Skip to navigation | Skip to content

The Cayapó people of central Brazil: on the nature of two

As one Mẽbêngôkre community moved west across Central Brazil in the early twentieth century, its young men, in their rovings and raidings and fightings, found a new people, attacked them, and kidnapped some children. Two years later, these new people attacked these same Mẽbêngôkre in their home village, to devastating psychological effect, for none of their other new enemies had ever attacked a Mẽbêngôkre village before. This led to several years of fearful flight and counter-raids. Eventually they made a new village far to the northeast of the new people, with two men’s houses at its centre explicitly named for social divisions of these new and now genuinely feared people. They had learned these names from one of the captive children, a boy who was apparently twelve years old.

Those Mẽbêngôkre people had just met the Panará people, who they came to call Krãjôkàr, “The People with Shaven Heads”. In their swaggerings and bullyings, those young Mẽbêngôkre men had not expected to meet people who give back as good as they get. Warfare was certainly important to Panará people, and nobody messed around with them with impunity, but they seem to have been more concerned with being left to their own devices than their new enemies were. They had no specific name for the Mẽbêngôkre people, and their current name for them is a recent borrowing from the language of the neighbouring Yudya people. The mutual attention between these two peoples is assymmetrical.

Was this really a first contact? In the colonial history of Central Brazil, the ancestors of both the Panará and Mẽbêngôkre were called the Cayapó. The word seems to be from an indigenous language unrelated to those of either of these two peoples. It has become habitual in recent anthropology to criticize such colonial usages, and to replace them with supposedly more authentic ones. It is therefore unlikely that readers who are not au courant with the recent ethnographic literature will recognize that the Mẽbêngôkre are the people otherwise known as the Kayapó, once made famous by Rauni, Sting, The Body Shop, and Terence S. Turner, or that the Panará are the people otherwise known as the Kreen-Akrore, a version of their Mẽbêngôkre name, made famous by Adrian Cowell’s remarkable film, The Tribe That Hides From Man.

There is no reason to believe that the ancestors of the Mẽbêngôkre and Panará peoples ever considered themselves to be Cayapó. But the colonial designation comes from a local indigenous perception which necessarily contained a grain of truth. Some other indigenous people, probably speakers of a Tupian language, at some point saw enough commonalities between these two peoples to tell Portuguese colonists that they were a single people, the Cayapó. The name is not a generic reference to hostile indigenous people, but a very specific reference to these two particular peoples, the Mẽbêngôkre and Panará. The Cayapó were seen to have something profound in common.

Indeed, Mẽbêngôkre oral tradition records that when their own ancestors lived far to the east, they had dealings with a people they called Krãjôkàr-re. From war captives, they learned an important ceremony from these people. Their descriptions of these ancient enemies tallies extremely well with accounts of the Panará, both old and modern. The early twentieth century unquestionably saw the reencounter of ancient enemies far to the west of their ancestral lands.

The south and the north

The foregoing account is largely based on Gustaaf Verswijver’s remarkable study of Mẽbêngôkre warfare [1992]. In a piece of methodological virtuosity, Verswijver was able to reconstruct Mẽbêngôkre history in immense detail by concentrating on what interested them: cycles of raiding and vengeance, sequences of name-transmission rituals, village formation, fissioning and migration. These issues interested his Mẽbêngôkre informants because they were vital to those informants for the understanding of their current social relations and their possible future ones. And I mean vital. For a Mẽbêngôkre person to get any of these details wrong could be a matter of life and death.

Anthropology is a much less vital endeavour than Mẽbêngôkre warfare, but apparently no less committed to accurate information. As anthropological knowledge of the people known as the Cayapó grew, it became increasingly clear that they were two distinct peoples speaking quite distinct, although very closely related, languages of the Northern Gê family. The literature started to distinguish them as the Southern Cayapó and the Northern Cayapó. These new monikers were not indigenous but anthropological, and based on alien geographical criteria. Thought to be extinct, the Southern Cayapó were relegated to historical accounts, while the Northern Cayapó went through an orthographic revolution to become the Northern Kayapó, and then, since there was apparently no contrastive southern population still in existence, to become the Kayapó tout court.

Meanwhile, two other things were happening. The Panará and Mẽbêngôkre people were coming into violent confrontation very far to the west of their historical homelands, and anthropology was professionalizing. These two sets of events were clearly not coordinated, and indeed until quite recently anthropological knowledge of the Panará and Mẽbêngôkre peoples was minimal. But anthropological knowledge was not innocent in this case, because as it professionalized, the older accounts of distinct “hordes”, or “nations”, or “peoples”, with a bewildering plethora of different names, became gradually professionally re-imaged as distinct “cultures” or “societies”, that is, as ideally bounded units amenable to ethnographic research. Such research generated monographic studies that could then be compared, but in this process potentially important social relations between the Panará and Mẽbêngôkre came to be seen as external to either culture/society.

In Tristes Tropiques [1976], his very famous account of his travels, Claude Lévi-Strauss records how he journeyed to the south of the region in which the Panará and Mẽbêngôkre were coming into renewed conflict. In From Honey to Ashes [1973], Lévi-Strauss makes clear that his itinerary was set by the impossibility of travelling due north of the city of Cuiabá because of the presence of the very bellicose Beiços de Pau people. These people were later called the Tapayuna, later still the Western Suyá, and more recently by their own self-designation of Kajkhwakratxi-jê. The oral history of their close relatives, the Eastern Suyá or Kisêdjê people, suggests that the Western Suyá had been forced south by intense raiding by the Panará people living to their north. Terrified of the Panará people, and consistently moving southward, the Western Suyá came into violent conflict with other local indigenous and non-indigenous peoples as they tried to carve out a new set of places in which to live well (Batista de Lima [2012]). Unbeknownst to Lévi-Strauss, the actions of the Panará people affected his life choices.

