Chien-p'ing
(15 January 5 BC) the house was suffused with brilliant light. Insome
i11fant'sf i:ither .had a specialist in divination by- shells
prognosticate the meaning of this strange event; it is hardly surprising to learn
For Huo Kuang's marriage of his grand-daughter to Chao-ti, see CC p. 115.
HSPC 97B.23a, commentary; Li chi 26.5a and 31.7a; Couvreur (1913), vol. I, pp. 592, 731.
HHSCC 6.6a, I0.2b and 44.8a, for Liang Na's admission as a kueijen and then as an empress.
SC 127, p. 8.
For the licence for infanticide in cases of certain types of illness or deformity, see Hulsewe
(1985), p. 139 D 56. 143 HHSCC 64. lla.
188 Divination, mythology and monarchy
that he interpreted it as a sign of the greatest possible good fortune.
Diviriafioii1s also mentioned in connection with the strange occurrences that
took place shortly after Shun-ti's accession, when steps were being taken to
re-instate Yang Chen posthumously (c. 126). Orders had been given for his
reburial with full honours, and some ten days before this took place, a flock of
huge birds had roosted in front of the coffin and sung a lament, flying away
only when the ceremony was completed.
Liu Ch'ang, king of Liang
(acceded AD 79) c_onsulted t11e. s,hylls the stalks to find
mt;aning of the \Vith which he was affiicted.
r -- -
On taking up his appointme11t in in AQ Ti-w_u Lun found
the population of the area was addicted to yin ssu, i:e., improper religigu,s
pradices, and that there was a general love of divination by shells al!.d
Other evidence for the practice may be seen in the references to a few
named diviners during Wang Mang's time or in Later Han. Wang K'uang, a
specialist in the use of shells, was one of those who had plotted against Wang
Mang in AD 21.
In AD 23 an apparition which appeared in the palace was
identified as the ghost of Liu Hsinb who had just committed suicide; Wang
Mang ordered Wang Hsi, who had won a fine reputation as a diviner with
hexagrams, to cast the stalks in this connection.
In the same year a diviner
named Wang Lang was falsely passed off as Ch'eng-ti's son Yii
and actually
established as emperor.
Hsii Chun (style Chi-shan) ancestor ofHsii Man is
decribed as being expert at prognostication with shells; very often his
pronouncements were vindicated in a conspicuous manner, to the extent that
his contemporaries compared him with Ching Fang.
Liu Wan, son of Liu
Yii (ji. 165) is mentioned as being skilled at casting the stalks in connection
with disaster and strange phenomena.
Attention should also be paid to the few references in Chinese sources to the
practice of divination by non-Chinese peoples. The Shih-chi carries the general
statement that although the Man, I, Ti and Ch'iang peoples have no ordered
establishment of rulers and subjects, they 11()i:!e 2ossess means .of
divination to decide matters that are in doubt, either with metal and_ stone, m
with wood and tree.s.
More specifically, we hear of the practice of the
southern tribes, as reported by Yung Chih shortly after the successful
campaigns fought against the Nan Yueh and Min Yueh in 111 BC. YungChih
was a native of Nan Yiieh who is alleged to have discussed the religious habits
of the Yueh peoples. Jn .. QfJ1_i Jalk he mentioned thatlon_ge_vity
attended those kings of the south who observed due reverence to the
HHSCC !B.23a, b; LH 'Chi yen' 9, p. 88; Forkc (1907--11), vol. I, p. 180.
HHSCC 54.7a.
HHSCC 50.7a.
HHSCC 41.2b.
HSPC 99C.12a; HFHD, vol. III, p. 408.
HSPC 99C.23b; HFHD, vol. III, p.452; identified as the son of Wang Kuang.
HHSCC 82B.2b; Ngo van Xuyet (1976), p. 11; for the two specialists, each named Ching
Fang, see Hulsewe (1986b). 153 HHSCC 57.14a.
SC 128, p. 3.
Divination by shells, bones and stalks
(kuei); laxity in doing sg had however led to an early demise. Wu-ti was
sufficiently well impressed to order the shamans from Yiieh to conduct their
own forms of worship. In their prayers to the spirits of Heaven and other
deities they practised divination by means of chicken bones. It was owing to
the faith that Wu-ti placed in these rites that religious forms of Yueh and
divination with chicken bones were introduced into China, 155 In commenting
on this passage, Li Ch'i (c. 200) explained that the chicken bones were used in
the same way as turtle shells. Chang Shou-chieh (ji. c. 737) was more explicit.
He wrote that a chicken and a dog were taken, live, and prayers were uttered;
thereafter the animals were killed and cooked, and sacrifice was offered. Two
special bones alone were removed from the chicken; if they had a fissure or
crack that resembled a human body, this was a sign of good fortune; but if the
marks failed to show these signs, they were marks of bad fortune. Chang
Shou-chieh concluded his note by observing that practices in Ling-nan were of
this type in his own time.
