. 1
( 15)


Ancient China and Its Enemies
It has been an article of faith among historians of ancient China that Chinese culture
represented the highest level of civilization in the greater Asia region from the ¬rst
millennium b.c. throughout the pre-imperial period. This Sinocentric image “ which
contrasts the high culture of Shang and Chou China with the lower, “barbarian”
peoples living off the grasslands along the northern frontier “ is embedded in early
Chinese historical records and has been perpetuated over the years by Chinese and
Western historians. In this comprehensive history of the northern frontier of China
from 900 to 100 b.c., Nicola Di Cosmo investigates the origins of this simplistic image,
and in the process shatters it.
This book presents a far more complex picture of early China and its relations with
the “barbarians” to the North, documenting how early Chinese perceived and inter-
acted with increasingly organized, advanced, and politically uni¬ed (and threatening)
groupings of people just outside their domain. Di Cosmo explores the growing ten-
sions between these two worlds as they became progressively more polarized, with the
eventual creation of the nomadic, Hsiung-nu empire in the north and Chinese empire
in the south.
This book is part of a new wave of revisionist scholarship made possible by recent,
important archaeological ¬ndings in China, Mongolia, and Central Asia that can now
be compared against the historical record. It is the ¬rst study investigating the antag-
onism between early China and its neighbors that combines both Chinese historical
texts and archaeological data. Di Cosmo reconciles new, archaeological evidence “ of
early non-Chinese to the north and west of China who lived in stable communities,
had developed bronze technology, and used written language “ with the common
notion of undifferentiated tribes living beyond the pale of Chinese civilization. He ana-
lyzes the patterns of interaction along China™s northern frontiers (from trading, often
on an equal basis, to Eastern Hun“Chinese warfare during the Ch™in dynasty) and
then explores how these relations were recorded (and why) in early Chinese histori-
ography. Di Cosmo scrutinizes the way in which the great Chinese historian, Ssu-ma
Chi™en portrayed the Hsiung-nu empire in his “Records of the Grand Historian” (99
b.c.), the ¬rst written narrative of the northern nomads in Chinese history. Chinese
cultural de¬nitions are explained here as the expression of political goals (for example,
the need to cast enemies in a negative light) and the result of historical processes.
Herein are new interpretations of well-known historical events, including the con-
struction of the early walls, later uni¬ed into the “Great Wall”; the formation of the
¬rst nomadic empire in world history, the Hsiung-nu empire; and the chain of events
that led Chinese armies to conquer the northwestern regions, thus opening a com-
mercial avenue with Central Asia (to become the Silk Road). Readers will come away
with an entirely new, more nuanced picture of the world of ancient China and of its

Nicola Di Cosmo is Senior Lecturer in Chinese History at the University of Canter-
bury (Christchurch, New Zealand). He has been a Research Fellow at Clare Hall,
Cambridge, and has taught at Indiana University and Harvard University. He is a con-
tributing author of The Cambridge History of Ancient China (Michael Loewe and
Edward Shaughnessy, eds., 1999) and State and Ritual in China (Joseph McDermott,
ed., 1999). He is a member of the editorial boards of the Journal of Asian Studies,
Asia Major, and the Journal of East Asian Archaeology.
Ancient China
and Its Enemies
The Rise of Nomadic Power
in East Asian History

Nicola Di Cosmo
University of Canterbury
Christchurch, New Zealand
To My Parents

published by the press syndicate of the university of cambridge
The Pitt Building, Trumpington Street, Cambridge, United Kingdom

cambridge university press
The Edinburgh Building, Cambridge CB2 2RU, UK
40 West 20th Street, New York, NY 10011-4211, USA
10 Stamford Road, Oakleigh, VIC 3166, Australia
Ruiz de Alarcón 13, 28014 Madrid, Spain
Dock House, The Waterfront, Cape Town 8001, South Africa

© Nicola Di Cosmo 2002

This book is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without
the written permission of Cambridge University Press.

First published 2002

Printed in the United Kingdom at the University Press, Cambridge

Typeface Sabon 10/12 pt. System QuarkXPress [BTS]

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

Library of Congress Cataloging in Publication Data
Di Cosmo, Nicola, 1957“
Ancient China and its enemies: the rise of nomadic power in
East Asian history / Nicola Di Cosmo.
p. cm.
Includes bibliographical references and index.
ISBN 0-521-77064-5
1. China“History “ to 221 B.C. I. Title.
DS741.3 .D5 2001
931“dc21 2001025577

ISBN 0 521 77064 5 hardback

Acknowledgments page vii

Introduction 1

Part I

1 The Steppe Highway: The Rise of Pastoral Nomadism as a
Eurasian Phenomenon 13

2 Bronze, Iron, and Gold: The Evolution of Nomadic
Cultures on the Northern Frontier of China 44

Part II

3 Beasts and Birds: The Historical Context of Early Chinese
Perceptions of the Northern Peoples 93

4 Walls and Horses: The Beginning of Historical
Contacts between Horse-Riding Nomads and
Chinese States 127

Part III

5 Those Who Draw the Bow: The Rise of the Hsiung-nu
Nomadic Empire and the Political Uni¬cation of
the Nomads 161

6 From Peace to War: China™s Shift from Appeasement to
Military Engagement 206

Part IV

7 In Search of Grass and Water: Ethnography and History
of the North in the Historian™s Records 255

8 Taming the North: The Rationalization of the Nomads in
Ssu-ma Ch™ien™s Historical Thought 294

Conclusion 313

Glossary 319
Select Bibliography 335
Index 361


So many times have I thought that this page would never be written, that
it is with great relief that I can now begin to thank all the friends and people
who have in one way or another given me assistance or inspiration. Because
in a previous incarnation part of this book was my doctoral dissertation,
my ¬rst debt of gratitude goes to the members of my doctoral committee
in the then Department of Uralic and Altaic Studies at Indiana University:
Christopher I. Beckwith “ with whom I ¬rst discussed my idea “ and Yuri
Bregel, György Kara, and Elliot Sperling, who allowed me to pursue an
interest that was at best tangential to the mainstream of the discipline. Lynn
Struve was an exceptionally scrupulous and insightful external member. I
must also thank Denis Sinor for encouraging me, while I was still a grad-
uate student, to present papers at several conferences. I did much of the
research that eventually went into this book at Cambridge University, where
I was a Research Fellow in the Mongolia Studies Unit (1989“92); my sincere
thanks to Caroline Humphrey and to the staff of the Mongolia Studies Unit
and the Faculty of Oriental Studies for having given me valuable and much-
appreciated support.
The dissertation being written, I had no intention of continuing my
research in ancient Chinese history. If I have persevered, the merit belongs
to Michael Loewe and Edward Shaughnessy. In different ways, they are
among the best scholars I have ever worked with. Loewe™s valuable works
were the ¬rst that I read in this ¬eld and also the last, given the inexhaustible
pace of his scholarship. Although not an Inner Asian specialist, Loewe (in
collaboration with Hulsew©) has done more to enlighten our understand-
ing of the ancient relations between China and Central Asia than any other
scholar, including Pelliot and Chavannes.
Loewe and Shaughnessy™s in¬‚uence on this book has also been essential
in a very direct way. I was thrilled when they asked me to contribute a

