Respected male household heads of the Rmeet (Lamet), an upland ethnic group of northern Laos, occasionally perform rituals in order to obtain ranked titles. These titles originally stemmed from lowland polities such as the Sipsòng Panna, Luang Prabang, or Nan, stratified societies in which they structured a hierarchy of officials and noblemen. Although these polities have now been absorbed into modern nation-states, the titles are still alive among some uplanders. Yet in the process of adoption, the Rmeet have altered the nature of the hierarchy they express. While titles formerly were conferred by kings or high-ranking officials, Rmeet ritual demands wife-givers, the relatives of a man's wife or mother, to perform this task. Often in the context of healing rituals, Rmeet approach these affines to bestow a title on them as part of an exchange of gifts. In Rmeet ideology, wife-givers are superior to wife-takers. They are the source of health and fertility of both humans and rice fields. Yet these wife-giving hierarchies do not form a pyramidal structure that integrates the entire society. Everybody has both wife-givers and wife-takers, and thus everybody is at once in a superior and an inferior position to other families (houses).

Thus, two contradictory types of hierarchy meet in Rmeet title-giving: a transitive hierarchy, in which the superiors of my superiors are also my superiors, exemplified by lowland principalities and signified by ranked titles; and an intransitive hierarchy, in which my superiors' superiors are not necessarily superior to me and might even be my inferiors—for example, when my wife-givers' wife-givers are my wife-takers. Transitive hierarchies have a top and a bottom, whereas intransitive ones do not (Parkin Reference Parkin1990). The question is, how do these two types of hierarchy relate in the present case? What does this tell us about communications between upland and lowland societies in Southeast Asia, and between predominantly “egalitarian” structures and predominantly “stratified” ones?

These questions have pervaded the study of highland Southeast Asia ever since Edmund Leach published his Political Systems of Highland Burma ([1954] Reference Leach2001; see also Kirsch Reference Kirsch1973; Lehman Reference Lehman and Steward1967; Robinne and Sadan Reference Robinne and Sadan2007; Russell Reference Russell1989; Scott Reference Scott2009). This region was shaped by centralized, stratified social orders, such as the “galactic polities” (meuang) of Tai speakers (Tambiah Reference Tambiah1985) and the Chinese state, as well as by village-based societies lacking discernible strata, such as the Rmeet today. Although the meuang have given way to nation-states, there are important continuities in interethnic relations. These continue to be expressed by highland–lowland contrasts, with uplanders usually seen as the receiving side of cultural influence or political force—a distinction made by both local policy makers and, although much more critically, academic scholars (e.g., Jonsson Reference Jonsson2005; Turton Reference Turton2000). In these accounts, the highlanders' relation with lowland polities is often seen as either imitation and adaptation or resistance. But all these studies face the question of how upland societies as self-reproducing systems manage external influences in their own terms (e.g., Proschan Reference Proschan2001; Tooker Reference Tooker1996).

The mere classification of some social orders as “states” and others as “nonstates” or “margins” does not say much about the relationships they have with each other and how these shape their reproduction. It is only by looking at the specificities of these societies that the way in which they process their relationships, internal and external, can be elucidated. The present approach focuses on ethnically defined social systems—conveniently called “societies”—in contact. It does not give initial priority to the difference between centralized societies (states) and those without supravillage centers or stratification. Even if one particular society might possess more military force, technological skills, or cultural splendor in contrast to another, the other society's response is not exhausted by compliance, passivity, or simple reversal. Processes such as Tai-ization, sinicization, or integration do not follow some kind of inevitable gravitational pull. First of all, they consist of structured action on the side of those who become Tai-ized or integrated. Integration, even when governments pursue an active interest in it, is ultimately acted out by the integrated. These actions need a specific perception of the situation and values in order to be performed. They arise from a specific society's view of the relation between different systems.

Diverging notions of hierarchy are crucial in the process of interethnic communication. Robert J. Parkin's (1990) terminology of transitive and intransitive hierarchy is helpful here. In the transitive hierarchies of stratified societies, each level is fixed in relation to the others according to one or more shared scales of value (e.g., Dumont Reference Dumont1980). But this does not mean that “egalitarian” societies lacking such features do not acknowledge status and power hierarchies at all. Equality and hierarchy are not mutually exclusive features that unify every representation of a given society, but qualities of particular relationships within a society as a whole. Thus, any society processes both egalitarian and hierarchical values. Apparent “egalitarianism” often consists of contextual hierarchies that complement each other. In particular, systems of ideas are structured by value hierarchies that can be discerned from communication and action (e.g., Barraud et al. Reference Barraud, de Coppet, Iteanu and Jamous1994). Such valorized action coagulates in social hierarchies that may be confined to particular contexts or subsegments of society, or to intransitive status.

