Abstract
The Hmong of southern China define themselves as an "unmarked category" in relation to two significant others: the Chinese and the Yi. This paper, based on fieldwork in Sichuan, examines some problems of Hmong ethnography, inquiring why color terms were used for some groups of Hmong, as well as subdivisions of them. An attempt is made to consider the Hmong not only in terms of their relationships with the Han, but also with the Yi. However, the weight of historical evidence is against this; the Hmong were a very isolated group. Many Hmong subdivisions did arise through intermarriage with Chinese, but if culture is often transmitted maternally, how were Confucian values disseminated to minority populations through intermarriages with Chinese males? The official classification of Hmong (together with other groups) as "Miao" posits a fierce opposition between "Han" and "Miao," yet the Hmong have stories of how Hmong and Han were originally two brothers worshipping at the same paternal grave, whose descendants lost touch with each other. This sort of "genealogical" model (variety out of unity) shows how idioms of patrilineality may overlap with diffusionist culturalist notions of assimilation to a greater Chinese identity. The genealogical model contrasts with a more existentialist and constructivist model that emphasizes the way identity may emerge from, or be imposed on, cultural differences. In fact both are mutually constitutive.
Keywords: Miao-Hmong-Sichuan-Confucianism-kinship-slavery-identityassimilation-nationlism
DAVID GRAHAM TELLS US that the Ch'uan Miao are an ethnic group living on the borders of Szechwan, Kweichow, and Yunnan Provinces, western China. The country is very mountainous, with numerous peaks rising 3,000 to 6,000 feet above sea level. There are many streams, forests, waterfalls, perpendicular or overhanging cliffs, natural caves and natural bridges, and deep holes or pits where the water disappears into the bowels of the earth. While the roads between the Chinese towns and villages are generally paved with stones, most of the roads are narrow footpaths up and down the steep mountainside or through fields and forests. (GRAHAM 1954, 1)
In terms of achieving a better understanding of the complex processes of identity formation that have taken place in southwest China, and the kinds of relations that have been forged with those defined as "others," it may be of interest to inquire into why some people of southwest China, who call themselves Hmong, were also known as the "Han Miao." In the above quotation, Graham is using an abbreviated geographical term to describe these people, who are clearly Hmong, as is evident from the translations he gives and the customs he refers to. The Hmong are a particularly clearly identifiable group who do refer to themselves customarily as Hmong, and who speak dialects of the Western or Chuanqiandian branch of the Miao language in the Miao-Yao language family; one of the three main branches of this Miao language.
HMONG GLE, BO, OR SUA?
GRAHAM (1937, 18) explicitly says that the Ch'uan Miao "call themselves Hmong," but adds that "The Ch'uan Miao are also called Hmong Swa or Chinese Miao because they have been much influenced by Chinese language and customs and the men dress like Chinese" (GRAHAM 1937, 20).
Elsewhere (GRAHAM 1954, 31) he remarks, "Swa is the Ch'uan Miao term for Chinese. It means `play or be idle' and is applied to the Chinese because they are landowners and live on the income from rentals paid by the Miao people, who do the hard work on the soil." This may be a nicely inventive informant's etymology, faithfully rendered by Graham, but there is no doubt that the Hmong term "Swa," or "Sua" as it is now written in the generally approved romanization for Hmong, does refer in particular to the Chinese. It has connotations of strangeness, of "otherness" against which Hmong identity is defined. In history and legend it is always against the Sua that the ancestors of the Hmong fought for suzerainty over the sovereign territory of China.
Yet in poetry and ritual verse it is almost always conjoined with the word for another people-"hang," meaning the Yi or Lolo of southwest China, who were sometimes overlords of the Miao. Thus, the "Sua-Mang." I do not know the derivation of this "Mang,"' but a not improbable etymology for the term "Sua" might relate it to the more common term for Han Chinese, "Hua" (t), since there are other examples of elision between the "f" of "Sua 112 and the aspirate of "Hua," and the tonology fits. This etymology would raise questions about when this term first began to be used by the Hmong for the "Chinese," and the significance of its doubling with the term for the Yi. I turn to this in the following section.
It is interesting that Graham does not use the term "Han Miao," which was in use, but instead the Hmong term "Hmong Swa," which he must have heard from Hmong themselves, perhaps not from the same group as was being referred to. It might have been a derogatory term, applied to groups of "sinicized" Miao-as we know many of the Ch'uan Miao wereby Hmong who considered themselves more properly Hmong than they. It could also have been an apologetic self-deprecation, even a proud boast, but the former is more likely.
For there is at least one other group of Hmong in China identified as Hmong Sua, or "Chinese" Hmong, in Wenshan (_Wl) district of Yunnan, close to the Vietnamese border, who live there in small numbers among the White Hmong and wear peculiarly dark clothing. The Hmong have their own terms for various subcultural divisions among themselves besides the White (Daw) and Green (Ntsua) Hmong of Southeast Asia, such as the Hmong Si, Hmong Pe, Hmong Pua, and Hmong Xau described as living with the Hmong Sua ("Chinese Hmong"), Hmong Daw ("White Hmong") and Hmong Ntsua ("Green Hmong") in Wenshan district of Yunnan (ZHANG, YANG, and SHEN 1988, 29-34).3 In Laos and Vietnam there are also groups of Hmong Si, Xau, and Pua. The terms themselves have no very clear meaning, except that they are clearly Hmong terms and do refer to real differences of costume, dialect and custom.'
GRAHAM (1954, 2) also says that the Ch'uan Miao are called "Hmong Gleh or White Miao because in some localities they wear or have worn white clothing." And LING and RUEY (1947, 22-23) clearly state there were "Pe Miao" (White Miao) in Gaoxian, Gongxian, Xingwen, Changning, and Yunlian counties of southern Sichuan, and that because the "White Miao" predominated over the "Flowery Miao" (Hua Miao) there, this entire area was known as the "Pe Miao" areas DE BEALIAR (1970, 20) concurs.
So why should the term "white" have been used for these "Chinese Hmong" or "Hmong Swa"? GRAHAM (1954, 3) mentions that formerly "most of their clothing was made of hemp cloth, which was undyed and therefore white, woven on their looms by the women" as the explanation for why these Ch'uan Miao were called the "White Miao," while at the time (that is, in the 1930s), most of their costume was blue cotton bought from Chinese. But, most confusingly, he then goes on to say that "in former years" their festive clothing was "beautifully embroidered and coloured," with the women's skirts' "pleated and dyed in a manner resembling somewhat the batiks of Java," and regrets the loss of this "fine and beautiful art."
While there is no confusion in Graham about the "Ch'uan Miao" being clearly and unambiguously Hmong, there is some confusion in his use of the term "White," because (1954, 19) he distinguishes one particular group of "Ch'uan Miao" near Chen Hsiung as the "Hmong Gls" or White Miao, from other groups of Ch'uan Miao like the Xiao Hua Miao of Yunnan, who he says are a small Ch'uan Miao group being absorbed by the Hua Miao, and also from the Ya Ch'iao (Magpie) () Miao of north Guizhou.
