Browse the book with Google Preview »
This work is a case study of the Wanano, a group defined by a distinct language, name, and ancestral history. I was initially drawn to the region of the Uaupés River basin in the Northwest Amazon by reading Irving Goldman's 1939 ethnography of the Cubeo, neighbors and in-laws of the Wanano. Good ethnographies raise thoughtful questions and Goldman's raised many. Among them was the character of a social ranking system not found elsewhere in Amazonia, a phenomenon unexplained in the literature on the region.
Like other Eastern Tukanoan-speaking peoples of the area, the Wanano think of themselves as a single descent group organized through common putative ancestry. The Wanano are internally subdivided into localized descent groups, known to ethnologists as patri-clans or sibs. A sense of siblingship unites members of the language group at every level. An emphasis on seniority, on the other hand, differentiates them: every individual within a sib, and every sib within a language group, stands in a fixed rank relation to every other. The order of seniority is reproduced through the inheritance of names in each succeeding generation. Rank, a scalar, sequential ordering of seniority, governs the interrelationships of subgroups within the whole.
My interest in the phenomenon of ranking was heightened by an incident that occurred early in my fieldwork. On arriving at the Wanano village of Wapu in 1978, I began to elicit terms of address in order to make sense of the relations among the 160 people in the village. I assumed that the terms themselves, such as "my father," would be neutral and public, even as they provided a key to the familial relationships of the members of the village. I had imagined this to be a sensible course and that my inquiry would be regarded as courteous and nonintrusive.
In many small-scale societies people are addressed by kinship terms, and this is true for the Wanano. The terms are ordinarily reciprocal: for example, if you call someone "my father," he will call you "my child," and so on. A survey of such terms should produce a symmetrical grid showing complementary pairs of terms for referent and referee or addresser and addressee.
My survey proved revealing, but not for the reasons I had supposed. One entire household was out of kilter with the rest of the village. Each member of this family reported calling everyone else a "younger" relative, which meant that the others should return the address by calling them "older." Instead, the other villagers claimed to call them "younger."
When I began listening to village conversations, however, I found that the unusual family was reporting one thing and doing another. Furthermore, there was a pattern to the discrepancy. In fact, all members of this family were called "younger" by all other villagers and returned the reply with kin terms preceded by "older." But why conceal something so seemingly innocuous as form of address?
I soon learned that address is the form of speech that carries information regarding status position. Individuals in a Wanano village are ranked, with no two individuals occupying the same position. In any dyad, one person is always the senior and the other the junior, indicated in speech by metaphors of relative age. These status positions are based upon inheritance from father to offspring. While thinking that I was making an innocent inquiry, I had unwittingly placed one family in a somewhat humiliating position. I later learned they were from a lower-ranked sib than the other villagers.
My investigation into terms of address eventually led me to discover that the low-ranked sib was in dispute with the other, more highly ranked villagers. Indeed, I later came to understand that the village was wrenched apart by factions and witchcraft accusations, reflective of and expressed through tensions related to rank.
Because I resided with a leading family of one of those factions, I was perceived to be allied with it. Once this became clear to me, I removed myself from Wapu to the refuge of nearby Yapima, the Wanano village where I established bonds of meaningful friendship and the feeling of being at home. My experience during those earliest days in Wapu, nevertheless, served to call my attention to the relevance of the dispute I had witnessed.
The discordances demonstrated by the family from Wapu are indicative not merely of personal emotions, though these are sincerely felt. On another level, they are representations of tensions that are inherent in the society itself. As such, they point to the fracture zones of a social system.
Anthropologists justifiably focus on agreed-upon systems of meaning. Rank is one such system. Yet divergence from standards is of equal importance to our understanding of society. The dissenting position of a sib at odds with higher-ranked villagers, presented in private to an outsider, suggests deeper, albeit covert, dissonance and opposition to an apparently coherent and unified social order.
The emphasis on rank at Wapu and the passions it generated continued to intrigue me. Other Amazonian populations had no such social forms. Rank appeared to be a central principle governing social relations of kin groups and individuals among the Wanano, as well as a locus of contested identities; thus, my initial interest in rank was affirmed and I continued to make it the focus of my inquiry.
