INTRODUCTION

Only after the gradual development of the Socialist economic, cultural, scientific, and technological enterprise and of a Socialist civilization with its own material and spiritual values, will the type of society and level of awareness that gave rise to the existence of religion gradually disappear. . . . At that time, the vast majority of our citizens will be able to deal with the world and our fellow men from a conscious scientific viewpoint, and no longer have any need for recourse to an illusory world of gods to seek spiritual solace. This is precisely what Marx and Engels have predicted—that there will be an age when people will have freed themselves from all alienating forces controlling the world and will have come to the stage when they will consciously plan and control the whole of social life. This is also what Comrade Mao Zedong meant when he said that the people, relying on themselves alone, will create a new age both for themselves and for the world. Only when we enter this new age will all that shows a religious face in the present world finally disappear.

—1982 Communist Party Document 19

More than twenty-five years have passed since the Chinese Communist Party (CCP) expressed its confident view on the eventual disappearance of religion in China.¹ Chinese people have certainly experienced breathtaking economic growth and modernization of their society under a period of “Socialism with Chinese Characteristics.” A policy shift from a planned to a market economy has had a tremendous impact on the living standard of the large majority of Chinese. Nevertheless, until now, there have been few signs that religious beliefs are declining. On the contrary, the relaxation of religious policies and the reinstatement of religious freedom in the 1982 revision of the constitution have set the stage for an explosive development of religions, whether traditional Chinese folk religion, institutionalized world religions, or new syncretic creations such as Falungong.² This development comes with a highly visible revival of ritual practices in Chinese cities as well as all over the vast countryside. Rather than being a waning adversary of unfolding Chinese modernity, it seems then that religion is an integral component of it. As Kenneth Dean has convincingly demonstrated in the case of Daoist ritual revival in Southeast China, “tradition” is not necessarily opposed to “modernity,” but reemerging ritual practices are in a “ceaseless negotiation with the forces of modernity” (2003: 358).

The process of modernization is taking place not only in the coastal regions but all over China, although not at the same pace everywhere. While the more remote corners of the territory such as the mountains of the Southwest still lag significantly behind the rest of China, in recent years, economic development has also started to pick up momentum here. In this part of the country, where a large portion of China’s ethnic minorities live, the policies of economic liberalization and ideological relaxation are having a profound effect on the lives and cultures of local communities. Interestingly, it is also possible here to observe a remarkable resurgence of religious activity.

Indeed, my initial worries about the imminent extinction of the hangui/anji tradition among the Premi in China’s southwestern provinces of Yunnan and Sichuan have been proved wide of the mark.³ When I visited Chicken Foot Village in Ninglang County, Yunnan, in 1998, I was told that the last hangui ritual specialist in the village had just passed away and none of the young people in the village seemed interested in taking over his job. On my latest visit in 2006, there was no doubt that changes had taken place in the village. Although basically still sticking to the traditional local building style, many houses now boasted windows containing real glass, and not a few were adorned with a satellite dish on top. In the enclosed courtyards, where once the family bicycles had been parked on the dirt ground, there now stood shiny motorbikes on neat concrete pavement. Other changes were less visible but no less remarkable, and some were certainly more unexpected: Chicken Foot Village now could claim seven practicing young hangui ritual specialists! They could be seen conducting rituals when people or animals were sick, when people had died and were cremated, when ancestral spirits had to be propitiated, or when offerings had to be made to the water deities, or lwéjabu, in order to bring rain for the newly planted crops. Often the hangui would be wearing their newly sewn, long yellow gowns, and during some rituals they would recite from photocopies of Tibetan-language scriptures. And this development was not limited to Chicken Foot: a dilapidated school building in the middle of the village had been turned into a veritable hangui school, training young Premi men from all over Ninglang County on becoming ritual specialists. From an almost obsolete tradition that, fifteen years ago, had been reduced to a memory in the minds of older people, the hangui ritual practice had turned into a vibrant cultural phenomenon spreading to all corners and layers of Premi society. Why, in a time of palpable social and economic development, were Premi villagers still not “able to deal with the world and our fellow men from a conscious scientific viewpoint,” and why did they still have “need for recourse to an illusory world of gods to seek spiritual solace”?

The hangui tradition was not the only important religious tradition found among the Premi. As with several other ethnic minority groups in Southwest China, such as the Na, the Naxi, the Rek’ua, the Namuyi, and the Shuhiⁿ, non-Buddhist religious practices seem to coexist with the strong presence of Tibetan Mahayana Buddhism. These different ethnic groups live in a mountainous region situated in what today are the Chinese provinces of Sichuan and Yunnan. The area constitutes the border area of historic Tibet, and throughout their history, the people living here have been exposed in varying degrees to Tibetan culture. This is most obvious in the presence of several monasteries belonging to different schools of Tibetan Buddhism. Nevertheless, until today, many of these ethnic groups maintained their own cultural distinctiveness, apparent in the use of a number of non-Bodic (or Tibetan) Tibeto-Burman languages and the existence of ritual practices and beliefs that can be defined as non-Buddhist.

