UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Sui-Tang foreign policy: four case studies Pan, Yihong 1990

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-UBC_1990_A1 P36.pdf [ 22.09MB ]
JSON: 831-1.0098752.json
JSON-LD: 831-1.0098752-ld.json
RDF/XML (Pretty): 831-1.0098752-rdf.xml
RDF/JSON: 831-1.0098752-rdf.json
Turtle: 831-1.0098752-turtle.txt
N-Triples: 831-1.0098752-rdf-ntriples.txt
Original Record: 831-1.0098752-source.json
Full Text

Full Text

SUI-TANG FOREIGN POLICY: FOUR CASE STUDIES By YTHONG PAN Diploma, University of International Business and Economics, Beijing, China, 1978 M.A., University of International Business and Economics, Beijing, China, 1981 A THESIS SUBMITTED EST PARTIAL FULFILLMENT OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES (Department of Asian Studies) We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September 1990 Yihong Pan, 1990 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial, gain shall not be allowed without my written permission. Department of /\sja.n S^lfW/'e^ The University of British Columbia Vancouver, Canada Date Ovk. t ^ O DE-6 (2/88) ABSTRACT The foreign policy of imperial China had two major aspects: 1) ideological purity, based on the Chinese cosmological view of the state, and emphasizing the all embracing rule of the Chinese Son of Heaven. 2) Practicality and flexibility, which provided imperial rulers the justification for conducting foreign relations on an equal footing with their neighbors, and allowed for retreat from claims of Chinese superiority, or even paying tribute to "barbarians." These two aspects have been noted and studied previously. In this dissertation I examine the interplay of the twin aspects in Sui-Tang foreign policy decisions and their implementation, how they clashed with or accommodated each other both when China was strong and when it was weak. Chapter I provides a survey of the tribute system, its roots in the pre-Qin period, its development in Han and the challenges it faced in the Period of Division. The Sui-Tang policy of resettlement of the Turks who had submitted, is the theme of Chapter II. The chapter examines the Tang system of the "subordinated area commands and prefectures." The Sui-Tang settlement policy was intended to bring the "barbarians" under Chinese administration and to use the nomads as a military force against other "barbarians." It also drew a distinct line between the non-Chinese and the Chinese so that the "barbarians" would not disturb the Chinese and would undergo a gradual process of sinification. But the success of the policy depended basically on the balance of power. The war policy of the Sui-Tang Chinese towards Koguryo, its motives and result are studied in Chapter III. For the better part of a century the Chinese made persistent efforts to establish their administration on the Korean peninsula through force. While there is a contrast between the pragmatism of Emperor Wen on the one hand, and the obsession with military glory of Emperor Yang and Taizong on the other, all three emperors insisted on Chinese superiority over the Koreans and all had considerations for frontier security. The differences in their attitudes lay mainly in the extent to which China should claim the superiority. Eventually, the Chinese were quite happy to withdraw beyond the Yalu River and accept Korea as a peaceful tributary. The alliance between Tang and the Uighur empire is the topic of Chapter IV. While before the outbreak of the An Lushan rebellion in 755 the Uighurs were at times subjects of Tang, the period after 755 saw the growth of the Uighur empire and the weakening of Tang superiority. In both periods their relations were characterized by an alliance based on common interests. In the latter period the Chinese had to treat the Uighurs as an equal power but the relationship was still maintained under the tribute system, which served to maintain the outward form of Chinese superiority. The seven Tang-Tibetan treaties are discussed in Chapter V. Compared with Tang relations with other peoples, the Tang-Tibetan relationship was remarkably equal. This was shown both in diplomatic reciprocity and in the conclusion of treaties. Nevertheless, some Chinese officials still held strongly to the idea that the Tibetans were "barbarians," which hindered the maintenance of the treaties. In the making of foreign policy in imperial China, the two major aspects, ideological purity and practicality, were reflected in two principles of Confucian doctrine: "the king leaves nothing and nobody outside his realm," and "having the various states of Xia within, and keeping the Yi and Di barbarians out." While the first principle represented the ideological purity and provided justification for Chinese expansion, the second stressed practicality, thus the two aspects achieved a balance. IV Contents Abstract ii Explanatory Notes viii List of Tables ix List of Figures and Maps x Chinese Dynasties and Sui-Tang Emperors xi Acknowledgement xiii Introduction 1 The Chinese Sources 5 The Sources of the Accounts of the Turks, Korea, the Uighurs and Tibet in the Jiu Tangshu 13 The Non-Chinese Sources 16 Modern Collections of Original Sources 18 Chapter I. Historical Background: the Tribute System 20 1. The Ideological Background of the Tribute System 20 The Late Shang Dynasty: An Embryonic Form of Tribute System 20 The "Feudal System" (fengjian): A Hierarchical Social Order 22 Competition for Supremacy 24 The Tribute System in the Confucian Works 27 Chinese versus "Barbarian": Sinocentrism 29 2. The Han Dynasty: Development of the Tribute System 34 The Heqin Policy in the Early Period 34 Han Expansion 37 Practices under the Tribute System 42 Different Types of Relations under the Tribute System 44 Universal Empire through the Tribute System 52 3. The Period of Division: the Challenge to the Tribute System 57 The Challenge to the Idea of Universal Kingship 57 The Legitimacy of the Sixteen Kingdoms 58 The Legitimacy of the Northern Dynasties 60 The Persistence of the Tribute System in Former Qin 62 The Tribute System in the Northern and Southern Dynasties 64 The Challenge of the Nomads to the Northern Dynasties 68 Conclusion 72 Chapter II. Sui-Tang Relations with the Tujue Turks: the Policy of Settlement 75 Traditional Chinese Measures to Deal with the Nomads 75 The Internal Conflicts among the Turks 80 The Sui Victory over the Turks 81 Ishbara Qaghan of the Eastern Turks under Sui Hegemonic Rule 84 The Eastern Turks under Sui Hegemonic Rule during the Time of Emperor Wen 86 Qimin Qaghan under Sui Hegemonic Rule during the Time of Emperor Yang 90 The Conflicts between the Turks and Sui 91 Sui and the Western Turks 93 A Period of Chinese Civil War 95 The Problems of the Turks in the Early Tang 98 Tang Conquest of the Eastern Turks 102 The Conquest of the Xueyantuo 104 Victories over the Western Turks 107 Heavenly Qaghan and the Debate on the Policy of the Resettlement of the Turks 108 The Resttlement of the Turks 113 The Jimi Fu-zhou 120 Administration of the Jimi Fu-zhou 122 Setback in the Turkish Settlement 127 China and the Revival of the Second Turkish Empire 131 The Change in Tang Frontier Defences 134 Xuanzong and the Last Phase of the Second Turkish Empire 137 Tang and the Turgish of the Western Turks 143 Conclusion 144 Chapter III The Sui-Tang Wars with Koguryd 150 Opinions of Modern Scholars 151 Historical Relations before Sui-Tang 154 Emperor Wen: A War of Deterrence 156 vi Emperor Yang: Three Wars of Aggression 160 Tang Gaozu: Peaceful Communication 166 Tang Taizong: the Early Peace 169 Decisions for Wars 172 Three Wars of Expansion 180 The Reasons for Failure 185 Tang Gaozong: the Conquest of Paekche 189 The Conquest of Koguryo 193 The Tang Abandonment of Conquest 195 Tang and Parhae 200 Conclusion 202 Chapter I V . The Sino-Uighur Alliance 205 Alliance: An Important Feature in the Sino-Uighur Relationship 205 The Name of the Uighurs 208 Sui-Tang and the Tiele before 647: A Traditional Alliance 209 647-679: the Tiele under the Chinese Administration 212 679-744: the Tiele between Tang and the Second Turkish Empire 215 744-755: the Founding of the Uighur Empire 220 Suzong: the Uighur Assistance and the Chinese Attitude 221 Daizong: the Uighur Assistance and the Chinese Attitude 228 Dezong: An Anti-Uighur Policy and Its Abandonment 236 788 to 840: A Period of Peace 241 The Horse-Silk Trade between the Uighurs and Tang 245 The Changes and Their Impact in the Uighur Empire 251 The End of the Uighur Empire 258 Conclusion 262 Chapter V. The Tang-Tibetan Treaties 265 Different Problems 266 The First Encounter and the Marriage of Princess Wencheng 271 A Period of Competition 274 The First Treaty in 706 • 283 A Peace Settlement in716 287 The Second Treaty in 732 293 The Third Treaty in 762 and Tibetan Invasion on Chang'an 300 The Fourth Treaty in 765 305 vii The Fifth Treaty in 767 306 The Sixth Treaty in 783 308 The Alliance with Tibet 316 The False Treaty of Pingliang in 787 318 Alliance with Nanzhao and the Uighurs against Tibet 322 The Seventh Treaty of 821/822 324 The Weizhou Incident 331 Conclusion 333 Chapter V I . Conclusion 339 Works Cited 353 Abbreviations 353 Traditional Chinese Works Listed by Titles 355 Modern Works Listed by Authors 358 Appendix 375 Explanatory Notes 1. The pinyin system is used for the romanization of Chinese. In the quotations of English works, except titles of books and articles and names of authors, any kind of romanization is changed into pinyin. 2. A full reference to all the books and articles cited in this dissertation and abbreviated tides is given alphabetically in three sections of the "Works Cited" at the end of the paper. Section one is "Abbreviations." Section two is "Traditional Chinese Works Listed by Titles." And section three is "Modern Works Listed by Authors." 3. Chinese years are converted to the western calendar according to the year to which the greater part of the Chinese year corresponds. The conversion is based on Wan Guoding. 1978. 4. Translation of Chinese official titles is based on Hucker, C. O. 1985. 5. Transcription of Tibetan names: h --ng; ri ~ ny; z — zh; s ~ s. 6. Chinese script for names, titles and terms are given in the text only the first time they appear. ix List of Tables 1.1. Foreign tribute missions to Han 380 2. 1. The Turkish rulers 385 2.2. The Participation in Tang expeditions by the Eastern Turks and the Tiele (634-669) -387 2. 3. The Jimi fu-zhou of the Eastern Turks and the Tiele (630-663) 388 2. 4. The Jimi fu-zhou set up for the Turks in Xuanzong's Time 390 3. 1. The Korean Rulers 391 3. 2. Tribute missions of the three Korean kingdoms to Sui and Tang (from 581 to 712) 393 4. 1. The Uighur rulers 395 5. 1. The Tibetan rulers 397 X List of Figures and Maps Figure 1. The nine-zone system 398 2. Two models of empire 399 Maps 1. Western Han and its neighbors 400 2. Korea at the height of Koguryo expansion in the 5th century 401 3. Chen, Qi, Zhou and their neighbors (572) 402 4. Sui and its nieghbors 403 5. The Protectorates (duhufu) of Anbei and Chanyu (669) 404 6. Tang and its neighbors (741) 405 7. Tang and its neighbors (820) 406 THE CHINESE DYNASTIES Late Shang (Yin): 1300 B.C.E. to 1028 B.C.E. Zhou: 1027 B.C.E. to 256 B.C.E. Qin: 221 B.C.E. to 207 B.C.E. Han: 206 B.C.E. to 220 CE. Former Han: 206 B.C.E. to 8 CE. Wang Mang: 9 A.C to 23 CE. Emperor Gengshi: 23-25 Later Han: 25-220 The Southern Dynasties Song: 420-479 Qi: 479-502 Liang: 502-557 Chen: 557-589 The Northern Dynasties Northern Wei: 386-534 Eastern Wei: 534-550 Western Wei: 535-556 Northern Qi: 550-577 Northern Zhou: 557-581 Sui: 581 to 618 Tang: 618 to 907 THE SUI EMPERORS Emperor Emperor Wen Emperor Yang Emperor Gong Reigned 581-604 605-617 617-618 THE TANG EMPERORS Gaozu 618-626 Taizong 627-649 Gaozong 650-683 xii Zhongzong 684 Wu-Zhou (Empress Wu) 684-705 Zhongzong 705-710 Wen Wang 710 Ruizong 710-712 Xuanzong 712-756 Suzong 756-763 Daizong 763-779 Dezong 780-805 Shunzong 805 Xianzong 806-820 Muzong 821-824 Jingzong 825-827 Wenzong 827-840 Wuzong 841-846 Xuanzong 847-860 Yizong 860-874 Xizong 874-888 Zhaozong 889-904 Emperor Zhaoxuan 905-907 Acknowledgement I would like to express my deep gratitude to Professor E. G. Pulleyblank, whose profound scholarship and penetrating insights have guided me throughout my academic work. His analytical criticism and kind advice have encouraged me in my study. I am very grateful to Professors Donald L. Baker and Jo-shui Chen, for their insightful and detailed comments and generous help. I would also like to thank Dr. Gareth Sparham, my friend, for his careful and painstaking efforts in editorial work on the dissertation. I am also grateful to Mr. Gonnami, of the Asian Studies Library at TJ. B. C. who has helped me with some problems in Japanese transcription. Introduction Any understanding of the foreign policy of imperial China must take into account two major aspects: 1) Ideological purity. By this I mean the established ideal pattern of relationships, from the Chinese point of view, that should obtain between China and other states. It was a pattern based on a particular cosmological view of the state and it emerged from, and evolved in, the particular environment of China, positing a hierarchical and fundamentally sinocentric order to the Chinese known world. It emphasized the all embracing rule of the Chinese emperor, the Son of Heaven, who was supposedly the only legitimate ruler of, not only China, but all-under-heaven, and it stressed the importance of the virtue and capability of the Son of Heaven. This pattern was reflected in ceremonial rituals under the tribute system. 2) Practicality and flexibility. This was a basic pragmatism which provided imperial rulers the scope, or the justification, for conducting foreign relations on an equal footing with their neighbors. It allowed for acceptance of the need to retreat from claims of Chinese superiority, or even to pay tribute to "barbarians." Such compromising measures could be carried out for as long as they were seen to be entirely for China's own security and stability. There are two collections of essays in English, examining Chinese foreign policy during imperial times, taking these two aspects as their respective focuses. These are The Chinese World Order edited by J. K. Fairbank and China among Equals edited by M. Rossabi.1 While the former collection of essays revolves around the theme of ideological purity and examines the Chinese world order as it is mirrored in the long-lasting tribute system, the latter collection concentrates on the realism and 1 Fairbank, J. K. 1968; Rossabi, M. 1983. pragmatism that lie behind certain imperial foreign policy decisions. It challenges the myth that, from the Han dynasty on, the Chinese uniformly and rigidly applied their system of foreign relations, that China lacked interest in foreign commerce and that China was ignorant of foreign lands.2 Each of these books analyzes the situation of a particular period. The former book is mainly concerned with the later imperial period in which the Chinese shocked the Westerners by their sense of superiority and sinocentrism, while the latter concentrates on the Song period when China was in fact in an inferior position in its relations with the Khitan-Liao and the Jurchen-Jin dynasties in the hierarchical international order. In this dissertation I examine the interplay of the twin aspects of ideological purity and practicality in Sui and Tang foreign policy decisions and their implementation, how they clashed with or accommodated each other and how they were both included in the decision-making process either when China was strong or when it was weak. The dissertation attempts to show how both aspects must be taken into account to explain fully Chinese foreign policy. The Tang dynasty, like the Han before it, was a universal empire which embraced and absorbed a wide spectrum of political, economic, military and cultural elements bequeathed to it by earlier dynasties and by foreign peoples. Unlike Han, though, whose only major organized rival was the Xiongnu empire, Tang faced an international situation hitherto unique for a Chinese empire. The situation was largely determined by a) the two consecutive nomadic empires, the Turks and Uighurs on the northern and northwestern frontiers, b) the three Korean kingdoms in the east which had gradually emerged as a force from the Later Han through the Period of Division, and c) Tibet, a newly emerged kingdom along the Chinese western frontier. 2 See also the reviews of these two books by Pulleyblank, E. G. 1969, pp. 423-425, and Zurndorfer, H. T. 1988, pp. 141-143. The dissertation is not intended to be a comprehensive study of all aspects of Tang foreign policy. That would require a much longer work. Following Chapter I, which provides a survey of the tribute system, its roots in ideas of the pre-Qin period, its development in the Han and the challenges it faced in the Period of Division, the dissertation restricts its scope to the four cases below, focusing on a particular type of policy in each. 1) The Sui-Tang policy of settlement of the Turks who had submitted is the theme of Chapter U. The chapter examines the Chinese attempt to incorporate the Turks who had submitted into their hegemonic empire and its result. Wolfram Eberhard, when defining some general patterns of historical development of the relationships between social groups, differentiates two types. One brings together two or more social groups into a new unit, and one is a breakdown of once-unified societies into smaller units.3 In the Sui-Tang policy of Turkish settlement, how did the policy fit into these types? How did the ideology of Chinese foreign relations lay a foundation and provide a theoretical justification as well as a cultural centripetal force for the tendency of unification? 2) The war policy of the Sui-Tang Chinese towards Koguryo, its motives and result are studied in Chapter TH. K. J. Holsti in his book International Politics points out that regardless of historical and geographical context, policy makers of different types of political units, whether tribes, city-states, empires, or modern nations, have attempted to achieve objectives or defend their interests by fundamentally similar techniques, mainly the use of force and the construction of alliances.4 This can also be applied to the study of Chinese foreign policy measures and can be particularly well illustrated in the case of Koguryo. 3 Eberhard, W. 1982, p. 4. 4 Holsti, K. J. 1983, p. 4. 3) The alliance between Tang and the Uighur empire is the topic of Chapter IV. The formation and maintenance of the alliance was an important feature in the relationship between the two states. Questions such as how and why the alliance could be achieved and maintained, and whether the Chinese adopted an attitude of equality in the alliance are dealt with. 4) The seven Tang-Tibetan treaties are discussed in Chapter V. Wang Gungwu holds that after the An Lushan rebellion Tang was struggling to survive as a military empire, but: it was probably unthinkable that the Tang court should have considered any dilution of their claims to superiority just because the imperial writ did not cover as large an area as it did at the height of the empire's power.5 To what extent is this true? In the conclusion of treaties with Tibet, how did Tang rulers give up claims to superiority in exchange for peaceful relations? Through the study of these four cases, I believe that one gets a deeper understanding of the methods by which foreign policy was conducted in traditional China. Besides the question of foreign policy as such, there is the separate question of the making of foreign policy which relates to domestic politics and the competing interests and aspirations of groups within the country, rather than to external factors. It concerns the role of the policy makers, how their values, beliefs, education, understanding and assessment of reality influence the decision-making process, and deserves a separate and detailed study which I hope to make in the future. In the present dissertation I discuss it only briefly. Some general remarks on the subject that 5 Wang Gungwu. 1983, p. 47. 5 can be illustrated from events referred to in the thesis are drawn together in the conclusion. The Chinese Sources This dissertation, which is concerned primarily with Chinese foreign policy, not the whole issue of the international relations of East and Central Asia in the Tang period, is based mainly on sources in Chinese. Traditional Chinese histories are centered around Chinese court politics. This feature makes them convenient for the study of Chinese foreign policy but at the same time places limitations on our view of foreign states. These histories usually only record those peoples that had direct contact with China and events that influenced its foreign relations in various ways. There are other defects in Chinese historical writings, such as the use of degrading language to describe non-Chinese. These features of Chinese histories have been discussed by previous authors and I will not discuss them further here. The numerous Chinese historical records, in spite of such shortcomings, however, are invaluable to our study of the history of China's foreign policy. The official histories provide the main body of information but other materials supplement them in various ways. For the Sui-Tang period, the Chinese works that I use are as follows. 1) The dynastic histories The Suishu f%%, compiled in the Tang Bureau of Historiography (shiguan Jr l f l ) by Wei Z h e n g Y a n Shigu^fif $ , Kong Yingda^k 6 b) the Suishu compiled in the Sui period by Wang Shao SJ^p ; and c) other materials collected by the Tang historians.6 The Beishi /ifc£_, compiled privately in Tang by Li Yanshou from 643 to 659. It was based on the manuscripts of Li's father Li Dashi ^ X£*p and enriched by Li Yanshou's own research in the Tang Bureau of Historiography.7 The Jiu Tangshu^ ^ , compiled in Later Jin first under the leadership of the Grand Councilor Zhao Ying^<^ and then Liu XuJ|\iJ B^J , written by Zhang Zhiyuanv5^^C^_and others, from 941 to 945. It was based on materials produced by the Tang Bureau of Historiography. For the period from 618 to 759, the "National Histories" (guoshi\^^), completed successively by Wei S h u ^ ^ _ , Liu Fang ^ % , Yu Xiulie ^I^SL and Linghu Huan^^Y^Jt, were the basis of the Jiu Tangshu and the compilers took over bodily from the already prepared histories without much change. For the period from 759 to 847, materials were drawn largely from the "veritable records" (shilu~p^ ) of the various reigns, and for the last period of Tang, the "daily records" (rili 13}ylf ) w e r e t n e basis, since the veritable records were either lost or had never been completed. In addition the compilers consulted materials collected at the time of the project8 The Xin Tangshu , compiled in the Song Bureau of Historiography by Ouyang Xiu^j^j^Song Qy^Sf1 and others, from 1045 to 1061. The basic sources for the work included: a) the Jiu Tangshu; and b) other materials available in the Imperial Library in Northern Song.9 The Xin Tangshu 6 Li Zongye. 1982, pp. 242-243. 7 BS, vlOO, pp. 3343-45; Li Zongye. 1982, pp. 253-255. 8 Nianershi Zhaji, vl6; Pulleyblank, E. G. 1950, pp. 448-457; Twitchett, D. 1956, pp. 48-49; Li Zongye. 1982, pp. 258-260. 9 Pulleyblank, E. G. 1950, p. 457; Li Zongye. 1982, pp. 265-266. creates a section of the "subordinated prefectures" (jimi zhou^-j^j^ in chapter 43B and has a chapter of "foreign generals" (Fan jiang^jftej ), both of which make the study of foreign relations more systematic. 