The Anthropology Of Marriage In Lowland South America: Bending And Breaking The Rules [1st Edition] 0813054311, 9780813054315, 0813053064, 9780813053066, 0813052890, 9780813052892

Table of contents :
Cover......Page 1
The Anthropology of Marriage in Lowland South America......Page 2
Title......Page 4
Copyright......Page 5
Contents......Page 6
List of Illustrations......Page 8
Introduction......Page 10
PART I......Page 22
1. Marriage Matsigenka Style: Some Critical Reflections on Theories of Marriage Practices......Page 24
2. Marriages, Norms and Structures: The Dilemma of Finding a Wife among the Piaroa of the Sipapo......Page 45
3. To Be Seen or Not to Be Seen! Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon......Page 64
PART II......Page 80
4. Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage......Page 82
5. Beyond the Norms: Marriage and Incest among the Ye’kwana......Page 94
6. Why Do the Ye’kwana Commit Incest So Frequently? A Discussion of Silva’s “Beyond the Norms”......Page 109
7. Why Do Women Run Away? Matrimonial Strategies among the Yanomami......Page 133
PART III......Page 164
8. “Poor Me, I Have No Cousin”: The Pragmatics of Marital Choice in the Northwest Amazon......Page 166
9. Why Was There a Transition from an Elementary Kinship Structure to a Complex One? A Short Ethnography of an Amazonian Village......Page 189
10. Changes in Canela Marriage over 70 Years: From Authorizing to Stealing......Page 230
11. Waorani Marriage......Page 256
Bibliography......Page 280
List of Contributors......Page 300
Index......Page 304

Citation preview

The Anthropology of Marriage in Lowland South America

•

University Press of Florida Florida A&M University, Tallahassee Florida Atlantic University, Boca Raton Florida Gulf Coast University, Ft. Myers Florida International University, Miami Florida State University, Tallahassee New College of Florida, Sarasota University of Central Florida, Orlando University of Florida, Gainesville University of North Florida, Jacksonville University of South Florida, Tampa University of West Florida, Pensacola

Lowland South America. Map by Samantha Elgersma.

The Anthropology of Marriage in Lowland South America Bending and Breaking the Rules

• Edited by Paul Valentine, Stephen Beckerman, and Catherine Alès

University Press of Florida Gainesville · Tallahassee · Tampa · Boca Raton Pensacola · Orlando · Miami · Jacksonville · Ft. Myers · Sarasota

Copyright 2017 by Paul Valentine, Stephen Beckerman, and Catherine Alès All rights reserved Printed in the United States of America on acid-free paper This book may be available in an electronic edition. 22 21 20 19 18 17

6 5 4 3 2 1

Library of Congress Cataloging-in-Publication Data Names: Valentine, Paul, 1942– editor. | Beckerman, Stephen, editor. | Alès, C. (Catherine), editor. Title: The anthropology of marriage in lowland South America : bending and breaking the rules / edited by Paul Valentine, Stephen Beckerman, and Catherine Alès. Description: Gainesville : University Press of Florida, 2017. | Includes bibliographical references and index. Identifiers: LCCN 2016049600 | ISBN 9780813054315 (cloth) Subjects: LCSH: Indians of South America—Marriage customs and rites. | Indians of South America—Marriage customs and rites—Amazon River Valley. Classification: LCC F2230.1.M28 A67 2017 | DDC 980/.01—dc23 LC record available at https://lccn.loc.gov/2016049600 The University Press of Florida is the scholarly publishing agency for the State University System of Florida, comprising Florida A&M University, Florida Atlantic University, Florida Gulf Coast University, Florida International University, Florida State University, New College of Florida, University of Central Florida, University of Florida, University of North Florida, University of South Florida, and University of West Florida. University Press of Florida 15 Northwest 15th Street Gainesville, FL 32611-2079 http://upress.ufl.edu

Contents

List of Illustrations vii Introduction 1 Paul Valentine and Stephen Beckerman

Part I 1. Marriage Matsigenka Style: Some Critical Reflections on Theories of Marriage Practices 15 Dan Rosengren

2. Marriages, Norms and Structures: The Dilemma of Finding a Wife among the Piaroa of the Sipapo 36 Alexander Mansutti Rodríguez

3. To Be Seen or Not to Be Seen! Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon 55 Daniela Peluso

Part II 4. Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage 73 François-René Picon

5. Beyond the Norms: Marriage and Incest among the Ye’kwana 85 Nalúa Rosa Silva Monterrey

6. Why Do the Ye’kwana Commit Incest So Frequently? A Discussion of Silva’s “Beyond the Norms” 100 Paul Valentine

7. Why Do Women Run Away? Matrimonial Strategies among the Yanomami 124 Catherine Alès

Part III 8. “Poor Me, I Have No Cousin”: The Pragmatics of Marital Choice in the Northwest Amazon 157 Janet Chernela

9. Why Was There a Transition from an Elementary Kinship Structure to a Complex One? A Short Ethnography of an Amazonian Village 180 Paul Valentine and Lionel D. Sims

10. Changes in Canela Marriage over 70 Years: From Authorizing to Stealing 221 William H. Crocker

11. Waorani Marriage 247 Pamela I. Erickson, Stephen Beckerman, James Yost, and Rosemary Diaz

Bibliography 271 List of Contributors 291 Index 295

Illustrations

Figures Frontispiece. Lowland South America ii 2.1. Piaroan kinship terminology (male ego) 39 6.1. Two residence groups A and B; the 0 generation composed of a brother and sister 111 6.2. Cross-cousin marriage, illustrating the circulation of bride service 116 6.3. Classificatory GP/GC marriage, illustrating the circulation of bride service 116 6.4. Parallel cousin marriage, illustrating the concentration of bride service 117 7.1. The choices facing the ego X–ego Y relationship 132 7.2. The choices facing the male ego–female ego relationship 133

Tables 5.1. 5.2. 5.3. 5.4. 6.1. 7.1.

Data classified into the “Old Period” and the “Modern” 92 Marriage types classified by kin and affinal links 93 The number of prescribed and incestuous (kaiña) marriages 95 Kaiña marriages by village and river basin in the 20th century 95 Kaiña marriages 109 Terms of reference and address for a female ego and a male ego 126 9.1. Marriage by ethnic group in San Carlos 181 9.2. Marriage between immigrant Baré and San Carleño Baré 182 10.1. Thirty-six marriages of Canela women from about 1935 to 2004 227

Introduction Paul Valentine and Stephen Beckerman

This is a book about marriage. Its ethnographic concentration is on lowland South America, and its topical focus is on the bending and breaking of the rules regarding marriage. Marriage rules, and questions of compliance with them, are a venerable topic in social anthropology. We know that marriage universally is buttressed by rules and norms that may be seen as constraining people’s choices. But at the same time, there are sociopolitical forces and ecological and demographic conditions that, together with people’s desires and preferences, influence their marital decisions. It is that play between structure and agency, between how things ought to be and how they are, that is the subject of this book. This topic has not been studied in a comparative way before in lowland South America. Because it is addressed here in terms of the resolution of the conflict between the socially constraining forces of structure and the individual choices of agency, we aspire to invigorate the topic of marriage in lowland South America and also to contribute to the ongoing debate in anthropology on the relationship between norms and practice. As both Mansutti Rodríguez and Picon point out in their contributions, the exceptional marriages, the ones that do not fit the norms, have too long been ignored or labeled deviant or “pathological” because they did not confirm to our generalizations or our structural models. However, if our theories are to have any validity, they must account for the full range of marriage phenomena, both rules and practice. Parallel with the issue of structure vs. agency in lowland South American societies, there is another theme running through this book: the way in which globalization is subverting traditional hierarchies, altering identities, and eroding ancestral marital norms and values—how the forces of modernization alter both structure and practice. In following this theme

2 · Paul Valentine and Stephen Beckerman

we see how the global forces of modernization, expressed in their impact on the Third World, come to work upon so-called Fourth World societies, the indigenous societies that still remain on the planet. It is a vision at once fascinating and more than a little chilling. There is also an intellectual genealogy behind this book. A generation ago Kenneth Kensinger (1984) published his Marriage Practices in Lowland South America, a volume containing 11 essays by well-known South American ethnographers (all but one of them American). With its detailed ethnography and sophisticated analysis, the book was influential among kinship theorists and South Americanists. It broached such subjects as the relations between kin terminology and marriage rules, the distinction between systems having prescribed vs. preferred marriage rules, whether affinal relationships are created by a marriage or are presupposed (and ratified) by a marriage, and so forth. All the authors were aware of the disparity between ideal marriage rules (whether articulated by informants or inferred by anthropologists) and actual marriage practice. Nevertheless, the general thrust of most of the essays was in the direction of elucidating the rules and the way in which they were understood by the people in question. The present work, like the Kensinger volume, is based on papers originally presented at a symposium. The chapters in this collection were to a considerable extent inspired by Kensinger’s compilation, although in their totality they tend to redirect the investigation of marriage in the South American lowlands toward issues less traditional in social anthropology. The thrust here is how individual motives and individual opportunities result in marriage choices that manipulate, circumvent, or outright violate marriage rules. Indeed, some of these essays come close to suggesting that marriage rules themselves are epiphenomenal. In this introduction, following a description of the themes and theories employed by the contributors, is an appreciation of the levels at which people strategize and sometimes deviate from or ignore marital norms to acquire a spouse. There follows a brief outline of the book’s organization.

Theory: Structure and Agency Identifying common themes in the book is straightforward. All the contributors address the same set of issues. Further, all the kinship and marriage systems described, except for the Canela, have or used to have

Introduction · 3

elementary kinship systems. Moreover, all the contributors agree that besides taking into consideration the “normative” kinship and marriage systems sensu stricto, other social spaces are equally relevant, if not more so, to an understanding of why marriage rules are sometimes ignored or manipulated. Sociopolitical, economic, and demographic forces and personal desires may all influence people’s choice of marriage partner. Crocker, Erickson and colleagues, Chernela, and Valentine and Sims also explicitly include the forces of globalization in their analyses. One of the most interesting findings is that irrespective of the theoretical orientation of the anthropologists, all the contributors, with the exception of Erickson and colleagues in their study of the Waorani, work toward the same goal and arrive at similar conclusions. They direct their attention to differentiating and exploring the relationship between the kinship and marriage structures as they ought to be and as they are. Significantly, their conclusions converge: in all the cases considered, there are or were mechanisms to impart flexibility to what appear to be rigid and intractable consanguinal and affinal structures. Rosengren, for instance, sets out to critique Lévi-Strauss’s (1969) formalized analytical models by targeting the way his structural model of reciprocal exchange does not correspond to Matsigenka palpable reality. His is a clarion call to move on from Lévi-Strauss’s grand design and describe “people as intentional subjects situated in the everyday world of their own experience.” He concludes that Matsigenka rules are less a normative system that governs people’s behavior and more a discursive convention. For instance, people who are mutually attracted define each other as crosscousins. It is true that sometimes Matsigenka marry their cross-cousins, but that is more a consequence of coming into contact with them more frequently than with other single people of their own age than a case of obedience to the rule that the Matsigenka marry their cross-cousins. Chernela’s subtle and probing analysis describes the consequences of globalization on Kotiria (also known as Wanano) kinship and marriage. Drawing on the insights of Robert Murphy (1972), she examines a number of the major themes that recur throughout this book, such as the dialectic between rules and practice, the relationship between structural intransigence and agency fueled by needs and desires, and the emergence of innovation and practical considerations. If one looks at the formal model of Kotiria kinship and marriage, one might predict that it allows for no flexibility; that the Kotiria are constrained by a narrow range of alterna-

4 · Paul Valentine and Stephen Beckerman

tives when they choose a spouse. Chernela carefully selects case studies that illustrate the range of possibilities actually open to the Kotiria. They must marry cross-cousins, but there are different kinds of cross-cousin. Mansutti Rodríguez adopts Needham’s (1972) scheme of distinguishing three analytical levels: (1) the jural rules, referring to ideals held by the studied people themselves—for instance, statements of marriage preference belong in this level; (2) the statistical-behavioral, referring to aggregate consequences of individual behavior—for instance, demographic, residential, and marital patterns; and (3) the categorical, made up of modes of classification and systems of nomenclature—the relationship terminology is perhaps the archetypical example. Moreover, Mansutti Rodríguez includes a computer simulation to gain time depth in his model, which provides the opportunity to view irregular behavior not as something residual and inexplicable (as could well have been the case when viewing kinship only in terms of structural models) but as necessary to the preservation of the marriage system. He employs Bourdieu’s (2005) distinction between official and private kinship and illustrates his approach with apposite case studies. Silva had the insight that a significant number of Ye’kwana marriages were incestuous, providing the opportunity to hypothesize why this is so (Valentine, this volume). Based on her extensive field data and Sawyer’s (2005) and Sims’s (2012) theoretical insights, it emerges that although the Ye’kwana have a kinship terminology that articulates unambiguously a set of norms, nevertheless, because each community wishes to retain its population, “wrong marriages” with the parallel cousin are a way of avoiding villagers leaving the village. On their marriage, parallel cousins are immediately reclassified as cross-cousins. Alès, in a tour de force, examines the way Yanomami manipulate the kinship and marriage system. She argues it would be mistaken to use only genealogical relationships for the categorization of marriage practices. There is no exclusive rule. Rather, people select one of several strategies that they apply to a set of structural variations to obtain the outcome they desire. Furthermore, her analysis confirms Silva’s observation of the frequency of Ye’kwana “incest”; among the Yanomami nearly one third of marriages are between classificatory brothers and sisters if the appropriate genealogical paths are taken, and not between classificatory husbands and wives.

Introduction · 5

A significant conclusion can be drawn from these cases. People are nearly always prepared to employ strategies that bend or break the rules to maintain their own ideal conceptual model of the kinship and marriage structure. These strategies reconfigure the elements of the kinship and marriage structure so that their own conceptual model of the structure still has weight. There are many examples of such strategies in this volume. Mansutti Rodríguez’s illustration of the link between demography and alterations in the Piaroa kinship nomenclature is one of the most significant. Alès’s description of parents deciding the relationship terms of the possible spouses for their children is another. She comments, “It is not about an affinity that would be totally determined mechanically from birth, but rather, one might say, an ‘elective affinity.’” The ideal conceptual model is honored even when there is a wide gulf between people’s thought world and the “real world” of people manipulating their kinship and marriage system to advance their own ambitions and desires. However, there is a proviso here. If the intent of the manipulation of the rules is in fact to change the rules, rather than simply rearrange the structural elements on which the rules operate, then the consequences can be transformative. This change of strategies occurred when some Curripaco (also known as Baniwa or Wakuénai) women migrated to the village of San Carlos de Río Negro, abbreviated to San Carlos in this volume, and married non-Curripaco men. At that point the women’s kinship and marriage system changed fundamentally (Valentine and Sims, this volume). However, we cannot assume that this shortage of women will irrevocably undermine the ideal conceptual model of those Curripaco living outside San Carlos, because as other case studies in this book illustrate, the people of lowland South America may go to enormous lengths to maintain their ideal. Among the Kotiria, when there was a shortage of women, people adopted girls for their sons to marry (Chernela, this volume). Among the Waorani a shortage of men is resolved by polygamy and accepting wide disparities in the ages of the spouses (Erickson et al., this volume). The ideal has survived in the face of enormous outside pressures and constraints. Valentine and Sims’s study of the Baré is different from the other essays in exploring the factors that can account for the transition from an elementary to a complex kinship structure. There is considerable historical material on the village of San Carlos; together with fieldwork on the Baré,

6 · Paul Valentine and Stephen Beckerman

who have lived there for at least two hundred years, the work provides a continuous history of the village from the latter part of the nineteenth century. Lévi-Strauss predicted (1969, 477) that an elementary structure would collapse and a complex one would take its place when certain characteristics were in place. Two of these, “the emancipation from relatives and the individualization of the contract,” are remarkably to the point in this case. But as Valentine and Sims demonstrate in their study, there are other significant factors he failed to include. In the cases of the Canela and the Waorani, Crocker and Erickson and colleagues describe how the forces of globalization, and its effects on the social forms of kinship, have particularly undermined the elders’ control of the sexual practices of the young and reduced their influence over choice of marriage partners. As elders’ authority has diminished, there have been more extra-marital pregnancies, more marriages for love, and fewer alliances between families. The next section provides a description of the strategies used to manipulate the norms of marriage, starting with individual strategies and moving to more collective efforts.

Strategies Individual Efforts Frequently the genealogical relationships between two people in smallscale societies can be “read” along different routes. The ways they are interpreted, individually or collectively, are signs of the locus of political control. For example, Mansutti Rodríguez describes the case of a Piaroa man who entered a secret adulterous relationship, which along one route was between cross-cousins, whereas along another it was incestuous. The circumstances of his love affair, carried on in his wife’s village, decided that their relationship was judged harshly and considered incestuous. Alès’s penetrating analysis of Yanomami strategies and manipulations through the genealogical route-ways, activating some while blocking others, illustrates the marital options a person faces throughout life. Agreeing with Chernela, Mansutti Rodríguez, and Picon, she cogently argues that “marriage practices should not be interpreted as the application of an exclusive rule but as a set of structural variations as evidenced by the diversified strategies of the social actors.” For instance, she describes

Introduction · 7

how the introduction of asymmetries in categories applied within a set of siblings and between sets of siblings can be set up. One arrives at the remarkable situation where the same individual will apply the term “sister” to part of a female set of siblings, and the term “wife” to the other part, while his brothers will employ the same designation for the entire female set of siblings. Such a strategy will potentially share out a female set of siblings among a group of brothers, so avoiding competition among them. She concludes, “Within the category of marriageables, the menu is in a way ‘à la carte’: one is no longer within the framework of elementary structures. . . . Dravidian terminology delivers a principle of reproduction of the marriage alliance, but it does not necessarily define the scope of this alliance” (author’s emphasis). As for the Piaroa and the Kotiria, it is the collective and individual choices Yanomami make that define their scope. Yet at the same time, as people form these alliances, they must also be able to rebut those who suggest they are “incorrect” or even incestuous. Public Strategies Often, with the dexterous use of strategies, personal interests can be transmuted into community interests. At the same time these strategies can demobilize those who fail to identify their particular way with the “general interests” of the community, reduce their status, and condemn them to leave the village. For instance, Peluso, in a superbly crafted chapter, describes how the Ese Eja have two distinct types of marriage, the public and the private, that correspond to two separate domains. When a father-in-law intervened and “forced” the transformation of his daughter’s private marriage into a public one, he did so to ensure he had a voice in her future. He transformed his personal, particular interests into public, collective, universal interests. This sort of strategy, aimed at manipulating the community’s views to correspond closely to the ideal norms and thereby win over as much community support as possible, appears to be a popular tactic. Similarly, Mansutti Rodríguez describes a strategy in which Hiwis and Piaroa came to live together in one village and established a number of inter-ethnic marriages, in spite of sharply contrasting value systems. A charismatic Evangelical preacher persuaded them that his universal, collaborative values were superior to their traditional ones. Mobilized by the religious services and celebrations, people had their frame of reference shifted, and their shared Evangelical faith became an important criterion

8 · Paul Valentine and Stephen Beckerman

that decided whom people may marry. Those who were unable to share these values were expelled from this inter-ethnic village. Expulsion from a village is one of the most common public strategies deployed when a fundamental norm is broken. People who marry their primary kin are usually obliged to leave their village (such expulsions are referred to by Mansutti Rodríguez for the Piaroa, and Silva for the Ye’kwana, and are known to occur elsewhere, including the Curripaco). As the incestuous spouses do not conform to the “general interests” of the community, they are reduced to the status of mere individuals and condemned to leave. Frequently, they go to live where they are unknown and establish new identities along genealogical routes that hide their incestuous relationship. In the societies described in this volume, none have political institutions that have an effective monopoly of legitimate power. However, there were those situations in the past, when women were seized on raids, and there still are cases where a feared shaman may pursue a claim on a young woman as a wife even though she does not want to marry him. “Fear of apparent manifest shamanic power,” comments Mansutti Rodríguez, “can be a decisive factor in choosing a spouse.” Likewise, there are instances where men of high status are able to ride roughshod over the marriage rules (Mansutti Rodríguez, this volume) in their desire to marry a particular woman or gain favorable terms in their son’s marriage. It is relevant to describe another case, although it does not appear in Valentine and Sims’s chapter—it would have been out of place there, as they describe the conditions Curripaco faced in the village of San Carlos, and this abduction occurred elsewhere. The case refers to an apparently straightforward Curripaco cross-cousin marriage that was, in fact, a choice inflicted upon the local patrilineal residents by a particularly feared shaman living in another village (Valentine 1991, 272–73). A young woman did not want to marry her future husband, the nephew (BS) of the shaman, and as had been prearranged, only after a tug-of-war did her village let her go. Tied up in her abductor’s canoe, she was taken to his village. She escaped several times but was sent back until, pregnant, she settled for her lot. Nevertheless, the villagers tried to pass it off as a simple choice, thereby making a virtue out of necessity. It was only after several spells of fieldwork in the village that Valentine was told this account. They had been forced to give up a daughter to an inadequate husband because of the shaman’s threats. It reminds us that marriage can never be fully

Introduction · 9

defined in genealogical terms. Reciprocal cross-cousin marriage can be the best kind of union. It is secure because it is between two long-term allies, and can be the least onerous, as it may involve no bride service if it is agreed that neither husband has to do it. Yet cross-cousin marriage can be one of the worst, if it is forced. Community-Wide Strategies Chernela provides three case studies that illustrate the ambiguities inherent in the rules by delineating the strategies the Kotiria use to manipulate rules so as to pursue their particular interests. In the first case, the most rigid of all, a sib held to the practice of reciprocal cross-cousin marriage over the generations with another sib in the same village and died out. In the second, a community practiced cross-cousin marriage but, as they also had multiple affines elsewhere, “diversified their in-law portfolio.” In the third case, two villages “took the long view” and adopted girls who might marry their sons. Chernela argues that the significance in the ambiguities in the crosscousin rule is a consequence of it being an object of manipulation. Under the appearance of obedience to the rule, Kotiria devise different strategies aimed at fulfilling their marital interests, based on their reading of the economic, political, and demographic conditions. Significantly, Chernela and Alès both conclude that these strategies ultimately provide the strategist with the maximum number of unmarried people available. This conclusion is worthy of further research to test whether, in general, kinship norms and rules are manipulated to gain the maximum access to potential spouses.

Ways Kinship Norms and Rules Are Manipulated As mentioned, the chapters in this book illustrate at least two ways in which strategists are able to exercise their choice when manipulating the norms and rules surrounding marriage. First, in small-scale societies everybody is usually related to everybody else along a number of genealogical paths. By manipulating the kinship nomenclature, as Alès describes, a particular path once blocked can be opened. For example, a Yanomami classificatory brother can be reclassified as a husband, or as among the Ye’kwana, a grandmother can become a cross-cousin and therefore a potential wife (Silva, this volume).

10 · Paul Valentine and Stephen Beckerman

The other way involves exploiting ambiguities in the norms themselves. For instance, take the statement “cross-cousin marriage is preferred.” Case studies in this volume illustrate that it can mean any of the following: (1) cross-cousin marriage is the ideal union, although almost impossible to achieve; or (2) cross-cousin marriage must be practiced in all circumstances; or (3) it should be practiced but not necessarily in all circumstances; or (4) cross-cousin marriage is one of many options; the only proviso is that the marriageable persons should behave as if they were cross-cousins. These ambiguities provide interest groups with a wide range of options to assist their strategizing for their preferred candidate and for single people to exercise their personal preferences. Nevertheless, and this is one of the key findings in the volume, there still remain some marital unions that are favored. In spite of the impact that globalization has had on kinship and marriage structures, the ideal conceptual model still continues to have weight in all but one of the societies described.

Discussion Some 75 years after Lévi-Strauss’s landmark publication on kinship and marriage we are still discussing his conclusions. Indeed, in a sense the findings in this book are compatible with his theories. However, to stop at that position would be an error. Our conclusion that the people of lowland South America hold on tenaciously to the ideal conceptual model of kinship and marriage is remarkable, a conclusion that would in no way surprise Lévi-Strauss. He formulated his ideal “models” in the form of kinship diagrams—marriage between the reciprocal, matrilateral, and patrilateral cross-cousins— never claiming that they would work without modification in actual practice. His theory of history, for example, was that conditions of actual life caused models to break apart and have to be reconstituted after the fact. This fits with the data in many of these chapters of groups reclassifying people into mental structures of “accepted” marriage rules after-the-fact; they are textbook structural transformations. Lévi-Strauss was clear that his notion of structures does not appear at the “empirical level” but rather that “the . . . elementary structures of exchange . . . are always present to the human mind, at least in an unconscious form” (1969, 464). In practice, according to Lévi-Strauss, the failure of the actual structure to jibe with

Introduction · 11

reality causes people to “justify” or “hide” these discrepancies. He also claimed that by studying numerous societies’ social structures, even when the complete model was absent in any one society, nevertheless, combining elements taken from them all, the ideal conceptual model emerges. For instance, although Rosengren convincingly shows that Lévi-Strauss’s reciprocal cross-cousin marriage model was of no use to him when he was engaged in fieldwork, nevertheless, even in that individualistic and ruleavoiding society, the Matsigenka future spouses are expected to behave as if they were cross-cousins. However, this book highlights two major problems with Lévi-Strauss’s approach to kinship and marriage. When we come to extracting elements from various societies to reconstruct the structuralist model then, as there is no clear method to the procedure, it cannot be falsified. Second, as the Baré case illustrates, elementary structures do collapse. Not only are the meanings of the kinship categories fundamentally changed but also, as in the case the Baré living in San Carlos (Valentine and Sims, this volume), the hierarchical social structure may exclude the practice of reciprocal cross-cousin marriages. There is another significant conclusion to be drawn. According to LéviStrauss, elementary systems of kinship and marriage function to lessen the gamble that when one group gives up a woman it will get one back. It is a gamble because it would be misleading to consider there is an unlimited supply of women in small-scale societies. By giving women away, men are gambling that they will get a return, but there are demographic and political factors among others that may limit the supply. Lévi-Strauss concludes that what makes the exchange less precarious and more orderly are sets of rules for the exchange of women with other groups. The counter argument, expressed in this book, is that it is not the rules that shape the practice; rather, the rules exist only in practice through the use made of them by the people who, by applying them repeatedly and intensely, keep them up or modify them to satisfy their indispensible interests. There are examples here of people who, although following the rules, have nevertheless made the wrong strategic marital choice and died out, whereas others, manipulating those same rules, have made the right choices and have survived and prospered. It is not the rules per se that make the exchanges more secure. Rather, by means of their manipulation, people attempt to maximize the number of their potential spouses (Alès

12 · Paul Valentine and Stephen Beckerman

and Chernela, this volume) and serve their own interests, those of their immediate families, and those of their societies at large.

Organization The main body of the book is divided into three sections according the emphasis the authors place on different aspects of the structure vs. agency issue. The first section features essays describing the motives behind breaking the marriage rules: by Dan Rosengren on the Matsigenka, Alexander Mansutti Rodríguez on the Piaroa, and Daniela Peluso on the Ese Eja. The second section addresses how the marriage rules are bent or broken. Here are essays by François-René Picon on the Guajiro, Nalúa Rosa Silva Monterrey on the Ye’kwana, Paul Valentine also on the Ye’kwana, and Catherine Alès on the Yanomami. The third section deals with the effects of globalization and recent changes on the marriage rules. Janet Chernela writes on the Kotiria, Paul Valentine and Lionel Sims on the Baré and Curripaco, William Crocker on the Canela, and Pamela Erickson, Stephen Beckerman, James Yost and Rosemary Diaz on the Waorani.

