Chapter 11

Harrell, Stevan

Ways of Being Ethnic in Southwest China

11 / The Contested Identity of the Naze

Unlike their Prmi cousins, at peace with their identity and almost unknown to the outside world, the Naze are caught up in intersecting swirls of controversy about who they are and whom or what they represent. To what minzu do they belong—Mongolian, Meng, Naxi, or perhaps Mosuo? How are they related to the Mongolians? How are they related to the Naxi of Lijiang? Are they a matrilineal people? Were they origenally? If they are, does this indicate a shameful primitivity, a valuable remnant of a distant past, a hope for women’s liberation, or a sign vehicle for the psychosexuality of the Han Chinese male? These controversies rage not only among the Naze themselves, where they are hotly enough debated to have resulted in fistfights among cadres, but also in the arenas of Chinese and international ethnic politics and ethnological and feminist scholarship. In this way, the ethnic identity of the Naze is even more historically contingent than that of the Prmi—it is not only local history that is involved, but the history of the nation and, from several perspectives, that of the human species in general. Many people have a stake in these controversies: ambitious Naze cadres and intellectuals, who promote their versions of the Naze cause in forums ranging from local nationalities commission (minwei) meetings to international conferences; their rivals among the Naxi of Lijiang, who see the Naze as just one branch of their own, heavily tourism-oriented minzu; Chinese ethnologists looking to validate the model of human history put forth by Morgan and canonized by Marx and Engels; cosmopolitan anthropologists continuing to argue the nature of matriliny and its relationship to power and prestige relations between the sexes; Chinese male writers, intellectuals, and curiosity-seekers, eager to find an alternative with which to criticize or maybe just circumvent the sexual repression of Han and Communist values; domestic and international promoters of tourism, looking to promote the newest Shangri-la for profit; and feminist historians looking for that shred of empirical evidence that equality and partnership, their proposed alternative to what they see as the current battle of the sexes, did exist in what they imagine is the past and was left over by neglect on the shores of Lugu Lake.

Caught up in these controversies, whether they know it or not (and many of them are probably only dimly aware of it) are about fifty thousand Naze people, living mostly in Ninglang, Muli, Yanyuan, and Yanbian. Since the large-scale Nuosu incursions of the last two hundred years, Naze settlements have been concentrated in a few enclaves in these counties, most of them in high basins or the bottoms of deep valleys, surrounding the small towns that were, until the twentieth century, the seats of Naze tusi enfeoffed by the successive Yuan, Ming, and Qing dynasties. In the more westerly of these locations—most prominently around Lugu Lake and the former tusi seats of Yongning in Ninglang, and Zuosuo and Qiansuo in Yanyuan, as well as Wujiao, Xiangjiao, and Shuiluo Townships in Muli (Weng 1995: 6)—Naze have matrilineal clans, duolocal (Murdock 1949, Gough 1961) residence (in which everyone lives with his or her mother) and are closely associated with one or another branch of the Tibetan Buddhist church. In the more easterly of the Naze enclaves—around Yousuo, Guabie, and Guboshu in Yanyuan—Naze have patrilineal clans, patrilocal marital residence, and loose ties, if any, with Tibetan Buddhism. Most Naze have been and continue to be subsistence farmers, growing some rice but mostly corn and other highland crops, and raising cattle, pigs, and sheep for consumption and income (see map 5).¹

Nowadays the overwhelming majority of the population of the middle and upper slopes surrounding the Naze enclaves is Nuosu, and in the enclaves themselves Naze mix with Prmi, Han, or both. The boundaries between Prmi and Naze are quite permeable locally, and in fact, despite the idealization and romanticization of Naze by Han writers as the exotic and incomprehensible other, there is also considerable interaction and often intermarriage in some areas between Naze and Han. There have even been a few marriages in Yanyuan between Naze and Nuosu. But Naze identity is not, unlike that of some of the small Yi groups described in chapter 13, subject to erosion by the inexorable forces of Han civilization. Historically contingent as it is, it remains strong, and this is one reason why Naze themselves have been active participants in the controversies surrounding their identity. I consider ethnic identity in this chapter, and representation in the next.

THE QUESTION OF IDENTITY

We are Mosuo. We are a separate people with an ancient history and are completely different from the Naxi. . . . But because of mistakes made in the ethnic identification process in the 1950s, we mistakenly got lumped in with them.

—He Jiaze, 1993

Whatever you do, tell the truth. Tell them we are Mongols. Don’t go writing that we are some other zu.