Along with Chicken Foot Village in Ninglang County, Yunnan, my main area of investigation has been the adjacent county of Muli. Even today, its high mountains and deep gorges have kept Muli relatively isolated from the outside world. Before the Chinese Communist takeover in 1950, Muli was a semi-independent monastic domain making up the southeastern border region of the Tibetan province of Kham. It was ruled by a Buddhist monk who carried the Tibetan title of “king,” or gyelpo. Buddhism had been established in this region since at least the fifteenth century, but it was only in the seventeenth century that Muli was firmly integrated into the monastic system of the Gelugpa, a school of Tibetan Buddhism founded by Tsongkhapa (1357–1419).

The county of Muli is situated in the southwestern corner of the Chinese province of Sichuan and is designated a Tibetan autonomous county. Despite its Buddhist history, the local people belonging to different Tibeto-Burman ethnic groups participate only sporadically in monastic Buddhism. Premi and Na people from Bustling Township⁴ in the south of the county visit the local Buddhist monastery once in a while during festivals, but only a few families in the township send their sons to the monastery to become monks. Nevertheless, when people in Bustling Township die or must make offerings to local deities, Buddhist lay priests or Buddhist monks living in the villages conduct rituals and recite Tibetan Buddhist scriptures. Interestingly, among the Premi within this township, such ritual occasions also involve the participation of non-Buddhist ritual experts—called anji—who might use the same Buddhist scriptures in their practice but are also likely to recite Bön⁵ texts or orally transmitted “texts” in the local Tibeto-Burman language. Anji is a local pronunciation of hangui, the word used in Ninglang County for designating Premi religious experts.

Similar cases of religious coexistence have been reported in other peripheral areas of cultural Tibet, such as Nepal (see, e.g., Mumford 1989; Ramble 1990; Holmberg 1989; Vinding 1998). Nevertheless, it is precisely the comparison with studies on religious practice in a country such as Nepal that brings a major difference to light. In studies on the relationship between Buddhism and local cults in Nepal, the state is almost completely absent. An approach that leaves out the role of the policies, institutions, ideologies, and discourses of the state would be impossible in the case of China. In China, people’s lives have been subject to one of the most widespread, radical, and comprehensive attempts at state-controlled social engineering in recent human history. Among the various aspects of social life, religious practice has been one of the selected targets of Communist attempts to change society. Even in remote Muli, in order to understand religious practice, it was necessary to take into account the area’s location within the territory of the People’s Republic of China. The presence of the state is therefore a connecting thread that runs through this book.

In the course of my research, it became clear that the initial questions about religious coexistence were based on debatable assumptions about religious duality and opposition between orthodoxy and heresy. They nonetheless provided an important heuristic device: by asking the “wrong” questions, I obtained invaluable knowledge and understanding and altered the direction of the research, replacing my initial exploration of religion and beliefs with a focus on the believers and their conceptions of religion. Rather than looking into the process and mechanisms of religious revival, I became concerned with Premi religion as it was practiced and how those involved understood and viewed this practice.

The monks of the major Gelugpa monastery in Muli did indeed perceive the relationship between anji practices and Buddhism in terms of opposition and orthodoxy. In this respect, their views were remarkably concurrent with those presented in the official Chinese state discourse on religion and ethnic identity in Muli, a discourse the local intellectual and political elites almost uncritically subscribed to and reproduced. However, once I left the monastery, the government buildings, and the state schools behind and entered the houses of Bustling Township’s villages, it did not take long to discover that conceptions of what constituted religion in Muli were quite different. This was the case both with regard to cosmology and beliefs and with regard to what were correct ritual practices. Villagers and local ritualists did not view religious practices in the village as contradicting Buddhist doctrine. They were unimpressed by the fact that many village practices were condemned as heresies by the monks from the monastery, such as the worship of local so-called living deities, and the ritual killing and offering of animals.

Over the years during which I conducted fieldwork for this book, participation in monastic Buddhism was waning among the villagers of Bustling Township. The local anji ritualists, in contrast, had many new students, and there was increasing demand for their religious services among the villagers. It would be no exaggeration to state that at the beginning of the millennium, anji practice was experiencing a real revival. The resurgence of local traditional ritual practices, at the level of both the household and the village, took place at a time when villagers were experiencing increasing socioeconomic remarginalization of their community. Without wanting to overstate direct causal links between religious revival and a sense—whether justified or not—of being left out of the wonders of the booming Chinese economy, resurgence of local religious practice among the Premi of Bustling Township cannot be fully understood without taking into account the context of more general processes of modernization and economic liberalization in China.