This book discusses rank as a major principle that, with intermarriage, organizes the social and economic life of the Wanano. The introduction orients the reader to the Wanano and to the major theoretical issues addressed in the book. In it I briefly outline two theoretical models that attempt to explain ranked society, then review the research of other Uaupés specialists and suggest how my study differs from and complements theirs. The distribution of rank structures and their relative importance in the upriver (Colombian) and downriver (Brazilian) sections of the Uaupés suggests, I believe, fundamental principles and features of rank systems in general.
Chapter 2 places the social system in a historic context of population movements, warfare, and encroaching European expansion, considering the relationship of political consolidation and marriage practice to warfare. This chapter locates politics not in the descent group, but in a different sphere: marriage. Here I argue that the unusual prevailing practice of linguistic exogamy in the Northwest Amazon can best be explained through an analysis of historic factors. I rely upon Wanano oral histories to demonstrate the role of obligatory language group exogamy for alliance formation in a context of neighboring predatory chiefdoms.
Chapter 3 reviews the activities of scientific travelers, missionaries, and state agencies in the Uaupés basin from 1760 to the present. The impact of European expansion upon Uaupés culture is considered in greater detail.
Chapters 4 and 5 are concerned with social organization and kinship, respectively. They describe in depth the two organizing principles of Uaupés social organization—rank, a system of internal subdivisions within exogamous, linguistically distinct patrilineal descent groups; and linguistic exogamy, with its countervailing matrices of social linkages that connect different language groups to each other.
Chapter 6 presents Wanano notions of "placement" and "displacement" and the sentiments attached to them as portrayed in spontaneously composed songs. This chapter conveys the experience of living in such a system, especially as it affects in-marrying women.
Chapter 7 takes up production, emphasizing the commensurability of abundance in the physical environment with the manifestation of "succulence," which indicates social seniority. Here social production and the production of the social are inseparable. This point is further explored in chapter 8, where the exchange system shows the requirements of seniors to give abundantly.
Chapters 9 and 10 reveal the dynamic processes that underlie an apparently static system. Chapter 9 relates the concepts of seniority, production, and exchange by examining the dramatic confrontation between an ineffective chief and villagers who attempt a work stoppage. In chapter 10 a low-ranking sib claims the site and status of a dying senior sib; its one remaining member is forced into an exile which she describes in a song she composes in flight. Here rank is shown to be a social product subject to negotiation and maneuver.
In foregrounding processes of change, these last chapters return the book to the issues of history and structure addressed at the opening. "Contradictions are vital agencies of change and are central to historic processes in all societies.... Indeed, the tension... may be the structure and not a sign of its demise" (Murphy 1986: 42). The insight sheds light on the "unrepresentable" events of history and their role in the construction and reconstruction of society. It also sheds light on the family at Wapu and its attempts to negotiate an ostensibly nonnegotiable social fact.
At a time when anthropology has become increasingly fragmented and specialized, this work is an attempt to return to the traditions of a comprehensive anthropology envisioned by Franz Boas—one that considers within its scope the full spectrum of social life. At the same time, the work attempts to utilize the specialized lexicons and discourse styles that have emerged from a number of subdisciplines. My attempt at multivocality is in response to the idea exemplified by the parable of the blind men describing an elephant: a multiplicity of perspectives is expected to yield greater insight than would any single viewpoint. In order to soften the contrast of discourse styles, I have divided the book into four parts entitled History, Sociology, Ecology and Economy, and Ordinary Dramas. A concluding discussion reworks the material and languages of the book into a summarizing, integrated statement.
A Tukanoan companion and I going to Rio de Janeiro once made a tourist stop at the summer palace of Brazil's first imperial family. The palace was full of the sumptuous display of nineteenth-century royalty, including a throne room. The room was dominated by the imposing presence of the throne—an ornately decorated chair located on a raised platform. My Tukanoan companion turned to me and asked, "Does the emperor kneel before his people or do the people kneel before the emperor?"
For the Wanano and other Eastern Tukanoans who are without figures of comparable authority, there is no assumption that a leader is inherently powerful. My companion's question places the locus of power not in a leader per se, but in a relationship. For this reason, knowing the implications of the kneeling gesture, she wonders which is the more appropriate: who does kneel before whom?