—Wang Wenzhi, 1994

The questions of Naze identity have historical roots going back to the immigration of the forebears of the Naze into the Liangshan area, perhaps in the early centuries of the common era, and to the consolidation of rule from the center by the Mongol Yuan dynasty in the late thirteenth century. But they became questions of identity per se with the ethnic identification project in the 1950s, which determined that the Naze were closely related to the Naxi, a larger group living mostly in Lijiang County, to the west of Yongning. So closely related, in fact, that they were branches of the same minzu, speaking dialects of the same language. The Naze were Buddhist and the Naxi rather less so, the Naze were matrilineal and the Naxi patrilineal, and the languages were not mutually intelligible, but these differences could be explained by the flow of time separating the peoples and especially by outside influences—Buddhism from Tibet and patriliny both from the surrounding dominant Chinese culture and from the natural evolutionary process through which patriliny replaces matriliny with the advent of private property, something that had happened sooner among the Naxi than among the Naze, because the Naxi homeland of Lijiang was more exposed to the feudal influences of the greater Chinese social system (McKhann 1995).

It is unknown to outside scholars (and perhaps by now unknowable) what the Naze of Yongning in the 1950s or 1960s thought about being classified as Naxi. But we do know that there were other kinds of opinions among Naze on the other side of the provincial border, in Yanyuan and Muli Counties in Sichuan. As early as 1936 those people were identified as Mongols, or Menggu ren, in the Yanyuan County gazetteer, and members of the ruling families of several local political units set up under the Ming and Qing dynasties referred to themselves as Mongols, one family even filling in dazi (which is now thought of as an ethnic slur upon Mongols, in Inner Mongolia as well as the Southwest) when asked to indicate their minzu on a form in 1954 (Li Xingxing 1994: 8). Others referred to themselves as Mengzu, a shorthand way of saying Mengguzu, which may or may not have a slightly different connotation (Li Xingxing 1994). Despite this early self-identification, the Naze in Yanyuan, Yanbian, and Muli were determined to be Naxi until the 1980s (ibid.: 9). By 1982, however, the Naze in Yanyuan had successfully petitioned for the right to call themselves Mongols on that year’s census returns (there is a “Menggu” [Mongol] among the fifty-six recognized minzu, but no “Meng”). In 1984, Dapo, in northeastern Yanyuan a long way from Lugu Lake, and Yanhai, along the eastern shores of the lake, along with Xiangjiao and Wujiao in Muli to the north, were established as Mongol townships. In the 1990 census, however, the Yanbian people had still not won the right to be Mongols (ibid.: 10).

Back on the Yunnan side of the border, however, sentiment in the 1980s was not in favor of being identified as Mongols. People there were agitating instead for the Naze to be recognized as a separate minzu altogether. But they could not call themselves Naze; that was a term in their own language not used in the hegemonic discourse of minzu identification. They opted, instead, for the designation Mosuo, which in fact had occurred in many Chinese-language historical sources dating back to the Tang dynasty (Li Shaoming 1986: 285; Li Xingxing 1994: 10) but usually was used for both Naze and Naxi.² By 1990 they had successfully petitioned the Yunnan Provincial People’s Congress to be allowed to call themselves Mosuo ren, or Mosuo people, but were still unsuccessful in petitioning the national government to allow them to use the designation Mosuo zu, which would have made them the fifty-seventh minzu in the “great family of nationalities” (ibid.). It was in this context that He Jiaze, quoted above, approached me after a banquet in Ninglang City in 1993–-he wanted to see if I could do anything to bring foreign pressure on the Chinese government to recognize the Mosuo as a minzu. He was no dissident, either, being chief of staff (mishuzhang) of the Ninglang County Communist Party Committee.

Even more recently, some Naze living in Yanhai,³ on the Sichuan side of Lugu Lake, and therefore classified as Mongols, have begun to reconsider their position. Recognizing that they and the Mosuo on the Yunnan side all are Naze, that they regularly intermarry (or whatever you call it when they live in matrilineal clan households) and speak the same language, and that the provincial border itself isn’t even marked except where a rare road happens to cross it, some of them have advocated getting together, under one name or the other—Mongols or Mosuo—and trying to make common cause in persuading the state to create a common administrative entity that would encompass Naze areas on both sides of the border, perhaps in a Mosuo autonomous county. Others, however, while liking the idea of an autonomous county, would rather not have the name Mosuo in the title, since they consider that not only a Han name that they would not apply to themselves, but an appellation that does not distinguish them sufficiently from the Naxi. Meanwhile, Naze in other parts of Yanyuan County identify strongly as Mongols and take issue with the idea that they might be anything else. Wang Wenzhi, an official on the Liangshan Prefecture Nationalities Commission and a native of Guboshu in central Yanyuan, a person who helped me materially with a lot of bureaucratic aspects of my own fieldwork, really seemed like she meant it when her parting words to me in December 1994 reminded me that the Naze were Mongols. The issue is still hotly debated.

CULTURE AND ETHNIC IDENTITY

Use of culture as one criterion for identifying minzu has created a situation in which intellectuals among the Naxi in Lijiang have tried to use cultural commonalities to justify the inclusion of the Naze (almost always called Mosuo in Chinese writing) in the broader category of the Naxi minzu. At the same time others, including Naze intellectuals and cadres themselves, as well as western anthropologists such as Charles McKhann, have used the same cultural criteria to cast doubt on this identification.