The end of the Cultural Revolution in 1976 and the subsequent Eleventh Party Plenum of 1978 heralded the end of a period of extremist policies during which most forms of religious practice disappeared from the public sphere. Nevertheless, it was not until the new constitution of 1982 guaranteeing freedom of religion and the circulation of the above-mentioned Party Document 19 that religious practice gradually reemerged into the open. A hesitant revival developed into a veritable religious craze during the 1980s and 1990s. Some researchers—especially those studying Tibetan communities—explained this as a spontaneous resurrection of beliefs and practices that had been forbidden by the powers in charge but had continued to exist in people’s minds (see, e.g., Germano 1998; Kapstein 2004). Others, focusing mainly on religion among the Han, interpreted the revival as a response to the ideological vacuum left by the debacle of the Cultural Revolution and the demise of high Maoism (see, e.g., Anagnost 1994; Feuchtwang 2000; Dorfman 1996).

Undoubtedly, all these approaches can contribute to our understanding of why and how religious practices resurfaced and quickly expanded all over China. In recent years, an increasing number of studies from different places across China’s vast countryside have been more concerned with the socioeconomic context of the religious revival. They place the revival—or, in some cases, reinvention—of local practices in the context of the social insecurity that followed economic reforms and the related breakdown of the system of state-organized social security. In a new chapter on religious practice in post-Mao China, added to the most recent edition of his seminal book The Imperial Metaphor, Stephan Feuchtwang attributes the revival of territorial cults in the countryside to farmers who, blaming the government for mismanagement of or inaction in response to recurrent natural disasters that threaten their existence, seek relief and protection in local cults. At the same time, he points to the social dimensions of the quest for an alternative soteriology: “the resurgence of religious activity forms other communities of faith, alternative sources of moral authority and senses of security” (2001: 240). Adam Yuet Chau argues that revived popular religious activities such as temple festivals of folk religion in northern Shaanxi “encourage and facilitate a kind of sociality radically different from that of the Maoist era” and that “instead of responding to state-imposed political ideals and campaign goals, villagers today are engaged in social interactions based on kinship or community obligations and responsibilities” (2005: 237).

In studies of communities classified by the Chinese state as ethnic minorities, religious revival is viewed as having an important social aspect, in relation not only to the local community but also to the larger ethnic group. Ben Hillman makes this point in his study of a Hui community in northwestern Yunnan. Even though Islamic practice had not been in evidence in the community for many decades, villagers decided in the late 1990s to build a new mosque and revive Islam. This project was set in motion by a sense of being left behind by the county leadership’s ambitious undertaking of promoting and commodifying the Tibetan character of the area. Even the name of the county has been changed from Zhongdian to Xianggelila (Shangri-la) in order to promote tourism. This Islamic revival—or, as Hillman conceives it, “invention”—becomes a socially meaningful experience for these Hui in that it produces a strong sense of community, creates economic opportunities, and enhances their sense of ethnic pride vis-à-vis the local Tibetans, Naxi, and Han (Hillman 2004).

The process of reviving local religious practice, especially traditions that were based mainly on oral transmission, is not a straightforward act of again taking up the threads that were severed in the tumultuous and traumatic years of the Mao era. The social and political context in which remembering takes place might determine much of the nature of such renewed practices. Jing Jun has compellingly laid bare the workings of social memory in a reconstruction of ritual in a village in Northwest China: “At first glance, the story of Dachuan’s Confucius temple may suggest that many traditional ideas and practices prohibited under the first decades of Communist rule have been revived. But a closer look reveals a far more complicated picture. These ideas and practices are not mechanically retrieved from the past; they are blended with cultural inventions, shaped by the local experience of Maoism, and permeated with contemporary concerns” (1996: 12).

The burning of ritual texts, the destruction of sacred objects and religious buildings, the public criticism sessions of religious believers and clergy, the ridiculing of traditional practices, the humiliation of ritual specialists—all these events have left deep scars on the memories of local communities. In Bustling Township, religion and ritual were a cause of latent stress throughout the Maoist period: either because people took their chances with facing prosecution or severe criticism and continued to pray or conduct rituals in frightened secrecy or because they took chances and risked otherworldly retribution for not conducting rituals. The community’s absorption of these traumatic experiences is an important factor in defining the shape of a revival of local religious practice. These are also recurrent conclusions from recent anthropological studies from very different parts of China. The dehumanizing consequences of Maoist collectivism are perhaps nowhere as telling and graphic as in Jing Jun’s study of Dachuan (1996). The members of the Kong clan of Dachuan witnessed the literal disappearance of the physical manifestation of their existence as a community when their entire village, with its Confucian temple and ancestral graves, was submerged as a result of the construction of a hydroelectric dam in 1960. The sufferings of forced resettlement and, in addition, the political humiliation of being associated with Confucius at the wrong point in history, infused the Kongs with a strong determination to commemorate and restore through the act of remembering. It was the reconstruction of a temple and the revival of the Confucius cult that signified communal recovery for the Kongs of Dachuan.