Oddly, there is evidence in Lévi-Strauss’ magnum opus of the ongoing survival of the Southern Cayapó/Panará. M6 of the Mythologiques concerns a war between the Bororo and their enemies, the Kaiamodogue [1970]. The suffix –dogue is a Bororo collectivizer for humans, and the root kaiamo is clearly cognate to the word Cayapó. Lévi-Strauss notes that the Salesian missionaries had originally identified these Kaiamodogue with the Kayapó in 1925, but had increasingly re-identified them with the Xavante people by 1962. Without having consulted the original sources, I think it is likely that the earlier Salesian identification was with the Southern Cayapó/Panará, rather than with the Northern Cayapó/ Mẽbêngôkre living very far to the north, and probably beyond the ambit of meaningful Bororo knowledge. The later Salesian identification of the Kaiamodogue with the Xavante people undoubtedly reflects the increasing proximity of the latter people to the Bororo, and their demonstrable enmity. Further, the Salesian tradition of ethnography was also professionalizing, as Lévi-Strauss noted, and becoming more rigid and dogmatic.

To get from their ancient territory to the area in which the Mẽbêngôkre first attacked them, the Panará people would have had to pass through Bororo territory. In doing so, they would have likely entered into conflict with these people. As such, it is the Southern Cayapó/Panará, and not the Northern Cayapó/ Mẽbêngôkre or the Xavante, who are the most likely candidates for the referent people of the Bororo’s Kaiamodogue. Bear in mind that no ethnographers were aware at that time that the Southern Cayapó/Panará still existed.

Had Lévi-Strauss known of the existence of the Panará people, and about how they live and how they think about things, the Mythologiques would have been an even richer text than it already is. Quite uniquely among speakers of Northern Gê languages, the Panará people do not tell the myth of origin of fire as a gift from the jaguar, as related in Chapter 2 of The Raw and the Cooked [1970]. They tell a very different story, but one that relates Bororo mythology to Northern Cayapó/ Mẽbêngôkre mythology much more directly than the rather circuitous route that Lévi-Strauss had to take in his argument. Lévi-Strauss’ intuition was correct, and had he been able to travel where he wanted and studied what he liked, he would have discovered remarkable things. But he could not travel where he wanted nor study what he liked because he would have gotten himself killed.

A frank declaration of personal interest

Before proceeding further, I should declare a few personal interests. One of the authors of the books under review here, Vanessa Lea, is an old friend of mine, and very likely my cousin in the manner of the Scottish Highland folk. The other author, Elizabeth Ewart, is my former doctoral student, first at the University of Manchester and then at the London School of Economics. She is also a very good friend of mine, although not, to my knowledge, a close kinswoman. Accusations of academic cronyism in this case would be very hard to put down.

That said, this review is directed at a point that is not explicit in either of these excellent ethnographic monographs, but which their being read together makes powerfully clear. The ethnographic monographic tradition, for all its remarkable descriptive possibilities, blinds us to important features of the real world. Panará sociality and Mẽbêngôkre sociality are not objects that are historically exterior to each other, but are instead deeply imbricated.

Long before I had met either Lea or Ewart, as a twelve-year old boy I saw on television Adrian Cowell’s The Tribe that Hides From Man. I was captivated. I simply wanted to be there, but I had genuinely no idea how this could be done. I was only twelve, after all, the probable age of the captive Panará boy who explained Panará social organization to his intrigued Mẽbêngôkre audience. But in that crucial moment, as I began to put away childish things, I formulated a life project. But how to implement it? The tools immediately to hand seemed very unpromising. A few years later, I read an early translation of Claude Lévi-Strauss’s Tristes Tropiques and his description of anthropology as a vocation. There, I discovered that my life project already had a name, and I was set on my way.

For all its manifold faults, Cowell’s film is in essence a fascinating indigenous commentary on a key theme in Lévi-Strauss’ work on indigenous Amazonian peoples, for at its core is a Mẽbêngôkre description of the Panará people. It is therefore about the ‘foreign policies’ or ‘international relations’ of two Central Brazilian peoples. This theme of ‘foreign policy’ was an early one in Lévi-Strauss’s work on indigenous American peoples [1949], and an enduring one. Admittedly, Cowell’s film is able to address the issue only from a Mẽbêngôkre perspective, since the Panará themselves appear seldom in the film and then only at a considerable distance. The film’s apparently ridiculous title is in fact a reference to a specific model of ‘Man’, the definition of humanity embodied in the United Nations or the Universal Declaration of Human Rights. But the film shows clearly that the Panará people, despite having no known peaceful relations with any other peoples, were densely connected to their indigenous neighbours, who knew a lot about them.

The present review is really an ethnographic commentary on a key conclusion to the Mythologiques. While relentlessly comparative, Lévi-Strauss progressively abandons the standard unit of the functionalists, the culture, or of the structural-functionalists, the society, for a very different kind of unit, the specific historical object constituted by the indigenous peoples of the Americas as a whole, which is in turn a specific sub-set of the history of humanity. As cultures/societies evaporate over the course of the analysis, Lévi-Strauss has continual recourse to the comparison of neighbouring peoples and communities, that is to social groups that necessarily have to take each other into account. The sociology of the relations between such neighbours is very poorly understood, and indeed very poorly served by classical social anthropological strategies of comparison. To posit the Mẽbêngôkre and Panará peoples as two different cultures or societies and then to compare them is simple enough. The more difficult but much more interesting question becomes, “What sort of higher-order ‘unit’ might their manifold relations constitute?” It is likely to look very different to anything like a culture or a society. OK, let’s go…

The view from the houses

There is a beautiful photograph in Lea’s book, of two adult women peering through the slats in the walls of a Mẽbêngôkre house at something that is going on inside. For this writer, there is something wrong with this image, something deeply unseemly. It looks like an invasion of privacy. The idea that what is going on inside that house might be of some serious import to those two women, and that they might have very good reasons to be doing what they are doing, does not immediately occur. From Lea’s book, one learns that these women might have every right to know what is going on inside that house. They may have transmitted their names and ritual wealth to children within it.

