Carolina’s Story. A case study of the diffusion of a myth in Southwestern Amazonia

ABSTRACT

Peoples can disappear, but not stories. This is what happened with a kuniba myth, narrated to Nimuendaju by Carolina, in a situation of forced removal promoted by the Brazilian state at the beginning of the 20th century. Such myth was about the origin of the moon, caused by the incest of a brother with a sister. Until the beginning of the 20th century, neighboring peoples of the Kuniba, such as the Cashinahua and the Kanamari, had very different narratives about the origin of the moon. However, throughout the 20th century, with the departure of the Kuniba, the theme of incest between brother and sister becomes part of the narratives, spreading throughout the region of the Juruá and Purus river basins. This leads to a reflection on the problem of “informational causality”, on the “foreign policy” that differentiates peoples, and on the “spatial dialectic” promoted by myths.

KEYWORDS
Ensemble; mythology; incest; history; moon

RESUMO

Um povo pode desaparecer, mas não uma história. Isso foi o que ocorreu com um mito kuniba, narrado a Nimuendaju por Carolina, em uma situação de remoção forçada promovida pelo Estado brasileiro no começo do século XX. O mito em questão falava da origem da lua, ocasionada pelo incesto de um irmão com uma irmã. Até o começo do século XX, povos vizinhos dos Kuniba, como os Cashinahua e os Kanamari, tinham narrativas muito distintas sobre a origem da lua. No entanto, ao longo do século, com a partida dos Kuniba, o tema do incesto entre irmão e irmã passa a estar presente nas narrativas, difundindo-se pela região das bacias dos rios Juruá e Purus. Isso conduz a uma reflexão sobre o problema da “causalidade informacional”, sobre a “política externa” estabelecida entre diferentes povos e sobre a “dialética espacial” promovida pelos mitos.

PALAVRAS-CHAVE
Ensemble; mitologia; incesto; história; lua

EDITORS’ NOTE

In December 2014, Peter Gow presented this article, still unfinished, for the appreciation of Revista de Antropologia. Given the enthusiasm of the editorial board, he was supposed to submit a finished version, but that ended up not happening. Gow died prematurely on May 18, 2021, leaving a work of great importance for the Anthropology of Lowland South America. As it could not be otherwise, we decided to publish the article now and, for that, we thank Marcos de Almeida Matos who, in addition to translating the text, recovered citations and bibliographic references, and wrote some notes in order to fill in gaps. “Carolina’s Story” is part of a set of texts writen by Gow, which is dedicated to pursuing and analysing mythical transformations - admittedly under the inspiration of Lévi-Strauss’ Mythologiques - in Southwest Amazonia. (One of them, “A mythical cline in Western South America: exploring a Lévi-Straussian ensemble”, was translated into Portuguese and published in the journal Tellus, in 2010). Following the proposal of his book An Amazonian Myth and its History (2001), Gow regards the way in which mythical transformations contribute to the understanding of historical transformations of indigenous peoples, and vice-versa. “Carolina’s story” is also the fourth article by Gow published in Revista de Antropologia (there is also an interview, included in volume 54/1, 2011). With its edition, we offer a small tribute to this bright anthropologist and contributor.

In the Mythologiques, Claude Lévi-Strauss argued for the existence of a heretofore unrecognized social object that he there called an ensemble. His translators into English, John e Doreen Weightman, called it a “larger entity”, but I here maintain the French original. What Lévi-Strauss was pointing towards with the concept of the ensemble is that tribal societies, the historic object of anthropology, are neither small-scale nor simple, but large-scale and complex. The problem is that they are large-scale and complex in ways that are very unfamiliar to our habitual modes of measuring scale, whether in size or complexity. Here I give an empirical example of one such ensemble, and explore certain of its sociological implications.

CAROLINA AND HER STORY

In December of 1921, the German-Brazilian anthropologist Curt Nimuendaju was given a list of 128 words in the Kuniba language by a woman called Carolina, who he met in the city of Manaus in Amazonia. He was also told a myth about the origin of the moon narrated by the Kuniba people. While not entirely certain, since the name of the teller of the myth is unrecorded, it seems most likely that it was told by this same woman, Carolina. Those 128 words and that myth represent a remarkably high proportion of what we know about Carolina, and about her people, the Kuniba. Carolina’s 128 words show, without a shadow of a doubt, that the Kuniba language was a dialect of the language of the Arawakan family most commonly known as Piro, spoken in various communities throughout Southwestern Amazonia, including the communities I have researched on the Urubamba river in Peru since 1980. The myth is the subject of this article.

The following is a summary of Carolina’s story given by Claude Lévi-Strauss, “What genipa is”.¹

A young Indian woman was visited every night by a stranger whose face she rubbed on one occasion with the bluish-black juice of the genipa [fruit]. In this way she discovered her lover was her brother. The culprit was driven out; as he fled, enemies killed him and cut off his head. Another brother who was trying to rejoin him, offered the head shelter. But it never stopped asking for food and drink; the man resorted to craft, abandoned it and fled. It managed to roll as far as the village and attempted to get inside his hut. As it was not allowed to enter, it considered in turn several possible metamorphoses; into water, stone, etc. Finally, it chose to be the moon, and rose into the sky unrolling a ball of thread as it went. In order to be revenged on his sister who had denounced him, the man, now changed into the moon, inflicted the curse of menstruation on her ([1968]1978: 94).

Lévi-Strauss’ version of Carolina’s story comes from a book by Herbert Baldus, which in turn was based on a letter to the latter from Nimuendaju himself. As I will explain later, this version critically changes what Carolina actually told him.

SOME MAPS AND DIAGRAMS

The ensemble that I am exploring here is discussed by Lévi-Strauss in The Origin of Table Manners, and which he there represents in two figures:

Figure 1
Tukuna and other tribes.

Figure 2
Group structure of Tukuna, Cashinahua and Munduruku myths.

The myth numbered M392 in Figure 2 is Carolina’s story.

Figure 2 clearly refers to some genuine social entity. It is made of three empirically existent human societies, the Cashinahua, the Tukuna, and the Munduruku, and four very concrete social actors, the venturesome husband, the shy virgin, the incestuous brother and the trusting visitor. Nobody except Claude Lévi-Strauss has ever suggested that Figure 7 is a real social object. My favorite representation of what I am looking at is this: “Sketch of the Zone of Litigation”. This map was prepared by the very famous Brazilian writer Euclides da Cunha²

Figure 3
“Sketch of the Zone of Litigation”, Cunha (1907).

