1Since the Yuan period (1279–1368), imperial states have moved several million Chinese people to Yunnan, establishing a new demographic pattern and transforming local society. Central administrative hierarchy, migrants, taxation, education, and economic and religious infrastructures introduced Chinese institutions into Yunnan, leading to the so-called sinicization (huahua) of the region. At the same time, native peoples' myriad influences on the Chinese immigrants served as a sort of indigenization (tuzhuhua) of the immigrants. Sinicization and indigenization were two sides of the process through which a middle ground was negotiated.

2This chapter will examine the administrative hierarchy that was imposed on Yunnan by Chinese states and that reorganized local power structures. To handle the so-called Southwestern Barbarians, Chinese states borrowed, created, adopted, and developed a special kind of administrative system, that is, the native chieftain system (tusi zhidu). Native chieftains were assigned imperial ranks, titles, and posts to rule their respective areas and peoples, and imperial states seldom intervened in the internal affairs as long as there were no large rebellions. In a word, rule based on native customs (yinsuerzhi) served as the principle on which the Chinese frontier policies concerning the Southwest relied.

3The power balance between imperial China and native regimes was dynamic, meaning that the middle ground shifted frequently. As soon as the imperial state felt strong enough, it was determined to transform the frontier society into another Chinese society. It was through this process that native regimes were suppressed, if not destroyed, and native peoples became the minority and their cultures, marginal. With several centuries of effort, direct central administration gradually took the place of the native chieftain system. The establishment, institutionalization, and eventual reform of the native chieftain system in the course of more than half a millennium revealed much about the Chinese imperial consciousness, determination, and power to "civilize" frontier societies and ethnic peoples.

The Frontier Commandery (Bianjun) System

4The frontier administrative system can be traced back to the Qin-Han period, especially during the Han Dynasty. As soon as the Chinese empire expanded into the frontier areas, it began to create an administrative system for these areas different from that of China proper. Although the guiding principle of Chinese administration of frontier peoples was to rule based on native customs, the final goal of the Chinese "civilizing" project was to reproduce a Chinese society on the frontiers. Therefore, whenever central states thought the time to be ripe, they quickly moved to abolish, regulate, or transform local systems, sometimes violently. By reviewing the historical process of frontier administrative institutions, this section will follow the increasing Chinese penetration into Yunnan.

5Zhuang Qiao, the Chu general who conquered the Dian Kingdom in the late third century BCE, established the ruling strategy for central states. It is recorded that after the conquest, Zhuang Qiao claimed himself king of Dian. However, he decided to follow local customs by changing dresses (bian fu cong qisu) and ruled according to local customs.1 His strategy and practices were followed by imperial rulers and his respect toward native customs was mentioned again and again by Chinese imperial rulers. Gradually, a kind of special administrative hierarchy was invented, reformed, and developed to deal with frontier areas and ethnic peoples. In the case of Yunnan, and even entire Chinese frontier areas, a clear historical logic could be identified, that is, to enhance Chinese control while managing frontier security and stability, even though the results varied locally.

6The Han Dynasty followed the junxian (commandery and county) system invented by the Qin while the tributary system was developed to handle many non-Chinese "barbarian" peoples stationed at the frontier areas. Since these peoples differed from the Chinese, and differed from the supposed barbarians beyond Chinese frontiers, the Han court created a new system to deal with these "semi-Chinese" and "semi-barbarians." The new frontier administrative system, as examined by Yu Yingshi, can be classified into three types: shuguo (subject state, or dependent state), junxian, and bu (military unit).2 Not only did the Western Han basically establish shuguo in the Southwestern Barbarian areas, it also added the junxian system. Four commanderies were established during the reign of Emperor Wu, namely, Zangke, Shenli, Yuexi, and Yizhou. According to Yu Yingshi, only relatively sinicized "barbarians" would be put under the junxian system,3 therefore, the Han considered the Southwestern Barbarians who were put under this system to be relatively civilized. The Eastern Han continued the cause of southern expansion, and established more administrative units with each community of "barbarians" it conquered.

7Normally Chinese people within China proper were ruled by the junxian system, and were subjected to taxation and levies of all kinds. But the junxian system in the Southwestern frontier areas differed tremendously. First of all, although central officials were assigned to these posts, they sometimes did not enter the designated commandery or county. Instead their governorship largely relied on native leaders, as native chieftains or princes were largely maintained and assigned honorary titles such as king, marquis, and the like. Second, the indigenous peoples were asked to pay relatively low tax under both shuguo and junxian in China proper. In some cases, the ratio of taxation was negotiated between the Han office and local peoples. This is the so-called frontier commandery system (bianjun zhidu) of the Han Dynasty.

8The invention of the bianjun system was designed to achieve a dual goal. On the one hand, the indigenous people were under the governance of the Han court, and were expected to become "civilized" by adopting Chinese culture; on the other hand, frontier "barbarians" were crucial to Chinese defense. Such a system demonstrated a vivid application of Han China's jimi strategy—ji is a halter for horses and mi is a halter for buffalos. With the ji and mi, herdsmen are able to have animals walk in the expected track while allowing them some freedom. To some extent, with the bianjun system, the Han court managed to impose reins on the Southwestern Barbarians.

9This bureaucratic imperial system led to more frequent interactions between the Southwestern Barbarians and the Chinese, which in turn caused many dramatic changes in native societies. Drastic changes within native societies consequently became a source of tension, conflict, and rebellion. While scholars of China think highly of Chinese culture and rule for supposedly bringing benefits to the indigenous peoples, local rebellions tell a different story.

10In 86 BCE, the Liantou and Guceng tribes in Western Yizhou commandery (Heqing and Jianchuan) killed Han officials. In the same year, over 30,000 Southwestern Barbarians in Zangke, Tanzhi (Xingyi and Zhenfeng, Guizhou), and Tongbing (Mile) rebelled. Three years later, in 83 BCE, the Guceng tribe and other tribes in Yeyu County (Dali) rebelled again. This uprising was of such a large scale that the assigned Han general did not dare to march forward. As a result, the governor of Yizhou was killed. Only with the support of the native troops of the Zangke Commandery was the Han army able to pacify this rebellion.4

11In the reign of Heping of Emperor Cheng (28 BCE–25 BCE), a war broke out among the king of Yelang, the marquises of Gouting and Louwo. The Han court was reluctant to dispatch its army because of the long distance. Instead, an official was sent as a peacemaker, but he was rudely ignored by the king of Yelang, who even shot at the wooden sculpture of the Han official. Then the second Han envoy deceived the king of Yelang and beheaded him.5 The shocked marquis of Gouting and the marquis of Louwo offered millet, buffalo, and sheep as signs of their submission. While the Han state intervened to stop the intratribal war, it intensified the conflict between the local and central. The father-in-law and the son of the king of Yelang revolted and eventually led to Han's involvement by force.6

12The next wave of native uprisings took place during the Wang Mang period. Wang, a utopianist who usurped the Western Han in 9 CE and attempted to resume the ancient Zhou system, carried out radical policies both in internal and frontier China, which made worse the already unstable Southwestern Barbarian area. In 9 CE, the Gouting people rebelled because the king of Gouting was degraded to the rank of marquis. This event ushered an unprecedented chaos in the Southwest. Many indigenous peoples in Yizhou, Yuexi, and Zangke stood up against Wang Mang's Xin Dynasty. In 16 CE, Wang Mang dispatched a large army to suppress the Southwestern Barbarians. It is recorded that 60 to 70 percent of the soldiers died of disease and famine, however, and the situation was not improved at all. The Ba and Shu areas were shaken. Another force of over 100,000 soldiers was mobilized, resulting in the death of several tens of thousands. This military campaign lasted for six years, until the collapse of Wang Mang's regime.7

13In 42, local people rebelled and defeated the governor of Yizhou, hence driving him to Zhuti. In the following year Liu Shang, the Han general, led an army of over 13,000 men, including "barbarian" soldiers from Zhuti, and marched into Yizhou. In 45, Liu Shang destroyed local rebels and about 5,700 natives were captured.8 In 76, a dispute broke out between the king of Ailao and a Han official on assignment. The king of Ailao killed the Han delegate and attacked Xitang (Yunlong County, Yunnan) and Bonan (Yongping County, Yunnan). The governor of Yongchang was forced to retreat to Yeyu (Dali). The Eastern Han organized a mixed army, including "barbarian" soldiers from Yuexi, Yizhou, and Yongchang to handle the rebellion. In addition, Lucheng, a noble of the Kunming people, led his force to join the Han campaign. The king of Ailao was killed, and Lucheng was awarded the title "marquis of destroying enemy and attaching areas" (polubangyihou).9

14In 118, a large uprising took place in the three commanderies of Shu, Yizhou, and Yongchang. The rebels, numbering over 100,000, killed Han officials, plundered cities and towns, and robbed common people. "Bodies were piled and no people alive within a thousand li" (haigu weiji, qianli wuren).10 Shocked, Yang Song, the Han general, did not dare to advance. Many native chieftains betrayed the rebellion that was eventually suppressed by the beheading of over 30,000 natives.11 Once more, the strategy of yiyizhiyi (using barbarians against barbarians) was put into effect.

15The massacre stabilized the Southwest for only fifty years. In 176, the indigenes in Yizhou revolted. The governor of Yizhou was captured and the whole of Yizhou became under the control of the rebels. Some officials suggested that the Eastern Han give up Yizhou, since the Han court was unable to suppress the rebels. It was only with the help of the Bandun "barbarians" (bandunman) from the Ba that the Han court managed to put down this event. This military conflict destroyed the local economy and population, and the indigenous elites had no resources with which to turn against the Han court, while the new Han official launched a relaxation policy.12

16The collapse of the Eastern Han Dynasty in the early third century left the Southwestern frontier areas autonomous. The three-century Han intervention, however, had tremendously changed the track of the Southwestern Barbarians. The frequent turmoil outlined above suggests that the Han presence disrupted native power structures and provided new sources for struggles of all kinds.

17The above military conflicts can be classified into two types, which, directly or not, were caused by the Han presence. The first type was the indigenous rebellion for independence and the second was the intratribal clashes in native societies, whereby the indigenes attacked each other and attempted to make use of the Han power for their own sake. Some conflicts were directly caused by the strategy of yiyizhiyi of the Han court. This strategy increased local tensions and hatred, as one ethnic group or regime was picked up and favored over another by the Han court. The group or regime became relatively powerful and influential. As such they sometimes attempted to conquer neighboring peoples. Consequently, a war was unavoidable in many cases.

18Economic factors were an important reason for native rebellions. The indigenous peoples generally were allowed to pay light taxes, and were sometimes exempted, as shown by the case of Ailao people under the administration of Zheng Chun. Zheng was appointed governor of Yongchang, and he negotiated the taxation with the Ailao elites. The regular annual taxes consisted of two pieces of cloth and one hu of salt from each rich and powerful family.13 Throughout Zheng's term there, the Ailao did not make any trouble.14 While taxation was somewhat light, the insatiable material demands of local Han officials in many cases ignited local revolts. For example, the 118 rebellion was sparked by exploitation, and when it was finally suppressed, the Han court punished ninety local officials who had agitated the indigenes. In addition, heavy levies or labors constituted another economic burden for the indigenes. Compared with taxation, the indigenes were usually subject to heavier labor corvée than Han Chinese.15 As soon as Emperor Wu subjugated the Yelang and Qiongzuo peoples, he ordered Tang Meng and Sima Xiangru to build the Southern Barbarian Road (nanyidao) and the Western Barbarian Road (xiyidao). The two projects called for the conscription of the indigenes and gave rise to their rebellion.

19Many other obligations, such as military obligation, served as causes for local revolts. In 112 BCE, when the Han conscripted the Yelang to join the campaign against the Southern Yue Kingdom, the king of Qielan killed the Han envoys and the governor of Jianwei commandery. He was afraid of the long distance of the Southern Yue campaign and the potential attack by neighboring regimes, had he joined the Han forces. Nonetheless, rebellions were only one of the many products of Chinese intervention. The emergence of yishuai and daxing was another cultural and biological product of interactions from the mid-Eastern Han Dynasty onward. Local chieftains (yishuai) and large clans (daxing) began to dominate Yunnan in the coming centuries and they managed to resist many Chinese military campaigns and maintained their autonomy.