That the villages of Northern Gê peoples are composed of a large circle of residential houses surrounding a central plaza is well known, and the Panará and Mẽbêngôkre are no exception. As Lea brilliantly shows, Mẽbêngôkre houses are not simply physical structures in which people happen to live. They are also specific places on the village circle that are jealously guarded, and are the containers of names and ritual wealth items and specific ritual rights. They also contain people, for women grow up in them and spend their entire lives there, while men also grow up in them but leave to marry elsewhere. The Mẽbêngôkre call these places on the village circle kikre djam djà, “house standing place”, and Lea calls them Houses, following Lévi-Strauss [1983]. Lea records details of twenty-four different Houses, but this list may well not be entirely complete for all Mẽbêngôkre communities. Any given House in any given community may be comprised of several actual residences located next to each other. At any given time, the names and ritual wealth items of a particular House may be lent out to the members of other Houses, and indeed such lending out is the essence of Mẽbêngôkre social life, but the precise House location of names and ritual wealth items is the valued knowledge of older Mẽbêngôkre women, and both ‘possessions’ must eventually be returned to where they belong. At its simplest, men lend their names to their sister’s sons, while women lend their names to their brothers’ daughters – following out the logic, men’s names endlessly return to their own House location, while women’s names return in a more recondite form. A woman with a lent name has to ideally return it to her father’s sister’s daughter’s daughter. That is the simple form, but it gets much more complex. The name giver/name receiver relationship is not restricted to mother’s brother/sister’s son for men, nor father’s sister/brother’s daughter for women, since the Mẽbêngôkre have a version of an Omaha kin terminology, such that a woman’s mother’s mother’s brother’s daughter is a “father’s sister”, and so on. A similar logic applies to men. Men can lend their names to their daughter’s sons, members of their wives’ House, which means the name must be returned in the next generation, minimally, to the holder’s mother’s father’s sister’s daughter’s daughter’s son.

The complexity of the system is real. Given that Mẽbêngôkre people may have up to thirty different names, many of which they may well not actually know (the older women do), and given that specific ritual wealth items are subject to endless permutations of material and form, the precise House location of any given name or ritual wealth item is often very unclear, and hence subject to serious dispute. Also, villages are constantly fissioning, and there may be no members of the appropriate House in the current village to whom to return lent names and ritual wealth, so the names and ritual wealth may have to be lent onwards to young people in the appropriate kin category but in the wrong House. Further, Houses can split up into daughter Houses, and while names and ritual wealth items of the original House are ideally divided up between the daughter Houses, another level of ambiguity is added. Such often violent disputes about names and ritual wealth items, no matter how serious, simply underscore the point that all names and ritual wealth items ideally belong to one or another House location. All Mẽbêngôkre people hold to this position, but there is, as Lea points out, no consensus as to the exact details.

The Panará are very similar but subtly different. They too have named house locations on the village circle, but only four of them. Again, the actual number of residences on these locations varies with the demography of the village. Identification with a house location is transmitted from a mother to her children, and is immutable: a Panará person is ideally born in his or her mother’s house location, and on death will be buried there. In the circumstances, Ewart has little hesitation in identifying these house locations as exogamous matrilineal clans. Compared to other Northern Gê peoples, the remarkable inflexibility of Panará clans looks quite odd. Unlike Mẽbêngôkre Houses, Panará clans possess little beyond a collective name and a location on the village circle, and while they act as units in certain ceremonies, they are not in competition for ceremonial prestige. For the Panará, there are four clans, and there always have been, and they neither divide nor re-group.

Panará and Mẽbêngôkre villages are circular, but they start in the east, “the root of the sky”. Like the world or a day, the spatial ordering of the village begins in the east, and everything else is calculated from there. For the Panará, who is in the east of a village is not a question, for this is always the Kwakjatantȇra clan. For the Mẽbêngôkre, who is in the east of a village is a question, and a source of considerable dispute. The sheer disputatiousness of the Mẽbêngôkre, as compared to the serene joviality of the Panará, has been regularly commented upon.

The view from the centre

Ewart’s account of dual organization among the Panará is masterful. On the face of it, the Panará have a relatively simple dual organization, with two moieties, “Those of the Root (East)” and “Those of the Tip (West)”, associated with men’s houses located in the centre of the village, exactly where one might expect. But Ewart goes on to note that the actual process of recruitment to a moiety is very vague, can change often over the life course, and seems to be largely down to ‘personal choice’, whatever that might mean in this context. Further, she notes that the imagery of the moieties opposes causes (the East) to consequences (the West). As such, and very unexpectedly, Panará moieties encode what Euro-Americans call ‘history’. Brilliantly, Ewart goes on to observe that the Panará opposition between panará, “we people, humans”, and hipe, “enemies”, has the same basic form. Self and Other are related to each other exactly like moieties, and in fact are moieties. Indeed, any pair of opposed groups, such as young and old or men and women, can operate as moieties related in predictable ways. Of course, Lévi-Strauss had already noted that point [1995], but Ewart provides its dense ethnographic proof for the Panará case.

The Mẽbêngôkre case is apparently quite different. The only person who observed a functioning moiety system among the Mẽbêngôkre was the remarkable German-Brazilian anthropologist Curt Nimuendajú, who described it for both the Pau d’Arco/Irã’ãmranh-re and the Gorotire communities (Lowie [1943]). He reported a system of eastern and western men’s houses associated with moieties remarkably similar to that of the Panará. No subsequent ethnographer has ever visited a Mẽbêngôkre community with a functioning moiety system, although the Mẽbêngôkre can describe the moiety system and are aware of which moiety they belong to. Lea has almost nothing to say about Mẽbêngôkre dual organization in her meticulous ethnography, which thereby reflects the basic irrelevance of this problem to these people. But that simply raises another question. Why, uniquely among Northern Gê peoples, is dual organization of such little interest to Mẽbêngôkre people? Terence Turner’s ethnography provides an answer [1979]: it is not that the Mẽbêngôkre have lost interest in dual organization, but that they have come to see it as far too problematic and far too divisive. Turner reports that Mẽbêngôkre people refuse to use dual organization because it leads to fighting. Yet it is there in every Mẽbêngôkre village, for the supposedly central men’s house is always actually either to the east or to the west of the real centre of the village, depending on the moiety affiliation of the village’s men (Verswijver [1992]). The Mẽbêngôkre maintain dual organization as a potent historical ghost. The Mẽbêngôkre conceive of their actual communities as fragments of earlier complete villages, such as Goroti Kumrẽnhtx or Pykatôti, and hence as intrinsically incomplete.