2) Other major historical works The Da Tang Chuangye Qijuzhu 7^%^%£l=Jfe.i£- , the court journal of Gaozu, compiled by Wen Daya>jl£#&overmg the period from 617 to 626. As a chronological resume" of the events concerning L i Yuan^^i j and his founding of the Tang dynasty from 615 up to the proclaiming of the new emperor, the book may be regarded as the first historical record of Tang. 1 0 The Zhenguan Zhengyao ^^jS^^r , a record of Tang Taizong's discussions with his niinisters, the memorials and suggestions from the ministers, and administrative measures of the period, collected and edited by Wu Jing^£^_, . The book was completed around 720.11 The Tongdian^A^ , compiled by Du Yo\iJ.±1£ from 766 to 801. It continued and enlarged the Zhengdian^X^. of Liu Z h i ^ ] ^ . 1 2 The Tang sections in the chapters on foreign countries were presumably taken from materials compiled in the Bureau of Historiography. It has the fullest account of the negotiation between Tang official Guo YuanzhenJ^x^^and Tibetan general Mgar Khri 'bring in Empress Wu's time.13 The Tang Huiyao , compiled by Wang Pu X _ > ^ a n d presented to the throne in 961 in early Song. It is a combination of the Huiyao of Su Mian ^ presented to the throne in 804, and the Xu Huiyao ^ of Yang S h a o f u ; ^ £ ^ J | j > r e s e n t e d in 853, with the addition of a very small 1 0 For a detailed study of the book, see Bingham, W. 1937, pp. 568-574. UZGZY, 1978, pp. 1-2. 1 2 Li Zongye. 1982, pp. 307-309. 13 TD, vl90, pp. 1023-24; Z2T7, v205, p. 6508. 8 amount of material for the period after 852.14 It has the most complete record about the discussions held early in Taizong's reign concerning the settlement of the Turks who had submitted.15 The Cefu Yuangui-ltfj-jft X-jJJulL » compiled by Wang Qinruoi. §ZJ%, Yang Yi 3^ jye. and others in Song under an imperial commission from 1005 to 1013. The materials concerning the Tang period were from the veritable records and Tang histories.16 It contains therefore more original sources than the Jiu Tangshu and the Xin Tangshu and more complete versions of documents abridged in other sources. With its organization of numerous materials under different topics, such as the section on "foreign vassals" (waichen bu$\-*j§j^> ), the Cefu Yuangui is particularly useful to my study. It records tributary missions from foreign countries which other histories have chosen not to mention. One example is in the case of the Turks. According to the Cefu Yuangui, during Tang Gaozu's reign, the Turks sent tributary missions to the Tang court almost every year. In the basic annals of Gaozu and the accounts of the Turks in the Jiu Tangshu and Xin Tangshu, on the other hand, it is reported that the Turks made incursions on Tang frontier almost every year, but records about the Turkish tributary missions are not complete. Some of the Turkish missions might have just arrived at the Chinese court to deliver messages, but the fact that they are still recorded as tributary missions implies that the foreigners had to conform to the rituals of the tribute system laid down by the Chinese in order to be accepted at the Chinese court. The Zizhi Tongjian^^i^^u. , compiled by a group of Song historians with Sima Guang ^J|zii_i as the chief editor from 1065 to 1084.17 1 4 Twitchett, D. 1956, p. 51. 1 5 THY, v73, pp. 1312-14. 1 6 Liu Naihe. 1983, pp. 1-28. 1 7 Li Zongye. 1982, pp. 335-336. When compiling the Tang section (chapters 185 to 265), Sima Guang consulted all the materials available at the time, including the histories mentioned above with a great, many additions. It is well recognized that the Zizhi Tongjian is superior to all other histories of the Tang period in terms of the amount of the material consulted, analysis of original sources and clarity of expression. The Zizhi Tongjian Kaoyi ^ Jjl^  is a series of notes made by Sima Guang while editing his history which frequently quotes from books that are now lost and provides clues as to the sources for information in his history. Here I shall discuss briefly some important works relevant to my dissertation that are referred to by the Zizhi Tongjian Kaoyi.l% Except for the veritable records, in the Tang part of the Zizhi Tongjian, there is no specific work mentioned concerning Korea. In the records referring to the Turks, only two works are mentioned by the Kaoyi:: a) passages about the Turgish in the Jinglong Wenguan i$£4Z->> compiled by Wu Pingyi , and b) a passage about the Eastern Turkish invasion of Chang'an in the Xiaoshuo /J^$Z_> of Liu Su^J i k -The Jinglong Wenguan Ji, ten chapters, compiled by Wu Pingyi, a scholar of the "literary institute" (wenguanXJtt7) in the time of Zhongzong (705-710). It is a collection of essays, poems and biographies of some scholars in the institute. Some records in this work referred to a contemporary event that paralleled the veritable records. Fragments of the work can be found in the Shuofu %Y^^ collection. The Xiaoshuo, three chapters, written by Liu Su, an official historiographer in the Tianbao period (742-755) and a son of the famous historian Liu Zhiji - f^ j^t? The Zhizhai Shulu Jieti ~M^%^t^J^-^ , a Song bibliographical work, mentions this work as of three chapters, but no further 1 8 Concerning these works, see XTS, v58, and bibliographical works compiled in the Song dynasty. They are Chongwen Zongmu Jishi, Junzhai Dushuzhi, Zhizhai Shulu Jieti, and Yuhai. information can be found,19 Though the Kaoyi quotes from the Xiaoshuo, it considers the information unreliable.20 The work is no longer extant As for the Uighurs and Tibet, the Kaoyi quotes many more works listed below. The Fenyangwang JiazhuanV^^x~^/\^ (= Guogong Jiazhuan ^2>^S^ ), eight chapters, a family biography of Guo Ziyijf^-|~ iBc (696-781) by Chen Kong^pj^, a staff officer under Guo for some time. Guo Ziyi was a very important high-ranking military official involved in many battles in alliance with the Uighurs in the suppression of the An Lushan ^ ^ J J U rebellion and in wars against the Tibetans. The Kaoyi often refers to this work. It gave a lot of additional materials from eye-witness accounts, but was not entirely reliable. Regarding one matter with reference to the Uighurs, the work contained the only information and was used by the Tongjian.21 Unfortunately the work does not exist any more. The Yehou Jiazhuan lij$i^%Kl^ , ten chapters, a family biography of the Grand Councilor Li Mi ^ ;ljk (722-789), by his son Li Fan . Although in many places the work contains unreliable information since Li Fan intentionally glorified his father,22 in some occasions relative to the Uighurs and Tibet, the Kaoyi refers to the work as an additional contemporary record. It is also an important source for the study of the militia (fubing ) system.23 The work does not exist integrally, but its fragments are included in the Shuofu and other collective works. 1 9 Zhizhai Shulu Jieti, vll . 2QZZTJ, vl91,p. 6020. 2 1 ZZTJ, v225, p. 7236. 2 2 ZZTJ, v231, p. 7456; Junzhai Dushuzhi, v9; Mackerras, C. 1972, p. 3. 2 3 Pulleyblank, E. G. 1955, p. 142, note 9. 11 The Beihuang Junzhanglu}\\i^jj%%^3^ , three chapters, also written by L i Fan, referred to by the Kaoyi in one or two places. The work is lost but from the title, one can see that it was a historical record about the non-Chinese chieftains to the north of Tang. The Duan Xiushi Biezhuan £&J%fe%l}1% (= D uangong Jiazhuan ), two chapters, a family biography of Duan Xiushi (718-783), compiled in the Bureau of Historiography by Ma Yu , who was the Vice Director of the Imperial Library in the Yuanhe period (806-819). Duan Xiushi was also a Wgh-ranking military official who was involved in frontier affairs particularly with the Tibetans. It is another case where the family biography provided additional information. The work does not exist any more. The Binzhirff^ , two or three chapters, by Ling Z h u n - ^ ^ - , an account of the history of the local armies in Binzhou ^ p>f|. The work covered Tang dealings with Tibet from 763 to 787 which involved Guo Ziyi. According to the Kaoyi Ling Zhun was unreliable because of his bias against Guo Ziy i . 2 4 Yet as an independent source, the work was often referred to by the Kaoyi to clarify some events. Ling Zhun was involved in the group of Wang Shuwen il^jXX. , 2 5 the famous reformer in the early ninth century. The work does not exist any more. The Huicnang Fapan / i ^ ^ f ^ ^ f S - . , one chapter, an account of the suppression of the Ze Lu j^a^ rebellion, the Wenwu Liangchao Xianti Ji "y^jK Jl^ ilF «2_> » three chapters, a collection of memorials, and Huicnang Yipin Ji ^ ^ - ^ ^ , a collection of works including official letters and edicts written on behalf of the emperors, all by L i Dey u ^ fj§^ <^ « (786-849). The latter two still exist while the first one is lost. In his long official career L i Deyu occupied various posts such as Hanlin Academician (Hanlin xueshi 2 4 ZZ77, v223, p. 7156. 25XTS, vl68, p. 5127. ^^J^-ZL ), Military Commissioner and Grand Councilor. He played an important part in the court decision making process and was very much involved in external affairs, particularly during the time of the collapse of the Uighurs. His various writings constitute an important source of information. The Tang Tongji /§£EtiL , compiled by Chen Yu fjfij^. It was a chronological account of the history from the beginning of Tang till 823 in one hundred chapters, but in Song there were only forty chapters left, ending at the time of Empress Wu.26 The Kaoyi refers to the work often. For one incident involving Tibet, the work was the only source and the Tongjian incorporates the information.27 Since the work is no longer extant and not much information can be found about the writer, we are unable to tell more about it The Bu Guoshi /p^l^}^ , six or ten chapters, compiled by Lin En 'fef^ who was a "presented scholar" (jinshijfe^L ) of the time of Xizong (874-888). The work does not exist any more but at the time when the Zizhi Tongjian was compiled, it was an important source for the later part of the Tang dynasty when few other materials were available. It had detailed accounts of events about Tibet at the end of Tibetan kingdom. Regarding one event involving Tibet, it provided the sole source and the Tongjian followed it 2 8 3) Collected works of Tang individuals These works include such government officials as Zhang Jiuling ^^/u j j j j ^ , Lu Zhi fesf^ , Bai Juyi ^/Jfr-JJ , and Li Deyu, who composed official documents on frontier affairs and relations with foreign countries, including official letters to foreign rulers on behalf of the emperors, and who wrote 2 6 Zhizhai Shidu Jieti, v4. 2 7 ZZTJ, v205, p. 6493. 2 8 ZZTJ, v246, p. 7938. memorials on foreign policy or otherwise referred to such matters in their private writings. There is a considerable amount of material relative to Tang history among the documents discovered in Dunhuangjf^ J?_ and Turfan. I find that they do not have much relevance for the purpose of my study, so they are not referred to in the dissertation. The Sources of the Accounts of the Turks, Korea, the Uighurs and Tibet in the Jiu Tangshu In their contacts with foreign peoples, especially their close neighbors, the Chinese showed great interest in collecting information on their land, geographical conditions, customs, histories and their communications with China. Chinese envoys to these foreign places, frontier officials, military generals and offices in charge of receiving foreign envoys used opportunities provided by their posts to collect information and produced numerous works and maps about non-Chinese peoples. In the monographs on bibliography in Chinese official histories, one finds many interesting tides of works concerning non-Chinese. Although most of them have unfortunately been lost in the course of time, nevertheless one can assume that they must have provided much information that was useful to contemporary Chinese governments in both diplomatic and military dealings with foreigners and for the compiling of official histories. For example, during Tang Taizong's time, a military officer, Wei Hongji JJL ^L$!fc, went on an embassy to the Western Turks but he was trapped there for three years, during which time he wrote a book, the Xizheng Ji ttii?fl£. f , about the various products and customs of the states on the route from Tang to the west. On Wei's return, Taizong inquired about foreign countries and when Wei presented the book, Taizong was greatly pleased.29 Under Tang Gaozong the court sent special missions to the Western Regions to collect information, and, as a result, the Bureau of Historiography produced the Xiyu Guozhi r^t^l^^^ with maps and, perhaps, other illustrations.30 The several geographical works compiled by Jia Dan fifj^L* > a Grand Councilor of the Zhenyuan period (785-804), provided a detailed study of the communication routes between Tang and foreign countries.31 Based on such original sources the "Accounts of Foreign States" (waiguo zhuan ^ ) in the Jiu Tangs hu embody a large amount of important information. By comparing the Accounts of Gaoli liv/f^  (Koguryo), Baiji ^'/f^~ (Paekche) and Xinluoz-pf^ ft. (Silla) in chapter 81 of the Suishu with chapter 94 of the Beishi, one finds that they have similar records, which shows that the Suishu is the basic source for the Beishi. The passages concerning the three Korean kingdoms in the Tongdian and the Zizhi Tongjian also show that the compilers of these two histories used the Suishu.32 In the chapters on the Eastern and Western Tujue x[yf^ Turi^ s and the Tiele^^^)of the Sui period, the same is also true.33 Both the Tongdian, compiled earlier than the Jiu Tangshu, and the Tang Huiyao, most of which was compiled earlier than the Jiu Tangshu, have chapters on foreign states, and therefore deal with the same subjects as those in the Jiu Tangshu. By comparing these chapters, one can come to the conclusion that the three books rely basically on the same sources, that is, records from the Bureau of Historiography, and the editor of the Jiu Tangshu did not simply copy the other two, but rearranged and amplified a great deal based on other different documents and materials. 2 9 XTS, vlOO, p. 3944. 3 0 XTS, v58, p. 1506. 3 1 XTS, v43B, pp. 1146-55; v58, p. 1506. 3 2 TD, vl86, Gaoli; ZZTJ, vl71-vl83. In the TD, vl85, sections of Baiji and Xinluo, the records concerning the Sui period are very brief. 3 3 SUIS, v84, the Accounts of Tujue, Western Tujue and Tiele; BS, v99, the Account of Tujue, Western Tujue and Tiele; TD, vl97; vl99; ZZTJ, vl71-vl83. Chapter 194, the Account of the Tujue (Turks), in the Jiu Tangshu is similar to, or the same as, the Account of the Turks in chapters 197 to 199 in the Tongdian with some additions. It is therefore probable that both sections were based on a common source. We can find few similar passages or fragments in chapter 199A, the Account of Koguryo, Paekche and Silla, of the Jiu Tangshu and chapters 185 and 186 of the Tongdian, and chapter 95 of the Tang Huiyao. It would appear therefore that the Jiu Tangshu compiled a new account in this case from different sources. A few passages in chapter 195, the Account of the Huihe fi^ Zl>(Uighurs), of the Jiu Tangshu and in chapter 98 of the Tang Huiyao have similarities. The section on the Uighurs in the Tongdian is very brief. While the editor of the Jiu Tangshu may have drawn most of his material from the veritable records,34 he may also have had the text of the Tang Huiyao in front of him, which he abridged and rearranged,35 adding other material from different sources. The same may also be true of the relationship between the chapter on Tibet in the Jiu Tangshu and related sections in the Tongdian and the Tang Huiyao. In general, the accounts of foreign countries in the Jiu Tangshu are more precisely dated and in some cases more clearly written than those in the Xin Tangshu, but the latter does add more materials, as the compilers of the Xin Tangshu claimed. These added materials are drawn from the biographies in the Jiu Tangshu, from the Tang archives, and from other historical works still extant in the Song dynasty when the book was compiled. Besides materials emanating from writings by Chinese officials, the chapters on the Eastern Turks in the Suishu, and on Paekche, Silla, and Tibet in the Jiu Tangshu contain a certain amount of correspondence sent to the Chinese emperor by foreign 3 4 Mackerras, C. 1972, pp. 2-3. 3 5 Pulleyblank, E. G. 1956, pp. 39-40. rulers. Some of these letters demonstrate that the foreign rulers insisted on equality with China, such as the letter by the Turkish qaghan in 584,36 and some were written in accordance with the Chinese rhetoric, the most obvious case being when the foreign rulers refer to themselves as "vassals" (chen £.). To answer the question as to who wrote these letters, one speculation is that the foreign rulers had them written by Chinese who were living at their courts or by their own people who had learned Chinese. For example, in Taizong's reign following the marriage of Princess Wencheng X_$£ to the Tibetan king, Tibet sent young people to Tang to study, and invited Chinese scholars to Tibet to compile official reports to the Tang emperor on behalf of Tibet37 Korea adopted not only the Chinese bureaucratic system but also the Chinese script. In the Chinese court there were "official interpreters" (yiguanlj^z^ ). But due to lack of evidence it is difficult to say whether such Chinese officials translated foreign correspondence into proper Chinese both in language and in rhetoric. The Non-Chinese Sources Besides the traditional sources in Chinese, the historical writings of the non-Chinese states contain invaluable materials, which corroborate, differ from, or supplement the Chinese sources. These sources provide a non-Chinese perspective which one can not afford to neglect Due to my lack of knowledge of the Turkish and Tibetan languages, I rely on translations. These works include: 1) Old Turkish inscriptions of the eighth century. In his Tujue Jishi y^Jfik. , Cen Zhongmian ^f~f has three inscriptions translated into 3 6 SUIS, v84, p. 1868; BS, v99, p. 3293; ZZTJ, vl76, p. 5476. 3 7 TD, vl90, p. 1023; JTS, vl96A, pp. 5221-22; XTS, v216A, p. 6074; THY, v97, p. 1730; CFYG, v978, p. 11496. Chinese with annotation, the Tonyukuk, Kul Tigin and Bilga inscriptions. Talat Tekin in his A Grammar of Orkhon Turkic has an English translation of these three inscriptions and two other inscriptions, those of Ongin and Kiili Chor, but he does not have annotation. 2) In Korean traditional works written in Chinese script, the one that concerns my study the most is the Samguk Sagi 3~\^^tS~Z^ compiled in the twelfth century by Kim Pu-sik<2Ll^ p^ C - 3 8 Kim Pu-sik adhered to orthodox Confucian standard of historiography. The Samguk Yusa JEL)ij? j | " compiled by Iryon / T 3 ^ towards the end of the thirteenth century is also an important history but it does not have an immediate bearing on my study and it includes more myth and other less believable tales than the Samguk Sagi. 3) Inscriptions of the Uighurs. The Shine-usu inscription in Old Turkish script records the account of Moyanchuo ]^5&xx. Qaghan (reign 747-759) and has a translation in Chinese and annotation by Wang Jingru i L l f f ^ . The other important inscription was found in Karabalghasun. It has three parts, one in Chinese, one in Sogdian and one in Old Turkish, and was written in the ninth century. The Chinese part is best preserved. In my dissertation I use the annotated text in Haneda T o ' r u^f^J , "Todai kaikotsushi no kenkyu" 7§ f< f § ^<0 4) The numerous Tibetan sources are in three forms: inscriptions, manuscripts, and wooden tablets. Those that are the most important to my study of Tang-Tibetan relations, especially the study of the treaties concluded by the two sides, are the 821/822 inscription in Tibetan and Chinese scripts, and the Old Tibetan Annals and the Old Tibetan Chronicle. For the inscription, I use mainly the translations and annotations by H. E. Richardson in Ancient Historical edicts at Lhasa and the Mu 3 8 Gardiner, K. H. J. 1970, pp. 1-42. 18 TsunglKhri Gtsug Lde Brtsan Treaty of A. D. 821-822, from the Inscription at Lhasa, and by Li Fang-kuei in "The inscription of the Sino-Tibetan treaty of 821-822." Richardson's book and Wang Yao's Tufan Jinshi Lu vj.^-^ Zj also discuss another important document, the "Ngan Lam Stag Sgra Klu Khong inscription." Wang Yao has a translation of the two Tibetan historical works in his Dunhuang Gu Zangwen Lishi Wenshu ^ j ^ . ? ^ , ^ _ X . " ^ , and a study of the wooden tablets of the Tibetan kingdom (early seventh to late eighth century) discovered in modern Xinjiang ^ f f ^J . with the title Tufan Jiandu Zonglu iX-^ f t f / f ; f But these works do not have records about the treaties concluded between Tibet and the Tang court. Modern Collections of Original Sources Modern annotated collections of source materials and other critical studies that have been particularly useful in preparing this dissertation are the following. 1) The Turks: Cen Zhongmian, Xi Tujue Shiliao Buqueji Kaozheng y^j^^ft$f^r^ffknd Tujue Jishi, and Edouard Chavannes, Documents sur les Tou-Kiue [Turcs] Occidentaux. These works are comprehensive in their collection of materials and have detailed annotations. 2) Korea: The Chosen shi ^ W^yt^ collected by Chosen shi henshu kai itj&b^it °f Japan. It is a collection of highlights from all available materials in Chinese, Korean and Japanese traditional sources arranged in chronological order. The editor notes discrepancies when they occur but does not go into a detailed discussion. 3) The Uighurs: There has not been a comprehensive collection of traditional sources but important works with emphasis on annotation are: Haneda Tom, "Todai kaikotsushi no kenkyu;" Colin Mackerras, The Uighur Empire; Feng Jiasheng 4^? , Cheng Suluo %§J$q ^  and Mu Guangwen Weiwuerzu Shiliao Jianbian&j£%fc^£]fy$&fa . 4) Tibet: Materials about Tang and Tibet are more abundant in both Chinese and Tibetan languages than about the above mentioned peoples. Collections of Chinese sources are: Zangzu Shiliaoji , a collection of sources from the Chinese official histories;39 Tongjian Tufan Shiliao^Jl^^i.^ >C__ | ^ - ; 4 0 and C«/u Yuangui Tufan Shiliao Jiaozheng ^CjfiL^I^^L t^-y-^ ff^  , with brief annotation.41 In his Kodai Chibetto shi Kenkyu t^f^-f^^l^t^flfL, Sato Hisashi f£j^-^ provides a detailed study of primary sources in both Chinese and Tibetan concerning the Tang period. The Xin Tangshu Tufanzhuan Jianzheng .ffi^^ilM?1t[ ^stvi£ b v W a n g Zhong xJ&j* is an annotated work with references to Tibetan and other Chinese sources. Yamaguchi Zuiho Ll\ t j ^Z ^ L in his Toban Okoku Seiritsu shi no Kenkyu " j L ^ - X - l ^ j^l'^.'y^S) 4if/L has carried further the study of the sources about the history of early Tibet, and Beckwith in his The Tibetan Empire in Central Asia refers to sources in Tibetan, Arabic, Old Turkish, and Chinese. 3 9 Edited by Chen Xiezhang, Suo Wenqing, and Chen Naiwen. 1982. 40SuJinren. 1982. 4 1 Su Jinren and Xiao Lianzi. 1981. Chapter I Historical Background: the Tribute System 1. The Ideological Background of the Tribute System The Late Shang Dynasty: An Embryonic Form of Tribute System As early as the Shang period (ca. 1600-1028 B.C.E.), the Chinese were faced with problems of dealing with other peoples. Based on a detailed and systematic study of oracle-bone inscriptions, D. N. Keightley proposes a set of thirty-nine "criteria" from which we can delineate the Shang king's concept of state activity in Late Shang. Keightley's criteria also indicate the nature of the relationships between the royal group at the core of state power on the one hand, and the various groups related to it: Shang allies, dependencies or member-states, and adversaries. Together, they formed the Shang world.1 The Zhou people, who were later to establish the Zhou dynasty (ca. 1027-256 B.C.E.), were, at least at one period, a member-state of the Shang. They received orders from the king and supported the king's affairs, sufficient reason to hold them to have been a member of the greater Shang state. One must mention, however, that in one inscription Zhou is referred to as Zhou-fang, fang ^ > being a designation that applies to non-Shang groups or lands. Nevertheless, weighing all the evidence concerning Zhou in the Shang divinations Keightley comes to the conclusion "that... the Zhou were one of the more distant groups that formed part of the Shang state and that Shang control over the Zhou was neither strong nor continual."2 This is a conclusion that one can accept Amongst the other groups interacting with Shang, the adversaries, in general, refer to the non-Shang people occupying those places whose names had the postfix 1 Keightley, D. N. 1979-80; 1983. 