The problem is not empirical, but rather methodological. These remarkable objects are out there, as the myths clearly reveal, but we do not yet possess the methodological devices to explore them.

The functionalist illusion. The mistaking of a methodological assumption for the absolute properties of an object.³

Lévi-Strauss’ analysis of this ensemble was based on myths collected in the early twentieth century. If we were to repeat this analysis based on myths collected in the late twentieth century or early twenty-first century, it would look very different.

CAROLINA’S PEOPLE, THE KUNIBA

Obviously, Carolina is unlikely to still be alive, and the Kuniba people are officially ‘extinct’: that is, they are not recognized as an existent indigenous people by the federal government of Brazil. It is possible, and indeed probable, that there are people alive today who are descended from the Kuniba, but there is no known Kuniba community in existence now. Indeed, Nimuendaju’s account of Carolina is almost the last mention of these people as living humans in the historical record.

Nimuendaju himself wrote of the Kuniba people,

This tribe, now extinct, lived until 1912 in the high ground between the leftbank of the middle Juruá river and the headwaters of the Jutaí. Following an attack they made against a rubber post, the main part of tribe was killed by Neo-Brazilians. Some survivors were transferred by the Serviço de Proteção aos Indios (the Indian Protection Service) to the Branco river [in the extreme north of Brazil]. Their language is Arawakan, of the Pre-Andine group.

The story was recorded in 1921 (1986:89).

Nimuendaju’s account here is somewhat inaccurate. Contrary to his account, the Kuniba had lived from at least the early nineteenth century on the middle Juruá river, and were the only people to live permanently on the mainstream, as opposed to the sidestreams like their indigenous neighbours, the Kanamari and Kulina. By 1867, according to the English geographer Chandless, the Kuniba were trading for metal tools with non-indigenous traders, who were already regularly ascending the Juruá to obtain forest products. They traded these metal tools to their kinspeople the Manchineri on the Purús river far to the south. The Kuniba population was apparently never very large, and seems to have consisted of a single village of less than one hundred people.

Historical records on the Kuniba people are absent between 1867 and 1912, but clearly they were drawn, in ways that we do not know, into the very rapidly expanding industrial extraction of wild rubber in the Juruá area, and it was this that led to their downfall. Given their long history of dealing with non-indigenous traders, it is likely that they shifted into the production and trade in rubber with relative ease: this may explain the lack of subsequent accounts of them until 1912. 1912 is a significant year, because it saw the beginnings of the precipitous collapse in world prices for wild Amazonian rubber, and this may explain the sudden shift in relations between the Kuniba people and the rubber bosses.

Judging from a local eye-witness testimony of apparently Kuniba origin and the records of the the SPI, the federal Indian Protection Service, we can piece together some account of what happened. Shortly before 1912, the Kuniba had been living on the Itucuman, a tributary of the Tarauacá, itself a southbank tributary of the Juruá, and on their traditional trade route to the Manchineri people of the Purús river. While they were living on the Itucuman, a rubber boss kidnapped the son of the Kuniba chief, Manoel Antônio, and of his wife, Wariman. The Kuniba fled north, across the Juruá river, to Icaraí, a settlement on the Jutaí river. In May 1912, they were followed by Cornélio Chavez, presumably the boss who stole the boy. The Kuniba killed him, his wife and one of his workers. They kidnapped Cornélio’s wife and their five daughters. The Kuniba were then tracked down by two other rubber bosses, João Rufino and Reinaldo Cavalcanti, who organized the massacre.

The SPI records note that only two Kuniba people, the chief and his wife, were killed in the massacre by the rubber bosses and their workers. The survivors of the massacre, seven men, five women and at least three children were forcibly brought back to the rubber station Restauração, on the Juruá: they included a woman called Carolina. Most of the survivors agreed to leave with the SPI, while three chose to stay at Restauração: two of the three children were given by their mothers to local white bosses. On the 3rd of November, the surviving Kuniba boarded a boat that took them away to Manaus, and then to a colony in northern Brazilian Amazonia.

The story of the Kuniba people goes cold, as Carvalho notes, until 1921, when Curt Nimuendajú met Carolina in Manaus (Carvalho, 2002). This is obviously the same Carolina as mentioned in the SPI report. In a letter to Theodor Koch-Grunberg, Nimuendaju expressed his disatisfaction with Carolina as a linguistic informant, suggesting that she had forgotten her language. There is some evidence of this same dissatisfaction in the published word list. Nimuendaju’s letter records that Carolina was living at a SPI posto, “post”, an SPI-supervised indigenous community, although no post was ever founded for the Kuniba. It seems likely, in these circumstances, that Carolina was no longer living in a Kuniba community, and no longer actively speaking Kuniba. That said, her command of the Kuniba language, judged by other data on Piro dialects, was clearly impressive.

JURUÁ-PURUS MYTHOLOGY IN THE LATE TWENTIETH CENTURY

Even if Carolina and her people are ‘extinct’, the story she told is not. Versions of this story are told by indigenous people throughout the Juruá-Purús area, peoples who were former neighbours of the Kuniba.

The Cashinahua, speakers of a Panoan language, lived to the south of the Kuniba in the early twentieth century, in the headwaters of the major southern tributary of the Juruá river, the Tarauacá. As noted, this was along the trade route that the Kuniba people used for travel to and from the Purús river. Sometime between 1910 and 1920, many Cashinahua moved south towards the Purús river, and into what is now Peru. Cecilia McCallum recorded the following late 20^th century Cashinahua version of Carolina’s story, on that river. It goes like this:

Yube used to visit his sister at night and make love to her. She did not know who her lover was, so she stained his face with genipa, and he fled from the village. Together with his brother-in-law he went to hunt their enemies the Bunkunawa dwarfs. Unfortunately the Bunkunawa killed Yube and decapitated him. His weeping brother-in-law buried him in the ground, but the head arose from the grave. In hideous fashion it followed after the terrified brother-in-law, bouncing along and pleading all the while for food and drink. All attempts at satisfying the head’s thirst failed, for the water poured out of the gaping hole in his neck. Eventually Yube realized that he must transform himself. He asked the women to throw coloured cotton into the sky and by grasping it with his teeth he ascended into the sky and became the moon (uxe). ‘Look!’ a little girl exclaimed, ‘There is Uxe!’ Furious, Yube’s head, who wanted to be known as Yube-Head-Foreigner (Yubenawanbuxka), caused all the women to bleed with red macaw tail feathers. After they stopped, they all became pregnant. After this, a rainbow appeared for the first time, transformed from Yube’s blood. As a result, humans are able to die, for there is a path from this world to the heavens (McCallum, 2001: 153-154).