The Hou Han shu and the San kuo chih refer briefly to the man tic methods of
other peoples. When the Fu-yii were concerned over questions of military
matters, they sacrificed to Heaven; they slaughtered an ox and divined by
means of the hooves: If these were uncurled, and loose, this signified good
fortune; if they were closed up they meant bad luck. 156 For the Japanese
Islands we are simply told that the peoples of Wo heated bones in order to
divine, thereby discovering what their fortune would be likely to be. 157
From the foregoing it may be seen that many Chinese of the Han period
retained a deep faith in the powers of divination. Deliberate attempts to search
for guidance in this way possessed a significance and a validity that was no less
forceful than that of other intellectual or religious activities undertaken in like
manner to plumb the secrets of the universe. The act of divination depended
on a belief in the t!ie U!J.iverse.H sought communication with
unseen powers through the medium o:f materiaI objects that were thought to
be inspired with IHU}1ipo11s_ m.u1liti.es_. It could be used to ensure good fortune
for the living and the dead; it could solve problems that affected human
destiny on earth; it could serve as a means of ensuring that the holy spirits were
being worshipped in an effective manner.
str_C?ngtradition, _establi_shed in pre-imperial times, that
regulated the forms and procedures which attended consultation of the turtle
shells and the yarrow stalks. Imperial governments included provision for
who performed the ceremonies correctly.
and in order to decide questions of policy: It was also practised at
" SC 12, p. 40; SC 28, p. 80; MH vol. Ill, p. 507; HSPC 25B. la.
HHSCC 85.4a; SKC 30, p. 841.
SKC 30, p. 856; HHSCC 85.12a; Tsunoda and Goodrich (1951), p. 12.
190 Divination, mythology and monarchy
a popular level with a view to obtaining assistance _As
the cenNries passed there grew up some measure of cntic1sm .that
being practised too frequently or for the wrong that was
ineffective and that it lacked an intellectual basis; or that 1t exercised a
detrimental effect on moral standards.
The oracles of the clouds and the winds
The subject of oracles and divination usually calls to mind two topics, the
shells and bones of the Shang-Yin period, and the cast of the yarrow stalks
with a view to constructing the hexagrams. Attention may then fasten on the
inscriptions of the shells and the bones, or on the Changes, and their respective
places in linguistic, religious, historical or philosophical studies. However, it is
by no means always recognised that divination and oracles continued to be
matters of considerable importance in their many forms in later ages, playing
a significant role alike in political decisions, religious practice and scientific
The discoveries of manuscripts and a few artifacts in recent years have
called for a new assessment of the whole subject. It is that
im,pe!"ial age Qfdivig(,ltion, in which man takes positi:Ve steps
cause certain. signs to appear in material form, and the consultation of
oracles, where mal1 seeks to interpreimessages that are inherent in natural
phenomena, were still an important means of enquiry or search for informa-
tion. In the early empires, for which this type of activity is well attested,
Chinese literature was still by no means voluminous, and intellectual activity
had not yet been entirely subjected to the imposition of an orthodoxy.
Emperors themselves were at times involved in mantic processes, and the
eslabflsiimeil1:"0r who were responsible these arts
and who were listed afongside other specialists, such as those of medicin.e or
. other aspeCfsofieiigi ciri. Acts of divina tio11 JhQ QQ DS..l!l tl!ti-911. gf i!J
all probabillty took place-at-aff l;;vels of Some of the practices were
supported by men of distinguished iriteBectual powers;
cri ticismf
- Divination and. oracles sometimes form a meeting place for philosophical or
scientific enquiry, religious ritual and the irrational response to undefined
urges of instinct or the call of myth. The a good ex(:lmple of the
combine the approaches of different disciplines, such as intellectual
history, archaeology and anthropology, and to take account of evidence of
many different types. These include records of incidents; protests voiced by
unbelievers; !Ilanusc;rir:it <;>fmanuals_; instruments
use; and entries in bibliographical cataiogues. In addition, the study of oracles
192 Divination, mythology and monarchy
and divination illustrates the facility of some Chinese of the Ch'in and the Han
periods both to concentrate on specialist practices and to fit them within a
major context. For example, instruments made for mantic purposes demand
the acceptance of some of the principles of astronomy and cosmology; at the
same time, the interpretation of oracles may call for a trust in myth and for
convictions that cannot be explained on rational grounds.
Consideration has been given elsewhere to certain aspects of mantic
practice which have been clarified by recently discovered archaeological
evidence. These aspects include the use of almanacs Uih shu) and instruments
(k'an-yii) and the interpretation placed on comets. Attention has also been
paid to the practice of divination with turtle shells and yarrow stalks during
the Han period.