20The words yishuai and daxing did not appear in early Chinese records such as Shi Ji and Han Shu. Much evidence exists to conclude that yishuai and daxing were biocultural hybrids of the indigenes and Han Chinese.16 First, many of the daxing were descendants of Han immigrants, such as the Yong, the Lü, the Huo, and the Meng families, all of whose family names were Chinese. Their ancestors were usually powerful officials assigned or forced to move to the Southwest. Second, yishuai usually referred to indigenous chieftains who were deeply influenced by Chinese culture. Usually they were chosen by the Chinese state to rule their people and territories, frequently with imperial titles, ranks, and posts. A typical example is the Cuan family, which dominated eastern Yunnan for several centuries.17 In Cuan Longyan's tomb tablet, the Cuans traced their ancestors to a famous official in modern Shanxi, although it is highly likely that the Cuans were native.18 The tablet statement seems to reveal how the native elites were influenced by Chinese culture due to extensive connections and how they made use of Chinese institutions to legalize and enhance their power.19 As true rulers of the Southwest or Nanzhong at the time, yishuai and daxing indeed illustrated precisely the double-sidedness of the process of interaction. Yishuai symbolized the process of sinicization of the natives, while daxing exemplified the indigenization of Chinese immigrants.20

21While it is true that political and military interactions altered local societies dramatically and quickly, these changes were preceded by commercial interactions that were transformative. The diverse indigenous economies included agriculture, husbandry, fishing, and hunting. The indigenes were organized into relatively small ethnic groups or regimes; hence, to a large degree, they were interdependent, but material exchange was crucial to their survival and development. Furthermore, Chinese records showed that merchants of the Ba and Shu made a fortune through trade with the Southwestern Barbarians, which in turn implies that the Southwestern Barbarians also benefited from the trade. Finally, tributes from native elites and the byproducts of wars turned out to be another kind of material exchange. Imperial sources often recorded that livestock counted by the thousands were taken by the Chinese army or paid by native peoples. In a word, trade among China proper and the Southwest characterized the indigenous peoples and their societies, with a slow but far-reaching influence over a long period. After conquest of the region, China did not hesitate to exploit Yunnan materially, as revealed by the Shu-Han regime.

Enriching the Shu-Han: Myth and History

22During the early third century, the Shu-Han regime in Sichuan claimed itself a true heir of the Eastern Han, and attempted to unify China. Zhuge Liang, the prime minister of the Shu, decided that the national strategy should be to ally with the Sun-Wu regime in Jiangnan and to launch northern expeditions to conquer the Cao-Wei in North China. However, yishuai and daxing in Nanzhong, Shu-Han's rear base, rebelled, which forced Zhuge to turn his attention to pacifying the rear base before any northern expedition could be launched.

23Zhuge is well known in the Chinese mass culture, largely because of the popular novel written in the Ming period, Romance of the Three Kingdoms. He has been regarded as the symbol of wisdom (zhisheng). A popular myth recounts the Nanzhong campaign in 225 as a major event that demonstrated Zhuge's wisdom. It claims that he pacified the Nanzhong people by the use of peaceful and kind treatment with a strategy to conquer their hearts before their bodies. It is believed that he left native chieftains to deal with local affairs and that local peoples did not revolt until Zhuge's death. Many ethnic stories and folk tales repeat that he introduced the advanced production technologies to help native peoples. However, a closer scrutiny of the Nanzhong campaign suggests that the above is more a Chinese invention than history.

24Unlike the common image of a peaceful conquest, there were bloody fights after Zhuge conquered Nanzhong—mainly by force. Second, Zhuge either appointed local elites as governors or moved them to the central government, but most of them were daxing, namely, descendants of Han Chinese, and some of them had been loyal to the Shu-Han regime during the rebellion period. By contrast, no yishuai was appointed to be local governor. Third, Zhuge restructured administrative units in Nanzhong. He expanded the four commanderies of Nanzhong into seven commanderies and set up several new counties. He installed the Laixiang military governor to oversee the seven commanderies (Yuexi, Yunnan, Zhuti, Xinggu, Zangke, Jianning, and Yongchang). By using the strategy of "dividing to rule," Zhuge managed to reduce the influence of local elites and thereby enhanced central authority. Furthermore, some indigenous groups were moved to Chengdu and recruited into state troops. Finally, local resources were exploited by the Shu-Han regime. Unlike in previous times, taxation by the Shu-Han was a regular occurrence and not just symbolic. Nanzhong provided the Shu-Han regime with laborers/soldiers, water buffalo, warhorses, gold, silver, leather, salt, cloth, iron, lacquer, and so on.21 In fact, abundant Nanzhong resources were material foundations for Shu-Han's frequent northern campaigns. Extraction of local resources placed heavy burdens on Nanzhong, which partially explained why the indigenous peoples still rebelled a few times after Zhuge's Nanzhong campaign.22

25The legend of Zhuge Liang represents the history of Yunnan as written by the Chinese. It advances the idea that the Chinese brought culture to the "barbarians" who purportedly behaved like children but whose nature could be transformed. Such a myth is widely accepted even by diverse non-Han ethnic groups in Yunnan. What is often missing in the story is that the Chinese conquered the indigenous groups by force and extracted local resources. Zhuge's strategy to use Nanzhong strikingly illustrates how he aptly continued and developed the Chinese idea of jimi. Indeed, many of Zhuge's ruling strategies, such as restructuring administrative units, dividing to rule, moving local elites to central posts, utilizing native forces, and exploiting local resources, could be found in future Chinese state polices.

Nanzhao and Dali: Out of China

26The Three Kingdom period initiated a chaotic time in China until the Sui Dynasty (581–618). During a period of over three hundred years, while Chinese states still established administrative units in Ningzhou (Yunnan), native chieftains ruled their own regimes. Indeed, many Chinese governors of Ningzhou were unable to enter their offices.23

27Immediately after its unification of China, the Sui began to handle Yunnan. South Ningzhou Commandery was set up around 590, with Wei Chong as governor general. Cuan Zhen, the head of the Cuan family, paid a visit to Wei Chong, expressing his loyalty to the Chinese state.24 In addition, West Ningzhou was established in Xichang, in coordination with South Ningzhou. Furthermore, three jimi commanderies (Gongzhou, Xiezhou, and Kunzhou) were established in northeast Yunnan and the Dian Lake area. Local chieftains were assigned to be prefects (cishi). However, the imprudent behavior of the Sui soldiers and officials provided a good excuse for the Cuan's resistance to the central power that in turn provoked the Sui Dynasty's two expeditions.25

28The short-lived Sui was unable to back up local officials to execute real administration. In one episode in his life, Liang Pi, the prefect of West Ningzhou, received a lot of gold from local chieftains who fought each other and tried to buy Liang. Liang returned the gold and asked them why they killed each other for an item that could neither feed nor warm people. These chieftains were very touched and changed their behavior slightly.26 The case of Liang is very revealing. First, it showed that the Sui authority had no power to stop local warfare. Second, local chieftains realized that nominal submission to the Sui was good for their own sake and that bribing local officials served as a crucial way to win central support. Such an intentional movement of local chieftains demonstrates how these indigenous peoples consciously made full use of the Chinese power that was a doubled-edged sword. For the short term, some local chieftains, elites, or ethnic groups benefited and received a lot of special treatment from the Chinese state. In the long run, Chinese rule provided a good opportunity for the state to intervene in local affairs. While local chieftains made use of the Chinese, the Chinese created and enlarged a chasm that pitted "barbarians" against "barbarians." Finally, Liang's success in reducing the warfare was achieved only by means of his Confucian ethics. The moral model on the one hand reveals the inability of Chinese authority; on the other hand, however, it probably illustrates that the local elites understood and even partially accepted Confucian ethics.

29The autonomy in Yunnan lasted until the arrival of the Tang forces. Many jimi commanderies, departments, and counties were established, with native chieftains as the heads. Almost at the same time, the Tubo Empire in Tibet quickly expanded its influence in northwest Yunnan. In fact, both the Tang and Tubo tried to win the support of local peoples. Interestingly enough, many local chieftains submitted to both the Tang and Tubo, hence taking advantage of the two powerful rivals. Nanzhao certainly was the most successful one. Based on the Dali plain, Nanzhao gained the support of the Tang and Tubo, and for the first time unified most of Yunnan. From then on, Nanzhao, Tang China, and Tubo engaged in a drama of international politics, which was discussed in chapter three.

30Nanzhao survived for over three centuries. Among the thirteen kings of Nanzhao, ten accepted imperial titles from Tang China, such as prefect (cishi), prince of Taideng (taidengjunwang), king/prince of Yunnan (yunnanwang), and king/prince of the Southern Kingdom (nanzhaowang). Political alliance certainly facilitated cultural exchange. For example, the Nanzhao court sent the royal youth to study in a school in Chengdu that was designed exclusively for them, which facilitated the spread of Confucianism. Some kings of Nanzhao and high-ranking officials were able to write fine Tang poems.27 Exchanging poems seemed to be a cultural activity at Nanzhao royal banquets or gatherings, just as it had in Chinese courts since the Han Dynasty. More important, the Nanzhao court copied the six-department system developed in Tang China.28

31Other foreign cultures came into Nanzhao as well. Tantric Buddhism, mixed with local shamanism, became the state religion of Nanzhao, and was surely spread from the south;29 besides, one Nanzhao king claimed himself the title piaoxin, or king of Pyu, which revealed the intimate association with the south. Nanzhao Yeshi records that the Kunlun Kingdom in mainland Southeast Asia presented beauties to the Nanzhao court.30 In addition, the Tibetan factor cannot be underestimated. All these facts serve to illustrate that Nanzhao aptly incorporated neighboring cultures and developed its own institutions. In so doing, Nanzhao not only achieved political independence but also cultural creativity. The conclusion by Chinese scholars that Nanzhao belonged to Tang China is not justified at all. It is at this point that Tang China's policy toward Nanzhao differed from the Han's, because Nanzhao was no longer a frontier but a rival of Tang China. Tang China's policies toward Nanzhao and Tubo treated them more as a foreign entity than a frontier. To some extent, south Sichuan, parts of Guizhou and Guangxi, and Annam were frontier areas for which Tang China, Nanzhao, and Tubo competed.

32Song China carried out an isolationist policy toward the Dali Kingdom in an attempt to reduce the interaction as much as possible. Although many tributary titles were assigned to the king of Dali, such as "king of eight states in Yunnan" (yunnan baguo junwang), "ruler of Dali Kingdom in Yunnan" (yunnan daliguozhu), "prince of loyalty and submission" (zhongshunwang), "governor of Yunnan" (yunnan jiedushi), "prince of Dali" (daliwang), and other honorary imperial titles, the context was indeed different, as the eastern Asian continent was a world of equal power during that period. The Nanzhao and Dali remained independent kingdoms until the arrival of the Mongols brought about drastic changes.

The Tripod Administrative System of the Mongols

Seizing Power

33The Mongols had a tremendous influence on Yunnan. It was during the Yuan Dynasty that Yunnan was annexed into China proper. And it was during the Yuan Dynasty that a tripod administrative system was created, which not only restructured local power structure but also shaped the Ming-Qing period.

34In order to reorganize the local power structure, the Mongols destroyed the Gao family in the Dali Kingdom. The Gao family had been the real rulers of Dali since the late eleventh century;31 and were the violent resistance force against the Mongols to avoid the loss of power.32 Second, unlike previous conquerors who had been unable to stay and control Yunnan for a long time, the Mongols succeeded in taking over the whole area of the Dali Kingdom and marched even further. One key measure was the utilization of the Duans. Perhaps because the Mongols destroyed the Gao family, which had bullied the Duan court during the late Dali Kingdom, the Duans greatly appreciated it and showed tremendous loyalty. As soon as the Mongols arrived, the Duans presented them with a map of Dali, a symbol of submission. The Mongols acknowledged the influence of Duan Zhi, former king of Dali, by conferring on him the title of mo-he-luo-zuo, or maharajah, literally "great king" in Sanskrit.33 Furthermore, the Duans were assigned to oversee the diverse groups within Yunnan, under the hereditary position of the governor general of Dali (dali zongguan).34 This post had been held by the eleven generations of the Duans until the end of the Mongol rule. Logically, Duan Zhi and his troops were utilized for the Mongols' further expansion into Southeast Asia and China. It is an amazing fact that when Uriyangkhadai, the Mongol general in charge of Yunnan, was ordered to siege Song China, his Mongol cavalry included only three thousand soldiers;35 local forces, led by the Duans, including the Cuan-Bo army (cuan-bo jun), or the Cuan-Bai army, consisted of the majority.36 The Cuan-Bo army took part in the bloody battles in the mid-Yangzi.37 In addition, the Duans also led its force to suppress a large rebellion in Mongolian Yunnan, since the Yuan Dynasty was unable to effectively impose authority over native groups at the beginning of its reign.38 This turmoil drove the Mongols to reorganize its governance hierarchy.