The most elaborate narration of how ritual is intertwined with a community’s suffering as a result of Maoist policies is undeniably Eric Mueggler’s The Age of Wild Ghosts (2001). The people of a Lòlop’o⁶ community in Yunnan imagine a ghostly return of an institution abolished shortly after the terrible famine of the Great Leap Forward (1958–60). This institution was a ritualized system under which local wealthy families took turns accepting responsibility for taking care of outsiders who passed through the area and carried out several of the tasks demanded by the state. In this way, the local community managed to keep some of the interference and burdens imposed by the state at bay. Several deaths during the Cultural Revolution of cadres who held posts of responsibility under the Great Leap are considered to have been caused by vengeful ghosts. These ghosts were assisted by the ancestral spirits of those who died during the famine and became wild ghosts. Mueggler’s analysis demonstrates how memories of past suffering and ritual practice have been linked in many subtle ways to produce a dream of community. Moreover, one of the great contributions of The Age of Wild Ghosts is that it convincingly recounts how the remembering of such experiences has fundamentally altered local conceptions of the state “from personified external Other to abstract internal Other” (2001: 288). Such conceptions are furthermore connected with current concerns, such as increasingly strict birth control policies. In short, the state has lost its ideological justification of being an all-encompassing provider of feelings of community.

Shifting Chinese policies also play a central role in the revival of religious practice in Muli. From the perspective of the villagers, because the Maoist state destroyed the monastery and the sacred texts of anji ritualists, it appropriated the responsibility of keeping order in Premi cosmology, providing for the people’s subsistence and protecting their health. This does not mean that Maoism replaced local beliefs and cosmology. For a limited but extremely confusing and disruptive period in the history of the villagers, several aspects of Maoist ideology invaded these beliefs and coexisted with them in an uneasy tension. A forty-eight-year-old villager of Downhill, in Bustling Township, asked about the consequences of not being able to hold certain rituals during the Cultural Revolution, answered in the following way: “They told us there were no evil ghosts, so performing [this ritual] was not necessary, and, anyway, we thought that if there were evil ghosts, Chairman Mao would protect us.”

As elsewhere in China, the reform policies of the 1980s initially raised living standards in Bustling Township considerably. At the same time, they also moved responsibility for subsistence from the collective to the single household. Now that households again owned their animals and disposed of the production of their privately contracted land, adversities like a bad harvest or disease among livestock were not borne by the community but by the individual household. When members of a household became ill, they could no longer count on free, if basic, medical care provided by the state.

But the reforms also brought ideological relaxation. With the revision of the constitution in 1982, religious freedom became officially reinstated and policies—at least nominally—respecting and protecting ethnic minority culture replaced previous campaigns of intense ideological indoctrination. Faced with the challenges of the post-Mao period, villagers in Bustling Township felt it natural to turn to their own local traditions—or what was left of them in their memories. Gradually, the fear dissipated, and they started to openly practice rituals that had been abandoned during the Cultural Revolution. They also revamped traditional forms of social organization that had been hard to maintain during the period of collectivization, class struggle, and political campaigns. Among the Premi of Bustling Township, the shift from the collective to the household reinvigorated the traditional basic unit of social organization, the dzèⁿ, or “house.” Besides constituting an economic unit, the inhabitants of a house make up a ritual community that is responsible for maintaining and mending the relationship with ancestral souls, the divine forces that control the elements of nature, and the evil spirits that bring diseases. In this task, they receive the help of local ritual specialists, the anji, or of Buddhist monks. But unlike the monks, the local elites, or the foreign researcher, they are not really concerned about the possible differences between these categories of ritualists or, for that matter, differences in the religions themselves. While views on religion among monks and local elites are heavily influenced by discourses of religious and political orthodoxy, such discourses have little repercussions among subsistence farmers trying to make a living in the remote periphery of the modern Chinese state.

RESEARCH

Conducting fieldwork in China, especially fieldwork based on lengthy stays in rural areas, is still a complicated matter. Since the early 1980s, the Chinese authorities have only hesitantly opened the doors to foreign fieldworkers, and there is still widespread skepticism at various levels of the government and the Party about foreign researchers staying for longer periods of time while rummaging about and asking questions.⁷ Permission to stay in a village is often difficult to obtain and/or subject to several limitations. Policies, and especially their implementation, change constantly, and what is allowed one year is out of the question the next. In my case, the ideal of a stay of at least one year in one village turned out to be impossible to realize. Nevertheless, repeated medium-length stays of a few months to a few weeks in the same village, combined with similar fieldwork at other sites and interviews of Premi scholars and other important actors, can constitute an acceptable alternative to a “true” Malinowskian stay in the field of at least one or two years of uninterrupted participant observation.

