National Institute for Asia and the Pacific | ANU Home | Search ANU
The Australian National University

Home  »  Publications  »  Rafe de Crespigny


South China in the Han Period

Internet edition 2004


Rafe de Crespigny

Faculty of Asian Studies, Australian National University

NOTE:  Much of the material in this paper, which was presented in 1988, was developed further in "South China under the Later Han Dynasty," the first chapter of my Generals of the South, published by the Faculty of Asian Studies at the Australian National University, Canberra, in 1990. More detailed discussion and annotation may be found there, together with a substantial bibliography and maps.
© The whole work is copyright. Apart from any fair dealing for the purpose of private study, research, criticism, or review, as permitted under the Copyright Act, no part may be reproduced by any process without written permission. Enquiries should be made to the Faculty of Asian Studies at the Australian National University


There is some awkwardness in my speaking, as a historian, to a conference of archaeologists, for the material we deal with is, by definition, very different, and the questions that we ask of it can again be very dissimilar. Moreover, the period of the Han dynasty and the short-lived Qin which preceded it, from the third century BC to the third century AD, is a part of the iron age, not the bronze age, of China.

     On the other hand, I hope it may be of some value if I offer a few comments on the expansion of Chinese imperial power into the lands south of the Nan Ling ranges and, in particular, if I present some account of the written material which is available in early Chinese sources and which may complement or illuminate the findings of physical material.[1]

     When the state of Qin unified the Chinese world in the second half of the third century BC, it included in that conquest the territories of the state of Chu, which had controlled the basin of the middle and lower Yangzi, including most notably the great tributary of the Xiang in present-day Hunan. However, although there was certainly trade across the mountains and along the coast, it was not until the armies of the First Emperor entered and traversed that region in the years between 220 and 214 that the region of present-day Guangdong and Guangxi came under political control from the north, together with northern Vietnam and, for at least a short time, part of Fujian.

     With the fall of Qin a few years later, Zhao Tuo, who had been appointed successor to the Commandant (wei) of Nanhai commandery about the bay of Canton, seized and blocked the passes through the Nan Ling and established his own kingdom of Nan-Yue.[2] This émigré Chinese state maintained its independence for almost a hundred years until the region was again conquered by the armies of Emperor Wu of Han in 112 BC, an achievement followed by acquisitions in the southwest of present-day Yunnan, notably the Tian kingdom in the region of the Tian Lake.[3]

     During the course of the civil war which brought about the destruction of the "usurper" Wang Mang in the early first century AD and the restoration of the Later Han dynasty, the far south of China established a short-lived independence, concluded this time by the campaigns of the celebrated general Ma Yuan between 40 and 43 AD. The name of Ma Yuan has been long remembered as that of a culture hero in the south, and his celebrated opponents, the Zheng or Tr'ung sisters, though themselves of Chinese origin, are still honoured in modern Vietnam as leaders of resistance against the north. [4]

     Finally, at the end of the second and beginning of the third century AD, as the government of Later Han collapsed into civil war, a local authority was established by Shi Xie (Sri Riep), with his headquarters in the region of present-day Hanoi.[5]  By 220, however, the successor state of Wu of the Three Kingdoms, based upon the Yangzi, re-established northern influence in the region, and the power of the Shih family was destroyed after Shih Hsieh's death in 226.

     Two particular points may be made about the history of far southern China during this period. Firstly, even during the periods of independence, the major initiatives in these southern regions were in the hands of immigrants from the north, who maintained a Chinese culture in contrast to that of the aborigines of the south. Zhao Tuo came south under the aegis of Qin, the Zheng sisters were of Chinese family, and the Shi clan, which arrived in the south at the time of Wang Mang, had maintained cultural contact with the north and included a number of local imperial officials.

     Second, the Chinese invasion of the south was a cultural, as much as a political or military, aggression. Most blatantly, this may be symbolised by Ma Yuan: not only did he kill the Tr'ung sisters and the other chieftains who had rebelled against Han, he had the sacred drums of his enemies melted down to be recast as the ideal statue of a horse, which was presented and displayed thereafter at the imperial capital, Luoyang.[6]  And just as the Chinese destroyed the alien iconography, so one may look further upon the accretion of the Ma Yuan legend as their attempt to establish dominance in the spiritual, as well as the physical, world of the south.