With the group Graham refers to as Magpie Miao, we are on strong grounds, since this group was intensively studied by LING and RUEY (1947), who recorded and transliterated their kinship terms, which were commented on by KROEBER (1958). These "Magpie Miao" who according to RUEY (1958) were called by others "Han Miao" or "sinicized Miao" and were one of the affiliated groups of the people called by Graham "collectively, but not accurately," the Ch'uan Miao (RUEY 1960,144; see also RUEY 1962 and 1967). It is very clear from these (linguistic and cultural) materials of Ruey that the "Magpie Hmong," who called themselves Hmong Ntsi, were Hmong, and that they were part of the group called by the Chinese the "White Miao" and by Graham the "Ch'uan Miao"(see RUEY and KUALA 1962). Yet Graham sets them apart from the "Hmong Gleh" of Chen Hsiung, and we know that they used batik (RUEY 1960, 145). It is equally clear that neither the "Magpie Hmong" nor Graham's "Hmong Gleh" were of the same group who refer to themselves in Laos, Thailand, Vietnam, and along the Yunnan-Vietnamese border as "White Hmong" (Hmong Daw), and with whom my first fieldwork was conducted.' These Hmong Daw (White Hmong) distinguish themselves strongly from the other main cultural division of the Hmong represented in Southeast Asia-the Green Hmong (Hmong Ntsua). Hmong Daw settle in separate villages from them, speak a consistently different dialect, with consistently different architectural structure in their households, and most importantly here, have no use at all for the batik excelled in by the Green Hmong.'
Besides other consistent dialect changes, the consonant "gl" or "dl" does not exist in the White Hmong dialect, but only in Green Hmong, which consistently converts it from the White Hmong "d."' The people who described themselves to Graham as "Hmong Gleh" or who talked of other Hmong in the region in this way (and those I later worked with in Gongxian),9 could not, therefore, have been speaking this distinctive White Hmong dialect of Southeast Asia which identifies the "White Hmong" there and in Wenshan (TAPP 1990 and 2001). The "Magpie Miao" dialect recorded by RUEY and KUAN in 1962 (see also RUEY 1958 and RUEY 1960) is much closer to the Green Hmong of Southeast Asia, while I have met Hmong from Yunnan, Guizhou and Sichuan who all claimed to have been "White Miao" and yet used what I recognized as the distinctively "Green Hmong" consonant "91" or "dl" to describe this.
So Graham's contradictory remarks on clothing here are important. They seem to suggest that while the people he was dealing with and whom he calls the Chuan Miao were certainly Hmong, either the "cultural divisions" so apparent and insisted on in Southeast Asia (like those between the White and Green Hmong) were simply different and perhaps more complex in China, or that archaic cultural distinctions of the type preserved in Southeast Asia between the "White" and "Green" Hmong were in the process, even at this early date in China, of amalgamating and disappearing. Groups of originally White Hmong were perhaps adopting batik techniques through intermarriage or simply buying cloth from the market, as techniques for making batik gradually became less important as markers of an identity.
Probably a long process of levelling and cultural intermixture has taken place through the present century, resulting in some erosion of traditional cultural distinctions within minority groups. Graham suggested that his Ch'uan Miao might be a mixture of "several smaller Miao groups," and LING and RUEY (1947, 18) described the difficulty of distinguishing between the dress of originally distinct groups such as the Hua (Flowery) and Qing (Green) Miao who lived together with the Pe (0, White) Miao in Mengzu in Yunnan: "several groups inhabiting the same area are likely to intermarry, so that after a considerable length of time has passed, it is difficult to identify them." But why were color terms used at all, in particular "White," for these groups? And how can we explain Graham's use of the term "Hmong Swa" to describe the Ch'uan Miao as a whole as a group that was adopting Chinese manners and customs? It is often said in Southeast Asia that the "White Hmong" were closer to the Chinese, more pacific in the past history of struggles and resistance against the Han, and that the "Green Hmong" are the older and more original group. Linguistic evidence would seem to support this; the lack of final nasals for instance in White Hmong by contrast with Green Hmong does appear to represent a historical loss (DOWNER 1963). Probably, then, "White" refers to a degree of assimilation to Chinese, and it would make sense then that it is frequently conjoined with terms like "Han Miao" or "Hmong Sua."
In the Chinese records the most general terms used for differentiations of the "Miao" (as opposed to the Hmong) are also color terms; in particular the Black (He), Red (Hong), White (Bai), Green (Qing) and Flowery (Miao) Miao, although hundreds of more specific terms such as Dog-Ear and Long-Horn Miao are used in the illustrated Miao albums of aboriginal people that were produced from the eighteenth century onwards for parts of Guizhou and Yunnan. Generally the He Miao (Black Miao) are taken to refer to the people who call themselves Hmu in Southeast Guizhou, who were prominent in eighteenth- and nineteenth-century rebellions, who tended not to adopt Christianity and speak the Qiandong (Eastern) branch of the Miao languages. While there are also Hmu in Hunan and Guangxi, the majority are centered around Kaili in southeast Guizhou. The "Red Miao" are generally taken to be the Miao of west Hunan, who call themselves Qho Xiong and speak what is classified as the Xiangxi (Eastern) branch of Miao (LEMOINE 1972, 195-200). The "Green" and "White" Miao of the Chinese records are generally assumed to correspond to the Hmongspeaking groups that designate themselves as "Green" or "White" Hmong in Southeast Asia and also in China, although there are problems in isolating the "White Hmong" dialect of Southeast Asia in most of China. The group referred to as Hua or Flowery Miao is generally taken to be the people who call themselves Hmo (sometimes written Hmao), or A Hmo, and live in similar locations to the Hmong in Yunnan and parts of Sichuan, Guizhou, and Vietnam. LING and RUEY (1947, 22) discuss these five main color classifications as if they were quite uncontentious, suggesting that originally they referred to different migrations and migration paths of the Miao towards the south of China (which may be true).
These terms are quite clearly Chinese classifications of the Miao, however, and fit well with other color classifications of minority groups in the region, and perhaps also with the Mongol-Manchu system of military orders, and the Red, White, Yellow and Green Banners of the (Ho or Haw Chinese) rebel armies who ravaged Guizhou from the mid-nineteenth century (McALEAvy 1968; JENKS 1994). Official Chinese interpretations claim these were originally Chinese terms that were then adopted by minority groups, but it does seem odd then that at least "White" and "Green" should be so firmly accepted by the Hmong of Southeast Asia and the China borderlands region as markers of distinctive types of identity.
Since we have these Chinese classifications of Red, Black, White, Green and Flowery Miao, as well as indigenous Hmong classifications of the White, Green, and, less clearly, the Black Hmong, let us consider some other color divisions found in the region. While color classifications seem almost never to have been applied to the Yao, closely related in culture and language to the Miao, the best examples of color classifications denoting cultural divisions among a single "people" are among the Lisu and Lahu peoples, and among the ("tribal") Tai. HUTHEESING (1990, 34) cites the Pai (White) Lisu, Hei (Black) Lisu, and Hua (Flowery, Flowering) Lisu as "official" terms for cultural divisions among the Lisu peoples, and gives a synonym for the "Black" Lisu as the "Old Lisu." The Lahu people, who share some surnames with the Lisu (who share some with the Hmong and Chinese), are divided by ethnographers (using mainly Burmese terms) into the Lahu Na (Black Lahu), Lahu Shi (Yellow Lahu), Lahu Nyi (Red Lahu, known to Thais as Red Mussur), Lahu Hpu (White Lahu) and Lahu Sheh Leh-who prefer to be known as Lahu Na (and are called Black Mussur by the Thai) although they have little in common with the great Lahu Na of Yunnan (WALKER 1974). Beyond the borders of China, and probably not therefore the result of Chinese classifications, we find the Southern Karen, the Sgaw Karen (known as White Karen to the Thai), and among the Northern Karen, the Kayah (known to the Burmese as Karenni or Red Karen). The color distinctions used for these groups do relate to clothing. Then there are the tribal groups of non-Buddhist, patrilineal Tai in the mountains of southern China, northern Vietnam, and Laos who call themselves the Black Tai and the White Tai, and those called by others Red Tai.