PART I

•

1 Marriage Matsigenka Style Some Critical Reflections on Theories of Marriage Practices Dan Rosengren

Lately much of the anthropological wisdom of our intellectual forebears has been subject to processes of reassessment that have aimed principally at breaking loose from formalized analytical models to reveal people as intentional subjects situated in the everyday worlds they experience. In spite of this, Lévi-Strauss’s work on kinship and affinity (1949) still wields a strong influence over much of how marriage practices in Amazonia are conceived, which reflects the compelling beauty and elegant simplicity of his structural model of exchange. However, in contrast to the analytical perspectives that have been inspired by Lévi-Strauss, and according to which the function of formalized kinship systems is to generate possibilities (and impossibilities) of marriage, I address in this essay two issues that when combined provide an alternative understanding of the marriage patterns that we perceive: one has to do with the perspective of marriage in elementary kinship systems as forming part of systems of exchange in which it is groups or individuals who are the principal actors; the other issue has to do with what is implicated in the meaning of relationship terms. Together, these queries combine in a more general reflection concerning the application of rules of behavior and how these relate to expressed motives for acting in determined ways. Thus this essay is, in the end, a reflection on the longstanding controversy between the significance of structure and agency in shaping our understanding of what motivates people to interrelate in the ways they do. Long ago, just before leaving for the field for the first time, I was told by my teachers and colleagues to pay particular attention to conflicts, since

16 · Dan Rosengren

the comments that they may provoke often disclose principles that suggest not only how people conceive of the world but also how they ought to act in it and in relation to others. In heeding this advice I take as my point of departure for the argument of this essay a dramatic incident that revolved around the struggle for a girl in which several parties were made to confront each other in public. Complementing this analysis, I discuss a myth segment that likewise deals with a disagreement; this time with regard to the choice of a suitor. Both the telling of the myth and the information regarding the particular dispute that I discuss were obtained among the Matsigenka, who live in the Urubamba and Manu River areas in the montaña, or high jungle, of southeastern Peru. Although geographically situated at the periphery of the Amazon tropical rain forest, the Matsigenka conform to the general Amazonian mode of subsistence, relying on a combination of hunting, foraging, and swidden agriculture. The division of labor follows the regionally common gender distinctions according to which men hunt and clear gardens, women are responsible for most of the household chores and do most of the cultivation, and both men and women occasionally fish and forage together. Their livelihood is today affected by the encroachment upon their lands and resources by migrants from the neighboring Andean highlands and, above all, by the huge and diverse impact on living conditions brought on by the commercial and large-scale exploitation of the finds of natural gas in the area around the headwaters of the Camisea River, one of the main tributaries to the Urubamba. In spite of the social and ecological turmoil that the external influences associated with the Andean immigration to the area and the exploitation of natural gas signify for many Matsigenka, their marriage practices and relationship system remain largely unaffected. Thus the Matsigenka relationship terminology, which is of the kind that has been described as Dravidian (Dumont 1953b), or as a two-section (or two-line) prescriptive marriage system (Needham 1962), has not undergone any notable changes during at least the last century (see, e.g., Aza 1923). At first glance Matsigenka marriages seem to conform perfectly to the rules of bilateral cross-cousin marriage that Dumont (1953b) associated with the Dravidian relationship terminology; people always marry those within the cross-cousin category. On further scrutiny, however, it soon becomes apparent that it really is the other way round. What conventionally are conceived of as rules are made to conform to actual marriage

Some Critical Reflections on Theories of Marriage Practices · 17

practices; people who are mutually attracted define each other as belonging to the same category that cross-cousins belong to.1 As a consequence, actual marriage practices, from both a genealogical and a categorical point of view, are much less systematic than they first appear. What had conventionally been seen to constitute a set of rules is, accordingly, less of a socially normative system that governs people’s behavior or a mental model governing people’s conceptions than it is a set of linguistic conventions used to make manifest certain social circumstances or intentions.2 The absence of an inherent regulatory structure is evident also with regard to another idea that for a long time has been cherished in many studies on marriage practices, that of exchange. Following the impact of Lévi-Strauss’s model, marriage has frequently been seen in terms of a relation where, for instance, families, clans, or moieties give their women in exchange for women from other families, clans, or moieties. Still today there is a tendency to view this kind of exchange between groups as a universal phenomenon within so-called elementary kinship systems. Accordingly, this perspective is frequently emphasized in analyses of the concentric dualism that is produced in Dravidian systems in Amazonia (see, e.g., Rivière 1993, 511; and most of the contributions to Kensinger 1984). Descriptions of how women are exchanged in marriage transactions have, moreover, often been given a certain male bias, which feminist scholars, reacting against the objectification of women, have tried to rectify. They point out that this bias is not a universal phenomenon, since women in many societies take an active part in the arrangements of their children’s marriages (Carsten 2004, 13). Notwithstanding the feminist criticism, families or households are still often represented in ethnographic descriptions as homogenous units composed of internally undifferentiated actors without taking into account that households are formed of members who may have different agendas. In studies of marriage, the formation of marital unions is still often seen in terms of exchanges between groups of the same kind, and as such these transactions are linked to issues relating to social structure and organization that are assumed to characterize entire ethno-linguistic groups.3 In contrast, comparatively little attention has been paid to what takes place within households partaking in marriage transactions. In effect, studies of marriage practices still tend to portray the family as a kind of social unit homogenous in its constitutions and harmonious in its conduct. Rarely is it seen as being formed of autonomous agents. This lack of

18 · Dan Rosengren

attention to internal differentiation tends to bias the study of families, and represent them as a confirmation of the structural models, models that presuppose the determinant force of formal rules, rules which, in turn, are seen as governing the form of the resultant marital unions.4 With regard to analyses of marriage among the Matsigenka, the exchange perspective is, as can be expected, prominent even though the understanding of the pattern of exchange differs. On the basis of the relationship terminology both Gerhard Baer (1984, 86) and Wayne Snell (1964, 6ff) observe that there is a rule of cross-cousin marriage.5 In agreement, Allen Johnson (2003, 163) concludes that the Matsigenka kinship system’s “most striking feature is that it tends to create a symmetrical exchange system consisting of two families intermarrying endlessly across the generations.” In contrast, France-Marie Renard-Casevitz (1998, 245) suggests the existence of complex systems with at least eight local groups that are united in networks of intermarriage. However, to understand actual marriage practices and why people act as they do, I find the abstract models of exchange to be of little help. In this essay I therefore examine the process of negotiating marriages and, as an alternative to more formalistic attempts at understanding the formation of marriage unions, I dwell particularly on how conflicting perspectives are handled. Thus the focus is on actual practices, “the messy realities” as aptly described by Janet Carsten (2004, 42), rather than on neat and tidy structural properties. Hence I adopt an approach that favors the perspective of the integrand members of households as independent decisionmakers. Discussing the analysis of “messy realities,” it is appropriate to mention the apparent indeterminacy that characterizes much of how social relationships are ordered among the Matsigenka. The way in which the Matsigenka relate to each other is liable to change when people find it convenient; those who one day address each other with the terms that are used for siblings may the next day treat each other terminologically with those that are used between in-laws. This indeterminacy plays havoc with formalistic analyses that model their understanding of social relationships according to genealogical notions, even though, as noted by Shapiro (1984, 4), the analysts may claim the contrary. The classificatory confusion that results has been noted not only in relation to the Matsigenka but also with respect to other indigenous Amazonian societies. With regard to the Piaroa of Venezuela, who similarly have a Dravidian terminology, Joanna

Some Critical Reflections on Theories of Marriage Practices · 19

Overing notes (1985, 173; see also 1984) that “it is probably a mistake, and a bad one, to reduce tribal kinship terminologies to kinship and affinity, or to social categories, or to consanguineal relations. They do, of course, usually refer to such things, as we understand them to be; but they may well have more important predicates.” For the Piaroa the meaning of these terms covers a wide social range that extends far beyond kinship and affinity and into cosmological relationships. In effect, the genealogical significance of such terminologies is only secondary as, arguably, they are indicative of more general modes of social relationships.6 The relationship terms are not employed merely to address and refer to people; arguably more important are the sets of rights and duties associated with them. Consequently the relationship terminology does not provide a fixed structure determining people’s behavior. It is not that one must marry certain persons because, according to the terminology system, they belong to the appropriate marriageable category. On the contrary, the set of relationship terms can be employed to define certain relationships. Other-sex cross-cousins all fall within the category of marriageable persons, but they are not the exclusive integrands since there is more to the terms than the genealogical relations. For the Matsigenka the terms used also indicate social relationships and intentions. Thus to find a spouse it is crucial to find someone who accepts the appropriate term of address together with the expected form of social treatment, rather than to identify someone who according to the terminology system genealogically fits the appropriate category of people who can be married. To the significance of koki and pagiro (conventionally given as MB, FZH, HF, and WF, and as FZ, MBW, WM, and HM, respectively) should thus also be added potential father- and mother-in-law. Similarly, to ani and pinato (conventionally given as mFZS, mMBS, WB, and mZH, and as fFZD, fMBD, HZ, and fBW, respectively) should be added potential brother- and sister-in-law. (Since there exist no terms for potential spouse or other-sex cross-cousin, it is the parents and siblings of the desired partner who are targeted for negotiating a redefined relationship.) Before examining the particular incident and the myth that serve as the ethnographic point of departure for this essay, I provide a brief account of the sociocultural background against which the two accounts of Matsigenka sociality unfold.

20 · Dan Rosengren

On Matsigenka Social Organization In spite of the common assumption that there exists a normative system shaping the actual pattern of marriage among Matsigenka, the organization of their social life is also frequently characterized as atomistic and outspokenly individualistic in nature. Matsigenka social organization is “loose” and flexible and, congenially with the two-line terminology system, they lack permanent groups such as unilineal descent groups and moieties. Moreover, after the initial year(s) of post-marital uxorilocality, the residence pattern varies as there is no explicit rule that determines where to establish new households. Parents strive though to retain married daughters with their families in the vicinity of where they themselves live. Accordingly, if a couple remain in the neighborhood of the wife’s parents’ household, the interaction between the two households is most of the time intense. From the perspective of young men this arrangement is often experienced as an attempt on the part of the young women’s parents to exercise control over their sons-in-law. The desire to avoid the parents-in-law’s dominance is consequently often a strong incentive for a man to argue for settling as far away as possible from his wife’s parents. The uxorilocal tendency, however, is strengthened since newly married women frequently wish to remain close to their female kin. The position of young, recently married men is usually too weak to enable them to enforce their own preferences in the face of such female coalitions. In spite of pronounced expectations to the contrary, some men manage to resist the pressure of their wives and in-laws and find ways to settle according to their own choice without it being considered a transgression of social rules. Indeed, in a few cases that I have come across, such bold action seems to have rendered the young men certain prestige, at least among their fellow age mates. However, when a Matsigenka couple establishes a new household, they commonly do so in the vicinity of the wife’s parents because this is how both she and her natal family want it, not because the couple is required to do so according to a system of formalized rules. It is principally the desire to maintain strong ties between households of closely related women that serves to form the actual social composition of neighborhood groups. Both the social and emotional ties between parents and their daughters frequently remain strong after the daughters’ marriages. The cautious and sensitive husband knows that his wife’s views in a family or inter-family

Some Critical Reflections on Theories of Marriage Practices · 21

conflict are not to be taken lightly since she may very well be expressing also the opinion of her parents. The maintenance of close bonds with parents means, therefore, that the daughters’ husbands cannot do exactly what they please. Siblings usually demonstrate strong ties of attachment, even after they have married and established families of their own—this also goes for brothers and sisters particularly if they manage to remain in the same neighborhood. Moreover, through the bonds of sibling solidarity, “sisters,” as a group, commonly exercise a considerable check upon their respective husbands. In this way women wield, directly and indirectly, a significant influence on the political life of the local group. The relative importance of the various actors’ influence over their own situation agrees with the strongly expressed individualistic ethos that permeates Matsigenka society. The actual social patterns that characterize each settlement emerge largely as the result of how those involved manage to negotiate their interests. Variations in the patterns of interaction can often be seen as the outcome of how different strategies accommodate to each other. With regard to relations between in-laws, for instance, it is apparent that the degree of subordination of the son-in-law is in general inversely related to his age; the younger he is, the more he is dominated and vice versa. Although age is involved in establishing the degree of inequality in in-law relations, the degree of interdependence has to do primarily with the son-in-law’s social status rather than with his age, which is of secondary significance. Usually it takes time for young men to gain the social status that can equal that of their wives’ parents.7 A more important factor is the effect of conviviality; the longer people consort, the closer and more similar they become. Cecilia McCallum (2001, 91) describes this as a process of blending, and with time affinity is turned into kinship. Eventually once young men reach a position when they are able to redefine their parents-in-law as “grandparents,” sháinka and sháonka, this not only marks the inversion of the relations of dependence but also signifies the ultimate annihilation of affinity.8 Even though the position of newly married women in general is stronger, for instance, with regard to influencing where to settle, than that of their husbands, the latter are not entirely without means of influencing their situation, even if they remain in their parents-in-law’s neighborhood. One option a man has is to promote one of his brothers’ marriage to a sister of his wife, which means that the two men have someone to whom they are close and upon whom they can rely to act in common against

22 · Dan Rosengren

their fellow superiors. Another option is to strengthen the bonds with fellow sons-in-law to the same senior couple—men whom they anyhow refer to as their “brothers,” ige. If these strategies are unsuccessful, divorce is the common solution, and recent marriages are often unstable as a result. However, if the couple manages to keep the family together for a few years, even though there are exceptions, their union usually remains (cf. Rosengren 1988). Within the residence group there are thus both centrifugal and centripetal forces, meaning that tensions between integrand households at times become strong.9 The potentially disruptive tendencies within residence groups are mitigated through frequent visits among the constituent households. Even if visits often are mutual, the system is far from balanced. Within a residence group the married daughters’ parents’ household usually becomes a focal point, receiving many visits from the other households of the group, whereas the members of the parents’ household correspondingly pay fewer visits to other households of the group. Women who remain living in the vicinity of their parents spend some time with them almost daily. Often it is a matter of just brief visits to chat about something, but occasionally it may be to give a hand in laborintensive tasks together with other sisters. Sons-in-law are generally not as frequent visitors at the parents-in-law’s households, but usually they pass by regularly enough to maintain good relations. In contrast, sons often have comparatively few opportunities to visit their parents. On a communal level relations of power and authority are generally weak and inconspicuous. Although they lack means of authority to impose decisions, the informally recognized leaders are men who aspire to influence conditions beyond the residence groups to which they belong. There is no formal system for the resolution of community conflicts, but residents are commonly prepared literally to go far to avoid conflicts. As among other conflict-avoiding Amerindian peoples (Overing 1989; Overing and Passes 2000), the Matsigenka commonly approve of virtues focusing on sociability, such as generosity and tolerance, while they disapprove of attributes that socially detach the individual from the social setting, such as anger, fearlessness, and aggressive bravery. Matsigenka sociability respects individual idiosyncrasies and freedom while minimizing authoritarian and hierarchical relations. The only alternative mode of conflict resolution is either intensive campaigns to convince the erring party to correct the behavior (Rosengren 1988) or the moving away of

Some Critical Reflections on Theories of Marriage Practices · 23

one of the conflicting parties until the conflict has been solved or forgotten (cf. Rosengren 2000). With the establishment of the administrative entity called the Comunidad Nativa (Native Community) conflict resolution campaigns have been launched in cases where disagreements have been considered to be potentially disruptive.10 When the seriousness of the argument has been deemed to require a certain degree of formality, then frequently the Asamblea Comunal (communal assembly) has been summoned to listen and, if possible, to guide the conflicting parties.

Negotiating Marriages One morning some years ago the Communal Assembly of Pemperoato gathered in the community’s school house to try to resolve a conflict that had arisen around a young woman, whom we here give the name of Rachel, and three young men: Isac, Benjamín, and Rubén.11 According to what seems to be the general opinion among the people of the community, the following had happened: Isac, who is married to María, an elder sister of Rachel, appeared during the middle of the night at the house where Rachel slept together with her younger siblings while their parents were away visiting a neighboring household. Members of the community generally held that Isac forced Rachel to come with him. First the couple went to an accomplice’s house, where people say Isac raped her. At daybreak the following morning they set off at great speed for Kemariato, the community where Isac was living, situated approximately a day’s rapid march from Pemperoato. When Moisés and Josefina, Rachel’s parents, returned home in the morning, they soon found out what had happened. Moisés was infuriated and wanted to leave immediately to retrieve his daughter. Worrying about what Moisés might do when he found Isac, Josefina stopped him from going and sent for Angel, their eldest son, whom she urged to go to Kemariato to bring back his sister.12 When Angel heard about what had happened to Rachel he immediately went to fetch his younger brother, and together they hurriedly left. The brothers arrived in Kemariato in the late afternoon and went straight to Isac’s house, where they found their sister. Besides Rachel, there was no one at home, which was a relief to Angel, who felt that the situation in Kemariato was somewhat awkward since, if it came to trouble, local people would certainly take the side of their fellow community member. Thus the brothers did not linger to wait

24 · Dan Rosengren

for Isac and set off as quickly as they could toward Pemperoato with Rachel in their company, even though this meant that they had to spend the night in a rough and simple lean-to along the trail. A few days later Alberto, a friend of Isac, arrived at the house of Elías, the president of the community of Pemperoato. Alberto said word had reached Kemariato that Rachel was back in her parents’ house. On behalf of Isac, he wanted to protest against Rachel’s retrieval, which it was claimed amounted to annulment of the marriage deal that Isac maintained he had made with Moisés on behalf of his younger brother, Benjamín. To resolve the conflict, Elías arranged a meeting between the contending parties some days later. At this meeting appeared Rachel, her parents, and Benjamín but not Isac. The meeting was moreover attended by Rubén, who also claimed that he had arranged with Moisés and Josefina to marry their daughter—a claim that was acknowledged by the senior couple. In his capacity as president of the community Elías served as mediator, trying to find a solution that would be acceptable to all. After discussing the issue for a while, during which the different standpoints were made clear, the meeting arrived at the position that the unhappy situation had arisen as a consequence of a misunderstanding on Isac’s part. It was declared that Rachel’s parents had intended all along to let her marry Rubén, which they claimed was what she really wanted. The meeting was concluded with Benjamín accepting this interpretation; he withdrew all claims on Rachel, while her parents, in return, agreed not to press any charges against Isac for the abduction and violation of their daughter. The conflict would probably have been settled at that point had it not been for the fact that only a few days later, rumors began to spread in Pemperoato, according to which Isac was very angry. It was said that Isac told everyone he saw that Moisés was a liar and afraid of his wife, who was the one who wanted Rubén to marry Rachel. It was also said that Isac was promising, the next time he met Elías, to give him a thrashing he would not easily forget. Hearing this, Moisés, encouraged by Elías, decided to disregard the agreement with Benjamín and file a formal complaint at the police station, one day’s walk downriver, for the capture and rape of his daughter. As a consequence of this development a new meeting was arranged in which representatives from both communities were present together with the contenders. This time Isac showed up, although he was careful never to stray far from the company of the delegation from Kemariato. When

Some Critical Reflections on Theories of Marriage Practices · 25

Isac gave his version of what had occurred he stressed that he had recently talked to Moisés several times about letting Benjamín marry Rachel, who he said was really enthusiastic about the prospect. His main argument, he stressed, was that it was María (his wife and Moisés’ daughter and Rachel’s elder sister), who had suggested Rachel as a suitable partner for Benjamín. Such an arrangement would be advantageous for María, since she then would have a sister living near her in Kemariato, a place where she often felt alone since the rest of her family lived in Pemperoato. After having discussed the proposition, Isac claimed that Moisés had seen the advantages of this union and had given it his blessing. Moisés admitted that he had been speaking to Isac about Rachel’s marriage, but he denied having accepted any union between her and Benjamín. Why, he asked, would he agree to lose another daughter in the same way that he had lost María, who now was living far away in Kemariato with this good-for-nothing son-in-law, who was never there to lend a helping hand when needed? Why indeed should he give in to this marriage with Benjamín, a man almost unknown to him, when there was another suitor, Rubén, whom he knew well and who had made clear his intentions, and who, moreover, had proven his qualities as a capable provider for Moisés’ daughter? Warming to the subject, Moisés stood up, first fastening his eyes upon Isac, and turning to the audience, he asked rhetorically: if Rachel had been intended for Benjamín, why then had Isac raped her? At this point Isac stood up and with raised voice denied that he had ever touched Rachel. Instead, he said, this allegation was something that Moisés had thought up to save himself because he was afraid of his wife. It was known to many, said Isac, that Josefina preferred Rubén to Benjamín, since Rubén was only a daft Matsigenka coming from downriver—thereby implying that Rubén was more “traditional” in contrast to the supposedly more urbane and modernistic people of Kemariato. Because of his weakness, Isac continued, Moisés did not dare but comply with his wife, and to save his own skin, he had lied and slandered Isac. This argument, however, failed to win Isac any support among those who were assembled. Then the meeting turned to Rachel and asked her whom she preferred of the two suitors; she nervously looked down and replied, almost whispering, that she would rather live with Rubén than with Benjamín. Finding this declaration decisive, the meeting, including the representatives from Isac’s own community, concluded that Isac had been proven wrong, that

26 · Dan Rosengren

there was no deal between him and Rachel’s parents with regard to her marriage to Benjamín, and that if there ever had been one, it was by now null and void. When Isac in his defense speech noted that Rubén, in contrast to Benjamín, was living in Pemperoato, he implied that if Rachel were to marry Rubén, she would remain in her parents’ neighborhood, in contrast to her sister María. This meant that she would form part of the matrilaterally related group of households with which Josefina interacted (Rosengren 1987a and 2004). As predicted by Isac, Rubén did indeed move into Rachel’s household only a few days after the communal assembly’s verdict, whereupon they started to live together as husband and wife. When Rachel gave birth to their first child not quite a year after this event, the couple had established their own household situated near enough to that of Moisés and Josefina to ensure that they could continue to interact on a daily basis. From the perspective of the exchange model this case may perhaps be regarded as unrepresentative because of the dramatic events. Although the case is admittedly atypical, the public discussion clearly brought forth arguments that illustrate how the various actors were thinking about marriage in general. Thus, since my focus is on the individual actors’ own manner of describing the event, this case is telling in several respects. What is perhaps most striking is the total lack of references to expectations regarding the exchange of women. The emphasis in the discourse around what happened, on the contrary, is on individuals and their doings and motive for acting in the way they did, and not on any relationship with another family or social group from which an exchange of marriage partners could be anticipated. Neither did anyone seem to show any interest in the relationship categories that the respective parties belonged to in relation to each other. All three of the young men referred to Rachel’s mother and father as pagiro and koki, and in return they referred to the young men as their notíneri (conventionally given as mZS, fBS, and DH).13 In the case of Isac the terms of reference employed were in conformity with the condition that he had married María, and by extension this mode of reference also applied to Benjamín being Isac’s brother. In the case of Rubén, no previous relations placed him in an in-law relationship position. In contrast, the relationship between Rubén and Rachel’s parents was the result of a mutual accord that included not only the terms of address but also a series of rights and obligations toward each other.

Some Critical Reflections on Theories of Marriage Practices · 27

Of the various protagonists the one with the least to say (at least in public) was Rachel, though she was not completely unheard. Although the woman’s parents, and particularly her father, seemingly had the final word, under more conventional forms of courting they will, as a rule, take into consideration the wishes of their daughters. If they should go against a strongly voiced refusal, the likelihood is that the enforced union will not be a long lasting one or, even more dramatic, the girl may even decide to end her life; heartache is a prominent motive for suicide among the Matsigenka (Rosengren 1988). The mere knowledge of what might happen seems to deter most parents from forcing their daughters into marriages against their wishes. Although it seems to be the parents who decide whom their daughters marry, this decision is usually preceded by the acquisition of the daughters’ consent.

Agents, Not Pawns Although attraction may be experienced in a variety of idiosyncratic and culturally influenced ways, it can be argued that its role in the acceptance of a potential marriage partner is universally of some importance. In the analyses that start from the position that marriage is either rule governed or a strategy to create or maintain alliances between kin groups, little attention is paid to feelings of attraction. Arguably, one consequence of this neglect is the representation of those who supposedly abide by the rules of marriage exchange as unaffected by any force of attraction. This view evokes an image of such people caught up in a rule-based system as unemotional automatons, radically different in this respect from ourselves. The potentially dire consequences that may result from parents’ disregard for their daughters’ emotions and desires are certainly well known among Matsigenka. Moreover, it is a theme that appears in myths and other narratives. One of the myths that touches upon this lack of attention tells of the woman who loved Worm. It is of interest in the present context as it articulates some of the issues brought out in the case described above and illustrates the attitude many Matsigenka have regarding how to choose a partner. The Story of the Woman Who Loved Worm Formerly there lived a family with a recently grown-up daughter. Together with the family there lived a man whom the father referred to as “my

28 · Dan Rosengren

son-in-law” (notíneri) and whom he wished his daughter to marry. The young woman, however, was not interested in this man because Worm was already her lover. Her father said, “You should prepare manioc beer for ‘my son-in-law.’” But she did not want to. Later her father said: “Now you should go to the fields and bring manioc to cook.” Again she refused. When she was alone she went into the women’s house and knocked on the worm hole. When her lover appeared from the hole she sat herself on top of it. When they were finished making love Worm returned into the ground while she rose to her feet and covered the hole with a stone. She loved Worm, whom she saw as a human, and she had no intention of marrying the other man. Her father said, “My daughter does not want ‘my son-in-law,’ rather she wishes that he is thrown out of the house.” The father, suspecting that something was going on behind his back, told his daughter to go and harvest manioc and then to prepare manioc beer. When she had left for the fields he hid himself so that he could spy on her when she returned. When she came home she quickly peeled the manioc tubers and cut them into pieces. When she had finished this task she entered the women’s house. From his hideout her father saw how she knocked on the stone on the floor. The worm appeared and she sat on top of him. After having finished making love she covered the hole and then she went to fetch water for the manioc beer. When her father saw her leave he quickly entered the house his daughter just had left and he said to himself, “It is because of Worm that my daughter is not interested in ‘my son-in-law.’” He went out and fetched his cauldron, began to boil water, and when it was boiling he brought it into the house. He knocked on the stone as he had seen his daughter do. When Worm appeared the father poured the boiling water into the hole. In agony Worm crept up onto the floor where he stretched out and died. When Worm was dead the father threw Worm out of the house and hid him. Then he hurriedly went to his fields to work and, as always, he came back in the afternoon. Meanwhile his daughter returned and entered the house. She knocked on the stone to call Worm, but no-one answered, and she said to herself, “For sure he has gone to see his mother.” In the afternoon she was sitting outside the house together with her mother. When the father saw his daughter tranquilly delousing her mother he fetched the body of Worm and threw it in front of the women. Angrily the father said to his daughter, “Look here! Here is your suitor! For the sake of Worm you did not want to live with ‘my son-in-law.’ Now go and throw him away!” The daughter stood up

Some Critical Reflections on Theories of Marriage Practices · 29

and wrapped a cloth around the body of her dead lover. She left the house and went down the trail, and after a while she threw the body into the forest, but she did not return home. Instead she continued along the path until she reached the river bank, which she followed, shouting to Jaguar to come and eat her. When Jaguar appeared the daughter accompanied him to his house. She stayed with him, and after a while she became his wife. The woman’s brother longed for his sister and he set out to look for her. Eventually he found her living with Jaguar. The brother stayed with the couple for a while then he returned home, but in the end he went back and settled permanently with his sister and his brother-in-law, leaving their parents alone.14 In this story we might perhaps have expected to hear about the remorse of the woman’s father for having lost his daughter; however, in a typical Matsigenka fashion, the myth does not contain any outspoken moral sense. The father’s behavior is clearly unforgivable to most Matsigenka, and that the daughter decides to leave her parental home comes as no surprise to the listeners. When serious conflicts break out, the Matsigenka usually avoid each other. The occurrence of a disruptive disagreement within a family signifies that one party leaves, only to return when the problem has been resolved or when it can be forgiven or forgotten (Rosengren 2000). Thus at the point in the narrative when the daughter sets out into the forest, the initiative to mend the relation lies with her. This does not signify that parents who lose a child because of a conflict do not grieve, and we may assume that the narrative’s audience takes that for granted. The young woman’s father surely longs for his daughter to come back safe and sound. That it is the brother who goes looking for her partly follows from the condition that he had no part in the conflict and confirms the close ties that conventionally exist between other-sex siblings—at least as long as they both live together with their parents—which makes his search for her a logical event. The self-evident way in which he sets out to look for his sister in the virtual reality of myth is, accordingly, paralleled in real life by Angel setting out to recover his sister Rachel. As the myth indicates, Matsigenka women have a greater influence over whom they marry than it may appear at first sight (see also O. Johnson 1978, 103ff). Moreover, it is clear that they do not necessarily share the same criteria of attraction as their parents. Obviously these criteria cannot easily be generalized, since they are highly idiosyncratic, which