This work is a case study of the Wanano, a Native American society in the Northwest Amazon, unusual in Amazonia in having rigidly bounded descent groups and ascribed social statuses. These amount to, in Wanano terms, vertical and horizontal "placements." The book takes this system of placement or rank as its focus.
Ranked systems, defining status difference, but lacking political centers, are found throughout the world, yet we know surprisingly little about them. Our models for civil society neglect this form, proceeding as they do from models of state development in which centralized political power is seen as essential in the structuring of rank. This is not so in the case of the northwest Amazon.
The Wanano in Uaupés Society
Approximately 1,500 speakers of Wanano, a language of the Eastern Tukanoan family, inhabit the middle Uaupés basin in Colombia and Brazil. The Wanano constitute one unit in a larger intermarrying population of twenty or more language groups who occupy the Uaupés and other headwater streams of the Rio Negro, a region known in the literature as the Northwest Amazon.
For the Wanano and linguistically related populations, language is considered a manifestation of descent, with speakers of the same language thought to be members of one patrilineal descent group within which marriage is prohibited.
As a result of the widespread practice of linguistic exogamy, approximately 14,000 speakers of diverse and sometimes distant language groups, inhabiting some 150,000 km2 in adjacent areas of Colombia, Venezuela, and Brazil, are related either by kin or in-law ties. The result is a uniquely coherent culture complex, which has patrilineal descent and cross-cousin marriage as major integrating structural principles.
On the basis of established literary convention (Jackson 1974, 1983) and the Wanano emphasis on language as the primary feature of personal and group identity, I refer to the Wanano as a "language group," a shorthand for the more precise, but longer, "linguisticdescent group."
Researchers (Jackson 1974; Sorensen 1967) have called attention to the remarkable degree of multilingualism and language group exogamy in the Northwest Amazon. The linguist Arthur Sorensen (1967) presents us with a useful classification of Eastern Tukanoan languages, identifying thirteen languages as members of the Eastern Tukanoan language family: Tukano, Tuyuka, Yurutí, Paneroa, Eduria, Carapana, Tatuyo, Barasana, Piratapuya, Wanano, Desana, Siriano, and Cubeo. Sorensen suggests that the Eastern Tukanoan languages are less closely related to each other than are languages of the Romance group.
Sorensen estimates at 10,000 the population of speakers of Eastern Tukanoan languages. The total intermarrying population of the Northwest Amazon increases to approximately 14,000 when non-Tukanoan speakers are considered. The cluster of Eastern Tukanoan societies is bounded on the north, south, and northeast by Arawakan speakers and to the west by Cariban speakers. These neighboring groups enter the Eastern Tukanoan system of extratribal marriage. Two Arawakan-speaking groups, the Tariana and the Baniwa, have intermarried with the Wanano over generations.
The Wanano regard four other language groups as agnatic kin and extend marriage prohibitions to them: the Piratapuya, Arapaço, Siriano, and Tuyuka (see fig. 1). Theoretically all other groups are marriageable, but some lack the established alliances that would make them ongoing in-laws. From the perspective of the Wanano and all other members of their multilanguage phratry, the world in this way divides into two complementary units: "brothers" and "marriageables."
Wanano both maintain ongoing affinity with other language groups over generations and forge new marriage alliances where ongoing affinity has not been established. In such a system new populations are easily accommodated as in-laws.
The Wanano call themselves Kotiria and are called Okotikana by the Tukano, Okodyiwa by the Cubeo, and Panumapa by the Tariana. The group, however, is most commonly known in Brazil as the Uanano or Uanana and in Colombia as the Guanano or Guanana; this naming follows the historic convention of translating the native name into língua geral. The spelling "Wanano" is perhaps the clearest transcription of the língua geral pronunciation and the one most consistent with established international norms.
Brazil's ten Wanano settlements are situated 3 to 24 km apart along a continuous stretch of the middle course of the main river. Each settlement contains from 30 to 160 persons; I estimate the total Wanano population in Brazil to number about 500 to 600. When we add the approximately 180 Wanano who live on the southern, Colombian bank of the Uaupés and the 800 Wanano cited by Nathan Waltz (1976: iii) as living in the Colombian Vaupés Territory, the Wanano population totals some 1,500 to 1,600 individuals.