The latter is the case with relations between wife-givers and wife-takers in asymmetric marriage alliance, in particular with a category of kin in which the mother's brother's daughter (MBD) is the closest relative. This is a common marriage preference in highland Southeast Asia (e.g., Kammerer Reference Kammerer1998; Lindell, Samuelsson, and Tayanin Reference Lindell, Samuelsson and Tayanin1979; Tooker Reference Tooker1991) and will be referred to as MBD marriage here. In fact, most marriages in accordance with the rule are not with the MBD, but with other relatives from the same category—that is, women from a former wife-giver's group. Wife-giver superiority is a common, though not universal, feature of these systems. Note that this is a matter of specific cultural value, not enforceable power. As everybody has wife-takers and wife-givers, the hierarchy does not add up to a top-to-bottom model, only to higher and lower status from any specific vantage point within the system.

MBD marriage can create long-term relationships between unilineal descent groups (patrilineages in the Rmeet case) because a woman moves from one line—that of her father—to another—that of her husband—along the trajectory of another woman in an earlier generation: her father's sister, when she married her husband's father. Therefore, MBD marriage repeats the alliance between two lines of the preceding generation, but the relation remains asymmetrical. At least in Rmeet ideology, there is thus no way to ultimately balance the gift of a bride by reciprocating something—a person—of the same value (Sprenger Reference Sprenger2006a).

Such systems tend to impede the permanent accumulation of prerogatives or resources in groups whose position is defined by marriage only (Leach Reference Leach1969, 279; Tambiah Reference Tambiah2002, 86–87 n. 8). Even the wealthiest and most privileged group has to recognize the superiority of their wife-givers. This principle pertains when MBD marriage is understood as a value idea that motivates action and judgments in any particular situation. In the Rmeet case, even when not realized, MBD marriage provides a central clue to kinship terminology, ritual, and social action. The relation of wealth and prestige with affinal hierarchy thus creates perennial ideological and practical tensions within such a system as a whole.Footnote ¹

Consequently, a hierarchy based on values that are independent of wife-giving and wife-taking creates a different kind of asymmetry. Referencing Louis Dumont (Reference Dumont1966), Parkin (Reference Parkin1990, 477) points out how transitive hierarchies of nobility or ritual purity upset marriage systems. In the Indian caste system, for example, the tendency of women to marry up within castes creates a surplus of men at the bottom and of women at the top of the hierarchy. Here, rank provides an external framework for marriage. Other than in MBD marriage systems of the present type, where marriages initialize intransitive hierarchical relations, they follow hierarchy in these systems.

Yet, if the Indian case provides an example of how hierarchies that are independent of marriage alliance can “swallow up” the latter, the following provides the reverse situation. Two types of society dominantly structured by opposing principles of hierarchy have interacted in northern Laos for at least two centuries: The first are the lowland Tai speakers, mostly wet-rice farmers and Theravada Buddhists, largely defined by centralized polities called meuang. These include the Lue (Lü, Lu), a group most conspicuously focused in the kingdom of Sipsòng Panna in Yunnan, who numbered 123,000 in Laos in 2005, and the Lao, the country's dominant ethnicity, founders of the kingdom of Lan Xang and later Luang Prabang. The second society are the Rmeet (Lamet), a population of 20,000 upland swidden farmers and non-Buddhists whose language belongs to the Palaung-Wa branch of the Mon-Khmer family (Steering Committee 2006, 15). The intransitive superiority of wife-givers over wife-takers structures much of Rmeet society and ritual.

There is also a less prominent hierarchy of ranked titles held by a minority of the population that has been partially adopted from the Lue, Lao, or the northern Thai principality of Nan (see Evans Reference Evans and Turton2000, 273). It is difficult to tell Lue, Lao, and Nan influences apart. The available historical records do not reveal much regarding Rmeet and lowland interaction, although a little is known regarding the Lue. Although the title system operates in distinction from the affinal hierarchy, it is dependent on the latter and, in terms of value, encompassed by it.

Data on Rmeet were collected during one year of anthropological field research in Nalae District, Luang Nam Tha Province, on the eastern border of the Rmeet settlement area, from 2000 to 2001, and during four subsequent trips to Luang Nam Tha and Bòkaeo provinces in 2002, 2005, and 2007, amounting to seven months. This represents the first in-depth ethnographic research on the Rmeet since Karl Gustav Izikowitz's study in the 1930s (Izikowitz [1951] Reference Izikowitz1979, 1985, [1944] Reference Izikowitz and Berngrim2004). I did not witness title-giving rituals during my fieldwork, but I collected data from title-holders and participants in recent occasions up to 2005. Data on lowland principalities were drawn from English, German, and French literature.