Revisiting places a second, third, or even fourth time, often under changed circumstances, yielded significantly different data in more than a few cases. Between 1995 and 2007, I spent a total of around thirteen months, divided over nine stays, in Premi areas, starting with a six-month stay in 1995 in the town of Lijiang, where I worked with mainly one informant and conducted short field trips to Premi villages. In 1998, I spent two weeks conducting interviews in Premi villages in Ninglang County in Yunnan. After obtaining an official permit in 1999 to do fieldwork in Muli County in Sichuan, longer stays in villages followed: one two-month stay in 1999, one five-week stay in 2000, and two three-week stays in 2001 and 2004. Finally, in 2006 and 2007, I visited other Premi areas in Yunnan and Sichuan but did not stay in villages more than a few days.⁸ During this period, I systematically collected most of the data that resulted in this book. Before that, ever since my first encounter with a Premi ritualist in 1987, I had traveled extensively throughout northwestern Yunnan, making short visits to many different Premi villages. Since Premi communities in this region were small and dispersed over a large mountainous area, this was a time-consuming activity that often involved long trekking tours but yielded rather limited and disparate ethnographic data.⁹

The basic method of gathering information in the field consisted of formal interviews, conducted in the presence of at least one assistant and recorded in notes or on tape (mainly for genealogies or ritual recitations and prayers). This resulted in 105 formal interviews. In the villages, the interviews were combined with informal conversation—during daily meals with government employees or with the host family—and observation (and, sporadically, participant observation, as during the harvest). In other places, such as the local towns of Muli in Sichuan and Ninglang and Lijiang in Yunnan, data were collected through longtime personal contacts with important Premi actors such as government cadres or scholars. Sometimes such contacts involved visiting their native villages and staying a few days with their families.

The majority of the interviews were conducted in Chinese, or in the local Yunnan or Sichuan version of it, and I asked the questions, sometimes rephrased in local dialect or Premi by assistants. Having spent half a year in the Lijiang area combining in-depth interviews with a prominent Premi scholar with the study of the Premi language in the mid-1990s, I was able to ask questions in Premi and understand some part of the answers. In a few cases, when no local assistants were present and we visited a house where only women were at home, Premi was the only possible means of communication. Complicating my task was the considerable local variation between Premi dialects, even between those spoken in villages quite close to each other. When I first arrived in Uphill and Downhill in 1999, I was disheartened to discover that I could understand only a little of the local dialect, although people could understand most of what I said. After visiting other villages and, finally, when conducting fieldwork in Walnut Grove, which is closer to Yunnan, I found out that it was easier to understand people there. This made it at least possible to get the gist of the conversation taking place around me and helped me significantly in understanding the finer intricacies of rituals and other aspects of Premi culture. Outside the villages, among cadres and intellectual Premi, my ability to engage in simple conversations created much goodwill. It also made it possible for me to challenge their views on the likeness between Tibetan and Premi languages, a frequent topic of discussion during my interviews with “urban” Premi. In the area of Walnut Grove, most people spoke three to four languages; in addition to Premi and Chinese, most also spoke Na and Naxi (Nahiⁿ in local pronunciation). Many of the women in Walnut Grove came from non-Premi villages but after marriage spoke only Premi at home.

Interviews with county government cadres took place either at their offices or in their homes and were often a combination of formal interview and requests for information on local conditions. The fact that the research was officially approved by the prefectural government made these cadres more inclined to hand out useful written material such as unpublished working reports and local statistics. Sometimes fortuitous circumstances made it possible to collect more in-depth information, such as the time when I traveled for two days with the cadres of the county Religious Affairs Bureau in order to visit the main monastery of Muli where we were to participate in a religious celebration. This proved to be a golden opportunity to observe how religious policy was implemented at the local level; at the same time, I would have the chance to be briefed on those aspects I wanted to learn more about or on issues I did not fully understand as they arose during the celebration.

One of the major informants for the study of Premi religion was Nima Anji, from the village of Walnut Grove. He was not only the most knowledgeable of all the anji in Bustling Township, with a steadily growing reputation extending well beyond the township’s borders, but also a good teacher. This quality turned out to be very fortunate for me, and through numerous visits and hour-long talks, he patiently explained the intricacies of Premi cosmology, rituals, and religious practice.

Over the years, through frequent return visits, I developed closer relations with some of the Premi cadres and intellectuals, and they became important informants for my project, which I also often discussed with them. One was a retired cadre from Ninglang County in Yunnan, Waizha Dorje Tsering. He was a key figure in the establishment of a hangui school in his hometown, a project I followed closely and discuss in chapter 5. It is a fascinating example of how Premi identity transcends its local meaning and becomes a larger, modern form of identity shaped and reproduced on the periphery of modern Chinese society, highly affected by that society’s prevalent discourses.

I carried out the main part of my fieldwork in two villages in remote Bustling Township in Muli County, Sichuan. The much more accessible village of Chicken Foot in Ninglang County, Yunnan, was a secondary field site. Except for a few occasions, I did not actually live in this village but made daily visits to conduct interviews while living in the nearby county-town of Daxingzhen. Although I did conduct more systematic household interviews here in 1998, my main method of research in Chicken Foot Village consisted of developing longtime relationships with a few people over a ten-year period in order to follow the village’s development over time and, in particular, the progress of the hangui school established there in 2000. Fieldwork in Muli was more traditionally structured and started with a two-month stay, mainly in the twin villages of Uphill and Downhill. The government of Bustling Township was situated in Uphill, and the government-assigned assistant, or peitong, and I lived in the building where local government officials lived and worked.¹⁰ When I arrived, I explained that I wanted to visit all the houses in the twin villages and register basic biographical data, family structure, and economic conditions; I said furthermore that I would like to ask people about religious practices and other customs and participate in religious ceremonies that would take place during my stay. Paying a visit to all families in a village in a systematic way made it possible to collect a large amount of comparable data in a survey-like way without actually carrying out a formal survey. As it turned out, insisting on talking to all the families gave me the opportunity to also talk to those families whom the local government might have preferred I not visit. Those included, for instance, polyandric families and other families typified as having “backward thinking” (you luohou sixiang de), or very poor families, or just families who did not particularly welcome government employees, which, maybe not surprisingly, were often families of the former categories.

