HOLIDAY: The press will be closed from Monday, December 22 through Thursday, January 1.

Vital Enemies

[ Native American Studies ]

Vital Enemies

Slavery, Predation, and the Amerindian Political Economy of Life

By Fernando Santos-Granero

A pioneering study of the enslavement of Amerindians by Amerindians in tropical America, outside the realm of colonial agents.

2009

$24.95$16.72

33% website discount price

Paperback

6 x 9 | 316 pp. | 40 halftones, 7 maps

ISBN: 978-0-292-71913-2

Analyzing slavery and other forms of servitude in six non-state indigenous societies of tropical America at the time of European contact, Vital Enemies offers a fascinating new approach to the study of slavery based on the notion of "political economy of life." Fernando Santos-Granero draws on the earliest available historical sources to provide novel information on Amerindian regimes of servitude, sociologies of submission, and ideologies of capture.

Estimating that captive slaves represented up to 20 percent of the total population and up to 40 percent when combined with other forms of servitude, Santos-Granero argues that native forms of servitude fulfill the modern understandings of slavery, though Amerindian contexts provide crucial distinctions with slavery as it developed in the American South. The Amerindian understanding of life forces as being finite, scarce, unequally distributed, and in constant circulation yields a concept of all living beings as competing for vital energy. The capture of human beings is an extreme manifestation of this understanding, but it marks an important element in the ways Amerindian "captive slavery" was misconstrued by European conquistadors.

Illuminating a cultural facet that has been widely overlooked or miscast for centuries, Vital Enemies makes possible new dialogues regarding hierarchies in the field of native studies, as well as a provocative re-framing of pre- and post-contact America.

  • Acknowledgments
  • Introduction
  • Part 1. Histories of Domination
    • Chapter 1. Capturing Societies
  • Part 2. Regimes of Servitude
    • Chapter 2. Captive Slaves
    • Chapter 3. Servant Groups
    • Chapter 4. Tributary Populations
  • Part 3. Sociologies of Submission
    • Chapter 5. Markers of Servitude
    • Chapter 6. Servile Obligations
    • Chapter 7. Dependent Status
  • Part 4. Ideologies of Capture
    • Chapter 8. Civilizing the Other
    • Chapter 9. Warring against the Other
  • Conclusions
  • Appendix. Assessment of Main Sources
  • Bibliography
  • Index

Browse the book with Google Preview »

On the very first day Columbus landed in America, he registered in his Diario (1991: 67) the first allusion ever made to the existence of captive slavery in native tropical American societies: "I saw some who had marks of wounds on their bodies and I made signs to them asking what they were; and they showed me how people from other islands nearby came there and tried to take them, and how they defended themselves." And he speculated: "I believed and [still] believe that they come here from tierra firme to take them captive." Since that fateful twelfth day of October 1492, wherever European conquistadores set foot in the American tropics, they found evidence of indigenous warfare, war captives, and captive slaves.

As early as 1509, to the north, Núñez de Balboa reported that native lords living in coastal Darien, Panama, were carried on the shoulders of slaves (López de Gómara 1946[1552]: 199). In 1515 another conquistador observed that the peoples to the east of Darien kept slaves—identified by the red and black designs that their masters tattooed on their faces—whose main duties were to extract gold, to perform agricultural tasks, and to do other menial services (López de Gómara 1946[1552]: 278). In the early sixteenth-century expeditions undertaken along the Paraguay River, to the south, the Spaniards found a great deal of interethnic warfare associated with the taking of enemy heads as war trophies, the cannibalistic consumption of war prisoners, the trading of captives for gold and silver objects, and the classification of entire populations as "enemy/slaves" (Schmidl 1749[1539]: 19; Martínez de Irala 1912[1542]: 347, 349, 352; Núñez Cabeza de Vaca 1585[1555]: 125v).

On the eastern slopes of the Bolivian Andes, to the west, highland informants told Garcilaso de la Vega (1963[1609]: 322) that since the reign of Tupac Inca Yupanqui, several decades before the Spanish conquest of Peru in 1532, the lowland Chiriguaná groups had been raiding the frontiers of the Inka empire to take captives. Spanish conquistadores told the Court historian Pietro Martire d'Anghiera (1966[1516]: 38v-39r), as early as 1516, that the Caribs of the Lower Orinoco River, to the east, undertook large maritime war expeditions to the Gulf of Paria to procure slaves for themselves from among the Arawak. Conversely, Martín López (1964[1550]: 45-46), one of the first Spaniards to explore the Orinoco River, wrote that the Arawak of the Orinoco delta carried out annual expeditions against the southern Caribs, killing everyone except children, whom they took as captives.