In the early twentieth century, the Kanamari people, speakers of a language of the Katukinan family, lived as immediate neighbours of the Kuniba on northern tributaries of the Juruá. Some remain there, but many moved further north during the twentieth century, into other river valleys. In a collection of Kanamari myths published by the Lutheran NGO Comin in 2007, the schoolteacher Paranem Manoel Kanamari from the Xeruã, a southern tributary of the Juruá, reports this myth as follows:

Very long ago, two children were born in a village, a boy and a girl. They grew up together. When the children had grown up, the brother went one night to his sister’s hammock and made love to her. He said nothing; so, the girl did not know that it was her brother who came each night to make love to her. The youth visited his sister only at night. The girl wanted to know who the youth was. She had a clever idea. She put genipapo in a pot and left the pot below her hammock, waiting for the youth in the middle of the night. After they had made love, she marked his face with the genipapo. At daybreak, she discovered that it was her brother who came every night. The pair were filled with shame and left each other. From that day on, the boy became the moon and the girl the sun, and they never met again. This story of the two children brought us a very important thing. They transformed into the light that always shines on us: the girl lights up the day, the boy lights up the night (Sass (org.), 2007: 53-54).

These three myths are clearly very similar. Obviously, Carolina’s story and the recently-recorded Cashinahua myths contain an element absent from the Kanamari ones: the visit to the enemies and the hero’s decapitation.

It might seem that Carolina simply told Nimuendajú a story common to indigenous communities of the Juruá-Purús region. She did not. Quite by chance, we know a remarkable amount about the mythologies of the Cashinahua and Kanamari peoples in the early 20^th century, and we know that neither of these peoples were then telling Carolina’s story. Instead, they were telling very different stories about the moon. It is only much later that they started to tell Carolina’s story, the Kuniba version.

THE EARLY 20^TH CENTURY VERSIONS

The Cashinahua data comes from the book Rã-txa hu-ni-kui. A lingua dos Caxinauas, published in 1914 in Rio de Janeiro by the polymathic historian João Capistrano de Abreu. This book is a remarkable document. Its title is bilingual, in Cashinahua and Portuguese, and means The Language of the Real People: the language of the Cashinahua. It consists of a long series of texts in the Cashinahua language, myths and others, along with a grammatical description and short dictionary of that language. The data derive from two young Cashinahua men from the Juruá valley, Bôrô and Tuxinĩ, who had travelled to Rio at Abreu’s request.

Bôrô and Tuxinĩ told Abreu three versions (two in Cashinahua, one in Portuguese) of the myth of origin of the moon very similar to that of Carolina but with this very salient difference: none mention B/Z incest. The initial scene of the moon-to-be and his sister, and the marking with genipa, is simply absent. Instead, both Bôrô’s and Tuxinĩ’s versions of this variant start with the war with the enemies, and the moon-to-be’s fate is motivated solely by his approach to, and decapitation by, those enemies, leading to his attempt to return home in monstrous form, his rejection by his kinspeople, and his decision to become the moon.

Bôrô also told Abreu a remarkably different version, although one which again contains no mention of incest. Here the moon-to-be is female, a young woman who refuses to marry. Her mother sends her away, and she wanders for a long time, weeping. When she returns, her mother refuses her entry to the house, saying, “You can sleep outside. That will teach you to not to want to get married!”. When the daughter insists on entering the house, the infuriated mother cuts off her head with a machete. The mother throws the body in the river, but the head rolls and moans around the house all night. As in the other versions, the head then resolves to become the moon and asks the mother for cotton thread, which it receives, and with the help of a vulture, it enters the sky and becomes the moon (see Lévi-Strauss, [1968]1978: 97).

Turning to the Kanamari case, a contemporary of Abreu, the French priest Constantin Tastevin lived in the Juruá area between 1905 and 1926, and produced a remarkable amount of primary linguistic and ethnographic description for this area, much of it still unpublished. His data is especially rich on the Kanamari people. As I will discuss further below, the Kanamari were and are divided into a series of ideally endogamous subgroups, each named after a totemic animal plus the classifier -djapa: this division was geographically instantiated in the location of subgroups along the east-west flow of the Juruá river. While all the subgroups shared the myths of origin of the world associated with the heroes Tamakori and Kirak, who are also associated with the east-west flow of the Juruá river, each subgroup had a distinctive vision of the moon. I summarize Carvalho’s (2002) account of Tastevin’s data as follows:

1. Amɛna Djapa version: The moon is the nasal ornament of the mythic creator hero Tamakori that gets bigger then smaller and then dies, and is replaced by a new one.

2. Wiri Djapa version: the moon is a married couple, the man called Dyuruyã and the woman Apohanyã, who sit in the sky singing.

3. Ben Djapa version: the moon is a young man, the son of Tamakori, who gets bigger then smaller then dies, and is replaced by the creator hero with a new son.

4. Unknown subgroup version: the moon is a beautiful young girl who swells up with retained menstrual blood, which causes her to become old.

Therefore, in the early 20^th century, the theme of B/Z incest in relation to the moon was absent from the mythologies of the Cashinahua and Kanamari peoples, while almost a century later it was dominant. Carolina’s story was spreading, and being borrowed by neighbouring peoples.

On the face of it, one might expect that in the early 20^th century, when the Kanamari and Cashinahua peoples were living closer together, and in contact mediated by the Kuniba, their myths would be more alike, and that as they came to lose contact with each other, and especially as they lost the mediation of the Kuniba, their myths would have increasingly become more different. That is a standard model based on accounts of biological and linguistic diversification, and of information transmission in general. But this is not what we find in this case. Instead, as the Kanamari and Cashinahua have moved further away from each other, and with the removal of the mediation of the Kuniba, their myths of the origin of the moon have grown to resemble each other more and more.