chapter to the Cambridge History of Ancient China, but I accepted the
task without a clear notion of how I was going to ful¬ll it. Having had
to train myself in the basics of archaeological scholarship to write the
chapter, my work for the History helped me to keep my interest in ancient
China despite pressure “to return” to my original ¬eld, Manchu and Qing
history. My participation in the making of the volume and the chance to
meet the greatest scholars in the ¬eld were an invaluable psychological
boost. My gratitude, then, goes to all the participants in the “Starved Rock”
preparatory workshop. By the time the chapter was written, I had had
some ideas that perhaps could be developed further. In talking with Ed
Shaughnessy I decided to try to consolidate those ideas into a book; Ed also
volunteered “ a sel¬‚ess act for which I am very grateful “ to read a ¬rst
draft. Needless to say, neither Shaughnessy nor Loewe is in any way respon-
sible for any shortcomings of this book, but their support and encour-
agement have been essential to keeping me in this ¬eld long enough to
¬nish it.
Over the years, I have become acquainted with many Early China schol-
ars who in different ways provided me with help, advice, and useful criti-
cism when required. Among these, I wish to mention Jessica Rawson, whose
scholarship, insightfulness, and enthusiasm I have always admired. I have
also pro¬ted from my acquaintance with David Keightley, Robert Bagley,
and Donald Harper. Lothar von Falkenhausen has been generous with
advice and assistance whenever needed, and his writings have been a source
of knowledge for me. I am most indebted to Emma Bunker among the art
historians working on the “barbaric” frontier. She has helped me to appre-
ciate the visual aspect of the material culture of northern China. Others
whose active research on the “northern frontier” of China has been
especially valuable to me are Jenny So, Louisa Fitzgerald-Huber, Fredrik
Hiebert, Victor Mair, Thomas Bar¬eld, Gideon Shelach, Katheryn Linduff,
and Yangjin Pak.
The greatest archaeologist I have known, during my time at Harvard,
was the late K. C. Chang. To my eternal regret, I was just a little too late.
Long before my arrival at Harvard, I had developed a revering admiration
for K. C. Chang, whose books were for me, as for everyone in my genera-
tion, the formative introduction to Chinese archaeology. When I came to
know K. C., a terrible illness had already started to erode his small, hard
physique. Over time, we had several conversations, which I will always
remember with great joy and great sadness. Yet the memory of the K. C.
Chang I used to talk with will survive: probably the strongest and most
generous man I have known.
Many of my former colleagues at Harvard provided me with advice and
help in my work in the ancient world; I wish to thank in particular Pro-
fessor Carl Lamberg-Karlovsky, with whom I had extremely rewarding
talks. Of the Early China scholars, I would like to thank Bruce Brooks and

the Warring States Working Group for keeping me informed about devel-
opments in the ¬eld as I was disengaging myself from my Early China
studies. I owe special thanks also to Professor Denis Twitchett, whose
unparalleled knowledge of Chinese history and scholarly energy, breadth,
and vision are inspirational.
I am also indebted to several librarians, in particular Charles Aylmer, the
Chinese Librarian at Cambridge University Library; the Librarian and staff
of the Harvard“Yenching Library; and ¬nally Martin Hejdra, the Gest
Librarian at Princeton University.
My stay at the Institute for Advanced Study, where I wrote the last part
of this book and tidied it up before ¬nal submission, was made especially
pleasurable by the acquaintance of several scholars whose fascination for
the ancient world I happen to share. Among them I should mention Pro-
fessors Glen Bowersock, Oleg Grabar, and Heinrich von Staden. Last but
surely not least, I must thank profusely the many valiant scholars in China
who study northern China™s archaeology. Some of them, like Wang Binghua
and Guo Suxin, I have had the pleasure to meet personally. Without their
efforts, work in this ¬eld would be impossible.
Among the institutions that provided me with teaching relief, assistance,
time, and support, all or part of which I used in preparing this book, I wish
to thank, ¬rst of all, at Cambridge University, the Mongolia Study Unit and
Clare Hall, which allowed me to work in blissful freedom for three years;
the Chiang Ching-kuo and Rockefeller Foundations, for postdoctoral
grants; Harvard University, which provided me with sabbatical leave on
two occasions; and, ¬nally, the Institute for Advanced Study, which is the
best working environment I have ever experienced. Cambridge University
Press has been marvelous in its care and assistance. I wish to thank in par-
ticular Mary Child, Camilla Knapp, and Mike Green. It is with enduring
admiration that I thank them for their patient and careful work.
Naturally I cannot ignore my wife, Lia, for her patience and support,
and my son, Francesco, for having had to share my time with an “older
brother” he could not see.
Whatever debts I have incurred in writing this book, responsibility for
it rests solely with me. This book is by no means an arrival point; rather,
it is a temporary stop on a journey that cannot be charted for sure. No
doubt our understanding of the “northern frontier” of China will become
increasingly rich, but this process of accumulating knowledge must be
guided by a sense of history that has sometimes been obfuscated, or simply
overwhelmed, by the combined weights of millenarian literary tradition and
quantities of archaeological data. Trying to ¬nd its way between the Scylla
of archaeology and the Charybdis of tradition, this book is an attempt to
recover that sense of history. In all, I must say that (while not without its
perils) it has been a marvelous voyage.


In the time of Duke Huan of Ch™i [the position of] the son of Heaven had
become humble and weak, while the feudal lords used their energies in attack-
ing [one another]. The Southern Yi and Northern Ti engaged the Central
States in battle, and the continued existence of the Central States seemed [to
hang by] a thin thread [. . .] Duke Huan was troubled about the distress of
the Central States and the rebelliousness of the Yi and Ti. He wished to keep
alive what was dying and to preserve what was ceasing to exist, to bring
esteem to [the position of] the son of Heaven and broaden civil and military
occupations. Therefore the Book of Kuan-tzu grew out of this situation.
(Huai-nan-tzu, 21:7a)1

It seems a shared human experience that the malleable substance at the
origin of “civilizations” “ a sense of cultural cohesion, shared destiny, and
common origin “ coagulates into a harder and stronger matter when the
peoples who belong to it are confronted, at times in a threatening way, by
other peoples who are seen as being different and “beyond the pale.” The
pale, the wall, the furrow in the soil are potentially dividing lines, demar-
cating the territory a community recognizes as its own, whose crossing, by
an alien entity, can generate con¬‚icts and threaten the stability of the com-
munity and, in extreme cases, cause its demise.
No wonder, then, that the antagonism between those who are “in” and
those who are “out,”2 and the criteria the community adopts to demarcate

