Ecological Need and Cultural Choice in Central Brazil

What specialists on lowland South American Indians call the Upper Xingú, or simply Xingú, is the basin in which the Ronuro, Batovi, and Kuluene Rivers join to form the Xingú River. The term “Xingúanos” refers to the nine Lydian villages located in the area.

The Geographical environment of the Xingú is characterized by a very marked separation between the dry and the rainy seasons. The region is transitional between the Amazon and the dry savannah of Central Brazil that is called the cerrado. Amazonian flora is represented in the forests, but the specimens are of smaller size. The open fields that separate the forests have a flora and fauna similar to that typical of Central Brazil.

Despite their speaking different languages, Upper Xingú tribes present an amazing degree of “General cultural homoge­neity”. Galvão (1960) classified them together as a single cul­tural area. The following tribal groups form the core of this area: the Tupi-speaking Await and Kamaiurá, the Arawakan­ speaking Waura, Mehinaku, and Iawalapiti, the Caribbean-speak­ing Kuikuro, Kalapalo, and Mantipú, and the isolated Truman. Two tribes, the Taxicão and the Suvá, are culturally peripheral to this core but have been strongly influenced by Xingú culture.

Xingú villages and houses are similar in their physical struc­ture Xinguanos share an economy based on fishing and slash- and-burn agriculture, a kinship system characterized by bilater­al descent, cross-cousin marriage, and Iroquois cousin terminol­ogy, and ritual and religious traditions. Village size varies from 50 to 150 people. Devices for intertribal integration include ritualized trade, intertribal rituals, and marriaGes. Intertribal relationships are usually peaceful.

The term “Gê” refers to a linguistic branch. Gê-speaking Indians have occupied a rather extensive area from the southern boundaries of the Amazon to the extreme south of Brazil. As in the Xingú, Central Brazilian Gê habitats present a very well-defined distinction between the dry and the rainy seasons. The Gê occupy the cerrado or savannah. Some Gê nowadays live in a forest environment where they were pushed by the expanding Brazilian frontier more than a century ago.

Central Brazilian Gê groups have been considered a cultural unit by several authors. A classic treatment is found in Lévi-Strauss (1958). Galvão (1960) places them at the core of a cul­tural area formed by three main nuclei. The first nucleus in­cludes the Timbira (Kanela, Apinayé, Krahó, Gavião, and Krikati), the second the Akwé (Xavante and Xerente), and the third the Kayapó (Gorotire, Txukahamãi, Kuben-Kran-Kegn, Kubén-Kragnotire, Diore, and Xikrin). A group in several ways similar to the Gê, the Boróro, is included in the area. Some non-Gê-speaking tribes, such as the Parakanã and Tapirapé, are also included, but are culturally peripheral to the three main Gê nuclei.

Central Brazilian Gê all have the same type of village and house. Village populations may reach several hundred, and there are historical reports of populations numbering in the thousands. There is, however, a strong tendency towards fission during part of the year. The Gê economy is based on agricul­ture in the gallery forests, hunting, and gathering. Social struc­ture is characterized by cross-cutting moieties, age-classes, and, in some cases, descent groups. Kinship terminological types vary. All Central Brazilian Gê are matrilocal.

I shall discuss some traditional ideas in ecological thought and suggest some new questions by comparing Gê and Xinguano groups.

A current theme in ecological thought relates the level of so­ciocultural complexity to the food supply in a given natural en­vironment. Steward and Faron (1959) carry this to an extreme in their discussion of the “family level of sociocultural integra­tion.” The main ingredients of the recipe are a small population, nomadism, a hunting-and-gathering technology, and a natural environment with a relatively low food supply. Carneiro (1968) emphasizes the importance of population growth in a limited geo­graphic area (independent of soil fertility) to the development of higher levels of sociocultural complexity. In the Xingú case (Carneiro 1968) and probably in the Gê case (Bamberger 1967), population pressure is absent; both Gê and Xingú traditional homelands could support much larger populations. Therefore the differences between the two in cultural organization must be understood on some other basis. The suggestion of other factors to explain these differences is one goal of this paper.