The following pages will concern the oracles of the clouds
and the winds and the ways in which they were thought to carry
provide __ ------
---f wo recenffy discovered manuscripts depict the clouds and the images
that can be recognised therein, accompanied by short texts which interpret
these signs and the activities that they were thought to foretell. Both the
illustrations of the manuscripts and the texts of the captions bear remark-
able affinities to references to the subject in the Shih-chi, Han shu, Chin shu
and Sui shu. The of certain types of cloud was.
perhaps emanaticrq of
energy; .. to . __
and s_qp_11J!ci.tjQIL.0Lthe thsc.iutere.L.
of eg:iperors and thei!. officials ll:J}d at certain key
of the year. The present chapter will look at the evidence tfiafnas
recently become.available in the two manuscripts, one of the Former Han
period and one dating from perhaps the tenth century. This evidence
extends our knowledge of augury from the clouds and shows the mainte-
nance of a traditional approach to the subject with a number of identical
features and characteristics that survived through a period of at least a
thousand years.
The oracles of the clouds
One of the manuscripts found in tomb no. 3 Ma-wang-tui, which may be
at some time before 168 BC, takes the form of a large sheet of silk, measuring
c. 48 by c. 150 cm.
The manuscript was divided into six horizontal registers,
each one of which included a number of separate entries for phenomena or '
' See chapters 3, 5 and 8 above and 10 below.
2 The manuscript has been entitled by modern scholars T'ien-wen ch'ihsiang tsa chan; see Ku
T'ieh-fu (1978), translated Harper (1979). For illustrations of the clouds, see Chung-kuo
po-wu-kuan, no. 2: Hu-nan sheng po-wu-kuan (Peking: Wen-wu, 1983), no. 138; Chung-kuo
wen-wu no. 1, 1979; and J yiian toying no. 9. It is possible that material comparable with that of
the manuscript may have been included in works entitled Yiin-ch'i t'u (see Hua-yang kuo chih,
SPPY ed., !OA.Sa), and Hou yiin-ch'i (see Sui shu 34, p.1020).

Oracles of the clouds and the winds
of the heavens. Of a total of some 350 surviving items, which include
11lustrat10ns of .these pheno.mena, over 300 also bore texts that gave the
phenomenon a title and provided an interpretation or prognostication of what
,1ts result would be.
has been given elsewhere to the entries on this manuscript for
types of comet. It has been observed that many of the
pred1ct1on.s ascnbed to the appearance of comets n!:lJure;
that the diagrams may be regarded as bearing a high standard of accuracy;
that the document .. <l-I.<:rl:!<i!Ic':lble oLthe painstaking
of records over a long period of time.
. The entries the concern clouds, as shown in figure 13,
mclude a notat10n or title, usually in the form of one of the names of the
states; an illustration, be it of animals, human beings or artifacts;
and a pred1ct10n, usually of a military nature. As with the entries for comets
the text takes the form of general statements and not of prognostications
ad rem for a particular occasion. Thus a series of depictions of clouds in the
shape of five four-legged animals carries the following text:
dog?. .,,. kt, J: -z: HF f h.
u:: /'' w. 1 t 1s appears over the city wall it will not be
horse ff ITT J: /g1
OX ff ffi J: \&
deer ff ffi J: J{J{
if this appears over the army, the general will
if this appears over the army there will be a
if this appears over the army there will be a
if this appears over the army it will take [or be
Probably the earliest reference to this subject is to be found in the Tso chuan
for __ .
.t() inthe clollds (yiin wu)as a means of'
Somewhat fafefllie Lu shih ch;un-ch 'iu refers
to the different shapes of clouds as follows, in a famous passage that describes
the attraction of like things for like: 4
Clouds from the mountains are [like] vegetation; clouds from the water are [like]
fish-scales; clouds that derive from drought arc [like] smoke or fire; rain-clouds are
[hke] the waves of water; in all cases they resemble that from which they originate as a
demonstration to mankind.
Attention now be. paid to the expression yijn ch\JheyitaI energy of
clouds, .which occurs m passages that concern philosophy or natural
m some which draw on mythology. The expression is seen at least
' Tso chuan, SSC 12.18a; Couvreur (1914), voL I, pp. 247--8; Legge (1861 72), vol. v, part I,
PP- 142-4; DeWoskin (1983), p. 8. 4 LSCC 'Ying t'ung' 13.4b.
