Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
The evolving vocabulary of otherness in pre-imperial China: From 'belligerent others' to 'cultural others'
(USC Thesis Other)
The evolving vocabulary of otherness in pre-imperial China: From 'belligerent others' to 'cultural others'
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
THE EVOLVING VOCABULARY OF OTHERNESS IN PRE-IMPERIAL CHINA:
FROM ‘BELLIGERENT OTHERS’ TO ‘CULTURAL OTHERS’
by
Uffe Bergeton
_______________________________________________________
A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
(EAST ASIAN LANGUAGES AND CULTURES)
May 2006
Copyright 2006 Uffe Bergeton
UMI Number: 1437581
1437581
2006
Copyright 2006 by
Bergeton, Uffe
UMI Microform
Copyright
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
ProQuest Information and Learning Company
300 North Zeeb Road
P.O. Box 1346
Ann Arbor, MI 48106-1346
All rights reserved.
by ProQuest Information and Learning Company.
ii
Dedication
To 小秧
iii
Acknowledgements
This thesis is one of the many products from my time in the department of East Asian
Languages and Cultures at the University of Southern California and I would like to start
by thanking the department, its faculty, staff, as well as my fellow students, for providing
me with an intellectually stimulating environment highly conducive to academic research.
I am particularly indebted to my faculty advisors, Edward Slingerland and Bettine Birge
for their continuous support of my various endeavors and for numerous valuable
comments on earlier versions of this paper. I also owe special thanks to Erica Brindley.
Not only has her own insightful studies on identity in early China been an important
source of inspiration for me, her willingness to read through and comment on half-baked
drafts has played a crucial role in the development of this thesis. I would also like to
thank George Hayden and Dominic Cheung for always being willing to offer helpful
advice and guidance. Finally, but most importantly, I am deeply indebted to my wife, Iris
Yim, for her unfailing support and encouragement, regardless of how quixotic my
projects might be. Needless to say, while all of the people mentioned above have
contributed to this thesis in one way or another, directly or indirectly, and share the credit
of the better parts, I alone am responsible for any inconsistencies, errors and other
shortcomings it may still contain.
iv
Table of Contents
Dedication
ii
Acknowledgements
iii
List of tables
v
List of figures
vi
Abstract vii
Chapter 1 Introduction 1
Chapter 2 Typological survey of compound ethnonyms in pre-
imperial texts
5
Chapter 3 Pre-Zhangguo vocabulary of identity: foreigners as
“belligerent others”
8
3.1 Rong 戎 as general term for ‘belligerent other’
10
3.2 Socio-political changes leading to diminishing
relevance of toponym compounds
13
Chapter 4 The birth of the cardinal direction system of
compounds in the Zhanguo period and the emergence
of yi 夷 as default term for ‘alien other’
18
4.1 Using the numeral compound si yi 四夷 to date the
cardinal direction paradigm
20
4.2 Greater combinatorial potential leading to Yi 夷
being established as default term for ‘barbarian’
23
4.3 Yi 夷 established as default term for ‘barbarian’ in
the Mengzi
28
Chapter 5 Ethnonym compounds 36
Chapter 6 Conclusion 44
Appendix 47
Bibliography 49
v
List of Tables
Table (8): Relative Frequency of Different Types of Ethnonyms 16
Table (14): Directional Compounds 23
Table (27): Ethnonym Compounds in Gugong Corpus 36
Table (28): Ethnonym Compounds 37
vi
List of Figures
Figure 1: Major States of the Spring and Autumn Period 47
Figure 2: Major States of the Warring States Period 48
vii
Abstract
In this paper I describe how changes in the concepts underlying the Zhou-barbarian
dichotomy in pre-imperial China are reflected in the form and distribution of the
linguistic expressions used to refer to foreigners. The main narrative that informed the
conceptualization of foreigners changed from warfare in pre-Zhanguo times, to culture
with the emergence of Confucianism in the late 5
th
to early 4
th
century BCE. In the earlier
period foreigners were referred to using expressions denoting specific peoples or with the
character rong 戎 ‘belligerent others.’ In the Zhanguo period the emphasis was on
cultural differences and a number of general words for anti-Zhou ‘barbarians’ developed
(siyi, yidi, manyi, etc.). Within these new paradigms of compound ethnonyms, the
morpheme yi had the greatest combinatorial potential and was occasionally used alone to
express the concept of ‘cultural others.’
1
Chapter 1
Introduction
Recent work on the pre-imperial origins of the Sino-barbarian dichotomy by Erica
Brindley, Di Nicola Cosmo, Yuri Pines, Poo Mu-chou and David Schaberg among others,
has provided detailed descriptions and analyses of how concepts Zhou and non-Zhou
identity were used in philosophical, political and historiographical writing.
1
This body of
work has pointed out a number of weaknesses in the traditional Sino- and euro-centric
historical models and thereby heightened our awareness of the complexities of this issue.
Schaberg (2001) warned against relying on pre-imperial historiography as an unbiased
source of information about non-Chinese peoples, by demonstrating that early Chinese
texts were interested in “distant and peripheral groups less as an end in themselves than
as foils for central culture.”
2
Brindley (2005) and Di Cosmo (2002) have highlighted the
fluidity both of the geographical borders and the conceptual distinction between Chinese
and ‘barbarians’ on the Southern and Northern frontiers respectively. Furthermore, as
convincingly argued by Poo Mu-chou (1994, 1998, 1999, 2005), Di Cosmo (2001), and
Pines (2005), early Chinese concepts of identity were grounded in culture rather than
1
E.g., Brindley, Erica. “Barbarians or not? Ethnicity and Changing Conceptions of the Ancient Yue 越 (Viet) Peoples
(~400-50 B.C.).” Asia Major 16.1(2005); Di Cosmo, Nicola. Ancient China and Its Enemies. The Rise of Nomadic
Power in East Asian History, 2002; Pines, Yuri. “Beasts or Humans: Pre-Imperial Origins of the “Sino-Barbarian”
Dichotomy.” In Amitai, Reuven and Michal Biran (eds.) Mongols, Turks, and Others. Eurasian Nomads and the
Sedentary World. Brill: Leiden, Boston, 2005, pp. 59-102; Poo Mu-chou 蒲慕州. Enemies of Civilization. Attitudes
toward Foreigners in Ancient Mesopotamia, Egypt, and China. Albany: State University of New York Press, 2005; and
Schaberg, A Patterned Past: Form and Thought in Early Chinese Historiography, 2001. Recent work on identity in
imperial China include Holcombe (1995), Fiskesjö (1999), Dikötter (1992, 1997), Crossley (1999), Eliott (2001),
among many others.
2
See Schaberg (2001:132), also quoted in Pines (2005:69). For the reliability of the speeches in the Zuozhuan for
Chunqiu history, see also Pines (1997).
2
biophysical nature
3
, and the main criterion for ‘Chineseness,’ i.e. ‘Zhou’ identity, was
defined as adhering to elite Zhou ritual practices and customs’.
4
In spite of the amount of interest in the pre-imperial origins of the concepts
underlying the Zhou-barbarian
5
dichotomy, a detailed study of the actual linguistic
expressions used to refer to foreigners is not yet available in English
6
. This paper is
intended as a small step toward filling this gap through a study of the emergence and
evolution of various paradigms of compounds used to refer to non-Zhou ‘others’ in pre-
Qin texts. Since the Chunqiu (722-453 BCE) and Zhanguo (453-221 BCE) periods
7
constitute a crucially important formative period in Chinese intellectual history, a clear
understanding of the way the Zhou-barbarian dichotomy was conceptualized during those
five centuries is a necessary prerequisite to the study of how these concepts were used
3
Dikötter’s (1992:1-30ff) claims about racially grounded identity in pre-imperial China may thus have to be revised.
4
I use the terms ‘Zhou’ and ‘non-Zhou,’ and ‘Zhou-barbarian dichotomy’ rather than the anachronistic ‘Chinese,’
‘non-Chinese,’ and ‘Sino-barbarian dichotomy.’ As observed by Erica Brindley (p.c.), Pines (2005), and Di Cosmo
(2002), the concepts of a shared ‘Chinese’ identity (referred to by the automyms Hua 華, Xia 夏, or Hua-Xia 華夏)
found in pre-imperial texts were to a large extent manufactured by the elite classes and political leaders in the Central
States (Zhongguo 中國) for political reasons. Throughout Chinese history the definitions of these terms were
continuously modified to adapt to changing historical contexts, see Endymion Wilkinson (2000) Chinese History: A
Manual, pp. 95-96, 132, 707-10, 750-53 and his bibliography. In pre-imperial times, ‘Zhou’ identity was mainly based
on adherence to Zhou rituals and elite culture and may thus not have extended to the lower classes in Zhou society, see
Pines (2005:65). Similarly, Loewe (1999:1002) questions the assumption that all the people inhabiting the area unified
in 221 BCE shared a unified culture and points out that population groups differed greatly with respect to a number of
factors ranging from social class, ethnic origin, and religion, to terrain and climate. In this paper I avoid these
complications by limiting myself to ‘otherness’ as conceptualized by the literate Zhou elite who were the authors of the
textual material studied here. For discussion of the importance of ritual norms in pre-imperial China, see Pines (2000).
As argued by Brindley (2005) and Pines (2005), the strongly defined Zhou-barbarian divide that can be ascertained in
Zhanguo discourse is most clearly formulated in Confucian writings, but can also be detected in other contemporary
texts.
5
I use the term ‘barbarian’ to refer to general concepts of ‘otherness.’ As noted by Pines (2005:61, note 8), many
ethnonyms in early Chinese texts, e.g. Yi and Di, were simply used to refer to non-Zhou peoples without any of the
negative connotations attached to the Western ‘barbarian.’ However, as shown below, certain Chinese ethnonyms
could also be used to refer to foreigners in general as ‘others.’ In such cases, I use either the general term ‘barbarian’ or
the more specific terms ‘belligerent others’ and ‘cultural others’ which will be defined in sections 2 and 3 below.
6
Di Cosmo (2002:100) briefly mentions similarities between certain uses of the Chinese word Yidi and the Greek word
for “barbarians.” See also Kruikov, Mikhail, et al. (1978) Drevnie Kitajtsy: Problemy Etnogeneza, pp. 274-76 and 287-
89, who discuss a number of names of non-Zhou tribes which were used as general concepts for ‘barbarians.’ For
resent research published in Chinese and Japanese, see Poo Mu-chou, “Gudai Zhongguo, Aiji yu Lianghe liuyu dui yizu
taidu zhi bijiao yanjiu,” Hanxue yanjiu 17/2 (1999), pp. 137-69, and the references therein; and Ochi Shigeaki “Ka-i
shisô no seiritsu,” Kurume Daigaku Hikaky Bunka Kenkyûjo kijô 11 (1992), pp. 43-137.
7
For convenience, I use the Chinese words Chunqiu and Zhanguo and their English equivalents, Spring and Autumn
and Warring States, interchangeably.
3
and abused throughout the rest of Chinese history till the present day. Similarly, just as
the concepts of identity which were formed in pre-imperial times have influenced later
periods, so has also the paradigms of linguistic expressions used to articulate these
concepts continued to shape the vocabulary of identity used ever since.
Focusing on compound ethnonyms based on the four words man 蠻, yi 夷, rong
戎, and di 狄, I show that there are a number of significant differences in the forms and
usage patterns of ethnonyms in the Chunqiu and Zhanguo periods. Through case-studies
of the ethnonyms used in the Mengzi, as representative of the Zhanguo period, and the
Shijing and Chunqiu, as representative of the Chunqiu period, I argue that developments
in both intellectual history (e.g. the emergence of Confucian universalism) and socio-
political history (e.g. the reduction in the number of non-Zhou statelets due to
acculturation and Zhou military expansion) contributed to these changes, and that the
most important of these was the creation of a concept of ‘cultural others’ in the Zhanguo
period. In the Chunqiu period warfare was the dominant narrative of identity that
informed the concept of foreigners as ‘belligerent others.’ Specific non-Zhou tribes were
referred to by compound ethnonyms consisting of place- or tribal names, denoting
specific geographic location and tribal affiliation, combined with mono-morphemic
ethnonyms (man 蠻, yi 夷, rong 戎, and di 狄), e.g. Huai-Yi ‘the Yi of the Huai river
area’, Shan-Rong ‘the Hill Rong,’ etc. In the Western Zhou and the early Chunqiu
periods, foreigners in general could also be referred to by the term rong 戎 ‘belligerent
others.’ It was only in late Chunqiu and early Zhanguo times that the Zhou developed a
strongly defined sense of cultural unity which served as backdrop for the formation of the
4
concept of ‘cultural others,’ who differed from the Zhou in cultural heritage and ritual
practices
8
. I argue that these changes in the concepts of identity, and in particular the
development of a concept of ‘cultural other,’ are reflected in changes in the relative
frequency of different types of compound ethnonyms based on the four terms
manyirongdi 蠻夷戎狄 in the period stretching from the Western Zhou to the end of the
Warring States.
It has been claimed that Classical Chinese lacked a single term corresponding to
Greek word βάρβαροι ‘barbarians,’ and instead relied on a number of different compound
terms, e.g. yidi 夷狄, manyi 蠻狄, manyirongdi 蠻夷戎狄, etc. to refer to ‘alien others’ in
general, or different sub-sets of ‘barbarians.’
9
While it is true that Zhanguo Chinese had
several compounds serving as general terms for ‘barbarians,’ I argue that morphological
analysis of paradigms of compound ethnonyms based on the four mono-morphemic
ethnonyms man 蠻, yi 夷, rong 戎, and di 狄, clearly shows that the single morpheme
ethnonym yi 夷 became the default expression for ‘cultural others,’ both in compounds
and when used on its own. None of the other three single morpheme ethnonyms man,
rong, and di were used in this way. Thus while there were several words referring to
aliens in general, only one morpheme, i.e. yi 夷, could be used in this way. In this sense,
Zhanguo Chinese can thus be argued to have had a single morpheme for ‘cultural other’
or ‘barbarian’.
8
This does not mean that common Zhou ritual practices did not exist in pre-Zhangguo times. Indeed, such common
rituals did exist, but they did not yet form the backbone of a culturally grounded Zhou-Barbarian dichotomy.
9
See Lattimore (1962:455); Di Cosmo (2002:95, n. 7); and Pines (2006:61, n. 8).
5
Chapter 2
Typological survey of compound ethnonyms in pre-imperial texts
The textual corpus on which the present study is based includes the 27 texts of the
thirteen Classics (Shisanjing 十三經) and the pre-Qin philosophers (XianQin zhuzi 先秦
諸子) in the National Palace (Gugong 故宮) online database.
10
It is only natural that a
very large number of different expressions used to refer to ‘barbarians’ and non-Zhou
peoples should be found in such a diverse body of texts which were composed for widely
different purposes and audiences over a period spanning more than five centuries.
However, due to obvious scope limitations I here limit myself mainly to the study of
compound ethnonyms
11
based on the four characters man 蠻, yi 夷, rong 戎, and di 狄.
These four words have all been used to refer to different non-Zhou peoples, either as
mono-morphemic ethnonyms or in compounds with descriptive prefixes indicating
geographical location or clan affiliation, see (1).