Bai-Yue or the Hundred Yue was the original term for one of the dominant cultural streams in southern China, and distinctions between more civilized (,, shu) Southern Barbarians as "Bai-Man" (white barbarians), and more savage and rebellious (I, sheng) barbarians as "black" or "Wu-Man" are of long-standing in Chinese historical records (BLAcKMoRE 1967). But these simple white/black, or raw/cooked contrasts, made by Chinese, fail to account for the complexity of cultural divisions, or for strongly felt distinctions between cultural divisions like the Green and White Hmong that result in their living in entirely different villages, or for the recurrent use of terms like "black," "green" and "red" in different contexts."
Besides christening all the Hmong of the region (including those over the border in Yunnan and Guizhou) as Ch'uan Miao, saying they are also called Hmong Sua or Chinese Hmong, and talking of them (sometimes in general, sometimes in reference to particular groups among them) as "White Hmong," GRAHAM (1954) also says that they referred to themselves as "Hmong Bo" or "Old Hmong." Where did this term come from? It is possible that the term "Hmong Bo" had something to do with the vanished "Bo" people who did once inhabit these parts of the Guizhou-YunnanSichuan borderland where the Hmong now are. The local Hmong tell stories of how their own ancestors, on behalf of the Ming army, vanquished the magical flying headless Bo Kings (TAPP 1996). The character for this "Bo" (locally pronounced "buo") has changed historically; while pre-Han references used a different character, the character "white" (M) was used from the Yuan to the Tang dynasties for the Bo people in Yunnan who practiced "hanging their coffins" from high mountains. The archaic pronunciation for this latter character, and indeed for the character for "hundred" cited in the Bai-Yue above, pronounced "bai" or "bei" today, was indeed "bo." It might therefore have been understood to have had something in common with other uses of terms for "white" in the southwest China borderlands to refer to particular kinds of identity. For example, there are the Bai (0) people of Yunnan, formerly known as Minjia, who are generally recognized to have been among the most "sinicized" minority groups of the area, and are sometimes taken to be descendants of these "Bo." Here "white" (written "Pai" in older European texts, but mostly pronounced "bei" rather than the standard Chinese "bai" throughout Sichuan, Yunnan, and Guizhou) seems to have connotations of an unmarked, or default identity, and perhaps of purity, of spotlessness, too. If the "Bo" or "Bo-ren" whom the Hmong are said to have fought and defeated (and perhaps assimilated) were thought of as using the term for "white," then the Hmong of the area might have referred to themselves as "Hmong Bo" to signify their assimilation of the Bo, and in this way have become known later as "White Miao" or "Pei Miao." This would argue against Graham's translation of "Hmong Bo" as "Old Hmong."" Moreover, ScoTT and HARDIMAN (1900-01, 597) referred to the "Miotzu" (who were clearly "Hmong") in Kokang as being called "Hke HpoV or "White Chinamen."' They add that "in some parts of Yunnan the name Hpo seems to be applied to them.""
However, a third explanation of Graham's term "Hmong Bo" is also possible. A group I have come across consistently in parts of China as well as Vietnam and Laos is the Hmong Pua; the "p" here is an unaspirated "p," like in the English "span," while the "ua" or "oa" frequently slides into an "o"; hence this "Pua" can often be heard as "boa" or "bo." This therefore might be the term that Graham transliterates as "Hmong Bo.""
The "Black Hmong" are a particularly elusive group (disregarding the misleading official classification used in Vietnam), as one of the Hmong patrilineal surnames, the Thao, are referred to by this name in a strange system of double classification of surnames. In the second part of this paper I consider the Hmong surname system in relation to these regional color classifications. The term "Black Hmong" ("Hmong Dlu" in Green Hmong) is sometimes almost jestingly used to refer to unknown groups of Hmong far away. It may be, however, that the Black Hmong, as opposed to the Black Miao of the Chinese records, who as we have seen are not Hmong at all, are a small group assimilated to the White Hmong along the Burmese and Vietnamese borders and distinguished by other groups of Hmong according to the darkness of their clothing. If the "Hmong Dlu" are a group categorized in this way by other groups of Hmong rather than by themselves, then the term "Hmong Sua" would also make sense as a term used somewhat derogatorily by some Hmong for other Hmong, perhaps because of their strangeness rather than their sinicization. It is even possible that Graham was using the Chinese term in general use at the time, "Han Miao," in discussions with his Hmong informants, which was then translated back to him (in Hmong) as "Hmong Sua," without this term ever having been in use by the Hmong of Sichuan to refer to themselves or other Hmong groups.
CULTURAL DIVISIONS AND SURNAMES
It was Robert COOPER (1984, 29) who first pointed out the oddity of these "cultural divisions" marked by color among the Hmong "that are not based on the kinship system and perform no apparent economic function." The explanation Cooper was given for this (by White Hmong) was that Green Hmong women adopted batik from the Chinese, while White Hmong men and women adopted Chinese trousers." COOPER (1984, 29) notes that this explanation may show that "the Hmong have been quite ready to borrow from the Chinese culture yet have done so in a way that maintains cultural distinctions between White and Green," considering it unlikely that the cultural divisions could have originated from different areas within China. It is however most significant that these cultural divisions should be linked with cultural borrowings from the Chinese in this way; White Hmong explanations of the differences between cultural divisions that I heard also seemed to trace the distinction between Green and White Hmong to different relationships to Chinese society, portraying the Green Hmong as more resistant to civilizing influences and the White Hmong as more mannerly, more akin to the Chinese.
But more complex sorts of cultural divisions within groups, often marked by color terms, are very widespread in the region, as we have seen. So too is the paternal descent system that the Hmong practice, with surnames that they share with the Chinese, with the Yao and Lisu people, and also "probably with some groups of Lahu" (DURRENBERGER 1970). So perhaps an explanation that goes beyond simple Hmong-Han relations is needed.
COOPER (1984) perceptively points out that there is a basic contradiction between Hmong ideals of patrilineal kinship, which demand hospitality between all members of the same surname and permit marriage with all surnames other than one's own, and the presence of these cultural divisions (like the "Green" and "White" Hmong). This means that in practice, marriages only rarely take place between the members of different surnames who are in different cultural divisions, and hospitality is seldom demanded from a "clan brother" of a different cultural division. In this section, therefore, I examine some of the relations between Hmong surnames and cultural divisions.
Strict exogamy is maintained between members of different patrilineal surnames, and local descent groups are formed by members of the same surname. In my own fieldwork I was surprised to find White Hmong in Thailand sometimes referring to themselves as a "Hmong Li Hmong" or as a "Hmong Li Sua" (that is, a "Hmong, Hmong Li," or a "Chinese, Hmong Li"), which denoted different ancestral sub-groups within local surname groups (TAPP 1989,167). These sub-groups were distinguished by a number of features to do with funeral and ancestral ritual, but particularly according to whether their burials were made after the "Chinese" fashion, with stones, or in the "Hmong" fashion, a simple grave without stone, marked only with brush and leaves. It seemed clear that these Hmong/Sua distinctions, which were normally external to the Hmong and marked for them a particular kind of "otherness" (as in the formulation, "the Hmong fought the Sua"), but which in some contexts were used to distinguish cultural divisions of the Hmong themselves (as in the "Hmong Sua" group rather than the "Hmong Ntsua" or "Hmong Daw"), had somehow also become internalized within the structure of some Hmong lineages." It was clear too that these fissions within the Hmong surnames were somehow related to, or conceptualized in terms of, historical relations with other peoples.
But this is not all. GEDDES (1976, 55) drew attention to the strange fact of the Hmong (by whom he meant only the Green Hmong, not the White Hmong) having two completely different names for their "clans" (surnames), an "insider" name and an outsider one. He lists Tang as becoming Hang to outsiders, Yang becoming Ma (probably wrong), Kloo (an alternative way of writing Dlu, "black") becoming Tow (Tho, the Chinese surname Tao), Tchai (Cal) becoming Lee (Li), and Mow (Mo) becoming Song (Xyong, the Chinese Xiong)."