194 Divination, mythology and monarchy
Figure 13 Figures seen in the clouds, from the silk manuscript from Ma-wang-tui.
four times in the Chuang-tzu, where it has been interpreted as referring to
natural phenomena in a same interpretation has been
applied in occurrences of the expression in the Lu shih eh 'un-ch 'iu, the
Huai-nan-tzu and in one passage in the Kuan-tzu. However, in a second
passage in the Kuan-tzu ('Nei yeh') the term is used according to one writer
metaphorically, with a spiritual connotation. In addition, while the expression
does not occur in the clouds are described there in
terms that show how they transcend material dimensions or limitations and
partake of spiritual powers. From these and other passages, and a much
longer one from the Shih-chi which will be discussed below, Kuroda Genji
traces a distinction between two views of yiin ch'i; one that was limited to
natural phenomena, and one that was extended to the concept of the clouds as
an instrument of prognostication, originating from military practice.
A much longer passage, which is the subject of a special study by Professor
Hulsewe, may be found in the Shih-chi and Han shu. At the time when
Hulsewe's article was published the manuscript from Ma-wang-tui was not
available for study and it may now be possible to amplify our understanding
of the passage.
The passage refers at the outset to observation of the vital energy that
informs the clouds (fan wang yiin eh 'i).
It mentions the visible signs of such
' Kuroda Genji (1977), pp. 165-72, cites passages from Chuang-tzu 'Hsiao-yao yu' l.7a, Graham
(1981), p.44; 'Ch'i wu Jun' 2.40b, Graham, p.58; 'Tsai yu' ll.34a; and 'Tien yiin', 14.47a;
Hsiin-tzu 'Fu p'ien' 26.13a (SPTK), Liang ed. p. 358; Kuan-tzu 'Shui ti' 39.2b; 'Nei yeh' 49.5a,
Rickett (1965), p. 167; LSCC 'Kuan piao' 20. l 7b; and HNT 3.30a. See also TP YL 6.4b for a
citation from Ta hsiang lieh hsing t 'u which concerns the control exercised over various climatic
phenomena including yiin ch'i.
SC 27, p. 72, MH vol. III, pp. 393f; HSPC 26.43a; Hulsewe (1979c).
7 Chavannes separates the two characters and renders 'observation des nees et des vapeurs'. The
expression yiin eh 'i is also seen in Shui-hu-ti strip no. 852R (Jao Tseng-i and Tseng Hsien-t'ung
(1982), plate 18). I am aware that many distinguished scholars prefer to leave 'ch'i' as it stands
rather than attempt a translation. The term 'vital energy' that is used here is comparable with
'vital force', as once suggested by Bernhard Karlgren (see Day (1972), p. 15 note 1). In Han
writings, eh'i had not been developed in the way that it appears later, principally in
Neo-Confucian thought. It signified a life-giving force or vitality that may remain latent within
certain natural substances or phenomena, or may become manifest in material form. In this way
the unseen force could be seen to appear in the clouds. A further example of the concept of eh 'i
Oracles of the clouds and the winds 195
energy, for example, as 'animals situated on top of the energy of the clouds'.
Types of vital energy are distinguished by appearance or colour (se); in
addition, account is taken of the different topographical origins of the clouds,
in terms of mountains such as Hua shan or Heng shan, rivers such as the San
ho or regions such as that between the Yangtse and the Huai rivers. The text
describes clouds in terms of artifacts such as spindles, axles or ladles, and the
prognostications are often of a military nature. The text's statement that
prognostication accords with the shape of the clouds is supported by examples
of clouds whose ch'i resembles domestic animals, tents, ships or banners.
There are only a few signs in the passage of attempts to draw a direct
relationship between the clouds and their messages and the system of
Yin- Yang wu hsing thought.
The way in which this passage depicts clouds and their imagery as animals
or artifacts is remarkably similar to their treatment in the manuscript from
Ma-wang-tui. Attention will be paid below to the even more striking identity
between the terms of the manuscript and those that appear in the next relevant
documentation of the Standard Histories, i.e., the Chin shu and Sui shu.
Consideration should, however, first be given to further evidence for the
practice of consulting the oracles of the clouds and the importance ascribed to
yiin ch'i during the Han period.
[Several passages allude to the way in which the vital energies of the clouds
accompanied Liu Pang. Thi_s W1!-.. .. .. to become .. then
..f!! ..ir QI< ..'!: ig11..o ..f
th,,<i:..teWCi*<i hi..m:}t was thanks to this phenomenon that, at
times when he was a fugitive in hiding, his wife, the later Empress Lii, could
always tell where he happened to be. .. .. .. !JihtYtQ.
.. __ ...denied .....tu ... .. ..!1..LQ!l ..
There are slight indications that, at least in imperial or official circles, it was
thought appropriate to carry out the rites of observation and consultation at
key points in the calendar. In tk_recorQ. .
e!11J? .. .. .. ....a,,!1i..IE:l:l!L()L.O..i?j,t;!C1Li11...1he
c!o1Jds,t11 A:P .. 222J .. <t11.<l9J..jt t,imed so ..as.