The case gets stranger when we realize that the early Cashinahua and Kanamari variants were collected after the effective removal of the Kuniba from the local scene in 1912. The Cashinahua and Kanamari people therefore ‘borrowed’ this myth from a former community of neighbours only after that community had disappeared. This seems utterly counter-intuitive, in defiance of all models of information transfer and hence of the Second Law of Thermodynamics. And it is possibly even stranger yet, for there is good evidence that the Cashinahua only took up the Kuniba variant in the late 1950’s. We therefore have a problem in informational causality.

This problem in causality is somewhat ressolved by the continued presence, among both Kanamari and Cashinahua peoples, of a few Kuniba people after the massacre and removal. The SPI sources show that three Kuniba people elected to remain in the Juruá area. Tastevin himself notes that at least two Kuniba remained among the Kanamari following the massacre in 1912 and the removal of the survivors: the tribe’s shaman, named Amador, and a woman named Kiama. These latter two individuals were presumably alternative names for the ones in the SPI records. For the Cashinahua, we have one piece of evidence of post-1912 Kuniba-Cashinahua coresidence: Tastevin’s diary describes a Kuniba man called Moysés living in the rubber post Porto Sérgio with his Cashinahua wife in the 1920’s.

So, while there are potential vectors of transmission from the Kuniba to the Cashinahua and Kanamari post-1912, we are still faced with a significant problem. Why did the Kuniba version of the myth become so widespread among the Cashinahua and Kanamari during the 20^th century? How could a handful of surviving Kuniba people, perhaps only three of them, have imposed their version of a myth on two neighbouring peoples, who spoke languages unrelated to Kuniba and to each other, and who were, at the time, effectively moving out of contact with each other?

As far as we know, this phenomenon applies only to the myth of origin of the moon. Obviously, this is the only Kuniba myth we have knowledge of, but there is no evidence in the extensive literature that Cashinahua and Kanamari mythologies were becoming more alike in general over the 20^th century: both mythologies show clear changes, some dramatic, but convergence is apparently occurring only with respect to the myth of origin of the moon.

THE JURUÁ PEOPLES IN THE EARLY 20^TH CENTURY

In the early 20^th century, the Kanamari, Kuniba and Cashinahua were neighbours in the Juruá basin, but they were neighbours of a particular kind. As Lévi-Strauss noted long ago for indigenous Amazonian peoples in general, the Kanamari, Kuniba and Cashinahua peoples were internally organized in ways that necessarily implicated their neighbours, such that internal social forms were diminished, domesticated, versions of their foreign policies, that is, of their relations with their neighbours. This feature of such peoples has been much obscured by the actual practice of ethnographic fieldwork in the area since the 1950’s, which has focussed on each people as a unique totality: “the Kanamari” or “the Cashinahua”. However, these recent ethnographies are of such high quality that it is easy to explore in some detail this isomorphism of the social interior with the social exterior.

In recent times, both the Kanamari and the Cashinahua peoples live in communities scattered over relatively large areas, often far from their pre-rubber industry territories. These movements have, on the whole, caused the Kanamari and the Cashinahua to move away from each other, a pattern exacerbated by state bureaucratic structures. Modern Kanamari communities are all in the state of Amazonas in Brazil, while modern Cashinahua communities are either in the state of Acre in Brazil, or in Peru. As such, there are very few bureaucratic spaces that would cause Kanamari and Cashinahua to come together, and there is no evidence that they do. That said, the international border between Peru and Brazil does little to minimize contacts between Cashinahua communities located on either side of it. Further, the Kanamari retain a strong interest in the Cashinahua, who are their paradigm of Dyapa, “enemies”. By contrast, there is no evidence that the modern Cashinahua show any special interest in the Kanamari.

As noted before, the Kanamari were and are organized into a series of endogamous subgroups, each named for a totemic animal species plus -dyapa, “group, enemy group”, such as the aforementioned Amɛna Djapa, Wiri Djapa and Ben Djapa. According to one of their ethnographers, Luis Costa, each Kanamari -dyapa subgroup was associated with a particular northern tributary of the Juruá river (Costa, 2007). Members of the same -dyapa subgroup considered each other as -wihnim, “kinspeople”, whether close or distant, while members of other -dyapa subgroups were considered as oatukuna, “other humans, strangers”, that is as other people whose co-humanity was recognized but not socially actualized. Mediating between such “kinspeople” and “strangers” was a category of “friends” (male-to-male-speaking: -tawari; female-to-female-speaking, -tawaro).

Costa’s account of the Kanamari makes no mention at all of the Kuniba people, perhaps because he worked among communities living on the Itacoaí river, a tributary of the Javarí, and hence at a considerable distance from their former territories along the Juruá to the south (Costa, 2007). Carvalho (2002) worked with Kanamari communities along the upper Jutaí river, very close to both their former territories and to those of the Kuniba (it is on this river that the massacre happened). Her ethnographic account has shown the considerable importance of the Kuniba people in the self-conception of at least some Kanamari. She records that her Kanamari informants spoke about these people often, and recorded the following comment: “There are two lost families that we have, Kuniba and Amɛna. The latter one lived on the Pau d’Alho river. When the white people arrived, they wiped it out completely” (Carvalho, 2002: 72, n. 31). The Amɛna-Djapa were unquestionably a Kanamari-speaking subgroup, while the Kuniba obviously were not. This informant, however, clearly included the Kuniba within the moral universe of Kanamari -dyapa subgroups, not as Dyapa, “enemies” like the Cashinahua, but as oatukuna, “other humans, strangers”, and as actual or potential -tawari, “friends”. Indeed, in the Kušitineri language, which was apparently identical to Kuniba, the word for “non-Kuniba, other indigenous people” was recorded by Tastevin as, precisely, tawari.

Figure 4
Kanamari spatial model

Costa’s account is from the twenty-first century, and from a river system to the northwest of the Juruá. However, travellers describe a very similar situation in the middle of the nineteenth century, as did Tastevin in the early twentieth century.