2 Keightley, D. N. 1983, pp. 529-531. 21 fang. In most cases the Shang king did not give orders to them, or ally with them, nor did Shang receive payments from them.3 One of the major activities organized by the Shang king was the military campaigns to attack those peoples, who were then often enslaved or killed as sacrifice after being captured.4 In the Shang relations with its world, the way of regulation implied an embryonic form of tribute system. The king organized campaigns, performed rituals and displayed his power by frequent travel, hunting and inspecting along the pathways of his realm. The dependents came to have audiences with the king, they supported the king's affairs, joined campaigns and sent in tribute including turtle shells or scapulas, horses, dogs, bovids and captives.5 The Shijing says: anciently there was Tang>^ > the Achiever; all from those Di and Qiang there were none who dared not come and bring offerings; there were none who dared not come to audience; Shang will have them forever.6 The inscriptions do not provide evidence to show that the Shang thought in terms of specific territorial units. "The polity seems to have been conceived in terms of personal power (who was in control) and kinship association (what relationship he had to the center) rather than land area (where he was in control)."7 But Late Shang rulers developed a sense of distinction between the inner and outer domains, a development which was to have an important impact on the later Chinese. The Shangshu says: In the exterior domains, the leaders of the hou 1%, dian^J , nanfi and wei^ states, and in the interior domain (i. e. the royal domain proper), all the functionaries and the many governors, the next-following officials, the 3 Keightley, D. N. 1979-80, p. 28. 4 Chen Mengjia. 1956, pp. 279-282. 5 Chen Mengjia. 1956, p. 639; Keightley, D. N. 1979-80, pp. 32-33. 6 Shijing, "Shangsong, Yinwu," p. 266, section 2. 7 Keightley, D. N. 1979-80, p. 26. managers of affairs, the honored officers and all those [members of] the many noble clans who resided in the village...8 There is not enough evidence to establish definitely a clear picture of the relationships at play amongst the different groups in the Shang world. There are differing opinions about who are to be included in the exterior domain: the hou and dian; or the hou, diem and nan; or the hou, dian, nan and wei.9 Whether the titles of the hou, dian etc. were conferred by Shang or were native to the leaders also remains an open question.10 One opinion holds that unlike in the future Zhou feudal system, those lords in the exterior domain possessed their own land and people, and were not enfeoffed by Shang.11 The "Feudal System" (fengjian A Hierarchical Social Order After the Zhou conquest of Late Shang, a new dynasty (ca. 1027-256 B.C.E.) was set up. The early Zhou rulers attempted to consolidate their power in several ways. They devised a political theory or doctrine based on their claim to possess the "Mandate of Heaven" (tianming ), a doctrine that was used to justify the Zhou take-over of the Shang, and the succession of Zhou rulers that followed. According to this doctrine the king, as the "Son of Heaven" (tianzi ), was conceived of as having received the Mandate of Heaven. He was in possession of all the land under heaven and was to rule "all-under-heaven" (tianxia Kj]^ )• It was he who bore the great responsibility of looking after the welfare of the people. Were he to fail in his responsibilities, he would lose the Mandate, thus the dynastic power. This 8 Shangshu , "Jiugao," p. 45, section 10. 9 Wang Guanying. 1984, pp. 80-81. 1 0 For example, different opinions are mentioned in Keighdey, D. N. 1979-80, p. 28; Yang Shengnan. 1983, pp. 128-172; Wang Guanying. 1984, p. 87. 1 1 Chen Mengjia. 1956, pp. 331-332; Wang Guanying. 1984, pp. 86-90. theory appears to have produced a psychological change on the part of the conquered Shang people which led them more easily to accept Zhou rule. At the same time the theory acted as a central cohesive force strengthening the loyalty of the Zhou vassals and officials and binding the entire people under Zhou authority.12 Based on their already existing clan structure, the Zhou established a.fengjian system, often translated as "feudal" system, though one should not understand, thereby, a feudal system identical with the Western model. It was "a system of government in which a ruler personally delegated limited sovereignty over portions of his territory to vassals."13 This system was not to divide up the Zhou realm into different kingdoms, but was to arrange feudal lords in the strategic regions with the aim of constituting "a fence and screen for Zhou."14 The Son of Heaven invested princes, members of the Zhou household, relatives linked by blood or marriage, and former military and civil officials with land and legitimacy. The conquered Shang people were also incorporated into the Zhou world as members in this system. In this fashion the Zhou king was able to centralize his political power as the head of kindred liege lords. The feudal lords within the system were obliged to carry out military activities ordered by the monarch, and to provide revenue to the king. Most important of all, their presence was intended as a display of Zhou power and as a force to contribute to the stability of Zhou authority. In order to ensure order in the feudal system and the unity of the Zhou house, a set of institutions, systems, stipulations, and ceremonial rules - "the rules of propriety" (Zhou — were worked out. Propriety consisted in a set of ceremonial rituals designed to pay respect, and to offer sacrifice to, deities and ancestors. They were, in essence, religious rituals, but with a political dimension. The concept lying behind the 1 2 Creel, H. G. 1970, pp. 81-100. 1 3 Creel H. G. 1970, p. 320. 1 4 Zuozhuan (Xi 24, quoting Fuchen), 189:: 192. ceremonial rituals was the concept of kindred, or extended family, based on which the fengjian system was established and functioned. The authority of the Zhou king, as the head of the clan, thus came to be combined with political power.15 These rules of propriety served as the foundation for the hierarchical structure of the Zhou realm, a structure which governed the relations of the extended family and regulated the relations between rulers and nobles. Within this structure, everyone, in his or her position, was to act in accordance with the rules of propriety so that harmony could be reached.16 Competition for Supremacy During the Chunqiu period (722-481 B.C.E.) Zhou feudal institutions disintegrated as a result of the increasing power of the feudal states on the one hand, and the decay in the political and military might of the Zhou monarchy on the other. The authority of the Zhou monarch declined and with it the ability to enforce rules. Zhou rules of propriety no longer served to bind all the states in a unity of peace and harmony. Some modern scholars hold that during the Chunqiu and Zhanguo (480-222 B.C.E.) periods the Chinese feudal states were independent of the Zhou king and dealt with each other as sovereign equals regardless of their size. Even treaties made with "barbarians" outside the Zhou feudal system were negotiated on a footing of equality since the treaties allowed these "barbarians" to retain their independence.17 The actual situation, however, was far more complex. While it is true that forms of interaction between states were as if among sovereign equals, equal in the sense that one state could not exert its power over others, in fact the predominant 1 5 Wang Guanying. 1987, pp. 75-83. 1 6 This ideal of harmony is stated in the Zuozhuan (Zhao 26, quoting Yanzi), 715::718-719. 1 7 Hong Junpei. 1971, pp. 26-35; Walker, R. L. 1953, p. 25. attitude was never egalitarian. In the Chunqiu period the Chinese states still attempted, in practice, to maintain the old hierarchical order with the Zhou king as the highest. He was represented or supported by the strongest of the feudal lords, who, as "hegemon" (fcapj), retained the power to assemble other feudal lords in a "treaty league" or "covenant" (meng^J). The dominant-subordinate form of relationship between the Zhou king and his feudal lords was still reflected in such a covenant Small states were to pay respect and to serve the larger ones, who in turn would show benevolence to the smaller ones. The ba acted for the king and was in charge of collecting tribute for the royal house.18 The member states in the covenant were required to come to the court of the ba to pay respect regularly, just as they had been required to come to the Zhou court. Failing to do so they would be punished.19 By the time of the Zhanguo period, as a result of wars, cunning diplomacy, and continuous annexations, only seven major powers survived. Small states had either been "swallowed up" by bigger ones or become dependents of them. These seven powers struggled among themselves for supremacy, trying to conquer the whole of the known world by force of arms. With the further decline of the Zhou house it was no longer deemed necessary to lay claim to supporting the Son of Heaven. They all assumed the title of "king" (wang 5.), a title which had formerly belonged solely to the Zhou Son of Heaven. It was a sign of the more equal basis on which they dealt with each other. In 288 B.C.E. the kings of the states of Qin and Qi even called themselves "emperor" (di ^ ), of the west and of the east, respectively, attempting to divide China into two spheres of power. They soon abandoned this plan, however, because the time was clearly not ripe for such a revolutionary step.20 Nevertheless 1 8 Zuozhuan (Xi 7), 148:: 149. 1 9 Zuozhuan (Zhuang 17), 96::96; (Wen 1), 228::229. 2 0 SJ, v46, p. 1898-99; Bodde, D. 1938, p. 128. there was even a suggestion, some two years later, that the king of Yan be made Northern Emperor, the king of Qin be Western Emperor, and the king of Zhao be Central Emperor.21 In spite of this new situation the idea of the absolute superiority of the Son of Heaven had become so deeply rooted that the concept of equality among sovereign states simply could not find much ground. The Chinese clung to the belief of a hierarchical order to their world and fought for supremacy. Among the various schools of political ideas, the Confucians, seeing how the wars for power caused people to suffer, were the most wholehearted in their insistence on the need to look back to the founders of Zhou to provide a model for the ideal pattern of relationships between the Son of Heaven and his subjects, both Chinese and non-Chinese. As a justification for their hierarchical social order, the Confucians developed a cosmological model, stressing that the order of society should be in accordance with the order of nature. Just as Heaven governs the universe, they argued, so the Son of Heaven rules the world. Confucius himself said: He who exercises government by means of his virtue may be compared to the north polar star. It keeps its place and all the stars turn towards it.22 And as everything in the universe has its own place or path of movement, as heaven is high above and the earth is down below, every person in the society has his or her own place, with the ruler high and subjects low. 2 1 5/, v69, p. 2270; Bodde, D. 1938, p. 128. 22Lunyu, "Weizheng," p. 12, ch. I. The Tribute System in the Confucian Works Drawing on the Thoufengjian system, Confucian scholars formulated detailed rules and regulations which supposedly existed in the early Zhou within a framework of a tribute system, adjusting relations between the Son of Heaven and his subjects including the non-Chinese. The practice under the tribute system can be summarized as follows. 1) The Son of Heaven enfeoffed feudal lords with titles and land. The feudal lords were required to send a minor mission to the court every year, and every three years a greater mission. Once in five years they had to appear there in person. Every five years the Son of Heaven made a tour of inspection of the fiefs.23 Visits made by the feudal lords to the Son of Heaven were called chao in spring, zong ^  in summer, jin^^xn autumn, and yu iltfv^m winter.24 2) In the Zhouli one finds the names of officers in charge of the work and rituals related to the tribute system. There are various protocols and rules to regulate and coordinate relations a) between the Son of Heaven and the feudal lords, and b) among the feudal lords themselves. The officers determined the duties and tribute due to the Son of Heaven from the feudal lords,25 and provided for the proper reception of visiting feudal lords, foreign rulers, and their envoys.26 3) The "Da sima " and "Zhifang s h i | ^ ^ l ^ " sections of the Zhouli record a nine-zone system (see FIG. 1): outside of the "capital" (wangji X i ^ ) with 1,000 li on each side, there were nine zones. Each was 500 li in each direction and each was outside the other in concentric circles, as it were. The houfu 1 ^ S « , w a s outside the capital, then the dianfu &JRJ^ , the nanfu $>%$^ , the 2 3 Liji, "Wangzhi," BK. Ill, p. 216, No. 12; No. 13. 2 4 Zhouli Zhushu, "Da zongbo," p. 652. 2 5 For details see the sections of "Da zai," "Da zongbo," and "Da sima" in the Zhouli Zhushu. 2 6 For details see the sections of "Huaifang shi," "Da xingren," "Xiao xingren," and "Xiangxu" in the Zhouli Zhushu. caifu , the weifu fifing , the man/a'f^ , the yifu^jR^ , the zhenfu, and the fanfu |§^Is^. 2 7 In the "Da xingren X^vT-^ " section one finds another design comprising six zones: houfu, dianfu, nanfu, caifu, weifu, and yaofuall of which were inside the "nine regions" (jiuzhou /L^l'j), as opposed to the "barrier kingdoms or beyond" (fanguo<^\%$ ). The feudal lords in the houfu were to come once a year to pay tribute, those in the dianfu, once every two years, those in the nanfu, once every three years, those in the caifu, once every four years, the weifu, once every five years, and the yaofu, once every six years. The lords of fanguo were to come once in a generation.28 The Guoyu and the Xunzi have a description of a five-zone system: 1) the "royal domain" (dianfu), 2) the houfu; which included the Chinese states established by the king, 3) the binfu ^R^or suifu&JZft^, which included the Chinese states conquered by the reigning dynasty, 4) the yaofu, which included Man and Yi, who were subject to Chinese control, and 5) the huangfu, which included Rong and Di who were non-Chinese and basically their own masters. Tribute and offerings from those zones were graded according to their distance to the Zhou capital.29 The Shangshu, "Yugong " attributes the five-zone system to a time as early as Yu-$§, 3 0 the legendary tamer of the flood who founded the Xia dynasty. The five or nine-zone system reflects idealization, imagination or even fabrication rather than actual reality. Yu Ying-shih dismisses the nine-zone system as largely fictitious but nevertheless holds that the five-zone theory has the support of 2 7 Zhouli Zhushu, pp. 1047-49; pp. 1193-96. The "Da sima" uses the term guoji instead of wangji, and ji instead of fu. The meanings of the names of the zones are disputed and obscure, and therefore translations are not provided here. For translations, cf. Biot, E. 1969, and Karlgren, B. trans. Shangshu "Yugong." 2 8 Zhouli Zhushu, pp. 1339-41. 2 9 Guoyu, "Zhouyu," p. 2; Xunzi Jijie, "Zhenglun," pp. 219-220; Yu Ying-shih. 1986, pp. 379-380. 3 0 Shangshu , "Yugong," p. 18, sections 33-39. historical realities, holding it to have been created based on a three-zone structure: namely, the royal domain, the lords' zone and the controlled zone. These three zones did exist during an early period in Chinese history, and the formation of the five-zone system, Yii argues, was under the influence of "Five Phases" (wuxing ) thought.31 If one considers the zone theory to embody the deeply held Chinese belief in an innate hierarchical and concentric structure governing correct political and strategic arrangement, one can see that the three-zone structure did provide at least a conceptual framework for China's external politics. But it is not clear why Yii would dismiss the nine-zone and accept the five-zone model. Both imply basically the same idea and both are idealizations of the actual reality of three zones. Chinese versus "Barbarian": Sinocentrism Although with the idea that "the king leaves nothing and nobody outside his realm" (wangzhe wuwai i^&Jtf^ ),32 there also developed a pervasive idea of sinocentrism and mistrust of foreigners, that is, of "having the various states of Xia within, and keeping the Yi and Di barbarians out" (nei zhu-Xia er wai Yi-Di t% Though the characters Hua lp and Xia j ^ , in the sense of "Chinese," do not occur in any of the Western Zhou bronze inscriptions, nor in the original text of the Yijng, Xia in this sense does occur twice in genuinely ancient parts of the Shujing and twice in the Shijing }A The Western Zhou thus did have a clear sense of community. They identified themselves with the Xia people, a people who, according to the Chinese 3 1 Yii Ying-shih. 1986, pp. 380-381. 3 2 Gongyangzhuan, Yin 1/6; Huan 8/6; Xi 24/4; Cheng 12/1. 3 3 Gongyangzhuan, Cheng 15/12. 3 4 Creel, H. G. 1970, p. 196. tradition, established the first dynasty.35 According to one theory the Xia and Zhou people originated from the people of Rong , and the Shang from the people of Yi. 3 6 As a state member of the Late Shang, Zhou was familiar with Shang culture, and had inherited some Shang political ideas and institutions. Through thefengjian system Xia and Shang descendants, as well as the Zhou people, became more closely connected, culturally and economically. Gradually they merged into a "new" group of people who differentiated themselves from other Rong or Yi people. This group of people was referred to as Hua-Xia people. Hua in the sense of "Chinese" first occurs in the Zuozhuan}"7 In the Zhou period by "Chinese" one means these Hua-Xia people. As for the non-Hua-Xia people or non-Chinese, the Chinese did not have a single term for them. Instead several words were used, the most frequent and general ones being Man f^^ , Yi, Rong, and Di. In the Shang oracle-bone inscriptions we find the words Yi and Rong. Man and Di first appear later in bronze inscriptions. These non-Chinese lived amongst, and in areas surrounding, the Chinese states. Many wars took place between them. The Chinese looked down upon these people in the same way as the Hellenes looked down on the barbaroi. The word "barbarian" is therefore often used in English works referring to these Man, Yi, Rong, and Di people. It is used here, not in a normative sense, but because it has gained wide currency and probably adequately conveys the attitude of the Chinese of that time. Several criteria were used by the Hua-Xia people to distinguish themselves from the "barbarians." 1. Language. The languages of non-Hua-Xia people were different from Chinese, although there are differing opinions about whether they were linguistically non-Chinese or linguistically closely related to Chinese. Pulleyblank holds the opinion 3 5 Shangshu, "Kanggao," p. 39, section 4; "Junshi," p. 61, section 12. 3 6 Cen Zhongmian. 1957 (1), pp. 117-118; Pulleyblank, E. G. 1983, pp. 420-421. 3 7 Zuozhuan (Xiang 14), 460::464. that while the Hua-Xia people had the same written and spoken language — Chinese, possibly the language of the Xia which was later accepted by the Shang and Zhou people, the non-Hua-Xia people used non-Chinese languages. This is shown by the fact that, though most of these people have eventually been sinicized, their remnants are still sizable even today, inhabiting, as minority ethnic groups with their own languages, the upland regions of southern, southwestern and western China.38 2. Material culture: clothing, food, style of hair, etc. To the Chinese, differences in clothing and hair style were not just matter of external appearances, but revealed differences in culture and the level of civilization. The Liji records: Where the statutory measures and the [fashion of] clothes had been changed, it was held to be rebellion, and the disobedient ruler was banished.39 In 307 B.C.E. the king of the State of Zhao made a decision to adopt Hu (a nomadic non-Chinese people) dress which was convenient for horse-riding. He was strongly opposed in this by his uncle and officials who insisted that any change in Chinese clothing to that of other people would mean changing the old doctrine and the "Way" (dao^J, going against the will of the people, against the learned, and departing from the central states which referred to the Chinese states.40 3. The practice of the rules of Zhou propriety. Zhou rules of propriety embodied the myths underpinning the institutions of the Chinese states, their social organization and behavioral norms. They were not followed by the non-Chinese and, according to the traditional explanation, were not required of them since they did not qualify as civilized people. In the Chunqiu period, for example, when the State of Chu referred to itself as Man-Yi and called the ruler "king" (wang)41 and when the States of 3 8 Pulleyblank, E. G. 1983, pp. 411-412. 3 9 Liji, "Wangzhi," BK. HI, p. 217, No. 15. 4 0 SJ, v43, pp. 1806-08. 4 1 SJ, v40, p. 1692. Wu and Yue also took the title wang, the Chinese did not try to punish them for this since they were "barbarians" outside the Zhou realm. To the Chinese their culture represented a higher level of civilization. It was, simply put, superior. It was not, on the other hand, confined to a certain group of people. The records from the Chunqiu and Zhanguo periods show that many tribes interspersed among the Chinese states became sinicized, mainly as the result of conquest by strong states. Seeing this reality, both Confucius and Mencius attempted to justify Chinese expansion with the argument that Chinese culture or the rules of propriety could convert other peoples to a more advanced civilization. Yet, it is clear that before they were transformed into, and accepted by, the Chinese these "barbarians" were considered as outsiders and inferior. These attitudes were not simply the result of Chinese feelings of superiority, but arose in response to continued "barbarian" attacks on both the Zhou monarchy and the Chinese states. The "barbarian" presence was so threatening that after the Western Zhou capital, Hao , was invaded in 771 B.C.E. by a Rong people living quite close to Hao, and king You was killed in the turmoil, the succeeding king Ping had to move the capital east to L u o ^ S , near modern Luoyang in Henan >%$D province. 