30 · Dan Rosengren

also means that they are unlikely to conform to an orderly exchange system in which women are principally used as pawns. There is, however, one criterion that seems to be important for most young women (as well as for their parents), a criterion that is indicated in the dramatic case of Rachel. They want to maintain close relations with members of their natal family, an aspiration that is instrumental in forming and maintaining the matrilateral core around which most residence groups are formed. We can only speculate about whether Isac intended Rachel for himself or for his brother, and it is hardly relevant to the present argument. More important is the motive Isac gives for choosing Rachel among all potential candidates; he claimed that María (his wife and Rachel’s elder sister) felt alone and far away from her family. As María realized, she probably could not persuade Isac to move to Pemperoato, which apparently had been an issue between them, and having Rachel come and live with her, or near her, would have been an acceptable substitute. And even though Moisés denied that this condition had made him accept Rachel’s marriage to Benjamín, it is an argument that certainly would be considered sensible to most Matsigenka. In contrast to parents’ involvement in their daughters’ marriages, they are, as a rule, fairly indifferent to their sons’ choice of marriage partners. The most immediate reason for their lack of interest is that most adolescent men make themselves independent of their natal families at an early stage. When they have reached 15 to 17 years of age they often move out and join another household where there are girls who in due time might become their wives—in former times this seems to have happened even earlier. Some young men take this opportunity to acquaint themselves with distant parts, and they roam over large tracts, joining for longer or shorter periods of time to different households until they eventually marry and settle down. Other men remain in the vicinity of where they were brought up. These latter adolescents seem most frequently to associate with people whom they know from before they left home, which often means families of their parents’ other-sex siblings, whose daughters, following the logic of the relationship terminology system, automatically belong to the category of women whom they can marry. The men who move in and the household members with whom they associate treat each other as “in-laws,” social categories to which they often, but far from always, belong. The senior couple of the household where the boy takes up residence is referred to as the “parents-in-law,” or

Some Critical Reflections on Theories of Marriage Practices · 31

koki and pagiro. The boy is referred to as notíneri, “my son-in-law.” The “mothers-in-law” and “fathers-in-law” are persons whose goodwill the young men depend upon, and as such the seniors are awe-inspiring, especially if there are other young men living in the household who have also established “in-law” relations with the senior couple. Many young adolescents feel uncomfortable approaching the parents with a request to marry their daughters if the “parents-in-law” have not made clear that this is their own wish. If the young man is doubtful about how the “parentsin-law” think about having him as an actual son-in-law, he may decide to use a go-between in the negotiations. This mediator seems always to be a man, although it can be anyone in whom the suitor feels that he can confide. Nowadays there are boys who do not attach themselves to other households and who remain with their parents. Generally they require the assistance of a go-between to approach the parents of a girl whom they desire to marry. The motives for staying in the natal household vary. With the introduction of land reserves, Comunidades Nativas, in the 1970s mobility has been made more difficult. The restriction on mobility is a consequence of the increased competition over land following from the sometimes limited territory extensions available to communities. In Pemperoato, which is a comparatively large community, people say that Benjamín was too timid to join a strange household. Because of his timidity, Benjamín did not himself dare to approach Rachel’s parents to arrange his marriage. Isac was the evident mediator to use since he was known to Rachel’s parents, and moreover he had a reputation of being persuasive. An additional perk to the timid young Benjamín was the reference to María’s wish to have someone from her natal family living nearby. If that occurred, it would make it possible for him to marry Rachel without having to leave the neighborhood where he was brought up. Since both young men and women usually know whom they wish to marry and act accordingly, they cannot be seen to act as representatives of social units in an exchanging network. The question then becomes why is it that young people find only some people attractive? Of course, this is something highly idiosyncratic; nevertheless, certain general considerations frequently seem to influence their choices. They do not concern ideas of beauty or charm—factors that are not without importance—but are of a more social character.15 Primarily and most decisively, the prospective mate must simply be someone with whom one thinks it is pos-

32 · Dan Rosengren

sible to live. The importance of this criterion is indicated by the high ratio of early divorces that take place when couples are unable to share their everyday lives because of incompatible personal chemistries. If a marriage has produced children and survived for a year or two, the two have most likely been able to adapt themselves to a life together, and their marriage is likely to develop into a stable union. To be someone with whom one can live contains at least two aspects that are of equal importance for both men and women. First, the ideal spouse should be reliable and hardworking, someone who can be trusted to do his or her fair share of the work to maintain the household; and second, to sustain a conjugal relation over time requires a degree of trust and “congeniality,” which will make the spouses overcome their mutual affinity and develop a “sameness” that goes beyond marriage to generate the family as a coherently functioning social unit in which peace and mutual confidence reign. Since siblings grow up together they are often quite close emotionally. During a period of their lives the relationship between brothers can be strained because they may be potential competitors for the same spouse. In contrast, other-sex siblings are generally close and supportive (Rosengren 2000, 228ff)—this intimacy is demonstrated both in the case discussed where Angel immediately sets out to rescue his sister from the vile Isac (according to Angel’s opinion) and in the myth that recounts a similar event, where a brother longing for his disappeared sister seeks her out. Following from the closeness of brothers and sisters, the children of other-sex siblings are often close, not primarily because they are genealogically cross-cousins but because of the ties between their parents. This relationship is often sufficient to generate the emergence of a degree of intimacy between their respective children that makes it easier to turn affinity into intimacy bordering on consanguinity. The development of conjugal congeniality is related to the very first observation made earlier; it has to do with the relationship between the terms of address used and the choice of marriage partners. Since potential spouses do not have to be actual cross-cousins it is necessary to establish an appropriate mode of treatment that is expressed both in the terms of address employed and in how people act in relation to each other. This could be considered a test, and should the prospective suitor fail to live up to expectations, his chances of being accepted as a husband and/or sonand brother-in-law will be slim. The condition that there are no terms

Some Critical Reflections on Theories of Marriage Practices · 33

of address for other-sex cross-cousins/potential spouses does not signify that those who relate to each other terminologically in this way are expected to avoid each other.16 There are alternative means by which such people can address and refer to each other; teknonymy is one way, and today the use of public personal names is widespread, a practice that was introduced by Christian missionaries and rubber barons at the end of the 19th and beginning of the 20th century.17 Obviously the relationship terminology system does not only denote relations of kinship and affinity in the conventional sense. The semantic field covered by these terms is evidently more comprehensive, and the actual use of the various terms suggests that they principally have to do with the description of actual social relations. Accordingly, when people choose to refer to each other by the same term that they use when they talk to their cross-cousins, this is an indication of the closeness of their relation, and it is not a manipulation of the terminology system, which has been suggested, for instance by myself (Rosengren 1987a and 1987b) as well as by others.18 As in similar terminology systems, the absence of a term relating to the other-sex cross-cousin indicates the processual quality of this relationship. This signifies that the large majority of people—that is, all those who are genealogically unrelated to ego and of the opposed sex—are terminologically treated in this way. The lack of a term suggests that this particular relationship is transformable, and most important, if the social congeniality is sufficient, a couple can turn into spouses; men turn into nojime (my husband) and women turn into nojina (my wife), and affinity becomes family.

Conclusion Based on a study of the formal characteristics of the relationship terminology system one might conclude that the Matsigenka have a kind of symmetrical exchange system, rather than a system where the choice of marriage partner is mainly governed by personal sentiments and not by structural restraints. If we pay attention to what Hornborg (1998, 169) refers to as “the elusive realm of social ‘performance,’” marriage among the Matsigenka is obviously not so different in this respect from marriage in so-called complex systems where individual agency rather than obedience to structural rules is supposed to be decisive for the choice of one’s

34 · Dan Rosengren

partner. And hence, the Other turns out not to be the odd and exotic creature suggested in the models that in their fundamentals were, to a large extent, developed during the early years of the last century.

Notes 1. See, for example, C. Busby (1997), who argues contra Dumont that the Dravidian terminology system defines marriageable categories and thus is prior to actual marriages. 2. On conceptions and the extensive and often technically complex discussion with regard to the significance of prescriptive and preferential models of marriage see Hornborg’s (1988, 32–35) admirably succinct but still complete review. Although Dravidian terminology systems may seem structurally simple at first glance, they have given rise to various and often contradictory analyses that in a no less admirable fashion are dissected by Viveiros de Castro (1998b). 3. Examples of such a generalizing perspective in recent analyses of marriage patterns among lowland South American peoples with Dravidian systems can be found in, for example, Dreyfus (1993) and Houseman and White (1998). 4. The recent effective critique of kinship studies’ old conceptual models has not had much impact on the thematically adjacent study of marriage. Arguably this is because recent interest has focused particularly on the impact of the emergence of new conditions—such as surrogate mothers, in vitro fertilization, and transnational adoption in the modern Western societies—on notions of kinship and kinship relations (see, e.g., Franklin and McKinnon 2001, 8). 5. Beverly Bennet (1991, 49) quotes Baer and Snell in agreement. 6. I believe the seeming “terminological messiness” to be only partly reminiscent of Viveiros de Castro’s (see, e.g., 2001) notion of the encompassing potential affinity according to which consanguinity is seen as an “inventional” differentiation that emerges from the affinal background. Viveiros de Castro’s focus is on affinity and consanguinity as conceptual categories rather than on the practice of everyday interrelations. His argument, therefore, principally refers to universalistic mental assumptions rather than to particular moral instances. 7. An alternative strategy for dealing with in-law relations is open for those who live in areas neighboring Wachipaeri settlements. In these parts there are Matsigenka men who marry Wachipaeri women because the post-marital residence ideal among the Wachipaeri is virilocal. In such marriage unions men do not need to bother to the same extent about their actual parents-in-law as they have to if they should marry a Matsigenka woman. 8. The term sháinka is used to address not only FF and MF but all men two generations or more above ego. Similarly, sháonka is used to address any woman, who from the perspective of ego, is two generations or more above ego. From ego’s perspective, all koki and pagiro will in due time be transformed into “grandparents” while, in contrast, those who are addressed as apa and ina (conventionally translated as F and FB and as M and MZ respectively) will always be addressed in that fashion.

Some Critical Reflections on Theories of Marriage Practices · 35

9. Elsewhere I describe the development of the residence group in more detail (Rosengren 2004, 95–98). 10. With the implementation in 1975 of the law creating native reserves, or Comunidades Nativas, the national authorities introduced the institutions of an elected board and a presidency as forms of local government. The first reserves were established among the Matsigenka during the last years of the 1970s. The process of implementation in practice has been problematical (Rosengren 2004) and is still unfinished. 11. Since I here describe a conflict that may still be sensitive to those involved I have changed the names of both the parties and the communities from which they came. 12. Josefina was worried, in part, based on Moisés’ reputation for being violent. In his youth he was sentenced to several years in jail for murder, and the Matsigenka say that someone who has killed another person is likely to do it again. 13. As the conflict intensified, Moisés and Josefina increasingly, but not exclusively, referred to Isac by his personal name, thus in a sense estranging him from the category that includes sons-in-law. 14. This myth was told in August 1996 by the late Dario Mahuantiari in Koribeni, where it was recorded and translated with the help of Mirian Piñareal. Here I present the first third of the narrative more or less as it was told, while the concluding two thirds are recounted in a condensed manner. 15. For Matsigenka men’s notions of female beauty see Yu and Shepard (1998). 16. In relation to the Matsigenka, this interpretation has been argued by, for example, O. Johnson (1979, 57ff) and Snell (1964). 17. The proper Matsigenka personal names are secret, since they are seen as intimately associated with the bearer, and knowledge of a person’s name confers power over him or her. Before the diffusion of Christian names nicknaming was common. 18. Others who have shared the view of a manipulative usage of the terminology system include Baer (1984, 87), Casevitz-Renard (1977, 126), O. Johnson (1978), and Snell (1964).

2 Marriages, Norms and Structures The Dilemma of Finding a Wife among the Piaroa of the Sipapo Alexander Mansutti Rodríguez

Kinship studies are one of anthropology’s most interesting, prestigious, and elaborated areas of knowledge. Considerable effort has always accompanied this approach. Morgan (1980), Lévi-Strauss (1981), Dumont (2004), and more recently Needham (1977), Heritier Augé (1981), and Godelier and colleagues (1998), to name only some of the most noteworthy, all illustrate anthropologists’ long demonstrated skill at modeling kinship systems. Perhaps due to the nature of the kinship phenomenon, its highly formalized study has centered more on the structural rather than the phenomenological approach. One of the collateral effects of this structural focus is a difficulty in understanding situations that differ from the norm. It is presumed that when behaviors differ frequently from that expected, the system is not functioning well. When these behaviors significantly contradicted the implicit logic of the system, they were treated as irregular. It was necessary to break from the dominance of kinship terminology structures, seen as closed, self-explanatory systems, in order to introduce the complexity of their interactions with other social spaces, so as to facilitate a reinterpretation of this area of study based on new principles. This approximation has been deeply influenced by Lévi-Strauss’s seminal 1949 work The Elementary Structures of Kinship, despite Lévi-Strauss himself having warned since the 1960s (2004, 148–49) that it is an error to as-

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 37

sume kinship systems are the principal means of regulating interpersonal relations. Taking up this suggestion and helping develop this new approach, Needham (1972, 166–81) made a valuable contribution by distinguishing three analytical levels: the first is that of rules, primarily juridical rules, that are derived from what the subjects believe should be observed; the second level is that of behavior and practice, expressed in statistical terms; and the third level consists of the terminology and structures that define the system. We understand, as Lévi-Strauss warned us, that each of these levels implies different determinants. Further, such an approach opens the doors to an analysis not merely alluding to the “pathological” but rather providing an opportunity to explain that which is left unexplained when viewing kinship only in terms of structural oppositions or societal ideals. No less useful has been the possibility of creating models to see how they function over a long period of time (Århem 1981b; Chagnon 1968b; Mansutti Rodríguez and Briceño Fustec 1993); that is, to incorporate time and its consequences into our models to arrive at new conclusions about the limitations of our constructs. We have applied one of these models, helped by mathematician Cristina Briceño Fustec (Mansutti Rodríguez and Briceño Fustec 1993), to see the increase in relative age discrepancy in successive genealogical generations in the Piaroan world, and to confirm that some of the “irregular” behaviors generated in daily life and marriages are necessary in order to diminish age discrepancy between intragenerational and intergenerational marriageable individuals and maintain the effectiveness of the governing kinship structure. Without allowing intergenerational or incorrect marriages, a Dravidian kinship system tends to become entangled by the growing dispersion of members of the same kinship generation along the pyramid formed by different demographic cohorts.1 For instance, as the years pass, ego’s son may be older than some of his classificatory grandparents, or ego’s daughter may be many years younger than all her potential husbands. When this pattern occurs, when the age distribution of the population and the kin terms are so out of kilter, then new irregular marriages, for example between cross-generational egos, have the capacity to redefine the allocation of relationship terms and provide a mechanism to readjust the kinship nomenclature to the demographic reality. With greater experience, and having developed kinship studies that include economic relations, demography, ecology, and politics, and with

38 · Alexander Mansutti Rodríguez

the crucial help of computerized systems, we are able to return to the interpretation of particular cases, without their being obscured by the randomness of individual case histories. In this chapter we describe and interpret a number of case studies that in other circumstances would have been considered irregular or deviations from the canonic norms of the kinship system.

The Piaroan Kinship The Piaroan kinship system can be defined as a Dravidian type due to the emphasis placed on generations and the contrast between affinal and non-affinal relatives in GO, G+1 and G-1 (Dumont 2004; Godelier et al. 1998). The most interesting aspect of the system is that it guarantees all the Piaroans may be consanguineous relatives even though the Piaroans themselves appear not to be conscious of these relationships. Automatically applied grammatical rules, acquired by Piaroans as they learn to speak, also apply to these same kinship norms in such a way that the relationships generated become deeply embedded in social experience. As figure 2.1 demonstrates, Piaroan terminology divides the ego’s kin into ego generation and the two generations surrounding ego as affinal and non-affinal. Affines are (i) those whom ego can marry in his or her own genealogical generation; (ii) their parents in the first ascending genealogical generation; and (iii) those with whom ego’s children can marry. The non-affinal relatives, referred to in the literature as consanguineous relatives, are (i) those whom ego cannot marry of his or her generation; (ii) their parents; and (iii) the members of the generation below ego’s whom ego’s children cannot marry.2 We are considering the simplest of Dravidian systems. The structural logic of this system is implacable. The generational world close to ego remains divided into two groups; those to whom ego and his children can be married and those to whom they cannot. Matrimonial norms establish that one should marry cross-cousins. If the system is to function with its structure rigorously intact, then the contracting parties should be bilateral cross-cousins. Distinct from other Dravidian lowland societies (Henley 1996, 35; 1987, 258; Silva Monterrey 2007; Thomas 1983, 346), cultural norms do not permit marriages between uncles or aunts and ego, nieces and nephews and ego, or between grandparents or grandchildren and ego. Finally, the most favored marriages are between the

Figure 2.1. The Piaroan kinship terminology (male ego). (Mansutti Rodriguez and Briceño Fustec 1993).

40 · Alexander Mansutti Rodríguez

cross-cousins, the closer the better. These marriages bolster alliances and endogamy. Seen this way, the system should promote a high percentage of marriages between consanguineous spouses. In fact, however, this is not the case. Of the 256 marriages about which there is information and which fulfill the requirement that their progenitors’ births were registered, only 24 (9.38%) can be demonstrated to conform to some degree of consanguinity between spouses.3 A more complete genealogy might raise this figure; the increase would be among those related in the second, third, and fourth ascending generations.4 However, it would not cause changes in the numbers of marriages between first cross-cousins because these genealogies have already been included. Of these 24, 17 were between first cross-cousins, five between second cross-cousins, one between uncle and niece, and one between father and daughter. Although the low frequency of consanguineous marriages appears to be counter to the norm, this is not the case. Our information indicates that all adults believe that their marriages conform to the rules. Young people say they have little knowledge or interest of these rules. There are other interesting data on the Piaroan marriage system. One kind relates to how the generations are defined by the kinship terminology of ego’s and the two ascending and descending generations, a feature that we refer to as genealogical generations. We consider that there is an age displacement within each of these generations, a product of the wide age differences accumulating from one generation to the next due to the births of groups of siblings born of the same mother (Mansutti Rodríguez and Briceño Fustec 1993). This displacement produces two interesting phenomena. The first is that the wide age differences among individuals of the same genealogical generation result in a scarcity of possible spouses of a marriageable age.5 The second feature is that marriageable generations in isolated areas differ in age from each other so that the young may be uncles, aunts, or even grandparents of young people in other places (Mansutti Rodríguez and Briceño Fustec 1993). In these instances, marriages among young people belonging to different genealogical generations, who under no circumstances could be considered as cross-cousins, act as an ideal mechanism to reclassify the kinship network on the basis of new marriages. These new marriages become a reference point to rename all the relatives associated with both family groups and reinitiate the cycle with individuals of comparable ages. Although these marriages may ap-

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 41

pear to violate kinship rules, in fact they are transformed into a tool that resets the system, correcting for inconsistencies. Demographic dispersion also reduces the availability of spouses in these communities. We estimate that the demographic density among the Piaroa in 1982 (when the population had already undergone at least 20 years of accelerated growth) oscillated around 0.25 persons per square kilometer. Fifteen years earlier, the time (1967–1968) Overing, Boglar, and Monod did fieldwork among the Piaroa, the biggest communities scarcely reached 30 people and formed part of regional groups, and each composite group’s area was about 400 square kilometers on average.6 In the harsh and mountainous territory of the Piaroa, this was a significantly large area of land for neighborhoods. If we consider the population as a stable pyramid, 100 people in one neighborhood (a very optimistic hypothesis) would have had to be distributed among all the age groups. If ego could only marry single women of his genealogical generation—that is, those women whom ego called chusapo isaju and who were part of the pertinent age category between 10 and 15 years of age—and if among the 16 single young people in that age bracket, eight were male and eight female, then four would be ego’s siblings or parallel cousins. This means that only four would be ego’s cross-cousins, for whom ego would have to compete with his three siblings to gain a spouse. Up to this point the situation seems feasible from ego’s perspective, but we know that these small groupings, less than 500 individuals, were highly dependent upon demographic whim; that the actual distribution did not correspond to the statistical model of age and gender just described. Neither does this model take into account the effects of other factors, like recurring epidemics, which can devastate a cohort of children, who along with the elderly are most likely to be subject to these scourges. Additionally, as differentiation by age can distribute members of the same generation into different generational categories, we can see that the four young women who could be ego’s marriageable cross-cousins may already be married, or be too young to marry, and that other young marriageable women are unavailable to ego because they have been classified as belonging to generations other than ego’s. In other words, the act of finding a mate becomes more and more complicated and might only be understood if we realize that in spite of the value the Piaroa place on close family marriages, we have many cases of young adults leaving their home regions to seek adventure and ending up marrying far away from their original villages.

42 · Alexander Mansutti Rodríguez

Besides the restrictions that are part of the system, and that originate in natural events like the dispersion of births, gender, and population, the new hierarchical relationships among settlements also are beginning to influence matrimonial alliances. According to the results of research carried out in the area (Mansutti Rodríguez 1988; 2002), the Piaroan settlements can be classified into three categories: farms inhabited by an extended family; communities where there are two intermarried extended families; and those villages where three or more extended families create complex networks of marriages (Mansutti Rodríguez 1988; 2002).7 According to this classification, of the 256 marriages registered in Sipapo in the 1980s, 190 were among inhabitants of villages, 35 were among members of communities, and 31 were among individuals living on farms. It is even more interesting to note that although 70 people came from family farms, 53 of them were married to people coming from the largest villages. In the two family settlements or communities, there were similar results: more than 50% of the alliances were established with people from the largest villages. Our data indicate that 80% of the marriages registered in Sipapo were with people from the large villages and that within this group, 59.5% of those marrying came from small settlements. This trend signifies that not only does the availability of marriageable people affect the distribution of Western wealth that comes with village life; it also creates the conditions that lead families in small settlements to seek alliances with the wealthier village people. Marriage is reaffirmed as a political act. Marriages are also affected by the new religious movements that have sprung up during these changing times. The most visible are the Evangelicals who have been building a new system of leadership and new relationships among communities based on their shared faith.8 The Evangelicals meet frequently in different places where they celebrate their services and rituals. These encounters have come to replace the Warime festivals (Mansutti Rodríguez 2006) as situations that offer opportunities to cement alliances. The Catholics, although less active in creating encounters, have also had an impact on the young people. They meet during the years that they are studying at mission schools, freed from family ties and fueled by a liberal climate that promotes affairs. When we look at this whole context of relationships bypassing the kinship system, the growing distance between norms and actual events is not surprising. The restrictions and new conditions that affect marriage

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 43

choices explain many of the alliances. That which seems to be abnormal about the logic of structure helps to mitigate the consequences of the weakness of the system and even normalizes it. At first glance it seems paradoxical that in the long term, those marriages outside the canon support the apparently cumbersome system and help enable the realization of the canonical ideal. The circulation and meeting of individuals, with their biological reproductive capacity, allow them to take advantage of the flexibility of the kinship terminology’s allocation and reconfiguration of kinship networks to assimilate and normalize Piaroa marriages between relatives who are not classified as crosscousins (chusapo or chusapo isaju), the ideal marriage category. They can be redefined to correspond to the ideal, and include older brother and older sister (chubuo or chubua), younger brother or younger sister (chijawa or chijawaju), nephew/son-in-law or niece/daughter-in-law (chujori or chujoriju), or even eldest brother/eldest sister (akarua or akaruaju), younger brother/younger sister (rujirua or rujiruaju), aunt/mother-in-law (chimiyaju) or uncle/father-in-law (chimiya). All these marriages violate the normative system. In this way, social practice corrects the distortions created by the kinship system operating in a small population randomly distributed. We examine a few case studies in which the marriage choices that individuals make do not correspond to the logic of their kinship terminology. We also explore individual histories in the light of a system of matrimonial alliances in which the logical choices of the participants are mediated by circumstances apart from the terminological system.

Case Studies Love: Shamanic Power and Forced Migration In about 1970 a Piaroan couple fell in love. They lived in communities along the Cuao River in times when matrimonial orthodoxy was the norm. The young man, named Paris, who lived with his parents, decided to declare his love for a young woman, and she accepted him. Immediately the young man ate with the young woman’s family and, with the permission of his in-laws, moved his hammock to their house. Both were classified as cross-cousins. What the two had not foreseen was that an old

44 · Alexander Mansutti Rodríguez

chanting-shaman, a meyeruwa, who had earlier announced his intention to marry the girl, would come persistently to claim her.9 The couple and the young man’s family felt themselves under attack by the meyeruwa, and for fear of the chanter, decided to flee and settle along another river system, where it would be more difficult for the chanting-shaman to harm them with his spells. They fled in fear, but their fear was never stronger than the emotional ties that bound them together. Their decision involved not only themselves but also the rest of the young man’s family. In this case, to save their marriage the young man and woman and his family group rose up against the exercise of arbitrary power employed by the chanting-shaman. Loneliness: Vulnerable Women and Post-Marriage Residence Yawika was young, intelligent, and capable. He was the only son among three groups of descendents of a great chief named Rafa, a Lord of Warime, a chanting-shaman (meyeruwa), and a chief of commerce and neighborhood (tjujaturuwa).10 He was renowned as the most powerful chief along one of the great rivers in the Sipapo watershed. In spite of having had various spouses he only had three children: two girls and a boy. Rafa’s territories were along the headwaters of the Autana River, near the commercial path to the neighboring Marieta River along the harder mountain route, where he had various business partners with whom he maintained close ritual relationships, both of friendship and commerce. In his neighborhood Rafa had three plots of land productively planted that he used alternatively. One day he was told that on the Marieta River, Miru Kuapo had died, leaving orphaned a prepubescent girl and a boy. He decided to adopt the children and arranged the marriage of his eldest son to the girl. The kinship relationships between the two varied depending upon which kinship network was selected to establish them. One was chosen that defined them as cross-cousins, and when the girl had her first menses, the marriage was consummated. In this way the great chief avoided having to send his eldest son, Yawika, to do bride service for some other eligible bride’s parents in their neighborhood or community.11 Yawika did not like his wife. Nevertheless, they were able to form a strong domestic group and had six children; three boys and three girls. Yawika’s eldest son married a woman who had separated from her first husband. The son, in a similar fashion to that of his father, took his bride to settle in his father’s community, going against norms prescribing that

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 45

couples recently united should live, for several years at least, in the bride’s family’s house. In these cases, the post-matrimonial residence rule was avoided. Passion: Sentiment and Manipulating Kin Connections In the 1990s Agapito was about 40 years old. Twenty years earlier, he was a highly eligible bachelor. He had recently graduated from the Salesian Missionary School at Isla de Ratón, where he had learned to service combustion engines and had been contracted by the state to build houses. Although he showed no great capacity for leadership, he had authority in spite of his youth. The families of young Piaroan women in his community and surrounding communities tried to win him over, but unlike most of the other young men of his age, he did not want to commit himself. He lived as adventurous a life as a young Piaroan man of his age and time could. His father was a liberal man; an extraordinary leader and a visionary for his time, who had decided not to influence his children’s marriage choices. Nevertheless, neither Agapito nor his father had reckoned on the stubbornness or persistence of Luisa Elena, Agapito’s father’s sister. She wanted a man like Agapito to be her daughter’s husband and her son-inlaw. She was married to an older man and felt she needed the services of a son-in-law, and so she began to put great pressure on her brother and urged him to persuade his son to marry this patrilineal cross-cousin. After a short time Agapito gave in and married his cousin. Very soon she was pregnant, as is usual among the Piaroa. The couple settled in his wife’s community, while Agapito’s paternal family nearly all lived in another community. Luisa Elena and her daughter demanded a lot from Agapito and made his life difficult. One day Agapito was discovered making love to a young woman. She could either be classified as his aunt or cross-cousin, depending upon the kinship links used to define the relationship. Since he was a married man and since it could be calculated that he had entered an incestuous relationship with his aunt—if the connection were defined through the young woman’s mother’s family line—Agapito was thrown out of his lover’s house and abandoned by his wife. Single again and subject to rumors of the local community, he was judged harshly. Nevertheless, Agapito and the young woman were in love and persisted in their relationship. In a short time the rumors died down and the young woman’s mother accepted them. Agapito and his parents-

46 · Alexander Mansutti Rodríguez

in-law decided that the way to legitimize the relationship was to classify it through the young woman’s father’s family line, according to which Agapito and his lover were cross-cousins. After 20 years, they have had five children. Shame: Incest and Kinship Terminologies Ramona is a middle-aged woman. She is raising two sons and two daughters. In the community where they live, their situation is marginal because they do not have active kinship ties with other family groups. When we interviewed her, she was living with a senile old man who had come to visit the area and had stayed with her. When we asked Ramona about her previous life, she spoke to us about her children’s father and how he had died many years before. He was, according to her, a cross-cousin. As the interview progressed, a reconstruction of her parents’ family history indicated to us that the link between Ramona and the children’s father was obviously a close classificatory father-daughter relationship. When we returned to the area, we visited her and raised the topic again. She began to cry and told us that the relationship had been father-daughter, and that from the outset her incestuous marriage had brought her bad luck and ruined her life. Nevertheless, she and her husband carried on their life together and constructed a kinship relationship that from the outside appeared to be legitimate. She explained her marginal situation in the village. If she had stayed where she was born, many people would have known of the incestuous relationship with her husband. They went to a village where their kinship relations were defined by the places that they occupied in the kinship network of her eldest daughter’s husband. This reference facilitated the creation of a terminological system that achieved an apparent consistency and legitimacy. She was no longer an incestuous wife. Rather, she had transformed herself into the mother-in-law of her daughter’s husband. From that point on she wove a network within her new community, while the old network, geographically distant, remained in place as a potential source of support if she were ever to need it. Solidarity: Evangelical Ties and Kinship The relation between the Piaroa and the Hiwi is tense. The Piaroa are semi-sedentary horticulturists who maintain fruit trees. The Hiwi were