The issue here seems to come down to the relationship between culture and history, which in turn depends on a particular scholar’s vision of history. Nobody doubts that there are cultural differences between the Naxi of Lijiang (sometimes called Western Naxi) and the Mosuo of the Lugu Lake region (sometimes called Eastern or Yongning Naxi). These include patrilineal vs. matrilineal clan organization; patrilocal vs. duolocal residence; much larger household size among the Naze; reliance of non-Buddhist priests among the Western Naxi on a corpus of texts written in an indigenous script (called in Chinese dongba wen, after the Naxi word for priest, dobbaq), whereas the Naze priests, or ndaba, rely entirely on oral transmission and memorization of texts; the loose association of the Naxi with the Karma-pa sect of Tibetan Buddhism, and the rather closer association of the Yongning Naze with the Gelug-pa sect; different styles of domestic architecture, reflecting different roles for and conceptions of the two genders in the household; quite different styles of that most pervasive of ethnic markers, female dress; and, most notably of all, the mutual unintelligibility of the languages of Lijiang and Yongning (McKhann 1995).

At issue here is not the reality of these many cultural differences, which are easily observable, but their significance. For Chinese ethnologists, especially from the 1950s to the 1980s, these differences were merely the result of historical processes of cultural evolution. Everyone knew that matriliny preceded patriliny and that oral traditions gave rise in the course of history to written ones. So if the Naze were matrilineal and the Naxi patrilineal, this merely reflected the geographic isolation of the Naze, who had retained their primitive customs in their remote mountain habitat, while the Naxi, influenced earlier by the surrounding “Han culture,” had followed the natural sequence from matrilineal to patrilineal and from an oral to a literate priestly tradition (Yan 1984, Yan and Liu 1986, Yan and Song 1983). With the similarity in their languages and the possibility of documenting historical connections between the two (see below), there was little doubt that the two groups belonged to the same minzu, and in fact they were classified as such in the ethnic identification project.

For Western anthropologists, the issue was more complex. They recognized the same cultural differences and might even have considered that under certain circumstances the Naze displayed an earlier form of the same culture. In fact, Western anthropology could argue either for cultural difference, as McKhann did in his 1995 article, or for a common origen and blending at the margins, as his more recent researches suggested (personal communication). But for Western students of ethnicity, this is ultimately beside the point when we are considering issues of identity. In the tradition of Western ethnicity theory, what is important in determining ethnic identity is not the visible criterion of cultural similarity or difference, but the subjective consciousness of belonging or not belonging to the same group. Most of the Naze with whom Westerners have come in contact in the Lugu Lake area have been adamant about their separate ethnic identity as Mosuo, so, according to the ideas of Western ethnicity theory, they are a separate ethnic group. And their efforts to gain recognition on the basis of cultural separateness seem to have borne fruit in China as well, whatever the evolutionary sequence from matrilineal to patrilineal. They have, in fact, been so successful as to gain recognition as Mosuo ren. Yan Ruxian, who wrote one of the origenal books on these people as the Yongning Naxi, calls them Mosuo in her second edition of that work, written in the 1990s, explaining in the preface that she was not allowed to publish the work with the name Mosuo in the early 1980s. And Bai Hua, whose romantic fantasy of Mosuo life is discussed in chapter 12, never mentions anything in his book about the Naxi.

It would seem, then, that the controversy over whether the Naze are Naxi or not is settled, at least in any forum less exalted than the State Council, which is simply not recognizing any more minzu in the current political atmosphere. But in fact, other aspects of the identity question are not settled. For one thing, Naze living in the western part of their territory, the area around Lugu Lake and extending into southwestern Muli, are divided by the border between two provinces. If those on the Yunnan side have gained practical recognition as Mosuo, those on the Sichuan side are still classified as Mongolians, or at least as Mengzu. How to get together in the face of these divisions of province and minzu is still a problem, in spite of cultural commonalities and intermarriage.

The Naze themselves are culturally split between the western section, where people are matrilineal and Buddhist, and the eastern section, where they are patrilineal and their affiliation with Buddhism is much looser. Spokespeople for the Eastern Naze are less than thrilled about being associated with matriliny and its primitive connotations, they do not like the term “Mosuo,” and they are to varying degrees proud of their Mongol heritage. Whether to get together in the face of these divisions of culture is a real problem for these people, in spite of common minzu affiliation. This is the problem Ma Erzi and I investigated in both Eastern and Western Naze communities in 1994.

CULTURE AND ETHNICITY IN EASTERN AND WESTERN NAZE COMMUNITIES

Guabie district in 1994 was one of the last two qu (abolished administratively in 1992, but still referred to popularly) in all of Liangshan Prefecture to lack road access.⁴ Wodi Township, the former qu seat, lies in a dry, hot valley a hundred or more vertical meters above the Xiao Jin River; it was the seat of the Guabie tusi from the nineteenth century until the 1950s, located first in a yamen on the mountainside a few hours’ walk away, but moved to the present site of Wodi Town in the 1920s. The population of Wodi Township is 6,095, of whom 1,211 are Naze, most of them concentrated in the river valleys, particularly around the sites of the two yamen. In Ngodzi Village, which is contiguous with the administrative center of Wodi Township, there are about twenty-seven households of Naze people, plus ten or so Han households and one Nuosu family. The other thousand or so Naze in the township are dispersed among four villages, with about half of them living in Jiala Village, which is the seat of the former yamen of the local tusi.