4 2 Hostility towards "barbarians," therefore, prevailed among the Chinese, and harsh militant action was justified by excluding "barbarians" as outsiders or lower beings of whom virtue and propriety could not be expected. This sinocentrism expressed itself clearly in political and geographical concepts, as well as in a deeper psychological attitude. In their world the Chinese considered themselves the center. The word "Middle Kingdoms" or "Central States" (Zhongguo y* |§ | ), was used to convey such an idea. In the bronze inscription dated the fifth year of King Cheng of the Zhou dynasty (ca. 1020) that word is first found referring to 4 2 SJ, v4, p. 149. the royal domain.43 In the Shijing, Zhongguo also refers to the royal domain. In the Zuozhuan it refers to the Chinese states as opposed to "barbarians." The word "Middle Land" or "Middle Plain" (Zhongyuan ^J^.), is also used in the Zuozhuan to refer to the Zhou realm.44 During the time of Chunqiu and Zhanguo, when the "barbarians," who had formerly been interspersed among the Chinese states gradually became sinicized or moved to the periphery, the idea that the Chinese were in the center surrounded by the non-Chinese seems to have gained more ground. The Mozi locates these non-Chinese as Di in the north, Rong in the west and Yi in the east The Liji adds the Man in the south.45 In the Zhouli we find the belief that Luoyang, the Eastern Zhou capital, is the center of the earth. It records that the "great director of the multitudes" (da situ 7^ ^ f ^ _ ) , using the gnomon shadow template, determined the distance of the earth below the sun, fixed the exact length of the sun's shadow, and thus found the center of the earth. The center of the earth was the place where the sun's shadow at the summer solstice was 1.5 Chinese feet (chi J^J.46 The observation of shadow lengths was used to determine latitude in fixing provincial and other territorial boundaries, and the exact center was said to be in Yangcheng f%?~$K, close to Luoyang.47 The five or nine-zone system, fictitious as it was, demonstrates the Confucian conception of a world arranged in concentric circles, or squares, with the Son of Heaven in the center. Though without much geographical significance, as discussed earlier, the idea of a unified Chinese empire with the Chinese in the center and the non-Chinese constituting a defence line around it, was an extremely potent myth providing, amongst other things, an explanation and rationale for the tribute system. 4 3 Tang Lan. 1976, p. 60. 4 4 Zuozhuan (Xi 23, quoting Chonger), 185::187. 4 5 Mozi Xiangu, "Jiezang," pp. 112-113; Liji, "Wangzhi," BK. Ill, p. 229, No. 14. 4 6 Zhouli Zhushu, pp. 361-364; Needham, J. 1959, p. 286. 4 7 Zhouli Zhushu, p. 363. Scholars and philosophers of some non-Confucian schools held different ideas about society and the world. The Agricultural School (nong jia^^ ) believed that rulers of worth should cultivate the land, eat what they produced, and prepare their own meals while carrying on the affairs of government. Followers of this school even saw no need for Sage Kings. Asserting that both ruler and subject should plough together in the fields, they overthrew the order of upper and lower classes.48 Zou Yan J^fy^, the leading thinker of the School of Yin and Yangp|? and of Five Phases, did not agree with the sinocentric idea. He maintained that there were "nine large continents" (da jiuzhou y^/U^rj ) in the world and each was divided into nine regions. What scholars called the "Middle Kingdom" (Zhongguo) was held to be but one part in eighty-one in the whole world.49 Hui Shi x^ffis, the leading thinker of the School of the Dialecticians (ming jiaj^^ ), also talked about the center of the world, which, according to one interpretation, might suggest the idea that the earth is spherical, and according to another interpretation, may imply that there were vast regions beyond the bounds of contemporary geographical knowledge.50 In any case these ideas challenged sinocentrism. Zou Yan's ideas, however, had different influence on later Chinese, especially the Han, as will be discussed later. 2. The Han Dynasty (202 B.C.E.-220 C.E.): Development of the Tribute System The Heqin%*$HL Policy in the Early Period After the unification of Qin (221-207 B.C.E.), the First Emperor (Shi huangdi •£*i$f? , re. 221-210 B.C.E.) lost no time in expanding his territory by a 4 8 Fung Yu-lan. 1952, p. 144. 4 9 SJ, v74, p. 2344; Fung Yu-lan. 1952, pp. 160-161. 5 0 Needham, J. 1956, p. 192. series of military expeditions. Through his success in driving the northern nomads, the Xiongnu , beyond the Great Wall, and in bringing the surrounding areas under Chinese control, it seemed that the rule of Son of Heaven was indeed extended to all-under-heaven. However, the tribute system, which was an ideal order to govern Chinese relations with foreign peoples, did not assert itself spontaneously as a result of this success. There were many hindrances to its full implementation and it was some time before it was accepted as a normal state of affairs. The frontier was unstable at the beginning of Han. The relations between the Chinese and their neighbors had undergone a drastic change from that of the Qin dynasty with the balance of power shifting toward the non-Chinese. The Chinese had lost control over the surrounding areas which they had established during the Qin period. Moreover the Xiongnu had grown into a formidable nomadic empire in the steppes north of the Great Wall. Under the rule of Modun ^fjk^ > they "reached their peak of strength and size, subjugating all of the other barbarian tribes of the north and turning south to confront China as an enemy nation."51 The Xiongnu, then, challenged Chinese superiority and authority as an equally great power. They made incursions over the Han border, and provided a refuge for dissident Chinese. In 200 B.C.E. the Xiongnu invaded Taiyuan • Emperor Gao (re. 202-195 B.C.E.) led troops in the battle, but fell into a trap at Pingcheng 2$£(near modern Datong fiJZ] in Shanxi il\ rirj? ). 5 2 The Xiongnu borrowed the Chinese political model and the concept of Son of Heaven. They named their chieftains "chanyui^^ ," short for "chengli gutu ^%zjL$k?JL chanyu," meaning "great one, Son of Heaven,"53 and insisted on dealing with China as an equal. In his letter to Emperor Wen (re. 179-157), Modun 5 1 SJ, vllO, p. 2890; Watson, B. 1961, p. 163; HS, v94A, p. 3751. 5 2 SJ, v8, pp. 384-385; vllO, p. 2894; HS, vlB, p. 63; v94A, pp. 3753-54. 5 3 SJ, vllO, pp. 2887-88; HS, v94A, p. 3751. Chanyu referred to himself as "the great chanyu whom Heaven has set up," "the great chanyu of the Xiongnu, born of heaven and earth and ordained by the sun and moon."54 Faced with this threat, Han adopted a policy of appeasement, and concluded peace treaties with the Xiongnu. The original items of the treaties were: 1) A Chinese princess was to be married to the chanyu, 2) Chinese annual payments of silk, liquor, rice and other kinds of food were to be made to the Xiongnu, 3) the Han and Xiongnu were to be "brotherly" states, and 4) the Great Wall was to be the border between the Han and Xiongnu.55 This was the famous "marriage alliance" (heqin agreement), although the giving of a Chinese princess in marriage was usually not its only feature. As one can see from the agreement, it was the Chinese, in fact, who paid tribute to the nomadic Xiongnu empire. From Emperor Gao to the beginning of Emperor Wu's reign (140-87 B.C.E.), heqin continued to be the main policy in Chinese contacts with the Xiongnu. Under the heqin agreement, it seemed that both Han and the Xiongnu hoped for a peaceful coexistence. In 176 B.C.E. the chanyu sent a letter to Han, expressing his wish to stop wars in order that the border peoples might enjoy peace and comfort, generation after generation.56 Emperor Wen supported this aim in his letter to the chanyu in 162 B.C.E., hoping that the peoples of the two states might be joined together "like the sons of a single family."57 The heqin policy, however, was never enough to guarantee peace in their relations, in which the dominant factors were always self-interest and balance of power. The general attitude of both sides towards each other was one of suspicion and 5 4 SJ, vllO, p. 2896; p. 2899; Watson, B. 1961, p. 167; p. 171; HS, v94A, p. 3756; p. 3760. 5 5 5/, vllO, p. 2895; p. 2902; HS, v94A, p. 3754; p. 3762; Yu Ying-shih. 1967, pp. 41-42. 5 6 SJ, vllO, p. 2896; Watson, B. 1961, p. 168; HS, v94A, pp. 3756-57. 5 7 SJ, vllO, p. 2903; Watson, B. 1961, p. 174; IIS, v94A, p. 3763. hostility. During Empress Lii's reign (187-180 B.C.E.), upon receipt of an insulting letter from the Xiongnu, she wanted to launch a military campaign but had to give up the idea because Han could not afford it Several years later, in Emperor Wen's reign, the court discussed whether it was feasible to attack the Xiongnu or whether to make peace. Again, because of Han's relative weakness, Han had to decide on keeping peace.58 The heqin policy was not only considered to be humiliating but also was a drain on the Chinese economy. Yet it did not prevent the Xiongnu from launching invasions, so as a policy it was severely criticized by a section of Chinese officials. The famous Han thinker Jia Yi ^ Iffi » framing on Confucian theories of a hierarchical and sinocentric world order and Chinese superiority, expressed the view that the Son of Heaven was the head of the empire who possessed the power to command the "barbarian." The "barbarians," he said, were the "feet" of the empire whose duty was to present tribute to the Son of Heaven but each year Han was providing the Xiongnu with money, silk floss and fabrics. It was as if the feet were on the top and the head at the bottom. Hanging upside down, as it were, like this, was something beyond comprehension.59 Ban Gu T^iJiJ also, in his Hanshu, commented that the heqin practice was a useless policy that cost a thousand jin Yf of gold every year and put the Chinese in the inferior position of serving the Xiongnu.60 Han Expansion Although the heqin policy did not prove to be very successful in preventing the Xiongnu from incursions, it did win time for Han to recover and to build up its 5 8 SJ, vllO, pp. 2895-96; HS, v94A, pp. 3754-55; p. 3757. 5 9 HS, v48, p. 2240-42; Yii Ying-shih. 1967, p. 11. 6 0 HS, v94B, pp. 3830-31. economic and military strength. By Emperor Wu's time Han enjoyed great prosperity, which enabled the emperor to engage in expansion on a grand scale. Military campaigns went on successfully in the south, southwest and east, resulting in a wide extension of Han political influence. After an initial brief period of peace with the Xiongnu, Emperor Wu started an offensive. As a part of the strategy to conquer his rival he sent Zhang Qian to the Western Regions (xiyu \^tt^ , Central Asia) in order to seek an alliance with the Yuezhi \=\ , 6 1 a former enemy of the Xiongnu. Though the mission did not realize its immediate goal, Zhang Qian's expedition discovered the Silk Road, the trade route across the Asian continent Finally Han secured a marriage alliance with the Wusun s^^i in Central Asia, also an enemy of the Xiongnu. Now the heqin policy was again adopted, but this time as a positive means and for a different purpose, namely to obtain and retain military and political assistance from the "barbarians." In 72 B.C.E. the Wusun despatched 50,000 cavalry in a joint force with Han, inflicting a heavy blow on the Xiongnu.62 During the prolonged wars between Han and the Xiongnu in Emperor Wu's reign, the Xiongnu were seriously weakened but Han also suffered great losses. When the Xiongnu asked for a peace treaty under the terms of a heqin arrangement, Han twice, around 119 B.C.E. and in 111 B.C.E., attempted to force them to submit and to acknowledge themselves Han subjects. Each time the Xiongnu resolutely refused, insisting on the marriage alliance, which implied equality.63 Not until the Emperor Xuan's reign (73-49 B.C.E.), when the Xiongnu were further weakened by internal struggles for the succession, and when Huhanxie Chanyu of the 6 1 For location, if it is not given in brackets see the list following Table 1.1. 6 2 HS, v94A, p. 3785; v96B, pp. 3903-05. 6 3 SJ, vllO, pp. 2911-12; HS, v94A, pp. 3771-72. Southern Xiongnu arrived in 51 B.C.E. in the Han court, did Han succeed in making the Xiongnu accept the status of a Han subject The "discovery of the West" by Zhang Qian and the consequent expansion of Han power enlarged Chinese contacts with other countries to cover almost all of Asia. Knowledge of foreign lands, via the transcontinental Silk Road and maritime trade route, broadened the Chinese view of the world. But this broadening of their horizons did not have much impact on the belief that China was the center to which other people should come, and that Chinese civilization was superior to all others. The expansion resulted, rather, in an extension of Chinese political influence and the further development of the tribute system, consolidating more the Chinese belief in their superiority. After Zhang Qian's mission to the Wusun in 119 B.C.E., states along the Silk Road as far distant as Anxi T^J|7 n (Parthia) are recorded to have despatched envoys and paid tribute to the Han court It seemed, at the time, that at last China could realize its hope of a full and unhindered tribute system. However, of the many missions which came from the Xiongnu during the period from Emperor Gao to Emperor Zhao (re. 86-74 B.C.E.), only one, in 100 B.C.E., is recorded in the Hanshu as having "sent a messenger to bring tribute."64 This was after the Han victory over Dayuan Xjf£i (Sogdiana in Han), and when a new chanyu had just come to the throne.65 It was not until some fifty years later, when the Xiongnu chanyu came personally to the Han court in 51 B.C.E. that the full development of the tribute system can be said to have finally become a fixture of the Chinese political landscape. Yu Ying-shih holds that, as Zou Yan's theory that China (Zhongguo) was just one part of a huge world gained currency in Han, and as the classical identification of China with "all-under-heaven" gradually gave way to a more realistic idea, the Han Chinese considered China as that which lay "within the seas" (hainei ). And 6 4 HS, v6, p. 202; Dubs, H. H. 1938-55, II., p. 103. 65HS, v94A, p. 3777. their sinocentrism was not a literal description of geography, but had politico-cultural overtones.66 It is true that the Han Chinese did have a more extensive knowledge of the world than the Chinese of earlier times, but Yii's argument neglects some important facts. First, the term "all-under-heaven" (tianxia), in traditional usage and as well as in Zou Yan's theory, included not only the "Middle Kingdom(s)" but also areas inhabited by non-Chinese. But to the Chinese in those days, China was identified with Zhongguo, not all-under-heaven. The Liji commentary says: All-under-heaven means extending outside to the four seas.67 The Erya says: [The world mcluding] the nine Yi , eight Di, seven Rong and six Man are called four seas.68 Second, the word "within the four seas" (hainei), is also an archaic term with the same meaning as "all-under-heaven," to refer to the Chinese known world. The Zhanguo Ce says: Today, if you want to incorporate all-under-heaven, overawe countries with ten thousand chariots, reduce enemy countries to submission, control all within the four seas and make all men your children and [turn] feudal lords [into] your vassals, it will not work without arms.69 The Xunzi also says: 6 6 Yii Ying-shi. 1986, pp. 377-379. 6 7 Liji Zhengyi, "Quli," p. 180. 6 8 Erya Zhushu, "Shidi," p. 270. 6 9 Zhanguo Ce, v3 "Qin One," p. 17. Crump, J. I. (1970, pp. 56-57) incorrectly translates the word tianxia as "great states." To comprehend the essentials of all-under-heaven and govern the people within the four seas.70 Third, as reflected in the famous discourses in 81 B.C.E. on the salt and iron monopoly policy, Zou Yan's theory was viewed by the Han Chinese in two ways. On one side, the Censor-in-chief (yushi dafwf-if^AJ^ ) Sang Hongyang^ J A ^ - . an advocate of the state monopoly, praised Zou Yan's theory and claimed that it justified the Qin territorial expansion which was aimed at extending its power over to the "nine large continents" (da jiuzhou). This shows that, far from leading those Han Chinese who were most involved in foreign policy to adopt a more modest estimate of the proper role of the Chinese emperor, Zou Yan's theory was simply incorporated into the old idea that the Chinese Son of Heaven had a god given authority over all lands. On the other side, the literary scholars, disillusioned by the Qin and early Han expansions, rejected Zou Yan's theory as absurd and criticized the unbridled Qin ambition which had led to its finally losing even its own country.71 The scholars still insisted on a sinocentric idea in the geographical sense: The frontier 'commanderies' (jun%f) are in the mountains and valleys. Yin and yang are not well balanced. Cold splits the earth. Blasting wind whirls the alkaline soil. Sand and stones are piled up and accumulated. The situation of the land has nothing suitable in it. Zhongguo lies in the middle between heaven and earth. It is where yin and yang come together. The sun and moon run their course across its southern parts and the big dipper and pole star rise from its north. It embraces all the mild and harmonious airs and produces all kinds of products.72 7 0 Xunzi Jijie, "Bugou," p. 30. 7 1 Yan Tie Lun Jiaozhu, "Lun Zou," pp. 331-332. 7 2 Yan Tie Lun Jiaozhu, "Qingzhong," p. 100. The ideas of these traditional scholars were not simplistic. Their sinocentrism was not just the result of a naive and unquestioning belief in the superiority of all things Chinese. Rather they were stressing the importance of Han concentrating on its own affairs, avoiding any action that would purposely attract the "barbarians." On the other hand Sang Hongyang articulated clearly this very real and practical issue which China was faced with all the time: the government simply had to solve or alleviate the problem of "barbarian" invasion either by actively attacking the "barbarians" or by defence in order that China could enjoy stability and peace. Practices under the Tribute System The tribute system, which regulated Chinese relations with non-Chinese, was an extension of the hierarchical feudal system existing within China itself. In Han, princes and nobles were given titles and fiefs as their own states by the Son of Heaven. Ceremonial rituals and various laws were imposed to control them. They were to come to the court regularly to pay homage and tribute.73 Han "commanderies" (jun^f) were also expected to present tribute to the court in addition to regular taxes.74 Practices under the tribute system as they pertained to Chinese relations with non-Chinese were as follows: 1) Rulers of non-Chinese states or their envoys came to the Han court to pay homage. 2) They presented symbolic tribute in the form of local products. 7 3 Zhang Weihua. 1980, pp. 185-244. 7 4 For example, in 196 B.C.E., an edict was issued saying "Henceforce the vassal kings and marquises shall regularly pay court and make offerings [to Us, the Emperor], in the tenth month, and each commandery shall [make an offering] in accordance with the total number of its people; each person per year [shall be taxed] sixty-three 'cash' (qian) in order to provide for the expense of making offerings [to the Emperor]." See HS, vl, p. 70; Dubs, H. H. 1938-55,1., p. 130. In 37 C.E., an edict repeated an earlier order prohibiting the commanderies and kingdoms from offering unusual foodstuffs, but allowing the presentation of foodstuffs to the ancestral temples as before. HHS, vl, pp. 60-61. 3) The Chinese emperor gave them imperial gifts in return. 4) The emperor conferred on the foreign rulers titles of nobility — "king" (wang), "marquis" (houf^), "lord" (junj& ), and "chieftain" (zhangjj^), an official title such as general (jiangjun^-^. ), and bestowed seals on them and sometimes on their officials. 5) The foreign rulers sent their sons as hostages in return for Chinese protection from outside aggression. The essential part of the practice consisted of items one, two, and three. Any state, so long as it sent missions with gifts to the court, is recorded as a tributary. The tribute system thus developed into a network embracing all foreign states which paid "official" visits to Han. Compared with the tribute system as it operated inside Han borders, however, there was a distinct difference. The tribute required of the Chinese, according to Schafer and Wallacker, in Tang times, "was a putative gift rather than a tax due. Though required of the local officials, its presentation retained an air of voluntary display of fealty to the throne."75 What the tribute in Han should be is not as clearly indicated in the Han records as in those of Tang or Song, but it had the same function. Failure of the Chinese to present tribute, or failure to pay court-visits at regular intervals would imply an act of disobedience. For the non-Chinese, on the other hand, who were outer subjects, tribute was not compulsory. The Han administration would not have to launch a punitive attack if tribute was not regularly forthcoming though the absence of tribute payment could be, and was, used as a justification for Chinese aggression against non-Chinese people. The policy towards non-Chinese tributaries was based on the maxim of "not interfering in the administration of those who had not been influenced by the rules of proper conduct" This policy was shaped specifically as 7 5 Schafer, E. and Wallacker, B. 1957-58, p. 213. a response to the Xiongnu on the eve of the arrival of Huhanxie Chanyu,76 but it became a general policy towards all non-Chinese in the Han period. To the Chinese rulers, then, the tribute system structured an ideal pattern for their relations with other peoples since the tributary visit implied the acceptance of Chinese rule over a universal empire. Also, because the "acceptance of imperial rule" involved not only different levels of "acceptance," but also different categories of "rule," which varied from one group of people to another,77 different practices were applied to countries according to their distance from China and relative importance. So, although the Han world view was just like that in the concentric five or nine-zone system, Yu Ying-shih concludes that compared with the tribute system in pre-Han period, the institutionalization of tributary practices and their systematic application in the realm of foreign relations was a unique Han contribution.78 Different Types of Relations under the Tribute System The relations between the non-Chinese tributaries and Han saw many ups and downs, but, generally speaking, five types of relationship can be discerned, and for a general view of these areas, see Map l . 7 9 Type One: Non-Chinese, who were linked to Han most closely in terms of political rule, were those in the Han "commandery-distrier" (junxian^f^^), and "dependent state" (shuguoMr Jf^ ) ) systems.80 They were governed by Chinese officials, but in 7 6 HS, v8, p. 270. 7 7 Yii Ying-shih. 1967, p. 189. 7 8 Yii Ying-shih. 1986, p. 381. 7 9 For locations and diplomatic activities, see Map 1 "Han and its neighbors" and Table 1. 1 "Foreign tribute missions to Han." 8 0 Yii Ying-shih, in his book Trade and Expansion in Han China, devotes a whole chapter to the analysis of Han China's general policy towards "barbarians," who having submited were ruled under the junxian and shuguo systems, as well as Han administrate apparatus of the two systems. See his book, ch. 4, pp. 65-91. the accordance with their old customs. Some chiefs, were granted the tides of king, marquis, lord, or chieftain by the Han emperors. Those under junxian administration were considered to be both outer and inner subjects.81 The shuguo system functioned as an adniinistrative device to take care of the "barbarians" who had submitted to Han rule. It may be regarded as a sort of compromise between the tribute system and the junxian system. The Qiang (in modern Qinghai and Tibet) were widely dispersed nomadic tribes with whom Han could not form a comprehensive tributary relationship. Different policies were adopted with different tribes, such as a) resettling them in Han territory under junxian or shuguo system, b) appointing Commandant Protecting the Qiang (hu Qiang jiaowei ^ ^ ^ J ^ ) to conduct relations with them, and c) suppressing them with military force. Type Two: In this category of non-Chinese peoples are Fnyu^tS^ > Gaogoulijl^ (Koguryo), Weimo/Jj^ fl? > a n d Han^L, lands situated to the northeast of the Han territory. Compared with type one and type three peoples, they were not aciministered closely by the Chinese, nor was there as much attention paid to them to ensure border security. Called "Eastern barbarians" (dong Yi ), they were considered by the Chinese to have a gentle nature and to be easy to control with teaching. Although Gaogouli and Weimo were administered under the junxian system, and other Eastern "barbarians" are recorded as asking to be attached to the Chinese commanderies of Lelang^^^^ and Liaodong"^^^ , they were much less closely supervised than the non-Chinese in the southwestern regions, who were under the Chinese commandery system and subject to a level of taxation in Later Han. 8 1 Kurihara. 