An important advance in cultural ecology took place when protein capture was introduced into the explanatory nexus of South American ethnology. However, this approach persists in relating the “poverty” of the environment (caused by low pro­tein availability) to a less complex organizational form The data presented here show that, in some situations, the opposite occurs. In areas where the natural environment is less easy to deal with, adaptation requires specialized devices, which suggest greater sociocultural complexity.

The criteria usually used for determining level of cultural complexity are those seen in Meggers’s (1977) study of the Amazon. They range from the size and permanence of villages to the presence of a more elaborate ceremonial life. Village size and permanence are considered the most important; the larger and more stable the villages, the higher the supposed level of cultural complexity. The value of these indicators has been suspect, however, since Carneiro (1960) established, in his study of the Kikuyu, that some tropical-forest groups, tradi­tionally supposed to be not very complex, have had more stable settlements than was first thought.

Oberg (1973) posits population increase as the fundamental factor in any analysis of cultural complexity, adding that this factor is affected by soil fertility and rainfall. According to his c1assification, two types of Brazilian Indian groups exist: homogeneous tribes with bilateral descent and corporateness at the local level and segmented tribes formed by several unilateral descent (corporate) groups. The latter type of social organiza­tion is correlated with larger population size and greater eco­nomic surplus. Although Oberg’s typology is an important step towards understanding the differences among South American Indian groups, he does not establish -and it does not seem that he intends to establish – a c1ear link between social structure and particular environmental conditions. The Xinguano and Gê tribes can be viewed in the light of this discussion. The Xinguanos are homogeneous tribes with bilateral descent and corporateness at the local level, while the Gê are segmented tribes consisting of several groups. However, this type of defini­tion is incomplete. For example, Xinguano social structure presents sexual opposition as the basic organizational criterion, while among the Gê such principles as age opposition are also important adaptive mechanisms. The use of such fundamental organizational principles and their relation to the natural en­vironment will be discussed in this paper to provide a new basis for comparing South American Indian groups.


Murphy’s (1960) study of the Mundurucú established the rele­vance of sexual antagonism for the internal organization of some South American Indian groups. My work among the Await has shown how the system of sexual antagonism, logi­cally expressed as an oppositional system, relates to the con­crete life of the Xinguanos (Zarur 1975). Sexual opposition has been seen as the basic organizational principle in the social structure of these groups, with generational opposition as a secondary principle.

It becomes apparent, when comparing the division of labor among Xinguanos and Gê, that the rules of sexual division of labor are more carefully observed among the former than among the latter. In both cultures, the men tend to do the hunting and fishing while the women gather food in the forest and harvest and process the crops. In Gê villages, however, one can see men as well as women carrying children; this is un­heard of in Xinguano villages. Possibly, the more flexible divi­sion of labor among the Gê groups is a consequence of inter­ethnic contact. Even if this relationship were established, how­ever, other differences arise when the analysis is taken to the ritual level. For example, Melatti (1975) has suggested that the expression of ritual sexual opposition does not have the im­portance among the Krahó (a Gê group) that it has among Xinguanos.

Other evidence of the organizational difference between Gê and Xinguano is the institution of the men’s house. In the Xingú, the men’s house belongs to all males of the village, re­gardless of age, while among the Gê it is usually reserved for un­married med For example, the Timbira have an approximation of the men’s house; bachelors sleep in the village plaza in the dry season and during the rainy season live in their own “bache­lor’s house.” Among the Xavante and the Kayapó, there are also men’s houses that, significantly, are referred to in the literature as “bachelor’s houses” (d. Maybury-Lewis 1974, Vidal 1977). Thus, the men’s house among the Gê does not have the same connotation as it does among the Xinguanos-the Gê male, in associating with this house, is not put in opposition to the Gê female. Further, among the Gê, the men’s house is not associated with rituals of strict sexual opposition as it is among the Xinguanos. For instance, among the Xinguanos, the Aweti’s men’s house gives its name, karytu, to the most important of their rituals. This contrasts with the Gê situation, in which the men’s house is associated only with a particular male age-class and with rituals focussing on this class. The violent sanctions against Xinguano women who enter the men’s house (gang rape) do not exist among the Gê.