Jp ..:1.vith"the ...be.gin,gi11g .. g(sp.ri.11g'.
This suggestion is possibly
in Han times is seen in the opening sentence of Wang Ch'ung's chapter on spontaneous
creation (LH 'Tzujan' 54, p. 775; Forke (1907-11), vol.I, p. 92), 'Thanks to the union of the
eh 'i of Heaven and Earth all things are spontaneously created, in just the same way as children
are spontaneously brought to life thanks to the union of the ch"i of man and woman.'
See SC 27, p. 77, HSPC 26.45a for a reference to yang ch'i; SC 27, p. 75, HSPC 26.44a for a
reference to the five colours, but without the assignment that accords with wu hsing theory.
9 HSPC IOOA.l la; Sung shu 27, p. 767; see also the addendum for a further reference to this
incident and for a few other passages.
10 Ming-ti's performance is recorded in HHSCC 2.4b and HHSCC (tr.) 8.4b for a day
corresponding with 20 February 59. In one passage he is said to have observed yiian eh'i, in the
other yiin wu. In a third passage (HHSCC 79A.l b), he is said to have observed yiin wu on an
unspecified date. It seems reasonable to suppose that the three passages refer to a single
196 Divination, mythology and monarchy
supported by a statement in the Shih-chi and Han shu, to which reference will
be made below, that the starting days of the four seasons were the appropriate
times for making observations, but it should be noted that this statement is not
put forward in immediate respect of observation of the clouds but in
connection with predictions for the harvest. Nevertheless, in the immediately
succeeding passage, 11 concerning observation of the winds with which the type
of harvest was supposedly linked, due account is taken of the circumstances of
the clouds. In addition, Professor Bodde cites a telling passage from the I-wei
t'ung-kua yen12 which shows the value placed on observation of the clouds at
the winter solstice, in order to determine the fortune of the incoming year. It
will be seen below that importance was likewise placed on carrying out the
observation of the winds at a comparable point in the calendar.
An incident in which a named observer of the vital energies, Su Po-e, was
acting on behalf of Wang Mang may also be noted.
In addition, a passage of
the Chou li has been explained by Cheng Chung (d. 83) and Cheng Hsiian
(127-200) as meaning that the duties of the Shih chin included consideration of
the messages of the clouds.
Several passages illustrate how attention to the vital energy of the clouds
could sometimes be paid in material form. We are told that the famous
intermediary Shao Weng once advised Wu-ti that if he wished to communi-
cate with the holy spirits he had best order
w of clouds were painted.
In a second instance,
1-r'i one of the protests recorded in the Yen-t'ieh lun against the extravagance of
the times, reputed! yin 81 BC, the man of learning ( wen-hsueh) cited the
fashion_ipg; .. or. i,?e. .l:l.s l:l11:
basic Traces of a painting of yiin eh 'i in red on black lacquer have
incident, and the expression yiian ch'i, which does not appear to be used elsewhere, requires
explanation. It may be noted that in the manuscript from Ma-wang-tui the yiin'
(speech) is regularly used for yiinb (cloud), and it is possible that in HHSCC 2.4b an ongmal
yiin' (speech) has been corrupted into yiian. The records of observation made by Chang-ti and
Ho-ti (78 and 93) on days corresponding with 18 February and 25 February of those years both
specify yiin wu. Consideration of the dating of these visits suggests that they may have taken
place on the day ofSpring's Beginning, i.e., the seasonal New Year (see Bodde (1975), pp. 45f).
For a further reference to the ruler's visit to the observatory to inspect these phenomena, see
Chung fun B.15a, b.
11 SC 27, pp. 79f; HSPC 26.46a.
12 Bodde (1975), p. 168. A work of this title, in two chiian, edited by Cheng Hsiian (127-200) was
known in Sung times, but it has long been lost. For collected fragments, see Shuo-fu 5, and Sun
Chiieh, Ku wei shu, pp. 265--95. For the dubious nature of the text, see Wei shu t'ung k'ao,
13 HHSCC 1B.23b; LH'Chi yen' 9, p. 90 and 'Hui kuo' 58, p. 831 (Forke (1907-11), vol. I, p. 181,
and vol. II, p. 206.
14 Chou Ii, SSC 25.4aff, Biot (1851), vol. II, p.84; Ho Peng Yoke and Ho Koon Piu (1985),
pp. 25f; see also Chou Ji, SSC 26.20aff, Biot, vol. II, p. 115, for the duties of this type that were
assigned to the Pao chang shih.
15 SC 28, p. 52, MH vol. III, p. 470; HSPC 25A.25a.
YTL Tung yu', 3, p. 22.