1980, p. 181. Among the five practices under the tribute system they were not required to send hostages, and no special offices were set up to mediate the contact with them. 8 2 Type Three: O n the northern and northwestern frontiers were situated non-Chinese states which were considered to be outer subjects. They received close attention because of their strategic importance to Han. Han contacts with them brought out all five practices of tribute system into play, 8 3 though sometimes they were treated in a special fashion in accordance with a particular situation. Han established administrative offices to run the military and political affairs which touched on their relations with these peoples, aiming thereby to maintain a "control by loose rein" (jimi^J^). In type three are the following three peoples: 1) The Xiongnu. Chinese treatment of the Xiongnu, who continued to be a major threat, was characterized by the granting of favorable terms, both politically and materially. Before Huhanxie Chanyu came to the court, a conference was held to discuss the protocol to be accorded to him. One opinion held that a chanyu could not be treated as a guest by a universal king: [According to] the rites and ceremonies, it is proper that he should acknowledge himself a subject like the vassal kings... and be ranked next below the vassal kings. 8 4 X i a o Wangzhi J | j§T £ , a H a n official, did not agree, arguing that the Xiongnu were a state of rival status, and therefore their chanyu should not be treated as a subject and should be ranked above the vassal kings. In this way, said he, if later the chanyu showed any disobedience by not coming to court regularly, Han would not have to start any punitive actions. 8 2 Details see HHS , v85, pp. 2810-20. 8 3 See Table 1. 1. 8 4 HS, v8, p. 270. Emperor Xuan adopted Xiao's opinion. An imperial edict was issued saying: [We] have heard that the Five Lords and the three [dynasties of] Kings did not touch in their administration those who had not been influenced by the rules of proper conduct [i.e., the outer barbarians]. Now the Hun chanyu had styled himself [Our] feudatory at the northern frontier and [is coming to] pay court in the first month. We are inadequate and [Our] virtue is unable to cover [the earth] widely. Let [the chanyu] be treated according to the rites for a guest and [let] his rank be above that of the vassal kings. When the chanyu arrived he was announced as a subject, but without the use of his personal name. An imperial seal and many gifts were conferred on him.85 This pragmatic attitude acknowledged that there were limits to Han capability to enforce strict adherence to the tribute system. Yet the Chinese still saved face by appealing to the example of the ancient "Sage Kings" (shengwang %j!j£_ ). Huhanxie Chanyu asked permission from Han to stay north of the Han frontiers as a safeguarding force, to which Han agreed. When he restored his own power under Han protection, however, he returned to the steppes. In 33 B.C.E. a palace woman, Wang Zhaojun .i.E7<?^ s?> was sent to wed Huhanxie, who then offered to guard the Han frontiers from Shanggu JzJif^ to Dunhuang, and asked that Han remove the frontier barriers and troops. But the Chinese refused, obviously because they could not trust the Xiongnu who could only be treated as an outer subject.86 In Later Han the former Southern Xiongnu were split again into southern and northern parts with the Southern Xiongnu having a pro-Han policy. In 49 CE. the chanyu despatched an embassy to the Han court and asked for subject status. He presented expensive gifts, requested a Han commissioner for protection, offered to send hostages and sued for resumption of the former agreement. In the following year 85 HS, v8, pp. 270-271; Dubs, H. H. 1938-55, II., pp. 256-258; HS, v78, pp. 3282-83; v94B, p. 3798. 86 HS, v94B, p. 3798; pp. 3800-05. Han permitted the Southern Xiongnu to move to Yunzhong g^f' commandery and again to Meiji. in Hexi^fifr commandery. Seven other Southern Xiongnu divisions were stationed in the Han frontier commanderies, each ruled by its own hereditary chief to watch for attacks from beyond the borders.87 Han appointed special officials to take charge of relations with the Xiongnu. In Former Han an office was created with the title of Vice Commandant in Charge of the Xiongnu (shi Xiongnu fu jiaowei^itj^i^xfr^^tf^ffi ), and in Later Han Leader of Court Gentlemen in Charge of the Xiongnu (shi Xiongnu zhonglang Jiang O&i^ffytft ) . 8 8 2) Oasis states in the Western Regions. These many agricultural states were situated along the important trade route into Central and South Asia, and since they were potential allies to both Han and the Xiongnu, both sides wanted to control them. Some Chinese officials wanted to draw a clear line between Han and these states with the intention of having nothing to do with them at all. Ban Gu commented that the various states of the Western Regions each had their rulers and their chiefs. They were cut off from Han and the intervening distance was great. If Han took possession of them they brought no profit, and if Han abandoned them they constituted no loss.89 In spite of this attitude and in spite of continuous debate about how Han should treat the Western Regions, many battles were fought for the control over the area. The Chinese could not afford to neglect the strategic importance of the region. From the beginning of Later Han to around 125 C.E., whenever Han lost control over the area and relations with the oasis states lapsed, the Xiongnu took control and used the area as a base for attacking the Chinese frontier. A real danger was that if the Xiongnu 8 7 HHS, v89, pp. 2943-45. 8 8 Bielenstein, H. 1980, pp. 110-112. 8 9 HS, v96B, p. 3930; Hulsewe, A. F. P. 1979, pp. 202-203. occupied the whole area they might enter into an alliance with the Qiang, the hostile "barbarians" living to the south of the Western Regions, thereby posing a great threat to the four commanderies in the Hexi region (in modem Gansu ^ and Qinghai), and that were they to fall it would be immeasurably costly for Han to rescue them. Having realized this, the court sent the general Ban Yong^/i^, Ban Gu's nephew, to the Western Regions. In 127 C.E. following Ban Yong's military conquest of Jushi ^ j f ^ and Yanqi^^ , which were allied to the Northern Xiongnu, seventeen states in the Western Regions came to Han for court-visit.90 Because of the necessity for close and peaceful relations Han had offices set up in the Western Regions, the most important of which was the Protectorate of the Western Regions (Xiyu duhu v&J) established in 60 or 59 B.C.E.9 1 In addition to the regular practices under the tribute system these states were to provide supplies to the Han troops in the Regions and to the envoys passing through, and to despatch their own troops when the Chinese required military assistance. As for Han side of the arrangement, the court conferred titles with official seals on the kings, whose rank was equal to the marquis inside China. Han performed the duty of a suzerain, protecting these states from invasion, settling disputes and punishing offenders.92 3) The Wuhuan*J?il2.(a people at the southern end of the Greater Xing'an Mountains -^?£H$t. , southeast of the Liaojj^ River) and the Xianbei j|^J^. (a people in the area between the Mongolia steppes and east of the Greater Xing'an Mountains). Both of these peoples also occupied an important strategic place in Han frontier thinking. Like the states in the Western Regions they were sometimes subjects of, or allied with, the Xiongnu. But unlike the agricultural oasis states, they 9 0 HHS, v88, pp. 2909-12. 9 1 Bielenstein, H. 1980, pp. 109-113; Yu Ying-shih. 1986, pp. 411-412. 9 2 For further details on Han relations with the Western Regions see Ise SentarO. 1968, pp. 37-89. were nomads, more capable of frequent and quick incursions into Han regions. Compared with the Xiongnu their threat to the Chinese was not severe since they were loosely organized. And since hostilities sometimes existed among the Xiongnu, the Wuhuan and the Xianbei, there were opportunities for the Chinese to use one against the other and thereby create a buffer area between themselves and the Xiongnu. In spite of the fact that conflicts with these two groups of people were frequent, China still tried to structure its relations with both people according to the tribute system. As for the Wuhuan and Xianbei, they too sought support from the Chinese. In Emperor Wu's reign in Former Han, the Chinese resettled the Wuhuan north of the Great Wall, outside five frontier commanderies (Shanggu _K<£s , Yuyang ^ v order that they could act as a sort of watchman over the Xiongnu. This was when the Commandant of the Wuhuan (Wuhuan jiaowei flcjffi) was appointed. The same position was again restored in Later Han when the Wuhuan resumed their peaceful relations with the Chinese after a period of hostility. The responsibilities of this Wuhuan jiaowei included a) supervision of the Xianbei as well as the Wuhuan, b) arrangement of rewards to the Wuhuan and c) administration and maintenance of border markets. Official relations between the Xianbei and Han started in 49 CE. in Later Han. It is recorded that, till 87, they functioned as a defending force on the frontiers without causing any trouble. Following a period of disturbances, some of the Xianbei around 107 resumed peace with Han. Han settled them along Ningcheng ^  t&t, in the commandery of Shanggu, where the Wuhuan jiaowei had his headquarters. A hundred and twenty tribes sent their hostages to Han.93 Youbeiping ^^h^ 4 9 3 / /HS , v90, pp. 2981-86. 51 Even though the states in this category were brought into the tribute system, and considered by the Chinese to be outer subjects of the Han Son of Heaven, in practical terms they were neither subjugated, nor, as in the case of the Xiongnu, did they ever really submit94 Rebellions against Han occurred from time to time. They maintained their independent status and Han, limited in its resources and unable to exert a strict and direct rule either politically or militarily, was powerless to change this state of affair. During his reign Wang Mang JL.^ (9-23 CE.) had attempted to make these non-Chinese peoples direct or "inner" subjects. He insisted that, "Heaven has not two suns, nor has Earth two kings" — this is the unchangeable way of all the kings. Some of the nobles of the Han clan were entitled Kings, and even the barbarians [beyond] the four [frontiers] followed [this practice]. It is contrary to the ancient institutions and absurd [in view of the principle that there is only] one sovereign [in the world].95 Generals were despatched as messengers by him to the surrounding areas to change the titles of king, conferred on the "barbarian" chiefs by the emperors in Former Han, to that of marquis. This move caused resistance from these non-Chinese rulers, who immediately broke off relations with Han or rose in revolt96 Emperor Gengshi (re. 23-25 CE.) restored the title of king of the Xiongnu, and his successor Emperor Guangwu (re. 25-57 CE.) restored the titles of the non-Chinese rulers in the surrounding regions.97 The step was a friendly gesture, intended to satisfy the imperative need to keep peace on the frontiers. Type Four: These were states in yet more remote regions, included in the outermost of the concentric circles of the tribute system, states whose tributary relations with Han 9 4 Bielenstein, H. 1967, pp. 90-92. 9 5 HS, v99B, p. 4105; Dubs, H. H. 1938-55, III., p. 274. 9 6 HS, v99B, p. 4115. 9 7 HS, v94B, p. 3829; v95, p. 3846; HHS, v85, p. 2814. followed the first four of the five practices. They were not supervised under any kind of Chinese adrninistrative device. Although the rulers received the titles of "king" or "marquis" from the Chinese emperors, because of their distant location they never had a great impact on Chinese frontier security, either as allies or as adversaries. These states were Wo f^ ? (Wa or Yamata in Japanese. Modern Kyushu, Japan) in the east, Shanguo^ qel^  (north Burma) in the southwest, small tribal states outside the Han southern frontiers, and Jibing) ^  (Kashmir) in the far west.98 Type Five: The extreme outer regions, such as Tianzhu ?\*=t (Northern India), Tiaozhif^-7^ (Seleucia),99 Anxi (Parthia), and Daqin (Rome). Missions from these places are also recorded as coming to pay tribute even though some envoys might, in fact, just have been merchants, simply coming for commercial benefit They were, nonetheless, purposely recorded as tributaries in accord with the mythology that "barbarians" coming to the court proved the good government of the Chinese Son of Heaven. Universal Empire through the Tribute System Edward N. Luttwak in his book The Grand Strategy of the Roman Empire describes two models of empire, the "hegemonic" and the "territorial" (see FIG. 2). His model of the hegemonic empire is quite similar to the idea of the Chinese notion of concentric zones. The hegemonic empire is centered around a central zone of direct control, and includes a) an "inner zone of diplomatic control," consisted of a series of client states surrounding the central zone, and b) an "outer zone of influence," 9 8 See Tablet. 1. 9 9 This is according to a personal communication from E. G. Pulleyblank. Tiaozhi is commonly identified with an ancient place near Bushire on the Persian Gulf. surrounding the inner zone. In the outer zone, client tribes defer to the power of the empire but are not under its direct control. The territorial empire, on the other hand, has a large territory under direct rule, around which are stationed imperial troops necessary for border defence and internal security. Client states in the hegemonic empire function as buffer states in the system of imperial security with their most important function, by virtue of their very existence, being to absorb the burden of providing peripheral security against border infiltration and other low-intensity threats.100 The Chinese model of empire bears some similarities to this hegemonic system. Yet the Chinese model, organized around the tribute system, reflects the typical Chinese belief that "the king leaves nothing and nobody outside his realm," and that "having the various states of Xia within, and keeping the Yi and Di barbarians out." In general we can say that those non-Chinese in type one were in the marginal areas between the Chinese territorial and hegemonic empires. Those in types two and three who were under Chinese administration or Chinese military protection were within the hegemonic empire. As we have seen, the Chinese hegemonic empire had various categories within itself, with different levels of Chinese rule, and with each state having different responsibilities and connections to the Chinese suzerain. Some states in the Chinese hegemonic empire were expected to undertake military tasks in a frontier security system. The word fan, sometimes found in Han as fanguo or fanchen^]§_ , which was used to refer to non-Chinese in the five or nine-zone system, means "fence", "screen", or "defence." But the non-Chinese states in the sphere of the hegemonic empire did not only serve a military purpose. Since the legitimacy of the Son of Heaven within China was bound up with the fiction that he was a world ruler, the existence of the hegemonic 1 0 0 Luttwak, E. N. 1976, pp. 19-30. Robert M. Somers gives a clear outline of these two models, see Somers, R. M. 1986, p. 988. empire also served as a proof of his virtuous and capable rule and justified his legitimacy and possession of the Mandate of Heaven. Finally, the tribute system, as a concentric network, served to build a model of universal empire. It included countries of type four, on whose rulers the Chinese conferred titles, and type five which sent visitors to the Han court. Larger than Luttwak's hegemonic empire, this universal model embraced all kinds of states. Fairbank presents us with an alternate model on which to build our understanding of Chinese dynastic history. He divides the peoples and countries which were included in the concentric hierarchy of "nations" in Chinese foreign relations into three zones: 1) the Sinic Zone, consisting of the most nearby and culturally similar tributaries, 2) the Inner Asian Zone, consisting of nomadic and semi-nomadic peoples, and 3) the Outer Zone, including peoples at a further distance over land or sea.101 These divisions based on cultural, economic and geographical criteria provide a convenient framework within which to sketch out a general picture of Chinese foreign relations. The divisions utilized in this dissertation, however, which see the Chinese empire in terms of territorial, hegemonic and universal circles, are based on political and military criteria, criteria which reflected the extent to which the Chinese had dealings with a certain people, and which determined the policy the Chinese possessed relative to them. The Chinese government exercised direct control over the territorial zone, was always concerned with its security and defended its interests whenever necessary. In regard to the peoples in the hegemonic zone of the empire, however, the Chinese government just tried to keep a peaceful and friendly relationship with those who were considered subordinates or allies, and only assumed a "potential" 1 0 1 Fairbank, J. K. 1968, p. 2. responsibility for their military protection. As to those in the outermost zone, forming a universal empire, the Chinese only had a diplomatic or trade relationship with them. The payment of tribute meant, from the Chinese point of view, a recognition of subjection even though the tributaries might have different ideas and aims. After all, it was reasoned, it was always the non-Chinese rulers who went to the Han court, and foreign envoys had to conform the rituals laid down by the tribute system when they went to the Chinese court. A tribute relationship, though anti-egalitarian, functioned to maintain a friendly or peaceful relationship or simply to facilitate diplomatic and commercial activities. A peaceful relationship was evidence of the virtuous and capable rule of the emperor, thus justifying his legitimacy and possession of the Mandate of Heaven, since the legitimacy of the Son of Heaven within China was bound up with the "fiction" that he was a world ruler. Having realized the importance of the system, or rather peace through the system, the Chinese had to maintain it even if it meant China had to bear a heavy burden in return. A tributary relationship, from the point of view of the non-Chinese, could imply a friendly or peaceful relationship or simply normal diplomatic activity. Sometimes, since the so-called envoys could be merchants hoping to gain commercial benefit, their arrival at the Chinese court did not even amount to diplomatic activity. The acceptance of the system by the non-Chinese, then, clearly depended on various factors. Some people were brought into the system unwillingly through Chinese annexation, and some willingly since they could gain political and economic benefits through their contacts with China. Even this latter sort of people, if they wanted to have relations with China, generally were required to conform to the rules laid down by the system. In a complex political organization, we can distinguish three types of power by which superiors attempt to establish or to extend their control over others: physical, economic, and moral.102 Some non-Chinese rulers adopted a positive attitude towards their relations with China in order to seek Han support through the tribute system to enhance their own physical power, especially when they faced rivalries. They would do this even though it implied an inferior status. Non-Chinese rulers also benefited economically from tribute and trade with China. The payment of tribute, and the return of Chinese imperial gifts, served the function of trade, and appeared more like "gift trade" to the non-Chinese. It was usually equal in an economic sense since the Chinese tried to return gifts of the same value, if not higher. Imperial gifts together with other financial support and the essential goods the non-Chinese obtained in this kind of "tribute trade" with the Chinese compensated for the less than exalted position they occupied in the tribute system.103 The Xiongnu chanyu's constant demand for Han goods did not stem from any personal greed, as the Han Chinese thought Rather it was the need to turn to the Chinese economy in order to obtain goods to reward tribal leaders and thereby maintain his nomadic state organizational structure intact. As nomads, the Xiongnu could maintain a subsistence standard of living, but with the economy based on a highly mobile pastoralism it was difficult to accumulate, or store, the wealth necessary to run a more complex and powerful state.104 The official titles conferred on non-Chinese rulers by Han not only made the non-Chinese states nominal Han subjects but also strengthened the moral power of those states. In 29 CE. the king of Suoju^ijz. in the Western Regions, for instance, was established. He was given the title of Grand Chief Commandant of the Western Regions (Xiyu da duwei T s ^ p ^ ) by Han. Fifty-five states of the Western Regions were placed under his control. In 41 CE. Han gave him the seal of 1 0 2 Commons. 1924, pp. 47-64. 1 0 3 Rossabi, M. 1983, pp. 3-4. 1 0 4 Barfield, T. J. 1981. Protector-general of the Western Regions (Xiyu duhu). Even though Han soon lowered his rank to General-in-chief (da jiangjun X./f^2^ ), the king still continued to use the former title so as to make other states his subjects and to levy heavy taxes.105 The prosperity of the tribute system depended mainly on the superior military and economic strength of the Chinese as well as on the relative weakness of the surrounding non-Chinese. Backed by the wealth of the nation, Emperor Wu strongly encouraged foreign peoples to come to China. Chinese missions were despatched to regions as far away as Parthia. By contrast, at the beginning of the Later Han period, Emperor Guangwu refused the request of the states of the Western Regions to establish tributary relations because China, then preoccupied with its internal problems, was unable to perform the duty of military protection as suzerain.106 3. The Period of Division: the Challenge to the Tribute System The Challenge to the Idea of Universal Kingship During the three and half centuries of the Period of Division, from the collapse of the Han dynasty in 220 until 589, when the Sui dynasty again unified the whole of China, the Han empire was divided up under the rule of different Chinese and non-Chinese leaders. Initially three Chinese regimes, Wei, Shu, and Wu, fought for supremacy. Then five "barbarian" (wuhu ) groups: Xiongnu, Jie^fp, Xianbei, Di ^ , and Qiang, who were looked down upon by the Chinese as "uncivilized," "savage," and "greedy," occupied the northern part of China, the so-called Zhongyuan, and established sixteen kingdoms at various times contrasting the Eastern Jin regime in the southern part. Referring to themselves as kings or emperors, the leaders of these "barbarian" groups took up the role of Son of Heaven, and ruled 105 HHS, v88, pp. 2923-24. 