Villages consist of mud or bark houses with thatch roofs located on high ground along the river edge. Houses are situated in a rectangle around a common plaza parallel to, and in view of, the river. Paths lead from a canoe landing to the residential area. Each house has a front entry onto the plaza and a rear door that opens onto paths that proceed through the surrounding forest to gardens and streams.
Wanano subsist largely on root-crop cultivation and fishing. Fish provides the principal source of protein, and manioc the principal source of carbohydrates. These items, and the utensils used to gather or process them, are essential to the sharing of resources, which occurs informally within a settlement on a daily basis and more intermittently and formally among different local settlements. Minimal exploitation of resources characterizes day-to-day life; intensive exploitation occurs prior to the occasional elaborate exchange ceremonies described in later chapters.
The Uaupés River at the center of Wanano life is a black headwater stream of the Rio Negro, a principal tributary of the Amazon. The Uaupés rises in the rainforests of eastern Colombia and follows a southeastward course through the villages of the Wanano into Brazil, where it enters the Rio Negro near São Gabriel da Cachoeira.
Although it drains a section of neotropical rainforest noted for its especially high rainfall, the forest cover here is sparse, relative to rainforests elsewhere. The impoverished soils through which the Uaupés flows give rise to forests and rivers of low nutrient content and relatedly low plant and animal life. Against this depauperate backdrop, the richness of cultural life stands in sharp contrast.
Descent Group Organization
The social organizations of the Wanano and other Indians inhabiting Brazil's Uaupés River basin differ from those of other Amazonian societies in that they are internally ordered into a series of scaled, named subunits proceeding from most senior, or first, to least senior, or last.
An origin myth shared by Eastern Tukanoan speakers tells of a sacred anaconda-canoe that journeys upriver from a primordial Water Door and swims underwater to the region of the Uaupés River. Reaching the headwaters, the anaconda-canoe turns around so that its head faces downriver and its tail upriver. It then slowly rises, and from the segmentations of its body emerge the first ancestors of each of the patrilineal kin groups of the Uaupés. Those who emerge from the head of the anaconda are designated as seniors, the "heads" of the local patrilineal descent groups or sibs. Analogously, sibs that emerge upriver are known as "tails."
The anaconda of the Northwest Amazon differs from other serpentine motifs, such as the ouroboros portrayed in Mediterranean iconography as a snake with its tail in its mouth, precisely in the directionality of the Amazonian usage. Positioning is crucial, for the birth order of sibs from the body of the ancestor becomes an order of status fixing the relations between distinct subsections of language groups.
The anaconda qua river, then, is an image that is used to represent the creation of a sociotopographical order and the emplacement of peoples. As a primary metonymy, the river may be said to be a birth canal, a "corridor" of ancestors, a boundary that separates differences and connects them. It is also a source of sustenance and social distinction.
Although deep genealogies are absent, the language group is conceptualized by its members as a group of agnates who trace descent from a set of ancestral founding brothers who emerged from the primordial anaconda canoe. Each of the twenty-five Wanano sib subdivisions recognizes one of the founding siblings as its focal ancestor. As the ancestral siblings are ranked according to seniority specified in the origin traditions of each sib, the entire language group is united in a comprehensive hierarchy.
Social distinctions within the Wanano language group are maintained and reproduced through a naming system in which each Wanano sib owns a set of exclusive ancestral names. The operative rule is that first-born is senior to second-born, and so on down the line; and that the descendants of first-born ancestors are senior to descendants of later-born ancestors. The basis of the terminology is thus inherited seniority, rather than actual age. Rank, then, is a language of social positioning, a linguistic and conceptual placement of individuals and groups in relation to one another.
Rank is manifest on a daily basis in the terms of relative address used by speakers in conversation and greeting. Also, rank statuses are associated with roles and expectations that are especially foregrounded in ritual. High-ranked groups are said to be "succulent" and are expected to manifest this trait through generous display in the sponsorship of large dance ceremonies. Such "succulence" is thought to be an attribute both inherent in status and a product of it.