An interesting recent article by Kao-Ly YANG (1998) deals exactly with this question of Hmong surnames. As Yang notes, LymAN's (1974) list gives the Green Hmong surname Hang? as becoming Tang6 to insiders, Haw2 changing to Dlua6, Khangl to Plua4, Kw3 to Nkw4, Li4 to CaiS, Mua4 to Zang2, Tho2 to Dlul, Tsangl staying as Tsangl, Vang2 changing to Vu4, Xyong2 to Mol, and Yang2 staying Yang2. MOTTIN (1978, 157-63), however, gives both Khangl and Tso7 as changing to Plua5, and both Tsangl and KongS as becoming Tsangl, implying that at least in these two cases, two external surnames represented themselves internally as members of the same group, who may not intermarry, although they bear different surnames externally.
Mottin also gives Vang2 as changing to Vu5/4; and while agreeing with Lyman that those externally surnamed Kw3 become internally known as Nkw4, he gives both of these as the Green Hmong equivalents for the White Hmong surname Lau2. Otherwise, where surnames are represented in both cultural divisions, the Green Hmong "external" term is normally the only one used by the White Hmong-with only a slight change expressed in the dropping of final nasals in White Hmong (so that the Green Hmong surname Vang2 becomes White Hmong Va2, for example). The White Hmong in Southeast Asia, that is, do not use these dual terms for surnames reported for the Green Hmong at all. Why not? How might these double terms used by the Green Hmong have come about, and what is their relationship to the cultural divisions among the Hmong?
We might note here that color terms are not the only ones used to distinguish Hmong sub-groups (nor are they with any other regional group) and that there is only one case of a color term being used as the synonym for a surname (the Green Hmong surname "Tho" becoming known to insiders as "Dlu," or black). There are three particularly relevant facts that need to be explained: (1) The Hmong term for the Han, "Sua," is not only used to distinguish the sinicized Hmong of Sichuan but also another Hmong group along the Burma border, and therefore acts as a marker of a distinctive Hmong cultural division identity. (2) Some Hmong surnames themselves are divided by burial practice and descent into "Hmong" and "Sua" groups. (3) The more common Sinitic surnames in use among the White Hmong are paralleled among the Green Hmong (who also use these surnames) by the use of "insider" terms for most surnames (in some cases referring to more than one "external" surname group).
We have only one color term (black, Dlul) that functions both as a surname and as the marker of a particular Hmong sub-group identity (and beyond that, of a Miao group in Guizhou who are not Hmong at all, in the Chinese term "He Miao"); but the term "Sua" functions (1) to mark distinctions between the Hmong and others defined as Han, and (2) to distinguish in certain cases cultural divisions (sub-groups) among the Hmong themselves. And (3), the term is also used internally in the "insider terms" for Green Hmong surname with outwardly sintic names like Wang (Vang2), Lee(Li4), or Xiong(Xyong2).
THE YI HYPOTHESIS
It is generally assumed that both dual terms for surnames among the one cultural division of the Hmong (Green Hmong), and the general Hmong use of Chinese-type surnames, can be explained by the historical imposition or adoption of a Han Chinese patrilineal descent system upon members of minority groups in southwest China. The only logical alternative view to this, as I have argued (1989), is that cultural identities, including that of the Han, or Hua, emerged subsequent to the formation of the patrilineal descent system, in terms of which, differences and identities of a cultural or ethnic kind were able to be imaged. From a historical point of view, I would stress how important it is to resolve this issue of the adoption of Chinese patronymics by different southwestern groups, at different times, and in different ways or for different purposes.
Sometimes tribal names were simply transliterated into Chinese, sometimes job lots were issued, as when four surnames were approved for use among all the Dai tusi ( local officials) of Tengchong-who referred to themselves subsequently as Dai Qie, "Chinese Tai" or "Chinese Shan" (HILL 1998). Patronymics had to be adopted for registration of land, or of households, in areas controlled by Ming dynasty tusi or Chinese officials. Sometimes lineages within cultural groups were distinguished according to whether they were original lineages or sinicized ones, as with the Nosu and the Lisu, who are both reported as having six original ones and nine "sinitic" ones.
But might not the dual Green Hmong surnames, which seem, like the Hmong surnames themselves, to originate from historic relations of the Hmong with the Han Chinese, be somehow related to the "cultural divisions" among the Hmong, which have never been adequately explained, such as between the "White" and "Green" Hmong, for which explanations are given that are also couched in terms of differential relations with "others" defined as Han Chinese, or "Sua"?
It is not beyond the bounds of probability that quite different types of affiliation and identity were originally involved; military or political, or local and regional, besides social and kinship-based, for example. Throughout much of southwest China a complex taxonomic hierarchy of nested ethnic groups existed which varied from region to region. LIN (1961) describes Lolo (Yi, Nosu) society in terms of three classes: the Black-Bone nobles who practiced hunting and pastoralism; the peasant White-Bone "subjects" who were sometimes house-slaves, but usually lived in Nosu villages under feudal terms; and "slaves" who were often captured Han Chinese, which the White-Bone could also own." He emphasizes the mobility between the White-Bone and the slaves. Some of the White-Bone were descendants of mixed liaisons between nobles and slaves." They could be granted a kind of freedom while retaining some feudal duties. Slaves (who could be of captured Han Chinese or other ethnic group origin, or former White-Bone subjects) were of two types: "house-slaves" and "separated-slaves." Farming land was allotted to separated-slaves and their children might join the houses of nobles as house-slaves, until a noble marriage took place when houseslaves would be divided by gender-female slaves following the bride, male slaves staying with their masters. They would then be paired off with other house-slaves in their new homes and perhaps later be sent out as "separated-slaves" or feudal "subjects."
It is hard to credit this occurring as a viable system, but this is what the literature reports. Yet there are also cases of the White-Bone living between the domains of two nobles and owing allegiance to both; and although the marriages of slaves are said to have been arranged by their masters, there are cases of White-Bone "subjects" (very much the majority) joining together to ensure a particular noble succession. The slaves were allowed to intermarry, but such arrangements, which would have attached for example brothers and sisters to separate noble households, must have considerably complicated whatever marriage system these slaves (some were Tibetan, some Hmong) practiced! Assume, for example, captured Han Chinese slaves in late imperial China, brought up to observe the precepts of a patrilineal descent system, attached to different noble Nosu landlords of different (patrilineal) surnames; these slaves must in a sense have had a double affiliation. They were not merely, perhaps, of the surname Li that their father had bequeathed them; they now also belonged to noble clan surnames X and Y respectively of the Nosu. Probably slaves of noble household X were encouraged to marry slaves of noble house X rather than those of noble household Y, so there would have been a double exogamy: Han Chinese surnamed Li of household X could neither marry Li nor the other surnames of household Y Perhaps they were then permitted to marry Li of their own households and archaic rules of patrilineal exogamy began to break down.20
Is it possible that Hmong of some original "Dlu" (black) kinship group were somehow attached to a Nosu lineage branch that had taken the surname Tao from the Chinese, and gained their dual affiliation in this way (Tao=Dlu)? Landlords of the Miao in Yunnan, Sichuan, and Guizhou west of Guiyang were mostly Lolo, probably Nosu or related Yi peoples." DE BEAUCLAIR (1970, 13) records how the Miao of northwest Guizhou would accompany their Lolo landlord (surnamed An) to the court of the first Ming emperor. So it is quite possible that some Chinese patrilineal surnames could have been adopted in this way and at this time from Yi landlords, of tusi or lesser rank, who had themselves adopted them politically as part of a sinicizing process (see TAPP 1989).2
Might we further assume that much of what the literature reports was in fact an elite view of the social system, taken from noble (Nosu) informants? In many areas it may have been the case that farmers of varying cultural allegiances were preyed on and extorted by the martial Nosu lords descending from their mountain strongholds, carrying off children as slaves and exacting tribute, and did not perhaps see themselves as a necessary part of this system at all. We do know that the Hmong in nineteenth-century Yunnan sometimes had Yi landlords, and it seems quite possible therefore that Hmong "tenants" could have been regarded by the Nosu as a part of their social system-perhaps as white-boned subjects or house-slaves, perhaps as "separated-slaves" (i.e., house-slaves who had been given a measure of independence). This would provide another potential explanation of the term "White Hmong."