35Symbolizing Yunnan's significance to the Mongols' expansion, Khubilai Khan in 1267 dispatched his fifth son to be the prince of Yunnan (yunnanwang).39 The arrival of "Prince Yunnan," as he was called, however, further fueled the extant power conflicts. In 1271, two high-ranking Mongolian officials poisoned Prince Yunnan and rebelled.40 Although suppressed, the rebellion revealed the weakness of loose central control and the lack of capable and loyal local officials. That is why Sayyid'Ajall Shams Al-Din (Sai-Dian-Chi Shan-Si-Ding) was chosen by Khubilai Khan to take charge of Yunnan.41

36Sayyid'Ajall Shams Al-Din, a Muslim confidant of Khubilai Khan, was asked by the latter to adopt a policy of "cautiousness and steadiness" (jinhou) to stabilize Yunnan.42 In 1274, the Branch Secretariat Council of Yunnan (yunnan xingsheng) was established with Sayyid'Ajall Shams Al-Din as civilian governor (pingzhangzhengshi). Since then, Yunnan has been a major and primary constituent of Chinese administrative units, fostering the development of local identity, which will be discussed in chapter five.

37Sayyid'Ajall Shams Al-Din, as the first governor of Yunnan Province, proved to be the right man for the job. First, he erased the tension with Mongol imperial princes. Second, he centralized the governing power, placing civil and military departments under the provincial government. For example, he managed to contain the Mongolian generals, thus avoiding bloody slaughters in military expeditions.43 Third, he changed current wartime administrative system into a regular one. The "ten-thousand household," "one thousand household," and "one hundred household" administrative entities were replaced with lu (circuit), fu (prefect), zhou (department), and xian (county). Finally, another crucial point he handled adeptly was making use of native chieftains, which some scholars believe inaugurated the famous native chieftain system.44

38Before I proceed to explore the native chieftain system, it should be emphasized that the terms "native," "indigenous," and "local" must be treated with caution. As we know, diverse interactions played a critical role in creating these powerful ethnic or pro-ethnic entities. Their emergence in Yunnan relied largely on biocultural, economic, and political exchanges with neighbors. In this sense, they were no longer native or indigenous. Likewise, the word "local" does not necessarily oppose "central" or "center." Although local kingdoms in Yunnan were somewhat within the Chinese tributary system, they, especially those in southern and southwestern Yunnan, sometimes paid tributes to Burma at the same time. And after all, paying tributes did not signify that they were inferior. By contrast, it reveals how aptly they utilized the outsiders to legitimize and strengthen their rule and reproduce their society. Hence, the dynamics of their societies and institutions should not be underestimated by their eventual submission to the Chinese.

Recruiting Local Help: The Native Chieftain System and Other Mongol Legacies

39While Sayyid'Ajall Shams Al-Din attempted to introduce a civilian system, he could not afford to ignore the existence and influence of native chieftains. Powerful as the Mongols were, they could not govern directly and had to rely on local elites and adopt local institutions to rule. The Duans were still employed to oversee other chieftains and their groups. The good treatment of the Duans established a model that comforted other chieftains who were appointed leaders of commanderies, departments, and counties. Indeed, many of them were given honorary titles and posts as soon as they surrendered themselves to Khubilai Khan in 1253 and 1254. This system of governance was later referred to as the native chieftain system.

40There were three general features to this system. First, it was launched under the provincial administration, so that no native chieftain post was higher than this level. Such a design is crucial in understanding the nationwide administrative hierarchies from the Yuan period onward, as provincial units constituted the fundamental structure of late Chinese empires. In addition, the native chieftain system was mainly launched in the Southwest.45 The reason is obviously because so many diverse ethnic groups lived in these mountainous areas.

41Second, the Yuan court assigned native chieftains with a number of ranks and posts. Many titles were created to display the activities of the native chieftain system, such as xuanweishi, xuanfushi, anfushi, zhaotaoshi (all literally meaning "pacification commissioner"), zhangguansi (regional commander), and so on. These posts were from rank 2 and downward.46

42 alt textMap 4.1 Yunnan in the Yuan Empire Third, as native chieftains were incorporated into the imperial bureaucratic system, their obligations and responsibilities were elucidated to them. Three kinds of obligations were required, with the military obligation being the primary requirement: The use of surrendered local troops made it possible for the Mongols to conquer and take root in tropical and subtropical climates that differed so much from their own. Second, local chieftains were deployed to put down local rebellions. Furthermore, as a symbol of their subjugation, local chieftains paid tributes (gong) to the court periodically. Tributes were required annually, biannually, or triannually, as determined by the distance to Beijing. The size of the mission was fixed, from several people to several dozens. Tributary items were usually local products, such as gold, silver, horses, elephants, tigers, knives, and other exotic goods. Tributary missions were rewarded with many gifts. Regular levies were demanded as well, including goods similar to the tributes.

43Finally, the Yuan court detailed rules concerning the appointment, succession, promotion, degradation, reward, and punishment of native chieftains. Whenever a chieftain was assigned, an item authenticating his rank (xinwu) was given to him by the court, symbolizing the establishment of the relationship. This item was the official acknowledgement that native chieftains were required to present when interacting with local or central government. The post of tusi was hereditary but the Yuan classified the order of succession. Usually the list of succession followed the line of son-nephew–brother-wife, but the successors had to be native (turen). Indeed, such an ordered procedure followed local practice, as the Yuan court stated that "[it] suits to follow original custom" (yi cong bensu).47

44The Yuan court also issued the promotions and punishment of, and rewards to, native chieftains.48 It stated that evaluations would be made every three years. Nevertheless, even when native chieftains committed crimes, the office would not be abolished, revealing that the Yuan court reigned lightly on native chieftains; either the Yuan feared destroying well-working local power structures and infrastructures or it lacked the resources and ability to rule with a heavier hand. In fact, the native chieftain system was a kind of compromise between the central authority and local regimes. As an acknowledgement of local power, it was the institutionalization of "rule based on native customs." Local power contributed as much as the central state to the formation and development of this system. To a great extent, local chieftains enjoyed their autonomy. The Duan family was one such case.49 Due to its historical importance and influence over the whole of Yunnan, the Duans were awarded prestigious honors and titles by the Mongols. Indeed, the Duans managed to preside over the Dali region into the late Yuan Dynasty, remaining autonomous until the arrival of the Ming army. Likewise, other native chieftains retained their power in many respects. Since they often took the post of prefectures and counties, the Mongol power seldom reached the level of county and below. To sum up, the native chieftain system in the Yuan Dynasty was in an early stage of its development, and regulations were made case by case, and were far from being comprehensive and systematic.

45The Yuan administrative hierarchy in Yunnan established a tripod pattern for the Ming and Qing. Theoretically, the imperial household was represented by Prince Yunnan and Prince Liang,50 both of whom administered power and enjoyed many privileges. For example, when Sayyid'Ajall Shams Al-Din arrived, the first thing he did was assuage Prince Yunnan's fear that his power would be reduced. The second leg of the tripod system was provincial administration of Yunnan, but the presence of princes took much power from the province and often challenged local officials.51 Finally, the existence of native chieftains made impossible the expansion of authority in Yunnan Province beyond urban cities. Beyond Kunming were numerous local chieftains who were the real rulers over the indigenes across mountains, forests, and rivers.

46The hierarchy of imperial prince, provincial authority, and native chieftains lasted till the Qing period and manifested the weakness of the central authority, a compromise the central state had to make on a frontier and ethnic area. However, when the centralization of power gradually took place, this tripod power structure was transformed into a single system. Gaitu guiliu—the reformation of native chieftains into part of the imperial administration—in the Ming-Qing period was a major and crucial step the central state had taken to centralize power and penetrate into local society.

47Another major repercussion of the Mongol rule left in Yunnan was the arrival of Central Asians. They arrived in Yunnan as a result of the Mongol campaign and administration. Muslims were obviously the largest group among them, though contemporary sources did not present an accurate figure. Represented by Sayyid'Ajall Shams Al-Din, many Muslims occupied administrative posts, fostering the Mongol control in Yunnan.52 Gradually they built their community and penetrated into local societies. The most astonishing fact was that they participated actively in the existing international trade network, and began to dominate the famous caravan trade to mainland Southeast Asia.53 Outsiders and immigrants generally tended to be more successful in trade.

48It is intriguing and somewhat perplexing that Muslims later were outspoken players in Yunnan when the Mongols were not. When Chinese immigrants flowed into Yunnan during the Ming-Qing period, tensions and conflicts grew between the two groups of latecomers, even when their economy and life were interdependent. Small quarrels eventually grew into a big clash in the mid-nineteenth century, which was the famous Muslim Rebellion (1856–1874) that hastened the Qing's collapse.54

Removing the Duans and Installing the Mus

49Gaitu guiliu by the Qing state had won a lot of attention. Nonetheless, similar efforts made by the Ming state preceded the Qing reform and had severely restructured and decreased local power and autonomy. The major progress made by the Ming was the establishment of a relatively unitary administrative hierarchy in Yunnan. Several crucial steps were taken, including the removal of the Duan regime, the installation of the Mu family, and the further institutionalization of the native chieftain system.

50Unlike the Mongols, who acknowledged the power of the Duan regime, the Ming court was determined to get rid of this influential ethnic polity, which had the potential to get in the way of central control. The eradication of the Duans symbolized the significant victory of central authority over local chieftains. The Duans had imposed its control over all or part of Yunnan for more than four centuries since the mid-tenth century, and its power was deeply associated with the local infrastructure. The practical and symbolic significance of the Duan regime in native societies was enormous. Successful as the Mongols were in establishing central administration over Yunnan, they were unable to evict the Duans. Instead they relied on the Duans to contain and comfort small native chieftains, and, if necessary, to suppress local rebellions. During the last several decades of the Yuan rule, the Duans managed to resume its autonomy in the Dali region. That is why the Duans thought that they had plenty of power and resources to negotiate with the Ming Dynasty, the newcomer in the southwest.

51However, the Duans' request for autonomy was turned down by the Ming generals. Duan Ming, the head of the Duan family, and his two sons were captured and taken to Nanjing. One son was assigned to Yanmen Passé (yanmenguan), a frontier station along the Great Wall; the other was assigned to Wuchang, a port city in mid-Yangzi. Certainly, their posts were honorary, an example of the courtesy the Ming showed to local chieftains. Other elites of the Duan clan were not as lucky. Some of them were killed in battles, some changed their family name and escaped to other provinces, and some were forced to move to provinces as far as Shandong in north China.55 Many Duan chiefs were taken to Nanjing and executed.56 In short, the Duan regime was wholly eradicated, which was reflected in the list of native chieftains during the Ming period: Within the Dali region there were twenty-nine native chieftain posts, and among the twenty-nine offices, only one bore the surname Duan.57 Whether or not this particular Duan belonged to the discussed Duan clan remained unclear. As a result, the Duans, a symbol of native peoples and power, were removed from Yunnan forever.