In my first fieldwork period, because I lived at the government building and because, as I found out later, I was perceived as being very closely associated with the people working there and therefore with the Chinese state as a whole, I was only partly successful in establishing more informal settings in which to conduct interviews or gather information. Having experienced the hospitality of ethnic minority people in Southwest China on numerous occasions, I was surprised by how few people invited me into their homes spontaneously. In the afternoon, after I had finished writing out my notes, I usually joined some of the small groups of people sitting in the sun outside their houses or shops along the path next to the government building. After some weeks, people got used to my presence, and I became, so to speak, part of the scenery. This was the closest I got to anything resembling participant observation during my first stay in Muli.

On my next stay in 2000, I revisited Uphill and Downhill and continued fieldwork in Walnut Grove, one day’s walk from Uphill and Downhill. This was also where I conducted most of my fieldwork in 2001. In Walnut Grove, my Nuosu assistant and I stayed with two different families, and in this way, we were much nearer to the life of the villagers. I developed close relations with the most knowledgeable and influential anji of Walnut Grove, and he became a major informant. Participation in several aspects of village life, mostly those related to agriculture (harvesting corn or threshing wheat), or larger events like the annual slaughtering of the family pig, provided excellent opportunities for talking to informants and getting a better understanding of social life.

As has been made clear, one of the major liabilities of the way my fieldwork was organized was its close association with agents of the Chinese government and the Communist Party. It should be reiterated at the outset that in China, conducting fieldwork without official permission is illegal and therefore, especially in the case of longer fieldwork in villages, practically impossible to carry out. In my case, since the county of Muli was officially closed to foreigners in 1999, I needed special permission even for a short visit. Conducting fieldwork that is officially sanctioned also entailed some consequences that I had not anticipated. Households in Communist China are very familiar with visits from official investigators, be they inspection teams for birth control or disease prevention units, registrars for the regular national population census, Communist Party propaganda teams promulgating new policies, or just members of the township government whose main function is to keep statistics on household economy or agricultural production. When a family therefore gets a visit from the local chairman of the Party accompanied by some unknown other people, this presents a familiar situation. The fact that a foreigner is part of the setup does not substantially alter the perception of the visit, namely, as an intrusion into matters that one might have good reasons for hiding from agents of the state, such as unregistered births, unlawful marriage customs, undeclared income, illegal land use, officially frowned upon practices, and so on. This realization struck me forcefully during one of my first interviews, when I tried to find out whether the family I was visiting carried out the widespread Premi custom of offering food to the ancestors before meals on a sacred stone at the fireplace. They did not understand my question, probably because I did not speak the local dialect well enough yet and because they were not prepared for such a question after having first been asked more standard questions on family composition, income, and the like. After several failed attempts at reformulation, Vugo, a Nuosu from the neighboring township and the local Party chairman, offered his reformulation: “You know, what he wants to find out is whether you practice all these feudal superstitious things” (Zheixie fengjian mixin huodong). That helped, but at the same time, it made clear to me that in a setup like this there would be severe limitations on the kind of information I would be able to obtain. Although ethnic minorities in China today are officially quite free to practice their “traditional customs” (fengsu xiguan), including religious practice, a certain level of official skepticism has survived the Cultural Revolution, and the belief that such customs are unscientific and hamper economic development is widespread among officials and Party bureaucrats at all levels. Somehow I had to make the best of the liability of being viewed as part of the intruding Chinese state. This was a precarious position, depressingly reminiscent of the colonial context in which late-nineteenth- and early-twentieth-century British anthropologists operated (Asad 1973).

Luckily for me, after some time, my local assistants got bored with hearing the same questions over and over again and often went out or started talking with other members of the household while I conducted my interviews. Gradually, after having learned more about local practices and sharpening my language skills and interview technique, I became satisfied with the amount of information I was able to collect. Nevertheless, it took another round of fieldwork to realize that this official arrangement limited not only the amount of information I could gather but also its reliability. In 1999, I visited the house of an old anji in Hill Village, which is half a day’s walk from Uphill. We were accompanied by a friendly young Nuosu, the government employee responsible for running Public Security (in other words, the township’s only policeman). This was my first visit to a practicing anji, and I was rather excited. My disappointment was therefore sizable when the old man seemed quite senile and was unable to tell me much. His son was plainly unfriendly to me and my Nuosu assistant from Xichang, so much so in fact that, at one point, the young policeman found it necessary to take him outside somewhat forcefully and keep him there. The whole visit deteriorated into a veritable anticlimax, and I would later remember it as one of the darkest moments of my research.