(1) Toponym compounds:
a. Huai-yi
12
淮夷 ‘the Yi of Huai’
b. Jiang-rong 姜戎 ‘the Jiang Rong’
c. Man-jing 蠻荊 ‘the Man of Jing,’ etc.
10
These 27 texts are: Zhouyi 周易, Shangshu 尚書, Maoshi 毛詩, Zhouli 周禮, Yili 儀禮, Liji 禮記, Zuo 左傳,
Gongyang 公羊 傳, Guliang 穀梁傳, Lunyu 論語, Xiaojing 孝經, Erya 爾雅, Mengzi 孟子, Xunzi 荀子, Laozi 老子,
Zhuangzi 莊子, Liezi 列子, Mozi 墨子, Yanzichunqiu 晏子春秋, Guanzi 管子, Shangjunshu 商君書, Shenzi 慎子,
Hanfeizi 韓非子, Sunzi 孫子, Wuzi 吳子, Yinwenzi 尹文 子, Lüshichunqiu 呂氏春 秋. See the Gugong database at
http://210.69.170.100/s25/index.htm. The date of composition and authenticity of many of these texts are still the
subject of considerable controversy. These questions will be addressed as they become relevant in the discussion of
each individual text.
11
The term ‘ethnonym’ is used here in a loose sense to refer to linguistic expressions that refer to variously defined
groups of people. Thus ‘ethnonym’ refers not only to specific tribes or peoples defined in terms of ethnic differences,
but to a wide range of terms ranging from specific tribal names to general terms for ‘barbarians’ and ‘alien others.’
12
All Chinese words are transcribed according to the Pinyin system. Although hyphens are not part of Pinyin
orthography, they are used here to indicate connections between sub-components of compound expressions. Other
transcription systems (Wade Giles, etc.) used in translations quoted here, have be all been changes to Pinyin.
6
In section 2, I show that this type of compound ethnonyms, which I will refer to as
toponym compounds
13
, was prevalent in pre-Warring States texts. In addition to such
toponym compounds, the terms man 蠻, yi 夷, rong 戎, and di 狄 can also be found in
compounds with numeral prefixes (e.g. si-yi 四夷 ‘the four Yi,’ ba-man 八蠻 ‘the eight
Man,’ liu-rong 六戎 ‘the six Rong,’ etc.), or directional prefixes, see (2).
(2) Directional compounds:
a. dong- yi 東夷 ‘the eastern Yi
b. xi-rong 西戎 ‘the western Rong
c. bei-di 北狄 ‘the northern Di
d. nan-man 南蠻 ‘the southern Man,’ etc.
In section 3, I argue that the paradigm of directional compounds only became fully
established in the early Warring States period. Furthermore, a detailed study of the
relative combinatorial potential of man 蠻, yi 夷, rong 戎, and di 狄 in different
paradigms of compound ethnonyms reveals that by mid- to late Warring States period yi
夷 was being used as the default term for ‘barbarian’ or ‘cultural other.’
In section 4, I argue that the evolution of compounds composed only of
combinations of the four terms man 蠻, yi 夷, rong 戎, and di 狄 show that the most
frequently used ethnonyms of this type, shown in (3), emerged as general terms for
foreigners or ‘cultural others’ in the Warring States period.
(3) Ethnonym compounds:
a. man-yi 蠻夷 ‘The Man-yi’
b. rong-di 戎狄 ‘The Rong-di
c. yi-di 夷狄 ‘The Yi-di’, etc
13
The term ‘toponym compound’ is thus not strictly restricted to compounds which include place names (toponyms). I
use it here as a general term for compound ethnonyms referring to specific non-Zhou peoples which contain a
morpheme marking either geographical location or tribal affiliation.
7
In addition to (1-3), many other types of compound ethnonyms can be found in early
Chinese text, e.g. color compounds (e.g. chi-di 赤狄 ‘the Red Di,’ bai-di 白狄 ‘White
Di,’ etc.), and loan word autonyms from non-Chinese languages, e.g. xun-yu 獯鬻 ‘the
Xunyu,’ xian-yun 獫狁 ‘the Xianyun,’ etc. However, due to scope limitations this paper
will focus mainly on the three compound types illustrated in (1-3).
8
Chapter 3
Pre-Zhangguo vocabulary of identity:
foreigners as “belligerent others”
In the Western Zhou (1045-722 BCE) and early Spring and Autumn period (722-453
BCE), the Central States (zhongguo 中國)
14
had not yet achieved a strongly developed
consciousness of cultural cohesion that would later set them apart as a unified block
against a backdrop of non-Zhou ‘barbarian’ peoples.
15
Politically the Central States were
also experiencing the disintegration of the Zhou court and, as a consequence, the rise of
individual states as powerful players in inter-state affairs. This led to a situation in which
each State had to fend for itself for survival through skillful diplomacy and alliance
formation. Under such circumstances real-politik overrode common cultural heritage and
often led Central States to enter into alliances with non-Zhou tribes or stateles in order to
engage fellow Central States militarily. Consequently, a culturally defined dividing line
between Zhou and foreigners was less important than practical security concerns and
14
In addition to Zhongguo, other Zhou autonyms include hua, xia, and huaxi. According to Poo Mu-chou (2005:47),
“scholarly opinion differ on exactly when this term came to be used in the sense of the central state(s), that is, a
geopolitical term denoting the state of Shang, Zhou, and their vassals located in the Central Plain area.” The reader is
referred to Chen Suizheng (1993) and Wang Ermin (1977:441-80) for more discussion of this issue. For discussion of
what may be the earliest attestation of the term Zhongguo in a Western Zhou bronze inscription, see Tang Lan (1934:6-
9). According to Poo Mu-chou (2005:47), it was “during the Eastern Zhou period that [the term Zhongguo] came to
mean the central state(s), that is, the Chinese states that existed in the Central Plain area of north China.” See also Yu
Rongchun (1986:76). The Zhou residents of the Central States (Zhongguo) referred to themselves as the Hua, the Xia,
or the Hua-Xia, see Poo Mu-chou (2005:47). For further discussion of the emergence and use of these terms, especially
in contrast to terms for foreigners, see Wang Zhongfu (1989:363-76), and Ogura Yoshihiko (1967). The etymology of
Hua and Xia is discussed in Huber (1983:181-3), and Zhou Fagao (1975:3952-54), respectively.
15
In this respect the development of a distinction between ‘civilization’ and the realm of ‘barbarians’ in pre-imperial
China mirrors that found in Ancient Greece. Before the 5
th
century BCE the Greeks had not yet developed a strong
sense of cultural cohesion and superiority vis-à-vis the rest of the world. They still viewed themselves as one people
among many, some of with both stronger and with longer cultural traditions than themselves, e.g. the Egyptians. See
Edith Hall (1989) Inventing the Barbarian: Greek Self-definition through Tragedy; Jonathan M. Hall (1997) Ethnic
Identity in Greek Antiquity; and Wilfred Nippel (2002) “The Construction of the ‘Other,’” in Greeks and Barbarians,
ed. Thomas Harrison, 278-310. Edinburgh: Edinburgh University Press, 2002.
9
pragmatic inter-state diplomacy.
16
In other words, warfare was the main narrative which
informed attitudes toward foreigners in pre-Zhanguo times,
17
who therefore were often
conceptualized as ‘belligerent others.’
18
This state of affairs is reflected in the vocabulary
of otherness found in texts from the period, e.g. the Shijing, and the Chunqiu, which
abound in ethnonyms referring to specific non-Zhou groups with whom the Central States
were interacting, either through hostile warfare or peaceful alliances.
Since historical records usually refer to specific battles or alliances the concepts
of ‘belligerent others’ was often expressed linguistically using toponym compounds
referring to specific non-Zhou enemies, e.g. the Huai-Yi and Man-Jing in (4).
(4) a. 淮夷 攸服 […] 既克淮夷 […] 淮夷 卒獲 […] 憬彼淮夷 。
here the Yi of Huai come to submit. […] We have quieted the Yi of Huai. […]
the Yi of Huai have all been dealt with. […] From afar the Yi of Huai come
(Ode 299, Waley 1996:310-312)
b. 蠢爾蠻荊 , 大邦為讎 ![…] 蠻荊 來威。
Foolish were you, Man of Jing, who made a great nation into your foe. [Fang-
shu] […] made the Man of Jing afraid (Ode 178, Waley 1996:153)
16
See Di Cosmo (2002) who discusses cases where members of the Central States labeled each other “barbarians”
(manyi, yidi) for political reasons. The lack of a clearly defined, hermitically tight division between Zhou and non-Zhou
is further illustrated by the fact that the Zhou often intermarried with non-Zhou and that people of non-Zhou origin
often rose to high administrative positions in Central States, as shown by the following Zuo passage: 又娶二女於 戎,
大戎狐姬生 重 耳,小戎子 生 夷吾。晉伐 驪 戎,驪戎男 女 以驪姬,歸 , 生奚齊,其 娣 生卓子。驪 姬 嬖,欲 立 其
子。“Subsequently, he [duke Xian of Jin] married two ladies from among the Rong, the one of whom he called Hu Ji
of the great Rong, bore Chong Er, and the other, who was of the small Rong, bore Yi Wu. When Jin invaded the Li
Rong, their chief, a baron, gave him to wife his daughter, Li Ji, who bore a son called Xi Qi, while her younger sister
bore him Zhuo Zi. Li Ji became the favorite with the duke, and wished to get her son declared his successor.” (Zuo,
Zhuang 28.2, Legge:114). As shown in this passage, off-spring of non-Zhou wives could aspire to the throne. See also
Pines (2005:84, note 78) for discussion of similar examples. See also Zuo, Xiang 29; and Zuo, Zhao 17.
17
It may be the case that peaceful interaction with foreigners was more prominent than extant records would lead us to
believe. As is well known, a large proportion of oracle bone and bronze vessel inscriptions predominantly deal with
warfare questions related to specific battles and alliances. Since early pre-Chunqiu writings on other materials (e.g.
bamboo) have perished, we do not know if they included evidence of other, more peaceful, interaction with and
attitudes towards foreigners.
18
For discussion of views of foreigners as enemies in Ancient Egypt, Mesopotamia, and pre-imperial China, see Poo
Mu-cho (2005:68-79).
10
The Huai-yi 淮夷 in (4a), who are one of the non-Zhou groups most often mentioned in
the Shijing and the Chunqiu, inhabited the Huai River area and the Man-jing 蠻荊 in (4b)
refer to a people living settled somewhere between the Han River and the Yangzi, in what
is now northern Hubei.
19
Another type of archaic location- or tribe-specific compounds
are formed with the suffix –fang 方 ‘side, border, region,’ e.g. man-fang 蠻方 “the Man
(attested in Shijing), gui-fang 鬼方 ‘the Gui’. Fang compounds date back to the Shang
dynasty when they were often used to refer to specific non-Shang or enemy countries,
20
but are also found in Western Zhou inscriptions, and survive into Spring and Autumn
times.
21
Like the toponym compounds in (4a-b), they are not general terms for foreigners,
but rather specific terms for particular non-Zhou groups.
22
3.1 Rong as general term for ‘belligerent other’
In the Spring and Autumn period foreigners were usually thought of as ‘belligerent
others’ and referred to by a number of linguistic expressions denoting specific groups
from specific geographical locations, among which toponym compounds were the most
19
See Waley 1996:154, footnote 1.
20
According to Keightley 1999:269, “Local, usually hostile, rulers with the title of Bo 伯 sallied forth from the Fang
(Y1205.2). The king divined about patrolling and attacking these border areas (Y1208.1-2). Fang was also used as a
suffix to refer to non-Shang or enemy countries that existed in and beyond the borders of the Shang polity.” See also
Chang, Shang Civilization, pp. 216-20, 248-59. For more discussion of fang-ethnonyms, see also Li Xueqin
(1959:61ff); Keightley (2000:66-72); and Di Cosmo (1999:907-9).
21
E.g. man-fang is attested in Ode 256, Waley 1996:263; and gui-fang in Ode 255, Waley 1996:262. In both cases they
are ambiguous between a toponym and ethnonym reading, e.g. ‘the Man region’, or ‘the tribes of the Man region.’
22
While the laconic Shang and Western Zhou inscription material is sparse and lack explicit statements about attitudes
toward foreigners, it seems safe to say that the existing textual evidence indicates a greater concern with practical
interaction with specific non-Zhou groups on a case by case basis, than with a culturally defined across-the-board
delineation of a civilized world vs. a barbarian periphery. The Shang and the Zhou were only one people among many
in the ancient world, some of which may have been technologically more advanced than the themselves (see Di
Cosmo’s discussion of metallurgy, chariots, horseback riding, etc.). In other word, they probably regarded themselves
and other peoples as (sometimes equal) players in a quest for control over material resources and territory. In sum, a
strongly delineated, culturally defined Zhou-barbarian dichotomy most likely did not exist in pre-Chunqiu times.
11
frequent, see table (8a,e-s) on page 10. Since the concept of a culturally defined Zhou-
barbarian distinction was still weak it is not surprising that Chunqiu period texts lack a
single term for ‘cultural other.’ In contrast, there is evidence indicating that rong could be
used as a general term for ‘armed enemy’ or ‘belligerent other.’
Both inscriptional material and received texts from the Western Zhou and Spring
and Autumn periods show that the character rong 戎 had two different meanings. First, it
was used as a name for a specific tribe or people, the Rong.
23
Second, it was used as a
word referring to various aspects of warfare ‘weapons, war chariot, battle, warfare’, etc.
In Chunqiu texts rong 戎 was often used as an adjectival modifier meaning
‘military-‘, or ‘war-,’ e.g. rong-che 戎車 ‘war-chariots’ (Odes 178, 299) and could also
be used on its own in the meaning ‘war chariot’ (Odes 128, 177). Other related meanings
include ‘arms’
24
or ‘weapons’
25
, and ‘attack’, or ‘war.’
26
I argue that this semantic field
also included the meanings ‘armed enemy,’ ‘belligerent,’ or ‘hostile other,’ as illustrated
in the bronze inscription in (5), where, according to Shaughnessy 1991:177-180, the term
rong, which he translates as ‘belligerents,’ actually refers to the people from the state of
Hu, which was a member of the Huai Yi confederation
27
.
23
E.g.: Ode 168, and Ode 300 in (17) and (32) below. In oracle bone inscriptions characters identified as rong, may
have been originated as family or tribal crest (Qiu Xigui (p.c.)).
24
E.g.: 以修我 戎 “and prepare my arms” (Ode 263, Karlgren 1950:234-236). Note that Waley translates the same
passage as “repair my war-chariots” (Ode 263, Waley 1996:281-283), where rong is interpreted as shorthand for the
modified noun rongche 戎車 ‘war-chariot.’ Both translations are linguistically justifiable.
25
E.g.: 戎 成不退,飢成不遂 “The weapons truly are not withdrawn, the famine truly is not relieved” (Ode 194,
Karlgren 1950:140). Note that Waley translates the same passage as “There is war, but we do not withdraw; there is
famine, but we do not progress” (Ode 194, Waley 1996:172-4). Both translations are linguistically justifiable.
26
E.g.: 用戒 戎 作 ,用逖蠻方。“[…] that you may be ready, should war arise, to keep at due distance barbaric tribes”
(Ode 256, Waley 1996:263).