The analysis proceeds in three steps. The first section reconstructs the ranks and classes of Sipsòng Panna Lue, as an example of how titles functioned in their original context. Other lowland principalities that interacted with Rmeet, such as Nan and Luang Prabang, had comparable ranks on the village level, often with similar nomenclatures. The next section documents how some of these titles feature in Rmeet society and how this practice has been transformed up to the present. The third section shows how the transitivity of rank is subordinated to the Rmeet kinship system in form of the bestowal ritual, which, time and again, reproduces the relationship between the two hierarchies. The conclusion addresses one question that structures the debate on egalitarianism and hierarchy: Why would an “egalitarian” society adopt ranks?

Figure 1.

Northern Laos, provinces and borders, modified from Bounthavy/Taillard (Reference Bounthavy and Taillard2000: 15).

Lue Hierarchies

It is difficult to reconstruct in which lowland principalities the titles used by Rmeet today originated. Before the French administration, they were subject to Nan and Luang Prabang, with some Lue polities as possible influences as well (Grabowsky and Turton Reference Grabowsky and Turton2003, 236; Izikowitz [1951] Reference Izikowitz1979, 11; Lefèvre-Pontalis [1902] Reference Lefèvre-Pontalis2000, 155–57). All of those principalities distributed titles to village officials and local leaders, with titles for rural elites strongly resembling each other (Bowie Reference Bowie and Durrenberger1996, 109; Reinach Reference Reinach de1911, 236). Among the receivers were members of non-Tai-speaking ethnicities, for example, Mien (Jonsson Reference Jonsson2005, 76) and, closer to the Rmeet, Khmu (Evrard Reference Evrard, Robinne and Sadan2007, 133, 144–45).

After the French expansion into Laos in 1893, part of the Rmeet area was administered directly by France, and another part—including the area of my research—by Luang Prabang (Izikowitz [1951] Reference Izikowitz1979, 39–40). The Lao kingdom gave titles such as phia and saen loong to local leaders, including non-Tai, both to the west (Kandre Reference Kandre and Kunstadter1967, 620) and to the east of the Rmeet (Aymé Reference Aymé1930, 31, 40, 66; Bouté Reference Bouté, Robinne and Sadan2007, 189). The latter area was just south of the meuang Ou Neua, where the Lue close to Rmeet originated. It is unclear which titles were used in Ou Neua and to what degree the neighboring Sipsòng Panna influenced them. Another influence may have been the Lue-dominated polity of Meuang Sing/Chiang Khaeng (Grabowsky Reference Grabowsky and Friedrich2006; Grabowsky and Renoo Reference Grabowsky and Wichasin2008). Thus, no perfect match between the Rmeet titles and any of the lowland systems can be found. But in general, the rank nomenclature was quite similar among the four large adjacent meuang, which shared close ties and cultural similarities until the end of the nineteenth century: Sipsòng Panna, Kengtung in Burma, Nan, and Lan Xang/Luang Prabang (Davis 2000; Hsieh Reference Hsieh and Harell1995; Peters Reference Peters1990: 343, see also Souneth Reference Photisane1996, 443). As the titles in others of these meuang are more difficult to reconstruct from the available literature, I choose the Sipsòng Panna as a starting point.

Lue arrived in the Rmeet area in the late nineteenth and early twentieth century. They settled in Tafá, an important point of access to a major road for Rmeet until today, and on the banks of the Nam Tha, where a number of Lue villages are still found (Izikowitz [1951] Reference Izikowitz1979, 25–26; Songsak Reference Watthanakul1988). Among them is Ban Mo, the closest lowland contact for the Rmeet whom I studied. According to some informants, Lue there have the same title system as Rmeet. Because of their ability to read and write, the French colonial administration preferred Lue to take on administrative positions. As Rmeet were engaged in trade and other contacts with Tai speakers they considered to be rich and powerful, Lue soon became sources of cultural influence. Izikowitz singles out how some Lue headmen in Tafá

Izikowitz ([1944] Reference Izikowitz and Berngrim2004, 188) mentions phia as one title thus sold and acknowledged in both systems. A comparison of titles among present-day Rmeet and historical Lue reveals further overlap. Lue hierarchy, exemplified by the Sipsòng Panna, can thus provide a model and even a possible source of the Rmeet title system at the time of its introduction.