Adapted from Alan Rickett, trans., Guanzi, vol. 1 (Princeton: Princeton Uni-
versity Press, 1985), pp. 5“6.
On the separation between “in” and “out” in the Chinese conception of world
order, see Lein-sheng Yang, “Historical Notes on the Chinese World Order,” in
The Chinese World Order, ed. John K. Fairbank (Cambridge Mass.: Harvard
University Press, 1968), p. 21.

not only its territory but also the characteristics that are assumed to be the
very basis of its raison d™être (a faith, a race, a code of behavior, a shared
set of values) are at the foundation of how a “civilization” de¬nes itself.
Although a sense of “belonging” to the community might exist prior to an
external challenge, the fact of being challenged makes its members acutely
aware of their common boundaries, forcing them to de¬ne cultural differ-
ences and leading them to build psychological and physical defenses. If there
is one characteristic that civilizations have in common, it is their ideologi-
cal need to defend themselves not just against their own enemies, but against
the enemies of civilization, the “barbarians.” This opposition between
civilization and its enemies can be recognized as one of the great ongoing
themes that we encounter in world history. Frontiers, however, are neither
¬xed nor exclusively defensive. With the expansion of civilization, the
opening of new spaces to investigation, the acquisition of broader geo-
graphic and cultural horizons, frontiers acquire ever-different meanings.
Because of their marginal yet critical status, frontiers are often gray areas,
liminal zones where habitual conventions and principles can lose value,
and new ones begin to appear. In this sense, the study of frontiers often
promotes a critical stance toward de¬nitions of “community,” “culture,”
or “civilization.”
The subject of the present work is the early history of China™s northern
frontier, the area that is understood as both crucial to a fuller understand-
ing of ancient China and the locus of one of the great themes of Chinese
history until modern times, namely, the confrontation between China and
the steppe nomads. The blueprint of this “theme” was ¬xed in the histor-
ical literature during the Han dynasty, as the Grand Historian Ssu-ma
Ch™ien composed, probably around 100 b.c., a monograph on the steppe
nomadic people called Hsiung-nu, which he included in his Shih chi
(Records of the Grand Historian). Ssu-ma Ch™ien based his history of the
north on the assumption (or the pretense of it) that a chasm had always
existed between China “ the Hua-Hsia people “ and the various alien groups
inhabiting the north. That assumption is still with us, re¬‚ected in modern
notions that the northern frontier has always been characterized by a set
of dual oppositions “ between pastoral and settled people (steppe and
sown), between nomadic tribes and Chinese states, between an urban
civilization and a warlike uncivilized society.
The main questions that this book explores are all about the historical
realities hidden behind these dualisms: how and when did pastoral
nomadism appear on the northern fringes of the Sinitic world? What was
the genesis of these two opposite principles “ what the medieval Arab his-
torian Ibn Khaldun (1332“1406) called the civilization ( umran) of the
settled and the civilization of the nomad “ in Chinese history? How did the
Inner Asian geographic, political, and ethnographic space become an inte-
gral part, consciously researched, of China™s ¬rst comprehensive history?

Given the primary need to contextualize the cultural and political dimen-
sions of this relationship, the historical circumstances of the northern
peoples™ interaction with China will form the arena of our ¬rst investiga-
tion. Two phenomena are particularly important here: the expansion
of Sinitic political power into alien areas throughout much of the pre-
imperial period, and the formation of the nomadic empire of the Hsiung-
nu that emerges soon after the imperial uni¬cation. We need also to examine
the cultural paradigm constructed by Ssu-ma Ch™ien to establish the
meaning of the north within the mold of a uni¬ed vision of China™s history,
a paradigm that could also be used by his contemporaries and by future
generations for gathering information about the north.
The main dif¬culty in discussing these issues is that early Chinese history
is an exciting but extremely ¬‚uid ¬eld of study: new texts and artifacts
regularly emerge from archaeological excavations, pushing new analyses
and interpretations to the surface. Because the material excavated is varied,
and the questions posed by archaeologists fan out in different directions,
the interpretive “surface” is continually bubbling with novel possibilities.
The historian is placed in suspension under these circumstances, as narra-
tives are constantly being destabilized. Striving to match archaeological
“narratives” and historical text-based narratives is a thankless task and
often of limited use given the intrinsic incompatibility of the two sets of evi-
dence. The textual sources often refer to an inherited tradition and, in any
case, incorporate the thought process of their authors; the material evidence
(as a body) is relatively accidental, and its interpretation and usefulness
depend on the questions asked by modern scholars. Yet all the information
available can be placed side by side to form a series of “contexts” that in
their interaction may provide useful leads. Thus data collected from dis-
parate sources “rebound” on each other within what is essentially a com-
parative analysis that tends to establish possible similarities, analogies, and
points of contact and, by a logical process, suggests scenarios for possible
These problems have, if anything, even greater cogency in the study of
China™s northern frontier, where the analysis of cultural contacts must span
huge geographical expanses and long periods of time. That alternative paths
of inquiry exist does not mean that the historian is forever prevented from
reaching any solution. To the contrary, it is the growing body of evidence
itself that offers the most exciting possibilities, while demonstrating that an
analysis is needed that moves away from the claustrophobic narrowness
of the Chinese classical tradition (largely endorsed by the modern Western
exegesis). This tradition has ¬rmly enclosed the analysis of cultural con-
tacts across the northern frontier between the Scylla of “sinicization”
and the Charybdis of “natural” (and therefore cultureless) behavior. The
“other” in the Chinese tradition seldom rose above a person regarded either
as someone who was suitable material for cultural assimilation or someone

whose nature was hopelessly different and impermeable to civilization
and thus destined to remain beyond the pale, often in unappealing or
dangerous ways for the Chinese. Under these conditions, a history that
critically examines cultural contacts and ethnic differences as part of
the formation of various cultures is written only with great dif¬culty.
In fact, the history of the northern frontier has frequently been reduced
to a recital of mutual conquests by peoples representing two opposite
This book is an attempt to expand that narrow space and to place the
history of the northern frontier on the level of a cultural history, by estab-
lishing various contexts in which such a history can be articulated. Let me
say from the start that these “contexts” are not meant to be exhaustive.
Nor do I try to espouse a single narrative. My principal aim is to provide
more than one key, in the hope of opening up different possibilities of
interpretation. Hence four separate, but interconnecting, contexts are intro-
duced here, each of which is examined as a separate problematique of the
frontier. Partly as a consequence of the type of sources available, partly as
a function of the historical discourse itself, these four contexts have
been arranged in more or less chronological order. Even though these
contexts (and the narratives that they produce) are still tentative and, as
already noted, intrinsically unstable, this is not to say that this type of
investigation necessarily leads to a blind alley. By moving from one set
of evidence to the next, asking questions that emerge especially from the
comparative analysis of the materials, and proposing answers that have not
been previously envisaged, I hope to see a rich context emerge that will
place the history of the relations between China and the north in a new
The book is divided into four parts, each having two chapters. Each
part represents a separate narrative of the frontier. Although other schol-
ars have treated these topics with great knowledge and competence, their
results are different from mine because they base them on radically differ-
ent premises. For instance, let us take two books, published in the same
year, that are classics in their genre and closely relevant to the subject matter
of this book, namely, Jaroslav Prusek™s Chinese Statelets and the Northern