106 HHS, v88, p. 2909. over the Chinese. Towards the end of the fourth century the Tuoba^ z; JJ&^ of the Xianbei people conquered all the other non-Chinese regimes and established a unified dynasty in the north. They competed with their counterpart in the Chinese Southern dynasties for the role of ruler of all-under-heaven. The very idea of the Chinese world order, universal kingship, was faced with a challenge. In disunity, there was no single Son of Heaven to rule both the Chinese and non-Chinese. Within so many different regimes, historians during the Period of Division all claimed that their own dynastic regimes possessed the Mandate of Heaven, and were, therefore, legitimately at the center and top of the hierarchical world order. The ideas of universal kingship and unity were considered as the only acceptable condition for China to be in, and provided justification for the different rulers in their competition for legitimacy and their conquering others in order to establish a unified rule in China like that of the Qin and Han dynasties. The questions that concern us here are how the Chinese historians regarded the states organized by the non-Chinese, the so-called "five barbarians" and the Northern dynasties, who threatened the ideology of Chinese Son of Heaven ruling the world, and how these regimes utilized the tribute system to strengthen their legitimacy. The Legitimacy of the Sixteen Kingdoms The non-Chinese rulers were not completely confident about their own strength, even though they spared no effort to demonstrate their power over the Chinese and to maintain or enhance their own ethnic superiority. In the uprisings at the end of Western Jin, there were certainly and-Jin feelings on the part of the non-Chinese, for the corruption and decay of the Jin Chinese on both central and local levels caused tremendous suffering to the people, particularly the non-Chinese who had settled inside China since the time of Later Han. However, to successfully rule an advanced agricultural society, which differed greatly from their own nomadic ones, it was necessary for these non-Chinese rulers to utilize prevailing Chinese political theories to establish and cultivate a belief in the legitimacy of their rule in the minds of the population at large. To that extent they had to conform to Chinese tradition, to the Chinese orthodoxies, or approved norms. Without doing so it was impossible to maintain the stability of the political order.107 They were also bound to organize their governments on existing Chinese models, which meant, on the practical level, relying almost entirely on a Chinese bureaucracy to carry out their rule. Legends, prophecies and signs from the gods were fabricated in addition to their adoption of Chinese political theories. Wittingly or unwittingly the non-Chinese regimes hastened the process of sinicization by their attempts to maintain their rule with the assistance of Chinese bureaucrats. Their official histories were compiled by Chinese traditional historians, who followed traditional rhetoric and patterns in court documents. Rulers and Chinese officials in the north, for example, claimed that, if the Son of Heaven failed in his responsibilities to guarantee the welfare of the people, whoever had virtue would have the Mandate of Heaven and could take over, thus using the most central and durable traditional Chinese ideology to justify their rule. There is, however, an almost unanimous opinion in Tang historical writings that the non-Chinese ruling north China before the Tuoba Wei's unification were simply "barbarians" usurping the Chinese power. For them it was a period of foreign invasion. Such invasions were not without precedent in Chinese history. The Tang historian Fang Xuanling/#]^J^ in the Jinshu^H} adopts the form of the "records of illegitimate rulers" (zaiji^iL) for the histories of these kingdoms. However, the Tang Chinese also acknowledged that the decay and lack of virtue of the 1 0 7 For a survey of the importance of legitimacy, see Wechsler, H. J. 1985, pp. 10-20. Western Jin rulers had provided the chance for the "barbarians." They saw the invasions as a punishment from Heaven, but still held strongly to the belief that "barbarians" should not rule China.108 The Legitimacy of the Northern Dynasties As for the Northern dynasties, the Northern Qi historian, Wei Shou^ ^Vx^  a Han Chinese, readily accepted and assisted the Tuoba Wei, which unified the north and set up the Northern dynastic power, in their claim to legitimacy. His Weishu^^L t£ adopts such a belief as its basic assumption. It begins with a myth about Changyi &1&>, the ancestor of Tuoba of the Xianbei tribe. He is said to have been a son of the Yellow Emperor who enfeoffed his sons with land, both inside the land of the various Hua (Chinese) and outside in the "outermost zone" (huangfu Changyi is said to have settled in the northern land, which had a mountain called Xianbei.109 Taking Northern Wei as the legitimate power, Wei Shou records all the other rulers, Chinese and non-Chinese, as falsely assuming the title of emperor. The histories of these regimes were compiled in accordance with the calendar of Tuoba Wei. Not only did he show a strong bias against other non-Chinese rulers who, according to him, were undoubtedly "barbarians," but his bias against the Chinese rulers of the South was, if anything, even stronger. Contrary to Wei Shou, Shen Yue^ Xl^ t? and Xiao Zixian of the Southern dynasties, the compilers of the Songshu and Nan Qishu respectively, took a clear stand on the side of the Southern dynasties. Shen Yue had been appointed to several high positions from the Song to Liang periods, 1 0 8 JS, vl22, pp. 3072. 1 0 9 WS, vl, p. 1. 61 whereas Xiao Zixian came from the Qi and Liang royal family. Obviously, it was then-obligation as well as in their political self-interest to defend the Southern dynasties as having received the Mandate of Heaven and to degrade the rulers of their Northern With the ascendancy of the Tang dynasty, the canon of "Standard Histories" for the Period of Division was filled in, partly by adopting existing histories including the Weishu, for the north, and the Songshu and Nan Qishu, for the south, and partly by compiling new histories, including the Jinshu (already discussed), the Bei Qishu Mt. and Chenshu p]^  f° r m e south. Except in the case of the Jinshu where there was no problem in adopting the view that the sixteen "barbarian" kingdoms of the north were illegitimate, the question of ascribing legitimacy only to southern or northern regimes was on the whole avoided. The Weishu, with its claim of legitimacy for the north, and the Songshu and Nan Qishu, which claimed legitimacy for the south, were both accepted. The later histories of both southern and northern dynasties were written from their own separate points of view. Perhaps this recognition by Tang historians can best be explained by pointing to the fact that it was in their own political self-interest to do so. The Tang dynasty was established following Sui, whose founder, Yang Jian $ ^ § 5 , was a Han Chinese official in the Northern Zhou. The royal house of Tang had close relations, through marriages, with the non-Chinese rulers of the Northern dynasties, and Tang officials came from various earlier different regimes. Tang unification of both north and south, and the consolidation of its rule, needed a unified legitimating theory acceptable to all. But although Li Yanshou was very careful not to use degrading terms in his wording of the "History of the Northern dynasties" (Beishi) and the "History of the Southern Dynasties" (Nanshi), he used the word beng ffl for the death of the emperors of the north, and the word cu for the death of the emperors of the south. rivals as the "rope-head caitiffs" (suolu %% ) or the Wei "caitiffs" (Weilu). north This may show that Li considered the southern rulers had a lower status, thus holding the Northern dynasties as legitimate. Among the Chinese scholars in the Sui, Tang, and Song periods we find basically three different opinions concerning the legitimacy of the Northern dynasties: 1) Non-Chinese regimes were as legitimate as native ones as long as they cherished native Chinese culture and preserved it for prosperity. 2) The Tuoba Wei were "barbarians" who did not practice propriety and righteousness. Their rule was, therefore, illegitimate. 3) Legitimacy was always a variable depending on whether one did or did not rule over a unified empire. Any regime, be it Chinese or non-Chinese, which did not rule over the whole empire was iUegitimate.110 The Persistence of the Tribute System in Former Qin In spite of continuous wars, turmoil, and the emergence and collapse of different regimes in the frequently changing situation, great importance was attached to the tribute system with its implications of hierarchical relationship of suzerain and subordinate. Both Chinese and non-Chinese rulers valued the tribute system as a basis for their dealing with other peoples, particularly because the smooth working of the system would enhance their prestige as Son of Heaven, enabling them to obtain considerable economic benefits through trade. Even though they might fail to impose one centralized government over China, the different regimes tried to develop the system within, and for the benefit of, their own territory. After the non-Chinese had established their states, along with Chinese political institutions they also adopted the tribute system, particularly the practice of political 1 1 0 Cf. Rao Zongyi. 1976; Zhao Lingyang. 1976; Wechsler, H. J. 1985, pp. 16-18. investiture and requests for tribute in their conduct with other peoples, who were in a weak position relative to them. One of the many examples is in the case the Former Qin state (351-394). learned from the Chinese in dealing with other non-Chinese. In 359 Weichen , the Wise King of the Left of the Southern Xiongnu, surrendered to Fu Jian, and asked for permission to occupy land in the interior of China. In agreeing, Fu Jian alluded to an incident in the Chunqiu period in which the Chinese state of Jin made an alliance with the Rong "barbarians" on its borders.111 This shows that he was deliberately adopting the role of a Chinese emperor, and treating the Southern Xiongnu as "barbarians." After Fu Jian won control over Liangzhou>^ *>rj in 376, Qin envoys were sent to the Western Regions and more than ten states came to pay tribute.112 The luxuries obtained from these tributaries were used to decorate Fu Jian's palace in order to show his power. Kings of Shanshan If-, and Nearer Jushi ji£.Jjip went in person to the court to pay homage.113 Around 382, kings of Nearer Jushi and Shanshan arrived again to pay homage to Fu Jian. They are recorded as having asked to send tribute annually. Fu Jian ordered them to pay tribute every three years and to appear in person every nine years.114 In the following year Qin sent troops to the Western Regions in order to bring all the oasis states under Qin's control.115 The campaign ended in the conquest of thirty-six states in the region and in the capture of a great amount of precious goods.116 The motivation of Fu Jian in the campaign may have been a combination of personal 1 1 1 WS, v95, p. 2055; JS, vll3, p. 2887. 1 1 2 / S , vll3, p. 2900. 1 1 3 JS, vll3, p. 2904. 1 1 4 JS, vll4, p. 2911. 1 1 5 /5 ,v l l4 , pp. 2914-15. 1 1 6 75, vll4, p. 2923. During his rule, Fu Jian ^ / of Former Qin adopted several measures ambition and a sort of what Mather calls "benevolent imperialism" and various other factors,117 but among them there was certainly one of consolidating his legitimacy as a Chinese emperor. The Tribute System in the Northern and Southern Dynasties In the period of the Northern-Southern dynasties when both sides needed to win the support of the non-Chinese in their competition with each other for supremacy, it was particularly important to attract non-Chinese peoples into the tributary relationship, and to keep the system flourishing. Many embassies were despatched to foreign lands to establish peaceful relations. The system of investiture with titles of nobility and offices as part of the tribute system developed with more titles being provided to the non-Chinese rulers together with rich imperial gifts, since these political and economic benefits would entice them to come and enter into a symbolic position of subordination to the Son of Heaven. Moreover, the tribute system would also enable the suzerain to obtain actual military support from tributaries. For example, in 439 Emperor Wen of Song (re. 424-453) decided to attack Northern Wei. He ordered the Korean state of Koguryo (Gaogouli Y%*J/$L M Chinese), which was considered a tributary of Song, to send horses and eight hundred horses duly arrived.118 The non-Chinese in turn were provided with opportunities to take advantage of the competition by playing off these two powers against each other. Some of them, such as Koguryo, Paekche (Baiji in Chinese, also a Korean state), Tuyuhun h^J^ >ifz. (in modern Qinghai), and states in the Western Regions, kept tributary relationships with both the Northern and Southern dynasties out of consideration for 1 1 7 Mather, R. B. 1959, pp. 2-6. 1 1 8 55, v97, p. 2393. their own interests, sending envoys continuously in order to obtain political and economic benefits. Between the Northern and Southern dynasties, diplomatic activities continued while wars went on. The two regimes had their different views on these communications. In the accordance with the claim of legitimacy the Weishu records the missions from the Southern regimes as coming to pay tribute to Northern Wei. However, these communications are either not recorded or recorded in a different wording as equal diplomatic activities in the Songshu and Nan Qishu, which were compiled in the Southern dynasties. For example, chapter 97, the Biography of Liu Yu$^/f&, the founder of the Song dynasty, in the Weishu records thirty-one tributary missions from Song during the fifty-nine years of the Song dynasty. Under the year 420 it says that Song requested peaceful communication several times, which Wei permitted. The following year Song despatched Shen Fan/^i^gj and Suo Jisun j^^&t-^ to pay a court-visit to Wei with tribute, chaogong J^^ sT in Chinese.119 Another record says that in 430 Song sent Tian Qi to pay a court-visit with tribute.120 In contrast, in recording the mission of 421, the Songshu says that after Liu Yu's victory over Wei in 417, Wei sent envoys to seek peace. From then on communications were kept up annually. Song under Liu Yu despatched Shen Fan and Suo Jisun to "return the embassy" (baoshi^^jij to Wei.121 For the one in 430 the Songshu records that Song sent Tian Qi to deliver a message to Wei.122 Chapter 95, the Account of Suolu (i.e. Northern Wei), records the mutual communications between the South and North as between equals.123 However, in the basic annals of 1 1 9 WS, v97, p. 2134. 1 2 0 WS, v97, p. 2136. 1 2 1 55, v95, p. 2322. 1 2 2 55, v95, p. 2331. 1 2 3 55, v95, p. 2334. the Songshu, two missions from Wei of 469 and 471 are recorded as "Caitiff (suolu) despatched envoys to present (xian^) local products," implying the payment of tribute.124 The Weishu does not record these two missions. Tang historians' writing on the Northern and Southern dynasties did not use the word "to come to pay a court-visit with tribute" (chaogong ^J|5) or other terms with the same meaning, as often as their predecessors. While in the basic annals of the Weishu twenty-eight missions from Song are recorded as having to come to pay tribute, Li Yanshou, in the basic annals of Wei in the Beishi, describes most of the missions as "the people of Song coming for a visit (Songren lai pin ^ . A ^ ^ )." There are, however, two exceptions. One passage written in 433 says: The people of Song came for a visit, 'presenting' (xian) a trained elephant. The other in 450 says: Envoys from Song Emperor Wen presented a hecatomb, offered (gong ) local products, and requested a daughter in marriage for the Imperial Grandson in order to seek peace.125 In describing other missions travelling between the North and South, Li Yanshou, in the basic annals of either the Beishi or Nanshi, uses the expression "people [of so and so] came for a visit." Generally speaking, the word chao in the sense of paying a court-visit was used by traditional Chinese historians, 1) when feudal lords or foreign rulers went to the Son of Heaven in spring,126 and 2) when rulers of small states went to visit rulers of large ones.127 They also used the word pin^%, in the sense of inquiry or visit, in two 1 2 4 SS, v8, p. 165; p. 168. 1 2 5 SS, v2, p. 48; p. 61. 1 2 6 Zhouli Zhushu, "Da Zongbo," p. 652; Liji, "Quli," BK. I, p. 111. 1 2 7 Zuozhuan (Cheng 12), 377::378; Zhouli Zhushu, "Daxingren," p. 1343; Guliangzhuan, Huan 9/4. A survey of the the Combined Concordances to Ch'un-Ch'iu, Kung-yang, Ku-liang and Tso-chuan in different contexts: 1) when envoys were sent by feudal lords to visit the Son of Heaven,128 and 2) when feudal lords visited each other.129 Li Yanshou, in both Beishi and Nanshi, uses the word pin for both parties, intending thereby to avoid making one side superior to the other. But it remains a question as to why he retained the usage of "presenting" (xian) and "offering" {gong) in the above two cases. One possible reason for his describing the Song mission as xian in 450 might be because at the beginning of that year Song launched a major military campaign against Wei but suffered a heavy loss. Song therefore sent a peace-seeking mission to Wei, which considered that act a surrender. Other Tang historians showed the same caution when dealing with the diplomatic activity between the South and North. The basic annals of the Liangshu and Chenshu compiled in Tang describe the envoys from the North to the South as "coming for a visit" (laipin). The Bei Qishu, compiled in Tang, record the missions from Liang as "paying tribute" (gong) but, since the tributary missions started in 550, this may simply reflect the fact that Liang was then undergoing an internal crisis caused mainly by an internal rebellion. Members of the royal family were struggling for support from outside by seeking to become dependents, or even subjects, of Qi . 1 3 0 The Beishi and the Zhoushu, compiled in Tang, also record the Liang's missions to Qi and Zhou as "tributary." Their opinion seems to be based on the fact that Liang became a dependent of Qi and Zhou. Harvard-Yenching Institute Sinological Index shows whenever chao is used to indicate a visit between feudal states, it is an envoy from the smaller state that is visiting the larger one. 1 2 8 Liji Zhengyi, "Wangzhi," p. 531, commentary. 1 2 9 Liji, "Quli," BK. I, pp. 111-112; Yili Zhushu, "Pinli," p. 543, commentary. 1 3 0 See the basic annals of the BQS. The Challenge of the Nomads to the Northern Dynasties After the nomadic Xianbei people had settied down and established dynastic rule over the agricultural Chinese, they were gradually sinicized and lost the advantage their earlier nomadic life-style had given them in warfare. During their rule, the period of the Northern dynasties, the different rulers had to adopt the heqin policy when confronted with threatening nomadic rivals like the Rouran J£^ $ (or Ruru^^ and the Tujue Turks. Under the conditions of a heqin agreement, it was the Northern regimes which paid tribute to these nomads although it was delivered in the name of gifts. At the beginning of the fifth century a nomadic people, the Rouran, rose as a vigorous power, extending their rule over the Mongolian steppes and Manchuria. Like the Xiongnu in the Han period, they became a fierce adversary of the Northern dynasties. The policy of the Northern dynasties towards them varied with the changing situation, but heqin remained a major means. In 429 Northern Wei launched a military campaign against the Rouran, who were defeated and weakened. In order to maintain peaceful relations as well as to consolidate the victory, Wei sent Princess Xihai tt£ T^ fk in 434 to wed the qaghan of the Rouran while the emperor himself married the qaghan's younger sister.131 However, in 506 when the Rouran were further weakened and asked Wei for a peace agreement, Wei insisted that the Rouran could only expect to become Wei subjects and not expect to conclude a peace agreement132 In 524 when Wei faced the internal disorders caused by the rebellion of the Six Garrisons, the Rouran helped suppress the rebellion. To show gratitude, in 528 Wei issued an edict granting privileged treatment to the qaghan. His personal name was not to be used when he was announced at the court, and he personally was not expected to 1 3 1 WS, v4, p. 83; vl03, pp. 2293-94; BS, v98, pp. 3253-54. 1 3 2 WS, vl03, p. 2297; BS, v98, p. 3257. refer to.himself: as "subject" when presenting documents. In 533 Wei permitted the marriage of Princess Langyai^ -^ f" to the qaghan's son.133 7 The marriage, however, did not take place because in 534 Northern Wei split into two::Eastern Wei and Western Wei. Enfeebled by their struggles with each other, both sides competed to conclude a marriage alliance with the Rouran so that they would be: able to maintain peace on their borders as well as concentrate their efforts against each other and against their southern enemies. Western Wei entered into two marriage alliances with the Rouran. In 535 Princess Huazheng'/cg^ went to wed the brother of Anagui f^^i^ Qaghan. Following the death of his empress, Emperor Wen of the Western Wei (re. 535-551) requested the Rouran for a marriage contract. Since at the time they were attempting to form an alliance with Eastern Wei, the Rouran initially did not show much interest in the request Emperor Wen then sent another mission, this time with rich gifts, and finally secured a marriage with qaghan's daughter who became empress in 538. 1 3 4 Eastern Wei concluded three marriages with the Rouran. In 541 a princess was sent to wed the qaghan's son. In 542 Anagui Qaghan asked to send his grand-daughter to wed the minister Gao Huan's f^^^son. Wei agreed.135 Apprehensive of the strong Rouran and their possible joint attacks with Western Wei, Gao Huan requested a Rouran princess for another son. But Anagui replied that it would only be permissible for Gao himself to wed his daughter. Gao Huan agreed after hesitation. The marriage proceeded in 545.1 3 6 These marriages succeeded in maintaining peace on the borders, and in posing some threat to Western Wei. 1 3 3 WS, vl03, pp. 2302-03; BS, v98, p. 3263-64. 1 3 4 ZS, v33, p. 571; BS, vl3, p. 507; v98, p. 3264. 1 3 5 WS, vl03, p. 2303; BS, v8, p. 281; v98, p. 3265. 