The kingdom of Sipsòng Panna, meaning “twelve panna-provinces,” with its capital at Jinghong (Chiang Rung), was situated in today's Yunnan, China. As usual in meuang polities, the king did not rule directly over the entire territory, but only over the capital and its adjacent regions. He had the allegiance of the rulers of the subordinate principalities (cao moeng [meuang]),Footnote ² numbering more than thirty and ordered into twelve panna (literally, but not descriptively, “thousand rice fields”) (Lemoine Reference Lemoine1997, 186–87). He also ruled over the heads of the panna and directly over particular villages in each meuang. Footnote ³

This graded structure of political entities contained several intertwined hierarchies, the most important being classes, ranked official positions, and so-called field official grades, which were assigned to the latter. The entire population was ordered into six classes:

1. I. mòm—the king and his close patrilineal relatives

2. II. wung—matrilineal and distant patrilineal relatives of the king; lords of the principalities (cao moeng) without blood relations with the king

3. III. luk lan thao phia/phaya—not considered nobility, but a somewhat intermediary class having little kinship relations with wung but eligibility for headmanship

The three lower classes were composed of the following:

1. IV. tai moeng/meuang—original inhabitants of villages, in charge of communal affairs

2. V. tai laek nòi, kun haeng cao—servants of the cao-lords, persons who had left the service of noblemen to do agricultural work

3. VI. xa/kha cao, xa/kha vat—slaves of lords and temples, respectively (Gao Reference Gao1990, 282–83; Hsieh Reference Hsieh1989, 113–14)

Officialdom differed from this order as it contained ranks, not class strata, and therefore showed more mobility. Yet the difference between nobles and commoners was observed, as the first three ranks could only be attained by noblemen. It was ordered into ten ranks:

1. 1. cao mòm ceng ha—head of government, only available to class I

2. 2. hòi dang—similar to ministers, available to class I and occasionally to class II

3. 3. cao long (military officers, high officials), pò lam (border-governing officials, mediators between moeng and center)—available to low-ranking class I members, class II, and occasionally class III

4. 4. cao long, cao bok—jurisdiction of a township, classes II–IV

5. 5. cao ho sip, cao sa ma—jurisdiction of an administrative village, classes II–IV

6. 6. phia/phaya long—jurisdiction of an administrative village or large hamlet, classes II–V

7. 7. phia/phaya—jurisdiction of a hamlet, classes II–V

8. 8. ca long—jurisdiction of a hamlet, assistants to phaya, classes II–V

9. 9. ca—assistants to phaya and ca long, common starting point for careers of classes II–V; many from class V

10. 10. saen—assistants to phaya and ca long (Gao Reference Gao1990, 281–83)

Within this framework, specific tasks were assigned to officials, resulting in a large number of individual titles. Thus, a saen laek was the saen in charge of supervising the blacksmiths (laek) (Gao Reference Gao1990, 283). This variability, if not the literal titles, is also found among Rmeet.

The official ranks were supplemented by an additional hierarchy, called “field official grades” in the literature. It was ordered according to the number of fields that were worked on the officials' behalf by serfs or through corvée labor. A percentage of the yield from these estates had to be given to the official, or they were worked communally (Kato Reference Kato1990, 366). The number of fields indicated in the title is not the actual number of fields an official benefited from. Gao Lishi (1990, 80) suggests that officials claimed land as opportunity made it available, probably with an eye to the amount controlled by others. It is possible that “field grades” and land were mostly assigned to officials by the meuang level of administration along with special tasks such as tax collecting, not always by the king (Liu Rui Reference Liu1988, 13). The title hierarchy contained seven ranks from na saen long (great 100,000 fields) to na sip nòi (less than ten) (Gao Reference Gao1990, 79–80; Ratanaporn Reference Sethakul and Turton2000, 324).

The three hierarchies of class, official rank, and field official grade differed in the way in which the levels were related. While officials regularly moved up the ranks of officialdom, according to their class, a leap from commoner to nobleman was much rarer. Only wealthy tai moeng (class IV) may have been able to buy their way into lordship (Song Reference Song and Narramore1989, 12). Furthermore, rankings did not always constitute chains of command. Field grades were probably rank pure and simple, while at least the lower five ranks of officialdom were linked as givers and receivers of orders (Liu Jie Reference Liu and Narramore1988; 8; cf. Chen Reference Han-seng1949, 20).

The Lue from whom Rmeet bought titles originated from a meuang where they had been the ruling ethnicity, but were no longer part of it. The titles were conceived as part of a fairly tight hierarchy in which clearly designated ranks were linked to privileges and power. Among Lue migrants, this order was probably present only in fragments, if at all. Thus, what they sold to Rmeet were not actual power positions, which might very well be the reason why they were open to sale to outsiders in the first place. Insofar as titles originally occupied an important position in the semantics of social order, they now had become “free” and underdetermined. This was an important condition for their transfer and subsequent transformation into a “Rmeet tradition.”