27
Throughout his book Sources of Western Zhou History, Edward Shaughnessy consistently translates rong 戎 as
belligerents when it occurs in phrases such as rong fa 戎伐 ‘the belligerents attacked’ and yu rong 御戎 ‘defend against
the belligerents,’ etc.
12
(5) […] The belligerents [rong 戎] attacked [fa 伐]. X. Dong led the supervisors and captains
to chase after and defend against [yu 御] the belligerents [rong 戎] at Yu Woods, striking
the belligerent [rong 戎] Hu. My cultured mother competitively and earnestly opened the
march, granting firmness to his heart, eternally giving inheritance to his body, and ruling
and conquering his enemies. (I) took one hundred heads, shackled two chiefs, and
captured the belligerents’ weapons: shields, spears, halberds, bows, quivers, arrows,
uniforms, and helmets, in all 135 pieces. (I) captured 114 belligerents [rong 戎] captives
and clothed (them?) […].
28
Here rong is used to refer to a people that presumable could also be referred to more
precisely by the toponym compound huai-yi. According to Poo Mu-chou (2005:45-46),
the fact that rong occurs interchangeably with a specific ethnonym such as Xianyun 獫狁
29
in bronze inscriptions such as (6), suggests that “either the Zhou people confused the
Xianyun with the Rong to the west, or the term Rong was used a general term referring to
the foreign barbarians.”
(6) Lord Bai said: Bu-Qi Yufang, the Xianyun are grievously troubling our western yu, (the
Xi-yu). The king has ordered me to move forward and pursue them to the west. […] You
attacked the Xianyun with great force in my war-chariots, at Gaoling. You cut off many
heads and took prisoners. The Rong gathered together and pursued you for a long was.
You fought hard against the Rong and graciously did not fall into trouble with my
chariots.
30
These examples of rong 戎being used to refer to peoples which already had other specific
ethnic names
31
, indicates that it could serve as a general term for ‘belligerent other’ in the
Western Zhou period.
32
This hypothesis is further supported by the fact that it often
occurs in compounds, preceded by toponyms or other descriptive modifiers serving the
28
The translation of this text, which is from the Dong gui dating from King Mu’s reign (956-918 B.C.), is by
Shaughnessy (1991:177-180) where a transcription of the original Chinese text can also be found.
29
According to Poo Mu-chou (2005:45), the “term [Xianyun] appears to have remained a specific ethnic name.” See
also Wang Guowei (1975:583-605); Pr ůšek, (1971:18-9); and Di Cosmo (1999:919). {See also Pulleyblank (1983)}.
30
“Jaroslav Prusek 1971:126-127. The author of this inscription obviously equated Xianyun with Rong.” (Poo
(2005:footnote 58, p. 167)
31
See Poo Mu-chou 2005:45, note 57.
32
See also Lynn’s (1994) translation of rong as ‘armed men/robbers’ in the Zhouyi: 惕號,莫夜 有戎 ,勿恤。( 周易
第 05 卷, 夬) “Despite cries of alarm that there are armed men night after night, this one need not grieve. […]”(Lynn
1994:405-7);自我致 戎 ,又誰 咎也﹖(周易第 04 卷, 解)“As this one himself attracts robbers, who else should bear the
blame?” (Zhouyi, Lynn 1994:383)
13
purpose of differentiating between rong 戎 ‘belligerents’ coming from different regions,
or different kinds of rong 戎 ‘belligerents,’ xu-rong, shan-rong, xi-rong, see the examples
in rows (e-s) in table (8) on page 10. It is plausible that some of these compounds should
be translated as ‘belligerent from Xu’ ‘Hill belligerents,’ etc. This way of translating
some of these toponym compounds finds support in Shaughnessy’s translation of Yan-
rong as ‘Yan belligerents’.
33
In his list of bronze inscriptions describing Zhou wars with
the Huai Yi, Shaughnessy also lists the Dong fangding II which refers to an attack a
against the Huai-yi 淮夷 referred to as the Huai-rong 淮戎 ‘Huai belligerents.’
34
While it is not the case that Rong was always used as a general term for
‘belligerent others,’ the evidence discussed above clearly indicates that it could be used in
this way. Given that rong was also used as a word for military accoutrements and warfare
it is quite appropriate that it should also be used to express the dominating pre-Zhanguo
concept of ‘otherness,’ i.e. ‘belligerent others.’
3.2 Socio-political changes leading to diminishing relevance of toponym compounds
As argued by Pines (2005:84-5) the mythological geography of a civilized center
(Zhongguo 中國) surrounded by increasingly uncivilized concentric circles of barbarian
peoples, found in the Liji and Shangshu, is in large part a late Warring States invention.
35
33
See “Ban gui” (Sh. 15.79:34), tr. Shaughnessy 1991:251.
34
See Shaughnessy 1991:178: “Dong leads the Tiger braves to repulse the Huai belligerents.” (Sh 49. ho 12:295)
35
E.g.: the “Yu gong” chapter of the Shang shu (Kong Yingda, annot., Shang shu zhengyi, rpt. Shisanjing 6, pp. 146-53;
and the “Wang Zhi” chapter of the Liji (Liji jijie 15, pp. 359-60). See also Pines (2005:84), Di Cosmo 2002, Schaberg
(2001), and Poo Mu-chou (2005).
14
In the Spring and Autumn period Man 蠻, Yi 夷, Rong 戎, and Di 狄 stateles were
frequently found inside the realm of the Central State, see (7).
(7) 公登城 以望 ,見 戎州 。問之, 以告 。公曰 :「 我,姬 姓也 ,何 戎 之有 焉﹖」 翦
之。
Once after Duke [Zhuang] had mounted the city wall to look around, he observed the
place called Rong-zhou. He inquired about it, and was told its name. He then said, “Our
surname is Ji. What have any Rong to do here?” and proceeded to plunder the place. (Zuo,
Ai 17.5, tr. adapted from Legge, p. 850)
The fact that the Duke of Wei can see the Rong state from the city walls shows that non-
Zhou people lived well inside the Central Plain area
36
. The distinction between ‘people in
the state’ guo ren 國人, referring to the Zhou colonizers living in the walled cities, and
‘people of the field’ ye ren 野人, which often referred to non-Zhou peoples living in the
open country between and within Zhou States, also refutes the over-simplified idealized
mythological worldview of a civilized center surrounded by a barbarian periphery
37
. In
other words, since in the Spring and Autumn period the Zhou often lived in close
proximity to non-Zhou groups, their concepts of foreigners were not so much structured
around a culturally defined across-the-board Zhou vs. non-Zhou dichotomy, but rather
based on various concrete forms of interaction, be they hostile or peaceful, with particular
tribes
38
.
36
See Pines (2005:84, note 77).
37
See Hsu (1999:545-586).
38
According to Hall (1989) a similar situation existed in the pre-5
th
century Greece when “many Greek families and
communities had relations of ritualized friendship with non-Greek dynasties, and Greek city-states often looked on
‘foreign’ cultures neighbouring them as partners, equals, and cultural leaders” (p. 12). Unlike the concept of uncivilized,
bestial, anti-Greek barbarians, which developed in the 5
th
century, cultural differences were not of prime importance in
historical narratives mentioning enemies of the Greek. “The social system of the Homeric poems reveals a pattern of
power relations and organization shared by Greeks and non-Greeks alike.” (p. 14) “Nor is there any difference between
the constraints imposed upon Agamemnon and Priam or Hector by the institutions of civic debate.” (p. 15) It is only
later, in the Greek tragedies, that barbarians were “made to behave in ways which fell short of the standards of Hellenic
virtue: they are emotional, stupid, cruel, subservient, or cowardly. Culturally their ways are barbarian; ethnographic
material is used to distinguish their customs from those of Greeks.” (p. 17, italics added here).
15
As typically happens when a large civilization expands through population growth and
military conquests, peoples not belonging to that civilization get pressed further and
further away from fertile lands and thus end up living in mountainous regions or
inhospitable swamps before eventually disappearing completely through intermarriage
and acculturation
39
. According to Pines (2005:85), “by the Zhangguo period, the Rong,
Di and Yi statelets began disappearing from historical accounts. In all likelihood, Zhou
states absorbed their neighbors. […] It seems that the absorption was cultural and not only
military, as by the age of the imperial unification we can no longer identify alien pockets
on the Central Plain. We may plausibly assume, therefore, that the majority of the aliens
living within the Zhou world were incorporated by the neighboring Zhou states,
providing thereby a good example of successful acculturation.” I find Pines’ acculturation
account very persuasive and argue it can be further corroborated by the changes in the
distribution of different types of compound ethnonyms that took place between Chunqiu
and Zhanguo periods. Rows (a-d) in the table in (8) lists the relative frequency of the
three types of compounds ethnonyms based on the terms man, yi, rong, di, investigated in
this paper (i.e. toponym, directional, and ethnonym compounds, see (1-3) above)
40
, in the
39
As argued in Pines (2005:85, note 79), the following Zuo passage illustrates how, in the Spring and Autumn period,
non-Zhou groups were relegated to infertile lands: 「來! 姜戎 氏!昔 秦人 迫逐乃 祖吾離 于瓜州 ,乃祖 吾離被苫
蓋、蒙荊寒 來 歸我先君, 我 先君惠公有 不 腆之田,與 女 剖分而食之 。[..] 」對曰:「 昔 秦人負恃其 眾 ,貪于 土
地,逐 我諸戎 。惠公蠲其大 德,謂 我諸戎 ,是四嶽之裔 冑也,毋是翦 棄。賜我南鄙 之田,狐貍所 居,豺狼 所
嗥。我諸戎 除 翦其荊棘,驅其狐貍豺狼,以為先君不侵不叛之臣,至于今不貳。“Come, you chief of the Jiang
Rong! Formerly, the people of Qin drove Wu Li, one of your ancestors, to Gua Zhou, when he came, clothed with
rushes and forcing his way through briars and thorns, and threw himself on our ruler duke Hui, who cut off from Qin
some poor lands, and gave them to you to afford you a subsistence. […]” The viscount replied, “Formerly, the people
of Qin, relying on their multitudes, and covetous of territory, drove out us Rong. Then [your] duke Hui displayed his
great kindness; and considering that we Rong were the descendants of the [chief of the] four mountains […], and were
not to be entirely cut off and abandoned, he gave us the lands on his southern border. The territory was one where
jackals dwelt and wolves howled, but we Rong extirpated the briars and thorns from it. Drove away the jackals and
wolves, and considered ourselves his subjects, who would not make inroads on his State, nor rebel. Nor to the present
day have we swerved from our allegiance.” (Zuo, Xiang 14, Legge, pp. 463-464)
40
As mentioned in section 1 above, other types of compounds existed, but are not discussed here.
16
Chunqiu (C) and the Shijing (S), in column (8iii), the Mengzi (M), the Xunzi (X), the
Lüshi Chunqiu (L), and the Zuo commentary (Z)
41
combined, in column (8ix), and the
Mengzi (M) and Xunzi (X) combined in column (8vi). Rows (8e-t) lists the different
toponym compounds found in those texts.
Table (8) Relative frequency of different types of ethnonyms
(i)
C
(ii)
S
(iii)
CS
(iv)
M
(v)
X
(vi)
MX
(vii)
L
(viii)
Z
(ix)
MXLZ
a. Toponym (+fang-) compounds 20/26 =
77%
1/20 =
5%
16/87 =
18%
b. Directional (row c) + ethnonym
(row d) compound (+si-yi)
6/26 =
23%
19/20 =
95%
71/87 =
82%
c. Directional compounds 3/26 =
11.5%
40/87 =
46%
12/20 =
60%
d. Ethnonym compounds 3/26 =
11.5%
31/87 =
36%
7/20 =
35%
e. Huai-yi 淮夷 2 8 10 0 0 0 0 2* 2/*0
f. Hun-yi 混夷 0 1 1 0 0 0 0 0 0
g. Kun-yi 昆夷 0 0 0 1 0 1 0 0 1
h. Yi-hu 夷虎 0 0 0 0 0 0 0 1 1
i. Yi-yi 裔夷 (ethnonym comp.?) 0 0 0 0 0 0 0 (1) (1)
j. Yi Hui 夷穢 0 0 0 0 0 0 (1) 0 (1)
k. Man-jing 蠻荊 0 2 2 0 0 0 0 0 0
l. Shan rong 山戎 1 0 1 0 0 0 0 2* 2/*0
m. Mao rong 茅戎 1 0 1 0 0 0 0 1* 1/*0
n. Luo rong 雒戎 1 0 1 0 0 0 0 0 0
o. Jiang rong 姜戎 1 0 1 0 0 0 0 2* 2/*0
p. Lu hun zhi rong 陸渾之 戎 2 0 2 0 0 0 0 2* 2/*0
q. Li rong 驪戎 0 0 0 0 0 0 0 1 1
r. Cao Zhong zhi Rong 草中 之戎 0 0 0 0 0 0 1 0 1
s. Quan Rong 犬戎 0 0 0 0 0 0 1 0 1
t. Man-fang 蠻方 0 1 1 0 0 0 0 0 0
u. Total Toponym 8 12 20 1 0 1 3 12 16/*7
41
Not surprisingly, the Warring States texts that display the largest number of toponym compounds are historiographic
and encyclopedic works like the Zuo zhuan and the Lüshi Chunqiu which discuss non-Zhou groups in earlier times,
some of which had ceased to exist in the Warring States period. The asterisks added in column (8viii) indicate instances
where toponym compounds found in the Zuo zhuan are part of commentaries on Chunqiu passages which mention the
same toponym compounds. In such cases, these compounds are clearly archaic relics from the past, rather than
representative of the productive Warring States vocabulary of identity, which will be studied in more detail in sections
3 and 4. As shown in (8uix), subtracting these ‘relics’ brings the total number to toponym compounds in Warring States
texts down from 16 to 7.
17
Row (8a) tells us that the frequency of toponym compounds fell from 77% in Spring and
Autumn period texts, see (8iiia), to 18% in the Warring States period, see (8aix), and only
5% in Warring States philosophical texts like the Mengzi and the Xunzi, see (8avi). This
decline in the use of tribe-specific toponym compounds is clearly a reflection of the
socio-political changes that occurred in the transition from the Spring and Autumn period
to the Warring States period, i.e. the reduction in the number of non-Zhou tribes and
statelets due to acculturation and military expansion (triggered among other things by a
growing Zhou population). While these demographic and socio-political factors may
account for decrease in the number of the toponym ethnonyms in (8a), they do not
account for the simultaneous rise in the frequency of directional compounds (e.g. dong-yi,
xi-rong, bei-di, etc.) and ethnonym compounds (e.g. manyi, rongdi, etc.), see (8b-d). In
the following two sections I describe the emergence of these two paradigms of compound
ethnonyms which were mainly triggered by changes in intellectual history.
18
Chapter 4
The birth of the cardinal direction system of compounds in the Zhanguo
period and the emergence of yi as default term for ‘alien other’
In Shang mythological geography the world was conceived of as a large square with the
Shang realm in the center (Zhong Shang 中商), surrounded by a peripheral domain
divided into four regions (si fang四方), or four lands (si tu 四土), corresponding to the
four cardinal directions.