There is reason to believe that sexual opposition is more clearly expressed in the kinship terminology of the Xingú (Zarur 1975), while the superposition of age categories is re­flected in the terminology among the Gê (Crocker nd., da Matta, 1976). Sexual opposition occurs among the Gê (Bamberger 1967), but it is found in conjunction with a strong age principle that is lacking among Xinguanos. The reason for this difference stems from the relevance of spatial mobility for the Gê and its lack of relevance for the Xinguanos.


In the model of the “Marginal” groups of South America de­veloped by Cooper (1942) and Steward and Faron (1959), the Gê of Central Brazil are categorized as small groups charac­terized by the “hunter’s pattern.” The interaction of a “simple” technology with a “poor” environment forced these tribes to break up into bands during that period of the year when the availability of food did not allow for a greater concentration of population. The bands, formed on different criteria, such as descent and postnuptial residence, circulated within the tribal territory, following hunting and gathering opportunities. When the animal population in a given area decreased to a certain level, the band moved to another hunting area, allowing for a return to optimal size of the animal populations in already exploited areas.

Ever since Nimuendajú’s (1946) demonstration of the im­portance of agriculture to the Timbira, it has been recognized that the Gê in general are not classic hunters and gatherers. Many works that followed, for example, Carneiro’s (1968) work on the Xinguano agriculturalist-fishermen and Sahlins’s (1972) work on primitive groups in General, have shown that most primitive economies can be characterized as satisfying the few needs of their members. On the basis of these data and this line of argument, the role of hunting as a source of food and as an economic activity among the Gê of Central Brazil has been minimized. An extreme expression of this point of view can be found in Murdock’s (1968) denial that the Gê are hunt­ers, using the argument that, for them, the hunt is most impor­tant on the ideological level.

The fact remains, however, that the Gê do break into small groups in order to hunt and gather. This behavioral pattern is not necessarily connected to any cyclical absence of food, as the game and agricultural products located close to the villages would be adequate to support the population. Turner (n.d.) has advanced the hypothesis that the pattern of “trekking” (periodic movement of subgroups of the village population) has the function of reinforcing the internal hierarchy or dominance system of the domestic group. Hunting, a traditionally masculine activity, reasserts itself as the dominant mode of produc­tion over gathering and horticulture, traditionally feminine activities. This, in turn, enhances the position of the male head of the domestic group, which includes daughters and sons-in­-law. This relationship between uxorilocality and “trekking” is of great importance for South American ethnology and for an­thropological theory in general, but the role of periodic group movement cannot be reduced to the ordering of relations within the family. Rather, the evidence suggests that group mobility in Gê cultures, though important for reinforcing family hier­archical relations, performs other functions directly related to subsistence and defense.

One General characteristic of almost all Gê tribes is log racing, described by Melatti (1975) and Nimuendajú (1946). The Timbira are divided into ceremonial moieties, one of whose func­tions is to serve as teams in the log race. The logs are carried by individuals from such teams, which compete against each other. As one individual loses speed, another from his team takes the log upon his shoulders to continue the race. The race is run on the periphery of the village, the central plaza being the starting point and the house of a ritually important person the finishing point.