52It sheds some light on the process of Chinese penetration into Yunnan to compare the Cuans in the Sui-Tang period with the Duans in the early Ming. The Sui conquered and took the Cuans to its capital. The Cuan chieftain was beheaded, but his son was later sent back by the Tang to take charge of local affairs. Such a case revealed the compromise between the central court and local regime. The court was able to conquer but could not rule effectively without the help of local chieftains. The case of the Ming was different. The Ming state conquered, and the Duans were captured in order to display the power of the court. More important, the Duans were either moved or killed, the roots of their power effectively destroyed. The removal of the Duan regime yielded much significance for the Ming and for the natives. It showed that the Ming had the ability, or at least the confidence, to rule directly. The central state no longer tolerated a regional challenger and potential threat even when the potential challenger could be of help. As for the natives, they lost a symbol, a representation of their history, tradition, culture, and power. No matter how powerful other native chieftains could be, there were no replacements for the Duans, who had enjoyed a great deal of prestige among native societies. Only the Duans could establish an alliance among such diverse ethnic groups and impose restraint on them, either for its own sake or for the central authority. With the removal of the Duan regime, the Ming Dynasty no longer faced a rival influencing the whole region of Yunnan, though it did meet with many problems regarding other local chieftains, especially in the southern borderlands. After all, the lack of leadership among native chieftains assisted the Ming control, since this situation accorded with Chinese wisdom, "dividing to rule."

53Although the Ming removed the Duan regime, to some extent, it continued the Mongol's tripod system, namely, imperial prince, provincial administration, and native chieftains. Instead of making use of the Duans, the Ming installed an outsider to oversee the new frontier. Mu Ying, one of the three generals who led the Ming campaign, was stationed in Yunnan, and his family remained in power until the last day of the Ming Dynasty.

54When Zhu Yuanzhang decided to conquer Yunnan, he might not have expected the hardships he would encounter. With over 300,000 soldiers, the Ming troops quite easily destroyed the Mongols and the Duans. Nevertheless, diverse local chieftains posed a continuous and exhausting problem. Native chieftains submitted themselves when the Ming army arrived but rebelled as soon as it left, keeping the Ming soldiers quite busy. Zhu Yuanzhang commented that the Ming army "is like the wind, the grasses immediately stand up when the wind leaves."58 In fact, within half a year the Ming troops took over Kunming and Dali, but it took a couple of years to pacify native chieftains. Hence Zhu Yuanzhang decided to enhance the military presence in Yunnan. About 90,000 soldiers were stationed in the nine wei (guard posts) in Yunnan.59 Each wei consisted of nearly 10,000 soldiers, which outnumbered by far the figure of 5,600 in a standard wei in China proper.60

55The large number of troops certainly required an experienced general who could be trusted by the extremely suspicious emperor.61 The arrangement of Mu Ying as one of the three campaign leaders seemed to be a deliberate decision by Zhu Yuanzhang. Among the three generals, both Fu Youde and Lan Yu outranked Mu Ying, but Mu Ying had one incomparable advantage over the others: He was an adopted son of Zhu Yuanzhang.62 In the time before Zhu Yuanzhang's rise, Mu and Zhu shared food, clothes, and a bed.63 Therefore, the intimacy between Mu and the founding emperor made Mu the best person to be Zhu's agent in the remote and recently conquered frontier province.

56Fu Youde and Lan Yu led the main Ming army back to Nanjing after the major campaign, while Mu Ying was ordered to curb local rebellions and succeeded in completing his mission. Mu not only put down local rebellions but also contained native chieftains in large southern frontier areas, and resumed social order, thus moving Yunnan out of wartime. Zhu Yuanzhang was so delighted with Mu's achievements that he praised Mu thus: "Since you are in charge, I do not need to worry about the Southwest."64 In 1384, Zhu decided to station Mu permanently in Yunnan. Since then, the Mu family was in Yunnan for over 270 years, until the end of the Ming, producing two princes (wang), one marquis (hou), one earl (bo), nine imperial dukes (guogong), and four military governors (dudu).65 These titles reveal tremendous imperial honors and trust, but the key position Mu held was the hereditary military position of zongbing, or commander in chief, the highest military post in Yunnan, that was granted to take charge of regional military commission (duzhihuishisi). Indeed, the Mu family served as the major leader of all military campaigns in Yunnan throughout the Ming Dynasty.

57The special role of the Mus established and fed influence and power in Yunnan. First, the Mus were awarded a large swath of land that came with laborers.66 All of their properties enjoyed tax exemption. Within half a century, the Mu family owned 360 farm units (qu).67 During the Wanli reign (1573–1619), sixteen of the twenty-five administrative units (commanderies) under the Provincial Administration Commission (chengxuanbuzhengshisi) were found to have Mu farms.68 And within those commanderies the Mu farms were found in each department (zhou) and county (xian).69 At the end of the sixteenth century, the Mu farms possessed over 10,000 qing of fertile lands, almost equal to that either under the Provincial Administration Commission or the Regional Military Commission.70

58The huge expanse of the Mu farms brought about many court impeachments. Angry officials argued for an inspection of the Mu farms. Imperial edicts were issued, but were never strictly executed. Indeed, the Mus appealed to the emperor to stop the inspection and their request was granted.71 In 1610, under the pressure of ministries, a survey was finally launched. It showed that among 8,842 qing of the Mu farms only 1,358 were granted by the imperial state, the rest having been either purchased (perhaps coercively) or expropriated.72 The survey was fruitless since no further action was taken. During the Tianqi reign (1621–1627), officials had asked to levy Mu farms for local relief on several occasions but these requests were put aside by the emperor.73

59The power of the Mus extended beyond the economy and into politics. In 1396, Prince Min, the eighteenth son of Zhu Yuanzhang, was stationed in Yunnan. Mu Cheng, the son of Mu Ying, reported the misconduct of Prince Min. As a result, Prince Min was deposed by his nephew, Emperor Jianwen, and moved to Zhangzhou (Fujian). A couple of years later, Emperor Yongle moved Prince Min back to Yunnan, which unavoidably led to tensions between the prince and Mu Cheng. To make peace between them, the emperor wrote to both of them. Prince Min was asked to keep in mind Mu Ying's contribution, and Mu Cheng was instructed by the emperor as well.74 The case clearly shows that the imperial prince did not win this conflict.

60The report of the misconduct of Prince Min by Mu Cheng implies that the Mus were probably assigned to oversee imperial family members and local officials, and this status often led the Mus to show disregard for local officials. Mu Kun, for instance, had three commissioners of Yunnan Province enter his house through the corner gate (jiaomen), which was a serious act of disrespect.75 Although this event might have been isolated, Mu's general attitude toward other Yunnan officials can be assumed.

61In fact, the arrogance and misconducts of the Mus were tolerated by emperors although the latter knew what the Mus did in great detail from testimony either by local or central officials. To appease the popular dissatisfaction and complaints, the emperors issued many edicts to warn the Mus not to abuse their power and privilege. While the tone of the edicts was harsh, little action was actually taken.76 Emperors treated the Mus with such generosity because they needed to ensure a secure frontier area. The Mus were heavily relied upon to maintain imperial control of Yunnan. According to the biography of Mu Ying in Ming Shi, most of the Mu bearers of imperial titles were dispatched to battlefields, their efforts having certainly accumulated more than enough capital for the Mus to ignore local officials and imperials laws.77 As long as the Mus fought in the battlefields, their misconduct, or abuse of power could be tolerated or forgiven. In short, personal loyalty to the imperial house overshadowed personal evils.

62Without any doubt, the wealth of the Mus exceeded that of Prince Yunnan and Prince Liang of the Yuan Dynasty. However, the Mus were not granted legal civil administrative power. As stated above, during the Yuan Dynasty, the central control was quite weak, but by the end of the Yuan Dynasty, Prince Liang dominated the provincial administration. Such a civil power, however, was not enjoyed by the Mus. Civil administration was left to the second pillar of the tripod system in Yunnan, that is, the provincial government.

63Interestingly, the provincial administration of the Ming Dynasty was a tripod system, too. It included Provincial Administration Commission (chengxuanbuzheng shisi), Regional Military Commission (duzhihuishisi), and Provincial Surveillance Commission (tixinganchanshisi). Each of these three commissions had two senior commissioners in each province, and these six commissioners formed a committee that shared responsibility for all provincial affairs. There were no provincial governors.78 The division of provincial power again revealed the decentralization of local officials while enhancing central control at the same time.

64An extraordinary feature of the provincial administrative system was the Wei-Suo system, which served military colonization (juntun), a critical step to stabilize the frontier province. Troops were stationed in big cities or crucial transportation towns and passes in Yunnan, so that native forces were contained and the provision was protected. Although figures and locations of the Wei and Suo in Yunnan varied temporally,79 a systematic hierarchy and infrastructure was established that served as the fundamental source of the central control and the reproduction of Chinese culture.

65Under the Wei-Suo system in Yunnan, 30 percent of the soldiers underwent drills while the rest participated in agricultural production, producing necessities for the whole guard. Fifty mu of land were assigned to each solider to guarantee enough supplies,80 but this quota could not be filled in Yunnan, due to the local topography. Soldiers and their families were classified as hereditary military households (junhu), although some soldiers managed to escape or were incorporated into the Mu farms.

66One distinguishing feature of Ming military colonization was that soldiers were encouraged to take their wives with them and bachelors were encouraged to get married before moving into new stations. Local governments were asked to escort soldiers' fiancées or wives to military garrisons. Sometimes bachelors in the garrison were even discharged.81 The state generally supplied military families with transport aids. These regulations were designed to encourage soldiers and their families to settle permanently in new places. As a result, the military presence not only managed to successfully annex Yunnan to China proper but also served as a civilian medium, reproducing Chinese agrarian culture. It was the Ming soldiers, their wives, and their offspring who dramatically transformed Yunnan, and started a new trajectory.

67Another major decision concerning the administrative hierarchy in the Southwest made by the Ming court was the establishment of Guizhou Province. The Guizhou area was part of Yunnan before 1413, when it was deemed an independent province.82 In 1413, the administrative pattern in the Southwest was created, and has continued until the present day.

68Although the Ming court had an unprecedented number of troops garrisoned in Yunnan, the Ming forces could only control major cities, passes, and market towns. Vast rural areas were left in the hands of native chieftains. Recognizing local power, the Ming state institutionalized and continued the native chieftain system. Indeed the Ming court took many steps to institutionalize this system, attempting to regulate and penetrate the native power structure.

The Institutionalization of the Native Chieftain System

The Yuan and the Ming: Strategies for Institutionalization

69The Yuan Dynasty began to construct the fledgling form of the native chieftain system. Regulations were issued to guide the activities of native chieftains, in terms of award, punishment, abolition, and succession. However, these regulations were seldom executed. By contrast, the Ming not only out-detailed the Yuan in terms of regulations but also was able to put them into effect.