     More insidiously, but perhaps even more effectively, the peaceful administrators of Han sought to "Confucianise" the non-Chinese people under their authority. In doing so, they not only established schools to teach the texts, the philosophies and the morality of the north, but they also attacked the influence and put down the manifestations of the traditional spiritual leaders of the southern community, described variously as wu, magicians or shamans. The official Chinese histories contain biographies of a number of these worthy and "lenient" officials, and we are told how they regulated the marriages and sexual conduct of the people, taught the care of children, and established proper customs of filial duty and family mourning. [7] From the Chinese point of view, this was an excellent and appropriate influence for civilisation -- but for the people of the south and for their spiritual leaders, this was the destruction of culture: birth, marriage and death are surely the central points of ritual and belief for any society, and it was these areas which were attacked and taken over by the Chinese invaders.

     So the standard histories of the time record chronicles of invasion and resistance, and biographies of missionary officials. Incidents and anecdotes are scattered through the texts of the histories, and some are collected in Accounts of the barbarians of various regions and tribes. At the same time there are also formal geographical descriptions, and these provide more general, and even statistical, information about the territories known to the Chinese.

     Among the standard histories, there are three types of descriptive texts. First, in chapter 129 of Shi ji, dealing with the period of the late second century BC, and in chapter 28 of Han shu, for the end of the first century BC, there are descriptions of the various regions within the empire. [8]  Second, in chapters dealing with non-Chinese peoples, and notably in the Accounts of the Western Regions of central Asia in Shi ji 123, Han shu 96A-B, and Hou Han shu 88, there are descriptions and itineraries of the lands beyond the frontiers of the empire;[9]while a concluding passage of Han shu 28 describes the countries which may be reached by sea voyage south from Hainan.[10]

     Thirdly, in Han shu 28 and in chapters 19 to 23 of the treatises compiled by Sima Biao now attached to Hou Han shu, there are lists of administrative geography, giving the names of the provinces (zhou), commanderies (jun) and counties (xian) for all the empire. These lists may be related to censuses compiled in 2 AD for Former Han and about 140 AD for Later Han, and they contain not only the names of the administrative units, with the population of each commandery, but also notes on the local production of such items as salt, iron and other minerals which were of importance to the economic management of the state.

     I shall not discuss the first categories of description in great detail. The regional description of China has been translated very considerably by Nancy Lee Swann, and the cities and routes into central Asia have been studied in detail by A.F.P. Hulsewé and Michael Loewe. There are, however, some points of confusion in handling these texts; in particular for the lands within the empire, there are elements of uncertainty as to how much the information given by Shi ji and Han shu should be differentiated according to their ostensible dates, and to what extent the geographical descriptions represent a conflation, with some rearrangement, of the same or similar bodies of material. The question is most difficult for central Asia, where accounts from Shi ji and Han shu must also be reconciled with parallel passages in Hou Han shu and in the book Wei lue of the third century AD, preserved in substantial fragments by the commentary of Sanguo zhi. [11]

For the lands and seas south of China, the short text at the end of Han shu 28 has been translated and studied by Wang Gungwu, together with supplementary information from Sanguo zhi, but a number of questions posed by the text can only be answered through future discoveries from archaeology. It may be interesting to note, however, that the trade which passed through the Red River basin in the time of Shi Xie at the end of the second century AD, and which is comparable to that described in texts dealing with earlier periods, included pearls, both from the Hainan coast and the countries of the southeast, shells, tortoise-shell, incense, kingfisher feathers, ornamental glass, lapis lazuli, rhinoceros horn, ivory, bananas, coconuts, longan and lychees, and cloth of vegetable fibres.

     One should also observe that the Chinese sources are, naturally enough, concerned primarily with the goods obtained by trade from overseas, and the main routes followed the valleys of the Xiang River and the other southern tributaries of the Yangzi, crossing the Nan Ling divide most notably at the Ling Qu "Magic Trench" constructed by the First Emperor of Qin, and thereafter following the West River and/or the southern coast. Though this was the main line of access for the Chinese, it should not be assumed that it was the only route - on the contrary, there is incidental evidence that local trade was regularly maintained along the coast southeast from the mouth of the Yangzi, with a Chinese trading station by present-day Fuzhou, and it seems clear also that there was trade and movement along the Red River into the west of China Proper and further north towards present-day Gansu and also west into Burma.