42
This worldview continued into later periods, with Zhongguo
中國 replacing Zhong Shang 中商 as the center. However, it was not before the late
Chunqiu or early Zhangguo period that the terms man 蠻, yi 夷, rong 戎, and di 狄
became systematically associated with the four cardinal directions (si-fang 四方) as
shown in (9).
(9) North
di 狄
West Center East
xi 戎 zhongguo 中國 yi 夷
man 蠻
South
42
“In cosmological terms the Shang conceived of a square world, oriented to the cardinal points, surrounding the core
area known at the Zhong Shang 中商, literally, “center Shang.” […] Beyond the core area, the Shang domain was
divided, ideally, into four areas, known as “the Four Lands” (si tu 四土) or “the Lands” (tu) named for the cardinal
directions. Thus, a series of Period V charges on a single scapula fragment (HJ 36975) starts with the general charge, [
今] 歲商受 [ 年] “[This] year Shang will receive [harvest],” followed by four subcharges about the East, South, West,
and North Lands receiving harvest. Similarly, prayers for harvest were directed to the Powers of “the Four Regions” (si
fang 四方) (HJ 14295).” (Keightley 1999:269) For discussion of the term Zhong Shang, see J.A. Lefeuvre, “An Oracle
Bone in the Hong Kong Museum of History and the Shang Standard of the Center,” Journal of the Hong Kong
Archaeological Society 7 (1976-8): 51: David N. Keightley, “The Late Shang State: When, Where, and What?” in The
Origins of Chinese Civilization, ed. David N. Keightley (Berkeley: University of California Press, 1983), pp. 533, 548-
51.” For more discussion of the interplay of cosmology and politics, see Wang Aihe. Cosmology and political culture
in early China. Cambridge: Cambridge University Press, 2000.
19
The association of ethnonyms with cardinal directions is perhaps most explicitly stated in
the Wang Zhi chapter of the Liji, see (10).
(10) 中國戎夷 , 五方 之民,皆有其性 也 ,不可推 移 。東方曰夷 ,被髮文 身 ,有不火 食 者
矣。南方曰蠻 ,雕題交 趾,有不 火 食者矣。西 方曰戎 ,被 髮衣皮, 有 不粒食者 矣 。
北方曰狄 , 衣羽毛穴 居 ,有不粒 食 者矣。中 國 、夷、蠻 、 戎、狄, 皆 有安居、 和
味、宜服 、 利用、備 器
The people of those five regions
43
– Central States, and the Rong, Yi, (and other wild
tribes round them) – had all their several natures, which they could not be made to alter.
The tribes on the east were called Yi. They had their hair unbound, and
tattooed their bodies. Some of them ate their food without its being cooked.
Those on the south were called Man. They tattooed their foreheads, and had
their feet turned in towards each other. Some of them (also) ate their food without its
being cooked.
Those on the west were called Rong. They had their hair unbound, and wore
skins. Some of them did not eat grain-food.
Those on the north were called Di. They wore skins of animals and birds, and
dwelt in caves. Some of them also did not eat grain-food.
The people of the Central states, and those of Yi, Man, Rong, and Di, all had
their dwellings, where they lived at ease; their flavors which they preferred; the clothes
suitable for them; their proper implements for use; and their vessels which they prepared
in abundance. (Liji Wangzhi 3.13, Legge, pp.229-230)
Together with the Central States, the ‘barbarians’ on the four sides comprise all the
subjects who should pay tribute to the Son of Heaven, see (11) where the directional
compounds dong-yi 東夷, xi-rong 西戎, nan-man 南蠻, and bei-di 北狄 are lumped
together with the Central States to refer to all the subjects of the Zhou ruler.
(11) 故東夷、西戎、南蠻、北狄 、中國諸侯莫不 賓 服。(管子, 小匡)
Hence, of the Yi, who live in the east (dong-yi), the Rong, who live in the west (xi-rong),
the Man, who live in the south (nan-man), and the Di, who live in the north (bei-di), as
well as of all the marquises of the Central States (Zhongguo), there were none who did
not pay tribute. (Guanzi, Xiaokuang)
43
Note that wu-fang ‘the five regions’ mentioned here include the four cardinal direction plus the Central States
(Zhongguo) as the fifth region.
20
As shown in (10-11) the directional compounds dong-yi, nan-man, etc. should be
translated as ‘the Yi, who live in the east,’ the Man, who live in the south,’ etc.
Translating these terms as ‘the southern Man,’ ‘the eastern Yi,’ etc. might lead the reader
to the mistaken assumption that there is a distinction between the southern Man, the
eastern Man, the northern Rong, etc. In other words, after the terms man 蠻, yi 夷, rong
戎, and di 狄 became firmly associated with the cardinal directions in the Warring States
period, as shown in (10-11), the single morpheme ethnonyms (man, yi, rong, di) and the
more redundant corresponding directional compounds (nan-man, dong-yi, xi-rong, bei-di)
could be used more or less synonymously. This change constitutes the first two stages in
the development of the paradigm of cardinal direction compound ethnonyms, see (12).
(12) (i) Stage 1 (ii) Stage 2
a. rong 戎 => rong 戎 ≈ xi-rong 西戎
c. di 狄 => di 狄 ≈ bei-di 北狄
b. man 蠻 => man 蠻 ≈ nan-man 南蠻
4.1 Using the numeral compound si yi 四夷 to date the cardinal direction paradigm
Although these terms si tu 四土 and si fang 四方, used to refer to the four lands or regions
surrounding the center, be it the Zhong Shang or Zhongguo, date back to Shang times, the
expression si yi 四夷 as a cover term used to refer to all the (uncivilized) ‘barbarians of
the four (directions/regions)’ does not seem to emerge before the Zhangguo period. In
21
the Gugong corpus the earliest attested occurrences of si yi 四夷 are from the Zuozhuan
and the Mozi, see (13).
44
(13) a. 凡諸侯有四夷 之功,則獻于王, 王 以警于 夷 ; 中國 則否。諸侯 不相遺俘。
When a prince has gained successes over any of the wild tribes (si-yi), he
presents the spoils to the king, who employs them to terrify the other tribes.
Spoils taken by one State from another are not so presented; and the princes do
not send of their spoils to one another. (Zuo, Zhuang 31.1, Legge, 119)
b. 王既已克殷 ,成帝之來 ,分主諸神 ,祀紂先王 ,通維 四夷 ,而 天下 莫不
賓,焉襲 湯 之緒,此 即 武王之所 以 誅紂也。
After King Wu had conquered the Shang dynasty and received the gifts bestowed
by God, he assigned guardians to the various spirits, instituted sacrifices to
Chou’s ancestors, the former kings of Shang, and opened up communications
with the barbarians of the four quarters (si-yi), so that there was no one in the
world (tianxia) who did not pay him allegiance. (Watson, 1963, Mozi, Against
Offensive Warfare, p. 58)
From the contexts in (13a,b) it is clear that the ‘barbarians’ in the expression si yi 四夷
refer to the ‘barbarians’ of the four cardinal directions, i.e. the man 蠻 ≈ nan-man 南蠻,
yi 夷 ≈ dong-yi 東夷, rong 戎 ≈ xi-rong 西戎, and di 狄 ≈ bei-di 北狄. In the Zuo
passage in (13a) the rituals of warfare with the uncivilized si yi 四夷 are contrasted with
the rituals regulating warfare among the Central States. By the time of the composition of
the Zuozhuan
45
, a culturally grounded Zhou-barbarian dichotomy had thus emerged and
was being used to distinguish between the civilized world of the Central States
(Zhongguo 中國) and the realm of the uncivilized barbarians of the four regions (si yi 四
夷). In the Mozi passage in (13b) si yi 四夷 is also clearly used as a cover term to refer to
all non-Zhou peoples. Interestingly, here the si yi 四夷 together with the Zhou are
44
That is, si-yi only is not attested before the 4
th
or 5
th
century BCE at the earliest.
45
The Zuozhuan commentary to the Chunqiu clearly reflects Confucian thought. It is therefore reasonable to assume
that is composed in the 4
th
century or later.
22
assumed to constitute all the people in the world; that is the ‘barbarians’ are included in
the domain governed by the ruler of everything under heaven (tianxia 天下)
46
.
The expression si yi 四夷 occurs 19 times in 10 different texts in the Gugong
corpus, i.e. four tokens in the Zuozhuan, three in the Shangshu and the Zhouli, two in the
Zhuangzi and the Lüshi Chunqiu, and once in the Liji, the Mengzi, the Xunzi, the Mozi,
and the Guanzi. In contrast, the logically possible numeral compounds si-rong 四戎, si-di
狄四, and si-man 四蠻 are not attested in any of the texts searched. This clearly indicates
that by the Warring States period yi 夷 had become the default term for all the non-Zhou
barbarians of the four directions.
47
Notice furthermore that all the texts in which si yi 四
夷 is found were composed in the Warring States period. That is, si-yi 四夷 is not attested
in the Shijing or the Chunqiu Annals.
In conclusion, not only can the distribution of the expression si yi 四夷 be used to
date the establishment of the paradigm of cardinal direction ethnonyms to the Zhanguo
period, it can also be used to argue that yi 夷was being used as the default term for
‘barbarian.’
48
This claim will be further supported in the following section where I
46
See Pines (2002) for more on the evolution of the cultural and political aspects of the guo vs. tianxia distinction, and
Pines (2005) for how it relates to the concepts of ‘otherness’. See also Pines (2005) for a discussion of the role of the
Sino-barbarian dichotomy in Mozi’s philosophy of impartial caring (jian ai). The fact that si-yi (as well as directional
and ethononym) compounds are used in the Mozi shows that, despite philosophical differences and different
evaluations of foreign cultures, Warring States texts tend to share the same vocabulary of ‘otherness.’
47
The same point can be made with the numeral compound jiu-yi ‘nine yi.’ Again, none of the the logically possible
numeral compounds jiu-rong, jiu-di, and jiu-man are not attested. Other numeral compounds are found, e.g. ba-man,
liu-rong, wu-di, but they are basically confined to few Liji passages. In contrast, both si-yi and jiu-yi are found in wide
range of Zhanguo texts.
48
Notice that the Zuo passage in (13a) also illustrates the use of yi alone. This use of yi as general word for ‘barbarians’
will be discussed in more detail in section 3.3.
23
discuss other evidence that yi 夷has the greatest combinatorial potential of all the four
terms man 蠻, yi 夷, rong 戎, and di 狄.
4.2 Greater combinatorial potential leading to Yi 夷 being established as default
term for ‘barbarian’
In a morphological paradigm of directional ethnonyms composed of one of the four
directional prefixes dong-, xi-, nan-, bei- followed by one of the four single morpheme
ethnonyms terms man 蠻, yi 夷, rong 戎, and di 狄 there are 16 possible combinations.
However, as shown in (14), only eight are attested in the Gugong corpus.
Table (14) Directional Compounds
(i) Compound (ii) Occurrences in corpus (iii) Prefix/ethnonym
a. dong-yi 東夷 20 (east/east)
b. xi-rong 西戎 8 (west/west)
c. bei-di 北狄 7 (north/north)
d. nan-man 南蠻 5 (south/south)
e. bei-rong 北戎 7 (north/west)
f. xi-yi 西夷 5 (west/east)
g. nan-yi 南夷 4 (south/east)
h. bei-yi 北夷 1 (north/east)
i. Total 57
The data in (14) allows us to make a number of observations. First, the most frequently
found directional ethnonyms are those in (14a-d) where the prefix and single morpheme
ethnonym both refer to the same cardinal direction, see (14iii). The fact that these
redundant or ‘matching’ compounds (14a-d) account for 40/57 = 70% of all the
directional compounds in the Gugong corpus supports the conclusion made above that
these compounds should be translated the ‘Di, who live in the west,’ ‘the Man, who live
24
in the south,’ etc. That is, rather than adding a distinguishing feature, the directional
prefix simply reinforces the cardinal direction which is already associated with the
following mono-morphemic ethnonym. Second, the data in (14) show that di 狄 can only
be found with the bei- 北 ‘north’ prefix. Likewise, man 蠻 can only be found with the
nan- 南 ‘south’ prefix. In contrast, rong 戎 is found both with the expected xi- 西 ‘west’
prefix, and with the bei- 北 ‘north’ prefix. Finally, and most importantly, the yi 夷 can be
found with all four directional prefixes.
While the redundant forms in (14a-d) are predicted by the analysis of directional
compounds proposed above, the existence of forms with mismatching cardinal directions,
see (14e-h), needs to be explained in other ways. First let us take a closer look at the
seven tokens of bei-rong. Of these, three tokens are found in the same Chunqiu passage
quoted in the Zuo, Guliang and Gongyang commentaries and should therefore only be
counted once.
49
The remaining four tokens are all found in the Zuozhuan. Of these three
refer to a specific group of rong 戎attacking the State of Qi from the north.
50
The last
instance refers to a rong 戎 attack on the state of Zheng, also presumably from the
49
This Chunqiu passage occurs as follows in the three commentarial traditions: ( 經十.三)夏,齊侯、許男伐北
戎 。(公羊傳, 僖公); (經十 .四)夏,齊侯、許男伐北戎 。(穀梁傳, 僖公); (經十 .四)夏,齊侯、許男伐北
戎 。(左傳, 僖公).
“In summer, the marquis of Qi and the baron of Xu invaded the northern Rong.” (Zuo, Xi 10.4, Legge, p. 156)
50
(i) 北戎 伐齊, 齊侯 使乞師于鄭。鄭太子忽帥師救齊。六月,大敗戎 師 ,獲其二帥
“The northern Rong had invaded Qi, which sent to ask the assistance of a force from Zheng. Hu, the eldest son of the
earl of Zheng, led a force accordingly to the help of Qi, and inflicted a great defeat on the Rong, capturing their two
leaders, […].” (Zuo, Xuan 6.4, Legge 1960)
(ii) 初, 北戎 病 齊,諸侯救之,鄭公子忽有功焉。 “Formerly when the northern Rong were distressing Qi, many
of the princes sent to its relief, and Hu, son of the earl of Zheng, acquired merit. […].”(Zuo, Xuan 10.5, Legge 1960:55)
(iii) 鄭昭公之 敗北戎 也,齊 人將妻之。 “When duke Zhao of Zheng [i.e. the earl’s son Hu, afterwards Duke Zhao]
defeated the northern Rong, the marquis of Qi wished to give him one of his daughters to marry. […].” (Zuo, Xuan
11.3, Legge 1960:57)
25
North.
51
In sum, all the instances of the compound bei-rong are either from the Annals, or
from the Zuozhuan, and are thus confined to the tradition of historiographic texts based
on the Annals.
In contrast, the ‘matching’ compound xi-rong (west-west) is found in a wide-
range of Zhanguo texts and is thus much more representative of the Zhanguo vocabulary
of identity. The eight tokens are spread out over eight works, i.e. the Shangshu, the
Shijing, the Liji, the Zuozhuan, the Liezi, the Guanzi, the Hanfeizi, and the Lüshi Chunqiu.
The fact that seven of these works can be shown to date from the Warring States supports
my claim that the cardinal direction system of (matching) compound ethnonyms
developed after the end of the Spring and Autumn period. The only instance of xi-rong
from a confirmed pre-Zhanguo text is from Ode 168, see (15).