Oracles of the clouds and the winds 197
been identified in a Former Han tomb of c. 140--130.
Such paintings arc also
said to have featured in the extravagant mansion built by the Liang family in
the suburbs of Lo-yang c. AD 150.
Paintings of this subject were also
prescribed in the regulations for funerals; they were to be used to decorate the
coffins of kings, princesses and certain grades of imperial consort.
For the next references to observation of the clouds and their oracles we
must turn to the treatises on astronomical phenomena in the Chin shu and Sui
shucompiled by Li Ch'un-feng (607-70).
The significantly different arrange-
ment of the material in these chapters from that of the Shih-chi and the Han
shu mayJ?_ossibly indicate a new approach to the subjectthfil conformed witli.
and treatment ofch'( . - ... ... ..... ..
Bodde has suggestecfihat it may have been Ching Fang the second (7937
BC) and Yang Hsiung (53 BC--AD 18) who had been concerned with
formulating the theory and technique of watching the progress made by vital
energy and measuring its passage in time by means of the twelve pitch-pipes.
Certainly it may be accepted that this technique was being practised during the
latter part of the second century AD, when it was described by Ts'ai Yung
(133-92). from the idea that eh 'i played a_pqrtjp ..
cosmic process and that the timing of !ts a matter of
profoundstgnificaiice;Toriftfiesecouid be observed and charted it would be
possible to uridersfa.ncfihe.movementsof the world of nature and tO harmonise
Ooservafion oftlieprogress maae by vital energy
had therefore become subject to a regularised procedure at court which was
carried out with considerable attention to detail. The implicationsof. t.h.e.
movements whicJ:t.were observed to take place subtly-inthe .. bore a
for and his of the government
of man. :Passi bly a sophisticated idea of eh 'i was emerging; as .compared
with that which underlay earlier practices and their documentation.
The passages of the Chin shu and the Sui shu referring to observation of the
clouds and their eh 'i are largely identical, and it is possible that the difference in
the arrangement of material there compared with that of the earlier histories
reflects the new way of assessing the importance of vital energy. In addition to
very short entries under the title yun eh 'i,
the two chapters include the greater
part of their information about clouds as part of a major section entitled tsa eh 'i.
17 See WWl979.3, lf, for an account of the tombofTs'ao Chuan near Ch'ang-sha (p. 3b includes
a reference to the painting). 18 HHSCC 34.12b; Bielenstein (1976), p. 73.
1 HHSCC (tr.) 6. lOb; see also HHSCC (tr_) 6. lla; 10.Sa and 29.13b.
20 See Ho Peng Yoke (1966), p. 13.
21 See Bodde (1959), p. 18. It is necessary to distinguish between two scholars, each named Ching
Fang and each concerned with the Book of Changes; (a) Ching Fang the first, teacher of
Liang-ch'iu Ho, at one time governor ofCh'i commandery, and known as a specialist in the I
ching by the time ofHsiian-ti (reigned 73-49 BC); Hulsewe (1986b) suggests dates of c. 140 to c.
80; see HSPC 88.9a; and (b) Ching Fang the second, pupil ofChiao Yen-shou, native of Tung
commandery, dates, as in Bodde, 79-37 BC; see below in connection with prognostication
from the winds; HSPC 75.9a and 88.lOb.
22 CS 12.16a; Ho Peng Yoke (1966), p.138; SS 20 p. 576.
198 Divination, mythology and monarchy
While interpretations of various manifestations of eh 'i are still largely couched
in military terms, the lay-out of the chapter treats the whole subject in a wider
context and in a more sophisticated way than hitherto.
The chapters of the Chin shu and the Sui shu include in a different order and
arrangement much of the text that appears in the Shih-chi and the Han shu, as
well as some details that are not in those chapters but which do appear in the
manuscript from Ma-wang-tui. As in the earlier histories, so in the Chin shu
and the Sui shu clouds are described as being shaped like artifacts, such as
spindles, axles and ladles.
In addition, the clouds are classified in terms of the
pre-imperial states and, as may be seen below, in many cases there is complete
identity in this respect with the entries in the manuscript:
Classification by names Comparisons drawn in Illustration in the
of states: manuscript, Chin shu, Sui shu manuscript
Chin shu and Sui shu
cloth cloth?
Chao lt11 ox OX
Ch'u ~
sun sun
carriage horse?
Lu ~
horse [not included]
Weia {ffj dog dog
carriage wheel [not included]
traveller traveller
Weib {(\! rat bird
crimson clothing [not included]
Ch'id n o t included crimson clothing [unidentified
in the Chin shu]
Yueh ~ dragon dragon
Shu ll
circular granary circular granary
Chung-shan i:p LiJ
[not included] ox
[not included] tree
[not included] {unidentified]
*For entries in respect of the foreign peoples of the north and south, see SC 27. 75-fi and HS
The manuscript from Ma-wang-tui and the evidence of the Standard
Histories cited above may now be compared with that of a manuscript of a
later date, found at Tun-huang.