This process would have been less likely if all titles had been officially distributed or recognized by the Royal Lao administration. Lue titles were not connected to administrative duties, and they were available to anyone who could afford them. Significantly, Rmeet do not remember titles as being linked to administrative positions. Village headmen used to hold titles, but still had to acquire them by ritual feasting. They could receive higher ranks by more feasting, but without changing their administrative position. What is more, although Rmeet still tell about relations with the king, he is not remembered for giving titles, but rather medals to accomplished villagers. Therefore, it is possible that the bestowal of titles to uplanders was not a major practice of the Lao administration in the Rmeet region. Also, the system would not have survived the 1975 Socialist revolution, which ousted representations of “feudal” hierarchies. Yet the profusion of the titles phia and saen loong today is probably attributable to the reinforcement that the title system received when Lao and Lue influences meddled.

There is a final point to be made. While there was a clear patrilineal bias to the aristocracy of meuang-building Tai speakers such as the Sipsòng Panna, Michael Moerman's unpublished notes suggest that the relations between lines were not determined by the wife-giving/wife-taking difference (LeBar, Hickey, and Musgrave Reference Lebar, Hickey and Musgrave1964, 209). Marriage was subordinated to class endogamy, but did not produce hierarchy by itself (Song Reference Song and Narramore1989, 11; Zheng Reference Zheng1990, 291; see also Lemoine Reference Lemoine1997, 175). For Rmeet, on the other hand, the main way to obtain a title is to acknowledge the superiority of wife-givers.

Rmeet Titles: Ruled Variance

Rmeet society did not produce any political organization beyond the village level. Before integration into the state, the only office that centralized and unified the village was the village priest (xemia, samaan), who was responsible for relations with the village spirit. Although he had the potential to regulate villagers' behavior, according to his character (Izikowitz [1951] Reference Izikowitz1979, 112–14), his responsibilities did not extend to relations outside the village. This was the task of the village headman (nai ban, a Lao term), an office that was introduced by a meuang polity, presumably Nan, and elaborated by the French and Luang Prabang administration at about the turn of the twentieth century (Izikowitz [1951] Reference Izikowitz1979, 112; [1944] Reference Izikowitz and Berngrim2004, 99). Today, both offices coexist, the village headman being more influential.

Besides official positions, there was an emerging class distinction, at least in the 1930s. Rich men joined a category called lem after an initiation ceremony performed by other lem. Lem enjoyed privileges in field plot assignment, rituals and the settling of disputes (Izikowitz [1951] Reference Izikowitz1979, 116; Sprenger Reference Sprenger, Robinne and Sadan2007).Footnote ⁴ Izikowitz pointed out that this institution is more pronounced among Upper Lamet and less so among Lower Lamet, but the nature of these ethnic subcategories is unclear, and no Rmeet whom I talked to recognized them.Footnote ⁵ Today, the lem distinction is absent in the villages I know, including Mokhala Panghay in the former French-governed region, where Izikowitz conducted most of his research. But in all of these villages, there were at least memories of a system of ranked titles, and quite a number of them still practice it. There are indications that the lem elite described by Izikowitz and title bearing—which he does not elaborate on—are basically identical. The rituals of bestowal are similar and called by the same word (plug), and a holder of the title phia was among Izikowitz's informants (Izikowitz [1951] Reference Izikowitz1979, 116). Rmeet acknowledge that the titles are in chlo yam, “lowland language,” which usually means Lao but can designate any Tai tongue. One interlocutor even said that the titles originated among Lue. Others claimed that Rmeet created them, a claim that follows a pattern of “indigenization” of cultural borrowings after several generations (Sprenger Reference Sprenger2006c).

Yet Rmeet took up the ranked titles with important modifications. Any kind of status recognizes particular values, and ranks or classes do so by (possibly multiple) lineal value scales (Luhmann [1980] Reference Luhmann1993, 26 n. 21). A rank can only be bestowed from a position that is either equal to or above it or that encompasses the entire ranking system. Yet in the Rmeet case, the person who actually bestows the rank on somebody does not need to have any title at all; even a poor man is able to bestow a title on a much richer one. It suffices that he is the wife-giver (taa) of the title receiver. The Rmeet converted ranks into a function of wife-giver superiority.