Descendants of the first of the twenty-five ancestral brothers are called "our eldest brothers" and are known as "chiefs." As descendants of the youngest ancestral brothers are the lowest in rank and are referred to as "younger brothers," "the lasts," or "servers," so the highest-ranked sibs are seen as "donors."
Although ranking produces a hierarchy of individuals, titles, and influence, it entails neither differences in wealth nor coercive controls of labor. Relatedly, the terms "server" and "servant," universally employed in the literature on the northwest Amazon, reference identities and roles that point to service as their prominent feature. Yet these must be understood in the very specific terms of Eastern Tukanoan society. Wealth is never accumulated by individuals or groups, nor service controlled. As a result of practices and values that generate reciprocity at every level of the social order, the ultimate outcome of representational ranking is substantive equality.
The Concept of the Ranked Society
The phenomenon of ranking in societies lacking both political centers and resource accumulation has been of theoretical interest to students of political and social systems for some time.
The concept of the ranked society emerges from centuries of scholarship on the nature of social inequality. Enlightenment scholars refuted the Aristotelian notion of inequality as an inherent, natural phenomenon, advancing in its place a two-stage model of social development, in which an egalitarian phase precedes stratification. Rousseau, Marx, and others believed that the introduction of private property spurs this transformation.
The traditional two-stage model does not account for the many societies reported by anthropologists that are neither egalitarian nor stratified. Although an extensive literature documents these societies (e.g., Sahlins 1958; Goldman 1957; Malinowski 1922; Leach 1954), we do not yet have a body of ethnographic data adequate to shed light on how such societies function or to provide a basis for comparative analysis. Furthermore, the scant available data derive primarily from island societies; few continental ranked societies have been documented.
The social organizations of the Indians inhabiting Brazil's Uaupés River basin, then, represent an exceptional case in the literature of a present-day ranked society inhabiting the continental tropical lowlands.
The operation of such ranked societies appears to be anomalous: by fixating degrees of prestige and identifying positions of authority, rank subverts egalitarianism. Yet, while a ranking system realizes an ideology of fixed statuses and levels, groups and individuals have little actual power and no military capacity to support their positions; it allows for maneuverability. In this sense, rank does not necessarily imply a strictly nonegalitarian, stratified society. Ranked societies, in short, defy categorization as egalitarian or nonegalitarian, for they are at once both and neither.
The anthropological literature contains extensive commentary on the distinctive criteria of ranked society and the roots of its development. One school holds that the economic factor—the redistribution of production—is critical; another stresses the political factor—the supralocal organization or chiefdom that results from warfare. As this work shows, these two major approaches may be complementary rather than contradictory.
In this study I take a different tack, by arguing that the unique feature characterizing ranked societies is a particular type of organization of work not found in egalitarian societies. This position encompasses both political and economic factors. Moreover, the study places the politics of warfare not in the realm of descent group organization but elsewhere: in marriage.
This book addresses not the origins of the rank level of society, but rather its dynamics. It emphasizes the key element of production and demonstrates the embeddedness of rank in contexts that are simultaneously political, economic, and ideological.
In Argonauts of the Western Pacific (1922) Bronislaw Malinowski speaks of rank as a form of social differentiation among the Trobriand Islanders and uses the "insignia of rank and authority" (1922: 52) as a basis for distinguishing them from other Melanesian societies.
Morton Fried uses "rank societies" to distinguish a discrete category of political organization—a level of complexity different from either stratified or egalitarian society. His definition is useful: "The rank society is characterized by having fewer positions of valued status than individuals capable of handling them. Furthermore, most rank societies have a fixed number of such positions, neither expanding them nor diminishing them with fluctuations in the population, save as totally new segmented units originate... or disappear..." (1960: 717).
In Sahlins' (1958) typology of stratification based upon degree, he speaks of "ranking systems" (1958: 2) to describe various types of stratification. In the same work, Sahlins identifies redistribution as the basis of chiefdoms and posits the adaptability of hierarchy in the equitable distribution of resources. His early positions have formed the basis for a continuing debate on the issue of rank.