Melatti (1975) explains that the log races represent the as­similation of the outside world by the village. There is no doubt that they are symbols representing the fusion of nature and culture, but it remains to be explained why this particular symbol was chosen and not some other. Melatti (1976:40) re­sponds by showing that the log races are fundamentally impor­tant as physical training exercises for a group that depends on hunting and that is often threatened by raids from warring groups: “The log races could be training for rapid retreats in the race of enemies while heavily burdened by baggage or by the wounded. The ability to run also allows for the more effi­cient exploitation of game in a given area and the more efficient pursuit of enemies in that area. Thus, it allows for subsistence activities in a much wider area than that which surrounds the village.” The races are strongly related to hunting, as the com­position of hunting parties and the division of the catch often follow the same lines as define the racing teams. One question that may be posed is to what degree the intensity of the present Timbira log races reflects the forced sedentarization of the nor­mally mobile Lydian groups due to the encroachment of non ­lndians on their lands.

The speed and endurance of the Timbira have impressed a number of observers. Nimuendajú (1946) referred to it, and Maybury-Lewis (1974:39) has remarked about The Shavante:

“they move at a quick, shuffling pare (which is extremely diffi­cult for an outsider to get used to), faster than a walk but not quite a trot. They can keep this up all day if necessary, inter­spersed with The darting or running after any game that may present itself.” The Krahó and other Gê tribes are probably some of the most mobile of all peoples. Log races as long as 6 km have been measured that the Krahó men covered in 20 min­utes. Other races, for both sexes and a number of age-groups, are approximately 30 km in length and take two and a half hours. When one considers that, in a modern military unit, a forced march of up to 40 km requires a day, the mobility of The Krahó appears fantastic. It has been calculated that a popula­tion like the Krahó (including women, children, and old people) can cover 100 km or more a day. Thus, the value of the daily log races becomes apparent.

Regrettably, in the study of social organization, technology, perhaps the key to understanding many essential aspects of social systems has been largely ignored. Anthropologists and archaeologists have long since established the correlation be­tween ceramics and sedentary life: ceramic objects, which are both heavy and fragile, are of negative value for a mobile population. .

The Gê of Central Brazil do not possess ceramics. Compared with that of more sedentary groups, Gê material culture is composed of only a small number of objects. These objects in General are light and shock-resistant, and approximately 65% of the items in the inventory are made of straw. The straw, primarily made from buriti and other kinds of palms, is easily found in the savannah. Because of the ease with which these objects are produced and the small quantity of goods that is actually necessary, each person is able to produce and discard at will several artifacts. For instance, Krahó baskets take from five minutes to an hour of work to make, depending on the care with which the basket is woven. The feathercrafts of the Krahó are not strongly developed, possibly because their semi­nomadic life-style cannot offer them conditions under which these craft articles can be preserved.

Gê technology is characterized by only a small amount of equipment, the absence of ceramics, slightly developed feather­craft, strongly developed strawcraft, and the ability to discard, and then reproduce later and at will, several material artifacts. The material inventories of these tribes are strong evidence of a mobile existence. In contrast, the artistry of the Xinguanos is rich in relatively sophisticated ceramic, wood, and feather work.


From the evidence just presented, I hope that it is sufficiently clear that Gê culture is organized around the concept of mo­bility. However, the problem set forth by Turner (1977) per­sists. Why was it necessary for this pattern of “trekking” to emerge when a satisfactory supply of food would otherwise be available? Beyond the necessity of reinforcing the hierarchy which exists in the domestic group, an answer can be sought in the adaptation of these groups to their natural environment. Initially, I will examine Gross’s (1975) theory on protein cap­ture and the location of South American Indian villages. It is possible that the Gê occupy an environment that would be un­inhabitable if not for their frequent changes of location. Unless the village population moved regularly, the protein sources might be quickly and irrevocably depleted. Here the protein resources would include fish as well as game.

Several of the Gê environments have only a small supply of fish, especially when compared with the tropical forest, for the savannahs have few rivers plentiful in fish. If the Xavante were to move to the banks of the Araguaia or the Krahó to the banks of the Tocantins, their protein needs could perhaps be fulfilled by riverine adaptation. However, though the tropi­cal forest is much richer in fish than the savannah, there is strong evidence that the savannahs are richer in hunting areas.