Barbarians in the Period 1400“300 B.C. (Dordrecht, 1971) and William
Watson™s Cultural Frontiers in Ancient East Asia (Edinburgh, 1971).
Prusek™s control of the classical sources exceeds that of anyone else who

has ever written on this topic, but he bases his narrative on certain premises
(the rise of pastoral nomadism in north China, for instance) that are out-
side the reach of the textual tradition and that can be con¬rmed only by
archaeological investigation. Prusek™s deep erudition provides a reading

that, in the end, goes far beyond the texts he so expertly analyzes, and
the resulting picture remains too close to a single set of evidence to be per-


suasive.3 In contrast, Watson™s archaeological work is extremely rich and
truly insightful, but if we look for answers to historical problems, this evi-
dence immediately shows its limits. The same can be said of other schol-
arly works that have provided much enlightenment on discrete issues and
problems but have remained limited to a particular period, set of sources,
or scholarly tradition and disciplinary training.4 All of them, of course,
provide a generous platform onto which one can climb to look farther
The ¬rst part, devoted to archaeology, is concerned with a frontier
de¬ned through separate material cultures, the “northern” and the Chinese,
that can cross borders and interact but that by and large represent two com-
pletely different traditions. The second part refers to a frontier de¬ned not
through material objects, artifacts, and burial rituals, but through written
words and the ideas they convey. This is a frontier that separates peoples
holding deeply divergent understandings of life, of society, of morality, and
of the values that inform and de¬ne them. It is also a frontier found between
those who write and those who do not (hence the one-sidedness of the evi-
dence). The third part describes a frontier that is more properly political,
one that is the result of political events, recorded in history, that led to pro-
found transformations in the concept of frontier. From a place frequented
by mythological and beastlike beings, the frontier became more concrete, a
place where soldiers were deployed, merchants went to trade, and politi-
cians sought to exploit. The fourth and last part deals with the historiog-
raphy of the frontier as it was “created” in the ¬rst historical narrative in
Chinese history, the Shih-chi of Ssu-ma Ch™ien. The in¬‚uence of this early
narrative cannot be overstated, as it colored deeply later understandings of
the formative process of the frontier, a process whose main lines have
remained largely unquestioned.

part i. The two chapters in Part I attempt to de¬ne the archaeological
context of the emergence of nomads in northern China. The ¬rst chapter
This is not a criticism of Prusek™s work, especially since Prusek did not have
°ˇ °ˇ

at his disposal the type of information we enjoy today, but rather a caveat on
placing excessive faith on the written sources when trying to articulate a histor-
ical hypothesis.
As a paragon of philological accuracy we could mention, for instance, A. F. P.
Hulsew© and Michael Loewe, China in Central Asia: The Early Stage, 125
B.C.“A.D. 23. An annotated translation of chapters 61 and 96 of the History
of the Former Han dynasty (Leiden: Brill, 1979); on a different level, Jenny So
and Emma Bunker™s archaeological expertise is brought to bear in the bold re-
evaluation of the trade between China and the North in their Traders and Raiders
on China™s Northern Frontier (Seattle and London: Arthur Sackler Galley and
University of Washington Press, 1995).


delineates the process through which pastoralism expanded in the Eurasian
steppe zone and the emergence of cultures that had developed advanced
bronze metallurgy and handicraft technologies. The introduction of horse-
back riding and wheeled transportation gave these cultures further impetus
and probably played a role in their ability to spread across Central Eurasia.
During the early ¬rst millennium b.c. mounted nomads, recognizable as
“early” or “Scythian-type” nomads, are evident in clustered cultural centers
throughout Eurasia. Northern China “ as we see in Chapter 2 “ was by no
means extraneous to this continentwide cultural process. Mixed economies
practicing both agriculture and stock rearing, culturally related to the Inner
Asian metallurgical complex, emerged between the world of the Shang and
the bronze cultures of Central Asia, Siberia, and the Altai. At this early
stage, the northern frontier societies constantly interacted with the Shang
and early Chou, and, even though a frontier did exist, no sharp demarca-
tion can be detected. In fact, China™s early frontier was permeable to the
introduction of forms of art and technology both from and through these
neighboring northern societies.
Gradually, northern China also experienced a transition to greater
reliance on animal husbandry. Here “Scythian-type” societies began to
appear, characterized by expert horsemanship, martial valor, and taste for
animal-style art whose formal conventions were shared across Central
Eurasia. These societies, which most likely developed a degree of internal
specialization, included farmers and herders and a nomadic aristocracy that
seems to have achieved a dominant position. Horse riding and iron tech-
nology gradually became widespread in northern China, possibly as a result
of a general evolution, among pastoral nomads, toward more sophisticated
forms of social organization. The ¬nal phase of the development of this
“archaeological” frontier in pre-imperial China unearths an abundance of
precious objects, mostly of gold and silver, which point to a commercial
role for the aristocracy and increased trade with China, dating, probably,
from the ¬fth or fourth century b.c.

part ii. If archaeology can help us to de¬ne cultural types in terms of their
way of life, technical abilities, local customs, and even spiritual realm, only
through written sources can we learn about the cultural and political per-
spectives of the Chinese regarding the north. This issue is inextricably linked
to a “culturalist” perspective that has long dominated the study of foreign
relations in early China. This perspective emphasizes the sharp dichotomy
between a world that is culturally superior and literate, with a common
sense of aesthetic re¬nement, intellectual cultivation, moral norms, and
ideals of social order embedded in rituals and ceremonies, and a world that
lacks such achievements. The boundary between these two worlds, sup-
ported by abundant statements in the early Chinese sources, was easily
interpreted as a boundary between a community that shared civilized values