The Cashinahua, according to Deshayes and Keiffenheim, experience themselves as Huni Kuin, “Real People” as opposed to nawa, “Enemies”, in a directly spatial and indeed a geographic manner. “Real People”, the Cashinahua, live in the headwaters, in the ridge country of the divisions between river systems. “Enemies” live along major rivers, or on the banks of the sea. In between live the Xutanava, the “Namesake Enemies”, the Yaminahua people, speakers of a closely related but mutually unintelligible language, and who share the same name-transmission system and name sets as the Cashinahua. In Cashinahua conception, the Yaminahua are key mediators, both spatially and socially, with the “enemies”, for it was through the Yaminahua that Cashinahua people accessed nawa goods. Yami, in some Yaminahua dialects, means “axe”, and the standard translation of their name is “Axe Enemies”. The Cashinahua do not have a very elaborate taxonomy of types of nawa, “enemies”, so it is unclear whether they recognized the Kuniba and Kanamari as separate and distinct peoples.

Figure 5
Cashinahua spatial model

Deshayes and Keiffenheim’s data comes from Cashinahua people in the late 20^th century, and indeed from people living quite far from their territory in the early 20^th century. But, again, this account does seem to correspond to the earlier situation.

TOWARDS A KUNIBA SPATIAL MODEL

Given all the things that we do not and will never know about the Kuniba people, it will come as no surprize to learn that we know virtually nothing about how they saw themselves in this social landscape. There are, obviously enough, no recent ethnographies of these people that would correspond with the studies of Costa and Carvalho or Deshayes and Keiffenheim. That said, Carolina’s story provides some clues, which are supported by Chandless’ account.

Carolina’s story opposes two kinds of journeys, and their relations to the interior of Kuniba society. The Kuniba version begins in a manner unknown from all later variants, and this mise-en-scène is actually absent from Lévi-Strauss’ summary in The Origin of Table Manners. For unknown but intriguing reasons, Nimuendaju truncated the story in his letter to Baldus. Carolina’s story in fact starts as follows: “A man had gone on a journey, leaving his wife at home. In the absence of her husband, each night she was visited by an unknown man in her hammock” (Nimuendaju, 1986: 89). We know, from Chandless and from other evidence for Piro-speaking peoples, what the husband’s journey meant. The Kuniba were long-distance traders. Therefore, the Kuniba had two valences of social distance and its attendant journeys: towards trade allies, and towards dangerous enemies.

The available documentary archive allows us to specify the directionality of these movements. The major trade route of the Kuniba was overland to the Manchineri people on the Purús river to the south. Chandless reports that the Kuniba originally travelled for trading purposes to the Purús river via the Juruá’s tributary the Tarauacá and then by its tributary the Envira and then overland to the other river basin. However, by the time of his visit, they had had to abandon that route due to attacks by the Nawa of the Upper Juruá, and had to follow a more easterly alternative. In the mid-19^th century, the Kuniba did not travel far up the Juruá river for fear of these same Nawa people.

The journey of the moon-to-be’s brother-in-law, which begins Carolina’s story, would therefore have been a journey by river, south towards the Manchineri people on the Purús, while the enemy village of the story, to which the moon-to-be runs away, is therefore likely to have been located among the Nawa people upriver on the Juruá. It is clear from the myth and the known context that the first journey is by canoe, the second on foot. Contemporary Urubamba Piro people would distinguish these two journeys as “to the other side/to the other river basin”, on the one hand, and “upriver/ into the forest”, on the other. A journey “to the other side”, for the Kuniba, was towards the Manchineri, trade partners and fellow kinspeople, while a journey “into the forest” is into the world of the enemy Nawa people. It is therefore likely that the Kuniba distinguished between journeys to the “other side/towards trading partners”, and movements “upriver/into the forest/towards Nawa “enemies””, as do the present-day Urubamba Piro.

It does not seem to be the case that the Kuniba identified the Cashinahua with the Nawa “enemies”. Kuniba traders had historically travelled close to Cashinahua territory as they went back and forth to the Purus river along the Tarauacá river and its tributaries, clearly without real fear of attack. The Nawa “enemies” they feared and avoided lived upriver on the Juruá.

In the Kuniba spatial model there is a third possible journey, unmentioned in Carolina’s story: a trading expedition downriver to the land of the payri, “white people”. By the mid-19^th century, such journeys seem to have largely ceased, and following Chandless, the Kuniba traded with white visitors who travelled to their homes (Chandless, 1869). By Carolina’s time, such journeys were probably no more than a vague memory, given the extent to which white bosses had come to dominate all commercial transport in the area. But, as we have seen, Carolina told her story precisely in the context of her own dramatic downriver journey to the world of the payri.

Figure 6
Kuniba spatial model

CAROLINA’S STORY AND INDIGENOUS FOREIGN POLICY IN THE JURUÁ-PURÚS

If we look back at Figures 4, 5 and 6, they look like societies in the classical definition of bounded units. But clearly they are not, because in each case the apparent ‘boundaries’ express not frontiers between “Self” and “Other”, but rather a specific point of transition along a gradient that goes from “Self” towards the “Other” and back again. As Viveiros de Castro (2002) has argued, for indigenous Amazonian peoples, the Self is precisely the Other of the Other. “Real People” know themselves to be “Real People” because they are seen as “Enemies” by their “Enemies”.

All the stories discussed here concern the moon. For the peoples of the JuruáPurús area, the moon is the celestial operator of these socio-spatial relations of Self and Other. In this area, all the major rivers follow a basic southwest to northeast trend in flow, one which is generally closest to a west-east orientation, and as such close to the apparent movements of the sun and moon in the sky. The Juruá and Purús rivers rise were the sun and moon set, and flow down towards where the sun and moon rise.

Carolina’s story and the other myths dealt with here concern the origin of the moon. As a celestial body, the moon is in marked contrast to the sun. While the moon shares with the sun the daily transit from east to west, it also shows a distinctive west to east movement in its time of rising and setting on a monthly basis: the new moon is first seen in the western sky, while the waning moon is last seen in the eastern sky. As such, in the Juruá-Purús area, the moon synthesizes both directions of riverine motion, downriver and upriver, and as we have seen, it is these riverine directionals that lie at the core of local models of social space. The moon therefore, in its contrasting daily and monthly transits across the sky, replicates on a celestial level the river-based spatial modelling of Self/Other relations among the indigenous peoples of this area.