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 47

nomads until quite recently and do not hesitate to chop down trees when they have fruit, or to enter and make use of others’ gardens when necessary. The values of both peoples are in sharp contrast and frequently result in conflicts. Nevertheless, within the Piaroan system of land settlement, there is an exceptional case. In the community of Cañó Piojo, the Hiwi and the Piaroa have established a number of inter-ethnic marriages. What brings together two such dissimilar peoples in the same communal area? The Evangelical religion and the charismatic leadership of MJ, who, as pastor, has encouraged these marriages and inter-ethnic collaboration. In spite of the fact that the village has hardly grown, and that a number of individuals have been expelled from the community, nevertheless, it is extraordinary that these two ethnic groups have lived together for over 40 years, even though there have been many conflicts. In fact, sharing the Evangelical religion has become a very important criterion, although not the only one, that decides whom people may marry. Opportunities to arrange marriages arise during religious services and celebrations, as we have already noted. Belonging to the same faith changes a family’s frame of reference and the active kinship networks determining whom ego is most likely to marry, which make possible alliances that once seemed impossible—villages half Piaroa and half Hiwi. Ambition: Kinship and the Pull of the City Juanita is a likeable young woman. She has shown lots of initiative: she studied in the Salesian Mission School through the second year of high school and then decided to explore the world. She went as far as San Fernando de Apure, a criollo city on the banks of the Apure River. There she had various jobs as a maid and established relationships with young criollo men. Nevertheless, she returned to Amazonia without children and met a Piaroan teacher with whom she lived and had two children. The man came from the Samariapo River, where he had his family. Because of the distances between the family groups, it was easy to create a system of kinship based on their marriage; in consequence, their marriage turned them into cross-cousins. Juanita grew tired of her husband and left him; he drank too much, mistreated her, and even raped one of her nieces. The family supported her. A little while later Juanita met another Piaroan teacher in Puerto Ayacucho, who came from an Evangelical district near the city. As Juanita’s

48 · Alexander Mansutti Rodríguez

father was a leader in his community and was interested in the growth of the village he had founded, she convinced her new husband to work as a teacher in her father’s village. They made their lives there and remained in the community until her father died and her husband retired. After that, Juanita and her nuclear family moved to Puerto Ayacucho, where they remain. It is clear that Juanita’s matrimonial decisions, supported by a father, were always oriented toward having at her side someone who would bring her close to the object of her desire—the West. Fear: Marriage and Shamanic Power In 2000 an old shaman, a master of Warime and well-known soplador or sorcerer (yuwawaruwa), decided to move from the Cuao River to the Autana River in the watershed of the Sipapo. This old shaman was considered to be very dangerous. Another shaman, who lived in the area to which he moved, was a chanter (meyeruwa). His authority in that sector of the river was incontestable. He was not a Lord of Warime, but his powers as a chanter derived from the “Lord of Báquiros” and protector of a poitjiana—that is, a “stone mother of the fish” who maintained the supply and variety of fish in the Autana River. When the old yuwawaruwa settled along the Autana River, the meyeruwa came to visit and learned that the yuwawaruwa had a 12-year-old son; he offered the yuwawaruwa his 18-year-old-granddaughter as a marriage partner for the son. The young woman came from the Salesian Mission School, where she had learned to dress in an occidental manner and enjoy the pleasures of the criollo world. She went about all the time with her sister, a year younger, whom she had to leave alone to attend to her young husband. The boy, scarcely out of childhood, had been raised in the most traditional Piaroan manner. The marriage was consummated and shortly afterward the young woman became pregnant. When her first child was born, she learned that her younger sister had just died. The family had found her hanged in her parents’ house; it was believed she died because of a malign spirit’s influence that was suspected to have been sent by the old shaman, the fatherin-law of the elder daughter. Fifteen days later the young woman married to the old shaman’s son was also found hanged. The meyeruwa’s family reclaimed her, took her lifeless body, and buried her near their community. The family of the

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 49

yuwawaruwa (soplador) kept the son of the dead woman. At this time rumors that the feared old yuwawaruwa had assassinated the two sisters were widespread throughout Piaroa lands. The old meyeruwa organized a vengeance ritual that had no effect. On the contrary, he had a stroke; and a year later, a second stroke killed him. The battle was lost. Fear of apparent manifest shamanic power, as we have seen, can be a decisive factor in choosing a spouse. Nostalgia: Sentiment and Political Decisions Don Fulgencio was an extraordinary meyeruwa. He was the central figure in an extensive and influential family network in the largest of the Piaroan communities along the Orinoco River. This network included an important domestic group whose lands were along the Urudei River. Don Fulgencio was originally from the Marieta, a neighboring river, but from a very young age he had established himself along the Autana; he had married the daughter of an old leader who had business dealings along both rivers. Because of established relationships, Don Fulgencio lived along the Autana but traveled often to the Marieta to visit his relatives and, more important, to maintain his commercial ties. When his father-in-law died, Don Fulgencio left the headwaters of the Autana to go to the Orinoco, and his relations with people along the Marieta became more distant. Meanwhile, his children grew up and became marriageable while he grew old. While living along the Orinoco, Don Fulgencio converted to Evangelism and traveled to church meetings. On one of these trips, he became reacquainted with his relatives along the Marieta. One member of the family network, PG, decided to move to Marieta, where he died. Don Fulgencio did not go there because he had consolidated a position of leadership in the large Orinoco village for which he was responsible, but he convinced two of his adolescent sons to marry in Marieta, which they did. Why would a leader like Don Fulgencio decide to return to his original home instead of strengthening his relations in the large village where he lived? In our judgment, it was nostalgia that played a fundamental role. Don Fulgencio always remembered that time of his life along the Marieta. He had gone to live by another river to find his wife, whose father was a well-connected businessman. In addition to the ties he formed through his in-laws, in competition with his brothers-in-law, he became the leader

50 · Alexander Mansutti Rodríguez

of his faction and assumed those responsibilities. Yet when he knew his life was coming to an end, he chose to return to his memories. He decided to reactivate the kinship and affinal ties of his childhood. These sentiments shaped his political decisions.

Discussion When we realize the random nature of individual responses to particular circumstances, which all occurred within the framework of the same society, we understand the importance of the three analytical levels of kinship developed by Needham (Parkin 2004; Godelier et al. 1998) or the simpler scheme developed by Bourdieu (1980) of “official kinship” versus “practical kinship,” also characterized as “hot kinship” versus “cold kinship.” In all of these approaches there is an effort to differentiate analytically between the systems of marriage and kinship as they ought to be and as they are. The structure alludes to synchronicity, consistency, and abstraction, while the actual processes are diachronic, leading to the dysfunctional and the concrete. An analysis of kinship systems can only be fully realized using a concrete and diachronic approximation that recognizes incoherencies as an active part of the processes and not as anomalous events. The individual histories reveal that marriage decisions in any given circumstance are not only, and not even principally, a matter of the internal logic of kinship terminology. Neither are they a purely individual choice. Rather, what happens is that the individual, and his or her personal history, are surrounded by determinations made by structures, norms, politics, networks, and personal feelings, to name just the more important factors that make it possible to understand the decisionmaking process. The Piaroa are today, and were yesterday, free to choose among alternative options offered to them by their system. Like our Paris, an individual can rebel against the arbitrary power exercised by a shaman, assume the risk of defying him, and escape with the woman he loves or, on the other hand, accept the decision of the shaman, give up the woman, and find another wife. Although the latter option appears to have been the more economically sound in this case study, it was not chosen. Our couple was free to opt among the different scenarios their system offered them. In this we can see the weight that sentiments contribute to the analysis, as Needham has already postulated (1977, 48–49).

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 51

The same applies to the incestuous couple. The couple decided to marry and, in time, abandoned the place where their transgression was well known to settle elsewhere, where they could recreate their kinship relationships to conform to the appropriate norms. To have an incestuous relationship is not necessarily a tragedy, nor does it have to threaten the integrity of the system, because the system is sufficiently flexible to replace itself with a different focus: a man can be redefined as a “son-in-law” instead of a “father.” Yesterday as today, the flexibility of the system permits playing with its rules depending on the correlation of forces, in conjuncture. A man can marry an orphan girl to avoid conflicts with norms requiring uxorilocal residence after marriage. If the bride’s in-laws are dead, with which family does the groom stay? The obvious response is to live with his father’s family. In this case the reasons for demands against the system come from a father who, for political reasons, persuaded his favorite son to marry an orphan so as to stay in his father’s village. Relationships that are appropriately defined in the terminology can be countered by the whims of a powerful person. In our field notes there are frequent accounts of old shamans demanding young women for themselves, in spite of the fact that they were inappropriately related to the women. It is also common that these demands are denied. Piaroa sensibility usually triumphs over desire, but sometimes it loses. These decisions are more related to power than to the rules. Seen in a broad emic perspective, the norms that promote matrimonial networks seem to aim for an impossible equilibrium, the ideal equilibrium of structure that is translated into the language of norms but cannot be translated into the language of sentiment. Don Fulgencio’s nostalgia leads him to marry his two sons to descendants of old allies from his youth. Curiosity and ambition lead Juanita to look for a mate among those who can satisfy her ambition to be near the criollo world, without having to renounce her family of origin. MJ founded a village where Hiwis and Piaroas can be allies; where there is inter-ethnic marriage, in a society that prizes endogamous intra-ethnic marriage as the norm. Fear motivates the most prestigious chant-owner of a river to arrange an inadvisable marriage with the aim of protecting himself—in the end unsuccessfully—from an apparently dangerous adversary. Rafa’s ambition leads him to break the circuit of reciprocity in marriage exchanges so that his

52 · Alexander Mansutti Rodríguez

eldest son will not abandon him. Agapito’s passion leads him to break the commitment made by his family and look for a woman to whom he is attracted, in spite of knowing the scandal this would cause when the union was discovered. Finally, “love,” understood as the unstoppable desire to be with another person, leads Paris, his wife, and family to flee the threats of an old shaman. In all these cases there were transgressions against the norms or an intent to transgress against them; in all the cases, the norms still apply even though violated, and the kinship network reconfigures itself from the new ego-centered nucleus that reordered it. The kinship norms are not a yoke; they are a language that orders and makes comprehensible and viable the relations between couples, including those who break the rules. The cases can be multiplied. Each individual case can be interpreted in terms of sentiment. Bourdieu (1980, 282), studying Arab kinship, said: “It is practical kinship which makes marriages; it is official kinship which celebrates them.” In this system the terminology and norms are languages that allow individuals to orient, interpret, and reinterpret the processes but that are always more or less distinguishable from the results of that same process. What is really fundamental to our task is not to quantify the number of marriages that demonstrably require a readjustment of genealogies but to see how many marriages are openly recognized as transgressions and how many manipulate kinship norms to impose correct solutions over incorrect ones, without open moral or practical sanctions. If, in practice, most abnormal marriages reconfigure the network and can act as if they were correct, the system is efficient. We must remember that we are describing living societies that, in order to reproduce themselves, have to mitigate the restrictions making this difficult. At this point it is important to consider and to justify the role of the anthropologist as analyst. When we describe and discuss kinship and matrimonial systems, we do not do it as members of indigenous cultures, nor simply to reproduce and repeat what they tell us. In fact, if that were the case we would not need anthropologists; indigenous people can do that for themselves. Our task is to take what our informants tell us and reflect on it, using and developing the theoretical tools anthropology provides us. To develop and create theory and interpret social data, based upon a comparative perspective, are how we repay, in part, our debt to the world’s remaining rich human experience. We welcome, therefore, the decisions

The Dilemma of Finding a Wife among the Piaroa of the Sipapo · 53

individuals make, based on their sentiments and interpretations, and in the understanding that they also form part of the theory. In order to analyze a kinship system, we have to include a study of its relationship to other social domains that it influences, while at the same time we study how these other social influences affect kinship. We are examining a kinship system in which decisions have influence on and are influenced by politics, demographic distribution, access to resources and services (including those of religion), old and new sources of wealth, and the profile of the individual who makes choices within this framework. The interpretations of those involved in these social processes do not necessarily explain them, but they certainly are part of the explanation. The same is true of their sentiments. And so we arrive at a complex approximation of kinship that we use to understand it. It is a system of kinship plagued with contradictions and influences, with the heaviest coming from the dynamics of demography, potential technological developments, politics, and most recently the processes of the intercultural domains fostered by industrial society. Such a system has needed to mix old and young, generation and generation, in order to maintain the system and open up possibilities of alliances that could be closed by the rigors of the system. Today, furthermore, the actors must interpret outside influences and adjust accordingly.

Notes 1. See Mansutti Rodríguez and Briceño Fustec 1993 for a detailed discussion. 2. We prefer non-affinal to consanguineous because, as Dumont has expressed clearly in his seminal work of 1954 (Dumont 2004), the category “consanguineous” includes its opposite sense “affinal” but not, we add, that to which the category actually refers. In practice, “affinal” and “non-affinal” are in many cases considered consanguineous in the sense that both may descend from common ancestors. 3. Data on family genealogies and marriages were processed by M. Marion Selz, a computer researcher at EHESS in Paris, to discover marriages that demonstrated some degree of consanguinity, as they were repetitive marriage alliances over the generations. We are grateful to her for her assistance. 4. Among the Piaroa, people who died are not evoked by name. This omission results in a rapid loss of genealogical memory as people only remember those whose names are known personally to them. At most, our informants can name their grandparents, and even then, only if they knew them. This situation is further aggravated for those who were orphaned, and who frequently do not even know the names of their parents. Our

54 · Alexander Mansutti Rodríguez

genealogies are therefore full of holes; there are individuals who we know existed but who cannot be named. 5. A genealogical generation is that group of individuals who share among themselves membership in the same terminological generation (G-n). In practice, they constitute all the descendents of the anterior genealogical generation. Thus, for example, in the Piaroan system all the following are of ego’s genealogical generation: father’s and mother’s sons and daughters; mother’s brother’s sons and daughters; mother’s sister’s sons and daughters; father’s brother’s sons and daughters; father’s sister’s sons and daughters; and mother’s and father’s cousins’ sons and daughters. All ego’s father’s genealogical generation are members of the father’s genealogical generation, and so on. In effect, the age differences among members of the same generation signify that there may be no coincidence between a demographic cohort and a genealogical generation. 6. Neighborhoods, called itso’pha by Kaplan (Overing 1975). 7. My doctoral thesis (Mansutti Rodríguez 2002) demonstrates the existence of a hierarchy of settlement patterns, which relates to the employment in the most important communities of a whole range of public servants—religious, political, civil, sanitary, and educational. This incursion into these communities, which I call commercial centers, has transformed their capacity to acquire Western goods and maintain large populations. 8. At the time we conducted our investigations, the Evangelical religious networks were extensive. Today their influence has been weakening and is substituted by the practice and ideology of the revolutionary political party, the Partido Socialista Unido de Venezuela (PSUV). We write in the ethnographic present. 9. Meyeruwa are owners of magical prayers and resources in opposition to yuwawaruwa mari, who are the owners of whispers (yuwa) and sorcery (maripa) (Mansutti Rodríguez 2006). The meyeruwa controls creative shamanism and the yuwawaruwa controls sorcery, the destructive. The maripa is very dangerous and is the speciality of yuwawaruwa mari or soplador but also can be used by meyeruwa or rezador. Yuwawaruwa are always more dangerous than meyeruwa. 10. Warime is a complex ritual of masks and sacred flutes that serves to consolidate the relation between shamans and the owners of natural resources. It is a modality of yurupari (Mansutti Rodríguez 2006; 2011a; 2011b). The man who inherited the right and instruments to direct the Warime is called warimeruwa or “owner of Warime.” 11. If Yawika had not married the orphaned girl, he would probably have had to leave his father’s house to go and live with his wife and her parents and would have had to stay there for several years.

3 To Be Seen or Not to Be Seen! Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon Daniela Peluso

Amazonian marriage unions are most commonly described and analyzed in terms of whom one marries rather than how one marries. Irrespective of change, these dimensions are consistently entrenched in Ese Eja community sociality and politics.1 With wavering intensity in time and place, Ese Eja couples have long been choosing to initiate their marriage, either publicly, hamatijawiaki, or secretly, ejakewawanaki.2 By comparing the different ways in which marriages commence, this chapter addresses the unspoken relationships between the individual and the group that Ese Eja marriages reflect, including the power relations in which marriages are embedded. In discussing public marriages, I emphasize how marriage is legitimized by what Ese Eja call “seeing”; the public involvement of kin and neighbors in the union. Contrastingly, in secret marriages, couples are united without any agreement or public acknowledgment from either side’s family. Here, I argue that secret marriages tend to fail because of the way that power is construed within indigenous Amazonian communities, given the absence of a central authority or state. Individuals who marry in secret deprive their family and neighbors of speech, action, and exchange, thereby revealing and challenging the underlying mechanisms of power. Despite their poor outcomes, secret marriages, like extra-marital affairs, persist and present a means for individuals to bask in a short-lived reprieve from group authority. Finally, the various ways in which marriages are initiated further accentuate the importance of marriage as a process

56 · Daniela Peluso

that mediates contemporary economic cooperation, production, and regeneration and often mark their success or failure.

Marriage as It Was, Marriage as an Ideal Marriage, hamatijawiaki (common law), is still a relationship that most individuals seek to attain and maintain.3 Similar to other Amazonians, Ese Eja individuals focus on establishing and maintaining a union, not on the act itself of getting married.4 Therefore marriage is not usually tied to a specific event but is the result of an ongoing process involving reciprocal collaboration between partners, a process that happens primarily after the initial union. As such, the way a couple initiates their marriage is part of that process and, in turn, is a public statement, an avowal toward the way they envision themselves amidst their respective families and their surrounding social worlds, a point to which I return. Marriage ideologies have changed dramatically since a few generations back when Ese Eja individuals described potential marriage partners as preferably belonging to the opposite patrimoiety. Nonetheless, there are still basic rules about whom one can and cannot marry. People of the opposite sex are in either wapa (marriageable) or wapa pojiama (nonmarriageable) categories, the latter term often translated into Spanish as “family.”5 Parents (all direct ancestors and descendants), children, samesex siblings of parents, and parallel cousins are all wapa pojiama. The category has recently been extended to include parents’ cross-sex siblings. Sex with wapa pojiama is memoo (incest) and can be the cause of illness, most often skin disorders. Although sexual affairs between wapa pojiama occur, marriages do not. Ese Eja also group themselves into two types of people, Icha and Kaka. Ese Eja, like other Macro-Panoans (Dole 1979; Kensinger 1995), are a dual organization society that were most probably once structured as a twosection patrimoiety.6 Icha/Kaka identity is a birthright and is transmitted patrilineally, from fathers to their sons and daughters.7 It appears these two types were once exogamous, requiring or tending to have members marry outside their group, thus upholding its boundaries. It is always a question of one or the other identity; they are never mixed. Children are born either Icha or Kaka; an identity inherited from their birth father.8 Correspondingly, the term for “father” and all of their classificatory fa-

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 57

thers is either Icha or Kaka. An individual speaks of “being an Icha-person or a Kaka-person” rather than as belonging to a group or moiety.9 Ese Eja communities no longer operate as exogamous patrimoieties, however. Polygynous households, once reflecting the status and prestige of a few men, are now uncommon. Etiikiana (elders) tell not only of a time when Icha could only marry Kaka and vice-versa but also of their own more recent experiences when it was preferred that toto (MB) and boyase (ZD) marry each other. These preferences complement Dravidian kinship address terms and reflect only a small portion of potential marriage partners. As Ese Eja increasingly began to see this oblique marriage relationship, whereby men married their sister’s daughters and women married their mother’s brothers, as incestuous, an FZD (paternal crosscousin) counterpart became favored.10 Now this FZD arrangement also is not as popular as it was a decade ago and has yielded to a much more open system. Just as anthropological kinship theory has moved away from its foci on descent and alliance, so do Ese Eja give less importance to exogamous patrimoieties and the types of alliance structures that result from ZD, FZD marriage arrangements. Whereas until relatively recently marriage was prescriptive and applied a positive kin-based rule, today individuals tend to form marriages as they see fit, with strict definitions of incest (memoo) being the only prohibition, despite shifts of who exactly is included in categories of wapa (marriageability) and wapa pojiama (non-marriageability).11 Like other modes of relatedness, marriage is characterized by a high degree of individualism and flexibility in face of increasing exposure to deja (non–Ese Eja) attitudes toward such ideas as avuncular and crosscousin marriage, romantic love, and individual choice.

To Be “Seen” Marriage frequently begins as a discreet but visible process following a mutual decision between a man and women, with the man subsequently moving into the woman’s household: uxorilocal residence is by far the most common choice. He brings his mosquito net and possessions to her family’s household, and she begins to cook for him. He then begins to hunt, fish, and plant for his wife’s household. These movements occur openly and deliberately for all to see.

58 · Daniela Peluso

Literally and metaphorically, ideas of seeing, learning, and knowledge interplay in significant ways. In Ese Eja, a Tacana language, when people say they know something they say “ekueya eba.’” The verb eba’ means “to see/know/learn” (as connoted by the verb stem [ba]). Where knowledge is experienced as a multifaceted interaction, one learns by observing and thus forefronts “seeing” as a primary source for knowledge. For many reasons, an Ese Eja epistemology of testimony does not privilege the acquisition of belief simply through what a speaker says.12 In fact, speech must be preceded by sight to be effectual. Less regularly, parents may also initiate a marriage by bringing their respective children together before the community so that their union can be openly witnessed by all. Similar to Cashinahua settings, these special meetings set young adults on the “right track” concerning their duties toward each other as husbands and wives (McCallum 1989). During political oratories by older individuals, women are told to cook, to bring food, to carry wood from their fields, and to work with their husbands.13 Men are told to hunt and fish so as to bring home food and to work well with their wives to make fields. Both men and women are advised to treat each other well, to cooperate, and especially to be wei’ama (not lazy). However, this public spectacle can be fraught with the same tensions as are scandals; this was the case in the first public marriage that I witnessed: Suddenly there was a big commotion as Naba’ai’s father called the community together. Naba’ai and Beyobeyo, ages 13 and 16, were rumored to be lovers for several months. When we [my “sisters” and I] went to see what was happening we found Naba’ai crying and trying to cover her face in embarrassment. Both of her parents were present. Only Beyobeyo’s mother was there, his father had gone off fishing days ago with his youngest co-wife and their children. I was used to seeing Beyobeyo looking tough and light-hearted but now he looked absolutely frightened. Naba’ai’s father announced that he had “seen” the young couple together and that now they were going to be “seen” together by everyone. The older men and women in the community came around to give advice. Then we all left and Naba’ai and Beyobeyo reluctantly went to live in her father’s first house, the house he shared with Naba’ai’s mother. (Fieldnotes 1994) I was surprised by how harsh it felt. Naba’ai and Beyobeyo seemed miserable. After all, so many adolescents were having sexual liaisons; granted,

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 59

many of them were with married individuals, which meant that their liaisons would not typically result in marriage. But still, why had Naba’ai and Beyobeyo been singled out? My female friends remarked on how Naba’ai’s father was tired of the way she had been ostentatiously running around with young boys. They explained that her father was still sad about his oldest daughter having married secretly, ejakewawanaki, a clandestine marriage that takes place without any public acknowledgment from the families involved. Their position was one of pity for Naba’ai; they felt sorry for her because she had been enjoying going to school and they knew that she did not want to settle down.14 That was the first time I had heard about the disappointment of having your children ejakewawanaki. I wondered what lay behind this pity for Naba’ai and what caused her father’s dismay over his eldest daughter’s secret marriage? The women laughed as they related stories about having to submit to their parents’ marital wishes. Amidst these stories were tales about losing a potential partner in fear of public retribution as well as tales of defiance. Some of the eldest women recounted how as young children they were forced to marry much older men and how they dreaded their obligations and eagerly awaited the men’s deaths. It did not matter that these men might have been good providers; they were seen as old, unattractive, and often mean-spirited. My friends viewed Naba’ai and Beyobeyo’s marriage as a “forced” marriage, a relic from the past. Although today some marriages are still occasionally arranged, people are no longer penalized for resisting them.15 All new unions are shaky and are never neutral; they create change and can potentially lead toward or exacerbate conflict.16 While couples forced together by their families may eventually grow fond of each other, quite often problems arise, and one or both of the partners may give up. It was only a matter of days and there was soon another uproar. This time I found Naba’ai looking down at the floor in embarrassment. Beyobeyo was off to one side surrounded by his closest friends where they watched his mother hurl high-pitched insults at Naba’ai. Apparently Naba’ai had been caught putting “poochi,” chicken shit, on the plantains she was preparing for her husband. Finally the commotion calmed down and Beyobeyo and a smaller crowd remained. Several of the more recently married men told him of their own personal experiences of the hardship as newlyweds. One young man

60 · Daniela Peluso

explained how his wife used to wash his clothes badly on purpose but that one day she began to care for him. Another told of how his wife pretended not to know how to cook. These men were encouraging Beyobeyo to be patient. (Fieldnotes, Bolivia, 1994) Naba’ai’s refusal to cook or have sex with her husband is a familiar strategy used by women throughout the world for breaking up marriages or achieving other aims (Strathern 1972; Siskind 1973; Rosaldo 1974; Holy 1985; Feldman-Savelsberg 1995; Holtzman 2001).17 When I told my friends in Sonene about Naba’ai’s provocative tactics, they laughed easily and shared their own similar and equally creative stories of dissuasion. One of the annoyed yet amused young men boasted that his counter-solution would have been to oblige Naba’ai eat the contaminated plantains. Another friend who was in love with a young girl who refused his marriage propositions felt that there was no solution. He remarked, “How can you live with a girl if only her father agrees. By force? That does not work!” Despite all the humor, scandal, and commentary, Beyobeyo was consoled and Naba’ai had been reprimanded—something that would not have happened had the couple married secretly. When I left Bolivia I was under the impression that Naba’ai and Beyobeyo’s marriage was ending, but when I returned a year later, I was surprised to find them living together with much ease and tranquility. It was only later that I began to appreciate the support that had been offered to them during their more challenging times together, support that people involved with “secret” marriages were not afforded; a support that is as much about individuals and families as it is about group and community power relations.

Not to Be Seen Ejakewawanaki, “secret” marriages, address broader questions about marriage. Secret marriages have long called into question assumptions about prescriptive marriage rules because in many cases they defied the rules outright. Yet now that prescriptive rules have given way to more flexible desires, secret marriages persist. Individuals who marry in secret shield themselves from vocal criticisms and acts of interference from their immediate families. In such arrangements marriage partners, as couples and as individuals, retain a high degree of independence and are not visibly held accountable for their actions in the marriage by their respec-

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 61

tive families and in-laws. Although this may provide a temporary and short-lived haven, it simultaneously denies speech, action, and exchange, thereby revealing and challenging the underlying mechanisms of power and how marriage is embedded in such relations. Ejakewawanaki mostly occur in early adolescence and among young adults with a fair amount of affinal and social distance who have not yet settled into a lasting relationship, and secret marriages often do not last.18 Once the man is discovered in his lover’s mosquito net, the family accepts the union but attempts to exercise limited direct influence over the new couple. One friend of mine built a fence around her house so that she could close it off at night to prevent such an ejakewawanaki for her daughter. Eventually, however, a boyfriend managed to open the fence and enter the house through a loose floorboard plank. When they were discovered at dawn they were considered ejakewawanaki. Months later, no one would reprimand this young man when, on drunken occasions, he would beat his wife for her alleged flirtations. In the early hours of the morning during an ongoing party, it was disconcerting to hear her screaming out, as her husband struck her, for her father’s and brother’s help and to see that no one went to help her. Her brother was angry with his sister’s husband but restrained himself from going to her side. When I asked him why he wouldn’t go, he said, “I never saw them—now I cannot speak!” Over and over again, I heard people refrain from helping their close relatives, always for the same reason, for having “not seen” them unite. Without sight there was no authority. The public visibility of a marriage renders it inclusive by constituting family and community members into its processual making. Visibility exposes the marriage to a group of witnesses who, in turn, may intervene to help or support them as individuals or as a couple. Contrastingly, when an individual chooses to marry secretly or “unseen,” it is a statement of independence not just from their families but also from their communities. Ejakewawanaki exclude the potential assistance of kin and thereby minimize the meaning and importance of individual family members’ roles in their marriage and all that entails. Non-relatives are even more reluctant to act openly or deliver opinions about instances that arise in the marriage. Although secret marriages cannot bar relatives or friends of a sense of responsibility toward individuals, they do effectively inhibit people from acting upon those feelings that would otherwise be a usual part of family and community life. As such it is understandable that Naba’ai’s

62 · Daniela Peluso

father felt such dismay over his eldest daughter’s secret marriage and that he aggressively sought to ensure that his youngest daughter was not in the same predicament—thus ensuring that he did not lose his voice and the potential for action in her future.