(15) 赫赫南仲 , 薄伐西戎 。[…] 赫赫南 仲,玁狁 于夷。
“[…] awe-inspiring is Nan-zhong, he attacks the Western Rong. […] Awe-inspiring is
Nan-zhong, the Xian-yun are pacified.”(Ode 168, Karlgren 1950:112)
Note that within the same Ode, the great military leader Nan-zhong is praised for having
defeated both then xi-rong and the Xian-yun. As mentioned, in section 2 Xian-yun is
probably a loan-word from a non-Chinese language which entered the Chinese language
as a transcription of an autonym used by a particular foreign people to refer to themselves.
The juxtaposition of the xi-rong with the Xian-yun as enemies of the Zhou defeated in
specific battles clearly show that xi-rong also refers to a specific tribes of Zhou
enemies.
52
51
北戎 侵鄭。鄭 伯禦之,患戎師 “The northern Rong [their seat was in pres. dep. of Yung-p‘ing, Chih-le] made a
sudden raid into Zheng. The early withstood them, but was troubled by the nature of their troops.” (Zuo, Yin 9.6, Legge
1960)
52
Given the fact that rong could be used as a general term for ‘belligerent others’ in pre-Zhanguo times, is not
impossible that terms xi-rong and xianyun in (15) refer to the same non-Zhou group.
26
The only other directional compound found in confirmed Chunqiu texts is nanyi 南夷,
see (16)
53
.
(16) 淮夷 來同,莫不率從 […] 淮夷 蠻 貊,及彼南夷 ,莫不率從。
“The Yi of the Huai River came to terms, there were none that did not obey. […] The
Yi of Huai, the Man, and the Mo, and the Yi of the south, there were none that did not
obey” (Ode 300, Waley 1996:315)
As in the case of xi-rong in (15) above, the expression nan-yi in (16) is listed along with
other ethnonyms referring to specific non-Chinese tribes, i.e. the Yi of Huai, the Man, and
the Mo tribes. The fact that nan-yi and huai-yi are mentioned in the same line as two
different peoples clearly shows that, at the time of the Shijing, nan-yi did not refer to all
the barbarians in the south, but rather to a particular kind of Yi, the southern Yi.
54
In sum, out of the 57 directional compounds in the Gugong corpus, only 3, or 5%,
i.e. bei-rong, xi-rong, and nan-yi, are from confirmed pre-Warring States texts. However,
the fact that two of these, i.e. bei-rong (north-west), and nan-yi (south-east), are
directional ‘mismatches’ indicates that the terms man 蠻, yi 夷, rong 戎, and di 狄 had
not yet become systematically associated with the four cardinal directions (si-fang 四方),
see (9), and that, as a consequence, the cardinal direction paradigm of compound
ethnonyms composed of matching directional prefixes and single morpheme ethnonyms,
see (10-11) and (14a-d), had not yet become productive. This conclusion is further
supported by the conspicuous absence of the compound si-yi from both the Annals and
the Shijing and other pre-Zhanguo texts.
53
Nan-yi is also attested in bronze inscriptions. See Shaughnessy (1991:178-9), Poo Mu-chou (2005:46), and Zhao
Tiehan (1965). {Search the CHANT database.}{DONG-Yi is also found in Western Zhou Bronze inscriptions.}
54
However, such sporadic occurrences of directional compounds do not prove that such a system was fully-fledged.
The expression nan-yi found in the bronze inscriptions mentioned in Shaughnessy 1991:178-9, refers to a specific non-
Zhou people, i.e. the Huai-yi, fighting specific battles with the Zhou, not to ‘all the barbarians living in the south.’
27
Now let us return to the most interesting fact about the data in (16), i.e. that while
man 蠻, rong 戎, and di 狄 are confined to matching directional prefixes
55
, the term yi 夷
can be found with all four directional prefixes, see (14a,f,g,h). That only yi 夷 has
realized its full combinatorial potential strongly suggests that it had become the default
term for ‘alien other’ or ‘uncivilized barbarian’ in by the middle of the Warring States
period. In other words, the data in (14a,f,g,h) lead us to assume the existence of a third
stage in the evolution of the vocabulary of identity, see (17). In (17) the main changes are
indicated by the forms outside the parentheses. The forms in parentheses are ‘relics’ from
earlier layers of the language that continue to function as quasi-synonyms in later stages.
(17) (i) Stage 1 (ii) Stage 2 (iii) Stage 3
a. (shan-)rong => xi-rong ( ≈ rong) => xi-yi ( ≈ xi-rong ≈ rong)
b. man-(jing) => nan-man ( ≈ man) => nan-yi ( ≈ nan-man ≈ man)
c. di => bei-di (≈ di) => bei-yi ( ≈ bei-di ≈ di)
d. (huai-)yi => dong-yi ( ≈ yi) => dong-yi ( ≈ yi)
The first stage is represented by the vocabulary of identity found in the Shijing and the
Annals, where, as shown in (8aiii) in section 3, toponym compounds etnonyms referring
to specific non-Zhou peoples dominate.
In the second stage, which begins in the Late Chunqiu or early Zhanguo period,
the terms man, yi, rong, di become associated with cardinal directions and a productive
paradigm of ‘direction-matching’ directional compounds emerge, see (10-11) and (14a-d).
As shown in (8c), the overall percentage of directional compounds increases from 11.5%
in the Shijing and the Annals, see (8ciii), to 60% in late Warring States philosophical
texts such as the Mengzi and the Xunzi, see (8cvi).
55
Except for bei-rong discussed above.
28
After the terms man, yi, rong, and di become firmly associated with the four
cardinal directions as shown in (9), the addition of a ‘matching’ directional prefix leads to
the formation of semantically redundant forms in which the cardinal direction is marked
twice, e.g. xi-rong (west-west), nan-man (south-south) etc., see (17ii). Since bei-di, xi-
rong, etc. are redundant expressions, they can be replaced by the yi-based directional
compounds in (17iii) without becoming ambiguous. The directional prefix now retains
the distinctive semantic feature needed to tell the barbarians of the four cardinal
directions apart, and we end up with the system in stage 3 in (17), where xi-yi becomes
quasi-synonymous with both xi-rong and rong, etc. The emergence of this new type of
directional compounds in (17iii) and the numeral compound si-yi clearly establishes the
expression yi’s role a default term for ‘cultural other,’ and leads to its being used with
this meaning on its own.
4.3 Yi 夷 established as default term for ‘barbarian’ in the Mengzi
In this section I show that the forms and usage patterns of compound ethnonyms found in
the Mengzi are representative of the vocabulary identity of the mid- to late Zhangguo
period. I also argue that the increasing replacement of specific ethnonyms, e.g. xi-rong
and bei-di, with the semantically more neutral yi-based ethnonyms (xi-yi, bei-yi), i.e. the
shift from Stage 2 to Stage 3 in (17ii-iii) above, reflects changes in intellectual history,
e.g. the hardening of a culturally defined Zhou-barbarian dichotomy
56
and the advent of
56
As noted by Pines (2005:90), due to the accelerating disintegration of the Zhou ritual system toward the end of the
Zhanguo period “the “Sino-barbarian” dichotomy became less pronounced and it was no longer an easy task to decide
who belonged to “us” and who became Other. Thus the pejorative treatment of Qin and Chu [as “barbarian” states] did
not prevent many leading thinkers, including Xunzi and Hanfeizi, from traveling to these states with the hope of
29
Confucian universalism. Finally I show that, in the Mengzi, yi was consciously used as
the default term for “barbarian” both as part of compound ethnonyms and when used on
its own.
The different types of man, yi, rong, di based compound ethnonyms found in the
Mengzi are listed in (18-20).
57
(18) Toponym compounds in the Mengzi (total 1/14=7%):
kun-yi 昆夷 ‘the Kun tribes’ (M 1B:3)
(19) Directional compounds (+si-yi) in the Mengzi (total 10/14=71.5%):
a. dong-yi zhi-ren 東夷之人 ‘Eastern barbarian’ (M 4B:1)
b. xi-yi 西夷 ‘Western barbarians’ (3M 1B:11, 3B:5, 7B:4)*1
c. xi-yi zhi-ren 西夷之人 ‘Western barbarian’ (M 4B:1)
d. bei-yi 北夷 ‘Northern barbarians’ (M 7B:4, Lau 1970:194-5)
e. bei-di 北狄 ‘Northern barbarians’ (M 1B:11, 3B:5)
f. nan-man 南蠻 ‘southern barbarian’ (M 3A:4)
g. si-yi 四夷 ‘barbarian tribes on the four borders’ (M 1A:7)
(20) Ethnonym compounds in the Mengzi (3/14=21.5%):
a. rong-di 戎狄 ‘Belligerent Di’ (M 3A:4, 3B:9, both from Ode 300)
b. yi-di 夷狄 ‘the northern and southern barbarians’ (M 3B:9)
Only 7% of the compounds are ‘archaic’ toponym ethnonyms, see (18), while the
remaining 93% are of the ‘new’ types which only emerged in the Zhanguo period, i.e.
directional (71.5%), or ethnonym compounds (21.4%), see (20-21). This is the exact
opposite of the distributional pattern prevalent in vocabulary of identity in Chunqiu texts,
see (8a) on page 10 above. As argued in the previous section, the presence of a
achieving appointment; serving so-called “barbarian” states was, therefore, entirely legitimate.” However, since
Confucianism is founded on the assumed superiority of Zhou ritual, a culturally defined Zhou-barbarian dichotomy is
still maintained in late Confucian texts such Xunzi and is strongly pronounced in earlier Zhanguo Confucian texts such
as the Mengzi and the Zuozhuan.
57
That is, not included in the present discussion of compound ethnonyms in the Mengzi are: di-ren 狄人 ‘Di
tribes’(Mengzi 1B:14, 1B:15), and xun-yu 獯鬻 ‘the Xun-yu’ (M IB:3).
30
productive paradigm of directional compounds (19) is one of the main characteristics of
Zhanguo vocabulary of ‘otherness’.
Why did an yi-based system of directional ethnonyms appeal to Mengzi? More
specifically, why are specific mono-morphemic ethnonyms like di 狄 and rong 戎,
replaced by the more general term yi 夷 in directional compounds? The Mengzi passage
in (21), which contains Mengzi’s extremely provocative claim that Shun 舜 and King
Wen 文王 - two of the most respected sage-kings in Chinese history - were of barbarian
stock, may help us answer this question.
(21) 舜 生於諸馮 ,遷於負 夏 ,卒於鳴 條 , 東夷 之人 也。 文王 生於岐周, 卒 於畢郢, 西
夷 之人 也。地之相去 也 ,千有餘 里 ;世之相 後 也,千有 餘 歲。得志 行 乎 中國 ,若
合符節。 先 聖後聖, 其 揆一也。
Mencius said, ‘Shun was an Eastern barbarian (dong-yi); he was born in Chu Feng,
moved to Fu Hsia, and died in Ming T’iao. King Wen was a Western barbarian (xi-yi);
he was born in Ch’i Chou and died in Pi Ying. Their native places were over a thousand
li apart, and there were a thousand years between them. Yet when they had their way in
the Central Kingdoms (Zhongguo), their actions matched like the two halves of a tally.
The standards of the two sages, one earlier and one later, were identical.’ (Mengzi 4B:1,
Lau 1970:128)
To most people in Mengzi’s target audience
58
, the statement that Shun 舜 was an ‘eastern
barbarian’ (dongyi zhi ren 東夷之人) and that King Wen 文王was a “western barbarian”
(xiyi zhi ren 西夷之人) would have been extremely shocking. As argued by Pines
(2005:73-4) there is indeed reason to believe that Mengzi intended to shock his audience
in order to persuade them to adopt his belief in an innate potential for goodness (ren 仁)
shared by all humans regardless of their origins. In other words, by arguing that the
famous sage kings started out as uncivilized barbarians but eventually through self-
58
Except perhaps for Zhuangzi, and perhaps also Mozi, who may have been of barbarian origin himself.
31
cultivation transformed themselves into paragons of virtue and founding fathers of the
much revered Zhou culture, Mengzi argues that everybody is born with the potential to
change his identity and become civilized or ‘Zhou’ through adoption of Zhou ritual.
While these aspects of the passage in (21) have been out by Pines (2005:73-4), we still
need to ask the question why Mengzi chose the expression xi-yi 西夷to refer to King
Wen as a ‘western barbarian’ rather than the quasi-synonymous, xi-rong 西戎 or 戎 rong,
which were also available in the vocabulary of identity of his time. Based on the above
analysis of the emergence and evolution of the paradigm of directional compounds we
are now in a position to answer this question.
Due to specific cultural and ethnic associations still attached to the term rong,
Mengzi preferred to use yi as a more neutral term for ‘alien other’ which was devoid of
specific cultural and ethnic connotations. As shown by the Zuo passage in (22), to the
literate Zhou elite in the Warring States period, Rong identity was defined by a number of
distinctive features, i.e. specific non-Zhou food, drink, clothing and language.
(22) 我諸 戎 飲食 衣服不與 華 同,贄幣不 通,言 語不達, 何惡之能為 ﹖不與於會 ,亦無
瞢焉。」
The food, drink, and clothing used by our various Rong tribes are not the same as those
of the Hua; gifts do not pass back and forth; and language is not understood. (Zuo, Xiang
14.1, Schaberg, 2001:133)
Since the main point of (21) is to emphasize the common human potential for change
through self-cultivation, a negatively defined neutral term like yi is more appropriate than
rong or xi-rong which would have invoked a number of distinctive cultural traits that are
both unnecessary and distracting in this context. Furthermore, by choosing a neutral term
devoid of reference to a specific existing non-Chinese tribe, Mengzi also avoids
32
undesirable political implications. In Mengzi’s lifetime, Rong tribes still existed as a
distinct non-Zhou people at the fringes of the Central Plain. A claim that King Wen
derived from the Rong might give this group a certain prestige that could be used
politically. Once again, by replacing xi-rong (or rong) with xi-yi Mengzi elegantly avoids
all these complications and leads the reader directly to the main point of the passage, i.e.
the cultural/inclusive
59
claim that anyone, regardless of his origins, has the potential to
transcend the Zhou-barbarian dichotomy through self-cultivation.
The Mengzi passages in (23a-b) provide another example of a specific directional
compound, i.e. bei-di 北狄 in (23a), being replaced by a more neutral yi-based compound,
i.e. bei-yi 北夷 in (23b).
(23) a. 東面而征 , 西夷 怨;南面而征,北狄 怨。曰: 奚為後我 ﹖ 』
“When he [=Tang, the founder of the Yin dynasty] marched on the east, the
western barbarians complained, and when he marched on the south, the
northern barbarians (bei-di) complained. They all said, “Why does he not
come to us first?” ” (Mengzi 1B:11, 3B:5, Lau 1970:69-70, 109-111)
b. 南面而征 , 北夷 怨,東面而征,西夷 怨。曰: 『奚為後 我 ﹖』
“[…] when he marched on the south, the northern barbarians (bei-yi)
complained, and when he marched on the east the western barbarians
complained. They all said, “Why does he not come to us first?’ […]”(Mengzi
7B:4, Lau 1970:194-5)
Except for the sequential ordering of clauses, the only difference between (23a) and (23b),
which according to D.C. Lau may derive form a lost passage of the Shangshu, is the
difference between bei-di 北狄 and bei-yi 北夷.