When I visited Uphill the following year, I decided to revisit the old anji in order to try to shake off the bad memories. This time I was accompanied only by a government employee who actually was from Uphill. The difference was striking, to such a degree that at first I did not recognize the hostile son of the previous year in the friendly smiling person who asked us in and prepared food for us while his father told one interesting story after the other. They explained to me that they had viewed my visit the previous year with much suspicion because I was accompanied by a government employee. Obviously, a local Premi who was also working for the township government was not viewed in the same way. He was not associated with the Chinese state, and my association with him cleared my record.

This opened my eyes to the kind of colonial-style power relations the foreign anthropologist enters into with his informants. It also alerted me to the possible means of resistance the villagers used against the unsolicited prying of all-powerful outsiders. In his “Everyday Forms of Peasant Resistance,” James Scott has pointed to the “ordinary weapons of relatively powerless groups” that can be used against anthropologists: “foot dragging, dissimilation, false-compliance, pilfering, feigned ignorance, slander, arson, sabotage” (Scott 1986: 6, referred to in Metcalf 2002: 56). Among the many possible motives for such actions, most of which are definitely opportunistic, Metcalf suggests less self-interested motives, such as “trying to turn the attention of his interrogators away from practices that had caused friction with the authorities in the past. According as he understood the predilections of his hearers, he might bowdlerize traditional rituals, or suppress references to magic or shamanism—the possibilities are endless” (2002: 56).

When we started working in Walnut Grove a week later, the experience was repeated, in that there was no one from the township government with us, and we were accompanied only by the local village head on our interview rounds. As a consequence, interviews were much easier, and both quantity and quality of data greatly improved; indeed, I had to thoroughly readjust some of my previously collected data. The “official” figures I had obtained from the township government, for example, on the number of polyandric and polygynic marriages proved to be completely wrong: they were much too low compared to my empiric data. My data on school attendance also had to be revised once more. According to statistics provided to me by the head of the township school, 405 out of the 408 children between the ages of seven and eleven in the township, or more than 99 percent, were attending school. Through my participation in classes at the school and through interviews conducted earlier, I had already found strong signs that this figure might be too high, but only when township government officials were not present did people admit that they often had their children go to school for just a short time in order to have them registered as attending school and in this way avoid the fines for keeping them at home. These instances proved once more how untrustworthy official statistics in China can be, even those obtained closest to the grass roots, let alone the thoroughly edited and published versions thereof. But my experience also proves that fieldwork in itself does not necessarily provide reliable data that compensate for untrustworthy statistics. Discovering and correcting faulty data unfortunately gives one only the fleeting satisfaction of having finally gotten to the bottom of a matter, because the question remains as to which data and conclusions will withstand the next round of fieldwork.

Although the major fieldwork stays related directly to this book were allowed by the relevant Chinese authorities, I have decided to conceal the identities of most of my informants by using pseudonyms for persons and places. The majority of these people would agree that my research topic was not especially controversial. Nevertheless, Western research on religious practice and on matters relating to Tibet or Tibetans more generally have been subject to official suspicion at different levels of the administration. The March 2008 disturbances in Lhasa and other Tibetan areas, and the consequent suppression and crack-down in all places with a population officially designated as Tibetan, testify to the unpredictable development of Tibetan-Chinese relations and the shifting Chinese policies on religion and minority culture. I have therefore chosen to err on the side of caution.

Furthermore, even for an informed outsider, it is not always obvious how different people might assess one’s research. I realized, to my surprise, that doing research on people officially classified as Tibetans that focuses in part on their non-Buddhist practices can be a sensitive endeavor, and not just because such research calls into question official ethnic categorizations. Even though I tried to argue that the categories of Tibetan and Premi do not have to be exclusive, I did not find much acceptance for this concept of ethnicity among members of the local elite in Muli, for example. Moreover, many intellectual Tibetans from outside Muli expressed their strong disagreement with this view, and some interpreted my studies on Premi culture as an attempt to split the Tibetan people.¹¹

But I had other reasons as well for choosing to make the identification of my sources difficult. Through my fieldwork, I uncovered matters that people sometimes were not entirely comfortable with having me know, either because these matters were thought of as embarrassing or were, strictly speaking, illegal: my continuous probing into kinship relations revealed some extramarital affairs and illegitimate children and, more problematically, that the preferred form of marriage in some villages was polyandry. Many of the interviews in Uphill and Downhill were conducted with the local Public Security officer or chairman of the Communist Party present. In this way, everything my interviewees said was already “on the record.” Nevertheless, even cadres at the local level might not be able to predict the impact on higher levels of authority of a systematic written presentation of data acquired under their auspices.