One of the genuinely remarkable features of Mẽbêngôkre ethnography is the simple absence of any association between naming and ceremonial moieties. This association is pervasive among their closest linguistic relatives, such as the Apinayé and the Suyá, but totally absent among the Mẽbêngôkre. That the Mẽbêngôkre are the exception to a more general pattern suggests that this absence of association between names and ceremonial moieties is a derived and recent feature of these people. Indeed, the moiety system that the Mẽbêngôkre describe for themselves has no association with the ceremonial system, and is instead focused on the opposition between the eastern and western men’s houses, and hence political factionalism. We might speculate that the original system of ceremonial moieties of the Mẽbêngôkre became increasingly associated with the male political domain and increasing dislocated from actual ceremonies and naming. Eventually the moieties became so divisive that no actual Mẽbêngôkre community was able to contain both, although they retained the memories of it as a sort of Arcadia, an image of what the Mẽbêngôkre should be like, but are not.

We might also speculate that as the Mẽbêngôkre moiety system broke free of the association with both names and ceremonial life, there was a parallel development in the importance of Houses, which are after all precisely the containers of names and ritual wealth items. As Lea shows, such Houses are incipiently present among other Northern Gê peoples, but none show the baroque development of this social possibility found among the Mẽbêngôkre. And while very reluctant to express the moiety system at community level in search of their lost completeness, the Mẽbêngôkre do try to persuade members of locally absent Houses to immigrate to their communities in order to facilitate ritual life. Desirable completeness has shifted from moieties to Houses.

While Mẽbêngôkre Houses show no evidence of overt dualism, there is some evidence of residual dualism in Mẽbêngôkre naming. For a start, there are eight classifiers for the prestigious “beautiful” names: 2x2x2 (the Xikrin Mẽbêngôkre are slightly different). Then, the name classifiers Bemp and Tàkàk, which can only generate masculine names, are confirmed during rituals that must occur in different, and ideally alternating, years. Largely feminine names with the classifier Nhàk are confirmed in the Tàkàk ceremony, while the largely feminine names with the classifier Bekwoj share a myth of origin with Bemp names, although they cannot be confirmed during the Bemp ritual (which is a male initiation ritual). The remaining four beautiful name classifiers (4 + 4 = 8) generate mainly female names confirmed during rituals that are apparently no longer performed.

And there is more. All beautiful names can be confirmed during two other ceremonies, Kwàrỳ-kangȏ and Mẽ-biȏk (the latter name means “The Painted People”, and comes in two forms, male and female). The former was adopted from the Yudya people, and is a surreal Mẽbêngôkre version of a Yudya manioc beer ritual. The latter was adopted from the Panará, and is presumably a Mẽbêngôkre version of some unidentified Panará ritual. This suggests that the transformation of Mẽbêngôkre ceremonial moieties was associated with the relatively recent and overt adoption of two rituals from two notable enemies.

Compared to the Mẽbêngôkre, Panará dual organization, while complex, looks more like that of other Northern Gê peoples. But there is a key difference here, for like the Mẽbêngôkre, the Panará make no association between moieties and personal naming either.  The moieties are central to much ceremonial life, but personal naming is a transaction between clans linked by father/child relations. The collective names of Panará people, the clan names, show a rigid matrilineal logic, and are central to other aspects of ceremonial life.

But following a consideration of the peculiarities of the Mẽbêngôkre as discussed by Lea, a parallel problem appears among the Panará. There are four Panará clans: 2×2. This is already suggestive. Further, Ewart argues that the four clans, despite being located at specific places on a circle, are actually located along a line, with two polar clans and two medial clans. The two polar clans have names that are remarkably similar to the names of the ceremonial moieties, and differ only in an explicit reference in the clan names to a key ceremonial activity, log racing (which is actually organized by the ceremonial moieties, not by the clans). The names of the medial clans are of uncertain meaning, even to the Panará. From a certain perspective, the four Panará clans along their line look like two ceremonial moiety systems, one nested inside the other.

While the overt moiety system of the Panará, the ceremonial moieties associated with east and west, is common enough in Central Brazil, the clans and their locations on the village circle look distinctly odd. The ‘eastern’ clan is always in the east of the village, but the presence of the two medial clans effectively shunts the ‘western’ clan to the north of the village. Richard Heelas, the first ethnographer of the Panará, attempted to associate the four clans with cardinal directions [1979], but the subsequent ethnographers, Steven Schwartzman [1984] and Ewart herself, have shown that this is not the case. The oddity is real. The ‘western’ clan actually is in the north of the village, while one of the medial clans is in the east of the village. The problem does not seem to bother the Panará, even though every day the sun rises over the houses of the ‘eastern’ clan, but never sets over the houses of the ‘western’ clan. Panará space, so central to Ewart’s account, is not that of the Euro-American mapmaker, nor of Euro-Americans in general.