The concept of movement in space suggests an explanation for the differences between the Gê and Xinguano cultures. While the former view movement in space as an organizational principle, the latter do not recognize it as such. Possibly, hunt­ers depend more upon a hierarchy determined by age-groups and generational differences, for the division of labor between age-groups becomes functional in organizing hunting activities. Among the Timbira, certain age-groups hunt more actively than others. Men with wives and children hunt more frequently than young bachelors, who, in turn, hunt more frequently than the old. Age-groups here are highly functional for hunting activities because the solidarity ties established between the individuals belonging to the same age-group are indispensable for the success of a hunting trip. Among the Timbira and the Krahó age moieties serve to organize hunting groups, log races, and the distribution of the game resulting from the hunts.

Another possible explanation for the mobility of the Gê groups as well as for the importance of age-groups and generational oppositions is the evolving sociopolitical environment. A culture that has both a stratified age-system and a hunting economy is automatically endowed with an efficient and aggres­sive military organization. Since the Gê tribes of the savannah shared the same aggressive’ organization, raids among them were common. Beyond this, the villages of the savannah are in general more exposed than those in the tropical forest.

The process of interethnic contact with its initial hostile phases reinforces the original pattern of mobility. As late as the 19405, disciplinary expeditions were still being leveled against the Krahó for hunting (cattle) on non-Indian lands. Military contingencies reinforce the importance of rapid at­tacks and retreats, and the log races help prepare Gê popula­tions to race these problems. To a group of Indians running across the savannah, carrying a 100-kg log presents much the same difficulties as would carrying a cow.

The original hypothesis of Cooper and Steward can be re­stated on the basis of the above discussion: Gê mobility pat­terns are less a result of a food-scarce environment (and simple technology) than a complex adaptation to protein supply, the nature of the sociopolitical system, and military necessities. These latter factors also illuminate the importance of the super­position of age opposition upon sexual opposition.


In the Xingú area, ecological factors necessitating such cul­tural adaptations as those just described are absent. The Xinguanos have chosen one major source of protein and simply ig­nore other sources. They are fishermen, and unlike the Gê, who have an ideology of the sensual pleasure that derives from meat, they hunt only small birds and monkeys. Yet, close to the Xinguano villages there is always an abundance of un­exploited game. In the forests can be found a large variety of animals, including tapirs, capybaras, deer, pacas, and others.

At a distance of about an hour by foot from the Aweti village, I found a vast field in which hundreds of deer were grazing. The Xinguanos consider the meat from these animals repug­nant. Such animals are considered to be “Indians who have gone into the forest and turned into animals,” and thus the use of their meat as food would be a type of cannibalism. The Xingú culture has chosen to exploit another source of protein. The area in which they live represents a transition between the savannahs of Central Brazil and the tropical forests of the Amazon region. Protein sources of the two environments are found in equal distribution. The area is covered by a web of canals and rivers in which fish are plentiful. Thus, considering the total Xinguano population, which is low in relation to the food-production ceiling (Carneiro 1968), we find that excluding hunting from production activities is a viable alternative.

War pressures never were as strong among the Xinguanos as they were among the Gê. Because of a series of historical and geographic factors, the Xingú area can be seen as a refuge area into which small groups from different tribes and linguistic groups were pushed by advancing Brazilian populations. The Gê adaptation to a more specialized environment, that of the savannah, has required cultural values that idealize the open fields of the area. The aggressive Gê are at a nice distance from the Xinguanos and, at the same time, have built a strong de­fense system against the incursion of other tribes and non­ Indians. In contrast, the intertribal system of Xinguano groups possesses, to a high degree, the ability to deal peacefully with new groups. This has developed over centuries of peaceful ac­commodation to foreign intrusion. For instance, the Suyá and the Trumai have been assimilated with relative ease into the region. Traditional warfare in the Xingú does not imply the massacres that characterize the intra and intertribal Gê con­flicts. For the Xinguanos, warfare is not much different from a violent sport. The capacity to defuse conflict manifests itself on the individual level and has led researchers to comment upon the “diplomatic” nature of the Indians of the Xingú (cf. Galvão 1950). The absence of any stronger hostility in the Xingú may be related to the absence of internal subdivisions of the village such as age-classes or descent groups. The poten­tial to organize rapidly for war does not exist in the Xingú as it does among the Gê. The Xingú’s more flexible system, char­acterized by sexual opposition and bilateral kinship, is not structurally compatible with these features. Thus, the Xinguanos, rather than the Gê, possess the simpler social structure, in spite of their living in an environment richer in agricultural productivity and protein potential (fig. 1).