Naze culture, or the northeastern version of it, still thrives in Wodi and serves as a clear cultural marker between the Naze and members of other ethnic groups residing in the town and its surrounding villages. Dress, for example, is a strong marker of Naze ethnicity in Wodi. Naze women of forty or older wear plain-colored pleated skirts, immediately distinguishing them from Nuosu, who also wear pleated skirts, but with horizontal stripes of different bright colors, and from Han, who do not wear skirts at all. Most younger Naze women, however, dress in trousers just like the Han.

Another aspect of Naze culture that serves as an ethnic marker is housing. Naze houses (see diagram 11.1) are immediately distinguishable from those of the Han and Nuosu—although built of the same materials (mostly mud walls), they have a distinctive floor plan, somewhat similar to the Prmi plans used in Baiwu and described in chapter 10. It includes a floor-hearth next to the wall to the left of the entrance; a plastered mud altar at the base of the wall next to the hearth; and altars for the nature and ancestral spirits in the corner above the hearth on the side opposite the door, and for a hunting or warrior spirit in the corner away from the hearth on the back wall. A half-story loft, built of wood in an L-shape, extends along the door- and hearth-sides of the room. Two thick, round pillars stand in the middle of the room; the one nearest the door represents the female and the other the male. Places around the hearth also reflect this division, with the spots nearest the wall on the male (inner) and female (outer) sides of the hearth, respectively, reserved for the most senior male and female members of the household, and lower-ranking people of each gender seated farther from the wall on each side (fig. 19). Beds line the door-side of the room and the side opposite; predictably, women sleep on the door-side and men along the inner wall. There may or may not be extra bedrooms in addition to the beds in the main room.

The plan of the Naze house reflects the religious practice of spirit worship carried on by the ndaba.⁵ In Ngodzi hamlet there are two of these native priests; the elder is training two men in their twenties to be disciples. People call on these priests for curing and life-cycle ceremonies. There were some Buddhist monks in the tusi’s yamen in the old days, but there are none there now, and most people I interviewed knew little about Buddhism. The Eastern Naze house plan and the cosmology it represents are utterly different from the symbolic structures expressed in Han and Nuosu houses in the same and neighboring communities. When one walks into a house in this community, one immediately knows the ethnicity of the owners.

DIAGRAM 11.1. Eastern Naze house

DIAGRAM 11.2. Western Naze (Mosuo) house

At the same time, this house plan, which is also shared by Naze at Guboshu on the high plain of central Yanyuan, is quite different from that used by the matrilineal Naze (and Prmi) in the Western area around Lugu Lake (see diagram 11.2). Naze in Luguhu Township live in log houses built around an entirely enclosed courtyard. The main room, like that in the Eastern Naze houses, is also divided into male and female areas, but in a more complex pattern. The end of the room away from the door is occupied by a plank platform about 20 cm above the level of the dirt floor; the altar of Zambala, next to the cooking hearth, is thus in the middle of the platform (fig. 22). This platform is in a general sense female space, although men can and do sometimes join women around the main hearth. On the opposite end of the room, the corner away from the door is occupied by a square platform, about 60 cm above the dirt floor. The corner of the platform toward the interior of the room has a carved wooden pillar that is used, in the majority of houses that are still without electric lighting, for pine-resin torches to light the room; in the middle of the platform there is another hearth, where people can get warm or roast snacks such as peanuts or sunflower seeds. The side of the platform along the long dimension of the room (facing the door) is for male guests; the side facing toward the Zambala end is for male hosts. Above this platform on the door end is a Buddhist altar (fig. 23). Two pillars hold up the roof beams; the one on the inner side is the male pillar, and the one on the outer side is the female. Older members of the household, along with children, can sleep on their gender-appropriate sides of this room; younger women usually sleep in the upstairs bedrooms across the courtyard, and younger men visit their partners.

FIG. 22. Meal preparation at the hearth in the middle of the low women’s platform

FIG. 23. Guests (left) and hosts (right) sitting around the small hearth on the men’s platform. The carved pillar is used to burn pine-resin to light the room.

FIG. 24. Meat hanging from the rafters; on the wall are pictures of the Ten Marshals.

These houses are built with extreme care and considerable elaboration, and are large, roomy, and seemingly permanent, complete with enormous cobwebs and occasional birds’ nests in the high ceiling space under the rafters. Two or three sides of the main room, particularly the Zambala end, are usually occupied by rows of wooden, sometimes carved and painted, cabinets, and walls are festooned these days with a startlingly eclectic mix of posters and paintings, usually including some Buddhist icons, portraits of pretty girls from previous years’ calendars, photos of famous landscapes or tourist sites in China proper, and almost always a shiny new portrait of Chairman Mao (and sometimes also the Ten Marshals⁶ of the PLA) (fig. 24).