and a community that did not recognize those values. This interpretation
has been so dominant as to preclude any other approach, even in the face
of notable contradictions, such as, for instance, that a single term analo-
gous to the European “barbarian” did not exist in ancient China. This is
not to deny that the world “outside” the Central Plain was at times por-
trayed, in the ancient literature, as a hostile and different environment or
that foreign peoples often were lumped together under an abstract concept
of “otherness” and regarded as inferior, uncultured, and threatening. But
we need to ask what this meant for the actual conduct of foreign relations.
How can we connect these statements about cultural difference to the his-
torical reality that produced them?
In my view, we cannot limit the discussion about the relations between
Chinese (i.e., Central Plain, or Chou) states and these other political
communities to a series of “cultural” statements retrieved from terse his-
torical sources open to diverse interpretations. Such an approach would
tend to establish that a system of cultural values existed, de¬ned both as
“Chinese” and in opposition to the system of “anti-values” supposedly
embraced by non-Chinese peoples, regardless of the historical context in
which these statements appear. But how can we accept that these statements
marked a true cultural boundary without analyzing the circumstances
under which they were made? To answer this question, Chapter 3 inves-
tigates the actual contexts of political relations between foreign states
and Chinese states. This chapter argues that the separation between a
“Hua-Hsia” Chinese cultural unity and an external barbarism, although
perceived of and expressed in those terms, was actually embedded in a
pattern dominated by the political and military strategies essential to
the survival of the Chinese states. Moreover, those states adopted a variety
of attitudes and strategies vis-à-vis the northern peoples, along a spectrum
ranging from virulent opposition to alliance, political equality, and
It is against a background of endemic warfare and ruthless conquest, and
within a logic ¬nely tuned to exploit every advantage that might promote
the survival of the state, that we must place the statements that we ¬nd in
Chinese sources stressing cultural differences. Recourse to arguments point-
ing to the inferiority of alien peoples served, at times, the political need to
escape norms regulating interstate relations and legitimize the conquest and
annexation of these peoples. At other times, the Chinese used foreign
peoples as resources for strengthening the state and as allies in interstate
Chapter 4 focuses on the early history of the relationship between
nomads and the northern Chinese states. During the late fourth century b.c.
a new type of protagonist appears in Chinese history: the mounted steppe
warrior. Contemporary sources hesitatingly acknowledged the existence of
horse-riding warriors, documented primarily through a famous debate in

which the king of the state of Chao expounds on the necessity to adopt the
methods of mounted warfare predominant in the north.
Analysis of events at this time reveals a new transformation taking place
on the frontier. The incorporation of various Jung and Ti peoples by the
stronger Chinese states did not exhaust the states™ need to expand or to
increase the resources at their disposal. In fact, the demands of the new mil-
itary situation, which resulted in the need to sustain prolonged, expensive
wars and in a great increase in the number of armies, may have been at the
root of the northern states™ expansion in the north. Offering a new inter-
pretation for the motives behind the construction of the early “long walls”
in northern China, this chapter will argue that the construction of static
defense structures served to establish ¬rm bases from which Chinese “occu-
pation” armies could control the surrounding, non-Chinese territory. Using
textual and archaeological evidence, this chapter will revisit the traditional
interpretation according to which the forti¬ed lines of defense, the precur-
sors of the Great Wall, were built to defend the Chinese civilization (or the
incipient Chinese empire) from the incursions of the nomads. Rather, walls
were meant as a form of military penetration and occupation of an alien
territory that the Chinese states could use in a variety of ways, including
horse breeding and trade, and as a reservoir for troops and laborers. Once
the Chinese began a more sustained pattern of relations with nomadic
peoples, the fundamental attitude they adopted toward the nomads shows
continuity with the policies and strategies that had dominated Chinese rela-
tions with the Jung and Ti, not the rupture that a purely defensive strategy
(implied by the erection of “defensive walls”) would entail.

part iii. The issues considered in Part II are essential to understanding
the next transformation of the frontier, which coincides with the emergence
of a uni¬ed nomadic power, the ¬rst such “empire” in world history and
precursor to the Türk and Mongol empires. The policy of occupation and
creeping expansionism practiced by the northern Chinese states in the third
century b.c. was endorsed with a vengeance by the uni¬er of China, Ch™in
Shih Huang-ti, who in 215 b.c. sent an enormous army to conquer and col-
onize the pasture grounds located in the Ordos region. Chapter 5 argues
that the relentless pressure of the Chinese states on the northern frontier
possibly acted as a catalyst for deep social transformations among the
nomads. In a partial reappraisal of the genesis of the Hsiung-nu empire, I
discuss in this chapter a pattern of state formation among Inner Asian
nomads that aims to be consistent with the events as they are narrated
in the historical sources. The rise of the Hsiung-nu empire forced radical
modi¬cation of traditional approaches to “frontier management,” as the
Chinese were now in a position of military inferiority. A new world order
thus emerged wherein the main powers split the world that they knew
into two large areas of in¬‚uence; although uni¬ed, China was no longer

hegemonic. The policy that dominated the relations between Hsiung-nu and
Han in the early Former Han period was one of appeasement and accom-
modation in which China became a virtual tributary of the Hsiung-nu.
Chapter 6 demonstrates why this policy eventually had to be abandoned
and why the Han dynasty needed to turn to more aggressive strategies. Two
factors emerge: ¬rst, the ripening of conditions that on the political, mili-
tary, and economic levels enabled China to invest more of its people and
resources in an all-out war effort; and second and most important, the
ideological shift that accompanied the realization that the ho-ch™in policy
of appeasement did not guarantee peace. Several explanations have been
offered to account for the Han endorsement of a military stance, and this
chapter will explore why the ho-ch™in policy did not work, by looking more
closely at the Hsiung-nu side. From an Inner Asian perspective, it appears
that the “appeasement” policy failed owing to a structural incompatibility
between Hsiung-nu and Han understandings of their mutual international
Chapter 6 ends with a survey of Han westward expansion and of the
Han motives in establishing a military presence in the “Western Regions.”
Again, the debates are not new, and most of the opinions I express here
coincide with those of other scholars. Yet my perspective emphasizes not
so much the economic factors as it does the military and political ones,
which seem to have prevailed in a context in which destruction of the
Hsiung-nu empire as a single political entity was the overriding concern.

part iv. A further, decisive, “transformation” of the frontier occurred in
the ¬rst century b.c., when the north ¬nally became an object of conscious
historical and ethnographic inquiry. The relationship between the Hsiung-
nu and China, as constructed by Ssu-ma Ch™ien in the Shih chi, became a
polarity between two antithetical principles whose genesis coincided with
the dawn of Chinese history. Together with the “crystallization” of Inner
Asian history into a pattern that had not been recognized before in any way
even remotely comparable to the grand scheme erected by him, the his-
torian Ssu-ma Ch™ien opened the door to an empirical investigation of
the north, made not of mythological accounts and moral precepts, but of
information that was as historically rigorous as one might expect from the
“Grand Historian.” He selected his sources carefully, acquired much infor-
mation from persons who had been closely engaged in Hsiung-nu affairs,
copied memorials and diplomatic correspondence, and narrated events with
precision and an abundance of detail. Part IV is based on the identi¬cation
of two chief strands in Ssu-ma Ch™ien™s narrative, one the collection of infor-
mation vital to understanding the Chinese confrontation with the Hsiung-
nu empire; the other, no less vital, the construction of a pattern that
rationalized the relevance of the north in Chinese history. Chapter 7 focuses
on the information that Ssu-ma Ch™ien incorporated in his monographic