Bride Service Indubitably, marriages initiated in secret also raise broader questions about bride service and a general economy of reciprocity. With ejakewawanaki there is minimal if any bride service, though this mostly depends upon whether the couple abides to a period of uxorilocality. Yet as an institution, even among marriages initiated in public view, bride service has long been a flexible, open-ended, and negotiable arrangement between an Ese Eja man and his wife’s household. No gifts, material or nonmaterial, are given to the bride’s family as a means of establishing the right to marry. The Ese Eja case contradicts what both Dean (1998) and Knauft (1997) refer to as a widespread growth in Amazonian “bride price.”19 Indeed, it further goes against a growing literature interpreting cases of elopement and “bride capture” as ways to avoid bride service or dowry payments and further pronounces them as ways to enact freedom of choice (Ayres 1974; Bates 1974; Lancaster 1981). Ese Eja bride service coincides with the beginning of a couple’s marriage and only informally concludes when a couple begins to raise their own children or set up their own household. During this time a man demonstrates his willingness to contribute to and assist in the new household to which he now belongs; on some days his input is more significant than others. The same family that complains about an individual’s bride service one day may boast about it the next. Ese Eja bride service also contradicts a “compensatory model” (Collier and Rosaldo 1981) premised on ideas about the “loss” of the bride, in terms of her labor and fertility, within her kin group. Ese Eja provide an ethnographic exception in Amazonia by combining residential uxorilocality with patrilineality.20 Ese Eja households, as I have argued elsewhere, consist of groupings of related females that I refer to as “female residential clusters.” Households usually consist of a married couple, all their unmarried sons and daughters, and their married daughters and sons-in-law as well as maternal grandparents, grandchildren and adopted children.21 As such, Ese Eja bride service defies classical definitions of bride price in which women are viewed as

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 63

being “lost” to the kin group; rather it is brothers, sons, and sisters’ sons who become dispersed among other households. By rooting the household and household production in the continuity of female residential clusters, uxorilocality tends to undermine the social and political significance of patrilineality.22 Consequently, residence rather than descent is the key criterion for Ese Eja community organization, enhancing women’s social roles and status and defying orthodox definitions and theories of membership in both cognatic and unilineal descent groups. These arrangements complement Ese Eja philosophies and practices of marriage and also contribute more broadly to an understanding of how bride price is not always a tactic to secure men’s rights over women, as reported elsewhere (Borgerhoff Mulder 1995). As the following discussion on marriage demonstrates, good working relationships within households reflect stable marriages, a mutually beneficial and loose arrangement of cooperation, and bride price is one of many strategies used in the formulation of sustainable households that come together to form larger sociopolitical groupings. Whereas visible marriages actively include others within their marriage and include themselves in broader family and household relations, secret marriage isolate the individual from these broader networks.

Regeneration Over time, it is through reciprocity and exchange that a marriage becomes complete, participating in a set of relations much broader than the marriage itself and that articulate with the surrounding households and community. Productivity and regeneration—be it having children, preparing fields, or in ritual life—are what prove the viability of a marriage. Households are the basic units of economic production, subsistence, and exchange. Yet it is the mature marriage relationship that sustains and anchors the household in this economic system. Couples consider a marriage to be stable when they begin to have children and to work well together in their fields.23 Smooth collaboration is both a source of satisfaction and a sign of friendship and harmony. Both men and women comment that it is good when a spouse is an epeeji (friend), a relationship in which affection and humor are displayed and valued. The process of marriage often creates these close relationships rather than presupposes them.24 This is why the

64 · Daniela Peluso

manner in which a marriage is initiated speaks directly to a couple’s intentions about how they envision reciprocity and productivity in broader terms. Typically, the couple begins cooperation by establishing small gardens while they are still living and working with the wife’s parents. Men and women work together to maintain a general division of labor, but nearly all of their activities overlap, and there are few exclusively gendered responsibilities (Peluso 2003).25 Nonetheless these gendered aspects form part of the interdependency and complementarity that are central to marriage. Expanded upon, interdependency and complementarity are dynamics that inform relations. By the time a couple has several children, they are usually able to provide for them as well as contribute to the general extended-family household. When a couple decides to build their own house, it is usually as close as possible to the woman’s parents’ house, adding to the residential cluster. In these first years of marriage, men not only help their fathers-in-law but also receive help from them in setting up their own fields and perhaps building a house. Yet with ejakewawanaki these reciprocal relationships, if existent, are perceptibly more tenuous than in other early marriages. The tensions became clear to me when Pasha was deserted by Api for another woman after a year of marriage without her conceiving a child. Pasha’s mother explained: Pasha was so annoyed with her father because he let Api keep the plantain field that he made next to ours. She was upset because Api had planted with our cuttings. My husband kept telling her, “Let him be daughter, let him be—our family has a lot of plantain.” Yet my husband did keep the house that Api was building; he told him, “This is because you did not live well with my daughter.” Others say that it was Pasha who went to Api’s fields while he was away and macheted down all of his saplings. I did not see this. Maybe it was the children. (Fieldnotes 1997) Pasha and Api’s marriage had initiated as ejakewawanaki. Pasha’s father only felt comfortable addressing Api forcefully once the marriage with his daughter had ended. Pasha was bitter with Api. She did not conform to the usual behavior of indifference that is ideally desired. Her alleged sabotage of the fields that she and Api had planted together provides a glimpse at how significant subsistence production, and a couple’s investment in it, is to marriage. The value of the fruits of a couple’s labor cannot be overem-

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 65

phasized, particularly in a primarily non-cash hand-to-mouth economy. When food is scarce, people tend to become irritated with each other. A couple who can enjoy a watermelon on a long hot and humid day, and have enough plantains to be able to keep a raceme hanging to mature and sweeten for roasting, or who are able to accompany cherished game meat with something from their fields have much working in their favor. These are the pleasures that working congenially together bring. Yet amiability is a socially influenced virtue that reflects the behavior and input of more than just the couple. There were many times in Pasha and Api’s marriage when their families and friends could have intervened in ways that might have helped the marriage through its bumpy start, yet by initiating their marriage in secret, they precluded any such help. If, as Lévi-Strauss (1969, 67) remarked about the Nambikquara, exchange indeed “peacefully resolves wars, and wars are the result of unsuccessful transactions” then we can examine ejakewawanaki and hamatijawiaki as initiatives that correspond to or belie broader systems of exchange that may result in conviviality or conflict. Clearly relatedness created by marriage constitutes a basic idiom for social and economic relations. If secret marriages do not acquiesce to exchange and deny speech and power to the individuals with whom the partners are meant to exchange, then they stand in opposition to positive community relations and struggles. Individuals who marry in secret shield themselves from vocal criticisms and acts of interference from their immediate families that may provide a temporary and short-lived haven, but they simultaneously deny speech and action, thereby challenging the underlying mechanisms of power, and they thus reveal how marriage is embedded in such relations.

No Chance to See I have stressed how the viability of marriage is linked to acts of sharing, producing, and consuming jointly as well as interdependently, so that marriage regenerates life and livelihood. All along, the unending process of relatedness and connectedness of households is generated through goods that are circulated by the active work of men and women coupled together through marriage (Gow 1995) as well as through the continuous act of living and sharing together (Abu-Lughod 1986). This does not contradict the reality that there simultaneously exists a high degree of de-

66 · Daniela Peluso

sire and antagonism between the sexes. Although marriage is a principal means for creating and maintaining relatedness, sexual affairs may also catalyze new ways for individuals to define relatedness. Affairs are mostly secretive relations that only interfere with marriage when they last too long and divert resources. The following example of an incident whereby a sexual affair breaks up a marriage serves to illustrate further how unions between men and women articulate with broader communities and to show the threat that invisibility poses to the social group at large. I was traveling with some friends when Sokue went missing in town on the day of our departure. We waited a few hours in the port and sent the children out as scouts to look around for her. They searched in all the likely places: the market, the local federation, and the streets leading to and from the port. Everyone was annoyed as we waited under the sun for her to appear; after a few hours we left. They suspected that Besha, Sokue’s husband’s brother-in-law (HZH), who had arrived in town days before we had, had convinced her to stay with him. Although no one had seen them together that was the verdict. Late that night when we arrived in the community, Sokue’s husband approached me. He was visibly upset, and softly asked, “Is it true that she has a new husband?” I felt bad because I knew he was hoping there existed another likely explanation for her absence at the port that morning and that I would be able to confirm his friends and family were somehow playing a cruel joke on him. I could only say, “We waited and the children looked for her; she never came.” (Fieldnotes 2005) This event was extremely upsetting to Sokue’s community to the extent that they decided to expel Besha from the village. Having seen more than a few marriages end, I wondered why this one caused such unusually outspoken anger. He had married into the village, so to speak, something that was not unusual. My male neighbor explained that what had happened was not fair play: “This is not how it is done. Although these things may not be good, we have our way of doing them! He did it without anyone seeing, without the chance that her husband might persuade her to change her mind. When Shaka [his ex-wife] left me she went to live with her mother so everyone could see. It is only then that Manko went to take her to live with him! That is the way it is done.” Indeed, Besha had deprived the community of sight, speech, and power over one of its members.

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 67

Seeing, Speaking, and Power Marriages actively shape social networks whether they are initiated openly or secretly, despite the fact that the latter often fail or pose problems. The different narratives have shown how it is not “marriage” or “kinship” per se that is significant but the way they are contrived and narrated as people claim relationships among themselves through sight, speech, and action. There is a tendency to view arrangements like ejakewawanaki as quixotic, but clearly marriage can have many roles that appear to be in conflict with each other yet have their own compelling regularities and contribute toward the broader meanings of marriage as situated within community and Ese Eja social and power relations.26 It is precisely within these dynamics that the key consequences of ejakewawanaki, the lack of sight and speech, can be juxtaposed against hamatijawiaki, where the processual aspects of unions are visible and where individuals have voice. The stories of ejakewawanaki I have recounted continually result in frustrations over a disabling of speech and action based on intentional secrecy that leads toward invisibility; “I never saw them—now I cannot speak!” Speech and power are clearly related, yet in Amerindian societies Clastres (1987, 130) has poignantly pointed out how “if in societies with a State speech is power’s right, in societies without a State speech is power’s duty.” He also reminds us that that power does not lie in the content of speech but in the act of speaking, as such; the locus of power lies not within an individual but within the community. Critiques of social evolutionism have long dispelled the notion of non-state societies as “overdue embryos” (Clastres 1987, 91). Instead, stateless societies have more nuanced delineations of power distinguished by Clastres as pouvoir and puissance, the former reflecting finitude, structure, and fixity and the latter multiplicity and regeneration. These notions of power for Amerindians contradict Weberian ideas of power as necessitating coercion through chains of command—obedience that typically operationalize state societies. Just as Clastres’ (1987) Guayaki chief is unable to exert his personal will on people, individuals who marry secretly cannot shape society around their needs. As Glendhill (2000, 28) notes, commenting on Clastres’ analysis of power, power in stateless societies is regulated against “the egoism of a few.” The Guayaki chief refuses inequality, and to do so he cannot separate power from the group; if he desires it for himself then he is simply ignored and eventually abandoned (Clastres 1994).

68 · Daniela Peluso

Egoism is precisely the form of individualism that marks secret marriages. By excluding family and friends from balanced reciprocal relations, couples who initiate secret marriages are ultimately viewed as being self-seeking. Although Ese Eja notions of power suggest that individuals do not submit themselves to authority, it is the authority of one individual that must be resisted. Hence when an individual excludes the group it a coercive act, one in which the very makeup of power is at stake. Ese Eja notions of power, as I have described elsewhere (Peluso 2003) certainly reflect puissance, particularly when combined with the subtleties that Viveiros de Castro (1998a; 1999) nuances in his elaboration of multinatural perspectivism that further opens ideas of agency to a wider universe of exchange relations. The singular—the egoistic individual—pushes the group via exclusion through negative reciprocity, whereas the multiple, reminiscent of Durkheim’s moral individualism, pulls the group through exchange and authorized action. Ese Eja marriages and the power relations they entail, like the Guayaki hunter who must kill anew each day, reestablishing his worth, ideally reflect a capacity for continual regeneration; rather than “signs,” it is “values” that are exchanged (Clastres 1987). It is toward such a poetry of language and its underlying power that the troubles of secret marriages point us, toward dialectics and exchange that are legitimized by the visual and authorized through acts of speech. For such reasons, when marriages are invisible and secret, they allude to the anarchy within anarchy, where they may potentially unravel themselves through their own negation of reciprocity and exchange.

Notes Author’s note: I am profoundly indebted to the Ese Eja for their friendship, guidance, kindness, and hospitality. It is an honor to be welcomed in their communities. My doctoral fieldwork (1993–1996) was supported by grants from the Social Science Research Council, Fulbright IIE, the Wenner-Gren Foundation for Anthropological Research, and American Women in Science. I am also grateful to the CIRABO (Beni, Bolivia), and MUSEF (La Paz, Bolivia). Special thanks are due to Miguel Alexiades and the late Steven Rubenstein for their comments and to Stephen Beckerman and Paul Valentine for their work in editing this volume. 1. Ese Eja are lowland Amazonian peoples comprising about 2,500 individuals living in several communities along the rivers Beni, Madre de Dios, Orton, Sonene, and Tambopata, in the border regions of Bolivia and Peru. The Ese Eja language belongs to

Marriage Choices among Ese Eja of the Bolivian and Peruvian Amazon · 69

the Tacana language family, itself related to other Macro-Panoan languages of Western Amazonia. Most Ese Eja are swidden field agriculturalists and hunt, fish, and gather forest products for their own consumption and sale. 2. On the extremely rare occasion that a couple may marry in a regional town through the state administrator, they call it ejame’iñaki, and it refers to a state-recognized “legal” marriage. 3. Homosexual unions, although uncommon, also mimic the notion of marriage. 4. Numerous lowland South American ethnographers found that the Amazonian groups with whom they lived did not have any special rituals or ceremonies marking marriage (Shapiro 1984; Viveiros de Castro 1992; Kensinger 1984). 5. These terms are also referred to as we’e and we’pojiama (Kimura 1983; Shoemacher n.d.). Wapa means “non-relative,” “not of the same place,” and pojiama means “is not other” (negation). 6. As a patrimoiety, the Ese Eja are patrilineal, recognizing descent through male links. Dole (1979) in respect to the Amahuacas and Kensinger (1995) regarding the Cashinahua, Capanahua, Marubo, Matis, Matses, and Sharanahua, also suggest patrimoiety organization. 7. Descent is a contestable term in Amazonia (Murphy 1979; Overing 1981). However, the link to Icha and Kaka is more than just a link to “fathers,” it is the link to an essentialized identity derived from two distinct ancestral human-animals (Peluso 2003). 8. I use the term “birth father” to refer to the birth mother’s husband. Elsewhere, I discuss the complicated issue of paternity in regard to the question of moiety membership and partible paternity (Peluso 2003). 9. Women are referred to as eponakaka, epona’icha, ichapona or kakapona (icha/ kaka-woman) and men are referred to as kakayawe or ichayawe, (icha/kaka-man). 10. The influence of deja (Mestizo/Ribereño) Christian morality with its taboo on avuncular marriage has certainly inhibited Ese Eja preference for it. Ese Eja now disapprove of marriage between a toto (MB) and his boyase (ZD) by classifying it as memoo (incest), often implicated as a source of illness. Also, see S. F. Moore (1963) for a discussion on how oblique marriages suggest a younger counterpart. In the case of ZD marriage FZD would be the younger counterpart. Another possibility is that a man relinquishes his potential marriage to ZD so that his son may marry her instead (FZD) (Viveiros de Castro 1998b). 11. Sex with wapa pojiama is memoo (incest) and can be the cause of illness and skin pigmentation disorders. 12. This may serve as a measure against exaggeration and gossip. 13. These talks are spoken in the imperative. 14. School is often referred to as a valid reason for not wanting to get married regardless of a particular individual’s interest in educational studies per se. Schools are often social spaces for young people to be together and not have to work in their families’ fields. 15. In many other Amazonian societies, such as the Achuar, women are still severely penalized for resisting marriage arrangements (Kelekna 1981). 16. See Clastres (1989 [1974], 223) for a discussion on how conflict is an aspect of the changes that marriages inevitably bring about.

70 · Daniela Peluso

17. Among the Jivaroan Achuar, Kelekna (1981) describes how female discontent is expressed through delayed beer service and intentionally slow responses to the husband’s desires. Among the Cubeo, Goldman (1979) tells of women who purposely seduced their brothers-in-law in order to provoke a crisis in their newly formed arranged marriages. 18. In the past, Ese Eja bride capture resulting in marriage occurred when raiding rival territorial groups. Hugh-Jones (1979) and Århem (1981a) note that among Pira-parana and Makuna, respectively, bride capture transpires between distinct territorial groups and thus represents great affinal distance, generally not leading to stable marriages. 19. Ese Eja typically give their first two children to their parents to raise, the bride’s family and then the groom’s. 20. Very few Amazonian indigenous groups are patrilineal/uxorilocal. Two examples are the Mundurucu (Murphy and Murphy 1985) and the Cashinahua (Kensinger 1995). 21. Many other possibilities for household composition exist but this is the general and most common framework. 22. Chernela (1997) argues that other residence forms, such as virilocality, subjugate Amazonian women. Also, see Turner (1979, 181) for a discussion on uxorilocality as a means to exchange sons for daughters’ husbands, exploiting men’s control over women to exploit men. 23. Raising children is so much a vital aspect of marriage that a couple’s childlessness will almost certainly break up a union unless they can adopt. 24. Basso (1975; 1984) also describes Kalapalo marriage as resulting in the close ties that are valued but not a prerequisite for marriage. 25. Kimura (1981) observed the same in Portachuelo. The Ese Eja differ from many Amazonian groups such as the Mundurucu (Murphy and Murphy 1985), who maintain a more rigid sexual division of labor with few overlapping tasks. 26. See Rubenstein (1993) for a discussion of “chain marriages,” illustrating how marriage can accommodate conflicting needs.

PART II

•

4 Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage François-René Picon

In the Guajira Peninsula there was a marriage that was not quite right. I became aware of it only slowly, over the years, and was never able to pinpoint the specific problem that made it irregular. It was not incestuous, or in any way an unambiguously outlaw relationship, but there was something about it that caused some people to comment on it. Her family was richer than his, and the impropriety may have had something to do with bride price. The observation of the history of this marriage revealed a good deal about how Guajiro society dealt with a couple who declared themselves married but whose union lacked the full approval of the appropriate third parties. Actually, the initial title of this essay was “Why did I marry her?”—a question that has an obvious answer, although in terms that one seldom finds in ethnographic description.1 But to address the focus of this volume, one must give voice to the third parties (the affines) and ask, “Why did they marry,” there being implicit the modifying clause, “in spite of what we wanted?” One must also insist on the contractual aspect of marriage, which “creates an alliance between two groups” (Leach 1966, 20). Sometimes, nevertheless, a marriage does not create this alliance—or, more precisely, the groups do not ratify this alliance, do not give it their approval. Meanwhile, the man and the woman live together and raise the children who are born to them. The case recounted here is the history of a “socially incorrect” couple. It was not a matter of Shakespearian tragedy, the failure of an impossible love—rather the contrary. Nor was it that the couple disobeyed a sacro-

74 · François-René Picon

sanct prohibition of their society; rather it was simply the case that two people ignored the matrimonial plans of their respective families. From a highly ethnocentric and modernist perspective—that is, from the vantage point of our own society—there was nothing extraordinary or exotic in this state of affairs. Nevertheless, ethnographic monographs about traditional societies rarely discuss matrimonial alliances that take place “in spite of ” or “against.” What I argue here is that as time goes by, there is a tendency to paint over this sort of situation—that bit by bit the impropriety by which the family was created is forgotten and that this gambit is the way to reproduce the group without needing to obey its wishes (or those of its allied groups) completely. Once more, one must remember that this sort of thing is an ordinary occurrence, that disagreements are not—cannot be—eternal, and that time produces forgetfulness, or selective memory, which is the same thing. Or, taking the thought further, if time produces order and erases individual conduct that failed to respect collective interests, one may think that there exist specific mechanisms or occasions that permit this return to the normal. Stretching from southwest to northeast into the Caribbean Sea, the Guajira Peninsula lies mainly in northern Colombia. In the south, one fifth of the Guajira is in Venezuelan territory; here the large urban-industrial area of Maracaibo is within an hour’s drive. On the Colombian side there is no really important town. The fact that the peninsula is and always has been shared by two countries should be stressed, because it seems to have created a favorable situation for the development of pastoralism. It offers the possibility of commerce between the two nations and also with the Caribbean islands, while avoiding taxes and customs duties (a situation lasting from colonial times to the present). To this feature, one must add that with respect to a population of any size, the climate and the natural resources of the peninsula allow only a limited amount of agriculture, hunting, and gathering. The Guajira Peninsula is not vast. Its Indian territory covers some 15,000 km2, two thirds of which has recently been declared “protected Indian land.” It is a typical semi-arid tropical landscape with irregular rainfall decreasing from an average of 600 mm/year in the south to 200 mm/year in the north. The population density of its pastoral lands is estimated at about 5/km2.

Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage · 75

The Guajiro call themselves Wayù and speak an Arawakan language, one of the major linguistic families of South America.2 The society is divided into some 30 matrilineal, non-exogamic clans called eiruku, each identified with a wild animal—bird, insect, or mammal. They are of uneven status, wealth, power, and population. The smaller clans (e.g., Jirnu, Sapuana) are limited to one region and are somewhat exogamic in practice, although exogamy is not prescribed. The larger clans (e.g., Uriana, Epieyú, Ipuana) are found in multiple zones and are less exogamic; they are locally identified and a local subclan may have a different “totem” from the same eiruku of a different zone. The smaller the clan, the greater its solidarity in external conflicts. Big clans have less cohesion. In both large and small clans, internal conflicts may and do arise—less serious ones in smaller clans, caused by “cohabitation incidents,” more serious ones between larger locally defined subclans of the same eiruku, often produced by competition for power. In fact, these latter conflicts show that with respect to the matriclan the distinction between internal and external is vague and contextual, and that it is more appropriate to consider Guajiro social organization from the point of view of the matrilineage. Even if the eiruku is a symbolically meaningful unit, it appears to be only that. It is more an indigenous theory than a concept that can explain the way things are really done. Stories are told about famous occasions in which the corporate ideal of the eiruku was fully operative, but (after having spent substantial time in the Guajira) those stories appear to me to have acquired the accents of myth. If such stories are historically true, they are the exception, not the rule. In this brief and impoverished image of overall Wayù social organization, there is a strong analogy to Evans-Pritchard’s (1937) classic description of the Nuer (such at least is my impression), with segmentary lineage systems in general, and also a parallelism to the classical criticisms made of the segmentary concept. On the household level the Wayù domestic group is made up of sisters and their mother as permanent residents, with brothers and husbands, who have clearly different statuses and interests, as temporary visitors. Men are highly mobile, since they are polygynous. The sibling relation is the backbone of Guajiro society; authority in is the hands of elders. Such is the “official” model. On the ground, what is found in Guajiro rancherias (clusters of adobe houses roofed with the dried inner parts of the stem

76 · François-René Picon

of the torch cactus (Cereus sp.) is a social composition that tends toward this model but does not necessarily achieve it. One also sees steadfast husbands—who are unsuccessful brothers—and more permanent brothers—who may be uninterested husbands. Sometimes a sister leaves her natal home and goes to her husband’s rancheria, because of insufficient bridewealth—insufficient from her maternal family’s point of view—or from her paternal family’s point of view if she is the oldest daughter, since when she gets married the bridewealth goes to her father’s side. In short, one has the impression, first, that all possible combinations can be found; second, that all those situations are unstable, in the positive sense of the word; and third, that they are all the specific expression of a temporary state of social relations. This reversibility of social arrangements seems to be linked to matrilinearity. It is not as “dogmatic” as patrilinearity because the former is “only a type of membership criterion,” as Aberle (1974, 658) puts it. And this state of affairs may have something to do with the Guajiro ability to cope with white men in the neighborhood: what is often described as instability and vulnerability may in fact be flexibility and versatility. The Guajiro breed cattle, horses, mules, donkeys, sheep, and goats. Pigs and chickens are also present around the rancherias. Each clan has a branding iron to mark cattle, horses, and mules; sheep and goats are given earmarks indicating the owner’s clan; donkeys, say the Wayù, are not worth marking. The animals are individually owned but collectively herded. A man has some of his animals with the herd(s) of his wife or wives and others with his sisters’ herds. The explanation given for this division of a man’s livestock is twofold: diversification of risk and the need to be present through one’s animals. The social network has to be activated and secured constantly since alliances are fragile and conflicts are common. There are considerable differences in herd size and composition, from a few goats to dozens of cows and horses. However, if Guajiro pastoralism is obvious, nomadism (i.e., the importance and frequency of movement of those herds) is less evident. Pastoral movements may amount to less than 100 km/year. But nomadic pastoralism is not only a matter of distance of travel; it is also seen in its relation to other economic activities. The diversification of resources and the movements of people related to these activities are high.

Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage · 77

The cash needed for modern Guajiro life is provided by wage labor in the Maracaibo area, in the state-owned salt pans, or in gypsum mines. The Guajiro also collect—on their own clan lands—the fruit of the dividivi tree (Caesalpinia coriara), used for tanning hides, and sell them to a Colombian factory. The Guajiro Peninsula is known as a region of contraband smuggling, but the Indians (contrary to the conventional wisdom) have only a secondary role there, for these activities are not in indigenous hands. Fishing and agriculture are the subsistence resources. Fishing is synonymous with poverty, even though fish are abundant in these Caribbean waters, and agriculture is considered as very marginal. The fundamental difference between cash and subsistence activities is that cash can be converted into livestock, while the products of fishing and agriculture cannot. They are for consumption, not conversion. Social life implies domestic animals. Power is related to wealth and animals are wealth. A man can be well above the level of physical existence and at the same time below the level of social existence if he has no animals. Livestock—particularly cows and horses—are the raw materials of social life: debts, alliances, and funerals are thought of and expressed in terms of animals and, significantly, eiruku is the word for clan and also for the meat of animals.3 Cash can be used to buy animals, and also to buy food, clothing, and whatever else is needed, in nearby town markets, never further than a half day’s journey by truck or donkey. Animals can also be sold, but there is an obvious reluctance to do so among poorer herders. The rich, on the other hand, sell animals and thus provide meat to the Colombian and Venezuelan markets. The parallels here with well-known mechanisms and characteristic features of other pastoral societies are evident.

Exception or Normalcy? This brief sketch of Wayù society allows a return to the question posed in my title: “Why did they marry?” What follow are the results of isolated observations and conversations. My primary research did not focus on social organization or kinship but on economics and particularly on the circulation of goods. Nevertheless, a situation—the price of the bride—in which goods did not circulate as they should have, or did so only to a very limited degree, should have de-

78 · François-René Picon

manded my attention. But given the academic training of the time—what does one do with exceptions?—I had neither the interest nor the necessary tools, nor was I willing to take the risk of confronting these questions.

From Where One Looks In the first pages of The Cubeo, Irving Goldman (1963, 23) notes how much the observations one makes depend on where one has managed to locate oneself and that these observations lead to generalizations. The choice, he writes, “was partly fortuitous.” That is what happened with the family with whom I began to share daily life—an atypical family, of lower middle economic and social level (which is important in a society as hierarchical as that of the Wayù), a family that had not complied with what is usually done. But the marriage existed (I do not know if one can say they had “gotten married”) because sons and daughter had been born. On the one hand, other people talked about this pair; on the other hand, they appeared to live normally. I do not know if methodologically it is a good idea to observe and describe from the point of view of a “marginal” unit (but what is the center— what is normal?). One can defend this vantage point with the argument that the margin has a certain grade of “exteriority,” and the information obtained is “in contrast.” In reality, as we all know, it is not methodology but rather that between the observers and the observed there come into existence bit by bit relations of proximity (or friendship) that are perhaps the most fundamental aspect of the anthropological method. One more reflection: one observes society from a point located in social space, but also, to come to know certain aspects, one must be situated in duration—not the longue durée of historians but that which permits one to see the evolution of relations between groups and individuals. It is true that one can achieve this temporal dimension through life histories; but to observe at distinct points in time, in different moments, seems to me to provide a supplementary guarantee of objectivity. The “moments” used here were three: between 1972 and 1974, 1985, and 1995.