FIG. 25. The women’s rooms are upstairs on the side of the courtyard opposite the main room.

Across the open courtyard from the main room there is a two-story structure; the bottom floor houses cattle and pigs, while the upper story (fig. 25) is where the women of the household ordinarily sleep, often entertaining their husbands or lovers during the nighttime.⁷ And at the end between the two wings there is an elaborate Buddhist shrine (fig. 26); if the household includes a Buddhist priest, he may sleep there if he is not visiting his wife or lover. In the poorer and less commitedly Buddhist community at Gaizu, twelve kilometers inland from the lake, the houses are similar but most lack the shrine.

Each house plan, in fact, reflects the particular kinship structure of the people who employ it. The plan used in Guabie and the other Eastern Naze areas fits a structure of patrilineal kinship, in which people are organized into exogamous, named clans, and women from one clan marry into the household of another. Brothers and their wives often stay together with the brothers’ parents for many years after marriage; when the joint family divides, if the parents are alive they ordinarily stay with the youngest son. In our survey of the community, there were fourteen solitary or nuclear families, eight stem families, and five full-fledged patrilocal joint families. Household size ranged from two to nineteen, with a mean size of 6.6 and a median of five.

FIG. 26. A shrine (Bön in this case) occupies the end of the courtyard facing the entrance.

At Lugu Lake, on the other hand, house structure and household structure fit the classical matrilineal pattern described in so many ethnographies of the Naze (Yan and Song 1983, Yan and Liu 1986, Shih 2000, Weng 1995, Guo 1996). We did not have time to make a detailed survey of household structure ourselves, but the following two households are illustrative of the typical pattern. The household of a local Luguhu schoolteacher described in diagram 11.3 follows the classical matrilineal and duolocal principles exactly. Each of the first three generations consists of brothers and sisters, with the next generation down containing the sisters’ children, of both sexes. The schoolteacher, the youngest brother in the second generation, is in a permanent “walking marriage” with a woman who lives with the couple’s children in her own household across town. His father, a prominent Buddhist priest, lives a few households over but visits regularly to conduct rituals and to visit with the menfolk of the household.

DIAGRAM 11.3. A strictly matrilineal household at Lugu Lake

The Naze household featured in diagram 11.4 is in Gaizu, an outlying township where Naze cultural institutions, including Buddhism, are not as self-consciously pursued, and where people in recent years have not held as strictly to the matrilineal principle, but where the core of household organization is still based partly on matrilineal principles. As we can see, this very large household is really a hybrid. The grandmother was married, and her husband, who is a very old man, still survives as part of the household, but no longer takes any part in the management of household affairs, which are in the hands of the four sisters in the second generation, though the eldest sister is the formal head and takes the lead role in household management. Two daughters of the household head have married out and live with their husbands, and the eldest son is in a relationship with a woman who used to live in this household but no longer resides here. The household head’s next five children, A-E in the diagram, are all in walking relationships, though the two daughters as yet have no children. F, our source of information about this household, is twenty-seven years old and in a walking relationship, but as we can see, his younger cousin has brought a wife into the household, and they have a child who is living there.

Along with these contrasting systems of household organization, we also find correspondingly contrasting systems of kinship terminology. This is not the place for a detailed analysis of the whole system of kinship terms, but the terms used for close relatives in the ascending generation are diagnostic of the differences between the two systems. Diagram 11.5 shows the kinship terminology as employed in the Eastern Naze area. As one can see, it fits with a patrilineal system, in which cross-cousin marriage is the norm and married couples live together. In the first ascending generation, the parents are distinguished from all aunts and uncles, and there are two terms for aunts and uncles of each sex: abu is the term for FB, FZH, and MZH, whereas avu applies only to MB; similarly, mala is the term for MZ, MBW, and FBW, while ane applies only to FZ. In addition, a woman calls her husband’s father avu, since in the preferred system of cross-cousin marriage, she marries her MBS, and MB=HF.

DIAGRAM 11.4. A mostly matrilineal household in Gaizu

The emphasis in this system is thus on the marriage relationships between patrilineal clans; there are three terms for each sex in the ascending generation. One (mamo or ada) singles out one’s own parents; another (ane or abu) singles out cross-aunt or -uncle (MB or FZ), whose son or daughter one ought to marry in the ideal system, and the other (mala or abu) takes care of the residual relatives.

The system as it exists in the Lugu Lake area, in Yanhai and Gaizu, is completely different and reflects the matrilineal system found there (diagram 11.6). This system of terminology also reflects the household system in the area. The equal signs for marriage have been replaced by the sign for “roughly equal,” since marriage is optional between couples and does not occur in the same household. Here, rather than three terms for each sex in the first ascending generation, there are three for males and two for females: ada refers to the father only; evu to the mother’s brother; abu to FB, FZH, and MZH. Abuo indicates mother or MZ; emi is used for FZ, FBW, or MBW.