     The third category of texts mentioned above, the treatises of administrative geography for Former and Later Han, is far more valuable than a mere list of names. Some fifteen hundred counties are given by the Former Han list, and twelve hundred for Later Han (the immense majority of them, of course, being the same in both texts), and over a thousand of those places can be identified onto a modern map. In some cases, the identification has been made possible by modern archaeology, but most are based upon local and scholarly tradition, and it has been possible to construct atlases of historical geography for the dynasties since Han with a very high degree of accuracy. [12]There are, of course, some variations, and there are areas of China on the northern frontier and on the North China plain where opinions may differ, but the general outline and most of the individual points are reasonably well agreed.

     One of the maps discovered at Mawangdui, near Changsha in Hunan, has provided an excellent test for the accuracy of this literary tradition. The map, compiled about 180 BC, provides a recognisable view of the southern frontier of Han against the kingdom of Nan-Yue based upon present-day Guangzhou, and it shows eight county cities in the region of present-day southern Hunan. A comparison of the Mawangdui map with modern reconstructions of the sites of those cities shows an almost perfect correlation. [13] On this basis we may feel quite confident of the ancient political geography.

     Considering some of the discussion which has taken place in other sessions of this conference, we may note again that the Han government was concerned with the production and sources of valuable minerals, and these were recorded for each relevant commandery and county. Such commodities could be obtained by overland trade to the west, but the empire of Han also expanded to take over sources of supply. In the territory of present-day Yunnan province there were many such sites, producing iron, copper, tin, silver and lead, and they provided a substantial motive for Han to expand against the Tian, the Ailao, and other peoples of that otherwise difficult region. [14] In Table 1, I present a list of mineral production other than salt and iron, as recorded in the two treatises.

     Finally, however, we can look upon the treatises of administrative geography as evidence for the expansion of Chinese power, and of the changing demography of the empire. Naturally enough, many of the Chinese who came to the south were seeking to evade the control of the imperial government, not to extend it, and they pressed beyond the effective frontier, occupying the lands of the aboriginal people and frequently intermarrying with them, in hope of escaping the obligations of taxation and labour or military service. The imperial government, however, followed these pioneers, and in many areas the establishment of a new county administration is a sign of formerly independent colonisation which has now reached the critical point where it is worth-while to bring under control. The process can be observed across many centuries. [15] 

     For the Han period, however, we have also specific figures for the registered population, arranged by households and by individuals, man, woman and child. Obviously, the figures were never perfectly accurate, but there is good evidence that regular serious counting took place. In the heartlands of the empire, we may accept the figures as indicating the real number of people in the territory concerned; on the frontiers, we may take them as the registered population, those under effective government control - with an unspecified number of non-Chinese people still remaining outside the reach of the government account.

     In Table 2, therefore, I present a summary of the population changes in different regions of the empire south of the Yangzi as indicated by the figures for Former Han, at the beginning of the Christian era, to those of Later Han, in the middle of the second century AD. With the exception of the region of present-day Yunnan, for which there must be some sense of disbelief, the picture presented by these figures is one of energetic colonisation and political expansion by the Chinese. South China at the end of Han was largely under the control of the newcomers.

     On the other hand, the very success of this development created tension within the unified empire itself. Though the lands of the south contained only about a fifth of the population of the empire as a whole (some eight million out of more than forty million), it appears that this was the "critical mass" required for the establishment and maintenance of the southern state of Wu under the Sun family in the early third century, and this separatist state confirmed the dissolution of the empire and paved the way for centuries of division which followed the fall of Han. By an ironic circle of development, migration and expansion from north China to the south during Qin and Han brought renewed independence to the region when central government collapsed in the north at the end of the second century AD. Indeed, from the fourth century to the seventh, as the north of China was all but overwhelmed by the political and military power of the steppe, it was the southern dynasties which played the major role in maintaining Han tradition in east Asia. [16]



Sources of minerals other than salt and iron recorded by the geographical treatises of
Former Han [2 AD] and Later Han [c.140]

Commodity                         F*                     L*                                         Commandery unit                   approximate
                                           HS 28               HHS 23                                  and county                                  Lat. Long.

copper                                                                                                       Yuexi
                                             F:1600              L: 3511                                Jiangdu                                      28°N 102°E

                                             F: 1601                                                          Laiwei                                       22°N 103°E
                                             F: 1601             L: 3513                                 Jianyuan                                    25°N 103°E
                                                                      L: 3513                                 Bengu                                       23°N 103°30'E