account of the Hsiung-nu (chapter 110), effectively starting an ethnogra-
phy and a literate history of the north that also served as a model for later
dynastic histories. Chapter 8 looks at how Ssu-ma Ch™ien rationalized the
history of relations between the north and China into a broad pattern
resting on two elements. One was the creation of a “genealogy” of north-
ern peoples that could match the historical “genealogy” of Chinese dynas-
ties and hegemonic states from the mythical beginning of history to the
historian™s time. The other was the insertion of the north and its inhabi-
tants within the system of correspondences between the celestial and the
human spheres that was believed, in Han times, to constitute cosmic order.
Events such as wars or the downfalls of rulers were regarded as manifes-
tations at the human level of the workings of that cosmic system, and, there-
fore, history was the “output” of a machinery of correlations that could
not exclude the Hsiung-nu or more generally, foreign peoples. Thus foreign
peoples and their lands become equal partners in the construction of
Chinese history, whereas in the past they had been (as far as I can tell)
excluded from the system of correlations and predictions upon which
historical causality was ultimately based.
Our knowledge of the genesis and earliest evolution of relations between
China and the north, down to the Han dynasty, is still gotten through the
lens of Ssu-ma Ch™ien™s “master narrative.” This narrative effectively made
the north into a historical protagonist. At the same time, it trapped the
history of the northern frontier into a dichotomous patterns from which
we have yet to free ourselves. By identifying the history of the frontier as
an artifact, as a “narrative” that must be placed in a given time and intel-
lectual milieu, and as the culmination (obviously not the end) of a long and
intricate process, we can also re-establish the northern “sphere” of Chinese
history as an area with its own autonomous, internally dialectical, histori-
cal, and cultural development.

part i
chapter one

The Steppe Highway
The Rise of Pastoral Nomadism as a
Eurasian Phenomenon

Geographic Features

A Note on Terminology

The terminology for the regions inhabited by the nomadic and semi-
nomadic peoples of Inner Asia in pre-historical and historical times is inher-
ently unstable, given that geographic areas such as Central Asia, Inner Asia,
the Northern Zone, and Central Eurasia are usually de¬ned ad hoc.1
Because the present work is concerned mostly with what Owen Latti-
more has called the “Inner Asian frontiers of China,” I have adopted “Inner
Asia” or “Inner Asian frontier” as a general term for the eastern part of
the continental mass of Eurasia. In practice, it includes three geographical
areas: in the east, Manchuria; in the center, Mongolia, including parts of
Kansu, northern Shensi, and northern Shansi; and in the west, not only
today™s Sinkiang but also the Minusinsk Basin and the northern part of the
Altai Mountains.
This central de¬nition must be accompanied by two others. The nar-
rower term, the so-called Northern Zone, is used, especially in China, to
describe the ecological and cultural frontier between China and Inner Asia.
Today this area is entirely within China™s political boundaries and runs from
the Liao Valley in the east, to the T™ai-hang Mountains up to the Ordos
region in the center, and to the Ning-hsia“Ch™ing-hai cultural region in the
west. This term often refers to the area of the Great Wall, but to avoid

For an extensive discussion of the de¬nition of Central Asia, see Shirin Akiner,
“Conceptual Geographies of Central Asia,” in Sustainable Development in
Central Asia, ed. Shirin Akiner et al. (New York: St. Martin™s Press, 1998), pp.


anachronisms, “Northern Zone” is clearly preferable to “Great Wall
The broader term, “Central Eurasia,” is particularly useful for referring
to the part of the Eurasian landmass that is crossed horizontally by a grass-
land belt stretching from western Manchuria to the Danube. Beginning in
the second millennium b.c., this region saw the development of pastoral
nomadic cultures that ¬‚ourished from the Pontic Steppe across the
Altai and to Mongolia.2 On the Asian side, this broad expanse incorpo-
rates, the region that Alexander von Humboldt called Central Asia in
1843 and Ferdinand von Richtofen later de¬ned as the part of continental
Asia forming a closed hydrological system, with no access to the open
sea. The boundaries he proposed were the Altai Mountains in the north,
the Khingan Range in the east, the Pamirs in the west, and Tibet in
the south.3 Others have de¬ned it in even broader terms, including the area
running from the Caspian Sea and the Ural River Basin in the west to the
Ferghana Valley and Pamir Range in the east, and from the limits of the
Kazakh Steppe belt in the north to the Hindu Kush and Kopet-Dagh in
the south. Today “Central Asia” has acquired a narrower meaning from
its use in the former Soviet Union, and it can be said to include the
territory of the Uzbek, Turkmen, Tajik, Kirgiz, and Kazak states, plus
the Sinkiang (Xinjiang) Uighur Autonomous Province in northwest China,
which, in ancient times, was closely connected with the rest of Central
Before we address the issue of the formation of pastoral cultures of
China™s Inner Asian frontier, it is necessary to survey the natural environ-
ments in which these cultures emerged, environments that placed limita-
tions on the directions and extents of their development. The vast territory
that separates China from Siberia and Central Asia can be divided into three
major geographic zones: the Manchurian Plain; the steppes and forests of
Mongolia; and the oases, deserts, and steppes of Sinkiang.4

Denis Sinor de¬nes Central Eurasia not only in geographical but also in cultural
terms; see his Inner Asia: A Syllabus (Bloomington: Indiana University, 1987, 3rd
rpt.), pp. 1“5.
See L. I. Myroshnikov, “Appendix: A Note on the Meaning of ˜Central Asia™ as
Used in This Book,” in History of Civilizations of Central Asia, ed. A. H. Dani
and V. M. Masson (Paris: Unesco, 1992), 1: 477“80.
The following geographic survey is based primarily on: George B. Cressy, Asia™s
Lands and Peoples (New York: McGraw Hill Book Co., 1963); Robert N. Taaffe,
“The Geographic Setting,” The Cambridge History of Early Inner Asia, ed. Denis
Sinor (Cambridge: Cambridge University Press, 1990), pp. 19“40; V. M. Masson,
“The Environment,” in History of Civilizations of Central Asia, 1: 29“44; Hisao
Matsuda, “The T™ian-shan Range in Asian History,” Acta Asiatica 41 (1981):

The Steppe Highway S I B E R I A



t tl

Khingan Range


Y en

R a.






Aral n
Kazakh Stepp Kh


AL ge R len
e an S e len ru
TA e
Ulan Bator

S yr

u n ta



Wh Long

L ia


L. Balkash
I li R

iver Urumchi

rt Peking

T™ai-Hang Mts.

Gobi D

™I E N

T SHAN Peninsula
Ferg lley


Yellow River

Va Kashgar Tarim R.