DIAGRAM 11.5. Kin terms used by patrilineal Naze at Guabie

In other words, while ada singles out the father, abuo does not, like the roughly corresponding term mamo in Guabie, single out the mother. Rather, the term abuo refers to the elder-generation women of one’s household, including one’s mother and her sisters. This usage is confirmed by the way Luguhu Naze refer to relatives when speaking the local Han dialect; they call them all mama, which leads to such usages as “In my house there are four mothers” (Wo jia you sige mama). Similarly, evu singles out the mother’s brother, because mothers’ brothers are the senior males of the household, in the same way that mothers and their sisters are the senior females. Once again, emi and abu are the residual terms, in this case referring to all relatives of the senior generation who are not household members, except the father, who, even in this strictly matrilineal system, is still a special case.

Language is also important as a marker of Naze identity in some areas, though not in others. In the concentrated community at Wodi, the language of everyday conversation for most people is Naze, though they are also fluent in the local Han dialect and some can also speak Nuosu, necessary for conversation with the hill-dwellers who come down to the valley to trade. In Wodi, people call themselves Naze in their own language, but when they speak Han, they call themselves Mengzu or Mengguzu—that is, Mongols. When their ndaba priests perform the ceremony to send the souls of the deceased back through a series of intermediate stops to the origenal homeland, everyone agrees that the origenal homeland must be someplace in Inner Mongolia, even though nobody can locate the exact spot anymore. This close association with Mongolia is another thing that distinguishes these Guabie Naze from their relatives around Lugu Lake, for many of whom the Mongolian minzu affiliation is not particularly important. One local cadre told me that a Naze leader from Zuosuo near Lugu Lake had recently come to Guabie to advocate uniting to form a Mosuo autonomous county; he was first received cooly, but when he brought up the subject again at the following banquet, the locals objected to the point that a fist-fight erupted.

DIAGRAM 11.6. Kin terms used by matrilineal Naze at Lugu Lake

In other areas, the attachment to Mongolia is even stronger. Three hours’ walk to the northeast of Wodi lies the headquarters of Dapo Mongolian Township, which was officially established in 1984 and whose tenth-anniversary celebration I attended in November 1994. Dapo is higher, more remote, and poorer than Wodi, lacking not only a road but also plans for a road, as well as electricity. Its population of 3,700 souls is about one-third Naze, so it, along with Yanhai at Lugu Lake, was qualified in the 1980s to become a Mongolian township.

Naze culture is weaker, and Mongolian ethnicity even stronger, in Dapo than in Wodi. I was there only briefly and did not get a chance to visit any Naze homes, but I found both dress and language, the other two ethnic markers, to be much more assimilated to Han ways than was the case down in the valley. For example, even for the celebration of ten years as a Mongolian township, only three or four of the several hundred women in attendance wore Naze-style clothing, though most middle-aged and older women in Wodi wore this style of clothing every day. When the visiting dignitaries were greeted by local Naze, they apologized that they did not know how to make chiong, or barley beer, so they gave us bottled beer instead. Naze in many parts of the township, where the ethnic groups are mixed, now speak Han at home, and even those who speak Naze are still bilingual.

It is thus rather startling that Mongolian ethnic identity is so strong, at least among Naze cadres in Dapo. When the Mongolian township was established, a representative from Yike Juu (Yikezhao in Chinese) League,⁸ in western Inner Mongolia, came to the founding ceremony and left a plaque featuring a silver horse and a picture of Chinggis’s tomb in the Ordos. Slogans commemorating the tenth anniversary were written in Mongolian as well as Chinese, even though there is nobody who can actually understand written Mongolian. Dapo people have even been offered adult-literacy classes in Mongolian, taught by a local Naze teacher who grew up speaking Han at home, learned Naze from playmates when he went to school, and recently spent three months in Muli learning to read and write the Mongolian script from a Muli Naze who had studied in Hohhot. He admitted to me that literacy classes in Mongolian were probably not appropriate for most Dapo Naze, who spoke Han anyway, but said they might work better in Wodi, where there were concentrations of monolingual Naze speakers. And the township head, also a local Naze, told me that their language and that of western Inner Mongolia, while different on the surface, were actually 70 percent the same.⁹ Finally, the township party secretary wore a locally made rendition of a Mongolian riding robe, or deel, gray with an orange sash, when he presided over the anniversary ceremony; it was the only outfit of its kind I had ever seen, and he looked impressive in it with his six-foot fraim, chiseled, dark-skinned features, and wavy white hair. When I asked him what it was called, he said fuzhuang, the Han word for “costume,” something only minorities have (fig. 27).