The moon is therefore the celestial counterpart to the riverine flows that govern the spatial relations between these neighbouring peoples. As such, the moon has an intrinsic significance for what Lévi-Strauss called the ‘foreign policy’ of the indigenous peoples of this region (Lévi-Strauss, 1949). By foreign policy, Lévi-Strauss meant the specific relationship of a society to its exterior conditions of existence in a field of other societies, and indeed any society’s attitude towards its neighbours. As he also noted, indigenous Amazonian peoples are characterized by the isomorphism of their social interior and exterior, such that each replicates and models the other, a theme developed in greater detail by Viveiros de Castro (2002).

A CELLULAR STRUCTURE

The myths collected by Capistrano de Abreu, Tastevin and Nimuendaju can be connected up to the socio-geographical system of the Juruá-Purús area discussed above. The early Kanamari versions are characterized by their heterogeneity, which directly corresponds to their association with specific endogamous sub-groups located on specific north bank tributaries of the Juruá river. Within the earlier Kanamari system, the Kuniba myth is simply another sub-group tradition. Here, Carolina’s story does not really stand out as special: it is simply one more story about the moon in a line of such stories strung along the Juruá river. The flow of this river, and the spacing of subgroups along it is, as we have seen, central to Kanamari sociology and cosmology.

Indeed, there is evidence that the differences between the early Kanamari versions of the moon are structured by this subgroup meaning of the flow of the Juruá river. From Carvalho’s (2002) map, following Rivet and Tastevin, we have the following sequence, upriver to downriver, on the Juruá:

Ben Djapa => Wiri Djapa => Kuniba

Therefore, we have the upriver-to-downriver sequence of the earlier Kanamari/ Kuniba versions:

1. Ben Djapa version, further upriver: the moon is a young man, the son of Tamakori, who gets bigger then smaller then dies, and is replaced by the creator hero with a new son.

2. Wiri Djapa version, in between: the moon is a married couple, the man called Dyuruyã and the woman Apohanyã, who sit in the sky singing.

3. Kuniba version, further downriver: where the moon is the secret incestuous lover of his sister.

It is significant that the Wiri Djapa version is both more similar to the Kuniba version of the two locatable Kanamari versions, and the geographically closer one as well. Indeed, in a footnote to Tastevin’s account of the Wiri Djapa version, Carvalho writes,

In a side note, Tastevin observed that Apohanya means “blowing, that which blows”. I suspect that Tastevin is mistaken and that the name refers to ypoanya, female parallel cousin, which makes complete sense in the context of the Wiri version, largely shared today, of the moon as the result of an incestuous union and which is, in general lines the same version of the myth that Nimuendaju collected in 1921 and identified as Kuniba (Carvalho, 2002: 283, n. 205).

It seems likely to me then that in the early twentieth century, the Kuniba version of the myth of the moon was set into Kanamari subgroup structure, such that each subgroup, including the Kuniba, had its own version of their ideas about the moon, ordered sequentially by their locations along the west to east flow of the Juruá river, and fully coordinated with the east to west and west to east movements of that celestial body.

That said, the Kuniba myth clearly points away from the Kanamari versions towards a different spatial axis, that of the enemies. In recent Kanamari accounts, “enemies”, Dyapa, refer to the Cashinahua, and this pattern is quite old. Thus, the enemies are located far to the south, away from the definitional flow of the Juruá river, along which the cosmogonic acts of the Kanamari creator heroes Tamakori and Kirak took place. I suspect that the Kanamari were unable to adopt the “visit to the village of the enemies” sequence of the Kuniba myth, and the whole story about the decapitation and return home, precisely because it would involve a journey towards the south, and hence falling outside of their spatial understanding of the relation between the moon, the Juruá river and their own social forms.

Turning now to the Cashinahua, their early twentieth century versions of the myth of the moon differ in two major senses from the Kuniba myth. Either, 1) they make no mention of the B/Z incest or the marking of the moon’s face with paint, and launch directly into the moon-to-be’s fateful journey to the enemies’ village. Or, 2) they invert the sex of the moon-to-be, and locate the key action within a domestic setting and within the intimate kin relations of a mother and her daughter.

In spatial terms, the Cashinahua locate themselves in the headwaters, such that any movement away from their communities is necessarily a movement downriver, and so the nawa, “enemies”, are always downriver. From the early 20^th century Cashinahua perspective, the Kuniba myth is a kind of synthesis of their two mythic possibilities. It begins within a domestic setting and within intimate kin relations, but now transformed from the relationship between a mother and daughter to that between a brother and sister, then moves to the relations between the male moonto-be and the enemies. One might say that while early 20^th century Cashinahua versions of the myth were vacillating between the two possibilities of a male moon and a female moon, the Kuniba achieved a dialectical resolution of the problem by integrating both possibilities, and innovating the unactualized possibility of a female moon in the form of the male moon’s sister and lover: the sister of a male moon is implicitly a female moon.

We therefore have two sets of sequences for the early twentieth century data. The Kanamari sequence emphasized that each sub-group along the Juruá river had a variant of the nature of ideas about the moon, and that such variation is graded or clinal. The Cashinahua sequence emphasized a binary choice between the sexes, male or female, of the moon. In both cases, the Kuniba myth acted as a dialectical synthesis of that sequence. This synthesis was not temporal in the Hegelian sense, but instead spatial: a Hegelian, temporal, dialectic has only one synthesis; while a non-Hegelian, spatial, dialectic has many. Under normal historical circumstances, such spatial dialectical resolutions saturate a given social field.

A GAP IN A PLENITUDE

The question then becomes: what happens when a specific spatial dialectical synthesis disappears for reasons that have very little to do with the existing social dynamics of that social field? What happens when the spatial dialectical synthesis is confronted with genocidal violence and state-sponsored population removal? In the case in hand, what happens to the wider system when the Kuniba people are removed? The answer, at least in this case, is that the two remaining peoples in the social field have to take the destroyed synthesis into themselves.

As a complex switchpoint between early 20^th century Kanamari and Cashinahua myths about the moon, and consequently ideas about foreign policy, the relations between neighbouring peoples, the Kuniba myth occupied a key mediating role in the local mythological systems. It was both a totemic version of Kanamari ideas within a wider totemic system, and a dialectical synthesis of two distinct Cashinahua mythic possibilities. As long as the Kuniba were around, telling their variant of the myth of origin of the moon, the neighbouring Kanamari and Cashinahua peoples were able to tell their own variants. When the Kuniba people were removed from that social landscape, it seems, both of their neighbours found themselves obliged to start telling versions of the Kuniba story.