Lop Nor Yel l ow
mak os
a Sea
akl NS
K u sh
ndu T O
Hi Koko Nor


0 500 mi
0 500 km

Map 1

Manchurian Plain

Located to the northeast of present-day China, the Manchurian Plain differs
from the Yellow River Plain in that it is not depositional, but erosional, and
it presents a rolling topography. Divided into a northern and a southern
half by the Sungari and the Liao River systems, the Manchurian Plain is
everywhere surrounded by mountains. To the east and southeast the Long
White Mountain separates it from the Korean Peninsula. To the north rises
the Little Khingan Range, running parallel to the Amur River, and to
the west the Great Khingan Range, which develops on a north-south
axis and separates the Manchurian Plain from the Mongolian Plateau. To
the southwest, a series of mountain ranges, such as the Ch™i-lao-t™u and
the Nu-lu-erh-hu, separate it from Inner Mongolia and from the Yellow
River Plain.
In the south the Liao River Valley has a hundred and twenty kilometers
frontage on the Gulf of Liao-tung. Between the mountains and the sea,
the strip of coastal lowland leads to the Yellow River Plain through the
Shan-hai-kuan, which served historically as one gateway for those seeking
to enter (or invade) China. In the northeast the Sungari enters the Amur
lowland through a narrow passage between hills, and in the northwest
a low section of the Great Khingan mountain range gives easy access to
In Manchuria three natural environments are found: forest in the
uplands, especially in the northern half; arable land in the river valleys; and
grassland in the west. Because of the continental climate winters are long
and bitter and summers short and hot. Snow falls from October through
April in the south, and September through May in the north. Precipitation
is concentrated in July and August, and amounts roughly to 630 millime-
ters in the east and 380 millimeters in the west. Soil is very fertile owing to
the natural cover of grass; the growing season is relatively short, but agri-
culture, precarious in the dry west, is possible in the east because of the
moisture from the sea.
The Manchurian uplands extend in the east from Liao-tung to the Amur
River, in between the mountains and the river valleys. Thanks to the greater
volume of rain and moisture at higher altitudes, we ¬nd vast forested areas,
which are deciduous in the south and coniferous in the east. This is the land
of hunters, and today local people still practice trapping, but agriculture is
also possible.
The western part of the geographical Manchurian Plain is today the
northwestern part of Inner Mongolia. The climate is more arid, unsuitable
for agriculture. The Great Khingan Range constitutes the eastern limit of
the Mongolian Plateau, and in fact both the environment and the lifestyle


of the people here resemble those of Mongolia.5 In terms of vegetation, the
north is a forest of Siberian larch and birch, while the south is a Mongolian-
type steppe. Traditionally, its inhabitants have mostly been hunters and pas-
toralists. The southwestern mountains are rugged and dif¬cult to cross,
serving as a natural boundary between two economic zones, the Liao Valley
in the east, suitable for agriculture, and the Mongolian Steppe and Gobi
Desert in the west. This mountainous area extends into northern China, in
particular, the provinces of Hopei and Shansi, where the T™ai-hang moun-
tain range acts as a natural divide running from north to south.
Moving westward from the northern part of the T™ai-hang Range, one
runs into the southernmost fringes of the Gobi, that is, the Ordos Desert,
circumscribed within the bend of the Yellow River. Surrounded by a rim of
mountains, the Gobi is the most northern and furthest inland of all the
deserts on earth, and for the most part it has a climate similar to that of a
dry steppe. The ground is covered with pebbles and gravel, and it has
enough water to sustain some vegetation and animal life. Extremely arid
patches, with sand dunes and almost complete absence of vegetation, cover
only ¬ve percent of the whole desert, mainly in the southwest.


Mongolia is divided into four vegetation zones, which run almost parallel
to each other from east to west.6 The southernmost part is a desert zone,
which is succeeded, going north, by a desert-steppe belt. North of this
is a dry steppe zone to the east and, to the west, a continuation of the
desert-steppe belt in the lower elevations and, in the higher elevations, a
mountain-steppe and forest-steppe zone alternating with patches of dry
steppe. The northernmost zone is heavily forested, though we also ¬nd
alpine meadows that provide excellent pastures interspersed with areas of
Siberian taiga. The southern Gobi extends from western Inner Mongolia to
eastern Sinkiang; to the north the Gobi occupies Mongolia™s southern half.
Mongolia also has several important mountain ranges. In the west, the Altai
Mountains extend northwest to southeast, and their southeasternmost
extension merges with a range known as the Gobi Altai Mountains, which
forms a series of ridges crossed by intramontane valleys and basins. North
of the Altai, in northwestern Mongolia, are mountain ranges that extend
further north into Siberia; to the east, a large depression known as the
Owen Lattimore, The Mongols of Manchuria (New York: John Day Company,
An extensive introduction to the topography, hydrographic system, ¬‚ora and
fauna of Mongolia is included in: The Academy of Sciences MPR, Information
Mongolia (Oxford: Pergamon Press, 1990), pp. 3“49.


Valley of Lakes is interposed between these ranges and the large moun-
tainous area known as the Khangai Mountains. This latter area has a rolling
topography, dotted by sand hills and dry river beds.
The central artery of the Khangai Mountains has a northwestern orien-
tation, similar to that of the Altai, and is crossed by several rivers, forming
a watershed between the drainage system of the Arctic Ocean Basin, into
which the northern rivers ¬‚ow, and the closed drainage system of Central
Asia. The relief of the Khangai is smooth and gentle in its northern and
northwestern parts; in the south, precipitous escarpments and breakages of
the plateau are more typical. In the north-central part of Mongolia, the
Khentii mountain ridge also forms a continental divide between the Paci¬c
Ocean drainage system and the Central Asian Basin. The eastern region of
Mongolia is a raised plain with abundant grassland, and an average alti-
tude of 800 to 1,100 meters above sea level.
The major waterways of Mongolia are concentrated in the north and
¬‚ow in the direction of the Arctic Basin: The Selenge is a tributary of Lake
Baikal, and the Orkhon is the main tributary of the Selenge and is fed by
the Tula. The rivers in the east of the country, particularly those ¬‚owing
from the eastern slopes of the Khentii Mountains, belong to the Paci¬c
Ocean drainage system; among these the Onon, a tributary of the Amur,
and the Kerulen, which ends its course in the Dalai Nor lake on the western
side of the Great Khingan Range, are the most important waterways and
natural avenues of communication between Central Mongolia, Trans-
baikalia, and northern Manchuria.


Sinkiang may be viewed as consisting of three major subregions: the deser-
tic Tarim Basin in the south, the vast T™ien-shan Range in the center, and
the semi-arid Zungarian Basin in the north. The Tarim Basin is drier than
any other desert in China, and it includes a totally dry desert in the center,
the Taklamakan, which is surrounded by a string of oases on its northern,
western, and southern edges. Amongst these oases, the largest are Yarkand,
Khotan, Kashgar, Aksu, Kucha, and Karashar. These oases are formed by
semi-permanent water streams originating from the glaciers at the tops of
the mountains encircling the Tarim Basin, that is, the T™ien-shan in the
north, the Pamirs in the west, and the Kunluns in the south. Irrigation
ditches allow water from the mountains to spread over the river™s alluvial
fan, creating relatively large stretches of farming land. Each oasis consti-
tutes a self-enclosed system that commands some of the desert around it,
an irrigated area with a principal city, barren foothills, and well-watered
mountain valleys upstream.