The few references mentioned above allude to the salient characteristics underlying native tropical American slavery: regular raiding of weaker peoples by powerful neighbors; classification of entire populations as enemy/slaves; taking of captives, particularly women and children; inscription of slave status through body marks; assignment of captives to household units; performance of menial services; and trading of war captives in exchange for luxury goods. Early chronicles also flag the existence of other forms of servitude—namely, servant groups and tributary populations. It is therefore puzzling that the topic of native forms of slavery in tropical America has never been directly addressed, despite the existence of an extensive and rich literature on the causes of warfare in the region (Chagnon 1968; Carneiro 1970, 1990, 1994; Lathrap 1970; Meggers 1971; Harris 1974, 1979, 1984; Gross 1975; Harner 1977; Ferguson 1990, 1995, 2000; Whitehead 1992; Kelekna 1998; Valentine and Julien 2003; Chacon and Mendoza 2007).

Reasons for this neglect may be multiple. The most important, and at the same time the most banal, has to do with the historical conditions of tropical American indigenous societies in the late 1930s, at the time when modern anthropologists (e.g., Claude Lévi-Strauss, Irving Goldman, Charles Wagley, and Allan Holmberg, to name but a few) began doing fieldwork in the American tropics. By then, native tropical American peoples had withstood centuries of foreign diseases, encroachment, displacement, genocidal policies, enslavement, and marginalization. They were thus but the stubborn remnants of their former selves. Even isolated peoples who were thought to have escaped the horrors inflicted by European agents were subsequently discovered to be regressive survivors of such process, experienced in a more or less remote past (see Holmberg 1969; Stearman 1989; Balée 1992).

The extant social organization of these peoples—reduced, scattered, culturally impoverished, and having endured forced acculturation, transfiguration, regression, and ethnogenesis—was only a shadow of earlier times. However, even anthropologists such as Lévi-Strauss in his Tristes Tropiques ([1955] 1974), who described the terrible consequences of the clash between Amerindian and European peoples, were unaware of the extent to which native tropical American peoples had undergone a fundamental transformation. Only much later would Lévi-Strauss acknowledge: "There, where we believed to have found the last evidence of archaic lifeways and modes of thought, we now recognize the survivors of complex and powerful societies, engaged in a historical process for millennia, and which have become disintegrated in the lapse of two or three centuries, a tragic accident, itself historical, which the discovery of the New World was for them" (1993: 9).

Powerful paramount chiefs, regional confederations, large political centers, elaborate temple ceremonies, extensive public earthworks, and native forms of servitude had ceased to exist by the time these scholars were writing. Consequently, features that had been corroborated by abundant archaeological and historical evidence were ignored or were simply disregarded as exaggerations made by zealous European adventurers eager to impress their royal patrons. This tendency was further reinforced by the structural-functionalist paradigm in vogue at the time, which stressed structure, homeostasis, and synchronicity over process, change, and history. The fundamentally altered lifeways of twentieth-century native tropical American peoples were thus essentialized, envisioned to be "typical" of these societies since times immemorial. This image has become so entrenched that, as Heckenberger (2003: 29) has rightly argued, "many feel that the very notion of native cities, regional bureaucracies, kings, priests, slaves, and the like, is untenable."

In the last decade, a few anthropologists and archaeologists (e.g., Dreyfus 1983-1984; Descola 1988; Whitehead 1993, 2003a; Roosevelt 1993, 1994; Heckenberger 2003, 2005) have insisted on the need to introduce the historical perspective of longue durée, providing a richer understanding of tropical American indigenous societies. This has slowly led to a "temporal revolution," by which it has become increasingly accepted that native tropical American societies have changed significantly since the beginning of the Christian era and particularly in the past five centuries (Fausto and Heckenberger 2007: 1; also Hill 1993: 10, 1996: 4). Even those authors who accept that these societies have experienced radical changes in the past five centuries, however, are reluctant to accept the existence of native forms of slavery. They argue, not without reason, that the terms "slave" and "slavery," as used by early European chroniclers, belong to a different semantic universe and thus distort Amerindian social practices (Whitehead 1988, 1990, 1999; Langebaek 1992; A. C. Taylor 1999; Wright 2002). Whereas they may accept that raiding, the taking of war captives, and the incorporation of captives in a more or less subordinate position were widespread practices in precolonial times, they deny that these can in any way be compared to slavery as it is understood in the Western tradition (Steward and Faron 1959: 187; Whitehead 1988: 181; Langebaek 1992: 145). In their view, Amerindian slavery became such only when the demand for slaves in the colonial market forced indigenous peoples to intensify raiding, turning war captives into commodities to be traded for European goods. Native slavery in the American tropics, they conclude, is not an indigenous pre-Columbian pattern but rather a colonial product.