70Scholars of China generally agree that the native chieftain system began with the Yuan Dynasty. A close reading of the Yuan imperial record, however, suggests that the native chieftain system in the Yuan was at a very early stage. First, while native chieftains were granted with titles and office, these titles were not different from those of assigned officials. Xuanweishi, xuanfushi, anfushi, and zhaotaoshi were given to native chieftains as well as to Yuan officials. Second, while the Yuan court issued some regulations to manage native chieftains, these rules were not systematic. More important, these regulations in essence were the official recognition of local customs. For example, in terms of inheritance, the Yuan generally accepted native practices.83 Finally, the most revealing fact was that the Yuan took a lenient attitude toward native chieftains who had violated imperial regulations or laws. During the Taiding reign (1324–1327) when the Huajiaoman (flower foot barbarian) rebelled, the emperor issued an edict to comfort them instead of launching a punitive expedition.84 In 1325, some native peoples in Dali and Weichu revolted, and the provincial administration asked the court to pacify them, but the emperor denied this request and dispatched his envoys to deal with the "barbarians."85 Since rebels could be exempt from punishment, small crimes or misconduct were always ignored. During the Yuan period native chieftains would be warned and even fined, but never removed (faerbufei).86 Obviously, such a tolerant policy resulted from a weak central control; it was not the case that the Yuan court did not intend to impose a tight reign; it was just unable to do so. Given this, John Herman is right to state that the native chieftain system was created in the early Ming.87

71After the Ming removed the Duan regime, all native chieftains in the Ming period were rendered "local," below the provincial level. Furthermore, the Ming state made serious efforts to institutionalize the native chieftain system. When the Ming launched the 1383 military expedition, many native chieftains submitted, and so were awarded the same titles that the Yuan had granted them.88 Native chieftains were classified as either civilian or military. Civilian native chieftains (tuguan) were supervised by the Ministry of Personnel while military native chieftains (tusi) were overseen by the Ministry of War.89 Military native chieftainship included xuanweisi, xuanfusi, anfusi, zhaotaosi, tu zhangguan (native commander), and native (tu) military titles of all kinds in the Wei-Suo system. Civilian chieftainship included native prefects, subprefects, magistrates, and so on. Usually civilian chieftains were located within inner areas while military chieftains lived in the frontier or remote areas where imperial authority rarely existed. As a result, Yunnan in the Ming period could be grouped into three areas. The area north of the Baoshan-Yuanjiang line was so-called inner land (neidi), where civilian native chieftains dominated areas, including native prefectures (tufu) of Chuxiong, Yaoan, Heqing, Xundian, Wuding, and Lijiang, and native departments (tuzhou) of Luoxiong, Zhaozhou, Lu'nan, Jianchuan, Mile, Shizong, Anning, Ami, Luliang, and Zhanyi, and native counties of Luoci, Yunnan and Yuanmou. South of the Baoshan-Yuanjiang was the "barbarian" area where military native chieftains were recognized, including Cheli, Babai, Luchuan, Nandian, Ganya, Longchuan, Lujiang, Gengma, Mengmao, Chanshan, Menglian, and so on. And the furthest south was the yuyi (containing the barbarians) area, where, although some commanders were stationed, the Ming hardly had any authority.90

72During the Ming period, there were 179 ranked military native chieftains and 255 ranked civilian native chieftains in Yunnan.91 Their ranks ranged from 3a all the way to 9b.92 The Ming court gave native chieftains imperial certificates, seals (silver or copper), hats, belts, and some credentials, including fu (tally) and pai (tablets). When the Ming governance took Yunnan from war to peace, it began to institutionalize the state-native chieftain relationship, and impose more regulations on native chieftains. Detailed decrees, codes, and rules were passed to regulate the inheritance, tributes, awards, promotion, punishment, and abolition of native chieftains.

73The key issue of regulating the state–native chieftain relationship was the inheritance process. The state stipulated the list of succession candidates, which to a large extent still followed the native custom. On the list were sons, brothers, wives, daughters, sons-in-law, and nephews. Nevertheless, the decrees concerning the procedure of inheritance showed how the state began to intervene and regulate the key tribal politics. In some cases, apparent heirs were required to present their genealogical charts to the concerned ministry, and local officials were also required to provide some kind of statement.93 To avoid confusion, chaos, and, possibly, bloody struggles, native chieftains in 1436 were ordered to provide genealogical charts that listed the names of sons and nephews so that the assigned heir could take over the office when the chieftain died.94 In 1441, native chieftains were requested to make four copies of the chart for the three provincial commissions and the Ministry of Personnel, and to update it every three years.95 In 1485, the Ming state proclaimed once more that native chieftains must update their genealogical charts every three years.96 In 1489, it was decided that an assigned official would help native chieftains to rule if the chieftain was younger than fifteen.97 In the early sixteenth century, the Ming state forged and distributed many credentials (xinfu) and gold seals (jinpai) to native chieftains.98 In 1530, more details concerning the chart were issued. Native chieftains were asked to list all their sons and grandsons, along with the names of their biological mothers, and their succession order; if the native chieftain did not have a son, he could list his brother, nephews, or daughter as his heir.99

74These decrees were mainly concerned with the succession procedure, namely, what native chieftains were allowed to take the chieftainship. Some other regulations detailed punishments, which included fines, demotion, removal, exile, execution, and abolition of inheritance. In 1535, it was stipulated that native chieftains would not be allowed to pass their office to their heirs if they fought one another, and/or persecuted soldiers or civilians.100 In 1555, it was stipulated that marriages of native chieftains were not allowed to cross provincial boundaries, and that native chieftains were not allowed to communicate with "outer barbarians" (buxu yu waiyi jiaojie wanglai). The harshest punishment produced by this type of violation was the abolition of chieftainship.101 Such regulations vividly revealed tensions and conflicts between ethnic connections and Chinese administrative units, as ethnic interactions had long preceded Chinese administrative units, and certainly crossed administrative boundaries. From then on, the Ming and, later, the Qing Dynasty attempted to shape and fit ethnic connections within administrative cages.

75The most significant sign of central control lay in the extent to which they were able to execute punishment. While the Yuan Dynasty generally forgave "guilty" native chieftains, which was merely a symbolic maneuver, the Ming actually executed its regulations efficiently. In 1395, Nong Langjin, the native subprefect (tongzhi, rank 5a) of Guangnan Commandery, was demoted to native tongpan (rank 6a) because his father had revolted.102 Before this degradation, his grandfather Nong Jijin had refused to pay a grain levy in 1385, and escaped, but in 1409 he confessed his crime, and was exiled to Liaodong (Northeast China).103 Although many forms of punishment were relatively light (sometimes in the form of fine or exemption), the death penalty was also implemented in the Ming period. In the reign of Zhengtong (1436–1449), Gao Lun, native prefect of Heqing, was executed.104 In 1528, Feng Chaowen and An Quan, the two native chieftains in Wuding Prefecture who rebelled, were killed and their families exiled.105

76The replacement of native chieftainships with direct governance by assigned officials (gaitu guiliu), another significant index of central control, was launched in the Ming Dynasty. The death penalty was just a measure to end personal life, but gaitu guiliu would end the whole native chieftain system in certain areas. Gao Lun, native prefect of Heqing, was probably the first case. After his execution, the native position was abolished in 1443, since all the possible candidates in his clan were thought not to be qualified.106 In 1478, a power struggle between the brothers broke out when An Cheng, native prefect of Xundian, died; thus the native prefect was abolished.107 In 1481, Ang Gui, native prefect of Guangxi Commandery, died, and the position was abolished.108 In the mid-1580s, the revolt of Zhe Jirong, native subprefect (zhizhou) of Luoxiong, was pacified, and Luoxiong was put under direct control.109 In 1607, after the chaos and rebellions of several decades, Wuding Prefecture was put under the direct governance.110 In 1621, the native department of Yunlong was abolished.111 Gong Yin counted that twenty-six native counties and above were abolished during the Ming period.112

77While many native chieftain offices ended, in a few cases native chieftainship was resumed to replace assigned officials. In 1433, local people in Nan'an Department appealed to establish native chieftainship. The Ministry of Personnel turned it down, but the emperor himself approved it.113 Ningzhou (department) was initially under direct administration, however, a native chieftain bribed Liu Jin, the influential eunuch, and a native subprefect was resumed that eventually would be abolished in 1522.114

78It is not a surprise that many crimes that would normally result in harsh consequences for Chinese officials were forgiven by the Ming state if those concerned were native chieftains. In 1384, the civilian commissioner of Yunnan submitted a report concerning legal codes for hereditary native chieftains. The decision made by the six ministries read that the supervisors were not allowed to question suspected native chieftains, that witnesses had to be questioned, and that the supervisor had to report his decision to the court.115 A 1430 report suggested that native chieftains be sentenced by the same legal codes as assigned officials (liuguan), but the emperor commented that the "barbarians" could not be ruled by the law applied to China proper (manyi bu keyi yi zhonguo zhizhi zhizhi), and therefore existing practices continued.116

79The institutionalization of the native chieftain system during the Ming period had an immense impact on the state-native chieftain relationship. The regulations concerning inheritance indeed initiated what would happen in the Qing reform. John Herman points out that the early Qing reform of the state-native chieftain relationship produced an unexpected consequence: inter- and intrachieftainship violence.117 Such violence, to be sure, was already quite frequent in the late Ming period. The replacement of native Wuding Prefecture was such a case.

The Struggle for Wuding

80Because in 1381 Shang Sheng, the female chieftain of Wuding, was the first native chieftain to surrender to the Ming court, her status was recognized accordingly.118 In 1528, Feng Chaowen, a native leader of Wuding, killed officials and rebelled. Feng Zhao, native prefect, and his mother, Madame Qu, managed to escape this slaughter. When the Ming troops defeated the rebels, they returned to Wuding City. In 1538, Madame Qu was approved to inherit the position of native prefect that was left behind by her deceased son. In 1563, Madame Qu recommended Suolin, her daughter-in-law (Feng Zhao's wife) for the inheritance of this position. After Suolin had taken over the position, Madame Qu felt angry that Suolin did not follow the propriety of the mother–daughter-in-law relationship, and adopted Jizu, a boy from a different family. With the support of An Guoheng (the husband of Qu's niece), a native leader (tushe) of Shuixi (Guizhou), and the Feng family (a native chieftain in Jianchang, Sichuan; a shared last name suggests a clan affiliation), Madame Qu attempted to depose Suolin but failed. Madame Qu then turned to the Ming government. She asked Jizu to submit her report to the state, in which she claimed to have been put in prison by Suolin. It is unclear whether Jizu arrived in Beijing, but it is highly likely that he did not, since he reportedly claimed that the state chose him to inherit, and led a force to surround Suolin for the official seal. Suolin ran to the provincial city with the seal. The commissioners solved this case by reconfirming Suolin's status, and she returned to Wuding.

81But the tension and conflict between Madame Qu and Suolin grew to such an extent that Suolin, under the suggestion of her assistant, Zheng Hong, planed to kill Jizu. The scheme was disclosed, and Jizu surrounded Suolin by force, and plundered the Hequ and Luquan counties. Suolin had no choice but escape to Kunming. This time the provincial authority did not stand with her. The official seal was taken away, and Zheng Hong was put into prison. Madame Qu was ordered to act as agent and Jizu was pardoned.

82In 1566, a new city of Wuding was completed. Zheng Hong was released and returned to Wuding. Jizu caught and killed him, and proceeded to attack the new city. Jizu succeeded in defeating local Ming troops, and killed and captured both military leaders and local leaders until he was forced to flee when he was surrounded by the Ming army. He fled to Dongchuan, planning to enter Sichuan because of his connection with the Fengs in Jianchang. The Fengs, however, betrayed Jizu when the overwhelming imperial force and native troops arrived. Finally, a native leader killed Jizu and presented his body to the Ming general. But the death of Jizu did not end the event.

83After this riot, the Ming state abolished the native commandery of Wuding and a prefect was assigned in its place, while the Feng regime was allowed to take charge of some unclassified position. Siyao, the son of Feng Li and related to Suolin, was chosen for the official position. However, Feng Li, who was dissatisfied with the abolition of the native prefect, allied with An Guoheng, the native chieftain in Guizhou, and other native chieftains in Sichuan, to attack the prefect city, thus proclaiming Siyao as prefect. Liu Zongyin, the assigned prefect, managed to defend the city, counterattack in the night, and destroy the rebels. Feng Li was eventually captured and executed.

84While Suolin and those loyal to her were silenced by imperial suppression, the other party of the struggle rose again. Ake, the nephew of Jizu, allied with other native chieftains, rebelled in 1607, took over and plundered Wuding, and captured Yuanmou and Luoci counties, thus demanding the seal of prefecture. The prefect happened to be away, so Ake could not obtain the seal; instead, Ake took the imperial hat, belt, and other credentials to legitimize his cause. The Ming's local military leader (zhenfu) was scared, as the Ming troops were far way, so envoys were dispatched to present the seal to Ake. Ake then retreated to Wuding City, and claimed himself as prefect. Within a short period, Ming officials mobilized troops and took over these counties and cities one by one. Ake was captured and presented to the imperial court in Beijing, where he was executed. With the removal of Jizu and his followers, the Ming abolished all the native chieftain positions in Wuding.119

85The case of Wuding shows how the institutionalization of the native chieftain system impacted the state-native chieftain relationship and the interchieftain power structure. First, the ascendance to the native chieftainship had to be approved by the imperial state, as Madame Qu reported to the state that Suolin would inherit her position; second, when a power struggle took place, both parties would seek support from the state. Madame Qu and later Suolin both turned to the state when they needed help in controlling their challenger; third, imperial symbols had been regarded as the source of legitimacy by the native chieftains. Ake first demanded the prefect seal. When he could not get it, he turned to other imperial credentials. Finally, the state took advantage of this power struggle and regarded it as a good opportunity to further central influence. The state did not simply support one party by smashing the other. In the first round, the state refused the claim by Madame Qu and continued Suolin's chieftainship, but it did not punish Jizu, who had lied to hold state support; in the second round, Madame Qu was supported, and Zheng Hong was put in jail. However, several years later Zheng was released and returned to Wuding to continue his service. Such a decision seemed rather imprudent, considering the previous tension between Zheng and Jizu. This story implies that the state played the parties against each other, a vivid enforcement of the strategy "to use barbarians against barbarians." Nevertheless, what happened later was totally unexpected. Interchieftaincy struggles developed into a revolt that involved even native chieftains from the other two provinces, Sichuan and Guizhou. After many exhausting battles, the Ming state finally pacified the revolt and brought the native chieftainship in Wuding to an end.