The Rmeet word for title is ngyai, “name.” Proper names and titles are differentiated by the verb used for their bestowal, ncheai ngyai (to give a name to a child) or plug ngyai (to give a title). Both rituals are performed by wife-givers but are otherwise quite different. Adult Rmeet are called by teknonyms, starting with words for “father,” “mother,” “grandfather,” and “grandmother” (identical for paternal and maternal). In case of titles, only taa (grandfather/wife-giver) and yaa (grandmother/wife-giver) are generally used as prefixes. Most titles add an individual component after this kin term and the rank indicator, thus functioning as a personal name. Titles extend to their holder's wives, who normally share the teknonym. Women themselves cannot buy titles. Even when a title-giving ceremony is performed as part of a healing ritual for a woman, it is her husband or son who receives it. Titles can not be inherited.

A major feature of ranked titles today is the range of difference in regard to the details and the general understanding of the system. Therefore, instead of providing a synthetic list of titles that summarizes the major congruences, I present two lists obtained from single interlocutors with comments. Titleholders are few in most villages. Therefore, the understanding of the rank hierarchy varies considerably between villages and individuals. Titles known in one village might be unknown in others, and titles that recognize not only wealth but also other factors such as education do not occupy a fixed position within the rank order.

The first list is the most differentiated one I received. The interlocutor is from Nam Thung, the village that today contains Mokala Panghay. He is a descendant of TaaPia, a prominent titleholder, lem, and informant featured in Izikowitz's monograph ([1951] Reference Izikowitz1979, 116).

1. 1. phia

2. 2. phei

3. 3. chajia

4. 4. saen loong

5. 5. all other saen titles (saen kham, saen monmii, etc.)

6. 6. meun

Few people acknowledged titles below saen, and these varied; some mentioned thao, meaning “elder” in Lao. Also, few people mentioned titles above phei. Phia, the title of my interlocutor's great-great-grandfather, was identified as the Rmeet version of the Lao title phanya, but also was phei on other occasions. All of these terms are related to phia/phaya, levels 6 and 7 of Lue officialdom. This places chajia in the structural position of ca, although phonetic resemblance is vague at best. Thus, the highest title in the Rmeet order belongs to Lue and Lao village administration, where it was subordinated to the village headman. Today, there is no such fixed subordination. Thus, the use made of the titles within Rmeet society was quite different from the Lao or Lue.

The second list, obtained from a Takheung villager, is shorter and closer to general consensus:

1. 1. phei, phia

2. 2. chajia

3. 3. saen loong

4. 4. saen kham

Generally, saen kham is acknowledged as the lowest title, available even to those without much wealth. It is transferred only on the occasion of an illness in order to increase a person's well-being and health. In other lists, additional titles starting with saen, such as saen monmii or saen beña, appear between this level and saen loong. These titles acknowledge some special ability of their holders. Saen beña, for example, indicates that a person is intellectually sophisticated, well educated, or has “good thoughts.”Footnote ⁶Saen loong indicates some wealth, but also the ability to act as a mediator. Rmeet understand loong, Lue for “grand,” as “to search,” thus befitting somebody who walks from house to house in order to solve disputes. Still, there is a marked difference between these titles and chajia or phei, which are reserved for wealthy elders.

Yet variation in the interlocutors' statements was rather great. For some, the function of blessing, promotion of good fortune, and the healing power of names was predominant. One even said that if an ill person does not get better after the ritual, the next higher title is bestowed on him. He explained that it is the high status of the titles that effects the healing. In this perspective, shared at least partially with most other informants, titles are sociocosmic operators that resituate a person in relation not only to society but also to those cosmological forces (often spirits) that made him ill in the first place.

But contrary to most other Rmeet healing techniques, cosmological repositioning by titles is not accessible to everybody. Most interlocutors stressed that the title always depends on a person's standing. Some even maintained that a person had to move through each of the ranks before acquiring the highest titles. One villager progressively gained the titles saen kham, saen monmii, and phei at about a decade's distance each. His father had acquired five titles during his long and illustrious life, each being higher than its predecessor. A tuned-down version of this view stated that a person had to acquire at least one lower-ranking title before being able to receive a high one. The distinction between the two classes of titles is drawn between saen loong and chajia.

Besides variations, there is a tendency to overtly manipulate the system, at least when talking to the anthropologist. Holders of titles such as saen loong or saen thuek would occasionally place their own titles higher in the hierarchy than other informants did. This might seem naive on their part, as if they did not consider that I would ask someone else. But more consistently, this strategy can be explained by the tacit acknowledgment that there is little general consensus on the order of ranks. Still, their active upgrading of their own title demonstrates that this is not a matter of indifference.

Today's rank order is not a rigid system that represents a single value scale and purveys the entire society with a transitive order. Rather, the title is supposed to reflect the social standing of its holder among his immediate peers or his social network. As the highest unit of social organization is the village, local title distribution often takes a particular village and neighborhood as the model for society as a whole and refers to current conditions of wealth, power, and status.