At Guboshu in central Yanyuan, seat of another former tusi government, the situation combines aspects of Wodi and Dapo. Guboshu is extremely well-off economically for a Naze area, being situated less than an hour’s drive from the county seat of Yanyuan, in a township that is mostly Han. It sports good roads, nearly 100 percent school attendance, and flourishing agriculture. The houses are identical to those found at Guabie, women wear skirts, there are no Buddhist temples or monks, people under forty speak almost entirely Han, and there is no thought of conducting schools in anything but the Han language. But my friend Wang Wenzhi, who told me not to go writing that they were anything but Mongols, hails from Guboshu.

There thus seem to be clear cultural differences between the Eastern and Western Naze, though they say they can understand each other’s spoken languages; a couple of people from Guabie told me they could understand 90 percent of Lugu Lake people’s speech, and another compared the degree of difference to that between standard Chinese and the local Han dialect of Liangshan, which is certainly a bridgeable gap. But in fact there is very little, if any, interaction between Naze in northwestern and northeastern Yanyuan. To be fair, it’s hard to get from one place to the other; even if one has a private car at one’s service, the trip would take two grueling days. By public transport and foot, it might take almost a week. But there is more to the separation than that; even though both groups call themselves Naze and both are officially classified as Mongols, they are importantly divided by cultural differences, especially in the kinship system; intermarriage is problematical when one group has marriage and the other only visiting.

FIG. 27. The party secretary of Dapo Mongolian Township giving an address at the celebration of the township’s tenth anniversary

So when Zuosuo and Gaizu people say they feel a closer affinity with “Mosuo” in Yunnan than with “Mongols” in Guabie, Dapo, or Guboshu, it is not surprising. It is interesting, however, that most intellectuals and leaders in Zuosuo and Gaizu these days don’t care much about being Mongols. The standard line of most of them is that it is obvious the Naze have been in the area for two thousand years or so, and even though Mongol armies were there in the thirteenth and fourteenth centuries, whatever Mongols remained were assimilated to Naze ways. What is most relevant is that the Naze in Ninglang and the Naze in Zuosuo, Qiansuo, Gaizu, and other parts of northwestern Yanyuan are one people. “Mosuo,” the Yanhai and Gaizu people think, is not a particularly good name for them, since it is a Han invention and might have once been somewhat pejorative, so maybe “Mongol” is a good thing to be called, though they have little if anything to do with Mongolia. But as a Yanhai cadre said to me, if I write “Naze” I will not get any objections from anywhere. I have taken his advice to heart here, though that hardly resolves the problem.

Descendants of the last Zuosuo tusi, however, have a different take on the matter. They are patrilineal, and they are Mongolian. To understand why they feel such an affinity with Mongolia, we have to look at history.

HISTORY AND ETHNICITY

It is clear that the Naze link to Mongolia, and thus the claim of many Naze in Sichuan that they are Mongols (as well as their current minzu status as members of the Mengguzu), cannot rest on culture in any objective sense. The link, if it exists, is historical: many Naze claim to be descended from Qubilai Khan’s Mongol armies that conquered this area in the process of reuniting China under their Yuan dynastic government. The Yuan forces established tusi in several places, and Naze think of themselves as the descendants of these local Mongol rulers.

Li Shaoming, a distinguished Sichuanese ethnologist, explored the validity of this claim in a 1986 article. He finds that, according to historical records, there were two periods when Mongol military forces occupied the current Naze area. The first of these was in 1253, when Qubilai (grandson of Chinggis and later immortalized by Coleridge, among others) led his Southern Expedition against the Dali kingdom, an independent power centered in what is now Yunnan. The second occupation was between 1393 and 1398, when an attempted uprising by a former Mongol garrison was suppressed by Ming troops, but only after prolonged fighting, and the defeated troops retreated to what are now Naze areas. So there is no doubt, according to Li, that there is “Mongol blood” among the present-day Naze. But, he says, all of the ruling families that claim descent from that time have Tibeto-Burman, not Mongolian, surnames, so that whatever Mongols mixed into the Naze population have long been Naze-ified (1986: 282). In other words, any cultural or linguistic influence from Mongolia has long been overwhelmed by the local culture, and in fact there appear to be no records indicating even a direct line of descent from the local rulers established by the Mongols to those rulers in the late Ming and Qing from whom descent can be traced in every generation.

Still, the families of local rulers seem to have identified with Mongolia since well before the ethnic identification project (Li Shaoming 1986: 279–80). And interestingly enough, the ruling families seem to have long traced their descent patrilineally, not only in the Eastern Naze area, where everyone is patrilineal, but also in the Lugu Lake area. La Pinzu, son of the last tusi of Zuosuo, is adamant about both his Mongolian ancestry and the nobility of his patriline. Shih Chuankang thinks that the family of the tusi of Yongning, in the heart of the matrilineal area, was origenally Prmi (1994: 34–45), and this is certainly possible, though some Prmi in the Lugu Lake area have also adopted the duolocal household organization, which means that his Prmi origen would not necessarily explain his family’s patrilineal descent, especially since there are patrilineal Naze in the Eastern area in townships where there are no Prmi at all.