Elman Service (1962) follows Sahlins in focusing on redistributive aspects of ranked societies. Using the term "chiefdom" to designate a society both denser than a tribe and organized through coordinating centers of economic, social, and religious activities, Service attributes the rise of chiefdoms to "a total environmental situation which was selective for specialization in production and redistribution of produce from a controlling center. The resulting organic basis of social integration made possible a more integrated society, and the increased efficiency in production and distribution made possible a denser society" (1962: 133-134).
Distinguishing political from economic factors, Fried faults Service for confusing the product of ranking (i.e., redistribution) with its cause. For Fried, the regularity of the redistributive role "conveys prestige and bolsters political status" (1967: 118). For him, "Rank has no necessary connection to economic status in any of its forms, though it does acquire economic significance. The point is that rank can and in some instances does exist totally independent of the economic order" (1967: 52).
Fried does not minimize the importance of redistribution to rank systems, but emphasizes intrasocietal conflict rather than integration as the more powerful explanation. While Fried, Sahlins, and Service all stress the key role of a redistribution center in ranked society, the three differ on how economic factors create, or reflect, the power structure.
Anthropologists have not unanimously viewed redistribution as the salient feature organizing ranked societies. Timothy Earle (1977), Christopher Peebles and Susan Kus (1977), and Robert Carneiro (1981, 1991) question this emphasis. Carneiro departs from the views of Sahlins and Service in the following way: "Service appears to have been strongly influenced by Sahlins' work on Polynesia, and like him, saw chiefdoms as having an economic rather than political basis.... Like Sahlins, Service saw chiefdoms as essentially economic in origin and function. He failed to perceive their basically political nature" (1981: 43).
Carneiro rejects the category "rank" and refers to chiefdoms," since in his view a more crucial factor is a level of political organization in which numerous local communities can be amalgamated under a single leader. He believes that hierarchy emerges in response to warfare, particularly warfare that results from population density and consequent circumscription (Carneiro 1970, 1987). Warfare in this way becomes the only mechanism that has the power to transcend the political autonomy of settlements.
Carneiro defines a chiefdom as "an autonomous political unit comprising a number of communities under the permanent control of a paramount chief" (1991: 168). By such reckoning, the Wanano would be intermediate between an egalitarian system and a chief - dom since they show more status differentiation than egalitarian societies but less political consolidation than chiefdoms. Societies such as this are neither fish nor fowl for they are not chiefdoms, but neither are they egalitarian bands or tribes.
Review of Rank in the Literature on the Uaupés Region
When the data reported in this study are compared to other research from the Uaupés area, a number of important patterns emerge. First, my data stress rank and horizontal linkages that cross-cut descent, whereas earlier works on the Uaupés, based upon Colombian cases, have emphasized segmentary descent and egalitarian principles. For example, Jean Jackson's rich ethnographic work on the Bard of the upper Colombian Uaupés (1972, 1983) devotes relatively little attention to rank. It appears, however, that many features of rank described for Colombian, upriver groups correspond to ranking among the Wanano. Jackson reports, for example, that a sib's rank is based upon a model of birth order and that lower-ranked sibs are considered to be recent: "Another way to indicate a sib's low rank is to say during a recital of the sib's origin myth that the sib's ancestor did not in fact come from. . . 'breast-milk river'. . . but was picked up later. . ." (1972: 68). Jackson finds, moreover, that "at times the low-ranking sibs of a given Tukanoan group will be described as being servants to the higher-ranking sibs—hunting for game and carrying out other tasks in a manner parallel to Makú activities" (Jackson 1983: 152). Bard informants state that low-ranking sibs were at one time savage, ("Makú-like"), and that they speak "ungrammatically" (1972: 67, 68).
For the Bard, seniority is a feature of terminology only in sibling terms (Jackson 1972, 1977, 1983). In this way the Bard differ from the Wanano and other Uaupés groups. Jackson mentions that among the neighboring Tuyuka, with a greater number of sibs, "a person [shows] ... that he is of a much higher ranked sib than the person he is addressing by calling that person grandfather and being called grandchild" (1976: 68).
Jackson nevertheless describes Uaupés society as "egalitarian," without "a significant amount of exclusive control of scarce resources by different social units" (1976: 68; 1983: 212). She also states that "no discernible social units occupy distinct economic niches or locate their settlements at particular strategic points in the landscape" (1976: 68). In the same article, she mentions "well-established rules of hospitality," but remarks that "excessive demands are rarely made on the hosts" (1976: 69). New data suggest that these are characteristic distinctions between upriver and downriver groups.