60
As in (21), where Mengzi preferred xi-
59
See Pines (2005:63) for discussion of the terms exclusiveness vs. inclusiveness and nationalism vs. culturalism.
60
“It is possible that this quotation is from the T’ang cheng (Punitive Expeditions of T’ang), one of the lost chapters of
the Book of History” (Lau, 1979/2003:47, note 17). See also Shangshu: 東征 西夷 怨,南征 北狄 怨;曰:『 奚獨後
予。』攸徂之民,室家相慶,曰:『徯予后,后來其蘇。』 ( 商書, 仲 虺之誥) “When the chief of Ge showed his
enmity to the provision-carriers, the work of punishment began with Ge. When it went on in the east, western
barbarians (xi-yi) murmured; when it went on in the south, northern barbarians (bei-di) murmured. They said, ‘Why
does he make us alone the last? […]” (Shangshu, Zhonghui zhi gao, Legge 1960: 180-1)
33
yi over xi-rong, the more general yi-based term (bei-yi) is best suited to express the way
his Confucian universalism constructs identity. Since for Mengzi Zhou identity depends
on whether one has internalized and follows the Zhou rituals (li 禮) in one’s behavior,
specific ethnonyms grounded in cultural or ethnic differences, e.g. xi-rong, and bei-di,
became less appropriate than neutral yi-based terms such as xi-yi and bei-yi in (21) and
(23). Why did Mengzi then not change all the instances of bei-di to bei-yi? I venture the
speculation that this may be because he is respecting the exact wording of the earlier text
he is quoting in Mengzi 1B:11, 3B:5 in (23a), while unconsciously rephrasing it using his
own vocabulary of identity in Mengzi 7B:4 in (23b). In any case, the fact that bei-yi and
bei-di occurs in identical contexts shows that they were quasi-synonymous in Mengzi’s
own usage. Notice also that he uses xi-yi, rather than xi-rong in both (23a) and (23b).
This bring the total number of semantically neutral yi-based directional compounds up to
6
61
which is twice as many as the number of semantically specific directional compounds
(i.e. bei-di 北狄 (1B:11, 3B:5) and nan-man 南蠻 (M 3A:4)). While this data-set is too
small to make any definitive conclusions, it does suggest that neutral yi-based compounds
were gaining ground at the expense of more specific compounds, and supports the
assumption of the third stage in the evolution of ethnonyms in (17iii) above. The fact that
Mengzi also uses the expression 四夷 si-yi ‘the barbarians of the four directions,’ further
strengthens this hypothesis, see (24).
61
The compound dong-yi is ambiguous between a ‘matching’ directional compound based on yi as a semantically
specific ethnonym ‘the Yi, who live in the east’, and a semantically neutral compound based on yi as a default term for
‘alien other’. As shown in the discussion of Mengzi 4B1 in (21), were Shun and king Wen are referred to as dong-yi
and xi-yi respectively, there is reason to believe that this instance of dong-yi is indeed a neutral yi-based directional
compound.
34
(24) 欲辟土地 , 朝秦、楚 , 蒞中國 而撫 四夷 也。
You wish to extend your territory, to enjoy the homage of Qin and Chu, to rule over the
Central Kingdoms (Zhongguo) and to bring peace to the barbarian tribes on the four
borders (si-yi). (Mengzi IA:7, Lau 1970:57-8)
In (24) the ‘barbarians of the four directions’ (si-yi) are contrasted with the Central States
(zhongguo) and is thus clearly used as a negatively defined term for non-Zhou
‘barbarians.’
Not only did neutral yi-based compounds gain ground, yi also stated being used
on its own to refer to ‘cultural others,’ see the Mengzi passage in (25) where the
‘barbarians’ (yi-zhe 夷者) are contrasted with the ‘Chinese’ (Xia 夏).
(25) 吾聞用夏 變 夷者 ,未聞變於夷者 也。[。。。] 魯頌曰: 『 戎狄 是膺,荊舒 是懲。』
周公方且 膺 之,子是 之 學,亦為 不 善變矣。
‘I have heard of the Chinese converting barbarians (yi) to their ways, but not of
their being converted to barbarian (yi) ways. […] The Lu song says,
It was the Rong and Di that he attacked;
It was Jing and Shu that he punished. (Ode 300)
It is these people that the Duke of Zhou was going to punish and you want to
learn from. That is not a change for the better, is it?’ (Mengzi 3:A:4, Lau
1970:103-4)
The Mengzi passage in (25) thus provides evidence of a fourth Stage in the evolution of
the vocabulary of identity, i.e. the emergence of the use of yi on its own to refer to
‘cultural others’ or ‘uncivilized barbarians,’ see (26)
62
.
(26) (i) Stage 2 (ii) Stage 3 (iii) Stage 4
a. xi-rong ( ≈ rong) => xi-yi ( ≈ xi-rong ≈ rong) => yi(-zhe)
b. bei-di (≈ di) => bei-yi ( ≈ bei-di ≈ di) => yi(-zhe)
c. etc.
Although Mengzi still uses directional compounds, preferring the neutral yi-based type,
xi-yi, and bei-yi in (21,23) and (26ii) over the more specific type, e.g. bei-di and nan-man
62
The term yi is also used alone as general term for barbarian twice in the Zuozhuan; once in Zuo, Zhuang 31.1, see
(13a); and once in Zuo, Ding 10.2: 裔不謀夏, 夷 不亂華 “Those distant people have nothing to do with the Xia; the
Yi must not be permitted to create disorder among the Hua States.” (Zuo, Ding 10.2, Legge:776-7) For discussion of
this passage in particular, and the Hua-Yi dichotomy in general see, Ogura Yoshihiko (1967).
35
in (19e,f) and (26i), his Confucian universalism ultimately makes a fine-grained
geographical distinction between ‘barbarians’ from different cardinal directions irrelevant
and leads him to use yi as a general term for ‘cultural other’, as shown in (25) and (26iii).
In conclusion, the combinatorial potential of yi, illustrated in (19a-d) clearly
indicates that Mengzi used it as the default term for ‘barbarian’. Independent evidence for
this comes from the fact that he also uses the compound 四夷 si-yi to refer to all
barbarians, see (24), and even uses yi on its own to contrast ‘barbarian’ ways with the
customs of the ‘Xia-Chinese’ (xia 夏), see (25). In the Zhanguo period, yi had thus
become the default morpheme for ‘barbarian.’ But as shown in the next section, yi was
not the only general word for ‘barbarian’ or ‘alien other.’ Compounds such as manyi,
rongdi, manyi-rongdi, and eventually, yidi also functioned in this way. However, yi was
the only single morpheme ethnonym that could be used both alone and in compounds as
default term for ‘cultural other.’
36
Chapter 5
Ethnonym compounds
In a morphological paradigm of bi-morphemic ethnonym compounds composed of the
four single morpheme ethnonyms terms man 蠻, yi 夷, rong 戎, and di 狄 there are 16
possible combinations. However, as shown in the table in (27), only six are attested in the
Gugong corpus.
Table (27) Ethnonym Compounds in Gugong corpus
(i) Compound (ii) Occurrences in corpus
a. yi-di 夷狄 41
63
b. man-yi 蠻夷 20
64
c. rong-di 戎狄 21
65
d. rong-yi 戎夷 2
66
e. di-rong 狄戎 1
67
f. rong-man
戎蠻 6
68
63
Most of these 41 tokens are from the Guliang (17 tokens) and Gongyang (20 tokens). Yidi is also found once in the
Liji and the Mengzi, and twice in the Lunyu.
64
These 20 tokens are distributed as follows: Shangshu (2), Zhouli (1), Zuo zhuan (9), Xunzi (1), Mozi (2), Yanzi
Chunqiu (1), Guanzi (1), Lüshi Chunqiu (2). The fact manyi is found in such a wide range of different texts indicates
that it was a productive part of the Zhanguo vocabulary of identity.
65
These 21 tokens are distributed as follows: Maoshi (1), Zhouli (1), Liji (1), Zuo zhuan (12), Gongyang (1), Mengzi
(2), Xunzi (1), Yanzi Chunqiu (1), Guanzi (1). The fact rongdi is found in such a wide range of different texts indicates
that it was a productive part of the Zhanguo vocabulary of identity.
66
The instance of rong-yi in the Liji passage in (10) is probably phrasal unit rather than a compound.
(i) 亦奚羨於彼 而棄齊國之社稷,從戎夷 之 國乎﹖“Why should you long for somewhere else, abandoning the altars
of Qi to follow a nation of barbarians (Rongyi)?” (Graham, 1960/1990:103)
(ii) 秦國僻陋 戎夷 “[…] Qin is a backward place, where Rong and Yi barbarians live.”(Lüshichunqiu, Knoblock and
Riegel 2000:607). Three instances of 戎夷 are emended as 式夷 Shi Yi (proper name) (Lüshichunqiu, Knoblock and
Riegel 2000:517; see also pp. 514-5).
67
司馬 起豐 、析 與 狄戎 ,以臨 上雒。( 左傳, 哀公) “The marshal raised the people of Feng and Xi, along with
[certain tribes of] the Di and Rong, and proceeded towards Shang-lou.”(Zuo, Duke Ai 4.2, Legge, p. 805).
68
Two tokens are from the Chunqiu, Zuo [Zhao 16.2] 戎蠻子 “the viscount of the Man-rong,” see Legge, p. 663. And
another two from the Chunqiu 16.1, 4.6 (in the Guliang zhuan) 戎蠻子 “viscount of the Man-rong.” The Gongyang has
曼 instead of 蠻. One token is from the Zuozhuan commenting on one of the Chunqiu passages with rongmanzi. Finally,
one token is found in the Liezi: 觀吾之鄉,如戎蠻 之國 ‘look at my own neighborhood as though it were the country
of the Man-rong’. Rongmanzi refers to a particular individual who was the ‘viscount’ (zi 子) of a particular group of
non-Zhou people. Hence it cannot refer to ‘alien others’ in general. That is, rongmanzi does not refer to ‘the viscount of
the alien others.’
37
With expressions composed of two ethnonyms, which each have the potential to refer to a
specific non-Zhou group, it is not always clear whether they should be classified as true
compounds or as phrasal units of conjoined mono-morphemic ethnonyms. Since
Classical Chinese grammar allows both constructions, only context can tell whether
sequences such as man-yi 蠻夷 and rong-di 戎狄 should be parsed as ‘the Manyi’ and
‘the Rongdi,’ or as ‘the Man and the Yi,’ and the ‘the Rong and the Di.’ I argue that only
the expression in (27a-c) are frequently used as true compounds which have the ability to
function as general terms for ‘cultural others’ or ‘barbarians’ and that productive use of
these compounds as general terms for ‘barbarians’ started in the late Chunqiu or early
Zhanguo late period. Detailed distributional data for forms and usage patterns of
ethnonym compounds in the Chunqiu and Zhanguo periods is listed in (28).
Table (28) Ethnonym Compounds
(i)
C
(ii)
S
(iii)
CS
(iv)
M
(v)
X
(vi)
MX
(vii)
L
(viii)
Z
(ix)
MXLZ
a. man-yi 0 0 0 0 2 2 2 9 13
b. rong-di 0 (1) (1) 2 1 4 0 12(11?) 16
c. manyi-rongdi
0 0 0 0 1 1 0 1 2
d. yidi 0 0 0 1 0 1 0 1 2
Total ethnonym + ethnonym 0 0 (1) 3 4 7 2 22 31
In the Chunqiu period texts only one potential ethnonym compound is found, i.e. rongdi,
see (29).
(29) 戎狄 是膺 The Rong and the Di he withstood.
荊舒 是懲 The Jing and the Shu he repressed’ (Ode 300, see Karlgren 1950:260)
While at first glance rong di in (29) may look like a compound, the parallelism with the
conjoined ethnonyms ‘the Jing and the Shu’ in the following line, indicates that it is also
a phrasal unit of conjoined mono-morphemic ethnonyms referring to two separate non-
38
Zhou peoples.
69
70
The data set in (28), with zero true compounds in the Chunqiu and the
Shijing (28iii) vs. thirty-one (= 35% see (8dix)) in the Mengzi, Xunzi, Zuo zhuan and
Lüshi Chunqiu combined (28ix), thus strongly indicate that true ethnonym compounds
first emerged in the Zhanguo period.
The Zuo zhuan passages in (30-31) illustrate the use of manyi and rongdi as
general terms for ‘barbarians’ which are contrasted either with the Central States
(zhongguo), as in (30), or with ‘all the Xia-Chinese (states)’ (zhu xia), as in (31).
(30) 吳伐郯, 郯 成。季文 子 曰:「中國 不振旅,蠻夷 入伐,而 莫之或恤 。 無弔者也 夫 !
“Wu invaded Tan, and Tan submitted to the terms of peace […]. Ji Wenzi said, “The
Central States (Zhongguo) do not array their multitudes, and the wild tribes of the
south and east (Man-yi) enter and attack them, while there is none to pity the sufferers.
[Tan] has no comforter.” (Zuo, Cheng 7.1, Legge, p. 363)
(31) 狄人 伐邢。管敬仲言 於 齊侯曰: 「 戎狄 豺狼,不可厭也 ; 諸夏 親暱,不可棄也 。
“The Di had invaded Xing. Guan Jing-zhong said to the marquis of Qi, “The Rong and
Di are wolves, to whom no indulgence should be given; within the States of the Great
land (zhu-Xia), all are nearly related, and none should be abandoned. ”(Zuo, Min 1.2,
Legge, p. 124)
While contrast with the Central States (zhongguo) and ‘all Xia-Chinese’ (zhu xia) in (30-
31) indicate that manyi and rongdi were general terms for ‘barbarians,’ these passages
also contain specific geographical information that show a semantic difference between
manyi and rongdi. In (30) the term manyi is used to refer to the state Wu, which was
69
According to Waley:315, note 3-4, “The Rong tribes raided the Zhou capital in 649; the Di attacked central China at
a number of different points toward the middle of the seventh century B.C. [..] Jing are the southern people known later
as Chu. The Xu (in south-west Shandong and Anhui) were regarded as non-Chinese, but at this period often fought in
alliance with the Chinese. We know from the Zuo zhuan that Duke Xi took part in an expedition against Chu in 656
B.C.”
70
Another case where context clearly tells us that a specific sequence of two ethnonyms is a phrasal unit of conjoined
single morpheme ethnonyms rather than a true compound is found in the following Liji passage: 昧,東夷 之樂 也;<
任>,南蠻 之 樂也。納夷蠻之樂於大廟 “[…] There (were also) the Mei, or music of the Yi of the East; and Ren, or
music of the Man of the South. The introduction of the Yi-man music in the grand temple was to signalize the
distinction of Lu all over the kingdom.” (Liji, adapted from Legge, p. 33) That is, yi-man can be compared to Modern
Chinese nonce compounds such as jie-di ‘older sister(s) and younger brother(s),’ and xiong-mei ‘older brother(s) and
younger sister(s),’ etc. which occur in specific contexts where the speaker is referring to his or her particular set of
siblings.
39
located south-east of the Central Plain area.