1 3 6 This is according to the BQS, v9, p. 124; BS, vl4, pp. 517-518. In the WS, vl03 (p. 2303) and the BS, v98 (p. 3265), the marriage is only recorded as that in 546 Anagui asked to sent his favorable daughter to wed Gao Huan. The competition between Northern Qi (formerly Eastern Wei) and Northern Zhou (formerly Western Wei) to win the support from the Tujue Turks is another example showing how these sinicized non-Chinese rulers would not insist on their claim to superiority when circumstances made it impossible. They tried to placate the Turks through alliances, both marital and military, and by acceding to their often unreasonable trading terms. Even though the various visits from the Turks are still recorded as tributary missions, a thinly veiled fiction by which the Turks were made to appear proper subjects, in reality, the Turks were speaking from a position of power, taking what they chose.137 The Tujue Turks became powerful in the mid-sixth century (see Map 3). Previously they are said to have been blacksmith slaves of the Rouran. Having broken away from the latter they turned to Western Wei for support. In 551 Western Wei consented to send a princess to wed Tumen i .T 'Jf (Bumi'n) Qaghan.138 To Western Wei the marriage was viewed as a successful alliance with an enemy of the Rouran. After 552 when the Turks won a crucial victory over the Rouran, they asked Western Wei to help execute all the Rouran refugees, a grisly task which Western Wei carried out. In the reign of Emperor Gong of Western Wei (554-556), Muhan Qaghan promised to marry his daughter to Yuwen Tai ^ X j ^ , who held the real power in the state at that time. Although the marriage did not take place because of Tai's death, the qaghan did agree to send another daughter to wed Tai's son Yuwen Yong The marriage was again put off because Northern Qi also tried to conclude a marriage alliance with the Turks by offering more gifts, and the Turks then wanted to break their promise. The Western Wei did not, however, give up in their attempt to make the Turks an ally. In 561, when Yuwen Yong (re. 561-578) came to the throne 1 3 7 Moses, L. W. 1976, p. 66. 1 3 8 ZS, v50, p. 908; BS, v99, pp. 3286-87. as emperor of Northern Zhou, he despatched several missions to the Turks. Finally he succeeded in securing a marriage alliance and getting the Turks to join in two military campaigns against Qi in the following years.139 In 568 the Turkish princess arrived to take up the position of empress of Yuwen Yong. In the following reigns of emperors she was shown great favor and given honorific tides.140 The marriage arrangement between Zhou, the successor state to Western Wei, and nomads was severely criticized by the Tang historian Linghu Defen^$>«V fjt-$l> w n o said that Zhou, abandoned those of their own kind and went to aliens, mixed barbarians with Chinese, abandoned the proper rules and order of marriage and sought external benefits from wolves. After a while those [of the Chinese] who paid [the bride-price] became wearied, but those who gave [the bride] were never satisfied. What was once called heqin before long resulted in enmity.141 Nevertheless, Northern Zhou made great efforts to maintain the alliance with the Turks, preventing them from invading by making an annual payment of silk products and by extending favorable treatment to their envoys in the Zhou capital. In 579 Zhou promised a princess in marriage to Tabo -^ -^ (^Taspar) Qaghan of the Turks.142 The marriage alliance worked in so far as it freed Zhou from worrying about the security of the northern frontiers, and, as a strategy to gain Zhou time so that under its successors, the cause of unification was completed. Northern Qi at first seemed to stick to the pro-Rouran policy against the Turks. It soon changed, however, as the situation deteriorated. In 552, defeated by the Turks, Anagui Qaghan of the Rouran committed suicide. His son Anluochen ^  ||[ J^and his cousin led their followers in flight to Qi. The following year saw 1 3 9 ZS, v33, p. 571; v50, pp. 910-911; BS, v99, p. 3289. 1 4 0 ZS, v9, p. 144. 1 4 1 ZS, v9, p. 149. 142 ZS, v50, pp. 911-912; BS, v99, p. 3290. Northern Qi launch a successful military attack against the Turks and Yanluochen was established as qaghan. Yanluochen soon turned against Qi, however, and QL thereupon abandoned the attempt to ally itself to the Rouran.143 Qi turned to the Turks and requested a marriage alliance around 555. Though Qi did not succeed, mainly because of interference from Zhou, communications with the Turks were established and this maintained a state of affairs, which posed a continuing threat to Zhou. Worried by the prospect of Turkish invasion, Qi also had to pay a heavy price for peace, "emptying their storehouses to supply the Turks." Boasting of his power, Taspar Qaghan was quoted saying: As long as my two sons in the south remain filial and obedient why should I be worried about lack of goods?144 Like the early Han rulers, the Northern Qi and Zhou could only restrain their nomadic neighbors from making incursions, to a certain degree, by paying regular tribute which cost them a great deal, not only in economic terms but also in political dignity. Conclusion The tribute system, rooted in the Late Shang and Zhou political systems, was formulated and held as an ideal model in regulating relations between the Son of Heaven and his subjects, Chinese and non-Chinese, in such Confucian classics as the Zhouli and Liji. Two key principles can be distinguished as underlying the tribute system: "the king leaves nothing and nobody outside his realm" (wangzhe wuwai), and "having the various states of Xia within, and keeping the Yi and Di barbarians out" (nei 1 4 3 BS, v98, pp. 3266-67. 1 4 4 ZS, v50, p. 911; BS, v99, pp. 3289-90. Cen Zhongmian has the opinion that the word wu (here translated as "goods") should refer to silk products. See Cen Zhongmian. 1958, pp. 8-9. zhu-Xia er wai Yi-Di). The first one represented the ideological purity of China's foreign policy and provided justification for Chinese expansion. The second one stressed practicality and flexibility and was utilized to emphasize border defence. The Han dynasty contributed to the development of the tribute system through the institutionalization of tributary practices and their systematic application in foreign relations. Generally, Han established five types of relations with other states, which can be grouped into three concentric circles, with a zone of direct control at center, a zone of diplomatic control in the middle, and an outermost zone on the periphery free of any Chinese control. The concentric circles can also be seen as territorial, hegemonic and universal circles, embracing "all-under-heaven." From the Chinese point of view, the working of the tribute system functioned to maintain a modicum of peace and security on the frontiers and to prove the benevolent rule of the emperor, thus justifying his legitimacy and possession of the Mandate of Heaven, since the legitimacy of the Son of Heaven within China was bound up with the fiction that he was a world ruler. The non-Chinese, on the other hand, no doubt held a different point of view in regard to the tribute system. They accepted the practices laid down by the system basically because they could gain political, economic and military benefits from it and use them to enhance their own military, economic and moral power. In the Period of Division, disunity challenged the idea of universal kingship, the basic conception in the Chinese world order. Different regimes competed with each other for legitimacy as the single ruler over China and all utilized the tribute system to strengthen their own legitimacy, thus making the tribute system sustain and play an important role. Another severe challenge to the tribute system came from the nomadic powers of the Rouran and the Turks, who greatly threatened the security of the Northern dynasties. Policy measures towards them varied with the changing situation, but "marriage alliance" (heqin) remained a major means. Under the conditions of heqin agreement, even though some of the visits from the nomads are still recorded as tributary missions, it was virtually the Northern regimes which paid tribute to these nomads. Chapter II Sui-Tang Relations with the Tujue Turks: the Policy of Settlement In 581 Yang Jian, a Chinese official of Northern Zhou, received the abdication of the child Emperor Jing of the Northern Zhou, thus inaugurating the Sui dynasty. In 589, with the Sui conquest of Chen — the Southern regime — China was once again unified under a centralized state power and had to face the perennial problem of the nomadic threat from the north. In this chapter, we shall discuss how they faced this challenge from the Tujue Turks (hereafter referred to as Turks), especially the Chinese policy of settlement of the Turks who had submitted. Traditional Chinese Measures to Deal with the Nomads Even in early Zhou, "barbarian" invasions were a major headache for the Chinese. It was a perennial northern frontier problem throughout imperial times. The famous debate, in 81 B.C.E. during the Han dynasty, over the salt and iron monopoly policy reflected the dilemma China faced with regard to the nomadic Xiongnu people frequently raiding and devastating Chinese territory. The effort required to build up defences or to launch attacks exhausted the Chinese and imposed heavy strains on the economy. Even though these strains caused internal disturbances such measures were required or else the forays and depredations never ceased.1 From early times, the Chinese discussed, argued about, and finally advocated various ways to treat the non-Chinese or "barbarians." They worked out different measures to prevent nomadic incursions, which have been surveyed comprehensively 1 Cf. Yan Tie Lun Jiaozhu and Gale, E. M. 1931. by Lien-sheng Yang and Wang Gungwu.2 The salient points of these measures are as follows. 1) Aggressive military measures to defend or expand Chinese territory. Qin Shihuangdi and Emperor Wu of the Han carried out expansive military campaigns on these justifications: the Son of Heaven was to rule all-under-heaven, Chinese culture could convert other people to a higher civilization, and the Son of Heaven had to use military force to protect his people from disturbance and to consolidate centralized power. The Han Chinese openly expressed these ideas. In the debate on the Han salt and iron monopoly policy, Sang Hongyang, the Censor-in-chief (yushi dafu), who was a Legalist statesman, insisted: When the Sage Ruler engages in expansion it is not for the sake of his private gains; when he uses military force it is just because he is enraged. They (expansion and military force) are the means by which he is to relieve disaster and avoid harm, thereby using farsighted thought on behalf of the common people.3 As well, frontier expansions could bring wealth, profit and goods China did not have, such as horses. The "censor" (yushi) in the debate said, as a result of the expansion of Emperor Wu, that precious novelties and foreign articles fill the Inner Palace, and fleet-footed palfreys and chargers pack the Outer Stables. Every common man can ride a fine mount, and the people feast to satiety upon oranges and pumaloes. This shows what affluence the profit derived from the frontier commandeering has brought.4 2) The adoption of a defensive attitude in order to draw a distinguishing line between Chinese and non-Chinese. This is the policy of "having all the Chinese within 2 Yang Lien-sheng. 1968, pp. 20-33; Wang Gungwu. 1968, pp. 34-62. 3 Yan Tie Lun Jiaozhu, v8, "Jiehe," p. 287. 4 Yan Tie Lun Jiaozhu , v3, "Weitong," p. 105; Gale, E. M. 1931, p. 92. and keeping all the barbarians out." The Han historian Ban Gu, an orthodox Confucianist, expressed this attitude through his analyses of "barbarian" problems. Discussing the issues within the framework of the five-zone system, he said that the "barbarians" were outer people and therefore, "Sage Kings" (shengwang) of the ancient time reared them like birds and beasts, neither making treaties with, nor attacking them. Based on precedents, Ban concluded that if one made treaties with the non-Chinese one would waste the gifts and be deceived; if one attacked them, one would weary one's army and invite invasions. Since their land could not be ploughed and made to produce food and since one could not keep their people as subjects, Sage Kings kept them at a distance and did not enter into close relationships with them. Ban Gu concluded the way the Sage Kings treated "barbarians" was thus: government and instructions did not reach their people, and the correct calendar was not applied to their country. When the "barbarians" came to invade, the Sage Rulers punished and resisted them and prepared for and guarded against them when they left. If, attracted by China's civilization, they came to offer tribute, the Sage Kings would treat them with courtesy and keep them under loose rein without severing the relationship, so that the blame of being crooked would always be on them.5 3 ) The attitude of appeasement. This was adopted in order to bring the non-Chinese into a peaceful relationship. An alliance would then be forged with one party against the other so as to divide and rule, i.e., to use "barbarians" against "barbarians." Measures introduced under this broad category include diplomatic missions for the maintenance and expansion of communications, peace agreements, political or matrimonial alliances, establishment of kinship relationships, political investiture and, finally, mutual trade under the framework of the tribute system and in border markets. The Chinese sometimes actively adopted such means in order to attract non-Chinese 5 HS, v94B, p. 3834. Part of the translation is from Yang Lien-sheng. 1968, p. 23. into an alliance or to bring them into the tribute system. Sometimes, however, when China was weak and the balance of power had tipped in favor of the nomads, such measures were forced upon the Chinese. There were even occasions when China was forced to pay tribute to the nomads in order to have peace on the frontiers. 4) Sinicization. The Chinese also attempted peaceful persuasion encouraging the spread of Chinese culture to their neighbors in order to convert them to Chinese ways. Whether conscious or not, this translated into a cultural expansion and incorporation of non-Chinese into the Chinese empire. The general spread of Chinese culture was seen not only in the implementation of the Chinese administrative system on the non-Chinese in the south, southwest and east where agriculture was a major way of life, but also in the resettlement of the submitted nomads on Chinese land. This resettlement policy was actively adopted in Later Han. It worked effectively, given the particular circumstances at that time, but later, in the view of Chinese historians in later imperial times, it led to a dark era, the catastrophe of the establishment of non-Chinese regimes inside China. In Western Jin, Guo Qin%f£X ,^ with foresight, urged the removal of the non-Chinese who had settled within the Great Wall since Han. He could see the danger of what soon actually happened when the non-Chinese rebelled and overthrew the Western Jin regime.6 Jiang TongxL^lli of the same period also expressed his concern and objections to the settlement of non-Chinese inside the Great Wall. In his essay "On the migration of the Rong" (Xi Rong L w n ^ ^ f ^ ) he argued that the Rong and Di (referring to the nomads) were the most greedy and ferocious among the four "barbarians," and that they submitted only when weak and rebelled when strong, so even the Sage Kings were unable to extend civilization to them and lead them to the 6 Quan Jinwen, v83, Guo Qin, "Shang yufang Xiongnu shu," p. 1941. proper path. In later Zhou, he said, some Chinese states had invited the "barbarians" to enter China and had appeased them in order to use them. The result of this was that disasters had occurred. However, from Han to Wei (of the Three Kingdoms), the Chinese had resettled even more "barbarians" inside China in order to employ their military forces which caused even more troubles. He urged that the Jin emperor move the "barbarians" outside China before they could accumulate strength and rise in revolt. His advice, however, did not receive much attention.7 Jiang Tong was correct in realizing that, in Chinese relations with the nomads, there existed a constant pattern. When the nomads were strong they challenged Chinese superiority and threatened their security; when they were weak they submitted to China and requested Chinese assistance in order that they could recover and accumulate their strength. Chinese rulers, themselves caught in the weaving of this pattern, would accept the submission of the nomads with the aim of utilizing their considerable military force. But Jiang Tong failed to see another important aspect of the settlement policy. In fact it was often the case that the Chinese rulers could not afford to refuse the "submission" and subsequent resettlement, particularly when it was less costly than engaging in wars for the purpose of keeping peace on the frontiers. In that manner they could buy a period when internal order would prevail. Besides, the "submission" proved that that the rule of the Son of Heaven was benevolent and virtuous, a factor that served to strengthen the legitimacy of the Son of Heaven. In Sui-Tang relations with the Turks, the Chinese adopted all the above measures designed to solve, or alleviate, frontier problems caused by the nomads, including the resettlement into the Chinese empire of Turks who had submitted. This policy is the focus of this chapter. 7 JS, v56, pp. 1529-34, The Internal Conflicts among the Turks The first Turkish empire started with Bumin Qaghan in 552 and split, in 583, into two qaghanates, the East and the West, both hostile to each other.8 In 581, with China entering a turning period in its history with the establishment of Sui, the Turks also faced their turning point: the death of Taspar Qaghan. His death set off a suicidal power struggle among the Turks which tipped the balance of power between the Chinese and the nomads in favor of the former. The Turks inhabited an extensive territory covering the steppes on the Mongol Plateau, Central Asia and the oasis states. Centralized rule was, therefore, very difficult. In 552, Bumin established his rule as Yili f^ M J Qaghan. His younger brother Shidianmi (Istami) ruled west of Altai as the chief of the western part of the empire from 562 to 576.9 The supreme ruling power was delegated peacefully, from 553 to 572, among the sons of Yili Qaghan. On the death of Taspar, however, conflict broke out The Turkish empire was at that time organized in a form of a feudal system. Shabolue (Ishbara) Shetu ^ f ^ s j ^ j j j f f i Qaghan, as supreme ruler, had his headquarters in the Otukan Mountains. Anluo ^|j |L, the son of Taspar, was enfeoffed in the territory in the basin of Tula River, called the Second Qaghan. Daluobian AJ2.J)S_ the Abo Qaghan occupied the territory northwest of the Otukan Mountains and east of Altai. His headquarters were known as the northern headquarters. Istami's son Dianjuefi^ /^  , called Datoujjjr gj? Qaghan, had his headquarters in what was formerly the land of the Wusun. East of the steppes lived the Xi ^ , Xi ^ and Khitan people, ruled by Ishbara's younger 8 Concerning the chronology of the Turkish rulers, see Table 2. 1. "Turkish Rulers." For the location of some important places, see Map 4. 9 Chavannes, E. 1969, pp. 160-163; Ma Changshou. 1957, pp. 16-20. brother ChuluohouyC .called Tuli^ -i((T61ish) Qaghan. In addition, there were other minor qaghans. Each of them had his own military capability, a state of affairs which was the source of continual rivalry and competition for power. On the surface they maintained unity but beneath the surface there were suspicion and ill feelings. There was an additional element that led to instability: the various groups of people conquered and ruled by the Turks were living under heavy oppression and being nomads themselves, these people were quick to rise in revolt when the opportunity presented itself.10 When Sui established its power, it soon made a correct assessment of the contemporary situation. This assessment concluded: 1) In the past both Northern Zhou and Northern Qi had been contending against each other and had divided north China between them. Both had communications with the Turks, not only because they feared a powerful enemy, but also because they hoped to decrease the burden of frontier defence on one of their borders by enlisting the support of the Turks. It was with that aim in mind that both had emptied their stores to supply the Turks. 2) The Turkish rulers were now competing among themselves. Their neighbors and their own subordinate tribes were all waiting for an opportunity to launch attacks or to rise in revolt 3) Furthermore the Turks had themselves to face a series of natural disasters, which greatly weakened their power.11 The Sui Victory over the Turks In the light of this assessment the Sui dynasty departed in various ways from the manner in which the Northern dynasties had conducted their relations with the 1 0 Ma Changshou. 1957, pp. 23-29; Cen Zhongmian. 1964, pp. 1-2. 1 1 SUIS, v84, pp. 1866-67; BS, v99, pp. 3291-92; Z277, vl75, pp. 5462-63. Turks. The first major step was Emperor Wen's abolition of the payment of silk products to the Turks. The change of attitude made Ishbara Qaghan feel humiliated. His wife, the Northern Zhou Princess Qianjin -f-'^ L , also resented Sui and wanted to use Turkish forces to take revenge for Sui usurpation of the throne from the Zhou house.12 In spite of their weakening power, however, the Turks still mounted constant incursions into Chinese territory. Emperor Wen insisted on meeting these incursions with a firm defence. This new, more defiant attitude was demonstrated by the stiff military resistance to the Turkish invaders by Chinese troops. The emperor also implemented two other measures: a) He built up and re-enforced frontier defences. On three occasions, in 581, 586, and 587, Sui improved the defences along the Great Wall on the northern frontiers.13 And b) he took advantage of the conflicts among the Turks to implement a policy of divide and rule. The major architect of the second policy was Zhangsun Cheng 0JC, who had stayed with the Turks for more than a year when he went on a mission to escort Princess Qianjin to the Turks at the end of Northern Zhou. At that time he had obtained first-hand information about the Turks' internal situation, a situation which suggested deep internal divisions. In 581 he suggested the policy of playing off the Turkish qaghans against each other. In accordance with the principle of making alliance with the far and attacking the near and abandoning the strong and joining the weak, he advised Emperor Wen to make Ishbara the striking target and to ally Sui with other qaghans.14 1 2 SUIS, v84, pp. 1865-66; BS, v99, p. 3291; ZZTJ, vl75, p. 5450. 1 3 SUIS, vl, p. 15; p. 23; p. 25; Cen Zhongmian. 1958, v2, p. 58. 1 4 SUIS, v51, pp. 1329-31; ZZ77, vl75, pp. 5450-51. The policy was soon put into practice. Embassies were sent to the other qaghans,15 and some results were achieved. In 582, when Ishbara led a large army to attack and plunder the Sui frontier, his plans to penetrate further south were abandoned when Datou Qaghan refused to follow him and left Zhangsun Cheng persuaded Tolish Qaghan to make a false report to Ishbara that Tiele and other tribes had revolted and were going to attack his headquarters. Anticipating trouble Ishbara retreated.16 The year 583 saw a major Sui military campaign against the Turks. The aim, as stated in an edict, was not to incorporate the Turks, only to extend Sui control up to the frontiers, thereby building up firm defences and barriers on the border passes and stopping the Turks from coming down to the south.17 Zhangsun Cheng continued to work on the qaghans. He succeeded in persuading Abo, who had suffered several defeats in battles with Sui, to stop his incursions. Abo then sent an embassy to the Sui court When Ishbara heard of Abo's double-dealing he attacked Abo's headquarters and killed his mother. With the assistance of Datou, Abo retaliated and recovered his territory. Tanhan Qaghan i^S^, who had formerly had a peaceful relationship with Abo, was deprived of his position by Ishbara and took refuge at Datou's place. Ishbara's younger brother Diqincha ^ClIP^^ also rebelled and went to Abo. Each of these qaghans, now being weakened, sent envoys to Sui, suing for peace and requesting Sui military assistance, but Emperor Wen refused.18 The Eastern and Western Turks became open rivals with the Eastern qaghanate headed by Ishbara and the Western qaghanate headed by Datou, Abo, Tanhan and Diqincha.19 1 5 SUIS, v51, p. 1331; ZZTJ, vl75, p. 5451. 