Two exceptions to the generalization that Uaupés societies are egalitarian are recent studies by Irving Goldman and Christine Hugh-Jones. Together with my own work, these exceptions suggest an explanation for the apparently different emphases the Bard and Wanano place on rank.
Irving Goldman's early ethnography, The Cubeo (1963), was based upon fieldwork carried out in 1939 among a low-ranking Cubeo sib called the Bahukiwa. This ethnography, the first in-depth account of a Northwest Amazon society, describes the distinctive Cubeo culture, while identifying elements shared with the larger cultural complex of the Northwest Amazon.
In 1981 Goldman revisited the Cubeo to work among high-ranked groups. Goldman's later analysis of the Cubeo (1981), based on this fieldwork, lays greater stress on rank than did his previous reports. These findings resulted in several publications on the key symbols and metaphors of Cubeo hierarchy.
In Goldman's later work he discusses the symbolic bases of "the linear order of hierarchy," conceived of by the Cubeo as "an organic and spatial concept pertaining to the body of the anaconda and the structure of the river" (1981: 10). He draws attention to the grandparent-grandchild dyad: "In this dyad there is reciprocity: the nurturing grandparents gain their personal immortality by giving their names, their specific spiritual attributes, to their grandchildren. They live again through them. But since these are also chiefs and servants to each other ... [it] is clearly ambivalent in its intimations for social status" (1981: 16). "Chiefs and servants are bound within the special metaphysical dyad of grandchild/grandparent with its functions of nurtured and nurturer" (1981: 21). Goldman's description of these complementary relationships is derived from informant "recollection" of a former period, when chiefly groups reportedly were a leisure class that neither cultivated nor foraged, tended to by their servants. Goldman suspects, as I do, that this description is distorted and "imbue[d] ... with some of the qualities of present relations with Colombian patrones" (1981: 21).
Goldman now characterizes the Uaupés system as an "elementary hereditary aristocracy" (1981: 1). His study reveals the symbolic analogies underlying the relationship of chief to servant, or grandchild to grandparent, as it obtained in an earlier, hypothesized period. Both of these reconstructions provide strong evidence for horizontal structures in the Colombian Uaupés, comparable, though not identical, to organizations I was able to observe in the Brazilian Uaupés.
The analyses of Stephen and Christine Hugh-Jones, based upon fieldwork among the Barasana of the Pirá-Paraná River in Colombia between 1968 and 1969, form a unit that integrates myth, ritual, and everyday reality of the Northwest Amazon. Christine Hugh-Jones (1979) posits a hierarchically organized system of five interdependent sibs, each having a specialized role. Of the five specialized roles-chief, dancer/chanter, warrior, shaman, and servant-she reports that only the dancer/chanter and shaman remain well defined; the chief, warrior, and servant roles having been lost with time. She locates all five roles in a symmetrical model of three domains: the politico-economic, the metaphysical, and the externally oriented domain.
In this model of five functionally interdependent, specialized sibs, Hugh-Jones finds sibs associated with two of these roles (shaman and chanter) and postulates the "missing roles of chiefs, warriors, and servants" (1979: 54) on the basis of informants' recollections. The Wanano provide evidence for the existence of sibs in the "chief" and "servant" categories, but suggest a pattern of localized interdependent pairs as opposed to local complements of five.
What accounts for the variation in different reports? Does it suggest patterning? To what variables can we link "more rank" and "less rank," and what does this reveal with regard to rank structures in general? The Wanano case, in its complexity, provides a fruitful setting for approaching these questions.
A number of significant patterns differentiate upriver and downriver groups. The most apparent is size. Christine Hugh-Jones gives population figures for thirteen upriver settlements, which range from four persons to twenty-nine persons (1979: 42); the average community size is fourteen. Irving Goldman (1963: 25) reports twenty-nine residence sites for the Cuduiari, "each averaging some 30 to 35 people." Jean Jackson (1976: 68) describes nucleated villages of "one to four small houses" separated from one another by two to ten hours of canoe travel. In the longhouse settlements of the upper Papuri, the area in which she worked, she describes "fewer than thirty individuals" (1976: 73) per settlement.