It was precisely these apparent oddities in the social organization of Central Brazilian peoples that so fascinated Lévi-Strauss [1963]. Two of the most prestigious Bororo clans are called “The Upper Village Chiefs” and “The Lower Village Chiefs”. One might expect them to be in opposite moieties. But, in fact, they are both in the same moiety, which is opposed to the opposite moiety by a completely transversal logic. Even if one posits another moiety system for Bororo people, as Lévi-Strauss did, based on an upper/lower contrast rather than the north/south contrast of the overt moiety system, this works really well for the ‘northern’ moiety, but not so well for the ‘southern’ moiety, which seems to work by quite different principles. It is as if, as Lévi-Strauss discussed for the North American Winnebago, the two halves of the same community experienced their basic social relations with each other in completely different ways.

Nimuendajú, the HCBP and the Panará

Our earliest professional anthropological account of either of the two Cayapó peoples is that of Curt Nimuendajú for the Pau d’Arco/Irã’ãmranh-re Mẽbêngôkre (Lowie [1943]). He reported that they were organized into exogamous matrilineal moieties called Upper and Lower, located in the eastern and western halves of the circle of residential houses respectively, and that they had two non-exogamous moieties, again associated with the east and west, located in the centre. Nimuendajú’s account is surprisingly close to the underlying model of Panará social organization just discussed: if the medial clans of the latter were removed, then Nimuendajú’s account of the Mẽbêngôkre would describe them perfectly.

No subsequent investigator of Mẽbêngôkre communities has found anything remotely resembling Nimuendajú’s account of exogamous matrilineal moieties, but we do not know if he simply misunderstood what he saw and was told, or whether his account was correct. The Irã’ãmranh-re Mẽbêngôkre people were almost totally exterminated in the exceptionally violent colonization of the Conceição de Araguaia region where they lived, and it is doubtful if their very few living descendants, if they still exist, would have much of any pertinence to say about the question, or whether it would be acceptable to ask them. That said, the basic similarity of Nimuendajú’s account of the Mẽbêngôkre to the underlying logic of Panará social organization suggests something deeper may be going on.

Nimuendajú was strongly influenced by Robert Lowie, whose approach to social organization was, of all the students of Boas, closest to what was to become classical British structural-functionalism. In an important way, Lowie professionalized Nimuendajú’s ethnography, leading him to see his data through a filter sensitive to concepts like patrilineality and matrilineality, patrilocality and matrilocality, dual organization, etc., concepts quite far from Nimuendajú’s original interests. At the same time, Lowie’s work became the initiatory experience of the young Claude Lévi-Strauss, as recounted in Tristes Tropiques.

In the 1960’s, David Maybury-Lewis, a British structural functionalist who had studied the Xerente and the Xavante peoples, speakers of Central Gê languages, organized a large project which sent doctoral students to study the Gê and other indigenous peoples of central Brazil [1979]. The project was known as the Harvard Central Brazil Project or HCBP (that, at least, is the standard American version of the story, the Brazilians having a rather different one). Ostensibly, the project was to empirically test Lévi-Strauss’ ideas about the social structures of these peoples, and indeed Maybury-Lewis’ students did restudy two of the peoples studied by Lévi-Strauss, the Bororo and the Nambikwara, but the major work of the HCBP was actually the restudy of most of the Gê peoples originally studied by Nimuendajú.

The studies of the HCBP showed that Nimuendajú made two serious mistakes about the social structure of Gê peoples. Firstly, he reported that the Xerente people were patrilocal [1943]. Secondly, he recorded a very unusual descent and marriage system for the Apinayé people, speakers of a language apparently mutually intelligible with that of the Mẽbêngôkre [1939]. His mistake about the Xerente seems to have been an entirely understandable false intuition about a social system that was utterly non-transparent to anthropologists at the time. What he recorded about Apinayé sociality is more germane to the topic at hand.

Nimuendajú argued that the Apinayé were organized into four named exogamous groups called kiye. These groups were recruited by the rather unusual mode of parallel descent, that is, men were recruited to their father’s group, women to their mother’s. Their exogamy was unusual too, since men had to marry the women of one specific other group, as did the women in turn. This strange social organization puzzled many anthropologists, and much was written about it.

Nimuendajú’s account of the Apinayé kiyé system was challenged by Maybury-Lewis [1960] and his former student Roberto Da Matta [1973] on logical and ethnographic grounds respectively, but he was clearly told something like this by his Apinayé informants: he could not simply have made all of this up. In essence, his kiyé system was one of four exogamous matrilineal clans, at least from a female perspective. Two of these clans had distinctively moiety-like names, while the other two did not. The analogies with actual Panará social organization, while by no means exact, can hardly be entirely fortuitous. When Da Matta reports that his informants told him that two of the kiyé names were simply descriptors of ritual wealth items of the real underlying non-exogamous moieties, one begins to wonder why these moieties, apparently unconnected with the naming system which is associated with a quite different moiety system, actually needed two names in the first place? Perhaps Nimuendajú was genuinely confused by what his informants told him, but it seems remarkable to me that his confusion took the specific form of an account of four exogamous ‘matrilineal’ clans, present among no other indigenous people he had worked with or knew about, but which is actually realized by the then very-little known Southern Cayapó/Panará people, who were at the time assumed to be extinct.

That Nimuendajú’s Apinayé informants’ accounts of the kiyé system are very similar to the actual social structure of the Panará people is worthy of some comment. While it is not true that Panará men are recruited to their clans by patrilineal descent, father-child relations are central to naming, to the genesis of joking relations, and the ultimate establishment of marriageability between clans. As Ewart makes clear, these are precisely panará jon soti, the “things of the Panará people”, their sense of their own complexity and uniqueness. But Ewart makes a deeper point about Panará social organization that can be extended to all of these Central Brazilian peoples. Their circular villages are clearly very important, and make for beautiful maps, which misleads anthropologists into thinking that these peoples think of their social relations in terms of a single overarching totality. They do not, however, see their social relations, so to speak, from above, in the manner of the Euro-American mapmaker. Each person sees their social relations from where they actually are in relation to specific other people. As such, Nimuendajú’s Apinayé informants probably did not understand his questions as being about Apinayé ‘social structure’, but about how their own social relations looked like to themselves. As we have seen, Northern Gê moiety systems are intrinsically unstable, and it is entirely possible that Apinayé people may at one time have found it logically possible that a system of moieties, each with two separate names, was actually a system of four different groups, and then tried to think through how these groups were related to each other using their own immediate social relations as a guide.