The key shift in Cashinahua and Kanamari versions of the myth of origin of the moon over the twentieth century was the emergence of the theme of brother/sister incest. In both cases, therefore, the origin of the moon was radically shifted into the heart of the social world, in the intimate nocturnal sexual relations of a brother and a sister. For both the Cashinahua and Kanamari in the late 20^th century, the origin of the moon has come to lie within the immediate community, in the incestuous desire of a “Real Human” for his sister.

In the early 20^th century, the Kuniba variant was implicitly present within the mythologies of both the Cashinahua and the Kanamari. Bôrô’s other version, where the moon is female, concerns a girl who is far too fearful of her potential spouses, who she treats as enemies. She is therefore the inversion of the Kuniba brother, who treats his sister as if she was his potential spouse. Equally, the early Kanamari variant from the Wiri Djapa subgroup alludes to a married couple, who are implicitly parallel cousins, that is, a less marked form of brother and sister. These early versions from both peoples are clearly ‘weak’ in relation to the ‘strong’ Kuniba variant, but are, so to speak, semantically headed in that direction.

I think that the reason that the early 20^th century Cashinahua and Kanamari peoples did not tell the myth of origin of the moon in the Kuniba manner was precisely because the Kuniba people did do so. That is, the early twentieth century Cashinahua and Kanamari peoples were relieved of the necessity to tell that story in that way because it was known to them in its Kuniba variant. It existed for both peoples as a Kuniba story, and its semantic content was available as a quotation, “The Kuniba people say that, but we say this.” As the ability to quote the Kuniba people evaporated with their effective local extirpation, this strategy was no longer possible. Both the Cashinahua and Kanamari mythologies were forced to do what only the Kuniba mythology was previously willing to do: locate the origin of the moon in the transgression of the incest taboo, the transgression of the very constitution of their own social interiors.

Because they did not have to tell the Kuniba variant of the origin of the moon, the early 20^th century Cashinahua and Kanamari peoples were able to explore other mythic possibilities. In the Cashinahua case, or at least the case of Bôrô, this allowed the possibility of telling of two very different myths, focussed on the separate potentials of whether the key protagonist is male or female. In the Kanamari case, the presence of the Kuniba variant allowed each subgroup to tell their own version of what the moon was, and to use this cosmological variability as a form of subgroup differentiation at a cosmological level.

CONCLUSION

What are the sociological correlates of this phenomenon? What does it have to say to social anthropology?

One of the problems in most social anthropological approaches to myth is to imagine these narratives primarily as commentaries on something else: they are a community’s conscious reflection on its own social forms. This is the position of what the French call “Anglo-structuralism”. That, of course, is hostile witness, so let me quote Pierre Clastres:

That the myths think themselves among each other, that their structure can be analysed, is certain. Lévi-Strauss brilliantly provides the proof, but it is in a secondary sense, for they first consider the society which considers itself in them, and therein lies their function. Myths make up primitive society’s discourse upon itself; they have a sociopolitical dimension that structuralist analysis naturally avoids taking into account lest it break down. Structuralism is only operative on the condition of cutting the myths from society, of seizing them, ethereal, floating a good distance from the space of origin (Clastres, 1978 [2004]: 201-202).

This is clearly totally inadequate to the current case, since under what possible historical conditions would a community abandon its own commentary upon its own social forms for a neighbouring community’s commentary on its different social forms?

It seems to me that this idea that myths are a commentary on a community’s social forms is a fundamental misunderstanding of what myths are for indigenous Amazonian peoples, and probably of myths more generally. At least for indigenous Amazonian peoples, myths are primarily intergenerational social transactions, and especially highly marked social transactions between grandparents and grandchildren. Their social transaction does not take the form of pedagogy, but rather the active constitution of the myth-teller as a grandparent, and the reciprocal active constitution of the myth-hearer as a grandchild. Myths are primarily phenomena of kinship.

For the indigenous peoples of Southwestern Amazonia, what constitutes a society is also primarily a phenomenon of kinship. Our Euro-American concept of Society is not how such people would formulate the issue. Their formulations are concerned with ideas about “Real Humans”, people among whom it is appropriate to live and to engage in ongoing everyday life. “Real Humans” are necessarily contrasted to “Unreal Humans”, people with whom it is impossible to live well and engage in ongoing everyday life. This contrast between “Real Humans” and “Unreal Humans” is not absolute, but rather transitive: to be ongoing as a project, “Real Humans” must contain a form of “unreality” or Otherness. “Real Humans”, people just like oneself, are in effect same-sex siblings to each other, and hence unmarriageable. To become marriageable, a relationship between two “Real Humans” must contain the seeds of difference, and this is most commonly the gender difference between an adult brother and sister. While sexual relations between such a brother/sister pair are, as we have seen, the paradigm of incest, their respective children, as cross-cousins, are the ideal marital partners.

As Viveiros de Castro has pointed out, in his essay “GUT Feelings about Amazonia: Potential Affinity and the Construction of Sociality” (2001), Amazonian kinship does not have a frontier, a point at which kinship simply stops and non-kinship begins. Instead, Amazonian kinship is characterized by elaborate spatial gradients of “close” versus “distant”, but with no mapping of the spatial gradient onto genealogy in the Euro-American sense. As one moves away from the community of everyday life, kin relations become increasingly affinalized, as is minimally true between an adult brother and sister, out towards actual affines, such as brothers-in-laws, towards strangers, who are virtual affines, and onwards out towards enemies, animals, the dead, powerful beings, the moon, and so forth, in an escalating series of potential affinity. At the edges of the cosmos, all beings are fully affinalized. The return journey is the inverse: affinity is progressively expelled from relations, rendering them increasingly consanguineal. Consanguinization does not stop at the relation between same sex siblings, but continues into the realm of what we might want to call the individual: in the case discussed here, the relation between the dead body and its decapitated head.