It is just possible that double Hmong surnames arose in this way, to signify a dual affiliation related to Yi peoples, which may originally have been of a regional, military, or class-based, feudal nature. The dual orientation of Hmong "otherness" towards both the "Sua" (Chinese) and the "Mang" (Yi), the repeated associations of different groups of Hmong with the "whiteness" that characterized the lower strata of Yi society, and the possible derivation of Graham's "Hmong Bo" from the Bo people, who were also associated with "whiteness" and the Yi people, all seem to point towards such a view, which would emphasize the role of Hmong relations towards the Yi in the historical past, rather than painting Hmong history, as the Hmong themselves so often do, purely in terms of Hmong-Han relations.
But I am afraid there are serious problems with trying to re-paint history in this way. For one thing, if it were the case that Hmong double surnames and cultural divisions arose out of more complex relations with both Yi and Han, one would expect to find much more overlapping of the Hmong insider/outsider terms, a whole variety of outsider terms perhaps representing one original insider term as the "original" Hmong kinship organizations were fragmented and cross-cut by different external allegiances. One would expect to find, for instance, the Chinese surname Li represented among the Hmong by at least two or three original Hmong kinship terms corresponding to it. Yet (with only two exceptions) there is a striking one-to-one consistency in the use of these insider-outsider terms, which would seem to argue against such an idea. Moreover, throughout history the Hmong have lived in extremely remote locations and valued their independence. It is unrealistic to think of them as forming a functional part of another social system, however the elite of that society may have viewed them. I would like to stress here the need for comparative research on the traditional kinship organizations of other non-Hmong, Miao groups in China, such as the Qho Xiong of Hunan and the Hmu speakers of southeast Guizhou, who may more recently than the Hmong have practiced entirely non-patrilineal systems of descent."
THE CULTURALIST MODEL
So we are reduced to the more usual explanation of the Green Hmong double surname terms (and the Chinese-type surnames used generally by the Hmong) through intermarriage with Han Chinese, rather than in terms of their relations with other peoples besides the Han (or of their adopting Chinese patronymics for strategic ends, since Hmong were rarely in a position to do this). This explanation supposes, for example, that Han Chinese of the surname Li married into the original Hmong kinship group Cai so extensively that the Chinese surname was attached to their own to signify the closeness of this relationship (Li=Cai). But there are problems with these assumptions too! Because one then has to assume that Chinese intermarriage occurred so very extensively with the White Hmong that they entirely lost their original kinship terms, since they do not use the internal surname terms used by the Green Hmong at all. If we assume such a degree of sinicization, how were they able to remain Hmong?
YANG (1998) notes that LEMOINE (1972) records there were some villagers surnamed Tho ("external term") among the Green Hmong who were not known as Dlu ("internal term"). YANG (1998) cogently argues that this could have come about through the uxorilocal marriage of a White Hmong man surnamed Tho, to a Green Hmong woman, whose children would be brought up as Green Hmong, speaking Green Hmong and wearing Green Hmong clothes. As the children of a White Hmong man with no "internal" surname, they may not have used the dual term used by Green Hmong and could have remained known simply as "Tho." The model Yang uses here is similar to the culturalist model of intermarriage according to which the Hmong originally gained their Chinese surnames through uxorilocal marriage with Han males. Actually a great deal of evidence supports this model.
Some of the smallest clans confined to a particular cultural division, like the White Hmong Tcheng2, Xe7, and Fa2, are said to have been fairly recently formed by intermarriages of Han Chinese men with Hmong women, while the Cho and (Green Hmong) Xong surnames are also known to be of peculiarly Chinese origin. Many other surnames have similar stories about remoter founding ancestors, like the Lau2 of north Thailand who traces its ancestry to a Chinese man who had married a Hmong to produce an heir (TAPP 1989, 169).
One sometimes hears of local prohibitions on marriage between members of different Hmong surnames because of an uxorilocal marriage in the past. The prohibition of marriage between the Li and the Mua in north Thailand, for example, is traced to the historic passage of the Li descent line through a Mua man's marriage to a Li girl who was the only surviving member of the Li descent group at that time. But one also hears of prohibitions between local Hmong and Chinese surname groups for very similar reasons. Hmong of the Yang surname in the Sichuan village where I did fieldwork said that in the first generation of their lineage their line had been founded not by a Hmong but by a Han Chinese, surnamed Deng, whose ears had been cut off in warfare and who had then taken refuge with a Hmong household. Deng later married the daughter of the household and took her Yang surname. For this reason, it is claimed, the Hmong surnamed Yang of the area have not been allowed to marry any of the Han Chinese of the Deng surname who lived nearby.
If the White Hmong man of Yang's example had married virilocally, his children might have married Green Hmong and eventually settled with Green Hmong, the sons uxorilocally. These mixed marriages are still exceptional, and there are no statistical studies of the residence of their children. But culture (and therefore cultural division) may tend to follow the mother in these cases, and perhaps more generally in the region.
Relatively common too is the sight of a lone Han or Yunnanese Muslim Chinese (Sua to the Hmong, but Haw or Ho" to the Thai) trader or settler who has settled down with a Hmong wife on the outskirts of the Hmong village, who may perhaps speak some Hmong and whose children can be indistinguishable from other children of the village. It could be that all or parts of all present-day Hmong clans have been formed in this way, through intermarriage with Han Chinese males (who may themselves have had first wives who were Han elsewhere), although this would amount to a reverse assimilation inasmuch as a separate Hmong cultural identity was perpetuated as a result. If this sort of intermarriage was so common in the past, minority cultures in the southwest, of this type, must be seen as hybrid cultures, whose "original stocks" have entirely vanished. Of course, the children of mixed marriages between Han males and indigenous women might in some cases have had a choice (Wu 1989), with the eldest sons perhaps receiving Chinese schooling and being assimilated back into a Han identity, while the remainder may have stayed with their mother's groups and referred to themselves in dual terms-surname Li, for example, of the Cai (mother's?) group (Li=Cai). But then the majority of these children would not have "become Chinese," and really such cases of Chinese schooling must have been quite exceptional for many minority groups. Processes of this kind, by themselves, could hardly account for SCHAFER's (1967) observation that most of the southern Chinese population has "non-Han" ancestry, although they might account for the presence of some Chinese-type surnames, and even dual terms for surnames, among the Hmong.
If we were to follow an earlier generation who argued for the regional assumption of the inheritance of fleshly, cultural characteristics (such as language and costume) through the mother, and bony, ancestral characteristics through the father," then non-Confucian mothers of the kind posited by all the assumptions of male Han marriages to minority women must have posed quite a problem to the diffusion of Confucian culture. It would be useful to have some statistical evidence of how often such marriages actually occurred, especially with regard to unions not considered as legal marriages, with women who might have been seen as concubines or mistresses. To the extent that "cultural" identity in southwestern China was maternally inherited, this does pose some intriguing historical questions about the historical diffusion of Confucian moral norms, and their relationship to the patrilineat ideology so firmly endorsed by Confucianism!