The members of the HCBP were clearly more professional in their approach to fieldwork than Nimuendajú, insofar as they had been more systematically trained and were, let’s be frank, less precariously financed. They were also, unfortunately, more dogmatic than he had been. They were out to prove a methodological point about anthropology, after all, and professionalized ethnography always wins out within the profession. I fully accept that Da Matta’s account of Apinayé society is more sophisticated and fuller than that of Nimuendajú, and that he provides a very convincing account of why the latter got things wrong. But two questions remain. Firstly, why is Nimuendajú’s ‘garbled’ account of the Apinayé so surprisingly similar to the actual social organization of the Panará? It is possible that Nimuendajú’s informants massively over-systematized for him what was actually going on, but they were doing so in a way that was both logically possible, and with a few necessary tweaks here and there, actually liveable. Secondly, Da Matta never addresses the possibility that the Apinayé might have actually changed in the decades between Nimuendajú’s fieldwork and his own. The possibility is far from unlikely, given that they were experiencing absolutely catastrophic population decline when Nimuendajú described them, and the beginnings of a demographic recovery when Da Matta did. In the parallel case of demographic catastrophe for the Panará, one clan came very close to complete extinction, with implications which we will never know, for luckily it survived.

This is an issue raised by Lévi-Strauss on several occasions. With his characteristic intellectual humility, he argued that we should treat earlier accounts with great respect because their writers were there before us, and often long before us. Things might well have changed between the time of the earlier accounts and what we later witnessed and were told. This is the hidden danger of professionalization. Professionalization is itself a process that takes time, so necessarily deals with a different historical data-set from its less professionalized predecessors. And the fact that people deny an earlier account of themselves does not, in itself, prove that that account was wrong. We tell ourselves and each other all sorts of unconscious lies about the past all of the time. In fact, these two questions and the problems they address are really one and the same. It is entirely possible that Nimuendajú’s Apinayé informants were, at the time they spoke to him, remembering a former state of things that was now lost and trying to work out the logical possibilities of a system that they experienced as currently chaotic, and that Da Matta’s informants had simply forgotten those earlier debates.

One of the most impressive contributions of an HCBP member, if relatively uncelebrated, is that of Jean Lave on the Krĩkatí people [1979]. Lave made a key methodological assumption: she simply assumed that what Nimuendajú had said about Eastern Timbira peoples was, basically, accurate. The task then shifted from the pointless project of proving that Nimuendajú (or Lévi-Strauss) was wrong, a hobby for the boys, to the much more interesting anthropological question about the genuine nature of the historical dynamics of the societies in question.

The Cayapó

Long ago, Nimuendajú noted a curious fact [1939]. The Mẽbêngôkre community who currently call themselves the Gorotire are actually calling themselves by the name that one half of the Apinayé people call themselves. Gorotire is a linguistic cognate of Kolti, one of the Apinayé ceremonial moieties associated with naming (the other is Kolre). Gorotire is the rump, so to speak, of the ancient community of Pykatôti, which is an icon of a lost Mẽbêngôkre unity: the earlier icon of maximal Mẽbêngôkre unity was the village of Goroti Kumrẽnhtx. But why was its name one that among the very closely related Apinayé logically implied the existence of its other side? We do not know, but it seems significant that the historical separation of the Mẽbêngôkre from the Apinayé people involved both the abandonment of a moiety relation and the retention of one of its terms. History is here the reduction of two to one.

Reading the ethnographies of Lea and Ewart in parallel, I began to have the sense that what the Mẽbêngôkre and Panará have in common as against other Northern Gê peoples might not be entirely fortuitous. Most of the ways in which Mẽbêngôkre social relations differ from those of their closest linguistic relatives, the Apinayé and Suyá peoples, make the former people more like the Panará. The disassociation between moieties and names, the association of moieties with two men’s houses, and the strong association of people with specific and sequential spatial locations on the village circle found among the Mẽbêngôkre make them much more like the Panará than they are like the Apinayé. In this light, my little historical hypothesis about the recent transformations in Mẽbêngôkre social organization look like attempts by the Mẽbêngôkre to coordinate their social organization with that of the Panará. These attempts can hardly be recent, since they are found in all Mẽbêngôkre communities, communities that have not shared a common history since at least the first half of the nineteenth century, ever since they moved west of the Araguaia river.

My speculations here are unashamedly conjectural, in the sense of the conjectural history that Radcliffe-Brown famously rejected as he set out the correct methodology for structural-functionalism [1952]. Conjectural history, in the sense of making up a “Just So” story to account for any apparent shortcoming in the fit between description and theory, must still following Radcliffe-Brown be rigorously rejected as intellectual laziness and poor science. But the present case is rather different. Here we do have actual historical data, albeit of a slightly unusual kind.

Firstly, linguists seem to be clear that the Panará and Mẽbêngôkre languages are less closely related to each other than the Mẽbêngôkre language is to those of the Suyá and Apinayé peoples. The issue is very complicated, but a generally accepted principle in historical linguistics is that relative difference between two languages reflects the time-depth of the separation of their speech communities. As such, the Mẽbêngôkre, Suyá and Apinayé people should form a meaningful historical unit, with the Panará as an outlier. Mẽbêngôkre oral tradition and action shows something very different. The Suyá and Apinayé people were apparently never particularly meaningful others for these people, while the Panará were.

As discussed above, Mẽbêngôkre oral tradition records their former relations with a people they called Krãjôkàr-re, and a brief description of them: these people were almost certainly the Southern Cayapó/Panará. And it is from these people that the Mẽbêngôkre learned one of the two ceremonies that can confirm all beautiful names of whatever classifier. This suggests that the violent contact with the ancient Panará was occurring at the time of the transformation in ceremonial moieties and name transmission hypothesized above. As such, these ancient Panará transmitted to the Mẽbêngôkre a new mode for name transmission.