Naturally, hiding the identities of places and persons has some drawbacks. First of all, it reduces the verifiability of the ethnographic data. In addition, for people who have done or are planning to do research in the same region, precise geographic and biographical information could be a useful reference for their own field studies. Yet in the end, opting for full verifiability in print would force me to weigh many of my statements very carefully and leave out information in order to be sure I was not compromising any of my informants. I might also be forced to modify some of the conclusions. And even then, mistakes might be made. To keep some spirit of authenticity, I have tried to stick to Premi names for people, replacing their real names with the names of ancestors given to me in the numerous genealogies I collected. Names of local villages have been changed to the English-language translations of existing villages, but their locations do not necessarily correspond to the villages in question. Clan names and house names have been altered when they would make identification possible.

Another category of problems related to research ethics stems from the obvious differences in social background between a researcher from a rich foreign country and people who depend on subsistence farming for their livelihoods. The people of Walnut Grove and Uphill and Downhill were, by most standards, poor. Although the basic necessities of food, clothing, and housing were fulfilled, very few families managed to generate enough extra income to improve their living standards. They also did not have many resources to draw on in case of adverse events, such as bad harvests or sickness in the family or among their domestic animals. In many families, my visit was naturally perceived as a rare potential source of extra income. In view of this apparent poverty, and knowing that I benefited greatly from talking to and living among these people, I found it very difficult to find a balance in contributing financially. Practical dealings over paying for food, renting mules, lodging, and so on were the daily source of much uneasiness and minor conflicts. The matter was further complicated because my Nuosu assistants were convinced that Nuosu people, through their tradition of hospitality, would not have considered me in the same way as the villagers did (that is, according to them, mainly as a source of extra income) and would never have accepted any money from me. On the contrary, according to them, Nuosu people would have held large welcome parties in honor of my visit. On several occasions, they told me that they were embarrassed by the way the villagers behaved toward me in terms of money.

A rather striking case in point took place in a village inhabited not by Premi but by people of related ethnic origin. This village was the poorest of the villages in which I had conducted fieldwork up to that point. It was situated in a predominantly Nuosu township, and the head of the township was related to one of my research assistants. When we arrived, the village held a large party on my behalf. A goat was slaughtered and prepared, and the whole village participated in a feast, with large amounts of beer and strong spirits. In view of the extreme poverty of the village, I tried to find ways of contributing toward the expenses of the party, but the government assistant consistently tried to undermine all my attempts. Finally, he admitted that the whole thing had been organized and paid for personally by his relative, the township head. They both had wanted to keep this a secret, but my insistence on paying the villagers left them no choice. It took me the rest of the day to convince the township head to let me pay half of what he had spent, which was almost his entire monthly salary. This is but one of the more extreme examples of how my visit could disturb people’s daily lives and create problems that I had few chances to anticipate and avoid.

Money would be a significant source of conflict and unease throughout my fieldwork. My assistants frequently got into agitated discussions with villagers over payments for goods and services provided to us in the village, such as board and transportation. With each revisit, these frictions seemed to increase, probably because people were now aware that my visit could mean extra income. In one of the houses where we stayed for a longer time, the daughter became ill, and I offered to pay more for our stay in order to help with her hospital costs, and this became the source of a major conflict between my assistants and our hosts.

Although I often disagreed with my assistants on the way they haggled with local people over small amounts of money, their position as intermediaries saved me from many time-consuming and embarrassing discussions. In the end, in their efforts to protect me from paying too much, my assistants took all the flak and earned a certain level of unpopularity among villagers, a fate I apparently was able to avoid. The issue of money was surrounded by a series of ongoing negotiations in which a delicate balance had to be reached between at least three interests: my interests in collecting data for my research project while also repaying the villagers somewhat for the problems my interference created and perhaps even gaining some friendship; those of my assistants, who had been given the task of helping me fulfill my goals without getting themselves or their institution into trouble but at the same time wanted to give me a positive impression of China; and, finally, those of the villagers who saw an opportunity to obtain some material advantage out of one more round of unsolicited prying into their lives.

This brings me to the final point, and that is the matter of the right of representation. Although many, mostly intellectual Premi, actively supported my project and expressed their opinions about it, I often wondered whether most people of Walnut Grove, Uphill, and Downhill would agree with my interpretations of what they said and did. I have no doubt that some of my conclusions would be contested were they able to read the final product. One reason may be that the villagers do not speak with one voice. There were families in Walnut Grove who were presumed to be possessed by evil spirits; these people were therefore excluded from common rituals in the village and were absolutely undesirable as marriage partners. They saw a religious revival of traditional practice with quite different eyes than did the anji whose knowledge not only was a means of making good money, by teaching anji rituals in Yunnan, but also constituted the grounds for his strong moral and social authority in his community and beyond. Those who criticize me for my lack of understanding of certain issues would certainly be justified in doing so in view of the relatively short time I lived in their villages. All in all, my fieldwork time was rather limited, and although it was long enough for me to get a deeper understanding of the hardships and uncertainties of life as a subsistence farmer, the realization that, at the end of the next month, I could travel back home to my secure and cozy little world was humbling.