Ultimately I am concerned with what being recognized as a "Han Miao," or a Hmong referred to as a "Hmong Sua," might have implied for local notions of identity and difference. Has this been a simple case of acculturation without assimilation (see SHEPHERD 1993), of emblems of another. culture or tradition being adopted that then reinforced an alternative identity? We know of the long process of acculturation in southwest China, which has contributed to the eventual assimilation of individuals and whole groups; we know what a high proportion of the present-day Han Chinese of the region had non-Han ancestors; we know to some extent how sinicization happened through the adoption of Chinese surnames and associated burial practices and the schooling of sons by local minority elites, as with the adoption of Chinese surnames and tombstones by Shan (Tai) chiefs in the Tengchong area of Burma (HILL 1982) and Kachin chiefs on the Burma border (LEACH 1954), or even through inmarrying Han son-in-laws (SHEPHERD 1993). In the traditional "culturalist" model, which is really an elite model, these sorts of phenomena were taken as evidence of the widereaching humanism of Confucian literati culture, its emphasis on manners and civilization as an index of social identity, rather than on origins or descent.
But, besides the problem of explaining the lack of double surnames among the White Hmong in Southeast Asia, there are general problems with this "culturalist" model, which fails to explain the violent suppressions of otherness that Chinese history repeatedly displays (cf TAPP 1995).
Largely in order to explain why the wholesale assimilation of minorities to Chinese culture did not occur, Shepherd makes the point that at times in history the central state may have had an interest in encouraging the political loyalty of minority groups through their espousal of Confucian ideals of conduct, while actively discouraging their cultural assimilation as groups. With a similar sort of motive, EBREY (1996) argues that, together with Confucian concepts of inclusive identity, there coexisted a patrilineal conception of ethnicity, which was exclusivist in that it excluded those not descended from a common ancestor. Ebrey thus suggests that we look to this notion of a patrilineal "we-group" to account for the active discomfort experienced by many Han Chinese with the idea of having had non-Han ancestors and their reluctance to admit to non-Han origins.
It does seem to me, however, that both Confucian and the patrilineal ideology are importantly inclusive in the sense that they posit a unitary origin from which variety and diversity-whether cultural or of genealogical branches-radiate outwards and spread. The capacity of the patrilineal system to project mythical marriages and identifications into the past, and forge alliances where there were in fact none, is surely one of its more remarkable features. The real alternative to these "genealogical" models, of variety emerging from an original unity, should be a more constructivist or existentialist model of an original plethora of cultural differences from which unitary identities and order emerge and evolve.
It is difficult to see Confucian culturalism "coexisting" with an ideology of patrilineal descent, since they were so often identified. Nor can we see one as inclusive and the other as exclusive since cultural or ethnic identifications may be portrayed precisely in terms of the patrilineal ideology that also differentiated them, as in common Hmong claims to a common paternal ancestry with the Han, expressed through legends telling of the separation of two original brothers from whom the Hmong and the Han respectively descended. It must therefore be a mistake to attribute what we might well call a form of racism to a patrilineal ideology, and exempt (patriarchal) Confucianism from that charge.
JUST PLAIN HMONG
The Hmong of Wutong in Gongxian, Sichuan, had no awareness of belonging to a particular sub-group of the Hmong, although some of the older men agreed that they might be called "White Hmong." The younger Hmong had heard of other groups of Hmong in far-away places like Guizhou, and America, and were keen to learn what sort of Hmong they might be. They knew they were Hmong, and they spoke recognizable Hmong, although it was not a dialect heard in Southeast Asia and most of the tones were different. But Hmong was only spoken in private, and Hmong rituals were mostly private domestic affairs not open to strangers. Outside, at work or on the road to the markets at Didong or Gongxian, the local Sichuan dialect was spoken loudly and as far as I could tell by almost all the Hmong. They had stopped wearing Hmong clothes (although one of the oldest women still span cotton), but had a hybrid minority uniform brought out for festive occasions for the women to wear, which included some traditional Hmong features like leggings, headdress, and embroidered armbands and sashes.
The relation between Hmong and Han was difficult to depict in purely ethnic or cultural terms. To outward appearances, these people might have been Han villagers and took pains to pass as such where they could. In the politicization of cultural forms that has taken place in China since 1949, Hmong "culture" had been a category fiercely attacked during radical periods, and there were large parts of it, such as shamanism and certain aspects of wedding ritual, which Hmong themselves feel ashamed of because they felt them to be backward and perhaps unnecessary. Yet the Chinese were still referred to in informal talk between Hmong villagers as "Sua," and stories were still told of fierce battles in the past with the Han, and Hmong shamans and ritual specialists still presided over distinguishably "Hmong" exorcisms and funerals. At the same time a tentative cultural revival, of the sort common in China at the time, was taking place-as is evident by the efforts made to recollect the Hmong past through recording and transcribing songs, legends, and ritual (see TAPP 2001).
Twenty-eight out of the 103 households in the three hamlets of the village were Han Chinese of six different surnames who coexisted amicably with the Hmong, claiming to derive benefits from living in an area classified as a minority one. A few intermarriages had occurred between the two groups, but members of both denied in general that this happened, because "our ways are different." There did seem, however, to be some coalescence of interests in shamanic rituals, since some Han villagers would consult Hmong shamans and the practices of the Hmong shamans had begun to incorporate exorcistic practices of the Han tradition.
These were the people Graham referred to as the "Ch'uan Miao," who were probably referred to as "White Miao" in the past, and who had been generally known since at least the 1920s as "Hmong Sua" or "Han Miao," although they did not recognize this term. From my research on often-forgotten traditions of costume and embroidery, it appeared that various smaller Hmong groups had coexisted in the region in the past, who had amalgamated in a kind of "levelling" process to form the common Hmong identity they claimed today. This process of the mixture and blending of smaller Hmong traditions was not solely the result of socialist education, since both Graham and Ruey referred to it in republican times, but must have been accelerated by the social changes, population shifts, and discouragement of local expressions of cultural identity since then.
THE HMONG TAKE-OVER OF THE "MIAO"
YANG Kao-Ly (1998) discusses her surprise at watching a video (Su 1996) of the Hmong in Burma, and says how hard it was to follow the dialect of these self-described "White Hmong," who have been cut off from other Hmong groups for most of the present century. Yet this dialect is no stranger than most of the Chinese Hmong dialects, which Hmong in Thailand and Laos complain they have difficulties in listening to. When a Southeast Asian Hmong meets a Chinese Hmong there is an immediate recognition that both are speaking Hmong, but also an immediate awareness of divergence in the way things are said, the loanwords, and some of the tones and pronunciations. And there is an immediate, commonsensical, recourse to a "genealogical" model to explain this-these Hmong have lived apart from us for so long, their language has inevitably diverged a bit from ours.
Speaking a language is of course no necessary index of cultural or any other kind of identity, but in this case it is: these are people who speak Hmong and recognize themselves as Hmong, and recognize themselves as Hmong partly because they recognize that they speak Hmong. Nevertheless, there is a feeling of strangeness, of distance, at first encounter, which is immediately imputed to the length of historical time and space of separation. It does seem almost self-evident that at some time in the not-very distant past the various Hmong dialects spoken were closer together if not identical, and have diverged because of spatial and temporal separation.
The term "Miao," loosely applied to present-day populations in China, has historically been a derogatory term, with connotations of dirt and barbarism. In Vietnam, Laos, and Thailand the dominant Tai-speaking groups adopted this Chinese term to refer to the Hmong as "Meo," and in this way it became embedded into official state discourse to refer specifically to the Hmong. It has always been resented by them, and since the seventies a concerted move by ethnographers of the Hmong has largely succeeded in establishing their name as "Hmong" rather than "Meo." The success of this ethnographic effort was assured by the hundred thousand Hmong refugees from the wars in Laos who were resettled in the US, Australia, and elsewhere, and insisted on the use of "Hmong."