There is more. As noted at the outset, the Mẽbêngôkre community whose young men re-found the Panará far to the west responded to the ensuing warfare by moving far to the northeast and founding a new village with two men’s houses explicitly modeled on what they understood to be Panará social categories, as told to them by the captive boy. This response is all the more remarkable given that this Mẽbêngôkre community must already have known about the disastrous consequences of having two men’s houses in a single village for its political viability. It is as if they were trying to re-inject the moiety-associated model of a village with two men’s houses with renewed Krãjôkàr-re-ness. It did not work. What the Mẽbêngôkre do not seem to have understood about the two men’s houses of the Panará, or perhaps actively rejected, was that these operate as ceremonial moieties, not as political factions (which are organized on a different basis). The two men’s houses of the Panará evoke the collective euphoria of log racing, while the two men’s houses of the Mẽbêngôkre evoke political divisions and tensions, and inevitable fission.

The Suyá people, themselves split between eastern and western branches, complicate this picture. While by no means certain, the split between the Mẽbêngôkre and the Suyá possibly postdates the split between the Mẽbêngôkre and the Apinayé. Certainly, Suyá social organization has more in common with Mẽbêngôkre and Panará social organizations than it does with that of the Apinayé. The Suyá have named houses on the village circle, which effectively function as matrilineal exogamous clans (Seeger [1981]). Among the Eastern Suyá, at least, one of these named houses is explicitly in the east. The horrific and fully genocidal demographic catastrophe that the Western Suyá experienced makes it very difficult to know if this was also true of them. What we do know is that the Eastern and Western Suyá retained the relation between ceremonial moieties and naming, even when the ceremonial moieties were associated with two men’s houses. This is a crucial difference to the Mẽbêngôkre, whereby the latter people more closely resemble the Panará than their closest linguistic relatives.

Careful case-by-case comparison of the Northern Gê peoples, of the sort already carried out by Marcela Coelho da Souza in her excellent doctoral thesis [2002], could be added to what is known about the history of these peoples to explore the mechanisms by which they have come to be as they are today. We know the recent history of the Mẽbêngôkre in considerable detail, thanks to Verswijver’s remarkable work, and the histories of the Eastern and Western Suyá are becoming better known too (Batista de Lima [2012]). The history of the Apinayé is well-known, and there are now at least three ethnographic snapshots of their social tranformations over the past century, from Nimuendajú, Da Matta and Giraldin [2000]. The pre-contact history of the contemporary Panará is more obscure, but it may be possible to enrich it, from sources both among them and among their neighbours. And for the earlier Panará, documentation is rich, and their history is known in great detail from the work of Giraldin [1997]. A new object beckons, an indigenous political history of Central Brazil.

The prospect of this new object can be founded on the remarkable historical datum with which I began. One community of Mẽbêngôkre were so impressed with the awesome bellicosity of the newly re-found Panará people that they founded a village explicitly modeled on what they understood Panará social organization to be. Their source was a twelve-year old boy, captured in war. It is clear is that the Mẽbêngôkre people were always impressed by the bellicosity of the Panará people, and their sheer vindictiveness when provoked.

Mẽ-biȏk/The painted people and their peanuts

From Ewart’s profound account of Panará dual organization, we can extract a deeper point about the comparison between these people and the Mẽbêngôkre people engaged in here. Following Lévi-Strauss’ insights into dual organization in Central Brazil, Ewart effectively argues that dual organization is simply how Panará people see all social relations. As such, the Panará and Mẽbêngôkre peoples see each other as the moieties of a social object that is of a very unfamiliar kind. Merely by thinking about the differences between each other, the Panará and Mẽbêngôkre necessarily call into being another social object of which they are the respective sides. What exactly this social object might be is very obscure to me, largely because it was until recently constituted by warfare. For this Euro-American here, warfare is an unpleasant form of activity that ideally should end. The feeling seems to be general to people like me: we peaceniks think it should end in world peace, while the warmongers feel it should end in victory. It comes to much the same thing. The Panará and Mẽbêngôkre peoples, as with most indigenous Amazonian people, do not seem to think like this at all. There is a great deal that we do not know about these people, but the two books reviewed here, in their immense wealth of detail, are beginning to reveal it to us. That is a great project for the future, but to conclude I return to an ethnographic present, or to put it less pretentiously, what I remember about being around the Panará and Mẽbêngôkre people in 1997.

In Nansepotiti, Kâranpô is painting me. It is taking forever. She is painting me in Mẽbêngôkre style. In response to my request to be painted in Panará style, Krentoma says, “Panará painting is ugly!” So, Mẽbêngôkre style it is going to have to be. When the session finally ends, I walk out onto the plaza of the village, to be greeted by loud and excited commentary. I ask Elizabeth to translate, and she tells me that one response was, “How have you come to be so beautiful?” Here I am, a Scotsman, painted in Mẽbêngôkre style by a young Panará woman, in a Panará village, being greeted by open expressions of aesthetic awe. I am become a very strange and interesting art work indeed.

A few days after the painting session, news came on the radio that Megaron, a high-up official in the Brazilian federal state bureaucracy tasked with indigenous people, was going to visit the village the next day. Megaron is Mẽbêngôkre. I was a bit nervous about this news, since my presence in the Panará village was technically illegal, and Megaron had the right to expel me. That night, the usually very genial Akâ made an angry speech in the village plaza, punctuated by the thumping of his club on the ground. Akâ was clearly very annoyed about something, which Elizabeth was unwilling to enlighten me about. The next morning, I awoke to discover that everyone had painted themselves completely black, head to foot, the body painting of war. Good grief! Apparently, the Panará had collectively decided that Megaron was coming to their village to expel Elizabeth Ewart’s older brother, who in his close-cropped hair showed his intense mourning for her recently deceased mother, and they were absolutely determined to prevent that for Elizabeth’s sake, this person for whom they had come to feel genuine affection. Needless to say, when Megaron actually did turn up in Nansepotiti, the two of us had a polite, inconsequential and entirely forgettable conversation. What else are reasonable people going to do in a village full of people literally dressed to kill?