The Model Because the marriage examined here is atypical, it is necessary to specify what can the called the “model” of marriage, keeping in mind that the

Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage · 79

anthropological literature on the Guajiro and the Wayù is abundant. This summary is divided into two parts: (1) how Wayù society is perceived and presented in the standard anthropological literature—that is, the “ethnographic present” (a notion that is difficult to defend these days); and (2) some points that I believe characterize the current functioning of Wayù society—that is, the context in which the “atypical couple” is located—and which may offer some keys to understanding this case. On the first matter, there are more than 360,000 Wayù. As already noted, there exist between 30 and 40 matriclans, some so large that they are divided into smaller and more “functional” units. Filiation is matrilineal. “Flesh” is transmitted by the mother (apushi) and “blood” by the father (oupayu). The important figure is the maternal uncle (MB, araura). Post-marital residence is matrilocal (the wife “stays home.”) For a marriage to be recognized, the family of the groom has to deliver goods on two or three occasions to the family of the bride, goods that are traditionally domestic animals and jewels (necklaces, rings, etc.). This bride price varies considerably according to the social and economic status of the parties involved. This is, that the groom pays in order that the group of the bride may reproduce itself, pays for not producing descendents for his own group; another example of the classical “matrilineal puzzle” documented in African societies. That may be—but in the Wayù case (as also already noted), this investment is recovered when the first daughter marries. Ideally, her price corresponds to the initial payment augmented by what the Wayù call, metaphorically and pedagogically, the “growth of the herd.” Animals and goods also circulate on other occasions, to resolve conflicts or when an individual dies (Picon 1996). In this last case, the theory is that a part of the herd of the deceased has to disappear (the eiruku loses a member and the animals that correspond to him), and some of his animals are given to those who participate in the funeral. The remainder of the animals are part of the inheritance of the nephews of the deceased. On the second matter, now some traits to begin to make out the dynamics of Wayù society. It is evident that it was radically transformed by the Spanish conquest—to a large degree, indeed, was born in that conquest. (What used to be called acculturation or deculturation is now called ethnogenesis!) And it is true that in order to understand how Wayù society functions, it behooves us to look at “classic” pastoral societies. In other words, there are only so many ways to raise cattle and, above all, to articulate the economic with the social.

80 · François-René Picon

Because they are very numerous and their territory is meager, many Wayù live outside the peninsula—for example in the famous Guajiro neighborhood Siruma in the city of Maracaibo—but they do not therefore cut relations with their relatives in the indigenous zone. It has been claimed that in Maracaibo, because residence there is neolocal and because of pressure from the larger society, matrilineal filiation has been weakened. I do not know, but I doubt it. In evidence, perhaps indirect: what I have called the “new clans”—when an alijuna (i.e., a non-Indian) marries a Wayù, the children function in both systems, and after two or three generations the descendants of such a union come to be considered as a legitimate clan. A child of such a marriage may be a doctor or lawyer in Maracaibo or Caracas, but his or her family of reference—in the indigenous context or when it is convenient in the larger society—will be the maternal line, and through it he or she will identify as Wayù. It must be remembered that brothers and sisters are those who share maternal flesh as well as paternal blood (be it “external” or not), and this sibling bond is the fundamental axis of this society from the point of view of kinship.

The Concrete Situation: Three Moments (1) In 1972–1974, the couple was living near the Colombian city of Uribia. He made a living buying and selling animals and she from domestic work for wealthier families. They had two daughters and two sons and a fifth child was born at the end of 1974. Twenty minutes’ walk to the east was the household of the husband’s three sisters; most of the time two of their husbands were present, and the third husband was rarely there. Proximity between brother(s) and sister(s) was not in any way irregular, and in this aspect the couple complied with the model of marriage. Farther north (about 45 minutes by car) lived the wife’s maternal uncle along with other members of her family. The only visit the couple made to this home had to be prepared and was obviously the product of negotiations. But here was the third party from the point of view of the wife. The people with whom the couple had permanent relations were the husband’s sisters and some relatives, in the widest sense of the word, for whom he and she performed services. These “honorary relatives” were

Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage · 81

mestizo entrepreneurs and rich Wayù. One cannot say the couple was isolated but rather that, with reference to the Wayù model of marriage, this case presented at the same time both normal and abnormal features—it was a bit marginal, but not marginalized, or at least if it were marginalized, that was not evident. (2) In 1985, during a few days verifying data, the husband of this couple, my informant, said to me, “Let’s go to my uncle’s.” “Which uncle?” I asked. “Taraura,” he answered—that is, his maternal uncle. At this moment, there appeared the missing figure, the hidden piece of the puzzle, the representative of his group—that is, the third party, from the husband’s point of view. I knew of his existence but also was aware of the absence of relations between MB and ZS and for that reason had not insisted on meeting him, accepting this absence in my data, to a certain extent very peculiar in a matrilineal society. So we went—15 minutes’ walking, toward the west, on the opposite side of where his sisters lived (and therefore away from his mother), another peculiar element in this morphology that expressed interpersonal relations—under the pretext of bringing a table and a bench. It was a pretext because, it turned out, the visit was really about a part of the uncle’s herd that he wanted to entrust to his nephew. The encounter was brief and lacked warmth. At the same time I had the impression that my informant wanted to demonstrate to me the normalcy of his kin relations, and that reconciliation—or more exactly the reopening of relations—does not necessarily require an expression of friendship or closeness. On the other hand, at this same time, the older daughters of this couple had reached marriageable age. In other words, the stages in the life cycle when goods were going to be circulated and interchanged were coming up. There are two such events: the inheritance of the nephew from the (maternal) uncle; and the bride price that comes to the father’s group when his oldest daughter marries. These moments are critical and (in a case such as this) it had to be decided if the family was going to continue being marginal or if, 15 years after the irregular marriage, everything was going to return to normal. (3) Ten years later, the couple continued living in Uribia but in another house. Daughters and sons had married and had their own children. During our very informal conversations, the wife insisted on the importance

82 · François-René Picon

and permanence of the maternal line—the children of her daughters were Epieyú, of her matrilineage—without my having to ask her if there were any sense of danger to matrilinearity. The husband’s maternal uncle had died, and his goods had been converted into a still and my informant into a producer and seller of rum (chirinchi)—which is drunk during funerals (among other occasions) and which may form part of debt compensation or bride price. The sons and daughters had left, but had not gone far, and their horizontal line, from which descent and filiation were to be assigned, occupied the leading edge of the social scene. The socially incorrect couple seemed to me to be relegated to second place, and as far as I know, people did not speak of, or had forgotten, the irregularity of their marriage, as if there were more important and more urgent matters to be dealt with.

Obscure Parts of an Anecdote; or, Representivity, Anecdote, Questioning, and Interpretation In considering and recording this history, its contents seem to me to be very anecdotal, above all because at the end of the day, things worked out. Now, I think one must ask two related questions: What is the level of representivity of this example? And what is the status of something that is, after all, an anecdote? Statistical ambition is extremely difficult if not impossible to satisfy in a society as numerous and hierarchical as that of the Wayù. It is true that a single case does not represent anything other than itself, but it may have the status of exemplary history, as can a certain type of anecdote—one that carries an unanticipated message. It can be exemplary also because it seems that the problem (to overstate the case) has been resolved in the Wayù way, that is, beyond tautology, that certain mechanisms particular to Wayù society have functioned. Even if one can draw the outline of this problem of representivity, there remain many points to clarify, and there is a great deal of information I do not have. But to know the questions is to know something about the answers. For example, I do not know what was paid—or if anything was paid—for this couple to “marry.” It was said that there was no payment, but I do not know if this statement is credible (when the rule of clan exogamy had been respected—Ipuana with Epieyú). Neither do I know if the third parties wanted him or her to marry someone else, or if the problem was economic.

Why Did They Marry? A Very Short Tale of a Lasting Wayù (Guajiro) Marriage · 83

Nor do I know the bride price received for the oldest daughter, nor what was paid in bride price by the three sons, nor who paid their bride prices. And I do not know if there exist incorrect marriages that continue in that status—that is, without the return to normal seen in this case. And one can continue to add unanswered questions. It is true that one cannot pretend to an exhaustive ethnography, and the more one knows, the more questions arise. But I do not believe, beyond the evident defects of this anecdote, that this situation makes invalid what could be called the moral of the story. So we come to this: Either we leave things as they are—the story of a couple who did not comply with the matrimonial strategies of their third parties—or we try to understand why, how, and for how long they were able to live, socially, even while being marginal (but not, I insist, totally marginalized). It could be time itself that hides (or fabricates) memory— but this process is universal, and consequently this type of explanation weakens the content of the story. Or it could be that there exist key moments and specific mechanisms that lead to—or compel—the re-equilibration of things. In other words, perhaps overambitious words, there might be a structural principle that comes into play in those moments, a principle that might be called “flexibility” or, following Leach (1970, 7) “equilibrium”—as opposed to “stability.” If two people, this couple, can live for a time—he without a maternal uncle (and without being a nephew), and she far from her maternal family both geographically and socially—there comes a time when there are goods to transmit, children to marry, and two groups (the groups of the third parties as well as themselves) that have to go on reproducing themselves. Then the rules must be honored because when the children marry, there are economic as well as social decisions to be made. And it is those children—the horizontal line, this “horizontal backbone” that is constructed as time passes, those who share flesh from the mother (apushi) and blood from the father (oupayu)—who at the symbolic and the material level represent the totality of the family as product and potential, who seem to me to make up this unity and this key moment in which the social order is reorganized to recover from the threat it felt for a while from two individuals who decided to live in their own way, decided to live what was in the end a love story.

84 · François-René Picon

Notes 1. An obvious reference to Luc de Heusch’s Pourquoi l’épouser? In this context, a few remarks: In addition to being, to my mind, an important contribution to the problematics of alliance, it is one of those rare anthropological texts the title of which is in the form of a question. When one knows that nearly all the titles of anthropology books are in the affirmative mode, and that questioning is the basis of ethnographic work, one can only conclude that the ethnographic questions have been resolved on the anthropological level. In addition, while the title in French is deliberately ambivalent—the gender of Ego is not specified—the English title Why Marry Her? carries a “male bias” that is certainly incorrect nowadays. Finally, to the degree to which this question implies the existence of a rule—and consequently the possibility of disobeying it—I think that in a situation such as the one I observed and commented on here, the question should rather be phrased as “Why marry him?” to do justice to the economic difference between the two families of origin, the two “third parties.” 2. The Guajira Peninsula is far from being an anthropological terra incognita. Methodical description started in the mid-18th century since colonial authorities urgently needed information to control Caribbean coasts open to foreign trade and influence. In the second half of the 19th century, well after independence was won, ethnographers, geographers, and travelers made important observations. The data collecting continued, and Armstrong and Métraux (1948) and Steward and Faron (1959) synthesized the information, including, however, some errors and misunderstandings. Later, systematic studies were carried out on magic, social organization, kinship, and myth—not to mention geography, pedology, etc. Around 1970 more thorough research began: on acculturation (Watson 1972), myth and shamanism (Perrin 1976), social organization (Goulet 1981; Saler 1988), and ethnohistory and economics (Picon 1983). More recently, among other investigations, it is important to cite those of A. Mancuso (2008) on gender and filiation. All fields have been covered, but more work is still needed inasmuch as the Wayú population is rapidly expanding and claiming and defending its identity. 3. This synonymy is one of two cases of native words present in the entire lexicon of breeding, the other being the word for horse (ama), which comes from the word irama, one of the two species of deer once abundant in the Guajira. The association is explained by the speed of the two animals.

5 Beyond the Norms Marriage and Incest among the Ye’kwana Nalúa Rosa Silva Monterrey

In this chapter I analyze the Ye’kwana marriage system focusing on the norms that lead to qualify marriages as permitted or incestuous (kaiña), defined in the emic sense. I present data that indicate the frequency of incest and discuss the consequences of breaking the rules. For the Ye’kwana, acceptable marriages are at the same time the preferred ones. From the emic point of view, all the incorrect or prohibited marriages—that is, the incestuous ones—fall within the category of kaiña. In Ye’kwana there is no term that signifies incest in the etic sense. Rather, all incorrect or prohibited marriages are considered kaiña.1

The Ye’kwana The Ye’kwana are known by a number of terms, of which Makiritare, Mayongong, Pawana, and Mayongomo stand out. They are Carib speakers, of whom about 50% also speak Spanish (Silva Monterrey 2010, 26). They number 4,472 people, distributed primarily in tropical, humid forest, in which are found pockets of savannah. This territory is traversed by large rivers, including the Cunucunuma, Ventuari, Padamo, Caura, Cuntinamo, and Orinoco, located not only in the southeast of Venezuela, in Amazon and Bolivar States, but also along Brazil’s northern frontier in Uraricuera basin region, in an area of some 80,000 km2. The Ye’kwana are located in adjacent basins interconnected by paths that traverse the jungle from one settlement to another. Such journeys take several days. For Ye’kwana, also known as canoe-people because

86 · Nalúa Rosa Silva Monterrey

of their skills in navigation, rivers constitute the most important travel routes. Every basin or sub-basin functions as a cultural block. The inhabitants of the different settlements of a basin relate in limited ways and engage in matrimonial exchanges. There exists preference of choice of partners, not only within the same basin but also between settlements of different basins. For example, such relations exist between the people of the Ventuari and some settlements of the Erebato (Silva Monterrey 2010, 236–43). The Ye’kwana Kinship System The Ye’kwana kinship system has been studied by several authors. ArveloJiménez (1974) gave us the first approximation, Heinen (1983–1984) meticulously described the terminology, Coppens (1981) gathered genealogical information of the Caura River basin, and I studied the terminology and the marriage system (Silva Monterrey 2007 and 2010). From a structuralist point of view, the Ye’kwana kinship system is similar to that of other ethnic groups of the Guyana region (Henley 1996, 33; Rivière 1984, 4; Shapiro 1984, 2; Silva Monterrey 2007, 217–21). It is Dravidian and is typical according to Henley’s definition of such a kinship system. Compared to other Dravidian systems of the region, what is particular to Ye’kwana kinship is the conjunction of (1) an uncommonly long and detailed genealogical memory for a Carib people; and (2) a particular language expressed in terms and establishment of certain types of marriages, formed from ordered, relational norms. As is to be expected in a Dravidian system, opposite sex cross-cousins are considered potential spouses, whereas parallel cousins are classified as siblings. In current usage the reciprocal term known for male and female cross-cousin, yetanadu, is imbued with shame and is almost a swear word. According to Heinen (1983–1984, 277), who in turn is following Arvelo, this term is arguably not in fact a kinship term—a position with which I agree. Indeed, Ye’kwana prefer to use the term that has no reference to marriage, because of the social and affective load that comes with the public recognition of a potential spouse. Nevertheless, I have been able to collect the female and male cross-cousin terms of address, which are wodi’chü and yamwadü respectively (Silva Monterrey 2007, 224–25).2 These are the correct kinship terms for cross-cousins. Wodi’chü is usually used by a male ego before his marriage and in an impolite manner after

Beyond the Norms: Marriage and Incest among the Ye’kwana · 87

his marriage.3 This term is in relatively common use, although it always carries a sexual load (Silva Monterrey 2010). The term yamwadü comes from the word yamwa, which signifies “boy.” Yamwadü unequivocally corresponds to cross-cousin for a female ego; women currently employ this term before a marriage. However, the term is not used after the marriage, not even when addressing the groom’s brothers. That is, once ego marries, ego calls his or her opposite sex crosscousins elder or younger “brother” or “sister”; for married ego the terms equivalent to “brother-in-law” and “sister-in-law” do not exist for opposite sex cross-cousins. Once the marriage has occurred, the Ye’kwana neutralize their relationship with their spouses, assimilating them into the categories of “brother” and “sister.” This verbal modesty contrasts with the high frequency of marriages between people who stand in a relationship to each other that, according to the kinship terminology, prohibits their union.

The Marriage System From the point of view of the terminology, Ye’kwana prescriptive marriages are focused on the achievement of preferred marriages between bilateral cross-cousins with uxorilocal post-marital residence. Also permitted are marriages between individuals classified as grandfather and granddaughter if they are approximately the same age. In addition, polygamous and polyandrous marriages are accepted, as are marriages between relatives with encompassing ties, such as groups of brothers with other families’ groups of sisters. In spite of the fact that these are the verbalized norms, recognized by everyone, there also occur incorrect or prohibited marriages, known as kaiña, between people of the same generation or adjacent ones. Kaiña encompasses all the marriages outside prescriptive marriages between cross-cousins or grandfather and granddaughter. Kaiña signifies “bad” or “incorrect.” It is used in a metaphorical sense; when people enter such a relationship, they are, as it were, burning themselves. Indeed, when someone is accidentally burned or scalded, then in a short ritual designed to alleviate the injury, another person shouts kaiña, kaiña, kaiña. It is not unusual, when a couple enter a kaiña marriage, to hear the comment that they are going to hell.

88 · Nalúa Rosa Silva Monterrey

The trans-generational marriages, grandfather-granddaughter—consanguines and classificatory consanguines—(Arvelo-Jiménez 1974,136) open the door to new marital possibilities; such marriages permit young people, classified in ascending generations, to be able to marry others of their own age who otherwise would be prohibited (Mansutti Rodríguez and Briceño Fustec 1993, 64). The Cultural Practices Associated with the Couple’s Selection The ideal Ye’kwana woman is modest, family oriented, and cares for her hearth. It is hoped that young men are not lazy or promiscuous. This belief in correct sexual behavior does not impede some sexual liberty among adolescents. The young of both sexes have frequent sexual relations before marriage, although this should be done discreetly. The parents consider such sexual experience less than ideal; they say the young should not engage in “sexual games.” Rather, once the couple like each other and after gaining the consent of his own father, the young man should go and live with the young woman’s family, where he should help his father-in-law. Usually the mothers talk together before the marriage, although this exchange of opinions is not something ritualized. The father should speak to the intended groom and say that he would dislike it if the young man were to “play” with his daughter. Among the Ye’kwana there is no marriage ritual. The marriage consists of living together, which becomes official when the young woman publicly goes to the community house (öttö), where the young men sleep, unties the young man’s hammock, and takes it to her house, after which the young woman attends to the family hearth. Although this act could be considered a sort of ritual, a marriage is only recognized as definite when the young woman becomes pregnant and gives birth to her first child, at which time the couple practice public avoidance (Mansutti Rodríguez and Silva Monterrey 2002, 203). According to the elders, in the past, young men were very respectful and stayed with their affinal family. Nowadays, old people also say young men do not take marriage seriously. They stay with a young woman for a time and then abandon her, much to the disgust of her parents, who can do nothing. The marriages of the present generation of mature people seem to be more stable than those of the young. In the majority of cases, today’s youths marry within the norm, in accordance with the Dravid-

Beyond the Norms: Marriage and Incest among the Ye’kwana · 89

ian terminology. However, there appears to be greater flexibility in the post-marital residence pattern, which is becoming progressively more virilocal. Factors That Influence the Selection of the Couple Until approximately the 1960s it was common for parents to force their children to enter a determined type of marriage. I know various couples who point out that in spite of their lack of choice, they have remained married to their partner all their lives; they say affection comes later. Some of these couples are now extremely old. Nowadays, physical attraction is an important factor in choosing a spouse. Young people fall in love and start to live together. The parents try to maintain some control over them, but the situation is different from how it used to be. Their control over girls and young women has weakened; many of them go to study in the cities, and it appears that the parents have less influence over their sons’ and daughters’ choice of marriage partner. The guide that defines relationships continues to be the kinship terminology, which provides the key to who are, and who are not, potential marriage partners; it applies equally to those who come from the same settlement as to those who live in different settlements. Young men and women of different communities meet during intercommunity activities. The frequent journeys to the cities to study increase the size of the marriage pool, and the chance of meeting a partner of the correct marriage category, although not necessarily of the same village or region. I have identified other additional factors that influence—or have influenced—the selection of a partner, besides that of physical attraction: age, ethnic origin, ethno-historical background, and the sociohistorical situation. Age Although all the factors noted play their part in selecting a spouse, the ideal is that marriage partners are of the same generation, or if they are of different generations, that they are of approximately the same age. Nevertheless, I have recorded young women’s accounts that describe how their parents have wanted them to marry their actual grandfathers or a classificatory grandfather because these old men wanted them. They also

90 · Nalúa Rosa Silva Monterrey

referred to the conflict these forced marriages have caused; it can reach such a pitch that the young women attempt to escape the community. This conflict also creates bad feeling in the community, given that although they know this kind of marriage is permitted, people consider it far from ideal. However, if the couple are of about the same age, although according to the terminology they are grandfather and grandchild, their marriage is seen as perfectly normal and acceptable. Region of Origin The ideal is to marry someone who is as close as possible, both spatially and in kinship terms. Therefore the best possible option is to marry someone from the same settlement who is an actual cross-cousin. This, however, does not exclude marriages between individuals coming from different settlements. According to our data there have been periods during which marriages between communities have been rare, and at other times these have been more frequent. The Ye’kwana say, for example, that during the rubber boom at the beginning of the 20th century (1896–1921 approximately; Iribertegui 1987), as their population was small, when men married into other communities, it resulted in the diminution and even the disappearance of their own community. The residence pattern demanded that the husband stay in his wife’s community. That is why, the Ye’kwana say, it is preferable to marry into one’s own local group and be with one’s own kind. Ethnic Origin Usually the Ye’kwana marry within their own group, whereas inter-ethnic marriages are scarce, although they do occur. The Dravidian kinship system is widespread throughout the Guyana region. The sharing of a similar kinship structure, together with its flexibility, facilitates marriages between different indigenous groups; this commonality assists the assimilation of individuals of different ethnic groups; they share mutually intelligible cultural references. Although at present inter-ethnic marriages are scarce, during the 19th century they were reportedly more common. Such “mixed” marriages are vividly remembered, yet individuals’ historical ethnic origin was dealt with discreetly and, in practice, kept secret (Silva Monterrey 2010, 133– 39). This discretion is hardly surprising given that on occasion, members

Beyond the Norms: Marriage and Incest among the Ye’kwana · 91

of one ethnic group were enslaved by another, or that members of one ethnic group were in open warfare with another. The current extant generations do not take ethnic origin into account, nor do they consider it an impediment in the selection of a partner.4

The Marriage Database The genealogies, marriages, and kinship data key to this chapter come from material that I collected between 1986 and 2000 and that collected by Dr. Daniel de Barandiarán in the 1950s.5 The data were drawn from the Caura River basin. The material was gathered from interviews with men and women of different ages, either in groups or individually. The genealogies provided by Barandiarán and the Foucault Brotherhood contain information concerning each person’s sex, name, ethnohistorical origin, and location of residence. The material I collected also includes, where possible, the age at death and place of death. Although the data are fairly consistent, that provided by Dr. Barandiarán comes primarily from the village of Kanarakuni, located at the headwaters of the Caura. In spite of the large quantity of information gathered, some material was unobtainable. For example, a person’s name would be forgotten, but from whom he or she is descended would be remembered; or it would be forgotten exactly how many children a woman had. In short, some data are missing, but not sufficient to impede the analysis. We took the oldest genealogies together with the present-day ones and established the links among the inhabitancies of the same village and among different villages. The large volume of data that I was handling required specialist support, which was provided by Dr. Marion Selz.6 The combination of my material with that collected by Barandiarán has facilitated the reconstruction of very long genealogies, given that a Ye’kwana alone can only normally remember five ascending generations. For some individuals the deepest genealogy reconstructed reached back to a maximum of 15 generations, although the average was approximately seven (Silva Monterrey 2007, 134, 217) and encompasses a period from the beginning of 19th century to the beginning of the 21st century.7 In this chapter we analyze 200 marriages, drawn from a database of 2,366 people, among whom 820 marriages are registered. The data refer to people alive and dead. These 200 data points were selected because

92 · Nalúa Rosa Silva Monterrey

they have the complete set of information on each spouse’s four grandparents. This material is important, as it allows us to calculate exactly the number of actual (that is, biologically related) cross-cousin marriages that occurred. As far as the material allows, I have drawn from the database the following information on each individual: Sex Place of birth River basin of birth The analysis takes into account the names that appear in the data. Names that appear in the genealogies only in Ye’kwana correspond to a period before the 1950s, as the general use of Spanish first names in the data corresponds to the period of missionization (1958–1970).8 This finding makes it possible to classify the data into two periods: the “Old Period” before the 1950s, and the “Modern,” from 1950 to 2000.9 The results of the analysis of the marriage data are summarized in table 5.2. We added the letter R before the name of the location to differentiate the regions from the villages.

Table 5.1. Data classified into two periods: the “Old Period” before the 1950s, and the “Modern,” from 1950 to 2000 Marriages in which we have data on each spouse’s four actual (biological) and named grandparents. Marriages in which we have data on each spouse’s four named grandparents, but I have been unable to establish an actual, biological kinship connection between the couple.

200

Marriages in which we have data on the 4 named grandparents of each spouse, and the couple are the descendants of at least one actual shared grandparent.

143

57

Table 5.2. Marriage types classified by kin and affinal links Old (before 1950) Modern Region or (1950–2000) settlementa

Types of relationships Equivalent in (based on biological the Dravidian terminology model of kinship)

No.

M=S

M=S

0

W=HS

M=S

0

MZ=ZS

M=S

1

1 old

R-Ven-Ven

F=D

F=D

1

1 modern

1 Kan-Kan

MS=MD (Siblings of the same mother; father unknown) FS=FD (Siblings of the same father; mother unknown) MZS=MZD (with a common grandmother)

B=Z

2

1 modern 1 old

2 Fede-Fede

B=Z

2

2 modern

1 Kan-Kan 1 Cha-Cha

Considered kaiña by the Ye’kwana

B and Z on mother’s side

7

5 old 2 modern

Considered kaiña by the Ye’kwana

FBS=FBD

B=Z

7

5 old 2 modern

MZS=MZD

B=Z

7

5 old 2 modern

FBS=FBD (with a B=Z common grandfather)

12

7 old 5 modern

3 R-Ven-Ven 2 R-Cau-Ven 1 Cha-Cha 1 Fede-Fede 2 R-Ven-Ven 2 R-Cun-Cun 3 R-Par-Par 2 R-Cau-Ven 3 R-Ven-Ven 1 Cha-Cha 1 Kan-Kan 1 Cha-Cha 2 R-Cau-Cau 2 R-Cun-Cun 3 R-Ven-Ven 3 R-Par-Par R-Ven-Ven

Grandfather= Granddaughter

Grandfather= Granddaughter

1

1 old

FZ=BD

Aunt=Nephew

2

1 old 1 modern

Notes Considered kaiña by the Ye’kwana Considered kaiña by the Ye’-kwana Considered kaiña by the Ye’kwana Considered kaiña by the Ye’kwana Considered kaiña by the Ye’kwana

Considered kaiña by the Ye’kwana Considered kaiña by the Ye’kwana Considered kaiña by the Ye’kwana

Considered normal by the Ye’kwana 1 Kan-Kan Considered kaiña 1 R-Cun-Cun by the Ye’kwana (continued)

Table 5.2—Continued Old (before 1950) Modern Region or (1950–2000) settlementa

Types of relationships Equivalent in (based on biological the Dravidian terminology model of kinship)

No.

MB=ZD

2

1 old 1 modern

2 R-Ven-Ven Considered kaiña by the Ye’kwana

30

16 old 14 modern

Cross-cousins Marriage between cross-cousins (with a common grandfather)

38

18 old 20 modern

Cross-cousins

31

20 old 11 modern

7 Jy-Jy 2 Cha-Cha 1 Bn-Bn 2 Ana-Ana 2 Ada-Ada 10 Jy-Jy 14 R-Ven-Ven 1 Kan-Kan 2 Cha-Cha 1 Bn 1 Ana-Ana 3 Ada-Ada 1 Says-Says 13 R-Ven-Ven 3 Cha-Cha 5 Jy-Jy 1 Bn-Bn 1 Ada-Ada 1 Ana-Ana 1 Kan-Kan

Marriage between cross-cousins (same grandfather and grandmother)

Marriages between cross-cousins (with a common grandmother)

aR

Uncle=Niece or Father-inlaw=Daughterin-law Cross-cousins

with the location differentiates regions from villages. Abbreviations: Ada = Adawañ R-Cau = Caura Ana = Anadiqueña R-Cun = Cunucunuma Bn = Boca de Nichare R-Par = Parime Cha = Chajudaña R-Ven = Ventuari Fede = Fedekuniña Jy = Jüwütünña Kan = Kanarakuni Saduñ = Sadduña Says = Sayu Shodiña Wasai = Wasaiña

Notes

Considered ideal by the Ye’kwana

Considered ideal by the Ye’kwana

Considered ideal by the Ye’kwana

Beyond the Norms: Marriage and Incest among the Ye’kwana · 95

The Number of Prescribed and Incestuous (Kaiña) Marriages Of the 143 marriages with direct biological connections there are 99 between cross-cousins and one between grandfather and granddaughter; that is, approximately 73% of the marriages are in accordance with the ideal as expressed in the terminology; they correspond to the marriage prescription. As can be seen in the third table, the proportion of marriages that conform to the norms have remained relatively stable at 67% for the Old Period and 73% for the Modern Period. Table 5.3. The number of prescribed and incestuous (kaiña) marriages Types of marriage

No.