                                                                                                                   Jianwei Dependent State
                                                                      L: 3516                                 Zhuti                                        27°N 104°E

gold                                                                                                           Yongchang
                                                                      L: 3514                                 Bonan                                     25°N 99°30'E

lead                                                                                                           Yizhou
                                            F: 1601             L: 3513                                 Lügao                                     24°N 103°30'E
                                            F: 1601             L: 3513                                 Bengu                                     23°N 103°30'E

silver                                                                                                         Jianwei Dependent State
                                            F:159                L:3516                                  Zhuti                                           27°N 104°E

                                            F: 1601             L: 3513                                 Lügao                                    24°N 103°30'E
                                            F: 1601             L: 3513                                 Bengu                                    23°N 103°30'E
                                                                     L: 3513                                 Shuangbo                               24°30'N 102°E

tin                                                                                                              Yizhou
                                           F: 1601             L: 3513                                   Lügao                                    24°N 103°30'E
                                           F: 1601             L: 3513                                   Bengu                                    23°N 103°30'E

                                                                    L: 3506                                   Xi                                          33°N 110°E


*    In this pair of columns, F indicates that the mineral is recorded in the treatise of Han shu 28, and the number following indicates the page of the Beijing edition on which the entry may be found; similarly, the designation L: 3506 indicates that the mineral is recorded in treatise 23 of Sima Biao, attached to Hou Han shu, and that the entry may be found on page 3506 of the Beijing edition.

We are told by Hou Han shu 86, 2849, that the territory of the Ailao people, west of the Dali Lake by the present-day Burmese frontier, incorporated into the empire in 69 AD, produced quantities of iron, copper, gold, lead, silver and tin, besides other rare and valuable commodities. The Later Han treatise, however, contains no reference to the exploitation of those minerals in that region.
Much further away, the land of Tianzhu, identified as the Indian sub-continent, also produced these metals.



Population increase in regions south of the Yangzi: 2 AD to c. 140

Modern provinces                        Han                                           individuals (millions)                       increase
or regions                              commanderies                                   2 AD            c. 140 AD                     per cent

Zhejiang/                               Wu and
south Jiangsu                          Kuaiji                                                  1.0                    1.2                             20%

South Anhui                             Danyang                                            0.41                  0.63                            55%

Jiangxi                                     Yuzhang                                            0.35                  1.67                           375%

South Hubei/                         Nan, Jiangxia
North Hunan                         and Wuling                                             1.1                    1.26                            15%

South Hunan                         Changsha, Lingling
                                             and Guiyang                                         0.53                  2.56                          385%

Guangdong                           Nanhai and Cangwu                                0.24                  0.72                          200%

Yunnan                                Yizhou and Yongchang                             0.6                    2.0                            300%


It is difficult to accept the figures for the commandery of Yongchang, which included part of the Former Han commandery of Yizhou but was separated by Later Han in 69 AD, ostensibly as a result of the accession of the Ailao people to the imperial sway. The Ailao people numbered some half a million at that time. There may have been further colonisation, associated particularly with the mineral works of the region and with trade along the route to Burma, but the population increase in this isolated region is still surprisingly high.



Map of the South China at the end of Han [GIF file 88 kb]



[1]     The Standard Histories for this period are Shih chi (SJ), compiled by Sima Qian early in the first century BC, Han shu (HS)compiled by Ban Gu in the first century AD, Hou Han shu (HHS), with annals and liezhuan by Fan Ye of the early fifth century, combined with treatises zhi from the Xu Han shu of Sima Biao of the third century, and the Sanguo zhi (SGZ)of Chen Shou of the third century, with commentary (PC) compiled by Pei Songzhi of the early fifth century.
These works are cited in the punctuated editions of Zhonghua shuju, first published in Beijing during the 1950s, and reprinted at intervals thereafter.

[2]     The history of Nan-Yue appears in SJ 113 and HS 95. On the general expansion of Han, see Yü Ying-shih, Trade and Expansion in Han China, California UP 1967, and his chapter "Han Foreign Relations" in The Cambridge History of China, Volume 1: The Ch'in and Han Empires 221 B.C. - A.D. 220 (hereafter Cambridge Han), 377-462. The discovery of a number of tombs of middle-ranking and minor officials of Nan-Yue is reported in Kaogu xuebao 40 (1974), and that of a royal tomb in Kaogu 1984.3.