A day or two later still, I seized the opportunity of another incoming plane to leave Nansepotiti. Time was very short, and there was an unholy scramble to give my things away to those I wanted to give them to, and to receive new things in turn. Somebody called to Pikon, who was in the kitchen house at the time, to give me a sackful of peanuts, and she shouted back (I paraphrase Elizabeth’s translation at the time),

If those Mẽbêngôkre people want peanuts, let them make gardens, let them plant peanuts, and let them harvest them! I am not doing it for them!

Everyone present was incredibly embarrassed, except for me who had no idea of the meaning of Pikon’s outburst. Quietly, someone said to her, “No, the peanuts are for Elizabeth’s older brother’s mother.” Extremely embarrassed in turn, Pikon rushed out with a huge sack of peanuts to give to me for my mother.

If the threat of ignominious expulsion from Nansepotiti was averted by the collective war paint of the Panará people, the ongoing consequences of Kâranpô’s wonderful artwork is another story for another time. To conclude, I focus on the peanuts. Some of those gifted peanuts made their way back to my mother (the entire sack was far too big, and the legalities far too complex). I told her, in her house in Edinburgh, “Pinkon sent these to you.” She enjoyed eating them, but was genuinely perplexed as to why a Panará woman would want to send her such a gift. I was incapable of explaining this in any terms that made sense to her. Thinking about it now, I suspect that she was worried about her own ignorance of how to compose a sufficiently polite thank-you letter to Pikon, this generous woman whom she knew that she genuinely did not know anything about at all. As an assiduous reader, I am sure that she would have found Ewart’s book fascinating. But she accepted the peanuts in good stead as an ongoing consequence of what her youngest son had found in his own wild rovings through the world.

Enemies, mothers and twelve-year old boys. This is a good little series of analytical terms, a sturdy little calculus, and perhaps the founding stone of a future science of what it means to be human. In the meantime, watch this, which records what another twelve-year old boy did next, and which is the perfect reprise to what I have written here:

The next time you meet a twelve-year old boy, give him the thumb’s up!



Batista de Lima, Daniela

2012 “Vamos amansar um branco para pegar as coisas”: Elementos da etnohistória Kajkhwakratxi-jê (Tapayuna), MA Thesis. Brasília: Universidade da Brasilia.

Coelho da Souza, Manuela

2002 O traço e o círculo: o conceito de parentesco entre os Jê e seus antropólogos, PhD thesis. Universidade Federal do Rio de Janeiro.

Da Matta, Roberto,

1973 “A reconsideration of Apinayé social morphology”, in David Gross (ed.) Peoples and Cultures of native South America, New York: Doubleday/Natural History Press.

Ewart, Elizabeth

2013 Space and Society in Central Brazil: A Panará Ethnography, London School of Economics Monographs on Social Anthropology, Volume 80: London: Bloomsbury Academic.

Giraldin, Odair

1997 Cayapó e Panará: Luta e Sobrevivência de um Povo Jê no Brazil Central, Campinas: Editora de Unicamp.

2000 Axpen Pyrak : história, cosmologia, onomástica e amizade formal Apinajé, PhD thesis. Campinas : Unicamp.

Heelas, Richard H.

1979 The social organization of the Panará, a Gê tribe of Central Brazil, PhD thesis,Oxford University.

Lave, Jean

1979 “Cycles and Trends in Krĩkatí Naming Practices”, in Maybury-Lewis (ed.).

Lea, Vanessa R.

2012 Riquezas Intangíveis de Pessoas Partíveis: Os Mẽbêngôkre (Kayapó) do Brasil Central, São Paulo: Editoora da Universidade de São Paulo.

Lévi-Strauss, Claude

1949 “La politique étrangère d’un société primitive”, Politique étrangère, 14(2):139-152.

1963 Structural Anthropology, New York and London: Basic Books, Inc.

1970 The Raw and the Cooked, London, Jonathan Cape.

1973 From Honey to Ashes, London, Jonathan Cape.

1976 Tristes Tropiques, London: Penguin Books.

1983 The Way of Masks, London: Jonathan Cape.

1995 The Story of Lynx, Chicago and London: University of Chicago Press.

Lowie, Robert H.

1943 “A note on the social life of the Northern Cayapo”, American Anthropologist, 45:633-6.

Maybury-Lewis, David

1960 “Parallel descent and the Apinayé anomaly”, Southwestern Journal of Anthropology 16:191-216.

Maybury-Lewis, David (ed)

1979 Dialectical Societies: The Gê and Bororo of Central Brzil, Harvard University Press.

Nimuendajú, Curt

1939 The Apinayé, The Catholic University ot America, Anthropological Series: Washington D.C.

1942 The Šerente, Publication of the Frederick Webb Hodge Anniversary Publication Fund, Los Angeles.

Radcliffe-Brown, A. R.

1952 Structure and Function in Primitive Society, London: Routledge and Kegan Paul.

Schwartzman, Steven

1984 The Panará of the Xingú National Park: The transformation of a society, PhD Thesis, University of Chicago.

Seeger, Anthony

1981 Nature and Society in Central Brazil: The Suyá Indians of Mato Grosso, Cambridge, Mass.

Turner, Terence S.

1979 “Kinship, Household and Community Structure amongg the Kayapó”, in Maybury-Lewis (Ed.).

Verswijver, Gustaaf1992 The Club-Fighters of the Amazon: Warfare among the Kaiapo Indians of Central Brazil, Ghent: Rijkuniversiteit Gent.

Please join our mailing list to receive notification of new issues