Old

Modern

Kaiña In accordance with the ideal Total

43 100 143

26 (32%) 55 (67%) 81

17 (27%) 45 (73%) 62

Marriage Exchanges between Villages Marriages between villages occurred throughout history, but during the rubber boom, at the beginning of the 20th century, they increased. In the case of Old Period marriages, the settlement names of the bride and groom are not recorded. The couple are only located by their region or regions. In fact, the majority of marriages took place between settlements of the same river basin. Turning to marriages between river basins, they were between matrilateral parallel cousins; that is, between ZS and ZD. Analyzing marriage between villages, as well as between river basins, there are cases of several brothers of one family selecting their partners from the sisters of another family of another village or basin. Such marriage exchanges between close family groups reinforced the links between them. Table 5.4. Kaiña marriages by village and river basin for approximately the whole 20th century

No.

Kaiña marriages within the same settlement

Kaiña marriages within the same river basins

Kaiña marriages between different river basins

11

28

4

96 · Nalúa Rosa Silva Monterrey

In relation to kaiña marriages within the village, one can say living together in the community would increase the number of marriages included in the prohibited categories. So it would be easy to assume that people who are found in different river basins have less frequent incorrect marriages because the geographic distance favors the manipulation of the terminology. But individuals, on leaving their own community and so increasing their pool of possible marriage partners, may nevertheless finish up marrying someone of a prohibited category, as reflected in the high frequency of people who marry with those in the same river basin but of a prohibited category. Why this occurs is unclear.

Incestuous (Kaiña) Marriages As can be seen in table 5.2, during the Old Period, there was an incestuous marriage, which according to the terminology corresponds to a marriage between a mother and son (actual MZ=ZS). The Ye’kwana consider such a union extremely serious. Similarly, in this category of close unions, a marriage occurred between a father (genitor) and daughter (child) (actual F=D); in this case they are biologically and directly related. Incest also occurred between FS and FD, and between MS and MD, father unknown. They occurred in approximately the same proportion in the Old and Modern periods. Even when the parties to the marriage manipulate the terminology and go ahead and marry, according to the rules those unions are prohibited and are still considered kaiña. The data record 32 marriages between parallel cousins, a prohibited union that, according to the Dravidian terminology, corresponds to the category of brother and sister. The majority of these marriages (70%) occurred in the Old Period, when the population was small and scattered. Demographic Change and the Frequency of Kaiña Marriages There are number of explanations for this phenomenon; one of the more persuasive has to do with demographic and social change. It is clear that from the 18th century and until the rubber boom at the beginning of the 20th century, the Ye’kwana suffered numerous calamities, with a consequent fall in population and with very small and dispersed settlements. The destabilizing factors, particularly the demographic one, could explain the numerous kaiña marriages of 32% for the Old Period. How-

Beyond the Norms: Marriage and Incest among the Ye’kwana · 97

ever, there is almost the same frequency of kaiña marriages (27%) in the Modern Period. As can be seen in table 5.3, the settlements are bigger, the means of transport are quicker (which promotes visits and exchanges with other settlements), the population is growing, it is a period of peace, and the culture is vigorous. In other words, it would appear that variations in the frequency of prohibited, incestuous, or kaiña unions cannot be explained as a consequence of the demographic change. The aforegoing description of kaiña, added to the little importance that is given to marriage in itself, are evidence of the value accorded to marriages between “the closest possible” kin, even consanguines, and above all from the same basin or within the same settlement. The act of marriage is a discreet event, the public recognition of which is made, as already noted, through silent gestures. The union is only considered an accomplished marriage when the couple have babies. The Consequences of Incestuous Marriages As has been described, intergenerational marriages are not infrequent between prohibited intergenerational kin. They are explained only in terms of desire to remain together or because they are in love. The prohibited intergenerational marriages involve difficult decisions, in which social values are put into play and obligate a readjustment of the kinship terminology among a whole network of people (see Silva Monterrey 2010, 229–31). They also involve ego taking decisions with respect to what the people in the social network will be called. Beyond the impact of the adjustments that have to be made to the terminology, which forces people to reflect and come to an agreement about the breach in the norms, the act of marrying kaiña has no other visible consequence at the social level. When the incorrect marriage is between very close relatives (for example, between an actual father and daughter), then people whisper; old people speak of the inconvenience of this type of relation, saying that it is bad, and the couple implicated directly in the relationship try to avoid the topic, being scared of the public recognition of the union, although the couple share a common roof. Ostracism is not practiced. Nevertheless, I have observed, in the case of the marriage between a classificatory father and daughter, that the groom was always trying to justify the marriage by referring to the non-indigenous definition of incest; whereas in the case of a union between an actual father and daughter, the couple tended to isolate themselves from the rest of the village—they censured themselves.

98 · Nalúa Rosa Silva Monterrey

The village community rejected, more evidently and intensely, this latter union, compared to the first example, although both are kaiña. Evidently, the attraction, the physical desire and the need to be close, brings one to the extreme, of breaking categories, of engaging in an incestuous union— an act that, although verbally rejected by the group, brings no other type of sanction.

The Current Situation Nowadays, in the majority of cases, the fathers do not force their children to marry, nor do they decide with whom. It seems that the pattern of marriage exchanges within or between villages or regions persists through its own inertia. The principal view that young people have with respect to marriage, besides the values associated with being the ideal man or woman and the respect for the norms surrounding marriage, is the importance of marrying within the village. I have noticed that among the young people marriage within the prescribed categories predominates; however, kaiña marriages continue to be frequent, although less so than in the past.

Conclusion Among the Ye’kwana, 32% of marriages have been kaiña from the beginning of the 19th century to the 1950s. In the second half of the 20th century, the actual proportion of such marriages is similar—only 5% less. This phenomenon indicates that incorrect or prohibited marriages occurred and continue occurring among the Ye’kwana. However, there is no clear explanation to account for this persistently high frequency of incest. The fact that the percentage has remained fairly constant over a long period of time, during which the political and social situation altered significantly, requires deep reflection; it indicates that there are always people who live beyond the norms.

Notes 1. I thank the Ye’kwana communities of the Caura basin for the help they provided me while undertaking this research; I also thank Dr. Valentine for his translation of this chapter, together with his valuable comments and revisions.

Beyond the Norms: Marriage and Incest among the Ye’kwana · 99

2. A similar discussion occurred with respect to the Pemón terms for cross-cousin. Thomas (1982, 64) was unable to discover the term. It was only later that Urbina (1983– 1984, 195) reported the term wirichi. Like the Ye’kwana, the Pemón cross-cousin term performs a similar function in the kinship system; like the Ye’kwana term, it is loaded with embarrassment, if not shame. Further, the two terms, wodi’chü and worichi, are phonetically very similar. The d is pronounced very much like the r, and the ü like the i. 3. Wodi’chü should not be confused with the term wodi’che, which signifies “little girl” and is used affectionately by men and women when referring directly to boys and girls. 4. I do not discuss here the relationships between changing marital strategies and changing sociohistorical conditions, including such events as epidemics, enslavement, and the incursions of the national economy. I intend to explore this topic, which requires a particular approach, at a later date. 5. I have been given access to the outstanding contributions of the Foucauld Brotherhood archive. Also, I received particularly generous support from Dr. Daniel de Barandiarán, who until the end of the 1950s lived as a missionary in the Caura region. I thank the Foucauld Brotherhood for their very kind assistance. 6. I thank Dr. Selz, of the Laboratoire d’Informatique de L’Ecole des Hautes Etudes en Sciences Sociales, Paris, for her kind assistance. 7. Details of the construction of the database can be obtained in my work (Silva Monterrey 2007 and 2010). 8. Although Spanish names make their appearance during the rubber boom at the beginning of 1900, they only make their general appearance in the 1950s. 9. The data are denser for the Modern Period than for the Old Period.

6 Why Do the Ye’kwana Commit Incest So Frequently? A Discussion of Silva’s “Beyond the Norms” Paul Valentine

In this chapter I attempt to solve a conundrum. Dr. Silva Monterrey, in her chapter on the Ye’kwana, throws down a challenge. She asks why the Ye’kwana apparently commit marital incest with such astounding frequency. Why do they appear to be so blasé about breaking one of their most profoundly held rules? In this chapter I pick up her gauntlet.1 As I construct my own argument, I return to Dr. Silva’s description of Ye’kwana life, sometimes very briefly to summarize, sometimes to add my own necessary gloss, and sometimes to reinterpret in the light of the ethnographic literature on the Ye’kwana. My essay can therefore be read as something that stands alone, without readers constantly having to refer back to its genesis, yet benefiting from being read in tandem with Dr. Silva’s stimulating contribution. The Ye’kwana, which is their auto-denomination, are Carib speakers living in small villages scattered over a vast area of southeastern Venezuela and northern Brazil.2 These villages—some 30 in all (Arvelo-Jiménez 1971, 13), rarely numbering more than 70 people and very often no more than 20 or 30 and sometimes even less (Arvelo-Jiménez 1971, 61)—are located in the tropical rain forest, a few abutting a patch of savanna. Each village (or settlement) typically consists of a single conical-shaped communal dwelling-house (öttö), which has a palm thatched roof and is enclosed by a low mud and wattle wall. At the house’s center is a circular area (annaka) given over as a ritual space and young men’s sleeping area. Around the perimeter of this central space is the women’s area (ösa),

A Discussion of Silva’s “Beyond the Norms” · 101

a circular band, divided into segments, each of which is allotted to one of the village’s resident groups (tünwannodö) (Heinen 1983–1984, 267). A resident group is typically composed of a father-in-law, his wife, their young children, and their married daughters with their husbands. Big villages consist of about six or seven such groupings, whereas small ones have only two or three. In each segment are clustered a number of hearths that correspond to the nuclear families of that resident group. The villagers perceive themselves as forming an autonomous, selfcontained living space. The political authority of those regarded as leaders rarely extends beyond the villages in which they live, and few, if any, privileges accrue to their role. Older men (usually fathers-in-law) who head their residence units tend to join the Council of Elders (inchonkomo). The village leader, “he who takes responsibility for people,” is one of their number (Heinen 1983–1984, 266). Community discussions are not limited to elders (inchomo), although their views are most likely to prevail (Heinen 1983–1984, 269). Nevertheless, their political authority is weak; they cannot sanction and punish any wrongdoing, such as breaches of the marriage rules, and disputes that cannot be resolved are likely to result in migration of some of the village inhabitants. Large villages are intrinsically fragile; this structural weakness manifests itself in a range of disputes, from adultery to food distribution to sorcery, and ends in village division and migration (Arvelo-Jiménez 1971, 26). Each village is deeply suspicious of “outsiders,” who are perceived as troublemakers and harbingers of conflict. There is sufficient evidence in the ethnography to indicate that the Ye’kwana suffer from the xenophobia typical of the region.3 Indeed, the Ye’kwana intentionally build their villages as far away as possible from their neighbors, at least a full day’s journey from the next village, but most of them are even farther apart (Arvelo-Jiménez 1971, 14). The Ye’kwana always settle on the banks of the rivers that traverse the region as these form the main trade routes (Arvelo-Jiménez 1971, 13, 50). However, these rivers are often difficult to navigate. Because of the nature of the seasonal cycle, “there are times of the year when the waterways are little more than fordable streams and other times when they are rushing torrents that flood over the surrounding terrain to head height” (Rivière 1984, 9–10). There are also paths that cut through the forest from one river basin to the next. It usually takes several days to reach a village in an adjacent river basin. It is strongly preferred that marriage partners are drawn

102 · Paul Valentine

from within the village (Arvelo-Jiménez 1971, 66), but marriages within the same basin or with partners from nearby settlements of other basins do occur. The villages that lie along a river section can be referred to as belonging to a certain “river area” or “cultural block”; however, according to Arvelo-Jiménez (1971, 53), this name has a “purely geographical connotation, without political or social significance.”

Kinship The Ye’kwana have a Dravidian relationship terminology that in many ways is similar to that of the Panare (Henley 1982), their Carib-speaking neighbors in the Guyana region. What is unusual about the Ye’kwana is that their genealogical memory goes back as far as some seven or eight and sometimes as many as 15 generations (Silva, this volume) a point to which I return in the section on the quantitative data base. Associated with each kinship term is a set of norms and values that establish a set of appropriate behaviors. The kinship system is prescriptive; in principle, cross-cousins intermarry, whereas parallel cousins cannot (Silva, this volume; Heinen 1983–1984, 276–77). The terms for parallel cousins and siblings are the same and are distinct from those for cross-cousins. It is also possible to marry individuals with whom no genealogical relationship can be traced. Prohibited marriages are labeled kaiña, a term that has sometimes been translated as “incestuous” (Arvelo-Jiménez 1971 and Silva, this volume), a point to which I return in the section on incest. When people come to live in a community they are reclassified as “real,” ne’ne, kin (Arvelo-Jiménez 1971, 174). Also, when cross-cousins marry, married ego calls opposite sex cross-cousin “brother” or “sister,” which are also the terms for parallel kin (Silva, this volume). Ye’kwana classificatory grandparents and grandchildren are prohibited from marrying unless they are approximately the same age, in which case married ego always classifies them as cross-cousins, and upon marriage ego reclassifies them as parallel kin (Arvelo-Jiménez 1971, 155; Silva, this volume). One further point can be made here. It is common to find that kinship labels are mutable in Amazonia. By coming to live in a village, by sharing food and perhaps sexual fluids with one’s partner, an individual can gradually become kin (see, for instance, Conklin 2001, 146; Pollock 2002, 45; Sahlins 2013; Valentine, this volume). In this sense, it would seem that

A Discussion of Silva’s “Beyond the Norms” · 103

Ye’kwana terminology shares this quality of mutability with others in the region. However, the Ye’kwana are unusual with respect to the speed at which they make such a switch. Immediately upon marriage a spouse calls his or her cross-cousin younger or older brother or sister. Indeed, they have no term for brother-in-law or sister-in-law (Arvelo-Jiménez 1971, 155–56; Silva, this volume). This practice is another point I return to later.

Marriage Ye’kwana marriage is regulated by a rule that constrains an individual to marry his bilateral cross-cousin. Unlike in a number of Carib groups, marriage with the sister’s daughter or any other Alter of an adjacent generational is decidedly wrong (Heinen 1983–1984, 276). As already mentioned, the Ye’kwana do practice marriage between the grandparent and grandchild generations, and although marriage between real grandparents and grandchildren is prohibited, all other members of these generations are potential spouses, if they are approximately of the same age. Young people have some say as to whom they choose to marry; nevertheless the parents have considerable influence too. When an agreement has been reached, the marriage takes place. In a simple ritual, without public fanfare, the future wife goes to the men’s space (annaka) at the center of the house, takes down her spouse’s hammock, and hangs it in her section of the women’s space (ösa), the house’s perimeter band. The process of becoming married is further cemented with the birth of their children. Although this ritual is in no way elaborate, it does not negate the profound importance of marriage and the significance of selection of an appropriate partner.4 Here I disagree with Silva (this volume) when she writes that “little importance . . . is given to marriage in itself.” In the past, the decision as to whom one should marry was taken by the parents, frequently when the couple were still children (Silva, this volume). Further, it is to the father-in-law’s profound interest that the future groom be suitable, because it is he who provides bride service on marriage. Moreover, the whole village prefers that the individuals who will become the new couple select their partners from within the settlement. As the Ye’kwana practice uxorilocal residence after the marriage (Arvelo-Jiménez 1971, 66; Heinen 1983–1984, 277; Silva, this volume), by engaging in village endogamy, they avoid having their young men forced to live elsewhere. Indeed,

104 · Paul Valentine

the best possible option is village endogamy. Marriages between villages of the same region are preferred to a lesser extent, whereas marriages between regions are the least favored. The Ye’kwana are typical of other Guyana Carib groups in other ways also. They lack any genealogically defined corporate group, nor are there identifiable and enduring groups that can be labeled as kin and affines that stand in a relation of spouse exchange. In other words, for the Ye’kwana, and for the Carib ethnic groups in general, there are no exchanges of women at an institutional level; for instance, no sisters are exchanged between groups of Ye’kwana men.

Kaiña, Incest, and Prohibited Marriages According to Silva, kaiña can be translated as “bad” or “wrong,” and both Silva and Arvelo-Jiménez use the term to describe “incestuous marriages,” although they recognize that there is no Ye’kwana word that signifies “incest” in the etic sense. Heinen (1983–1984, 276) avoids the word “incest” altogether; rather he defines kaiña marriages as unions that “do not conform.” I follow his example and use similar phrasing, rather than use the word “incest,” which in this context can be misleading. All ethnographers agree that the Ye’kwana use the phrase kaiña to describe marriages that are incorrect or prohibited; that is, marriages between primary kin, between adjacent generations, between parallel cousins, and between grandparents and grandchildren biologically defined (Arvelo-Jiménez 1971, 137; Heinen 1983–1984, 276; Silva, this volume). Silva comments (this volume) that marriage between “very close kin” meets intense disapproval; in this context she is probably referring to marriages between parents and their children; between adjacent generations, the most common of which are those between mother’s brother and sister’s son or between father’s sister and brother’s son; and between siblings, including half-siblings, biologically defined. One could perhaps conclude from her remark that marriages between parallel cousins are looked on with less severity. Heinen and Arvelo-Jiménez agree with Silva that prohibited marriages are considered wrong and that the couple face considerable disapproval. On occasion, they may even isolate themselves from the rest of the community, rather than face the public shame and scandal. Unfortunately no anthropologist has specifically described Ye’kwana attitudes toward parallel cousin marriage. It seems unlikely that

A Discussion of Silva’s “Beyond the Norms” · 105

such marriages meet the same level of public disapproval as those between primary and secondary kin because, as we shall see, a fairly large proportion of all marriages are between parallel cousins.

The Quantitative Data Base One way to assess Silva’s data, and the use to which she puts them, is to answer the following questions: 1. Are the data a good representation of the general population? 2. Are they reliable and fit for the purpose to which they are put? 3. Are they analyzed satisfactorily? I discuss these three questions but limit my comments to only those areas that have a direct bearing on my own argument. Combining her data, which she collected from 1986 to 2000, with that obtained by de Barandiarián (1966) in the 1950s, Silva has a bank of over 2,000 data points. Each data point includes information on a person’s name, sex, genealogy, ethno-historical origin, and place of residence. The data cover an enormous time frame from the early 1800s to 2000. From this information, Silva has selected 200 marriages, which refer to people alive and dead. These 200 data points were selected because each has the complete set of information on both spouses’ four grandparents.5 She asserts that this information provides her with the opportunity to calculate the number of prohibited marriages that have taken place. Selecting data points on the basis of the known names of the four grandparents of each spouse is probably an unbiased sampling technique, given that we know the sample was taken from several villages. There is, nevertheless, one minor caveat. Morrill and Dyke (1980, 6) have shown, by comparing genealogies collected in the field with recording census data, that an alarming number of individuals are forgotten, whole generations may disappear, and genealogies may become telescoped when researchers are collecting field data. Consequently it is possible that when Silva relied on informants’ memory of those long dead, the named grandparents may, in fact, have been more distant ancestors, promoted to that position because the generations in between have been forgotten. Turning to explore the reliability and quality of the data themselves, a couple of points are relevant. Anthropologists are well aware that genealogies collected in the field in other cultures do not necessarily represent the

106 · Paul Valentine

facts of descent and relatedness. Indeed, Morrill and Dyke (1980) show that the discrepancies between the ethnographic approach and the documentary one comply with certain clear patterns and indicate the kinds of demographic information likely to be missed. Especially liable to bias is material that relies on informants’ memory of people now dead. Silva does not distinguish the data collected from those still alive from the material that relies on informants’ memories of the past. This oversight creates another layer of uncertainty and makes it more difficult to estimate the degree of reliability of her material. Nevertheless, Silva has attempted to eliminate errors by comparing genealogies drawn over a wide geographical area and correcting for inconsistencies. It is unclear whether there was sufficient replication of material for her to compare her data based on informants’ recall of genealogies of the 1950s with those collected by de Barandiarián in the 1950s from informants alive at that time. If so, it would provide a golden opportunity to estimate the degree of reliability in her survey. However, as anthropologists, many of us, like Silva, are stuck with informants as the sole source of genealogical data, and these reconstructions are far better than the ignorance we would face without such studies. Nevertheless, wherever possible it is worthwhile to calculate the degree of error, because if it is significant then it may invalidate our conclusions. The existence of such biases suggests that we should be wary of over-interpreting our results. The degree of error may be greater than the apparent differences in the data sets. For instance, when I start comparing and contrasting prohibited marriage rates, as discussed later in this essay, that may be pushing the data beyond the useful limits. The weaknesses that have been reviewed so far are fairly typical of the kinds of problems anthropologists encounter when collecting genealogical material. However, there is one area where the bias is peculiar to the Ye’kwana. Ye’kwana genealogies are known to go back some 15 generations. Such a finding is extremely unusual in Amazonian ethnography and perhaps accounts, in part, for Silva’s sample spanning such a long historical time period. But how reliable are these genealogies? Are they perhaps used to rationalize the social, political, or economic needs of the Ye’kwana themselves? It so happens that Arvelo-Jiménez (1974, 264ff) has provided a brilliant description of verbal duels when these genealogies are wheeled out as ammunition to bombard one’s adversary. The purpose of the duel (literally called “to speak about one’s own head”) is to show off the purity of one’s own Ye’kwana line compared to the dubious ancestry

A Discussion of Silva’s “Beyond the Norms” · 107

of one’s adversary. Having a long ancestry can only have been acquired if one’s ancestors were Ye’kwana stretching back to the beginning of time. Not only that, they demonstrate the purity of the whole village, or at least the core members, while at the same time subtly undermining the credentials of the opposing protagonist. “Every known non-Ye’kwana ancestor is a cause of embarrassment and is kept a ‘deep secret’ even though such information is well known in the community and often beyond” (Heinen 1983–1984, 272). In this context one would expect it likely that genealogies would be massaged, kinship denotations changed, village origins strategically forgotten, and prohibited marriages glossed over. How significant this factor is in undermining the reliability of Silva’s data is unclear. As she acknowledges, it is the early material that is the most sketchy and questionable. Perhaps, therefore, she has factored in these sources of possible error, selecting the most persuasive options in the construction of her data base. Nevertheless, in spite of these weaknesses, and in part because of the paucity of Amazonian numerical data on kinship and marriage, the material Silva provides is a fascinating insight into how the Ye’kwana see their world. I now turn to explore her use of the data and limit myself to comments that have a direct bearing on my own argument and explain the differences between my conclusions and hers. The 200 data points that Silva selected on the basis that they are the only ones available in which four grandparents of each spouse are named seems to me an unbiased sampling technique. However, Silva discards 57 of them because these couples do not share a named grandparent, thereby reducing her sample size to 143. As this exclusion brings bias into the sampling procedure, I ignore it, and instead use the original sample of 200 data points. This inclusion reduces the proportion of prohibited marriages and, in part, explains the discrepancies in my findings compared to those of Silva. There is another way in which discrepancy occurs. Silva calculates the proportion of kaiña marriages based on the total number of such unions. In my analysis I prefer to focus on parallel cousin marriages and ignore unions between “close kin.” Close kin marriages refer to unions between primary kin; between immediately adjacent generations, particularly between the mother’s brother and sister’s daughter; and between actual grandparents and grandchildren. I do this for the following reasons. First, there are a large number of parallel cousin marriages and they vary over

108 · Paul Valentine

time, whereas marriages between close kin are few in number and, as Silva points out, are approximately constant throughout the period. Marriages between close kin occur because the couples love each other and desire to stay together (Silva, this volume) rather than, it would seem, for any structural reason. Second, such marriages are rejected by the community, even though the couples attempt to redefine their kinship relationship. A number of “wrongly” married couples are living in towns now. Rather than assume as Silva does that the couple met in town and married there, it seems more likely that some, at least, migrated from their villages in an attempt to avoid the stigma of their union. To what extent do these differences in Silva’s and my calculations affect the findings? Because Silva restricts herself to a small and biased sample, she overestimates the proportion of kaiña marriages. She considers that from a sample of 143, 43 or 30% were “incestuous,” whereas according to my calculations there were 200 marriages of which 43 or 21% were “out of order.” In other words, Silva overestimates the proportion of kaiña unions by some 10%. Second, she does not take into account the quantitative differences in the kind of kaiña unions. We know the number of unions between close kin was small, only 10, or 5%, and as far as the numbers permit us to tell, there seems to be no variation in the frequency of such unions over time, whereas there were 33 parallel cousin marriages, or 16.5%, and as we shall see later, over time there may have been significant fluctuations in their numbers. Interestingly, with these minor calibrations to Silva’s data, they correspond very closely with those provided by Arvelo-Jiménez. Arvelo-Jiménez offers material collected in 1968–1969 on 160 marriages that took place in seven Ye’kwana villages. She includes figures on the number of kaiña marriages between actual and classificatory parallel cousins. She also lists the number of kaiña close kin marriages. From Arvelo-Jiménez’s material it is easy to construct a table that compares her data set with my revised version of Silva’s findings (see table 6.1). From it we can draw the following conclusions: approximately 20% of all marriages are kaiña; a number fluctuating around 15% are between parallel cousins; and 4% or 5% are between close kin (i.e., between primary kin, those in adjacent generations, and between actual grandparents and grandchildren). Arvelo-Jiménez does not discuss why some 20% of all Ye’kwana marriages break the rules, and it is to Silva’s credit that she picks

A Discussion of Silva’s “Beyond the Norms” · 109

Table 6.1. Kaiña marriages Total marriages

Kaiña marriages

Silva

143

30%

Valentine

200

Arvelo-J

163

Marriages between parallel cousins

Marriages between close kin

21%

16.5%

5%

19%

15%

4%

up on this point. This surprising Ye’kwana characteristic is discussed in the latter part of this chapter. I now turn to a related topic. A cursory glance at Silva’s table 5.4 may give the wrong impression. The table provides numbers of three kinds of kaiña marriages (including incestuous marriages between primary and secondary kin): (1) within a settlement, (2) between settlements of the same river basin, and (3) between river basins. Reading the table one may gain the impression that there have been more prohibited marriages within the same river basin than within a settlement. However, this table is, in part, based on de Barandiarián’s data, and he fails to distinguish between prohibited marriages within a village and those within a river basin. He simply conflates them into one category. Therefore, there are an unknown number of cases of prohibited marriages in a village within the category of “prohibited marriages within the river basin.” In short, given the quality of the data Silva is using, it is impossible to calculate the frequency of prohibited marriages within a village. However, and this is important for the argument presented here, Heinen and Arvelo-Jiménez have collected data on village endogamy. Heinen writes that completed tabulations of 1978 data show that the Ye’kwana of the Erebato basin are essentially village endogamous (1983–1984, 272). As mentioned, marriage is uxorilocal; however, ideally young married sons do not leave the community but move into another extended household of the communal roundhouse. He notes that less than 10% of the men migrate to different villages. Further, the data collected by Arvelo-Jiménez (1974) in the late 1960s in the headwaters region of the Venturari River seem to confirm this estimate.6 I now explore the factors that may account for the frequency of prohibited marriages, the most prevalent of which are between parallel cousins.