This was a scroll of paper, measuring 31 cm
23 CS 12.19b; Ho Peng Yoke (1966), p.144; SS 21, p. 584. In expressing this view I may be
differing somewhat from Professor Hulsewe; it does not seem to me that the text of the Shih-chi
or the Han shu reads abruptly or that a lacuna should be suspected; see Hulsewe (1979c), p. 42.
24 SC 27, p. 74; HSPC 26.44a; CS 12.23a; Ho Peng Yoke (1966), p.147; SS 21, p. 59.
25 The most comprehensive study of this document will be found in Ho Peng Yoke and Ho Koon
Piu (1985). Reference is made therein (introduction, pp. 2-3 and p. 134) to earlier studies by
Hsiang Ta, Ch'en P'an and Hsia Nai, which concern both the document under consideration
Oracles of the clouds and the winds 199
Figure 14 Figures seen in the clouds, from Chan yi.in ch'i shu.
in height with a length that is given variously as 277 or 299.5 cm. The scroll
consists of seven pieces that were gummed together, being damaged at both
the beginning and the end. After discovery and retention in private hands
since 1943, the scroll finally came under the custody of the Tunhuang County
Museum (item 58).
One side of the scroll carried the text of monetary accounts that date from
the T'ien-pao period (742-56). On the other side there was written part of an
astronomical chart which is distinct from the text that is under consideration
now. This bears the title Chan yun ch'i shu; the text is further divided, by
subtitle, into (a) Kuan yun chang and (b) Chan ch'i chang. On the basis of
calligraphic evidence it has been suggested that this side is considerably later
than the monetary register of the T'ien-pao period and that it dates from the
second half of the tenth century. The text is by no means free of error and it is
regarded as the work of a poor copyist rather than the author of the text,
which may consequently have originated some time before the tenth century.
Like the manuscript from Ma-wang-tui, the Chan yun ch'i shu includes
illustrations of clouds with textual captions (see figure 14). In (a) Kuan yun
chang there survive twenty-nine items, of which three consist of text only. Of
the fifty-one items in (b) Chan ch'i chang, twenty-two have both illustrations
and text; two have text only; twenty-seven consist of illustrations only. The
text that is attached to the illustrations is mainly a set of prognostications
here and the astronomical chart which precedes it. See also Ma Shih-ch'ang (1982),
pp.477-508, and Ho Peng Yoke(1985), pp. 146f. Illustrations of parts of the document appear
as follows: (a) the astronomical chart Chung-kuo ku-tai t'ien-wen wen-wu t'u-chi p. 121, colour
plate 10 and plate63; and Ho Peng Yoke (1985) fig. 58 and endpapers; (b) Chanyiin ch'ishu; in
colour, in Seki Kafuku and To Kengo (1978), pp. 170 1; and Chugoku Tonka ten no. 85; in
black and white in China Pictorial 1980.3, p. 17. The illustrations of the clouds are also
reproduced, individually alongside their text, in Ho Peng Yoke and Ho Koon Piu's
200 Divination, mythology and monarchy
based on the type, colour or shape of the clouds with particular reference to
military fortunes, as may be seen in the following examples (interpretations
follow the suggestions of Ho and Ho and their correction of the text on the
basis of comparable passages from other sources which are mentioned below):
Ho and Ho
p.69, no. 8
p. 71, no. 10
p. 81, no. 21

If there are clouds like a carriage's canopy over the enemy
he should not be attacked.
fl 1ti 1Jl'. l: !ffi s ((J) r (ttl)
L' /F PJ
If on observation from a distance there are clouds like
fighting cocks above an army [or a camp] with red and
white colours intermingled, the army [camp] has obtained
heaven's will and should not be attacked.

If there is a clear sky in all directions with an isolated
cloud rising, large forces of troops will arise in the area
where it is seen.
Regarding the relation of this manuscript to other texts, Ho and Ho state
(English summary, 2) that textual comparison shows that all the entries in the
Chan yiin ch'i shu manuscript have their parallel items in the astronomical
chapters of the Chin shu and the Sui shu. Indeed, the Chan yiin eh 'i shu, because
it is less complete and developed, seems to be more rudimentary and could
well have existed before the time of the chapters of the two Standard Histories
(compiled by Li Ch'un-feng, early seventh century). Ho and Ho also raise the
question of whether the Chan yiin eh 'i shu was perhaps related to the T'ien-wen
chan yiin ch'i t'u, another lost text which is mentioned in the bibliographical
chapter of the Sui shu, although they admit that this can only be a conjecture.