71
In (31) the term rongdi is used to refer to
the di tribes who attacked Xing in the state of Qi
72
, which was located in the far north-
eastern corner of the Zhou realm (see the map of Major Spring and Autumn period states
in Appendix I)
73
. In other words, these passages indicate that manyi was the general term
for ‘barbarians’ coming from the south or from the east, while rongdi was the general
term for ‘barbarians’ from the north or west. In this respect, manyi and rongdi are similar
to the Modern Chinese terms xiong-di “brothers” and jie-mei “sisters.” Xiong-di and jie-
mei can be considered general terms for ‘siblings’ in that they refer to sub-sets of siblings,
see (32e-f).
(32) Words for ‘siblings’ in Modern Chinese:
a. xiong 兄 ‘older brother’
b. di 弟 ‘younger brother’
c. jie 姐 ‘older sister’
d. mei 妹 ‘younger sister’
e. xiongdi 兄弟 ‘brothers’
f. jiemei 姐妹 ‘sisters’
g. xiongdi-jiemei 兄弟姐妹 ‘brothers and sisters, siblings’
Similarly, man-yi and rong-di are general terms for ‘barbarians’ which refer to different
sub-sets of non-Zhou peoples see (33e-f).
71
Archaeological evidence shows that the material culture of the states of Wu and Yue differed from the other Zhou
states, see L. von Falkenhausen (1999:525-39).
72
For discussion of another Zuo passage relating an attack by the Di, see Zhang Huaitong (2001:21-27).
73
For further discussion the geographical location of the Rong-di, see Zhao Tiehan (1965). For Rong and Zhou
interaction in the Chunqiu period, see also Kod ō Kinpei (1962).
40
(33) Words for ‘barbarians’ in Classical Chinese:
a. yi 夷 ‘the Yi barbarians (who live in the east)’
b. rong 戎 ‘the Rong barbarians (who live in the west)’
c. di 狄 ‘the Di barbarians (who live in the north)’
d. man 蠻 ‘the Man barbarians (who live in the south)’
e. man-yi 蠻夷 ‘the barbarians to the south and the east’
f. rong-di 戎狄 ‘the barbarians to the west and the north’
g. manyi-rongdi 蠻夷戎狄 ‘the barbarians to the south, east, west and north’
As indicated in (32g) and (33g), the parallel between classical Chinese ethnonyms
compounds used as general terms for ‘barbarians’ and the different modern Chinese
words for siblings can be pushed even further. Modern Chinese lacks semantically
general mono-morphemic words for ‘brother(s)’ and ‘sister(s)’ and uses the compounds
xiong-di ‘(lit.) older brother-younger brother’ and jie-mei ‘(lit.) older sister-younger
sister’ instead. Modern Chinese also lacks a mono-morphemic word for ‘sibling(s)’ and
instead uses a combination of the compounds for ‘brothers’ and ‘sisters,’ i.e. xiongdi-
jiemei ‘(lit.) older brother-younger brother-older sister-younger sister’ which is
semantically equivalents to the English word ‘siblings’, see (32g). In other words, the
Modern Chinese word for ‘siblings’ has the internal structure in (34).
(34) /\
/ \
/ \ / \
xiong-di jie –mei 兄弟姐妹 ‘siblings’
As the Xunzi passage in (35) shows, Warring States Chinese used a similar four-
morpheme compound, composed of the ethnonym compounds manyi and rongdi, as a
general term for ‘barbarians,’ i.e. manyi-rongdi 蠻夷戎狄.
41
(35) 故 諸夏之國 同 服同儀 , 蠻夷戎狄 之 國 同服不 同制 。封內 甸服 ,封外 侯服 ,侯衛賓
服,蠻夷 要服,戎狄 荒服。
Accordingly, all the states of Xia Chinese have identical obligations for service to the
king and have identical standards of conduct. The countries of the Man, Yi, Rong and Di
barbarians perform the same obligatory services to the king, but the regulations
governing them are not the same. Those who are enfeoffed within [the royal domain] do
royal service. Those who are enfeoffed without [the royal domain] do feudal service.
Those who are in the feudal marches zone do guest service. The Man and Yi nations do
service according to treaty obligations. The Rong and Di do irregular service. (Xunzi 18,
Zhenglun 4, Knoblock 1994:38-39)
74
In (36) the ‘barbarians’ (manyi-rongdi 蠻夷戎狄) are contrasted with the ‘Xia-Chinese’
(zhuxia 諸夏). Furthermore, in the immediately following context, the manyi barbarians
are distinguishes from the rongdi barbarians with respect to their service to the Zhou
court. This indicates that the internal structure of manyi-rongdi as a general term for
‘barbarians’ is similar to that to xiongdi-jiemei as a general term for ‘siblings’ in modern
Chinese, compare (34) and (36a).
(36) /\
/ \
/ \ / \
a. man -yi rong –di 蠻夷戎狄 ‘barbarians’
b. … -yi … -di 夷狄 ‘barbarians’
74
Evidence that manyi-rongdi was used a ‘compound’ can also be found in the following Zuo passage: 蠻夷戎狄 ,不
式王命,淫湎毀常,王命伐之 “When any of the wild tribes, south, east, west, or north (man-yi-rong-di), do not obey
the king’s commands, and by their dissoluteness and drunkenness are violating all the duties of society, the king gives
command to attack them” (Zuo, Cheng 3, Legge 1960:349). Note however, the existence of potential counter-examples,
e.g. the passage from the “Wang Zhi” chapter of the Liji in (10) above, where these four term are found in a different
order: 中國、 夷、蠻、 戎、狄,皆有安居 “The people of the Central states, and those of Yi, Man, Rong, and Di,
all had their dwellings, where they lived at ease […]” (Liji Wangzhi 3.13, Legge, pp.229-230). The fact that we do not
find manyi-rongdi here could potentially be used to argue that manyi-rongdi is a phrasal unit rather than a compound
on a par with modern Chinese xiongdi-jiemei ‘siblings’. Note, however, that this unexpected order is due to the fact that
it mirrors the order in which the ‘barbarians’ of the four directions are discussed in the preceding passage: 東方曰 夷,
[…] 。南方曰蠻 ,[…] 。西方曰 戎,[…] 。北方 曰狄,[…] 。”The tribes on the east were called Yi. […] Those on
the south were called Man. […] Those on the west were called Rong. […] Those on the north were called Di. […]”.
(Liji Wangzhi 3.13, Legge, pp.229-230). In other words, in this context the sequence yi-man-rong-di 夷蠻戎狄 is not a
compound but rather a phrasal sequence of mono-morphemic ethnonyms. See also the discussion of the sequence yi-
man in footnote 67 above.
42
As shown in (36b), I propose to analyze the compound yi-di as an abbreviated form of
manyi-rongdi.
75
Interestingly, of the forty-one occurrences of yi-di in the Gugong corpus,
37/41=90% are found in the Guliang (17 tokens) and Gongyang (20 tokens), where it is
used as a general term for ‘cultural others’ which is often contrasted with the Central
States (Zhongguo), see (37).
(37) 楚有王者 則 後服,無 王 者則先叛 。 夷狄 也,而亟病中國 , 南夷 與北狄交。中國 不
絕若線, 桓 公救中國 ,而攘夷狄 , 卒怗荊, 以 此為王者 之 事也。
When Chu has a [true] King then they will submit [to the Son of Heaven]; when they do
not have a [true] King, then they will be the first to revolt. They are Yidi, they often
violate the Central States. The Southern Yi and the Northern Di join together [to attack
the Central States]. And the Central States are imperiled like a string that is about to
break. Duke Huan saved the Central States, and resisted the Yidi, and finally pacified
the Jing Chu. And thereby he completed the task of a [true] King. (Gongyang, Xi 4.3)
76
In contrast, in the Zuo zhuan, which has the highest number of the compounds manyi (9
occurrences) and rongdi (12 occurrences), not a single instance of yi-di is found. This
complementary in the distribution of yi-di, man-yi, and rongdi, in the Zuo vs. the
Gongyang and Guliang is all the more surprising since these three works are
commentaries on the Chunqiu, and may thus be an indication that they were composed at
different times. While manyi and rongdi are both attested in a wide variety of Zhanguo
texts
77
, the bulk of all occurrences of yidi are found in the Guliang and the Gongyang,
with only four tokens found elsewhere, i.e. two in the Analects
78
and one in each of the
75
See Di Cosmo (2002:100) for discussion of the yidi as general term for ‘barbarians.’
76
Notice that in this Gongyang passage, the State of Chu is branded as a ‘barbarian’ (yidi) state. See also Di Cosmo
(2002) for further discussion of the use of terms for ‘barbarians’ in inter-state political discourse. For Zhou perceptions
of the State of Chu, see also Pines (2005:88).
77
Manyi: Shangshu (2 tokens), Zhouli (1), Zuozhuan (9), Xunzi (1), Mozi (2), Yanzi Chunqiu (1), Guanzi (1), Lüshi
Chunqiu (2). Rongdi: Shijing (1 token), Zhouli (1), Liji (1), Zuozhuan (10), Gongyang (1), Mengzi (1), Xunzi (1),
Yanzi Chunqiu (1), Guanzi (1).
78
(i) 子曰:「 夷狄 之有君,不 如諸夏 之亡也 。」 “The Master said, “The Yi and Di barbarians (yi-di), even with
their rulers, are still inferior to the Chinese states (zhu-Xia) without their rulers.” (Analects 3.5, Slingerland, 2003:18);
(ii) 樊遲問仁。 子曰:「居處恭,執事敬,與人忠;雖之夷狄 ,不可也 。」 “Fan Chi asked about Goodness. The
Master replied, “When occupying your place, remain reverent; when performing public duties, be respectful; and when
43
Mengzi
79
and the Liji
80
. Assuming the Gongyang and Guliang to have been written later
than the Zuo zhuan,
81
this seems to indicate that yidi became widely used as a general
term for ‘alien other’ at a later time than manyi and rongdi.
dealing with others, be dutiful. These are the virtues that cannot be abandoned, even if you go to dwell among the Yi or
Di barbarian.””(Analects 13.19, Slingerland, 2003:148)
79
周公兼 夷狄 、 驅猛獸而百姓寧,孔子成《春秋》而亂臣賊子懼。《詩》云:『戎狄是膺,荊舒是懲;則莫我
敢承。』‘[…] the Duke of Chou subjugated the northern and southern barbarians, drove away the wild animals
and brought security to the hundred surnames; Confucius completed the Spring and Autumn Annals, and brought
terror into the hearts of rebellious subjects and undutiful sons. The Book of Odes says, It was the Rong and Di that he
attacked;It was the Ching and Shu that he punished […]’ (Mengzi 3B:9, Lau 1970:115)
80
素 夷狄 ,行 乎 夷狄 “Situated among barbarous tribes, he does what is proper in such a situation. […](Liji,
“Zhongyong”, Legge:306)
81
See Cheng (1993).
44
Chapter 6
Conclusion
In pre-Zhanguo times the Zhou had frequent interaction with a number of non-Zhou
peoples, often living inside or in close geographical proximity to the Central States, who
were conceptualized as ‘belligerent others’ and were referred to by toponym compounds
denoting specific tribes living in specific geographic locations, e.g. Quan-rong ‘the Dog
Rong,’ Huai-yi “the Yi of the Huai region,’ etc. Due to its semantic field related to
warfare, the term rong could also be used, either alone or in compounds, to refer to
‘belligerent others’ in general. Through military conquest and acculturation many non-
Zhou groups gradually vanished and, as a consequence, the concept of foreigners as
‘belligerent others’ became less relevant. When the concept of foreigners as ‘cultural
others’ emerged in the Zhanguo period with the advent of Confucian universalism, the
need for a corresponding linguistic expression was felt. Rather than creating a new
morpheme
82
(as the Greek onomatopoeia βάρβαροι ‘barbarians’
83
) two different
strategies were adopted:
(i) using the single morpheme ehnonym yi, which had hitherto been used
to refer to specific tribes (e.g. Huai-yi, Kun-yi, etc.), as a general term
for ‘barbarian’ or ‘alien other’, both in compounds (si-yi, xi-yi, bei-yi,
82
Or using the term rong ‘belligerent other,’ which was semantically unfit because of its association with warfare. In
addition to it is use a general term for ‘cultural other’ (in Zhanguo period Chinese) and other ethnonym uses, the term
yi 夷 could also be uses as nouns, verbs, and adjectives with meanings ranging from ‘level (both concrete and abstract
senses),’ ‘harmonize,’ ‘equal’, ‘common,’ etc.
83
According to Hall (1989) the word βάρβαρος is a reduplicative onomatopoeia which “originally was simply an
adjective representing the sound of incomprehensible speech” (p. 4). It was not until after the Persian wars that it came
to be used in as “the universal anti-Greek against whom Hellenic –especially Athenian – culture was defined” (p.5). For
further discussion of the etymology of this word and its possible origin as a loan word, see Weidner (1913:303-4) and
Porkorny (1959:91-2).
45
nan-yi, dong-yi, etc.) and when used on its own, e.g. yi-zhe in the
Mengzi passage in (25) and (26iii)
(ii) using a multiple-morpheme compound in which the combined
meanings of the individual etnonym components add up to the totality
of all ‘barbarian’ peoples (e.g. manyi-rongdi or the corresponding
abbreviated form yidi).
Interestingly, both these strategies were products of the cardinal direction system
illustrated in (9). That is, the creation of the redundant directional compounds (xi-rong,
bei-di, nan-man, etc.) laid the ground for the creation of the less redundant, and less
specific yi-based directional compounds (xi-yi, bei-yi, nan-yi, etc.). The use of yi as a
negatively defined default term for ‘barbarian’ in compounds, led to it becoming devoid
of specific ethnic meaning components and thus enabled it to function as a general term
for ‘barbarians’ on its own. Similarly, it was because of the association of man, yi, rong,
and di with the four cardinal directions, that the sum of the meanings of these four terms
added up to the totality of all non-Zhou peoples. Otherwise a very long combination of all
existing ethnonyms (xian-yun, rong, man, yi, di, mo, jing, shu, etc.) would have been
needed. Needless to say, such an unwieldy compound would never have been viable in
actual speech.
The results of this study strongly suggest that the specific changes in the form and
use of individual terms for ‘barbarians’ cannot be understood as isolated phenomena, but
rather should be viewed as the combined result of many factors conspiring to trigger
paradigmatic changes in the larger formal systems of linguistic expressions of
46
‘otherness.’ Since our source of knowledge about notions of identity in pre-imperial
China comes mainly from textual evidence, the importance of empirical linguistic studies
of this type cannot be overstated. Indeed, as the analysis of the vocabulary of ‘otherness’
in the Mengzi shows, a thorough understanding of the structure and evolution of formal
paradigms of ethnonyms vastly enhances our ability to interpret individual ethnonyms in
any given text since it enables us to predict expected patterns and thus explain deviances
from standard usage patterns in alternative ways. Only by simultaneously studying the
structure and evolution of the formal expressions that make up the vocabulary of
otherness, and the political, philosophical, or historiographic narratives that inform them
and give them meaning, can we hope to advance our understanding of how attitudes
toward foreigners were conceptualized and expressed in pre-imperial China.
47
Appendix
84
Figure 1: Major States of the Spring and Autumn Period
84
The Spring and Autumn and Warring States maps from The Cambridge History of Ancient China, eds. Michael
Loewe and Edward L. Shaughnessy (Cambridge: Cambridge University Pres, 1999), p. 548, 594.
48
Figure 2: Major States of the Warring States Period.