110 · Paul Valentine

The Selection of Marital Partners: How the Kinship Rules Work in Practice Arvelo-Jiménez, Heinen, and Silva all agree on one important point. The guide that defines relationships continues to be the Ye’kwana kinship terminology, which provides the key to who are, and who are not, potential marriage partners. To recapitulate briefly, the rules are simple: marriage with the cross-cousin is considered ideal; and marriages between primary kin, between adjacent generations, between the grandparent and grandchild when the age difference is significant, and between parallel cousins are all out of order or kaiña. Classificatory grandparents and grandchildren, providing they are approximately of the same age, are always classified as cross-cousins and can marry. For the Ye’kwana, as for many other ethnic groups in Amazonia, traditionally wealth is not measured in terms of material possessions but rather in terms of the number of inhabitants living together or the control a leader has over his following (Rivière 1984, 93). Since marriage is seen as a mechanism that has the potential to redistribute the population throughout a whole area, marriage has a greater significance than an end in itself. Indeed, the Ye’kwana consider that to marry out of the village, irrespective of the emigrant’s gender, brings impoverishment to the community (Arvelo-Jiménez 1971,136). How do these rules and preferences work in practice? Arvelo-Jiménez (1971, 66) writes that the greater the village size, the greater the chance that a person will find a partner within the village. In a sense this is broadly true, but it avoids the question, how did the village become large in the first place? If we approach the problem in the following manner we can see what options are open to the villagers themselves. Each village has a core. Usually this consists of a pair of siblings and their partners, their children, and perhaps their children’s children (Arvelo-Jiménez 1971, 109–34). Assuming that the pair are a brother and sister, which, from the Ye’kwana point of view, is the ideal (Arvelo-Jiménez 1974, 86, 123), their children will be cross-cousins and able to marry within the village (figure 6.1).7 Now assume that the village’s core consists of two brothers or two sisters with their partners and children. Also assume that there are no cross-cousins available in the village, which is not improbable, given that some villages have only some 20 or fewer inhabitants. In these cases their children will be unable to marry each other

A Discussion of Silva’s “Beyond the Norms” · 111

Figure 6.1. Two residence groups A and B; the 0 generation composed of a brother and sister.

because they are parallel cousins. What are the options available for these young people? The young man could marry outside the village. Here he is following the uxorilocal rule; nevertheless, there are key disadvantages. On the one hand, his natal village will be displeased by his moving to another village, because as the other village gains a member, his home village will lose one, thereby reducing its potential for autonomy and resilience (Arvelo-Jiménez 1971, 75; Heinen 1983–1984, 267). On the other hand, from the young man’s point of view he has to migrate to another village where he will be perceived as a foreigner, he may have no kin support, and he will possibly be viewed as a center for conflict and village division (Arvelo-Jiménez 1971, 66). Further, he will be under the control of his father-in-law (Arvelo-Jiménez 1971, 104). So from various perspectives these options have serious shortcomings, and that is not taking into account that, as mentioned, villages are far apart, so the young man will be living a long way from home. What are the options if the young man stayed at home and tried to bring a woman into his village? In fact, the only young women who would be prepared to come to his village as his wife are those who had been adopted at a young age and live in their foster parents’ village (ArveloJiménez 1971, 75). But they form only a small pool, and sometimes the young man is unaware of such a woman’s location or unable to claim her as his wife (Arvelo-Jiménez 1971, 77). Further, it is most unlikely that other young women would agree to leave their own village and come to live with him in his: it would break the uxorilocal rule of residence; the

112 · Paul Valentine

woman’s father would be displeased because he would lose not only her labor but also the labor of a son-in-law; and other men in the woman’s natal village would feel jilted by her action. They consider that she should marry one of them. If there are no single men, it is no matter—from the natal villagers’ point of view it would be preferable if she married polygamously rather than leave the village. Indeed, Arvelo-Jiménez writes that for a young woman to leave her natal village is regarded by the Ye’kwana as “sinful” (Arvelo-Jiménez 1971, 70), and in the same way as when a man goes to live elsewhere, she will feel a foreigner in a distant land, perhaps unsupported by kin and a possible center for contention later. As all these options may block or bring serious disadvantages to the marriage, there is only one remaining option left—and that is to marry the parallel cousin. Although this option is seen by the Ye’kwana as out of order, when we review the data this choice is frequently selected. We know that some 16.5% of marriages (or 33 of 200 marriages) fall into the prohibited category. What are the repercussions of such prohibited marriages to the kinship terminology? Does it throw the network of relations out of kilter? One of the common features of prescriptive terminologies is that when the prescriptions are broken, then the kinship terms are altered post facto to correspond to the ideal. For the Ye’kwana the ideal is that everyone is consanguineously related. This ideal is sustained by the acceptance of a number of fictions. First, by gaining residence in a village, ego becomes a consanguine (Arvelo-Jiménez 1971, 171). Second, as we have already noted, on marriage parallel and cross-cousins are lumped together as brother and sister. At one stroke this renaming reaffirms the ideal in two ways: (1) by incorporating cross-cousins into the consanguine fold they become co-residents in the village, and (2) at the level of the terminology, by lumping together parallel and cross-cousins as brother and sister, it avoids the disclosure of which of these marriages were prohibited; it keeps the ideal of cross-cousin marriage intact. The substitution of kin for affinal terms masks the high frequency of parallel cousin marriages, which according to the kinship terminology are prohibited. Turning to marriages between grandparents and grandchildren, if the couple are of markedly different ages, then the union is considered to be out of order, whereas if they are of the same age, they are classified as cross-cousins, and the marriage becomes prescriptive. The fiction here is that a relationship that was once seen to be based on genealogical descent is transformed into a collateral and affinal connection.

A Discussion of Silva’s “Beyond the Norms” · 113

Looking at both the parallel cousins and the grandparent–grandchild relationships together, we see an overriding principle of the kinship terminology. It is the genealogical and consanguine categories that are overridden by those of age and residence. Being of the same age and living in the same village trump descent and marital connections. That these two relationships are stressed is significant, for they reflect the collective idea about the nature of consanguinity and affinity, the way fiction can be smuggled into the terminology and the way flexibility is maintained to guarantee that the ideal is kept intact. One other point has to be added in part to avoid the accusation that an ethnocentric vision of kinship has been adopted. I have defined these relationship manipulations as “fictions,” but people’s place in society can be viewed in different ways, and it is the focus of the analysis that has led to this view. If we were to switch our perspective and not emphasize the genealogical, lineal, and affinal interconnections of kinship but rather the manner by which individuals are classified and reclassified, then these transformations are no longer seen as fictions; rather they reflect better the way Amazonian people categorize and make sense of their social universe. Neither way of looking at kinship is exclusive of the other, yet by approaching the subject in a different way we arrive at the same conclusion but with a different emphasis. This second approach is apparent in the way that one category can be substituted for another. The Ye’kwana are a small-scale society, where everyone knows everyone else and everyone is related to everyone else (Arvelo-Jiménez 1971, 169; Heinen 1983–1984, 272) thanks to the classificatory kinship system and the rule stating that co-residents are consanguineous kin. When the Ye’kwana reclassify a grandfather into a crosscousin, thereby nullifying the old classification or substituting the new one, that change has a ripple effect on ego’s relations with everyone else. By redefining a grandfather as a cross-cousin, ego becomes a cross-cousin to others, and a brother to another set of people, and so on. This reclassification has not involved recourse to genealogy or affinal relations—do not forget that most people will be related to others by a number of different pathways. Rather, it involves an attempt to gain some consistency among the dyadic classifications that make up ego’s social network. In other words, the Ye’kwana are not engaging in a fiction; they are applying their rules of classification as they know them, in the light of the political and economic context and the sentiments of those involved. In short,

114 · Paul Valentine

it seems likely that the Ye’kwana perceive themselves under threat from outsiders, and because they prefer village endogamy, when there is no suitable cross-cousin available, they marry their parallel cousin.

Why Do the Ye’kwana Not Sleep with Their Enemies? The explanation that I have put forward is still partial and inadequate. It faces the objection that I have simply reiterated that which the Ye’kwana themselves probably say; we do sometimes engage in marriages that are out of order, we are afraid of our enemies, and so we are forced to marry within our own communities. The case I have put so far does not develop the argument much further than that. It does not explore the underlying interrelations involved: how are kinship rules and obligations linked to politics and production processes? And what happens when there are marriages out of order? We have not begun to explore what the deeper consequences would be of engaging in prohibited marriages. Indeed, why do the Ye’kwana not marry their enemies? In many other societies people do. Here I have to stake out my theoretical position, which is in no way original but a necessary clarification before we move on. I think that the multitude of decisions and actions that people make interact, and in the interconnectedness of these parts new properties—such as institutions and culture—emerge. In addition, such group effects transcend and causally constrain individual action but do not negate individual agency. Many anthropologists do not agree, and see social life merely as a series of processes and strategies made up of a myriad of individual decisions and actions that people make all the time. They criticize the view I have put forward for its dualist ontology, depicting society as driven by groups and collectives when all that exists are individuals. However, Sawyer (2005) and Sims (2012) have shown how the new consensus among the philosophers of the mind (who are not merely relabeling Durkheim’s ideas about collective phenomena) is focused on the way individual actions are changed by their association in the complex whole, and how they can be drawn upon to critique post-modern and post-processional approaches to the humanities. Returning to the Ye’kwana: their economic life is based on a mixture of hunting, fishing, slash-and-burn cultivation, and gathering that gives rise to extensive rather than intensive use of resources. Mutual obligations

A Discussion of Silva’s “Beyond the Norms” · 115

play an important part in production. The men often cooperate in a work group, conceptualized by the Ye’kwana as “working together,” building a house; moving from garden to garden felling and clearing the trees and undergrowth; and sharing their fish and game (Heinen 1983–1984, 265). At the same time as the men take on their shared obligations, the women—who have their own domain, the outer ring of the round house— form their collective groups. The elder women take charge of work groups for the gardens; indeed women spend some 26% of their time working gardens other than their own (Hames 1987, 267). Political leadership is based on the control of young men, and these young men go and work in groups at the behest of their fathers-in-law. The father-in-law, “the one who has (potential) sons-in-law,” is likely to head the resident unit and have a place on the Council of Elders (Heinen 1983–1984, 269). Men do not participate in communal activities as individuals but generally as members of extended households together with the old father-in-law or his delegates. Hunting territories and fishing areas are not privately owned but are claimed as “habitually visited by an extended family unit” (Heinen 1983–1984, 266). Thus kinship relations are expressed in work groups (tujuumoto), where people work in a collective cooperative non-individualistic manner. In other words, we can see that politics, kinship, and production work are, as it were, of a piece; there is a pattern of collective action and sharing within an egalitarian framework. It is true that the old control the young, but this is a cyclical pattern and the young, in time, will have their day. Their key social and economic unit, as Heinen (1983–1984, 265) ably describes, is the relationship between father-in-law and son-in-law and in particular the control that the father-in-law exercises over his sonin-law’s labor. The argument is that young men provide services to their father-in-law and the father-in-law responds by supporting them. Over time the junior couple’s own daughter marries, and then the junior family can break away and form its own extended household and the old fatherin-law and his wife can come and live with his son-in-law, a situation the very reverse of what happened when the young man first married his father-in-law’s daughter. What we see here, I suggest, is the old men controlling the young men’s labor through enticing them to have sex with and marry their daughters. As we see below, when there is cross-cousin marriage there is a circulation of men, the women stay put, and this pattern is repeated through the generations.

116 · Paul Valentine

Figure 6.2. Cross-cousin marriage, illustrating the circulation of bride service.

We see in the diagrammatic representation of cross-cousin marriage (figure 6.2), a sharing and distribution of services between the son-inlaw and father-in-law between residence units. The same pattern emerges when marriages take place when there is a two-generational gap between bride and groom (figure 6.3). This sharing and circulation of services between residence units is possible because, in the Ye’kwana case, it is impossible for real grandfathers to marry their real granddaughters. In short, when there is cross-cousin marriage the cyclical pattern of relationships facilitates collective economic activity; economic and political life is egalitarian and non-hierarchical, and sons-in-laws’ services are shared among the extended households. After this very brief sketch of Ye’kwana economic and political life based on the ethnographic literature, permit me to pursue an even shorter thought experiment, and ask what would one expect would be the consequences if all or nearly all marriages, instead of being between cross-

Figure 6.3. Classificatory GP/GC marriage, illustrating the circulation of bride service.

A Discussion of Silva’s “Beyond the Norms” · 117

Figure 6.4. Parallel cousin marriage, illustrating the concentration of bride service.

cousins, were between parallel cousins? What effect would that have on economic and political life? Judging by the many cases recorded in the ethnographic literature—for instance, Malinowski’s (1959, 77–8) famous account in Crime and Custom—the legitimate suitor, in our case the cross-cousin, would feel aggrieved and resentful that a parallel cousin has taken the potential bride. Further, one can predict that a marriage out of order would create a blockage in the circulation of men, and by the same token instead of the services of sons-in-law being shared by various fathers-in-law among the different residence groups in the village, there would be a concentration of them (see figure 6.4). If such practices persisted, one would expect a separation of residence groups and a decline in collective production and shared food distribution. One might assume that because the son-in-law is also the classificatory son perhaps marrying into his own residence unit, the flow of services is going in one direction without their circulation; there is a monopoly of services. Such patterning would facilitate individualistic action and not collective behavior; it would atomize rather than harmonize the community; it would not create an egalitarian society but rather a hierarchical one. Without doubt, such an outcome has not occurred. For instance, there is no description of such friction in Arvelo-Jiménez’s excellent book on Ye’kwana politics, nor in Heinen’s comprehensive article on Ye’kwana economics and kinship does he describe the consequences of prohibited marriages on the economy. Indeed, he specifically writes that public disapproval of kaiña marriages strongly discourages such unions (Heinen

118 · Paul Valentine

1983–1984, 277). How is it that the Ye’kwana have been so successful in covering up the high frequency of prohibited marriages that they have gone almost unnoticed, at least until now, by all the extremely competent anthropologists who have carried out fieldwork among them? The answer must be that the diagrams or simple models depict extreme situations that palpably do not exist. In the second example it was assumed there was a very high proportion of parallel cousin marriages. But we know from the data that the number is much less; about 15% of marriages are between parallel cousins, much fewer than would be able to channel the services of sons-in-law to a few powerful leaders. Second, even if there are parallel cousin marriages in one generation, this occurrence does not preclude the reemergence of cross-cousin marriages in the next. So any tendency toward hierarchy in one generation can be counteracted in the following one. The figures depict the relationship between bride service and cousin marriages and focus on the extremes; in a way this representation is similar to economists’ models of perfect competition and monopoly. Perfect competition predicts the free circulation of money for goods and services, whereas monopoly predicts friction in the flow of money, which flows toward the monopolist. Ye’kwana social and political life is located, in the same way that the economist’s models locate market behavior, somewhere between the two extremes of their constructs. What the models are designed to do is facilitate the construction of hypotheses that can be tested. So, for instance, from our discussion of the frequency of parallel cousin marriage, one can predict that with an increase in hostilities between and toward the Ye’kwana, villages would divide and scatter as the Ye’kwana fled into the depth of the forest. With a decline in village size there would be an increase in parallel cousin marriage. Our simple model predicts that with the increase in parallel cousin marriage, there would be the emergence of elites as the flow of women and bride service gravitated toward them. (Obviously the introduction of parallel cousin marriage could not be seen as the only causal factor leading to these social, political, and economic changes.) With the decline in hostilities, it would be predicted that the power of the elites would be dissipated and there would be a return to the predominant pattern of cross-cousin marriage. To what extent do the data support these predicted oscillations? To attempt to answer that question we turn to Ye’kwana social structure, and their history and geography, which, as we will see, also all throw light on

A Discussion of Silva’s “Beyond the Norms” · 119

why the Ye’kwana do not follow Tylor’s advice and marry their enemies (1889, 267). First, I turn to the structural feature that, in part, accounts for why the Ye’kwana do not marry their enemies. Unlike the Tukanoan Indians of the northwest Amazon, groups of Ye’kwana men never exchange their sisters or daughters to broker peace deals. As we have seen, the cornerstone of Ye’kwana social, political, and economic life is the relationship between the son-in-law and the father-in-law. It is a person to person relationship, one between individuals, not one between groups. The result is that such a bond does not involve enough people to sustain an alliance between villages. In part, as a consequence of such a weak connection, young men must perform a very long period of bride service, a duty they can only fulfill by living uxorilocally. Such an arrangement contrasts dramatically with the Tukanoan practice of marriage exchanges between groups, which do indeed form the basis for alliances between villages that can persist for generations; and incur only short periods of bride service, sometimes altogether avoided if both parties exchange brides at the same time; and are followed by virilocal marriages. In other words, it is impossible for Ye’kwana women to marry their enemies because of the structural position they play mediating the relationship they establish between men. I now turn to the contingent factors, the peculiarities of Ye’kwana geography and history, to account for why Ye’kwana men stay at home. Moreover, they provide data to test the predictions. The Ye’kwana live in an enormous area of some 30,000 square kilometers. Their population was approximately 1,500 in 1968–1969 (Arvelo-Jiménez 1974, 11)—and probably much less in the early 1900s—with a density of population of approximately 20 square kilometers per person.8 They live in the forest isolated from their neighbors; choosing to live as far away as possible from other settlements, often they are separated from them by major rivers, which do not facilitate movement between these river basins. And second, their history is full of threats and violence. As Arvelo-Jiménez notes, during these periods of “crisis registered by the ethnography, the Ye’kwana either split their villages into even smaller groups, blending into the forest . . . or concocted temporary interregional alliances that resulted in the centralization of political power” (Arvelo-Jiménez 2000, 732). In the early part of the last century the crisis was caused by the gangs of rubber traders’ henchmen out to seize the indigenous population to work for them. These incursions led the Ye’kwana to fragment into smaller and

120 · Paul Valentine

smaller units and flee farther into the forest. Arvelo-Jiménez writes that with this fracturing and forced migration of the Ye’kwana, there was an increase in village endogamy. During the rubber boom the resident pattern demanded that the husband marry into his own community because otherwise it might have diminished and even disappeared (Silva, this volume). “That is why, the Ye’kwana say, it is preferable to marry into one’s own local group and be with one’s own kind” (Silva, this volume). In short, the Ye’kwana have taken the option of marrying in and surviving rather than marrying out and losing the protective force of their young men.9 We know, extrapolating from Silva’s data, that there was a greater number of parallel cousin marriages before the 1950s than after—approximately 19% compared to 13%.10 However, as discussed, whether this difference is significant given the level of possible error in the data themselves is a moot point. Further, that the two events occurred at the same time is only indicative and not conclusive of any causal link. Nevertheless, the data do support the hypothesis. More fieldwork is required. Coinciding with the splitting of villages into smaller units during times of crisis, as mentioned there was also a tendency toward a centralization of political power. Arvelo-Jiménez writes (2000, 732): “The social system has an ingrained potential for centralization” that was triggered by the external threat of violence.” From about the 1870s to the 1950s, the Ye’kwana were subject to attacks by the Sanema, who stole from their gardens and kidnapped their women. The Ye’kwana responded by creating a political elite and a war leader, who organized the scattered villagers together into regional alliances and coordinated a corps of warriors who successfully repulsed these invasions (Arvelo-Jiménez 2000, 732–33; Heinen 1983–1984, 274).11 Based on the Yanomamo literature (Chagnon 1992, 158), it is possible that the leader disregarded the Ye’kwana marriage rules and married his parallel cousins—he was powerful enough to do so. What we do know is that such leaders often had several wives. Further, Ye’kwana shamans are frequently powerful enough to marry polygamously and flout the marriage rules. It is not uncommon for them to have several wives in different communities whom they have bullied into marrying them (Arvelo-Jiménez 1971, 101). Again, it is an area that calls for additional research. In conclusion, I consider the proportion of prohibited or “incestuous” marriages to be somewhat less than Silva calculates. Based on Silva’s stimulating research, it is posited that there are links between village en-

A Discussion of Silva’s “Beyond the Norms” · 121

dogamy, the scarcity of cross-cousins, and the consequent marriage with parallel cousins. Structural and contingent factors are offered to account for why the Ye’kwana have “married in” and survived. Also, several hypotheses are suggested that could be tested with additional historical research and fieldwork. A number of tentative generalizations are made with respect to the bending and breaking of the marriage rules. A significant number of Ye’kwana marriages are incorrect or “incestuous.” However, parallel cousins can become cross-cousins through the retrospective manipulation of the kinship terminology. Residence and age take precedence over descent and marital connections. Therefore, one’s kinship identity is not fixed but mutable; it is not a given but, as it were, a process dependent on where one is living and with whom one is sleeping. But there are limits to the accommodation. The wild card is love. When very close kin, such as the mother’s brother and the sister’s daughter, fall in love and start to live together, although they attempt to mask their true kinship identities, that attempt fails. It seems likely that it is they who are expelled from the community and make their way to the towns. In other words, by bending the rules, but within limits, the ideal kinship structure, “we marry our cross-cousins,” is kept intact.

Notes 1. I thank Dr. Silva for allowing me the opportunity to resolve this problem. She has done so on the commendable desire to promote debate on her data. 2. In the literature they are frequently referred to as Maquiritare, but this is a term completely foreign to the Ye’kwana. Durbin and Seijas, after an intensive review of the literature, collected 53 terms to refer to the Ye’kwana or Du’kana (Arvelo-Jiménez 1974, 1). 3. Arvelo-Jiménez (1971) emphasizes this point. She writes that the attitude toward “non-villagers is one of covert suspicion and distrust at its best” (323). “In-marrying men are not always accepted . . . gladly” (75); “there are problems of assimilating a foreigner into village life and satisfying the group ideal of internal solidarity” (76); “an in-marrying individual might encounter opposition from one or more family groups within the recipient village” (76); “out of all the recorded cases of artificial attachment to another village there are only three cases of individuals . . . whose attachment to highly solidary family groups [was] quite successful” (85). Indeed, “Ye’kwana deliberately establish their villages great distances apart” (47). 4. According to Arvelo-Jiménez the marriage ritual is not as simple as Silva suggests. It is true that the marriage ceremony involves the bride taking the groom’s hammock from the men’s area, but Silva does not mention that before this the young woman cooks

122 · Paul Valentine

for her future partner every night for 10 days or so, and they talk in private for the first time. Only if the young couple’s doubts are cleared, particularly with respect to how the groom will treat his future parents-in-law, will she, on the orders of her mother, unhitch her husband’s hammock and carry it to her place. 5. Silva fails to inform us as to whether these 200 cases are the only ones she has that contain the information on both spouses’ four grandparents. If there are other examples, then one would have to ask why she has selected these 200 cases to the exclusion of others? Probably it is an oversight on her part, and indeed there are only 200 relevant data points, because if there had been more, surely she would have included them to enrich her study. 6. Ye’kwana territory is traversed by major rivers that more or less divide the area into regional groups, one of which is the Erebato basin and the upper Venturari River. Nowadays, with the speeding up of communication and as a result of past incursions, regional formations are barely discernible (Heinen 1983–1984a, 273). 7. Alternatively the model can be elaborated in a different way, the better to visualize how it relates to village life, yet showing how the core of a village may very possibly be a sibling pair. Imagine two residence groups (figure 6.1.), which Heinen considers to be the minimum constituents of a village (1983–1984a, 265–66). Each would be residing in a different segment of the perimeter ring of the house. One resident group could be composed of ego, his wife, and her mother and father. And the other group would consist of ego’s sister, ego’s and his sister’s mother and father, and ego’s sister’s husband, who, for instance, could be a man marrying into the village or ego’s wife’s brother. The village core remains the same, the brother and sister, but now placed in the context of two residence groups. 8. Their population has since risen rapidly. For instance, it was estimated to be 3,000, according to the 1982 Indigenous Census (Heinen 1983–1984, 263). 9. Through my own fieldwork among the Curripaco, I know the fear and sense of vulnerability induced by the rubber boom can persist long after the original threat has gone. The Curripaco also tell me that some Baré suffered greatly during the rubber boom and hid in the depths of the forest to escape the rubber traders’ depredations. When they go to certain distant regions, they see signs that the Baré are still there, totally isolated from the outside world, too afraid to show themselves. 10. Because Silva excludes a key piece of data there is some difficulty estimating the percentage of parallel cousin marriages that occurred before 1950 and those that occurred subsequently. Nevertheless, an approximation can be calculated. Silva takes as her total a sample of 143 marriages. In this sample we know that there were 22 parallel cousin marriages pre-1950 and 11 occurred post-1950. However, Silva focuses her analysis on these 143 unions only. But, and here is the problem, we know that these 143 marriages are, in fact, a subset of a larger population that includes 57 marriages in which the names of the eight grandparents are known, but the spouses have no common grandparent. In other words, in this subset there are no cross- or parallel cousin marriages. The total sample is 200 (143 + 57), and to avoid bias, it is this population that I analyze. However, Silva does not indicate how many of these 57 marriages occurred before 1950 and how many occurred after. If we assume they are distributed in the same ratio as the 143

A Discussion of Silva’s “Beyond the Norms” · 123

sample, then an approximation can be calculated. We know that of the 143 marriages, 81 occurred before 1950 and 62 after 1950; that is, they are distributed in the ratio of 3:2 (81:62). If we assume that the 57 marriages are also distributed in the same ratio of 3:2 or 34:23, then approximately 115 marriages (81+34) occurred pre-1950 and 85 (62+23) occurred after 1950. In other words, approximately 19% of marriages were with the parallel cousin pre-1950, and 13% were with parallel cousins post-1950. 11. The Ye’kwana were not alone in this ability to centralize political power effectively and quickly. The Shuar, like the Ye’kwana, were divided into very small settlements over a very large area, with only loose and flexible kin and affinal ties between settlements. Nevertheless, in 1599, seeing the strategic sense of putting on hold the feuding among themselves, they united to throw the Spaniards out of their territory. They chose a war leader from one of their settlements, entrusted themselves to his direction, and then, in a coordinated attack, massacred nearly all the colonists and drove out the rest. They then returned to their houses to carry on their lives much as they had done before (Harner 1984, 20, 77–79).

7 Why Do Women Run Away? Matrimonial Strategies among the Yanomami Catherine Alès

Beyond the formalism reflected by many kinship studies, I proposed to analyze the functioning of a kinship and marriage system associated with a terminology of Dravidian type, yet according to the way the people themselves practice this terminology.1 Within the frame of this system, which seems to correspond to the application of simple terminological rules, the complex and often contradictory connections between different nomenclatures of address and reference, and also between the latter and the system of attitudes, in fact leaves open formal possibilities of kin classification and, consequently, individual and collective choices. That allows for different configurations of positive and negative marriage rules that can be activated depending on the sociopolitical, demographic, or historical context. Such results demonstrate the inadequacy of using only genealogical relationships for the categorization of marriage practices within the context of a Dravidian alliance terminological system. To frame it in another manner, in the same way that Dravidian kinship systems are ego-centered, the corresponding marriage practices should not be interpreted as the application of an exclusive rule but as a set of structural variations, as evidenced by the diversified strategies of the social actors. Parallel to these strategies developed ultimately to increase the number of possible marriageable women for an individual, one of the traits that stands out when one considers the question of marriage among the Yanomami is the frequently high number of conjugal unions contracted by an individual during his lifetime. One might assume that the practice

Why Do Women Run Away? Matrimonial Strategies among the Yanomami · 125

of polygamy fosters this phenomenon, but it is not only that. The number of separations and secondary unions is particularly important. Daily life, just like conflicts within and between communities, often revolves around “love affairs” and can lead to duels or collective battles creating opposition between villages. The marriage practices and marital lives of the Yanomami are especially hectic and fuel social and political dynamics. It would thus seem that wives are much easier to obtain than to keep.

Dravidian Kinship and Marriage Strategies My interests focus on the structural and symbolic methods by which the system of kinship and marriage functions in the Yanomami family, among the Yanowami of Venezuela.2 This system, as in many Amazonian societies, is of a Dravidian type: it opposes the relatives of an ego as consanguines and affines and identifies cross-cousins with allies.3 All persons are distributed into two classes, the “consanguines” and the “affines,” with several corresponding categories based on sex and generation: those of “father,” “mother,” “elder brother” or “younger brother,” “elder sister” or “younger sister,” “son” and “daughter,” for the consanguines; those of “husband,” “wife,” “brother-in law,” “sister-in-law,” “father-in-law” and “mother-in-law,” for the affines.4 Grandparents and grandchildren are designated by means of terms belonging to central generations, which can be either affinal or consanguineal, according to different Yanomami subgroups (see table 7.1).5 We must however distinguish the terminology of reference from that of address; this distinction opens up possibilities for variations of appellations that will be codified and reinforced by the system of attitudes. In this context consanguinity and affinity do not appear as strictly determined genealogically as the reference terminology might indicate. So here I underscore the particularities of applying vocabularies of address and reference—in effect, they may not completely overlap—and the multiple arrangements that could ensue in order to adjust the “preferential” rule with a view to creating allies, in other terms “marriageables”: avoidance of marriage to a first cross-cousin; the option for the affinity relationship when a relationship can be deduced from several combinations; the changing of distant relatives into affines; and the introduction of asymmetries in the categories applied within a set of siblings and between different sets of siblings.

Table 7.1. Terms of reference and address for a female ego and a male ego*

G-2, G-3 G-1

G0> Ego

Consanguines Affines Man / Woman Man / Woman (grandfather / grandmother) aiwë / amiwë aiwë / amiwë (father / mother) (father-in-law / mother-in-law) fayë / nayë shoayë / yesiyë fape / nape shoape / yape (elder brother / elder sister) aiwë / amiwë aiwë / amiwë

(husband / wife) feãrõyë (F) / suwëpiyë (M) (brother-in-law / sister-in-law) heriyë (M) / natofiyë (F) shoriwë (M) / amawë (F) shoriwë (M) / amawë (F) ote (F)

G0