Ross (1978) has discussed Bamberger’s comparison between the Kayapó and the Xingú Indians. Given the similarity of the Upper Xingú and some of the Kayapó environments, Bamberger claims that the adaptive systems of the two cultural units depend upon their world views. Ross tries to demonstrate that the Upper Xingú environment differs from the Kayapó’s and that Xinguanos fish because it is the most productive activity.

Bamberger’s interpretation of the Upper Xingú seems quite correct. The absence of environmental pressure gives the Xinguanos the capacity to organize along much more environ­mentally independent lines than many other Indian groups. She does not emphasize, however, that the Kayapó carne from a clearly defined savannah zone more than a century ago. In a different way from the savannah Gê, the Kayapó, although lacking canoes, use riverine resources with a great deal of sophistication. After a century in this new environment, they have generated new adaptive processes. The same is true of the Boróro, a Gê-related group. They use canoes, have an elaborate culinary tradition based on fish, and use many water motifs in their mythology. Thus the intensity of periodic changes of location and General mobility of the Kayapó and the Boróro express differing degrees of adaptation to the river.

Ross’s evidence is insufficient to show a higher productivity of fishing for the Xingú. In fact, the Kayapó and Xingú en­vironments are not that different, but the Kayapó both fish and hunt, hunting more than fishing.

Other transitional cases like the Kayapó may be found in the intermediate area between the Amazon and Central Brazil and serve as test cases. The Tapirapé, a Tupi-speaking group living near the banks of the Araguaia River, share several char­acteristics with the Gê, although lacking much of their social complexity. The Tapirapé came from a well-defined forest area farther west several centuries age (Baldus 1970). They have age hierarchies, go on trek, and hunt (Wagley 1977), but they also fish, though this is not as important as hunting. Across the river, their neighbors, the Karajá of Bananal Island, choose only to fish.

These transitional cases show that environmental determin­ism operates only when protein sources require specialized de­vices for their capture. This would be the typical savannah-Gê case. When there is more than one protein source available, the culture may choose one of them (Xinguanos and Karajá) or all of them (Kayapó, Boróro, and Tapirapé). In a more di­versified environment, the organization of culture will be much more dependent upon its own internal logic than on ecological constraints.


The absence of environmental pressures requiring specializa­tion for protein capture in the Xingú has permitted a simplifica­tion of the social system of these indigenous groups. This simpler system, defined by the opposition of the sexes, can be viewed as a basic form of group organization among lowland South American Indians. More complex cultural forms, such as those of the Gê, may be understood as complications of these elementary structures by the addition of other criteria of or­ganization, such as age stratification or the formation of de­scent groups.

To verify this hypothesis, the ethnographic sample will have to be enlarged. With the inclusion of other indigenous groups, I believe we will arrive at conclusions with important implica­tions for South American ethnology:

1. Theories concerning the rationale for different degrees of complexity among indigenous groups, especially those that correlate geographic potential and food production, must be carefully scrutinized.

2. The main reason for different levels of sociocultural com­plexity stems from the necessity to specialize for protein cap­ture and for warfare. The more diverse and plentiful the re­sources, the less complex and specialized the cultural institu­tions need be. It follows that the less diverse the ecological resources, the more complex the culture.