[3]     The history of the people of the southwest appears in SJ 116, HS 95 and HHS 86. On the conquest of the region by Han, see also SJ 117, the biography of Sima Xiangru, translated by Yves Hervouet, Paris 1972.

[4]     The biography of Ma Yuan is in HHS 24; see also Hans Bielenstein, The Restoration of the Han Dynasty, volume III, in Bulletin of the Museum of Far Eastern Antiquities (BMFEA) 39 (1967), 63-65, and K.W. Taylor, The Birth of Vietnam, University of California Press 1983, 37-48.

[5]     The biography of Shi Xie is in SGZ 49 (Wu 4). See also Taylor, The Birth of Vietnam, 70-80, and Jennifer Holmgren, Chinese Colonisation of Northern Vietnam: Administrative geography and political development in the Tongking Delta, first to sixth centuries A.D., Canberra 1980, 72-77.

[6]     HHS 24, 480. For a summary of this form of activity, see the article by Miyakawa Hisayuki, "The Confucianization of South China," in Arthur F. Wright (editor), The Confucian Persuasion, Stanford UP, 1960, 21-46.

[7]     See, for example, the biography of Wei Sa in HHS 76, of Diwu Lun and Song Jun in HHS 41, and of Luan Ba in HHS 57.

[8]     The relevant text of SJ 129 has been rendered by Nancy Lee Swann, Food and Money in Ancient China, Princeton UP 1950, 437-451.

[9]     On these texts, see particularly A.F.P. Hulsewé and M.A.N. Loewe, China in Central Asia: The early state 125 B.C. - A.D. 23, Leiden 1979.

[10]    See Yü, Trade and Expansion, 172-175, and also particularly Wang Gungwu, "The Nan-hai Trade"; A Study of the Early History of Chinese Trade in the South China Sea," in Journal of the Malayan Branch of the Royal Asiatic Society 31.2 (1958), 1-135 at 19-20, and Paul Wheatley, The Golden Khersonese, Kuala Lumpur 1961, 8-13.

[11]    SGZ 30, 858-63 PC quoting the Wei lue of Yu Huan of the third century. The major early work on the question is that of Edouard Chavannes, "Les pays d'Occident d'après le Wei lio," in T'oung Pao 6 (1905), 517-531. K.H.J. Gardiner of the Australian National University and D.D. Leslie have lately been working further on this field: e.g. their article "Chinese Knowledge of Western Asia during the Han" in T'oung Pao 68, 4-5 (1982), 254-308.

[12]    The most useful of these reconstructions is Zhongguo lishh ditu ji, edited by T'an Ch'i-hsiang ; volume II, Shanghai 1980, deals with the Qin and Han periods.

[13]    Rafe de Crespigny, "Maps from Mawangdui," in Cartography (Journal of the Australian Institute of Cartographers) 11.4 (September 1980), 211-222.

[14]    Though jade (yu), nephrite, was traded from central Asia, and it is recorded that the county of Lantian in Jingzhao,, near the present-day city of the same name in Shensi, produced a "beautiful jade" (meiyu: HS 28A, 1543, and HHS treatise 19, 3403), we are also told that hills in the east of Huiwu/Kuaiwu in Yuexi commandery (about 26°30' N, 102°30' E) produced a jade (pi), of a blue-green (qing) colour, presumably nephrite: see HS 28A, 1600, and HHS treatise 23, 3511 and 3512 note 8; cf. also the entry for neighbouring Qingling county, which refers to "golden horses and jade hens."

[15]     On the census of Han and later dynasties, see Bielenstein, "The census of China during the period 2-742 A.D.," in BMFEA 19 (1947), 125-163. On the use of records of the establishment of counties as a guide to colonisation, see Bielenstein, "The Chinese colonisation of Fukien until the end of T'ang," in Søren Egerod and Elsie Glahn (editors), Studia Serica Bernhard Karlgren Dedicata, Copenhagen 1959, 98-122, and de Crespigny, "Prefectures and Population in South China in the first three centuries AD," in Bulletin of the Institute of History and Philology, Academia Sinica, Taiwan, 40.1 (1968), 139-154.

[16]   A recent work on this theme is Charles Holcombe's In the Shadow of the Han: literati thought and society at the beginning of the Southern dynasties, Hawaii UP, 1994.


Top of page