In China, however, "Miao" was established as an official category of state minority discourse after 1949 and has largely lost its derogatory connotations (it remains very much an official, and Chinese, term). Moreover the situation in China has been more complex than in Southeast Asia since in China there are other groups speaking languages distantly related to Hmong who have historically been, and are now, also classified as "Miao."26 Groups like the Hmu, the Qho Xiong, the A Hmao, and the Hmong are all classified together as "Miao," although for millennia they may have had no actual relations with each other and, without their official classification together as "Miao," would recognize no affinities at all with each other.27
There is thus an uneasy and vexed relation between the official term "Miao" and the local, vernacular or colloquial use of terms like "Hmong," "Hmu," "Hmo," or "Qho Xiong." The Chinese generally do not know these terms, so that they have no official meaning at all. In Gongxian the local Hmong villagers were highly aware of being officially classified as "Miao" and of the various advantages and disadvantages this setting apart of them, from their Han Chinese neighbors, gave them. "Miao" was indeed the way they would have presented themselves to any outsider who asked questions about their identity.
With the recent encounters between groups of visiting Hmong refugees from France and the US to China to meet their long-lost brethren, some very odd shifts of identity have taken place, to which SCHEIN (1998) has drawn attention. I wish to present these shifts of identity here in a slightly different context, however. When official visits were first arranged, whether these were academic conferences, guided tours or trading congresses, the authorities in China endeavored to ensure that Miao of China met these "Miao" from Southeast Asia, and the result was that the Hmong originating from Southeast Asia were not initially introduced to the Chinese Hmong at all, but to Qho Xiong "Miao" who knew no Hmong and were not Hmong, or Hmu who were in the same position. At the same time of course they met a number of "Miao" cadres from all groups who had lost their original languages and only spoke Chinese.28 It is now the case that the visiting Hmong from overseas have gone beyond these original contacts and managed to meet groups of "really Hmong" villagers, often outside official channels, with whom they can speak their Hmong language, discuss details of ritual and wedding procedure, sometimes trace family relationships, and generally recognize each other as Hmong brethren. Yet most of these Hmong dialects are still strangely distorted and at variance with the dialects of Hmong they have grown up with and speak. This often causes a surprise for them similar to that experienced by the film-maker Su Thao and Yang Kao-Ly when they heard the Hmong spoken by the Hmong of Burma. To some extent the visiting Hmong have accepted that in China they may without great prejudice be identified as "Miao," while at the same time nationalistic Miao sentiments, mostly among non-Hmong groups in China like the Hmu of southeast Guizhou, have resulted in a growing tendency to refer to nonHmong Miao groups in China as also "Hmong."
Professional linguists have contributed to this confusion by adopting the ethnonyms of particularly vocal cultural groups like the Hmong and Mien" to represent the entirety of other Miao and Yao languages, and now refer to the "Hmong-Mien" rather than the "Miao-Yao" group! Both for the Yao and the Miao language families, just one ethnonym, powerfully represented by a vocally articulate and prominent group outside Asia, has gradually crept up the genealogical ladder of branching relationships to dominate the whole. In this way the ethnonym "Hmong," which after all is only one of a number of similar related ethnonyms of a particular cultural group in China, is slowly climbing up the genealogical ladder of cultural, linguistic, and historical relationships to the point where, if this process continues on its logical course, it will finally culminate in replacing the (general, alien) term "Miao" altogether. And at least some of the non-Hmong cadres in China also appear to be going along with this process of proto-nationalism.
CONCLUSIONS
I have raised some problems here, of ethnography and history, to which I cannot pretend to have answers. But some general conclusions do emerge. When the Hmong define themselves as "Hmong" in relation to others defined as "Sua," or as "Sua-Mang," or when they define themselves in relation to other groups of Hmong, as "White" or "Green" Hmong, "Hmong Leng" or "Hmong Pua," "Hmong Si" or "Hmong Xau," differential positionings of identity are being made that may be seen to celebrate an original diversity of languages and cultures. When they hail other, non-Hmong, Miao "cousins" as "Hmong," on the other hand, a more genealogical logic is being appealed to, which emphasizes the unity that has been diverged from, and may be seen to mark a historically nationalistic narrative. And when the Hmong identify themselves as members of patrilineal surname groups that they share with the Chinese and may often trace to Chinese paternal ancestors, or tell stories about a common paternal ancestry with the Han Chinese, or refer to themselves or to other Hmong groups as "Hmong Sua," the genealogical descent system itself is being used to express important metaphors of identity and whatever should be the opposite of metaphor: the discerning of differences in likenesses.
NOTES
1. The term man applied to the Yao in northern Indichina probably derived from the
Cantonese pronunciation (man) of Chinese min (K, people) rather than the older Chinese term for southern barbarian (9, man).
2. This is a husker, rather than a kisser, like "sh" in "shoe," and would be written xua in pinyin. Both the "s" in Sua and the "h" in Hua are unvoiced fricatives.
3. I am using the Romanized Phonetic Alphabet developed by Heimbach and Smalley, with Bertrais's assistance (see HEIMBACH 1979) for Hmong terms from Southeast Asia, omitting the final consonants that indicate tone values, and converting the double vowels indicating final nasalization to "ng" endings.
4. Both the terms "Hmong Xau3" and "Hmong Paw6," which I have heard could have the meaning of "numerous," while the "Hmong Sil" might refer to the "scattered" Hmong. Here I have added numbers for the tones; 1 =b (high), 2=j (high falling), 3=v (mid-rising), 4=s (low), 5= (mid), 6=g (low breathy), 7=rn (low glottally stopped).
5. Although their language is classified together with Hmong in the third Miao branch, the (Da) Hua Miao (*Ag) speak a language unintelligible to Hmong and call themselves "A Hmao" or "A Hmo." The Xiao Hua Miao, however, may be a Hmong group classified by the Chinese with the other Hua Miao.
6. My first fieldwork with the Hmong was conducted in North Thailand in 1981 and 1982 with assistance from the Economic and Social Science Research Council and the Central Research Fund of the University of London.
7. A smaller group known as the Hmong Qua Npa, or those with "striped armbands," also speak White Hmong.
8. The White Hmong "d" becomes the sound variously written as "gl," "kd," or "dl" in other Hmong dialects. I believe that the Green Hmong voiced velar stop (k) combined with a lateral fricative (13) becomes a pre-glottalized dental in White Hmong (-d).
9. Post-doctoral fieldwork was conducted in Sichuan, Yunnan, and Guizhou between 1989 and 1992 with assistance of the British Academy.
10. For example, Green Hmong on the Vietnamese border are locally referred to as Red or Red-Headed (UFO, Hongtou) Miao; Qing Miao does not equate with Green Hmong in Guizhou.
11. On surviving "Buo," see GoULLART (1955).
12. This could be from the Burmese te-phyu. There is a related Tai term for the Chinese, ceg, which HILL (1998) spells in pinyin qie.
13. DAvIES (1909) confirms Hmong in Burma were called "White Chinamen," spelling it "Che-hpok." "Pai-Yi" (Baiyi, White Yi) was on the other hand the older Chinese term for the Tai-speaking Dai of Yunnan.
14. On the other hand, pua in a low tone would give "one hundred Hmong," which, like the Chinese statement laobaixing (ff '), refers to the common people. The statement pua3 1eng2 means "altogether," and given a tone change from 3 to 6 after Hmongl this is another possible derivation.
15. It was only in the past that White Hmong women wore undyed skirts, now kept only for special occasions; today they wear pantaloons.
16. The Xyong and Li surnames refer to the "Chinese" division as loj, or "greater," and the "Hmong" division as me, the "lesser" (or "younger").