86Many events in late Ming Yunnan replicated the case of Wuding. Interactions between the state and native chieftains eventually characterized local politics and created a new society in late imperial Yunnan. Looking back at the frequent revolts of the Southeastern Barbarians during the Han period, we can see that the advent of the China "factor" dramatically changed local power structure and the ways struggles and fights played out.

The Native Chieftain System: A Civilizing Instrument

87While the Ming state made efforts to enhance its control over Yunnan, native chieftains aptly made use of the Ming's resources to enhance their own legitimacy, power, and influence. Since the Ming court regarded the native chieftain system as a form of control over the native elites and general populace, native chieftains accepted this bureaucratic imposition for their own interests. The Ming claimed imperial sovereignty over native chieftains and their indigenous peoples, while native chieftains took imperial posts for their political legitimacy over their subjects, thus containing neighboring regimes and sources of wealth. In short, the cooperation between the imperial state and the frontier chieftains was based on their misperceived cultural practices, which benefited both parties. Furthermore, under the native chieftain system, local connections by tribe, clan, or other associations still functioned, and were sometimes utilized for local resistance against the state. The case of Wuding is very revealing in terms of this issue as well.

88When Madam Qu came to regret her decision to have Suolin take over as native prefect, she did not immediately appeal to the Ming state for a change. Rather, viewing it as a family/clan/tribe issue, she invited the Ans in Guizhou and the Fengs in Sichuan, both her relatives, to intervene. Only after her attempt failed did she then turn to the imperial government. In the final stage, Ake, again, turned to the Guizhou and Sichuan connections for help. The Fengs in Sichuan, as the sources imply, had promised support, but even this was betrayed in the face of overwhelming imperial forces. Obviously, native chieftains in Yunnan remained in close connection with those in Guizhou and Sichuan. The cross-provincial boundary clan networks had indeed lasted for ages, and were not destroyed during the late sixteenth century, even though it was weakened. The Ming decree that forbade cross-provincial boundary marriage illustrated the role of such an ethnic/clan association. Cross-provincial tribal/clan/ethnic connections were still evident in the Qing period.

89Education served as another key measure for the Ming institutionalization of the native chieftain system. Initially, children of native chieftain families were selected to go to the imperial university (guozijian) in Beijing, as a special favor bestowed upon them.120 In 1395, Zhu Yuanzhang ordered that native youth were to be selected to receive Confucian education so that they would understand Chinese rituals.121 In 1481, it became a regulation to select native boys for state schools.122

90Sending native boys to Beijing killed two birds with one stone. These boys were sons, brothers, or close relatives of native chieftains who could serve as hostages whom the Ming could use to curb any potential betrayals of native chieftains. Furthermore, instilling in these youth the state ideology would transform these "barbarians" into civilized subjects, who in turn would civilize their "barbarian" subjects, since many of these young students came from elite families and many would succeed leaderships within the native chieftainship.

91Since the capacity of the imperial university in Beijing was limited, it was ordered that schools be established in native chieftain areas.123 In 1395, an imperial edict was issued to establish Confucian schools in native chieftain areas.124 Although we do not know to what degree this regulation was executed, it underscores the understanding that Ming rulers had that education was an effective way to civilize "barbarians." On the other hand, many native chieftains understood that the mastering of Chinese culture would bring about many benefits; therefore they actively responded to the state. The Mu family, which was native prefect of Lijiang, exemplifies this case. During the mid-Ming period, the Mus were well known for their performance of Confucian culture, and were even regarded as the most "civilized" among the native chieftains.125

92John Herman has detailed how the early Qing enhanced the regulation of the inheritance of native chieftainship either by introducing the Han patrilineal pattern or by incorporating education into the inheritance requirement.126 My close reading of the Ming sources has uncovered that these efforts were indeed initiated by the Ming state, though their degree of success needs further scrutiny. It is clear that the Ming institutionalized the native chieftain system, and that the Ming efforts were followed by the Qing state with more resources and power. In short, a decrease in local autonomy accompanied the increase of central penetration into the frontier area as the native chieftain system was established, institutionalized, regulated, and finally abolished, step by step, and case by case.

93The native chieftain system was a remarkable breakthrough not only in Yunnan history but also in Chinese frontier and ethnic history. For the first time, the central state managed to borrow the indigenous bureaucracy and implant and regulate it in Yunnan. Such an institution, as a result of the long-term development of Chinese frontier policy, manifests the maturity of Chinese state-building. As briefly discussed, the Chinese began to formulate their own bureaucratic system to handle frontier areas and ethnic peoples as early as their first empire. The idea of "rule based on native customs" was applied to create distinctive offices and infrastructure in frontier areas, for example, the frontier commandery system in the Han times, and the jimi system in the Tang period. The native chieftain system was essentially the systematic development of these previous efforts.

94Underlying such an institution were some unavoidable tensions. The native chieftain system was both strategic and tactical. It was a long-term strategy as long as frontier and ethnic minorities existed, and it was a short-term tactic as long as the imperial state thought it was time to reform frontier areas. It is the first and necessary step to develop a suitable way to establish and consolidate central control over non-Chinese frontier areas, but the eventual imperial aim was to civilize and transform ethnic peoples. Consequently, as long as conditions were thought ripe, the replacement of this frontier and ethnic institution with the direct administrative system launched in China proper was unavoidable. Tensions and contradictions between the long-term strategy and the short-term tactics accounted for the frequent establishment and abolition of native chieftains as soon as this system was formed. While the Ming Empire continued this trend, the largest and seemingly most "radical" move was taken during the Yongzheng reign (1722–1735), which ushered the decline of native chieftain system in late imperial China.

The Decline of the Native Chieftain System in the Qing Period

95The beginning of the Manchu rule in Yunnan adopted the tripod system. In 1661, Wu Sangui, the Manchu ruler of Yunnan, killed Emperor Yongli, thus ending the Southern Ming regime. Because of his achievement, the Manchus granted Wu with the exceptional honorific of "prince of pacification of the west" (pingxiwang). Indeed, Wu had, to a large degree, made Southwest China his regime. Despite the imperial titles, however, the undercurrent of suspicion and wariness grew with the increasing power of Wu Sangui. In 1673, Wu rebelled, together with the other two feudatories in Fujian, Guangdong, and Guangxi. Eight years later, the so-called Three Feudatories Rebellion was eventually put down by the young emperor Kangxi (1661–1722), and the Manchus put Yunnan under its direct governance.

96The Ming's provincial structure, developed in the early fifteenth century, was accepted by the Manchus, except that the Qing established the post of general governor of Yunnan-Guizhou. Such a position, while showing the close association between Yunnan and Guizhou, demonstrated another remarkable step of centralization, since the presence of a prince or hereditary family regime, such as the Mus in the Ming Dynasty, no longer existed. As a result, the tripod system of administrative hierarchy in the Yuan-Ming period was discarded by the Manchus. Logically, the Qing's goal of centralization aimed at what were below the provincial level, namely, the native chieftains. Throughout most of the Qing period, the power of the imperial state gradually increased while the space native chieftains controlled was correspondingly diminished. The major step that the Qing state had taken was the project of gaitu guiliu.127

97In its early period, the Qing state regulated the native chieftain system by narrowing the parameter of inheritance to the native chieftain office, annual meetings between chieftains and local officials, and Chinese elementary education as a compulsory requirement for the apparent native chieftain heir.128 These regulations eventually weakened the social connections between native chieftains and indigenous societies, as native chieftains began to attach themselves more closely to the imperial state to avoid the loss of title and power. The alienation of the indigenous society and the enhancement of the central presence took place at the same time.129 In so doing, the Qing furthered what the Ming initiated, that is, to legalize imperial interventions and influence the frontier society.

98The Qing Empire not only accepted the submissive native chieftains, it also acknowledged many small native chieftainships in areas where the Ming had barely accessed, which showed further central penetration.130 In total, there were about 273 native chieftains of all kinds in Yunnan during the Qing period,131 compared to 434 during the Ming.132

99Table 4.1 Ranks of Native Chieftains in Yunnan during the Ming and Qing Periods Table 4.1 reveals the increasing central power and the declining native chieftain system throughout the Ming-Qing period. In the first place, the number of officially acknowledged native chieftains in the Qing (273) was much smaller than that in the Ming (434). Second, the number of high-ranking native chieftains (rank 3 downward to rank 6) was greatly reduced over time. For example, in the Ming there were 15 native chieftains of rank 3, while only 1 existed in the Qing. This trend revealed that native prefects and those of higher rank were being replaced by assigned officials, a major effort to remove native chieftains from the provincial level to the prefectural level and downward. The third point in table 4.1 corresponds with this trend as well. The figure of native officials of rank 7 increased more than three-fold, from 28 in the Ming to 85 in the Qing. It demonstrated that native chieftainships in the Qing were largely assigned with the rank of county-level and below (zhixian, or magistrate of county, ranked 7a), since the number of native chieftains with rank 7 constituted more than one-third of all ranked native chieftains in the Qing Yunnan. When we include native chieftains of ranks 8 and 9, one third becomes one half. In other words, the Qing authority had begun to penetrate into counties below.

100However, how can the dramatic decrease of native chieftains lower than rank 9 from the Ming to the Qing (151:43) be explained? A close look at the source suggests that the statistics are misleading. Among the 151 native chieftains of the Ming, 95 were native xunjian, which was ranked 9b in the Qing. Moreover, there were another 25 native postal officials (yicheng), which occupied similar status to the xunjian. In addition, the source lists another 19 native chiefs (tushe) as Ming native chieftains. The figures of the Qing counterparts were far smaller (3 yicheng, and 6 tushe). Therefore, this comparison does not oppose the general trend that characterized the Ming-Qing transition.

101As I have shown, the tremendous decrease of native chieftains was not natural but state-designed. While the Qing in the beginning recognized the submissive native chieftains, those loyal to the Ming were smashed without hesitation. For example, the native prefecture of Yuanjiang was abolished in 1659 because Na Song, the native prefect, resisted the Qing arrival. Many rebel native chieftains in Ningzhou (Huaning), Zhee (Eshan), Mengzi, Wangnongshan, Wenshan, and Shiping were removed as well.133 The 1720s witnessed gaitu guiliu under the aggressive policy issued by Emperor Yongzheng and his trusted O'rtai, governor general of Yun-Gui. The first year of the Yongzheng reign (1723) saw the abolition of the native prefecture of Lijiang, followed by the disappearances of the native department of Weiyuan in 1724 and also of the native prefecture of Yaoan in 1725. The six years between 1726 and 1731 saw the climax of this state project, when many areas of northeastern and southern Yunnan experienced the removal of native chieftainship.