As such, ethnographically known collectivities such as the Kuniba, Cashinahua or Kanamari do not conceive of themselves as separate in a genetic sense, as diversifications from a temporal point of common origin, but rather in a spatial sense. That is, this community of “Real Humans”, where we live, is the localized precipitate out of a universe of otherness.

What are myths in this scheme? In the indigenous languages of the region, myths are called “ancient people’s stories, the stories of the ancestors”. It would be easy to misinterpret that definition, and to give it a genetic rather than a spatial interpretation. Myths are not strictly “the stories of the ancestors” in a genetic sense, the parallel reproduction of ancestors and of their stories. Myths are the ancestors. Ancestors are not temporally distanced, but rather spatially distanced. The myth about the moon is not primarily a Kuniba, Cashinahua or Kanamari myth. It is primarily the moon’s story about himself or herself, told from a position that is epistemologically completely inscrutable.

In the “Overture” to the Mythologiques, Lévi-Strauss wrote:

Durkheim has said (p. 142) of the study of myths: “It is a difficult problem which should be dealt with in itself, for itself, and according to its own particular method.” He also suggested an explanation of this state of affairs when later (p. 190) he referred to the totemic myths, “which no doubt explain nothing and merely shift the difficulty elsewhere, but at least, in so doing, appear to attenuate its crying illogically.” This is a profound definition, which in my opinion can be extended to the entire field of mythological thought, if we give it a fuller meaning than the author himself would have agreed to ([1964]1970: 5).

As Nancy Munn brilliantly noted about Walbiri myths, who is the witness to mythic action? The answer is: no one. Who is this “no one”, this non-existent person who saw the moon becoming the moon? Society? I think not. We need a new sociology.

• 1
  T. N.: Gow refers to “M392: Kuniba, the rolling head and the origin of the Moon”, which was analysed by Lévi-Strauss in The Origin of Table Manners: Introduction to a Science of Mythology 3 ([1968]1978: 94.)
• 2
  T. N.: The map Gow is refering to is in the first edition of 1907’s Peru versus Bolivia.
• 3
  T. N.: The original text is quite laconic, but we can understand Gow’s critique of functionalism by referring to an excerpt of Lévi-Strauss in The Naked Man: “It is high time that anthropology freed itself from the illusion gratuitously invented by functionalists, who mistake the practical limitations imposed upon them by the kind of studies that advocate for the absolute properties of the object which they are dealing. An anthropologist may confine himself for one or more years within a small social unity, group or village, and endeavour to grasp it as a totality, but this is not reason for imagining that the unity, at levels other than the one at which convenience and/or necessity has placed him, does not merge in varying degrees into larger entities, the existence of which remains, more often than not, unsuspected." ([1971]1981: 609).

REFERÊNCIAS BIBLIOGRÁFICAS

• ABREU, João Capistrano de. 1941. Rã-txa hu-ni-ku-i: Grammatica, textos e vocabulário Caxinauás Edição da Sociedade Capistrano de Abreu.
• CARVALHO, Maria Rosario Gonçalves de. 2002. Os Kanamari da Amazônia Ocidental. História, mitologia, ritual e xamanismo Salvador: Ed. Casa de Palavras.
• CHANDLESS, William. 1986. “Notes of a Journey up the River Juruá”. Journal of the Royal Geographical Society of London, Vol. 39: 296-311.
• COSTA, Luiz. 2007 As Faces do Jaguar: Parentesco, História e Mitologia entre os Kanamari da Amazônia Ocidental Rio de Janeiro, Tese de doutorad, Museu Nacional da Universidade Federal do Rio de Janeiro.
• CUNHA, Euclides da. 1907. Peru versus Bolívia Rio de Janeiro, Livraria Francisco Alves.
• DESHAYES, Patrick; KEIFENHEIM, Barbara. 2003. Pensar el Otro entre los Huni Kuin de la Amazonía Peruana Trad. Balbina Vallejo. Lima: CAAAP.
• LÉVI-STRAUSS, Claude. 2019 [1949]. “La politique étrangère d’une societé primitive”. In: DEBAENE, Vicent (ed.). Anthropologie Structurale Zéro Paris: Éditions du Seuil: 201-219.
• LÉVI-STRAUSS, Claude. 2004 [1964] O Cru e o Cozido (Mitológicas I) Trad. Beatriz Perrone-Moisés. São Paulo: Cosac Naify.
• LÉVI-STRAUSS, Claude. 2006 [1968]. A origem dos modos à mesa (Mitológicas III). Trad. Beatriz PerroneMoisés. São Paulo: Cosac Naify.
• LÉVI-STRAUSS, Claude. 2011 [1971]. O homem nu (Mitológicas III) Trad. Beatriz Perrone-Moisés. São Paulo: Cosac Naify.
• MCCALLUM, Cecilia. 2001. Gender and Sociality in Amazonia. How real people are made Oxford: Berg.
• MUNN, Nancy. 1973. Walbiri Iconography: Graphic Representation and Cultural Symbolism in a Central Australian Society Ithaca: Cornell University Press.
• NIMUENDAJU, Curt. 1986. “Mitos Indígenas na Obra de Curt Nimuendaju”. Revista do Patrimônio Histórico e Artístico Nacional n. 21: 64-111. Tradução: Marcos de Almeida Matos Revisão: Renato Sztutman
• SASS, Walter (org.). 2007. Tâkuna. Nawa Bûh Amteiyam Amkira. Mitos Kanamari. São Leopoldo: Ed. Oikos & COMIN.
• TASTEVIN, Constant; RIVET, Paul. 1923/1924 “Les langues du Purús, du Juruá et des régions limitrophes. Io Le groupe arawak pré-andin (Fin)”. Anthropos, Bd. 18/19, H. 1./3.: 104-113.
• VIVEIROS DE CASTRO, Eduardo. 2001 “GUT feelings about Amazonia: potential affinity and the construction of sociality”. In: RIVAL, Laura; WHITEHEAD, Neil (eds), Beyond the visible and the material. The Amerindianization of society in the work of Peter Rivière Oxford: Oxford University Press: 19-43.
• VIVEIROS DE CASTRO, Eduardo. 2002 A inconstância da alma selvagem e outros ensaios de antropologia São Paulo: Cosac & Naify.

Publication Dates

• Publication in this collection
  06 July 2022
• Date of issue
  2022

History

• Received
  16 June 2021
• Accepted
  21 Sept 2021