17. In the parentheses I have converted Geddes's transliteration to the authorized one (again without tone indications or double vowels).
18. Others describe four distinct classes.
19. Children of unsanctioned liaisons between male nobles and White-Bone women were, however, called "Yellow-Bone."
20. LIN (1961) says the "white Lolo do not have a clan system of their own," although he
later describes them evolving one.
21. Although Tai in east Guizhou were also called Yi, Yi of the Nosu ethnonym dominated west Guizhou around Dading, much of Yunnan, and parts of Sichuan from the Daliang range to the northwest.
22. These Lolo were from Qianxi in Guizhou; Lolo also had a fortress at Dafang, where Hmong probably also lived.
23. SHIRATORI (1985) for example recorded among the He Miao of Kaili, Guizhou, the linked Tibeto-Burman system in which the son takes his father's last name as his first.
24. A possible derivation of this term is also "Hua" (*).
25. LEvi-STRAUSS (1969) followed Shirokogoroff in suggesting that the distinction between bone, expressing characteristics inherited through the father, and flesh, expressing those inherited through the mother, was of pan-Asian significance.
26. These non-Hmong "Miao" groups are only represented in China, and (apart from small numbers of Hua Miao) do not exist outside its borders.
27. Hmong is typically among the first (Chuanqiandian) dialect of the Chuanqiandian branch of Miao, but there are also Guiyang, Huishui, Mashan (Mang), Luobo River (AHmyo), and Chonggan River ("Gedau") dialects of Hmong classified under this Western branch (not counting the Punu dialects nor the Diandongbei dialect). The extent to which they are mutually intelligible is not known.
28. See also my "Exiles and Reunion: Nostalgia among Overseas Hmong" in The Anthropology of Separation and Reunion, ed. Charles Stafford (forthcoming, Routledge).
29. The majority of the people who are called Yao in Southeast Asia refer to themselves as Mien, or Iu Mien. But this is only one ethnonym among the Yao Min, Byau Min and Kim Mun or Landian Yao, who speak clear Yao languages, to say nothing of the Han-speaking "Pingdi Yao," the Punu ("Hualan Yao"), Kiong-nai (who speak languages now thought to be closer to Miao or She [NIEDERER 1998]), and the Lakkia or "Chashan Yao" (who speak a Taiic language, but who are all classified as "Yao" in China (see LEMOINE and CHIAO 1991).
REFERENCES CITED
BERTRAIS Yves
1964 Dictionaire Hmong-Franfaise. Vientiane: Mission Catholique. BLACKMORE, Michael
1967 The ethnological problems connected with Nanchao. Symposium on historical, archaeological, and linguistic studies on southern China, South-East Asia and the Hong Kong region, ed. F S. Drake. Hong Kong: Hong Kong University Press. COOPER, Robert
1984 Resource scarcity and the Hmong response: A study of settlement and economy in northern Thailand. Singapore: National University of Singapore.
DAvIES, H. R.
1909 Yun-Nan: The link between India and the Yangtze. Cambridge: Cambridge University Press.
DE BEAucLAIR, Inez
1970 Tribal cultures of southwest China. Taipei: The Orient Cultural Service. DOWNER, Gordon
1963 Chinese, Thai, and Miao-Yao. In Linguistic comparison in South-East Asia and the Pacific, ed. H. Shorto, 133-39. (Collected Papers in Oriental and African Studies No.5). London: School of Oriental and African Studies.
DURRANBERGER, E
EBBEY, PATRICE
FITZGERALD, C.P.
GEDDES, WILLIAM
Ap of t, phi *: TfY ld": *k
`cw Jim -, 0: 3!, S-day 30. 44sft*, Paper 75 Maw
1982
1998
HOTESSNING, OTHERMARE
1954 Political systems of highland Burma. London: The Athlone Press. LEMOINE, Jacques
1972 Un Village Hmong Vert du Haul Laos. Paris: C.N.R.S. LEMOINE, Jacques and CHIEN Chiao, ed.
1991 The Yao of south China: Recent international studies. Paris: Pangu, Editions de I'A.RE.Y
LEvi-STRAUSS, Claude
1969 The elementary structures of kinship. Boston: The Beacon Press. LIN, Yueh-hwa (Lin Yaohua)
1961 The Lolo of Liangshan. New Haven: Human Relations Area Files. LING Shun-Sheng and RUEY Yih-Fu
1947 Xiangxi miaozu diaocha baogao [aa***o [A report on an investigation of the Miao of western Hunan], trans. Human Relations Area Files 1963. Shanghai:
LYMAN, THOMAS
MCALEVY, HENRY
MOTTE, JEAN
NEIFERBER, BARBARA
RuEY, Yih-Fu
1958 Terminological structure of the Miao kinship system. Bulletin of the Institute of History and Philology, Academia Sinica 29.
1960 The Magpie Miao of southern Szechuan. In Social structure in Southeast Asia, ed. George Murdock. Chicago: Quadrange Books.
1962 The Miao: Their origins and southwards migrations. In Proceedings of the International Association of Historians of Asia Second Biennial Conference, 6-9 October Taipei.
1967 A Study of the Miao People. In Symposium of historical, archaeological and linguistic studies, ed. R S. Drake. Hong Kong: Hong Kong University Press.
RUEY, Yih-Fu and KUALA Tung-Kuei
1962 Chuannan Yaqie Miao de hun sang lisu J I IfIIl 'F9 OURY*EW6 [Marriage and Mortuary Customs of the Magpie Miao, southern Szechuan]. Taipei: Ser. A, no. 23. Institute of History and Philology Monographs, Academia Sinica.
SCHAFER, E.
SCHEN, LOUISA
SCOTT, JG
SHEPARD, JOHN
SHIRATORI, Yoshiro
1985 A provisional theory towards the understanding of the minority nationalities in China and the ethnohistorical study of South China. Proceedings of the International Conference on China Border Area Studies, April 1985, pp. 1335-48. Taipei.
Su, Thao
1996 The Hmong of Burma, Part I. California. ST Universal Video. TAPP, Nicholas
1989 Sovereignty and rebellion: The white Hmong of northern Thailand. Oxford: Oxford University Press.
1990 Milieu and context: The disappearance of the White Hmong. In Proceedings of the 4th International Conference on Thai Studies, 11-13 May, 1990 (Vol.3). Kunming Institute of Southeast Asian Studies.
1995 Minority nationality in China: Policy and practice. In Indigenous peoples ofAsia, ed. R. H. Barnes, Andrew Gray, Benedict Kingsbury, 195-221. Michigan: The Association for Asian Studies Inc.
1996 The kings who could fly without their heads: "Local" culture in China and the case of the Hmong. In Unity and diversity: Local cultures and diversity in China, ed. Tao Tao Liu and David Faure, 83-99. Hong Kong: Hong Kong University Press. 2001 The Hmong of China: Agency, context. Leiden: Brill..
WALKER, Anthony
1974 The divisions of the Lahu People. Journal of the Siam Society 62 (2): 253-68. Wu, David
1989 Culture change and ethnic identity among Minorities in China. In Ethnicity and Ethnic Groups in China, ed. Chien Chiao and Nicholas Tapp, 11-22. Hong Kong: New Asia College, The Chinese University of Hong Kong.
YANG Kao-Ly
1998 "Problems in the interpretation of Hmong surnames." Paper presented at the International Conference on the Hmong/Miao in Asia, Aix-en-Provence, 11-13 September.
ZHANG Yuan Qi, YANG Shao Long and SHEN Zheng Kun
1988 Hmoob Nyob Paj Tawg Teb [The Hmong of Wenshan]. French Guyana: Association Communaute Hmong.
NICHOlAS TAPP
The Australian National University, Canberra
Copyright Nanzan University 2002