49
Bibliography
Allen, Joseph Roe III The Book of Songs. Translated by Arthur Waley. Edited with
Additional Translations by Joseph R. Allen. Foreword by Stephen Owen. Postface by
Joseph R. Allen. New York: Grove Press. 1996
Brindley, Erica. “Barbarians or not? Ethnicity and Changing Conceptions of the Ancient
Yue (Viet) Peoples (~400-50 B. C.),” Asia Major 16.1 (2005).
Brown, Miranda. “Men in Mourning: Ritual, Politics, and Human Nature in Warring
States and Han China, 453 BC-AD 220.” Ph.D. dissertation. University of California,
Berkeley, 2002.
Chang, K. C. Shang Civilization. New Haven, Conn.: Yale University Press, 1980.
Chen Suizheng 陳穗錚. “Zhongguo cicheng de qiyuan yu yuanyi 中國詞稱的起源與原
義.” Shi Yuan 史原 19(1993):1-38.
Cheng, Anne. “Ch’un ch’iu 春秋, Kung yang 公羊, Ku Liang 穀粱 and Tso chuan 左
傳.” In Michael Loewe (ed.), Early Chinese Texts. A Bibliographical Guide. Berkeley:
The Institute of East Asian Studies, 1993
Cho-yun Hsu. “The Spring and Autumn Period,” in The Cambridge History of Ancient
China, eds. Michael Loewe and Edward L. Shaughnessy. Cambridge: Cambridge
University Pres, 1999.
Crossley, Pamela K. A Translucent Mirror: History and identity in Qing Imperial
Ideology. Berkeley: University of California Press, 1999.
Di Cosmo, Nicola. “The Northern Frontier in Pre-imperial China.” In The Cambridge
History of Ancient China, eds. Michael Loewe and Edward L. Shaughnessy. Cambridge:
Cambridge University Pres, 1999.
Di Cosmo, Nicola. Ancient China and Its Enemies. The Rise of Nomadic Power in East
Asian History. Cambridge: Cambridge University Press, 2002.
Dikötter, Frank. The Discourse of Race in Modern China. Stanford, California: Stanford
University Press, 1992.
Dikötter, Frank. The Construction of Racial Identities in China and Japan. Honolulu:
University of Hawaii Press, 1997.
Elliott, Mark C. The Manchu Way. The Eight Banners of Ethnic Identity in Late Imperial
China. Stanford: Stanford University Press, 2001.
50
Falkenhausen, Lothar von. “The Waning of the Bronze Age: Material Culture and Social
Developments, 770-481 BC.” In The Cambridge History of Ancient China, eds. Michael
Loewe and Edward L. Shaughnessy, Cambridge: Cambridge University Pres, 1999, pp.
450-544.
Fiskesjö, Magnus. “On the 'raw' and 'cooked' barbarians of imperial China,” Inner Asia 1,
no.2 (1999):139-168.
Gentz, Joachim. Das Gongyang zhuan: Auslegung und Kanoniesierung der Frühlings
und Herbsannalen (Chunqiu). Wiesbaden, 2001.
Hall, Edith. Inventing the Barbarian: Greek Self-definition through Tragedy. Oxford:
Oxford University Press, 1989.
Hall, Jonathan M. Ethnic Identity in Greek Antiquity. Cambridge: Cambridge University
Press, 1997.
Harrison, Thomas, ed. Greeks and Barbarians. Edinburgh: Edinburgh University Press,
2002.
Holcombe, Charles. “Re-Imagining China: The Chinese Identity Crisis at the Start of the
Southern Dynasties Period.” JAOS, 115 (1995), pp. 1-14.
Hsu Cho-yun and K.M. Linduff. Western Zhou Civilization. New Haven: Yale University
Press, 1988.
Huber, P. “The Relationship of the Painted Pottery and Lung-shan Cultures.” In The
Origins of Chinese Civilizations, ed. David Keightley. Berkeley: University of California
Press 1983.
Karlgren, Bernhard. The Book of Odes: Chinese Text, Transcription and Translation;
Stockholm: Museum of Far Eastern Antiquities, 1950 (Published originally in BMFEA 16
(1944) and 17 (1945)).
Keightley, David N. (ed.) The Origins of Chinese Civilization, Berkeley: University of
California Press, 1983.
Keightley, David N. “The Late Shang State: When, Where, and What?” in The Origins of
Chinese Civilization, ed. David N. Keightley. Berkeley: University of California Press,
1983.
51
Keightley, David, N. “The Shang: China’s First Historical Dynasty. In The Cambridge
History of Ancient China, eds. Michael Loewe and Edward L. Shaughnessy (Cambridge:
Cambridge University Pres, 1999, pp. 232-291
Keightley, David, N. The ancestral landscape: time, space, and community in late Shang
China, ca. 1200-1045 B.C. Berkeley: University of California, Berkeley, Center for
Chinese Studies, 2000.
Knoblock, John, tr. Xunzi: A Translation and Study of the Complete Works, Vol. 1-3,
Stanford University Press, 1988, 1990, 1994.
Kod ō Kinpei 後藤均平. “Shuj ū jidai no Sh ū to Ron 春秋時代の周と戎.” In Ch ūgoku
kodai shi kenky ū, 中國古代史研究( 第二), ed. Ch ūgoku kodai shi kenky ūkai 中國古代史
研究會, pp. 71-102. Tokyo: Yoshikawa, 1962.
Kruikov, Mikhail, et al. Drevnie Kitajtsy: Problemy Etnogeneza. Moscow, 1978.
Lefeuvre, J.A. “An Oracle Bone in the Hong Kong Museum of History and the Shang
Standard of the Center,” Journal of the Hong Kong Archaeological Society 7, 1976-8.
Legge, James, trans. The Chinese Classics. 5 vols. Hong Kong: Hong Kong University
Press. 1960.
Legge, James. Li chi: book of rites. An encyclopedia of ancient ceremonial usages,
religious creeds, and social institutions. Translated by James Legge. Edited with an
introd. and study guide by Ch‘u Chai and Winberg Chai. New Hyde Park, N. Y.,
University Books, 1967.
Li Xueqin. Eastern Zhou and Qin Civilizations. Translated by K.C. Chang. New Haven
and London: Yale University Press, 1985.
Loewe, Michael. “The Heritage Left to Empires.” In The Cambridge History of Ancient
China, eds. Michael Loewe and Edward L. Shaughnessy. Cambridge: Cambridge
University Pres, 1999, pp. 967-1032.
Lynn, Richard John. The Classic of Change; a new translation of the I Ching as
interpreted by Wang Bi. New York, 1994.
Mattos, Gilbert L. “Eastern Zhou Bronze Inscriptions.” In New Sources of Early Chinese
History: An Introduction to Reading Inscriptions and Manuscripts, ed. Edward L.
Shaughnessy. Berkeley, 1997, pp. 85-124.
Nippel, Wilfried “The Construction of the ‘Other.’” in Greeks and Barbarians, ed.
Thomas Harrison, 278-310. Edingburgh: Edinburgh University Press, 2002.
52
Ochi Shigeaki 越智重明. "Ka-i shisô no seiritsu" 華夷思想の成立 (The Formation of
the "Sino-Barbarian" Ideology). Kurume Daigaku Hikaky Bunka Kenkyûjo kijô 久留米大
學比較文化研究所紀要, 11 (1992), pp. 43-137.
Ogura Yoshihiko 小倉芳彥. “I I no toriko: Saden no Ka I kannen 裔夷の俘—左傳の華
夷觀念.” In Chūgoku kodai shi kenky ū 2, 中國古代史研究( 第二), ed. Ch ūgoku kodai shi
kenky ūkai 中國古代史研究會, pp. 153-87. Tokyo: Yoshikawa, 1967.
Pines, Yuri. “Intellectual Change in the Chunqiu Period: The Reliability of the Speeches
in the Zuo Zhuan as Sources of Chunqiu Intellectual History.” Early China
22(1997):100-16.
Pines, Yuri. “Beasts or Humans: Pre-Imperial Origins of the “Sino-Barbarian”
Dichotomy.” In Amitai, Reuven and Michal Biran (eds.) Mongols, Turks, and Others.
Eurasian Nomads and the Sedentary World. Brill: Leiden, Boston, 2005, pp. 59-102.
Pines, Yuri. “Disputers of the Li: Breakthroughs in the Concept of Ritual in Pre-imperial
China.” Asia Major (Third Series), 13 (2000), pp. 1-41.
Pines, Yuri. “Changing Views of tianxia in Pre-imperial Discourse.” Oriens Extremus,
43.1-2 (2002), pp. 101-116.
Pokorny, J. Indogermanisches etymologisches Wörterbuch, I (Berne/Munich), 1959.
Poo Mu-chou 蒲慕州. “The Emergence of Cultural Consciousness in Ancient Egypt and
China: A Comparative Perspective.” In Essays in Egyptology in Honor of Hans Goedicke,
ed. Betsy M. Bryan and David Lorton, 191-200. San Antonio, Tex.: Van Siclen Books,
1994.
Poo Mu-chou 蒲慕州. “Encountering the Strangers: A Comparative Study of Cultural
Consciousness in Ancient Egypt, Mesopotamia, and China.” Proceedings of the Seventh
International Congress of Egyptologists. Louven: Peters, 1998.
Poo Mu-chou 蒲慕州. "Gudai Zhongguo, Aiji yu Lianghe liuyu dui yizu taidu zhi bijiao
yanjiu" 古代中國、埃及與兩河流域對異族態度之比較研究 (A Comparative Study of
the Attitudes toward the Aliens in Ancient China, Egypt, and Mesopotamia). Hanxue
yanjiu 漢學研究, 17 (1999), pp. 137-69.
Poo Mu-chou 蒲慕州. Enemies of Civilization. Attitudes toward Foreigners in Ancient
Mesopotamia, Egypt, and China. Albany: State University of New York Press, 2005.
53
Pr ůšek, Jaroslav. Chinese Statelets and the Northern Barbarians in the Period 1400-300
B.C. Dordrecht: Reidel, 1971.
Pulleyblank, E.G. “The Chinese and their Neighbors in Prehistoric and Early Historic
Times.” In The Origins of Chinese Civilizations, ed. David Keightley. Berkeley:
University of California Press, 1983, pp. 411-66.
Schaberg, David. A Patterned Past. Form and Thought in Early Chinese Historiography.
Cambridge and London: Harvard University Press, 2001.
Shaughnessy, Edward L. Sources of Western Zhou History. Berkeley: University of
California Press, 1991.
Slingerland, Edward. Confucius Analects. Indianapolis/Cambridge: Hackett Publishing
Company, 2003.
Tang Lan 唐蘭. “Siguo jie 四國解.” Yugong 禹貢 1, 10(1934):6-9.
Wang Aihe. Cosmology and political culture in early China. Cambridge: Cambridge
University Press, 2000.
Wang Ermin 王爾敏. “Zhongguo mingcheng ji qi jindai quanshi 中國名稱及其近代詮
譯.” In Wang Ermin, Zhongguo jindai sixiang shilun 中國近代思想史論. Taipei: Huashi
chubanshe, 1977.
Wang Guowei. “Gueifang Kunyi Xianyun kao 鬼方昆夷玁狁考.” In Guan tang ji lin 觀
堂集林: 20 juan / [zhu zuo zhe Wang Guowei zhuan]. Taibei: Yi wen yin shu guan,
Minguo 47 [1958], pp. 583-605.
Wang Zhongfu 王仲孚. “Shilun chunqiu shidai de zhuxia yishi 試論春秋時代的諸夏意
識.” Dierjie guoji Hanxue lunwenji 第二屆國際漢學會議論文集 vol. 1, pp. 363-76.
Taipei: Academia Sinica, 1989.
Watson, Burton. Mo Tzu: basic writings. New York: Columbia University Press, 1963.
Weidner, E. “Βάρβαρος,” Glotta, 4(1913):303-4.
Wilkinson, Endymion. Chinese History: A Manual. Cambridge, MA, 2000.
Yü Rongchun 于溶春. "Zhongguo yici de youlai yanbian jiqi yu minzu de guanxi 中國一
詞的由來演變及其與民族的關係." Neimenggu shehui kexue 內蒙古社會科學 2:75-80,
1986.
54
Zhang Huaitong 張懷通. “Chunqiu Zhuanggong sanshier nian faxing zhi ‘di’ kao 春秋莊
公三十二年伐邢之`狄'考.” Zhongyuan wenwu 中原文物 3(2001):21-27.
Zhao Tiehan 趙鐵寒 “Chunqiu shiqi de Rong-Di dili fenbu jiqi yuanliu 春秋時期的戎狄
地理分析及其源流.” In Gushi kaoshu 古史考述, pp. 314-47. Taipei: Zhengzhong shuju,
1965.
Zhou Fagao 周法高. Jinwen gulin 金文詁林. Hong Kong: Chinese University Press, 1975.
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
A commercial and optimistic worldview of the afterlife of the Song people: Based on stories from the "Yijian Zhi"
PDF
Ito Noe: Living in freedom. A critique of personal growth in Japanese society
PDF
Representations of the mother as origin and force of life in Japanese literature and history
PDF
China's periphery in perspective: A comparative look at the Yanbian Korean Autonomous Prefecture and Shenzhen Special Economic Zone
PDF
National character studies in America and Japan: Toward a new understanding of nihonjinron
PDF
China's telecommunications race: An attempt by the CCP to achieve technical legitimacy and hold onto power
PDF
If you build it...: A "different story" of the re-emergence of baseball in China, the people who play it, and why
PDF
Japan's modernization and troubled identity: Grappling with the West and other foreigners
PDF
Disparity of power: The United States engagement with Korea
PDF
How China's economic reform changed domestic circumstances for exports during the 1980s
PDF
Maternal devotion: the symbiotic relationship between mothers and sons in Yi Jian Zhi
PDF
The collapse of Japan's bubble economy and its consequences
PDF
The impact of ethnic mobilization in postwar Japan: A reflection of Japan's two Korea policy
PDF
The Japanese mainstream media in the Sino-Japanese rapprochement, 1964--1972
PDF
From denationalization to patriotic leadership: Chinese Christian colleges, 1920s--1930s
PDF
The independence of binding and intensification
PDF
Property and subjectivity: Women's property rights in Joseon Korea (1392--1910)
PDF
Religion and literary practice in the essays and fiction of Endo Shusaku
PDF
Women'S Education In The Tokugawa Society
PDF
A picture of human composition: An interpretation of Mencius' theory about human nature
Asset Metadata
Creator
Bergeton, Uffe
(author)
Core Title
The evolving vocabulary of otherness in pre-imperial China: From 'belligerent others' to 'cultural others'
Degree
Master of Arts
Degree Program
East Asian Languages and Cultures
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
History, ancient,history, Asia, Australia and Oceania,Language, Ancient,OAI-PMH Harvest
Language
English
Contributor
Digitized by ProQuest
(provenance)
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c16-51284
Unique identifier
UC11328855
Identifier
1437581.pdf (filename),usctheses-c16-51284 (legacy record id)
Legacy Identifier
1437581.pdf
Dmrecord
51284
Document Type
Thesis
Rights
Bergeton, Uffe
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the au...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus, Los Angeles, California 90089, USA
Tags
history, Asia, Australia and Oceania
Language, Ancient