102 alt textMap 4.2 Yunnan Province in the Early Nineteenth Century In 1726, O'rtai presented a report to the emperor on the reform of native chieftains, which was to "free" ethnic peoples from the "brutal rule" of native chieftains and enhance central control over frontier society. His suggestion was immediately accepted by Emperor Yongzheng. As governor in charge of Yunnan, Guizhou, and Guangxi (1728), O'rtai was assigned to take charge of this large project in which Yunnan was the main focus. In summer 1726, O'rtai arrested the native subprefect (tu zhizhou) of Zhanyi and the native prefect of Zhenyuan, confiscated their farms and assets, and exiled them to Jiangnan.134 Under enormous state pressure, Dao Liandou, commissioner (zhangguan) of Zheledian, "voluntarily" handed over his seal and pleaded for direct administration.135 At the same time, O'rtai dispatched troops to Dongchuan, thus abolishing native chief (tumu) who had controlled native forces there. Forces were again used to get rid of native chieftains in Wumeng and Zhenxiong.136

103While native titles and offices were immediately abolished, the power and influence of native chieftains could not be eradicated so quickly. Lu Dingkun, a native chief in Wumeng, took advantage of the local abusive Qing troops, and rebelled in 1730. Lu was the uncle of Lu Wanzhong, former native prefect of Wumeng, and he helped the Qing to suppress Lu Wanzhong and other native chieftains in Zhenyuan and Cheli. He probably expected to be promoted to the position of native prefect; however, the Qing decided to move him to Henan. Disappointed, he and his son, Lu Wanfu, called on the native chief to revolt. The rebels took over Wumeng and killed the Qing generals. This event immediately sparked rebellion in Zhenxiong and Dongchuan, the two recently reformed prefectures.137 As a result, the whole province was shaken. O'rtai responded by mobilizing a force of 20,000 soldiers from Yunnan and Guizhou, including imperial and native troops. The Qing brutally brought down the rebellion while some rebels fled to the Big Cool Mountains (Daliangshan) in Sichuan.138

104In addition to northeastern Yunnan, where Wumeng, Chuxiong, and Dongchuan prefectures were located, southern Yunnan became the other major focus. In 1726, the native chieftain of Zhenyuan prefecture was replaced with Liu Hongdu, who was assigned as subprefect (tongzhi) in charge of the prefecture. He attempted to implant the Chinese systems to Zhenyuan, which led to much discontent among the indigenous people. This overtaxation made the indigenous population angry, vowing secretly swore to kill him. Liu was duly informed but he did not believe it. On the seventeenth night of the first month (1728), when Liu was enjoying a local opera, the rebels flowed in. Liu hid himself near the stable. When discovered, he told the rebels, "What you wanted was nothing but the seal. Just take it," and presented the seal. The rebels answered, "We want the seal as well as your death!" Liu was killed and his heart was reportedly eaten. Many other military officials were killed as well. When the Qing dispatched troops, Dao Ruzhen, the leader of this uprising, surrendered himself, as he claimed his hatred was quenched after killing Liu. Dao was persecuted.139 After suppressing the Zhenyuan uprising, the Qing troops marched southward to put down another native rebellion in Cheli. As a result, a new prefecture of Puer was established. It was the furthest southward reform made by the Qing.

105While gaitu guiliu was a radical and aggressive state move to abolish native chieftains, it was by no means arbitrary and rash. O'rtai followed his own principles identifying areas that would benefit from reform. He chose only those areas in which he thought the conditions for reform were ripe and the successful implementation of direct administration would be almost certain. The general principle of the project boiled down to one sentence, that is, lands east of the Lancang River were suitable for direct administration, while lands west to the Lancang River were suitable for native chieftainship (jiangnei yi liu bu yi yu, jiangwai yitu bu yi liu). This principle was based on local and societal changes. With this spatial boundary in his mind, the fate of native chieftains in Yunnan was determined. By the early eighteenth century, Han migrants and their offspring had scattered throughout many areas of Yunnan. Social and ecocultural changes had taken place, and provided the demographic and cultural basis for administrative changes.140 Furthermore, the Qing Empire had just entered its golden age. With abundant resources, the Qing had a military force powerful enough to smash native uprisings.

106Gaitu guiliu was completed during the Yongzheng rein, and it drastically transformed the administrative structure in Yunnan. The number of native chieftains was reduced to twenty-two, and most of them were located in south and southwest Yunnan, where many native chieftains had remained loosely related to the Qing.141 Some of them simultaneously paid tribute to Burma. A few in interior Yunnan who kept their posts were largely contained. Although most native chieftains were abolished, which was a major victory by the central authority over local power, one should bear in mind that the influence and prestige of native chieftains lingered for quite a while. Thus, a kind of informal power and cultural prestige among ethnic people died away very slowly. Native chiefs still dominated daily lives in ethnic villages, clans, and tribes.


107The historic changes in the relationship between China and Yunnan, between the state and the frontier, and between the state and native chieftains demonstrate the strengthening of central presence and authority over time on the frontier. The Han Dynasty initiated military colonization, but native chieftains enjoyed a high degree of autonomy, and resistance and rebellions broke out one after the other only after the court increased exploitation, or the Han officials squeezed too much from the local peoples. The Shu-Han regime, after the southern expedition, established a relatively effective policy, utilizing native elites, troops, and resources for its national unification project.

108From the mid-third century onward, Chinese states seldom exerted influence over Yunnan. The Sui and early Tang both began to resume and further the Han achievements; however, their efforts were stopped either by the collapse of the Sui or by the rise of Tubo. The unification of Yunnan by the Nanzhao Kingdom began a new era. From then until the Mongol conquest, Yunnan was independent from Chinese empires. The Mongol expedition was a turning point as it brought Yunnan into China proper. The rise and fall of Chinese empires did not change the status of Yunnan as an administrative part of China. In addition, the Mongols successfully initiated a series of policies in Yunnan that were followed and developed by late imperial states.

109The transformation of the state–native chieftain relationship during the Yuan, Ming, and Qing periods attests to the advent of the central penetration in Yunnan. In the Yuan period, there was a provincial-level chieftain, namely, the Duans; in the Ming period there was not such an influential native figure, but there were many prefectural-level native chieftains. During the Qing period, native chieftains largely held county-level posts and below. Consequently, the areas they ruled were rarely larger than a county. In addition, the decreasing numbers and ranks of native chieftains uncovered the increasing central state control.

110The presence of imperial princes in the Yuan and Ming periods also demonstrates this growing central control. The prince of Yunnan and prince of Liang were powerful in the Yuan Dynasty. In the Ming period, while the Mus enjoyed military leadership and many privileges, they were limited and contained by administrative institutions and imperial codes. In the Qing, except for the Wu Sangui regime, there was no imperial prince in Yunnan.

111Finally, the establishment of Guizhou Province in 1413 divided the vast frontier area into two provincial units, which made Yunnan much more manageable. The decision signaled the basic administrative pattern in Southwest China. Since then, this provincial pattern in the Southwest has remained.

112The process of administrative institutional changes in Yunnan has been accompanied by the growing infiltration of the state, with the general trend indicating that central control moved from north to south, and from urban to rural, mainly along trade routes. What supported such a dramatic administrative change was the demographic transformation brought about by Chinese immigration.


Note 1: SJ, juan 116, in YNSLCK 1: 4. back

Note 2: Yu 1967, 65-82. back

Note 3: Yu 1967, 77. back

Note 4: For the above rebellions, see HS, juan 95, in YNSLCK 1: 32. back

Note 5: Ibid. back

Note 6: Ibid. back

Note 7: HS, juan 95, in YNSLCK 1: 33. back

Note 8: HHS, juan 86, in YNSLCK 1: 58. back

Note 9: Ibid. back

Note 10: HHS, juan 86, in YNSLCK 1: 63. back

Note 11: Ibid. back

Note 12: HHS, juan 86, in YNSLCK 1: 58. back

Note 13: HHS, juan 86, in YNSLCK 1: 61. back

Note 14: HHS, juan 86, in YNSLCK 1: 63. back

Note 15: Yu Yingshi 1967, 83. back

Note 16: For Yishuai and Daxing, for instance, see Fang Guoyu 2001, 1: 355-389; You Zhong 1990, 82-90. back

Note 17: For discussions of the Cuans, see Fang Guoyu 2001, 1: 459-501. back

Note 18: YNSLCK 1: 233-240. back

Note 19: Charles Backus 7. There is a debate whether the Cuans were native or not. Fang Guoyu, for example, regards the Cuans as native. See Fang Guoyu 2001, 1: 458-501. back

Note 20: I will discuss these interactions later on in the chapter. back

Note 21: SGZ, Shu Shu, juan 13, in YNSLCK 1: 104; HYGZ, juan 4, in YNSLCK 1: 254. back

Note 22: SGZ, Shu Shu, juan 13, in YNSLCK 1: 104; SGZ Shu Shu, juan 3, in YNSLCK 1: 108; Li Hu, "Shuhan 'Nanzhong' Zhengce Ersanshi" (On a couple of issues of Shuhan's Nanzhong policy), Lishi Yanjiu 4 (1984): 153-166. back

Note 23: The only exception was Xu Wensheng, the governor of Ningzhou, who seemed to have some achievements. He even enlisted local people to suppress rebels in China proper. See Liang Shu (History of Liang) juan 46, in YNSLCK 1: 152. For Chinese governors of this period assigned to Ningzhou, see YNSLCK 1: 157-160. back

Note 24: XTS, juan 222, in YNSLCK 1: 403. back

Note 25: For the two expeditions, see Backus 1981, 10-13. back

Note 26: Sui Shu, juan 62, in YNSLCK 1: 337-338. back

Note 27: Tai Ping Guang Ji, juan 483, in YNSLCK 2: 194. back

Note 28: Man Shu, in YNSLCK 2: 76. back

Note 29: For Buddhism in Nanzhao and Dali, for example, see Fang Guoyu 2001, 2: 507-522; 523-540; 541-557; 557-562; 563-582; 583-601; 601-636; Chapin, Helen B., "Yünnanese Images of Avalokitesvara," Harvard Journal of Asiatic Studies 2 (1944): 131-187; Chapin, Helen B., and Alexander C. Soper, "A Long Roll of Buddhist Images," Artibus Asiae 32 (1970): 5-41, 157-199, 259-306; and 33 (1971): 75-142; and Li Linchan, A Study of the Nan-Chao and Ta-Li Kingdoms in the Light of Art Materials Found in Various Museums (Taibei: National Palace Museum, 1982). Buddhism was a key feature in the Nanzhao and Dali kingdoms and apparently a field that has not been fully explored. It is a pity for the author that the current book has not yet scrutinized this issue. back

Note 30: Nanzhao Yeshi, in YNSLCK 4: 782. back

Note 31: For the Gaos, see Fang Guoyu 2001, 2: 470-506. back

Note 32: Yuan Shi, juan 4, in YNSLCK 2: 484; juan 121, in YNSLCK 2: 546. back

Note 33: Yuan Shi, juan 126, in YNSLCK 2: 567. For the Duan-Yuan (Mongolian) relationship, see Fang Hui 1999. back

Note 34: Yuan Shi, juan 126, in YNSLCK 2: 567. back

Note 35: Yuan Shi, juan 121, in YNSLCK 2: 547. back

Note 36: The Bo people are regarded as the predecessors of modern Bai ethnicity. Other native troops were the Moxie army (naxi), the Luoluosi Army (yi), and the Huoni Army (hani). back

Note 37: Some of the Cuan-Bo army settled in Sangzhi, Hunan Province. He Long, one of the ten marshals of the People's Liberation Army, is said to be a descendant of the settlers. back

Note 38: Yuan Shi, juan 6, in YNSLCK 2: 485. back

Note 39: The Mongol Yuan Dynasty paid extreme attention to Yunnan. Xia Guangnan counted the officials in Yuan Shi (History of the Yuan Dynasty) and found that a hundred officials, or one-seventh of those who had their biographies recorded in Yuan Shi, had served in Yunnan. See Xia Guangnan 1968, preface, and 75-105. back

Note 40: Yuan Shi, juan 7, in YNSLCK 2: 485; juan 167, in YNSLCK 2: 562. back

Note 41: For Sayyid'Ajall Shams Al-Din and his experience in Yunnan, see Armijo-Hussein, Jacqueline Misty, "Sayyid'Ajall Shams Al-Din: A Muslim from Central Asia, Serving the Mongols in China and Bringing 'Civilization' to Yunnan," unpublished dissertation, Harvard University, 1996. back

Note 42: Yuan Shi, juan 125, in YNSLCK 2: 556. back

Note 43: Yuan Shi, juan 125, in YNSLCK 2: 557. back

Note 44: For the above measures, see Yuan Shi, juan 125, in YNSLCK 2: 556-557; for the tusi system, Gong Yin in his review article has listed Chinese sources and works; see Gong Yin 2002, "20 Shiji Zhonguo Tusi Zhidu Yanjiu de Lilun yu Fangfa" (Theories and approaches in studies of the Chinese native chieftain system in the twentieth century), Sixiang Zhanxian 28, no. 5 (2002): 100. back