3. A new framework based on such organizational principles as age and sexual oppositions may be suggested to identify South American Indian social types. This framework takes into account the environmental backgrounds.

4. It can be suggested that population size does not determine the level of complexity of indigenous groups. Rather, popula­tion size is one consequence of the needs of corporate groups to capture protein and to defend their home territory.

References Cited

BALDUS, H. 1970. Tapirapé: Tribo Tupi do Brasil Central. São Paulo: Editora Nacional.

BAMBERGER, JOAN. 1967. Environment and cultural classification: A study of the northern Kayap6. Unpublished Ph.D. dissertation,

Harvard University, Cambridge, Mass.

CARNEIRO, ROBERT. 1968. “Slash-and-burn cultivation among the Kuikurú and its implications for cultural development in the Ama­zon

Basin,” in M an in adaptation. Edited by Yehudi Cohen, pp. 131-44. Chicago: Aldine.

COOPER, J. M. 1942. “The South American Marginal cultures.” Pro­ceedings 0f the 8th American Scientific Congress, Washington, voI. 2,

pp. 147-60.

CROCKER, W. n.d. “Canela age and sex hierarchies, including parallel transmission: An analysis of changing authority structures,” in

Estudos em homenagem a Eduardo Galvão. Brasília: Centro Nacional de Referência Cultural. In press.

DA MATTA, ROBERTO. 1976. Um mundo dividido. Petr6polis: Vozes.

GALVÃO, EDUARDO. 1950. O uso do propulsor entre as tribos do Alto Xingú. Revista do Museu Paulista 4:353-68.

-. 1960. Areas culturais indíGenas do Brasil, 1900-1959. São Paulo.

GROSS, DANIEL. 1975. Protein c1pture and cultural development in the Amazon Basin. American Anthropologist 77:526-48.

LÉVI-STRAUSS, C. 1968. “Les structures élémentaires dans le Bresil Central et Occidental,” in Anthropologie structurale. Paris: Plano

MAYBURy-LEWIS, D. 1974. Akwe-Shavante society. New York: Oxford Universitv Press.

MEGGERS, BETTY. 1977. Amazônia, a ilusão de um paraiso. Rio de Janeiro: Civilização Brasileira.

MELATTI, J. C. 1975. Ritos de uma tribo Timbira. MS. –.1976. Corrida de taras. Atualidade IndíGena 1:38-46.

MURDOCK, G. P. 1968. “The current status of world hunting and gathering p_nples,” in Man the Imnter. Edited by R. B. Lee and I. DeVore.

Chicago: Aldine.

MURPHY, R. 1960. Headlmnters’ heritaGe. Berkeley: University of California Press.

NIMUENDAJÚ, CURT U. 1946. The eastern Timbira. Berkeley: Univer sity of California Press.

OBERG, KALERVO. 1973. “Types oí social structure among the low­land tribes of Central and South America,” in PeoPles and

cultures o/ na/ive South America. Edited by Daniel Gross. New York: Doubleday.

Ross, ERIC BARRY. 1978. Food taboos, diet, and hunting strategy: The adaptation to animais in Amazon cultural ecology.


SAHLlNS, MARSHALL. 1972. Stone Age economics. London: Tavistock. STEWARD, JULlAN H., and LOUIS FARoN. 1959.

Native peoples o/ South America. New York: McGraw-HiIl.

TURNER, TERENCE. n.d. “The Gê and Bor6ro societies as dialectical systems: A General model,” in Dialectical societies.

Edited by David Maybury-Lewis. In press.

VIDAL, Lux. 1977. Morte e vida de uma sociedade indígena brasileira. São Paulo: Hycitec.

WAGLEY, CHARLES. 1977. Welcome o/ tears. New York: Oxford Uni­versity Press.

ZARUR, GEORGE C. L. 1975. Parentesco, ritual e economia no Alto XingúzÍ, Brasília: Fundação Nacional do Indio.