It is in connection with this manuscript that Ho Peng Yoke explains augury
by the clouds in terms
cloud is .. regard.ed as.beiIJ.z.comparable ..
or indicati11g battle .. 911ditions. By careful observation an alert general
\voufa be able . to plan his campaign with foreknowledge of the likely
This practice depends on the view that eh 'i is produced as a
result of the interaction of heaven and earth and the pressures of Yin and
Yang; in due course ch'i accumulates and forms clouds. In addition ch'i may
be produced by the presence of individuals or groups in forms that correspond
with their qualities, intentions, or circumstances. The clouds above can
therefore convey information concerning the characteristics of events or
Ho Peng Yoke (1985), pp.146f.
Oracles of the clouds and the winds
personal.iti?s below. . to military
ma .?.ne --
?.ther words which concerned augury by the clouds may be traced.
Fujiwara Sukeyo's catalogue Nihon koku genzai sho mokuroku of 889 to 97
includes for Yiin ch'i ping Ja under section 33 (ping fa), together with a
entitled T'ien mu ching annotated by Li Ch'un-feng. In addition, the
subject features in Yu Chi-ts'ai (Northern Chou), Ling-t'ai pi yiian eh. 4 and
in Ch'u-t'an Hsi-ta (fl. K'ai-yiian period 713 to 41), T'ang K'ai-yiian 'chan
ching ( chs. 94-7). attention should be drawn to two other manuscripts
fr.?m .which it has not been possible to examine: one entitled Feng
yun eh 1 hou chan m P. 3784; and P. 3794 Chan ch'i shu, whose contents include
a work that concerns both the clouds and the winds.
The oracles of the winds (feng-chiao)
F'eng-chiao 'the corners of the winds', forms another example of oracles that
were seen to exist in the normal phenomena of nature, if man would but take
the trouble to look for them. In the Shih-chi and the Han shu the passages on
oracles of the clouds are followed by a section that concerns prognostica-
tion for the harvest, and it is in this connection that the oracles of the winds are
Together with the oracles of the clouds they are mentioned in the
mtro.ductory passage to the biographies of specialists in the occult arts (fang
m the Hou shu, along with a host of other mantic techniques. 27 As
will be is .. .to this. ty12e.
before the .. but it is only from
feng:c'liii:io''comesiiito .. prominence. In this connection publication is early
awaited of a document written on fifty-one wooden pieces, found at tomb no.
1 Yin-ch'iieh shan Lin-i (Shantung province) and dated between 140 and 118
BC, the content of which is described as Jeng-chiao chan. 28
and practice ofccH1sul ting the or.acles of the winds

........ . --:-- .P.,," ...12 ......... gg_ .................... P.tff! . ..1.!l .a.
It may perhaps be asked whether a development took
place from an original type of naturalist prediction, based on observation of
the direction, force and timing.oftne ...w"lridS,toi'-combination of such ideas
the theories of the cosmos and its operation which were being formulated
durmg the Former Han period. By the eighth century the theory and practice
had reached a highly. stylised form.
...}! .. ..
..1!1:en .w1tb .... SJJ.glJ. ... . 19.J!fe ... Chang Heng
(78:-139) and Ts'ai Yung (133-92).
_ -- -- - - !<;"<'"'- ,' .. e 'cc""
HHSCC 82A.lb; Ngo Van Xuyet, p. 74; DeWoskin (1983), p.43.
WW 1974.2, pp. 18, 32 and 35.
See Ngo Van Xuyet p. 188.
202 Divination, mythology and monarchy
( i) Origins and theory
ConsultatiOn of the winds is mentioned in the Shih-chi and the Han shu
immediately after the passage that concerns the oracles of the clouds.
subject is introduced as being a matter of concern on all occasions when
awaiting signs of the type of harvest that could be expected'fan hou sui mei e',
and the use of the term hou may perhaps be significant; for }1'.Jl11.g
t!ie implies
waiting. It was in the context of waiting for signs of the progress of eh 'i that the
lwe!Vepitch-pipes were set up, and prognostication from the winds may
perhaps best be understood in the same way, i.e., as predictions attendant
upon signs of an expected development.
In the case of the clouds that
development had already occurred.
type of It was in such circumstances that Wei
name was singled out as a specialist skilled in predictmg the

on the first day of the first month, in order to determine the
or 'tiie .. were.as"roTIOW-S:--
.. ,...,., ..... '"" ........ ,., .... ...,,."' >. '' "'- - f '' , . '"' "'- <.o. .
Direction of wind's origin
major drought
minor drought
ripening of jung beans; slight rainfall; a levy to
north harvest of medium quality
north-east harvest of goo