1 6 SUIS, v51, p. 1331; ZZTJ, vl75, pp. 5456-59. 1 7 SUIS, v84, pp. 1866-67; BS, v99, pp. 3291-92; ZZTJ, vl75, pp. 5462-63. 1 8 SUIS, v51, pp. 1331-32; v84, pp. 1867-68; BS, v99, pp. 3292-93; TD, vl97, p. 1068; ZZTJ, vl75, pp. 5463-65. 1 9 Ma Changshou. 1957, pp. 26-27. Ishbara Qaghan of the Eastern Turks under Sui Hegemonic Rule With the two Turkish qaghanates in a state of open warfare, there came a turning point in Chinese relations with the Turks. In 584 Ishbara decided to make peace with Sui, and his wife Princess Qianjin asked to change her surname to Yang, the Sui royal family name, as a gesture of submission. Yang Guang the future Emperor Yang, proposed that Sui should seize the opportunity to attack the Turks, but Emperor Wen, ever cautious, followed his usual course in relations with the nomads and chose to assist Ishbara Qaghan. For an overview of the goegraphical locations of various places at that time, see Map 4. Unlike the situation towards the end of Former Han, however, when Huhanxie Chanyu of the Xiongnu went in person to the Han court to make peace with the Chinese, thus marking the end of the period of constant incursions into Han territory, Ishbara Qaghan had never paid any court-visit. Sui for its part, not as strong as Han, had yet to accomplish the unification of China. Ishbara, therefore, in his request for a peace settlement in 584, insisted on an equal footing with the Chinese. In his words: Our sheep and horses are the emperor's domestic livestock, and your silk products are our property. There is no difference between you and us. Nevertheless, the Sui envoy managed to demonstrate Chinese superiority by persuading the qaghan to perform the kowtow when he received the sealed letter from Emperor Wen, and to accept the status of vassal of the Chinese emperor.20 Having accepted Ishbara as an outer subject under his hegemonic rule, Emperor Wen, as a suzerain, took on the responsibility of assisting Ishbara politically, economically and militarily. In exchange he was able to use the Eastern Turks for 2 0 SUIS, v40, p. 1174; v51, p. 1332; v84, pp. 1868-69; BS, v99, pp. 3293-94; ZZTJ, vl76, pp. 5475-77. frontier defence, particularly as a deterrent force against the Western Turks. Then, in 585, when Ishbara was faced with repeated incursions from Datou of the Western Turks and from the Khitan, he requested permission to emigrate to the south of the Gobi Desert, to Baidaochuan f ^ J j / / outside the Great Wall. Sui permitted this, and sent troops to assist him, providing food, and clothing, carriages, uniforms, and drums and flutes for military music. Ishbara, now strengthened, launched a successful attack against Abo.- When his people were attacked by the Aba /*P$<^  (Apar) tribe, upon Ishbara's request Sui troops were sent in and inflicted a defeat upon the Apar.21 It was after this victory that Ishbara presented a letter formally abandoning equal status, recognizing himself as a Sui vassal and offering to send hostages and annual tribute to the Sui court Still he expressed the wish to maintain his own culture, that is, not to change the Turkish dress, hairstyle and language. Greatly pleased at this turn of events, Emperor Wen, in his edict claimed that China and the Turks were no longer two countries bound by a peace agreement but that the relationship was now one of unity between a ruler and subject He ordered that the news be spread widely throughout the country and that a ceremony be held at the ancestral temple to announce the Turkish submission. The honorable treatment of not directly referring to his personal name in any official document was granted to Ishbara. Princess Qianjin was given the Yang surname and the title of Princess Dayi Xj> . The titles of the Pillar of State (zhuguo ), and of Duke Who Pacifies the Country (anguo gong^)^zX^ ) w e r e conferred on Ishbara's son.22 In their relations with the Eastern Turks, the Chinese followed ceremonial practices under the tribute system. In 586 the Sui calendar was introduced to the 2 1 SUIS, v54, 1368; v84, p. 1869; BS, v99, p. 3294; ZZTJ, vl76, p. 5482. 2 2 SUIS, v84, pp. 1869-70; BS, v99, pp. 3294-95; TD, vl97, p. 1068; ZZTJ, vl76, p. 5483. The SUIS and the BS record that the change of the name of Princess Qianjin was in 585 while the following sources record it as in 584. Cf. SUIS, v51, p. 1332; ZZTJ, vl76, pp. 5475-76; CFYG, v974, p. 11440. Turks,23 as a symbol of their acceptance of Chinese rule. In the following year, the Chinese agreed to Ishbara's request to hunt in the Heng /fJL and Dai'fC commanderies. When Ishbara died in 587, the Sui court business was suspended for three days for mourning while an embassy was despatched to express condolences.24 The Eastern Turks under Sui Hegemonic Rule during the Time of Emperor Wen Ishbara's brother Chuluohou was established as qaghan, and Ishbara's son Yongyulii f^J^-M] (Ongul) became yehu ^ . f ^ (= yabghu, vice qaghan).25 Emperor Wen sent Zhangsun Cheng as an envoy to bestow the title of Mohe Hi?/ (Bagha) Qaghan on Chuluohou, and the tide of yehu on Ongiil.26 The Eastern Turks continued to act as a force against the Western Turks which was exactly what the Chinese had hoped for. Chuluohou went to battle under the Sui banner against Abo Qaghan, whose troops surrendered, thinking that the Eastern Turks had Sui military assistance on their side. After capturing Abo, Chuluohou requested Sui to decide on his treatment Some ministers proposed that he be executed. This was opposed by Zhangsun Cheng and another minister, however, and Emperor Wen avoided making a decision, hoping thereby to "cause the people from far away to come."27 2 3 SUIS, vl, p. 23; ZZTJ, vl76, p. 5485. 2 4 SUIS, v84, p. 1870; BS, v99, p. 3295; TD, vl97, p. 1068; ZZTJ, vl76, p. 5489; CFYG, v974, p. 11440. 2 5 SUIS, v84, pp. 1870-71. 2 6 SUIS, v51 (p. 1332) records that Sui invested the tide "Yehu Qaghan" on Yongyulii. However, the word "qaghan" after Yehu is crept in according to the commentary in the BS, v99, p. 3308, note 31. According to the ZZTJ , vl76 (p. 5490), Chuluohou was established as Mohe Qaghan and Yongyulii as yehu. 2 7 SUIS, v51, p. 1332; v84, p. 1871; BS, v99, p. 3295; TD, vl97, p. 1068; ZZTJ, vl76, pp. 5490-91. Chuluohou died in his western campaign in 588, and was succeeded by Dulan Qaghan who kept up a close tributary relationship with Sui. Border markets were set up at the Turks' request28 Emperor Wen, at this time, still could not trust the Turks with the task of frontier defence. After the conquest of Chen in 589 there was a proposal that the Turkish qaghan should be appointed "watch guard" (houzheng/f>^ J£), a proposal which Emperor Wen rejected on the ground that the Turks were not familiar with the geographical conditions in the frontier regions.29 Tributary relations entered a rocky period in 593 when a Chinese called Yang Qin ffitX^ fled to the Turks and told them the false story that Liu Chang ^] with his wife who was a Zhou princess, was planning to rebel against Sui power. He asked Princess Dayi to assist by sending troops to attack the Sui borders. Dulan Qaghan, believing this, stopped sending tribute and began making forays across the border. Emperor Wen sent Zhangsun Cheng to the Turks twice and finally got hold of Yang Qin. The court now made plans to get rid of Princess Dayi.30 Even though Sui was not very strong, since the situation of the Eastern Turks was not stable an opportunity for Sui arose when, in 593, Tolish Qaghan asked for a marriage with a Sui princess. He was Chuluohou's son, called Rangan ^Jf~ (Zamqan),31 and a potential rival to Dulan Qaghan. Sui laid out its conditions. If Tolish helped get rid of Princess Dayi, who had attempted to make an alliance with the Western Turks and had engaged in other anti-Sui activities, Sui would agree to the marriage. Tolish succeeded in persuading Dulan to kill Princess Dayi. Dulan then himself requested the hand of a Sui princess in marriage. Zhangsun Cheng favored granting Tolish's request rather than that of Dulan on the grounds that Dulan's 2 8 SUIS, v84, p. 1871; BS, v99, pp. 3295-96. 2 9 SUIS, v40, p. 1171; ZZTJ, vl77, p. 5517. 3 0 This is according to the record in the SUIS, v51, pp. 1332-33; ZZTJ, vl78, pp. 5542-43. The SUIS, v84 and BS, v99 have a different record which say that Yang Qin went to the Turks before 589 when Sui conqured Chen, and Dulan Qaghan caught Yang Qin and informed Sui of the incident. 3 1 SUIS, v51, p. 1333; TD, vl97, p. 1069; ZZTJ, vl78, p. 5543. The SUIS, v84 (p. 1872) says that Rangan was the son of Shetu. This is probably wrong, cf. Cen Zhongmian. 1958, vll , p. 512. submission to Sui was simply because of his ill feelings toward the Western Turks, and that, if the marriage with Dulan went ahead, Dulan would then make use of Sui power to oppress other Turkish rulers who would eventually rebel against Sui. Zhangsun Cheng suggested that Sui try to entice Tolish to move down to the south since Tolish did not possess a strong force and was, therefore, easy to deal with. As well, Sui could use him for defence against Dulan Qaghan. Emperor Wen agreed.32 A close relationship was established between Sui and Tolish. Following the marriage between Tolish and Princess Anyi^J^ , which went ahead in 597, Emperor Wen deliberately granted favorable treatment to Tolish with the intention of further stirring up hostilities between Dulan and Tolish. Tolish and his people moved south to the old headquarters of the Turks in the Otukan mountains. When this happened Dulan suspended tributary relations with Sui. Now with Tolish secure as a vassal, the Chinese were quick to use his people as a ready force to watch over the frontiers and guard against Dulan's incursions.33 In 599 Dulan Qaghan was making ready to invade China at Datong in northern Shanxi. This was reported by Tolish to the Chinese, who organized a military campaign. Hearing this Dulan allied himself with the Western Turks to attack Tolish. When Tolish suffered defeat at the hands of Dulan he considered taking refuge under the Western Turks, but Zhangsun Cheng persuaded him to go back to Sui. Greatly pleased, Emperor Wen conferred on Zhangsun Cheng the title of the Left Cavalry General (zuo xunwei biaoqi jiangjun if%p^%Z%^^%^^~ ) commissioned with special power to protect the Turks.34 Emperor Wen then pursued the strategy of appeasing Tolish by conferring on him the title of Yili zhendou Qimin 4<rJ %£f2j& R, Qaghan and the 3 2 SUIS, v51, pp. 1332-33; v84, pp. 1871-72; BS, v99, p. 3296; ZZTJ, vl78, pp. 5542-43. 3 3 SUIS, v51, p. 1333; v84, p. 1872; BS, v99, p. 3296; TD, vl97, pp. 1068-69; ZZTJ, vl78, p. 5558; CFYG, v978, p. 11494; 3 4 SUIS, v51, pp. 1333-34; ZZTJ, vl78, pp. 5563-64. building of the town of Dali X.<k!} for the Turks in Shuozhou • After Princess Anyi died, Princess Yicheng was sent to wed Qimin,35 and 20,000 Chinese troops were stationed to help Qimin Qaghan guard against the Western Turks.36 Zhangsun Cheng proposed that the Turks under Qimin should be resettled in Wuyuan i£J$- , between Xiazhou %J+\ and Shengzhou fj^J+j with the Yellow River to the north, as a line of defence against the incursions of Dulan Qaghan. The area was turned over to Qimin and his people for pasture land.37 In 602 still more walled towns were built in Jinhe^;^" and Dingxiang / ^ J | ^ for the Eastern Turks,38 and it appears that Qimin moved to this area.39 Now the relationship between Sui and the Eastern Turks under Qimin Qaghan became even closer. Like the second Huhanxie Chanyu of the Southern Xiongnu in Later Han, Qimin and his people were settled on the Chinese frontiers under the supervision of Chinese officials with military troops. His title of qaghan was maintained. In 600, when Dulan was killed by his own men and a time of civil strife ensued among his people, Qimin's men were despatched to entice those Turks to Sui, and many more Turks submitted to Sui.40 In the following years when Sui continued to battle against the Western Turks, Qimin Qaghan joined forces with the Sui military campaign of 601 against his Western rivals. After the Western Turks suffered a severe defeat Zhangsun Cheng advised Qimin to despatch his envoys to the Tiele and other tribes of the Turkish people who had been under the rule of the Western Turks in order to try to win them over. The envoys were successful and soon the Tiele and more than ten other tribes came to Sui. 3 5 SUIS, v2, p. 44; vSl, p. 1334; v84, pp. 1872-73; BS, v99, p. 3297; ZZTJ, vl78, pp. 5568-69; CFYG, v978, pp. 11494-95. 3 6 SUIS, v74, p. 1697; 77J>, vl97, p. 1069; ZZTJ, vl78, p. 5569. 3 7 SUIS, v51, p. 1334; v84, p. 1873; BS, v99, p. 3297; ZZ77,.-vl78, p. 5568. 3 8 SUIS, v74, p. 1697; Z277, vl79, p. 5572; Cen Zhongmian., 1958, v2, pp. 83-84; 1964, p. 12. 3 9 Cen Zhongmian. 1958, v2, pp. 83-84. 4 0 SUIS, v51, p. 1334. As a result Datou's rule of the Western Turks collapsed in 603 and he fled to Tuyuhun in modern Qinghai province. With the submitted nomads brought under his sway Qimin's power increased.41 Qimin Qaghan under Sui Hegemonic Rule during the Time of Emperor Yang After Emperor Yang ascended the throne, he attempted to incorporate Koguryo, some part of the Western Regions, and Linyi <r^-Cj (in present Vietnam) into the Chinese commandery-district system. But he drew a distinct line between Sui China on one hand and the nomadic Turks on the other, and intended to keep Qimin Qaghan and his people under loose administrative control as an outer subject within the greater Chinese hegemonic empire. At the beginning of his reign, when the Khitan made incursions into Yingzhou , Emperor Yang ordered Wei Yunqi Jfls^^L to lead 20,000 Turkish cavalry in a counter attack. The Khitan suffered a crushing defeat at their hands.42 Emperor Yang's intention of using those Turks who had submitted for frontier defence was shown more clearly when, in 607, Qimin Qaghan made repeated requests to be allowed to adopt Chinese dress and hair style. Even though his ministers suggested granting permission since it would be tantamount to a proof of successful sinicization, Emperor Yang refused and insisted on retaining the line between the Chinese and nomads, justifying his decision by the idea of different zones in the tribute system. In his letter to Qimin he explained his decision by saying that since the north of the Gobi Desert had not yet been made tranquil, expeditionary forces still had to be sent in from 4 1 SUIS, v51, pp. 1334-35; v84, pp. 1873-74; BS, v99, pp. 3297-98; ZZTJ, v l79, pp. 5571-72; p. 5590; p. 5600. 4 2 JTS, v75, pp. 2631-32; ZZTJ, v l80, pp. 5621-22. time to time to quell disturbances. As long as Qimin was sincere and obedient, it was unnecessary to change dress. To retain the good will of the Turks, Emperor Yang continued to grant Qimin the favor of not using his personal name while announcing him as a subject, and of giving him a rank above that of the Chinese princes. Emperor Yang also ordered the building of houses and a walled town for Qimin in Wanshoushu fa ^ , between Jinhe and Dingxiang.43 The balance of power between the Sui Chinese and the Turks changed gradually in favor of the latter during Qimin's later years, particularly with the incorporation of the Western Turks, the Tiele and other tribes into Qimin's sphere of power after the Western Turkish Datou Qaghan met his defeat in 603.44 Some Chinese ministers began to express their worry. They were concerned that Qimin, now with inside knowledge about China, might cause troubles in the future,45 and pointed out that it was not feasible for Sui to accept Qimin as subject, provide him with economic and military aid, and place the Turks inside the Great Wall. A suggestion was put forward that Sui should move the Turks outside of the Great Wall, and establish frontier garrisons so as to prevent a dangerous situation from developing.46 But the emperor was not persuaded. The Conflicts between the Turks and Sui In 609 Qimin Qaghan died and was succeeded by his son Shita-^-^ Qaghan. In accordance with the Turkish practice, Emperor Yang permitted his 4 3 SUIS, v3, p. 70; p. 71; v84, pp. 1874-75; BS, v99, pp. 3298-99; ZZTJ, vl80, p. 5627; p. 5632; p. 5641; CFYG, v974, pp. 11440-41. 4 4 SUIS, v84, pp. 1873-74; BS, v99, p. 3298; ZZTJ, vl79, p. 5600. 4 5 SUIS, v41, p. 1184; ZZTJ, vl80, pp. 5632-33. 4 6 SUIS, v60, p. 1459; ZZTJ, vl81, p. 5661, CFYG, v990, p. 11632. Concerning the year of the suggestion, see Cen Zhongmian. 1958, v3, p. 93. marriage to Princess Yicheng who had formerly married to Qimin.47 Turkish power continued to grow to such an extent that minister Pei J u ^ £ & suggested fomenting discord among Shibi's people as a way of weakening his force. As a means to this end, the court offered a Chinese princess to Shibi's younger brother and proposed to confer the tide of southern qaghan on him, but he was afraid to accept When news of the Chinese plan reached Shibi he quite naturally began to distrust Sui. Pei Ju further suggested that the court should entice Shishu Huxi ~£%%., a capable assistant to Shibi, to come to Mayi J§> &J to trade with the Chinese. The Chinese circulated reports that whoever arrived first would get the best deal. Huxi and his men therefore hurried there without first informing Shibi, only to stumble into a Chinese ambush and get killed. When Sui tried to convince Shibi that Huxi was going to defect, and that since the Turks were Sui subjects, the Chinese had the right to punish whoever rebelled against the qaghan, Shibi saw through the subterfuge and suspended tributary relations with Sui. This time not only did the chicanery of Pei Jue fail, but it also brought disaster on the Chinese. In 615, when Emperor Yang went on a hunting trip, Shibi led his troops, besieged Emperor Yang in Yanmen/^ J9^ , and captured thirty-nine of forty-one walled towns in the commandery of Yanmen. Emperor Yang had to order the recruitment of rescue troops throughout the empire, and to announce the abandonment of the costly campaign against Koguryo, which had gone on for several years. To alleviate the situation Sui had to ask Princess Yicheng to give false information to Shibi. She told him that there was an emergency on the northern border, so Shibi finally withdrew his troops.48 4 7 SUIS, v84, p. 1876; BS, v99, p. 3299; ZZTJ, vl81, p. 5647; CFYG, v974, p. 11441. The CFYG, v999 (p. 11718) says that Qimin died in the tenth year of Daye (614), which might be wrong. 4 8 SUIS, v4, p. 89; v63, p. 1492; v67, p. 1582; v84, p. 1876; BS, v99, p. 3299; TD, vl97, p. 1069; ZZTJ, vl82, pp. 5697-99; CFYG, v990, p. 11633. The Turkish attack was the final straw. Coming on the top of unsuccessful campaigns against Koguryo, and the turmoil inside the country caused by various internal uprisings, the Sui dynasty went into a steep decline and soon after met its end. Sui and the Western Turks After 583 the Western Turks, whose traditional territory was not so close to China as was that of the Eastern Turks, extended their sway over a large area in the Western Regions. In 585 Emperor Wen sent a mission to Abo Qaghan to appease this increasingly powerful neighbor who had joined the Western Turks.49 But wars continued between Sui and the Western Turks. Abo himself was defeated in 587 and Datou, the supreme leader of the Western Turks, was defeated in 603 under the joint forces of Sui and the Eastern Turks. In Emperor Yang's time, the qaghanate was under Chuluo Qaghan whose lands were in the Iii Basin. His mother was a Chinese. After his father Nili i/tLlRj Qaghan died she married Nili's younger brother, according to Turkish custom. Both went to pay a visit to the Sui court at the end of Emperor Wen's reign and remained there because of the turmoil in the qaghanate. At the beginning of Emperor Yang's reign Chuluo launched attacks on the Tiele and exacted heavy taxes from them after his success. The Tiele then rebelled and set themselves up as an independent entity.50 Seeing that Chuluo was in a difficult situation, the ever scheming Pei Ju urged the court to seize the moment. Acting on his advice, in 608 Emperor Yang sent Cui out to the Western Turks that the Eastern Turkish qaghanate had submitted to Sui in as an envoy bearing an imperial letter to Chuluo. Cui pointed 4 9 SUIS, v l , p. 22; ZZTJ, v l76, p. 5482. 5 0 SUIS, v84, pp. 1876-77; BS, v99, pp. 3299-3300; ZZTJ, vl80, pp. 5622-23. order to use Sui forces against the Western Turks. He warned the qaghan that allied Chinese and Eastern Turkish forces were massed for an attack against the Western Turks, but Chuluo's mother had begged the emperor to persuade Chuluo to submit. If Chuluo accepted the status of vassal, Cui continued, his country would be at peace and his mother would enjoy a long life. Otherwise Chuluo's mother would be executed and the Sui-Qimin joint forces would eliminate the Western qaghanate. Faced with what seemed to be an inescapable reality, Chuluo had no alternative. He performed the kowtow and received Emperor Yang's letter on his knees. As well he agreed to join the Chinese in an attack on the Tuyuhun and sent a tributary mission to Sui.51 But Chuluo had not actually fully accepted Chinese suzerainty. When Emperor Yang, in his tour of 610 of inspection in the west, asked Chuluo to pay a visit to him in Dadouba Xrf-^v valley, Chuluo refused to do so. Emperor Yang was furious at this act of outright disobedience, but had no way to deal with the problem until a timely falling-out among the Western Turks provided the Chinese with an opportunity for attack. When, in 611, the tribal leader Shegui ^ J j | of the Western Turks sent an embassy to Sui requesting a marriage alliance Pei Ju pointed out to Emperor Yang that Shegui was the grandson of Datou, but he now found himself in a subordinate position to Chuluo. In all probability, Pei Ju argued, he would wish to ally himself with Sui, so Pei suggested that Sui should give Shegui generous gifts and confer on him the tide of great qaghan. His scheme of divide and rule succeeded in making Shegui launch attacks against Chuluo, which in turn made Chuluo come in person to pay respect to Emperor Yang at the end of the year.52 The following year saw the incorporation of Chuluo's people into the Sui hegemonic empire. They were divided into three groups: a) a group of more than 5 1 SUIS, v84, pp. 1877-78; BS, v99, pp. 3300-01; ZZTJ, vl81, pp. 5636-37. 5 2 SUIS, 84, pp. 1878-79; BS, v99, pp. 3301-02; TD, vl99, p. 1077; ZZTJ, vl81, pp. 5654-55; CFYG, v990, p. 11633. 10,000 physically weak people led by Chuluo's brother, who were settied in the commandery of Huining , b) some stronger forces who were stationed in Loufan ^ / c f ^ for frontier defence, and c) Chuluo in command of 500 calvary men. This latter group escorted Emperor Yang on his tours of inspection, and joined in the military campaigns against Koguryo.53 In 614 Sui gave the hand of a princess to Chuluo in marriage agreement with imperial gifts of tens of thousands of pi ]A of silk. Sui also expressed its intention of helping him recover his territory, no doubt with the long-term aim of making Chuluo responsible for frontier defence. The plan aborted, however, with the collapse of Sui.54 A Period of Chinese Civil War After the siege of Yanmen Sui was on the verge of total collapse. The Turks soon recovered their comparatively strong regional position and the Western Turks again expanded, incorporating the Tiele and the various oasis states, which in turn became tributaries of the Turks.55 The Eastern qaghanate under Shibi Qaghan also expanded to bring under its sphere of influence the Khitan, Shiwei ^~ in the east and the Tuyuhun and Gaochang (Karakhoja) in the west. This revival of Turkish power was due mostly to Chinese decline. In the turbulence of civil war at the end of Sui, the Chinese