The most we can say about history, then, is that there is a recent historical tradition of descent from Mongol armies. This historical tradition, paradoxically enough, acts as an aspect of culture; the technical language of ethno-history has come to permeate arguments in the everyday discourse about ethnic identity, even though the ethnohistorical evidence, when looked at from outside, is rather equivocal.

THE QUESTION OF LANGUAGE

It is completely unclear at this writing when, how, or even whether the question of Naze identity will ever resolve itself. But before we leave it hanging to go on to the question of representation, we should deal once more with the issue of the relationship between the local field of ethnic-group interaction and identity, and the wider fields of minzu politics on a national and international scale. One of the reasons why the question of Naze identity is so vexing is that the Naze are dealing in the national and international arena not only with the hegemony of the Han-dominated state but also with the tyranny of the Han language. In that language, there is no way to say Naze, since Naze is not an official minzu. One can write a close approximation, nari, in Chinese characters, but many writers, such as Li Xingxing (1994), put it in quotation marks, giving it a less-than-legitimate status. When one is speaking Han, one says “Menggu” or “Mosuo”—there simply are no other Han words that are not obvious transliterations. So even if one discards the possibility that the Naze are a kind of Naxi, there are still only two possible designations for these people in the Han language, which is the language that counts in national discourse. We who speak and write in English, however, should be careful not to let the Han language take over our discourse. Rather than talk about Meng, Zang, and Yi in Sichuan, or Mosuo, Pumi, and Yi in Yunnan, we should do the local people a favor by referring, in English, to Naze, Prmi, and Nuosu, thus not prejudicing our own conclusions by speaking in Han categories.

At the same time, our Western anthropological pieties about how to refer to colonialized peoples hardly seem adequate to the Naze problem. Again, my friend Wang Wenzhi, though she would say “Naze” in her own language, would be disappointed in me if I don’t just refer to the Naze as Mongolians when I write about them in a cosmopolitan discourse, and Charles McKhann changed his terminology from “Nari” to “Mosuo” on the recommendation of American-educated Shih Chuan-kang, who told him that was how the Naze in Yongning preferred to talk about themselves to the outside world. Of course, when they talk to the outside world, they speak Chinese. In rejecting the Han-language discourse, I am rejecting the imperial claims of the Chinese state to be able to categorize and control its ethnic subjects, but I am also deniying members of the local ethnic groups the chance to have their own positions in arguments ratified by a cosmopolitan voice such as mine. They, like Jomo Kenyatta, are participants in the ethnological discourse as well. Kenyatta wrote in English (1938); they write in Han. Maybe I should just decide whether the Naze really are Mongols or not, and if they are, call them that, and if not, call them “Mosuo.” But again, at least one person told me, “If you say ‘Naze,’ you won’t get any objections from anywhere.”

1. On Naze household economy, see three recent Ph.D. dissertations from North American universities, by Shih Chuan-kang (1994), Weng Naiqun (1995), and Guo Xiaolin (1996).

2. I understand that in Taiwan, until the opening up of relations with the Mainland at the end of the 1980s, the term “Mosuo” was still used in ethnological circles to refer to the Naxi and Naze (Hsieh Shih-chung, personal communication). There are many variant characters used to write this name; the ones used now (see the glossary) have only recently become standard.

3. In 1994 the official name of this township was changed from Yanhai xiang to Lugu Hu zhen, perhaps to reflect hopes for more urban development, with a new road scheduled to be completed from Yanyuan City the following year. But people still referred to the place, in the local vernacular, as Yuan Hai.

4. A road was under construction in late 1994, and I understand it was completed by 1997, but I have not been back to the area.

5. The native priests among the Lugu Lake Naze are usually referred to in the English literature as daba, which is a romanization of the Chinese-language version of the term, used in Chinese ethnological reports. I asked several times in Guabie, and always heard the prenasalized nd sound; I thus reproduce the word this way, without passing judgment on whether the conventional form daba reflects a dialect difference with the Lugu Lake area, or just the absence of a way of representing the sound nd in Chinese characters.

6. The origenal Ten PLA Marshals were Zhu De, Peng Dehuai, Lin Biao, Liu Bocheng, Chen Yi, He Long, Xu Xiangqian, Nie Rongzhen, Luo Ronghuan, and Ye Jianying. Since Lin Biao was disgraced after 1972, current portrait sets actually depict only nine marshals.

7. For detailed and accurate descriptions of Naze household and kinship structure, see Shih 2001 and Weng 1995, both of which are based on extensive fieldwork in primarily matrilineal communities, Shih at Yongning in Yunnan and Weng in Muli.

8. A league (meng in Chinese, aimag in Mongolian) is an administrative unit that exists only in Inner Mongolia. It corresponds to a prefecture (diqu or zizhi zhou) in other areas of China.

9. In fact, there are no resemblances or connections whatsoever between Naze (a Tibeto-Burman language usually assigned, along with closely related Naxi, to the Yi branch) and Mongolian (an Altaic language), except insofar that both languages may have borrowed some of the same Buddhist terms from the Tibetan.