Note 45: Native chieftains were also set up in the northwest, but they mainly were military in nature. back

Note 46: Imperial officials were ranked based on the prestige and duty of their positions. Each rank was divided into two classes, a and b. Within this nine-rank system, rank 1a was the highest and rank 9b the lowest. back

Note 47: Yuan Shi, juan 103, 2635; juan 26, in YNSLCK 2: 503. back

Note 48: Yuan Shi, juan 103, 2635. back

Note 49: For the Duan family in the Yuan Dynasty, see Fang Hui, Dali Zongguan Duanshi Shicinianli jiqi yu Mengyuan Zhengquan Guanxi Yanjiu (A study of the Duan family and its relations with the Mongol court) (Kunming: Yunnan Jiaoyu Chubanshe, 1999). back

Note 50: Prince Yunnan was stationed in Dali, and Prince Liang in Kunming. Since 1330, Prince Yunnan had lived in Kunming, implying that the Duans had controlled the Dali region. In addition, other princes (zhuwang) were stationed either in Dali or in Kunming. Altogether, twenty-seven princes stayed in Yunnan during the Yuan Dynasty. See Xia Guangnan 1968, 68-71; Xin Fachun, Ming Mushi yu Zhongguo Yunnan zhi Kaifa (The Mus and the exploitation of Ming China's Yunnan) (Taibei: Wenshizhe Chubanshe, 1985), 8-9. back

Note 51: Since the 1330s, Prince Liang indeed had been the real Mongolian ruler of Yunnan. The court of Prince Liang took the place of the office of Yunnan Province. back

Note 52: For the Mongolian and Muslim immigration, for example, see Xia Guangnan 1968, 43-60. Today many Chinese Hui communities in Yunnan claim themselves descendants of Sayyid'Ajall Shams Al-Din. back

Note 53: For the Muslim caravan, for example, see Andrew Forbes and David Henley 1997. back

Note 54: A deep pity of and a possible challenge to this book may be the missing discussion of the Muslims in Yunnan, especially of the uprising (1856–1874). But a close examination would confirm my argument that a new local identity was the key to the understanding of modern Yunnan. The Manchu Qing was able to suppress the Muslim Uprising only because it won the support of several key figures of the Muslim elites. back

Note 55: Nanzhao Yeshi, in YNSLCK 4: 798; Fang Hui 1999, 105. back

Note 56: Nanzhao Yeshi, in YNSLCK 4: 798. back

Note 57: "Tuguan Dibo Yunnan Tuguan" (The list of native chieftains), in YNSLCK 5: 387. Duan Bao was assigned as native subprefect (tu zhizhou) of Yunlong Department. back

Note 58: Tan Qian, Guo Que, juan 7, in YNSLCK 5: 10. back

Note 59: The Wei-Suo system basically follows the Mongol Wanhu-Qianhu (myriarch-chiliarch) system. back

Note 60: Lu Ren 2001, 13. back

Note 61: Zhu Yuanzhang was one of the most suspicious emperors in Chinese history. Almost all his crucial generals and assistants who helped Zhu ascend the throne were either executed or forced to death by various means. back

Note 62: Ming Shi, juan 126, in YNSLCK 3: 390. back

Note 63: Some imperial scholars suspected that Mu was a natural son of Zhu, but this suggestion was refuted by the relative age difference between them. Because of this relationship, Mu was one of the few generals who died in peace. back

Note 64: Ming Shi, juan 126, in YNSLCK 3: 390; Mingtaizu Shilu, juan 258, in YNSLCK 4: 18. back

Note 65: Ming Shi, juan 126, in YNSLCK 3: 389-394. back

Note 66: The Mus sometimes just took lands reclaimed by troops into its farms. A1529 source reads that the soldier of the six Wei in Yunnan had become its private labors. Mingshizong Jiajing Shilu, juan 98. back

Note 67: Ming Shi, juan 126, in YNSLCK 3: 391. back

Note 68: Wang Yuxuan, "Ming Qianguogong Mushi Zhuangtian Kao" (An examination of the Mu farms in the Ming Dynasty), in Mingdai Shehui Jingjishi Lunji (Collections of the social economic history of the Ming Dynasty) (Hong Kong: Chongwen Bookstore, 1975), Vol. 1, 116. back

Note 69: Wang Yuxuan 1975, 117. back

Note 70: Wang Yuxuan 1975, 118. Qing, equal to 100 mu, was a unit of area in imperial China. The size varied over time. One mu is 0.0667 hectares in modern time. back

Note 71: Lu Ren 2001, 128-29. back

Note 72: Deng Mei, Nanzhong Coudu; cf. Lu Ren 2001, 127. back

Note 73: Mingxizong Shilu, juan 49 and juan 58, in YNSLCK 4: 94. back

Note 74: Ming Shi, juan 118, in YNSLCK 3: 387; juan 126, in YNSLCK 3: 391. back

Note 75: Ming Shi, juan 126, in YNSLCK 3: 392. back

Note 76: Ming Shi, juan 126, in YNSLCK 3: 389-393; Xin Fachun 1985, 77-220. back

Note 77: Ming Shi, juan 126, in YNSLCK 3: 389-393. For the Mus in Yunnan, also see Fang Guoyu "Ming Zhenshou Yunnan Mushi Shiji," in YNSLCK 3: 685-691; "Ming Qianningwang Mushi Shixi Shilue," in YNSLCK 3:691-695. back

Note 78: During the Ming Dynasty, governor general (zongdu) was a temporal position, for example, for a military expedition. It was only during the Qing period that the governor general became a regular office. back

Note 79: For the discussion of the Wei-Suo in Yunnan, see Fang Guoyu 2001, 3/145-318; Lu Ren 2001, 40-41. back

Note 80: Ming Huidian, juan 18, in YNSLCK 3: 705. The above ratio and figure vary locally. back

Note 81: Lu Ren 2001, 45. back

Note 82: The reasons for the establishment of Guizhou were not clear. The short-term consideration may have been to secure the provision line for the Ming campaign in the southern border land of Yunnan. Moreover, Yunnan was territorially too large to govern, and dividing it into two parts also contained the power of the Mus. back

Note 83: Yuan Shi, juan 26, in YNSLCK 2: 503. back

Note 84: Yuan Shi, juan 29, in YNSLCK 2: 504. back

Note 85: Yuan Shi, juan 29, in YNSLCK 2: 505. back

Note 86: Yuan Shi, juan 103, 2635. back

Note 87: John Herman 1997, 50. Dynasty histories may support his point. There is no juan in Yuan Shi devoted to native chieftains; while ten juan in Ming Shi were for biographies of native chieftains. back

Note 88: Ming Shi, juan 313, in YNSLCK 3: 436. back

Note 89: Between 1397 and 1530, both these two kinds of chieftains were supervised by the Ministry of War. back

Note 90: Gong Yin, Zhongguo Tusi Zhidu (The native chieftain system in China) (Kunming: Minzu Chubanshe, 1992), 460-461; for tusi in the Ming, also see Jiang Yingliang, Mingdai Yunnan Jingnei de Tuguan yu Tusi (Native chieftains in Yunnan of the Ming Period) (Kunming: Yunnan Renmin Chubanshe, 1958). back

Note 91: Gong Yin, 1992, 58-61. back

Note 92: Gong Yin, 1992, 65-66. back

Note 93: Mingyingzong Shilu, juan 27, in YNSLCK 4: 49; for the regulations concerning the inheritance of native chieftainship, also see Ming Huidian, juan 121, in YNSLCK 3: 729-732. back

Note 94: Ming Huidian, juan 121 in YNSLCK 3: 729-730. back

Note 95: Ming Huidian (The Ming statues), juan 8, 4b-5a, in Yingyin Wenyuange Siku Quanshu (Taibei: Taiwan Shangwu Shuguan, 1983), Vol. 617. back

Note 96: Mingxizong Shilu, juan 273, in YNSLCK 4: 52. back

Note 97: Ming Huidian, juan 121, in YNSLCK 3: 730. back

Note 98: Mingwuzong Shilu, juan 7, juan 178, in YNSLCK 4: 52. back

Note 99: Mingshizong Shilu, juan 112, in YNSLCK 3: 56. back

Note 100: Ming Huidian, juan 121, in YNSLCK 3: 730. back

Note 101: Ming Huidian, juan 121, in YNSLCK 3: 731. back

Note 102: Ming Shi, juan 313, in YNSLCK 3: 443. back

Note 103: "Tuguan Dibo Yunnan Tuguan" (The list of native chieftains), in YNSLCK 5: 420. back

Note 104: Ming Shi, juan 314, in YNSLCK 3: 450; "Tuguan Dibo Yunnan Tuguan," in YNSLCK 5: 416. back

Note 105: Ming Shi, juan 314, in YNSLCK 3: 452-453. back

Note 106: "Tuguan Dibo Yunnan Tuguan," in YNSLCK 5: 415-416. back

Note 107: Ming Shi, juan 314, in YNSLCK 3: 452. back

Note 108: "Tuguan Dibo Yunnan Tuguan," in YNSLCK 5: 403. back

Note 109: Ming Shi, juan 313, in YNSLCK 3: 449. back

Note 110: Ming Shi, juan 314, in YNSLCK 3: 451-453. back

Note 111: Ming Shi, juan 313, in YNSLCK 3: 446. back

Note 112: Gong Yin 1992, 461. back

Note 113: Ming Shi, juan 313, in YNSLCK 3: 441-442. back

Note 114: Ming Shi, juan 313, in YNSLCK 3: 440. back

Note 115: Mingtaizu Shilu, juan 167, in YNSLCK 4: 26-27. back

Note 116: Ming Xuanzong Shilu, juan 71, in YNSLCK 4: 49. back

Note 117: Herman 1997. back

Note 118: For the Feng family that occupied the hereditary position of native prefect of Wuding, see Fang Guoyu 2003, 3: 559-580. back

Note 119: Ming Shi, juan 314, in YNSLCK 3: 451-453; Ming Shilu, "Wudingfu," in YNSLCK 4: 253-255. back

Note 120: Mingtaizu Shilu, juan 204, in YNSLCK 4: 492-493. back

Note 121: Mingtaizu Shilu, juan 239, in YNSLCK 4: 493. back

Note 122: Ming Huidian, juan 78, in YNSLCK 3:720; Mingxianzong Shilu, juan 212, in YNSLCK 4: 499. back

Note 123: Mingtaizu Shilu, juan 239, in YNSLCK 4: 493. back

Note 124: Mingtaizu Shilu, juan 239, in YNSLCK 4: 493. back

Note 125: Ming Shi, juan 314, in YNSLCK 3: 454. back

Note 126: John Herman 1997. back

Note 127: For the list of native chieftains of Yunnan during the Qing, see QSG, juan 301, 4056-4062; Gong Yin 1992, 457-742. back

Note 128: John Herman 1997. back

Note 129: John Herman 1997. back

Note 130: For the list of those native chieftains, see Gong Yin 1992, 462. back

Note 131: Gong Yin 1992, 113-114. The figures of native chieftains in Yunnan during the Qing provided by Gong are not consistent. The figure 273 is based on pages 113 and 114 (157 military native chieftains and 80 civilian native chieftains), plus 36 lower than rank 9. On page 462, Gong's conclusion was about 210. back

Note 132: Gong 1992, 58 and 61. Gong gives the number of 332 on page 461. back

Note 133: You Zhong 2001, 517. back

Note 134: Ni Tui, 588. back

Note 135: Ni Tui, 586. back

Note 136: Ni Tui, 584-589. back

Note 137: QSG, juan 301, 4054-4055; Ni Tui, 582-585, 588, and 608-612. back

Note 138: You Zhong 1991, 521. back

Note 139: For this event, see Ni Tui, 592-593. back

Note 140: For migrations and the transformation of society, see chapter five. back

Note 141: Gong Yin 1992, 463. The twenty-two tusi were in Cheli, Gengma, Longchuan, Ganya (Yingjiang), Nandian, Menglian, Zhefang, Zhanda, Lujiang, Mangshi (Luxi), Mengmao (Ruili), Nalou, Kuirong, Shierguan, Menghua, Jingdong, Mengding, Yongning, Fuzhou, Wandian, Zhenkang, and Beishengzhou. back