Part of this flexibility presumably originated during a period of transition that resulted in greater variety among villages. Present-day interlocutors remember titleholders and the bestowing ritual of the past in a way that resembles Izikowitz's description of a village upper class. The bestowals were major social events; elderly informants showed great excitement remembering them. According to Izikowitz ([1951] Reference Izikowitz1979, 116), in the 1930s, the fellow lem performed the ceremony, while today wife-givers fill this role. Yet as the lem class—at least among the so-called Upper Lamet—was more or less endogamous, this is not necessarily contradictory (ibid.: 99). Interlocutors maintain that it was always wife-givers who gave titles, but some of the most prominent titleholders of the past had their ceremonies performed by unrelated dignitaries from other villages. This indicates a shift of titleholding from an institution that closely mirrored lowland hierarchies to one that is part of Rmeet status reproduction. What was established as a communicative device to relate to lowland elites and societies became an institution that embedded crucial outside representations into village society.

The period of prominent titleholding was disrupted when the Second Indochina War spread to the Rmeet area in the second half of the 1960s. Fighting, migration, and the loss of wealth and livestock robbed village life of the stability necessary for the maintenance of the rank order. After the revolution of 1975, the new socialist government discouraged rituals, in particular among highlanders, and suppressed displays of prerevolutionary power structures. Especially those Rmeet villages that moved to roads after the return of peace reacted strongly to these pressures. It seems that most villages which had settled on the road from Ban Houeisay to Luang Nam Tha in 1976 abolished title-bestowing on this occasion. Yet some villages that had moved earlier or stayed in the mountains revived it in the following decades, and title bestowals were performed as recently as 2005. But the connection between high status and titles is more tenuous now. Even long-established and wealthy village headmen do not necessarily hold titles. As one interlocutor commented, the titles that count in the present are nai ban (village headman) or pathan (president), positions in the state administration. By this, he expressed how forms of communication between Rmeet society and the centralized hierarchies outside it have changed.

The effect is twofold: The old title system became increasingly assimilated into the specifics of Rmeet sociality, which enabled further modification. Each new bestowal today refers to earlier bestowals among Rmeet, not to any outside order. This way, the order of titles assumes the character of a “tradition,” action that is signposted as repetitive and classified as part of internal reproduction. From this perspective, the claim of many interlocutors that the titles originated among Rmeet is justified. The ethnic Other that functioned as a potential check, the meuang hierarchy, has lost relevance compared to the modern state. On the other hand, modern state administration offers new semantics for inside–outside communication to be integrated into Rmeet society. Village headmen or teachers become subject to local modes of status transmission and ritual blessing that work as relay points between state offices and status construction within villages.

Connecting Hierarchies: The Ritual of Bestowal

Rmeet wife-givers have important functions in most life-cycle and fertility-inducing rituals: at birth, death, house warming, sowing, harvesting, and some healing rituals. Their superiority is often described to the outsider in terms of transitive political hierarchy: “Wife-givers are like lords, like the president,” “wife-takers have to obey them.” In practice, though, wife-givers have neither the physical means nor the social support to force their will upon their wife-takers. They do have the right to demand ritual gifts, in particular extended bridewealth, and also to prolong bride service periods after engagement, when their son-in-law lives and works in their home. Yet the superiority is not transitive. While wife-givers of wife-givers do have a claim to bridewealth, this gift is less than the one given to the bride's household. They are thus not considered to be higher up the hierarchy than immediate wife-givers.

Therefore, affinal hierarchy is relative. While Rmeet do not think of their marriage system as a circle, the hierarchy is exclusively constituted by the marriages between two particular houses or lines. A reversal of the relationship by another marriage is possible through a higher bride price and a gift to the old wife-givers. In this case, the hierarchy between the intermarrying houses is effectively reversed, former wife-givers becoming wife-takers. Rmeet explicitly acknowledge this as a break of the rules that needs balancing by additional gifts. Even then, dangerous spirits may emerge from such reversals. In all respects, MBD marriage preference is a powerful value that organizes structural features and guides social action and judgments but, as often with central values, can only be approximated.

Bestowers of titles are chosen from two types of relations: wife-givers and officials. Yet in any normative statement, Rmeet stress that titles are given by wife-givers. Also, the ritual exchange that accompanies title bestowing is modeled upon bridewealth, and thus the official's superiority is encompassed by the model of the wife-giving relation. In fact, some Rmeet explain the choice of officials by saying, “they are big like wife-givers,” while they apparently do not feel the necessity to legitimate the choice of wife-givers in the first place.Footnote ⁷

In most recorded cases, the wife-givers approached for the ritual were close affinal relatives, such as the wife's father, wife's brother, mother's brother, or mother's brother's son. If these persons were not available in the village, genealogically more remote wife-givers were chosen. Although a wife-giver does not need to be wealthy or hold a title himself in order to perform the ritual, considerations of personal status occasionally override the affinal relationship. In a few cases, wealthy wife-takers or a lowland acquaintance were chosen for the task. The latter corresponds to a pattern of outsiders being integrated as wife-givers into Rmeet social classification (Sprenger Reference Sprenger2006b, chap. 9.5).