Into this southern, more desertic region ¬‚ows the main river of Sinkiang,
the Tarim, which is the ¬nal destination of the streams ¬‚owing from the
surrounding mountains, although many evaporate or disappear under-
ground before reaching it. Owing to the aridity of the climate, there is no
cultivation on the Tarim™s banks. Eventually, the Tarim ¬‚ows into the Lop
Nor lake, located in the eastern part of the region. Directly to the north of
Lop Nor, close to the southern slope of the Bogdo Mountain in the eastern
Altai, is the Turfan depression, 266 meters below sea level. North of the
Tarim Basin, the T™ien-shan extends east into China for 1,600 kilometers.
Elevations reach 6,686 meters in the west and 5,089 meters in Bogda Ula,
north of Turfan, in the east. The orography is rugged, although there are
elevated plains and broad valleys covered with alpine meadows in some
The northern half of Sinkiang is occupied by an arid zone known as
Zungaria. This is a desertlike area, but it is less arid than the southern
part and closer in appearance to the Gobi. Some oases are along the
northern slopes of the T™ien-shan, but they are smaller and less richly
irrigated than the southern ones. To the west, the T™ien-shan splits into
two branches that embrace the fertile valley of the Ili River, which
¬‚ows to the northwest, draining into Lake Balkash. North of the Ili, the
Zungarian Gate, at 304 meters of altitude, is a deep corridor between the
northern edge the T™ien-shan and the Tarbagatai Range in the northwest.
This is the lowest pass in all Central Asia, and it was used by nomads
as a gateway to the Kazakh Steppe. The extreme northern and northeast-
ern limits of the region are marked by the Altai. The foothills of the
Altai form a rolling plateau with excellent pastureland. The valley of the
Irtysh, in the far north, between the Tarbagatai and the Altai, at an
elevation of approximately 430 meters, forms another gateway to Central
In addition to these mountain chains, the southern edge of the Tarim
Basin meets the Altyn Tagh mountain chain to the east, whereas the south-
central and southwestern sides of the Taklamakan are blocked by the lofty
Kunlun Mountains, extending down from the Tibetan Plateau. On moun-
tain slopes, precipitation is suf¬cient to allow growth of a relatively dense
grass cover. Indeed, the best pastures to be found in this region are on the
slopes of the Altai and in the intermontane valleys and alpine meadows of
the T™ien-shan; nomads can pasture their herds in these areas through the
year. Forests also grow above the steppe belt, at altitudes of between 1,400
and 2,500 meters.
Finally, an important area for the development of early metallurgy and
pastoral nomadic culture is de¬ned by the Altai and Sayan Ranges, which
begin near the Zungarian Gate, close to Lake Baikal, and extend east for
1,600 kilometers. The central ridges of both ranges are rolling uplands,
which reach an altitude of about 2,586 meters. The Altai system, coming

into Siberia from northwestern Mongolia, is enclosed between the Irtysh
and Ob Rivers, where we ¬nd the Altai Mountains proper, culminating in
Peak Belukha, at about 4,300 meters. East of the Ob lies the eastern Altai
Range, reaching almost to the Yenisei. The two ranges of the Sayan system
encircle the Minusinsk Basin: the eastern Sayan Range extends from Lake
Baikal to the Yenisei, while the western Sayan Range cordons off the basin
in the south. Here, too, the prevailing orography is of rolling hills. Steppe
vegetation covers the lower slopes of the Altai-Sayan mountains up to some
860 meters; above it is a forest of Siberian larch, cedar, ¬r, pine, and birch
up to and above 1,720 meters, followed by alpine meadows to the snow
line at around 2,580 meters.
Sinkiang commands the communication routes between China and
Central Asia. Before the advent of modern rail transportation, the caravans
going west from Hsi-an (Shansi province) en route to the western êntrepots
and markets reached Lan-chou and then began to cross the arid Kansu
region following the base of the Nan Shan (Ch™i-liang) Range and travel-
ing from one irrigated oasis to another. The so-called Kansu Corridor “ a
depression less than 80 kilometers wide and over 960 kilometers long “ is
dotted with oases drawing water from the Nan Shan Range. At the end of
the corridor, Jade Gate (Yü-men) opened the way to Sinkiang, after passing
the cities of An-hsi and Tun-huang. This area, at the western end of the
Gobi, is today a barren desert, but there are signs that in antiquity the cli-
matic conditions were more favorable and that it was then possible to travel
along a line tangential to the southern edge of the Tarim Basin.7 The better-
known route, however, crossed the desert and proceeded northwest to
Hami, on the eastern fringes of the T™ien-shan mountain range, and only
then divided into routes to the south and to the north of the T™ien-shan
To the south, two routes developed, skirting, respectively, the northern
and the southern fringes of the Taklamakan Desert. They joined in the
western part of the Tarim Basin, where the large oasis of Kashgar is located,
and proceeded to the Terek Pass, and through this to Ferghana and Trans-
oxiana. North of the T™ien-shan, the route passed through Urumqi, and
from there, via Kulja, reached the Ili Valley and the Zungarian Pass. Finally,
yet another gateway to Central Asia is located farther north, where the
uninterrupted steppe belt along the base of the Altai provides a passage to
the valley of the Irtysh.

See Hou Can, “Environmental Changes in the Tarim Oases as Seen through
Archeological Discoveries,” in Between Lapis and Jade, ed. F. Hiebert and N. Di
Cosmo, Anthropology and Archaeology of Eurasia (Spring 1996): 55“66;
Mutsumi Hoyanagi, “Natural Changes of the Region along the Old Silk Road in
the Tarim Basin in Historical Times,” Memoirs of the Research Department of
the Toyo Bunko 33 (1975): 85“113.


Pastoral Nomadism in the Steppe: Preconditions

The forests, deserts, and especially grasslands of Central Eurasia have his-
torically been associated with the rise of pastoral nomadism. The ¬rst his-
torical descriptions of these nomads, the Scythians, come down to us from
Greek historians and geographers. Although their individual conceptions
of the Asian nomads varied substantially, they clearly believed that in the
prairies of Central Asia a different strain of people had developed, one
whose customs and lifestyle were incompatible with those of sedentary
I praise not the Scythians in all respects, but in this greatest matter they have
so devised that none who attacks them can escape, and none can catch them
if they desire not to be caught. For when men have no established cities or
fortresses, but all are house-bearers and mounted archers, living not by tilling
the soil, but by cattle rearing and carrying their dwellings on wagons, how
should these not be invincible and unapproachable.8
How had this different way of life arisen? In the nineteenth century, fol-
lowing Darwinian and positivist theories, scholars believed that nomadism
was an evolutionary stage, more advanced than hunting, from which it was
supposed to have sprung, but less developed than agriculture, in the pro-
gressive march of humankind toward civilization. This idea can be traced
back to Lewis Henry Morgan™s in¬‚uence on positivistic ethnographical and
sociological thought: the people who ¬rst domesticated animals became
accustomed to pastoral life before learning to cultivate cereals.9
At the end of the nineteenth century, scholars began to criticize this view-
point, arguing that “the domestication of animals was possible only under
the conditions of a sedentary way of life.”10 The domestication of animals
requires a long process of experimentation and accumulation of technical
knowledge, and presupposes the existence of other sources of economic

. 1
( 15)