In contrast, downriver sites are large. The largest Wanano settlement held 160 persons between 1978 and 1981 and was located less than an hour's walking distance from a village of 82 persons. Koch-Grünberg (1909) reported a population of 200 for the same settlement in 1902. The population of the settlement in which I was based was 77. In short, the average size of downriver settlements is three to four times the size of upriver settlements.
Resource availability is another factor distinguishing upriver and downriver settlements. Downriver locations are the more strategic in terms of environmental and technologically related factors. Due to the dendritic pattern of rivers, river size and fish supply are related. With each bifurcation of the river, the fish population theoretically divides in corresponding proportions. Furthermore, spawning fish migrating upstream, and their pursuing predators, arrive at the downstream sites first. With fish-fence technology, a downstream group may block fishes from proceeding upstream.
Researchers from the Colombian Uaupés have noted this distinction in riverine resources. According to Christine Hugh-Jones, "Towards the headwaters the availability of fish gradually declines . . . those who live in the headwaters are described as 'eaters of tiny fish.' The headwater people are forced to rely more on forest products, not necessarily because these are more abundant, but because the fish are scarce" (1979: 239).
Hugh-Jones offers a further distinction: she associates downstream sites with "political dominance," which she attributes to greater involvement in ceremonial exchange. This, in turn, she explains through the "superior cultivation" of downriver settlements (1979: 239). While Hugh-Jones accurately notes the connections among political relations, exchange rituals, and a surplus of the garden products manioc and coca, she overestimates the differences in the quality of cultivable land and overlooks a more critical variable: the capacity to attract labor. Agricultural land is equally or more available to upriver settlements, which are smaller in size, than to larger, more permanent downriver settlements. The ability to mobilize the labor to create an abundance of labor-intensive agricultural products increases a group's capacity for participation in exchange rituals.
The spatial distribution of ranked groups places high-ranked groups downriver and low-ranked groups upriver. The Brazilian Uaupés (downriver) is inhabited by the highest-ranked sibs and their accompanying servant sibs. My own focus reflects the emphasis placed on rank by the chiefly sibs with which I worked. The deemphasis on rank in studies of the Colombian (upper) Uaupés may be attributed to the absence of high-ranked groups in that area. It appears likely that rank structures will be preserved where differences in rank exist and privileges relating to rank are at stake. This observation at once supports and explains Christine Hugh-Jones' "missing" chiefly and servant sibs: they are not located upriver. Furthermore, the Wanano data suggest that the chiefly groups are corporate, or at least "fixed," in terms of riverine territory.
The Dynamics of Rank
Although this study focuses on observations derived from one ranked society, its main purpose is to make observations that are generalizable to other ranked societies as well. For example, a characteristic noted among rank societies elsewhere is geographic or cyclical variation, to the extent that seemingly different political models are found within a single cultural entity. The Northwest Amazon typifies this type of society but exhibits "hierarchical zoning": upriver groups maintain an ideology of hierarchy with few manifestations of it; downriver groups, however, are immersed in the day-today practice of ranking.
Gaps in our understanding of the Northwest Amazon as a "ranked" system have stemmed in part from the fact that each individual study has treated its group as a closed system, without fully taking into account the interdependencies and fluxes characterizing the larger socioeconomic whole. Furthermore, previous Uaupés studies have focused only on upriver settlements, which demonstrate, but do not emphasize, ranking. The earlier reports are by no means anomalous or incorrect. They are, however, insufficient. Without a consideration of the downriver chiefly groups, a comprehensive understanding of Uaupés political, economic, and social systems cannot be achieved. This work is the first to be based upon data gathered in downriver (eastern) settlements. Furthermore, the study is aimed at the interactions of widely disparate ranked persons and groups. Specifically, it considers in some detail chiefly sibs and servant sibs living in symbiotic relation to one another. In summary, I find that the weak emphasis on rank upriver—and its strong emphasis downriver—suggests a holistic model for Uaupés society that encompasses patterns of variation and enables us to offer general statements about such societies.