The following description refers to title bestowing in the mountain village Takheung in the 1990s and 2000s. Occasions were either an illness, the erection of a new house, or, in case of higher titles, a buffalo sacrifice for the house spirit (often also for healing). The person who wants to “buy” (wei) a title asks his wife-giver to perform the ritual for him. If the wife-giver decides to comply, he meets with his own relatives and the village dignitaries, the village headman, the priest, and the elders of the council (thao kae). Together they decide which title to bestow, according to the receiver's wealth, special abilities, and previous titles.

On the assigned day, the wife-givers arrange a gift bowl (khan) in either the village ritual house or their own workshop building (both called cuòng). The bowl contains four banana leaf cones, a boiled chicken, and a number of Piasters (French colonial currency)—most commonly four, but it depends on the number of coins that the receiver wants to return. The wife-givers, accompanied by relatives and dignitaries, then proceed to the house of the title receiver, firing shots from a rifle in the air. Upon their arrival, the receiver answers by firing the same number of shots or double from his rifle—often, the count of shots matches the number of Piasters being transferred. After performing a sword dance inside the house to drive out malevolent spirits, the wife-giver puts a cap on the receiver's head and places a towel on his shoulder. This act corresponds to the winding of a new turban around the head of a new lem in Izikowitz's account ([1951] Reference Izikowitz1979, 116). In both cases, the most sacred part of a person's body, the head, is touched. This is normally avoided and even leads to compensation payments if done in anger. This way, wife-givers confirm their familiarity with as well as their superiority over the title receiver. At the same time, they are forced to acknowledge the receiver's rank. As part of the ritual, they utter that they should be struck dead if they do not address him with his proper title. In fact, the wife-givers who perform the ceremony are sometimes the only persons who actually use the title when addressing its holder. The term of address acknowledges both rank and kin relationship. For example, bei saen loong combines “sister's husband” with the title.

After this, the wife-givers perform a wrist-binding ceremony for the wife-taker and his household, a common form of blessing not restricted to this particular kinship relation. In some cases, the wife-taker's forehead is smeared with steamed rice and pieces of chicken, another unusual gesture of familiarity and subordination—this type of feeding the “soul” (klpu) is usually done on the palms or the wrists. Finally, the gift bowl is handed over to the title receiver. A meal of pork or buffalo offered by the new titleholder follows.

The titleholder delivers his payment immediately after this or on the following day, when he goes to the wife-givers' workshop. It consists of a gift bowl of the same type, containing the original Piasters of the wife-givers and either the same or double the amount from the title receiver, doubling being a requirement for higher titles. He also transfers a cow or half a buffalo, the latter as money or shared ownership. The wife-givers may reciprocate—often not immediately, sometimes never—with an assortment of animals and objects: a chicken, a dog, uncooked rice, a sickle and a machete, a trouser and a shirt, and, most conspicuously, a clay jar and a blanket. The last-named are typical gifts of wife-givers, the jar promoting the taker's fertility. Some interlocutors even received a mortuary blanket, an important item used to cover a dead man's body before burial.

Except the exchange of Piasters, the gift of a cow and the (optional) answer in different objects exactly parallels bridewealth exchange, although of a minor type. While title bestowers are generally close wife-givers, the bestowal resembles exchanges with the wife's mother's house of origin, that is, wife-givers of wife-givers. These can claim bridewealth amounting to about half of what the wife's house receives. Title-giving thus appears as less important than wife-giving, but it is still part of the transfer of fertility and social identity along affinal hierarchies. In this framework, the link between the healing power of titles and their source in a superior social position becomes meaningful. As gifts from wife-givers, titles convey life-giving completeness to persons and households, just as wife-givers create new households by marriage (Sprenger Reference Sprenger2006d, Reference Sprenger, Lorrillard and Goudineau2008).

After the ritual, the new titleholder is announced to the neighboring villages. In the past, this was done by cutting a Piaster into as many pieces as there were neighbors and sending one piece to each village. Today, letters written by the headman and accompanied by a few thousand kip (a Piaster today is worth 50,000 kip) fill the same function. Note that the announcements refer to a nonhierarchical social order: it is spatial proximity, not a relation of superior center to periphery that determines who is being notified.