The dismissal of native forms of slavery seems to stem from a certain moral reluctance to discuss an institution of which few traces remain, one that does not fit with the highly egalitarian ideologies and practices of present-day native tropical American societies, and that projects an image of "savagery" that could be (mis)used to deny Amerindians their rights as autonomous peoples. This bias, I would suggest, is also linked to a generalized and persistent perception of slavery as the institution developed in the plantations of the southern United States (Meillasoux 1975: 12), a sort of "ideal type" against which to measure other forms of what Condominas (1998) has called "relations of extreme dependence."

Following this line of argument, scholars have identified five elements that, because they are absent from preconquest native tropical American peoples, prevent us from characterizing war captives as slaves. First, captives were eventually incorporated through marriage or adoption into the families of their captors. Second, captive labor did not free their masters from their productive obligations and was not crucial to the reproduction of their economic system. Third, slaves were not subjected to systematic exploitation and were generally well treated. Fourth, they were not considered to be property, and thus could not be bought, sold, or traded as chattel. Lastly, their status was not hereditary and, hence, they did not constitute a permanent social class. Indeed, some authors suggest that because certain native terminologies liken the status of so-called slaves to that of potential affines, the Amerindian institution of war captives had a kinship dimension that was alien to slavery as it was practiced by contemporary Europeans.

Whereas I agree that the use of the highly charged term "slave" in the context of native tropical American societies might be misleading, I firmly believe that we cannot content ourselves with negative characterizations, explanations of what Amerindian "slaves" were not, rather than explanations of what they were. This, in short, is the main objective of this book: to describe a widespread native institution, and to discuss its importance from a native point of view.

Slavery is only one among various native forms of servitude, often combined in diverse permutations. The focus of this book will therefore be on what I have called native regimes of capture and servitude: the set of patterned relations of extreme dependence that are found within a particular regional system of interethnic power relations at any given juncture. Native regimes of servitude are regional and interethnic because they always involve more than one settlement and more than one people. And they are inscribed in a system of power relationships, for the settlements and peoples involved are engaged in a variety of power struggles, alliances, and associations. These do not respond to random forces—as in a Hobbesian kind of war of "every man against every man"—but are framed by preexisting interethnic hierarchies, balance of forces, and trajectories of interaction. Native regimes of servitude are an ongoing historical product and are thus in permanent flux. This quality of dynamic transformation will become apparent when I analyze the changes experienced by these regimes as the result of the colonial encounter.

Following in the steps of recent scholars of slavery in the Old World (Condominas 1998; Villasante-de Beauvais 2000), I have tried to avoid the use of terms that correspond to other historical realities when discussing Amerindian forms of servitude, favoring instead more historically neutral terms—thus, "tributaries" instead of "vassals" or "serfs"; "commoners" instead of "plebs"; "servant groups" instead of "helots." In other cases, however, such as with the use of the term "slavery," its replacement by other, more general terms, such as "extreme dependence," would have been more confusing than clarifying. To use native terms, as advocated by some scholars, would have been extremely useful if this were a study of a single society. But since it is a comparative work dealing with six different Amerindian peoples, this option would have been more cumbersome than helpful. For these reasons, I have opted to use the term "captive slavery" instead of "slavery," a solution that has the advantage of emphasizing that slavery in tropical America was always the result of warfare and piracy. Likewise, instead of "slave," I have opted to use the terms "captive" and "captive slave," which, as we shall see, are more faithful to the meaning of the native terms used to refer to this kind of dependent.

The present book focuses on the different forms and regimes of servitude existing in tropical America before the arrival of Europeans. Emphasis is thus placed on tropical American indigenous societies at the time of contact. Two problems for those studying the history of Amerindian societies ensue from this perspective: first, what should we consider to be the time of contact? and, second, what should we understand by "contact"? When considering America as a whole, the time of contact is taken to be the time at which Columbus first landed in America. At a more regional level, however, the time of actual contact between Amerindian populations and Europeans varied widely throughout the Americas, depending on the whims of European adventurers and the vagaries of colonial history. Thus, the time of contact is not a precise point in time; rather, it is a long period characterized by multiple, intermittent, and temporally variable phases of contact, culminating in the conquest of the contacted native peoples and the settlement of their lands. In this book, I use the notion of "time of contact" precisely in this way, to signify the period in which a given indigenous society came in contact with Europeans but still retained its political autonomy.

This brings me to the second caveat. It has long been accepted that Amerindian populations experienced the effects of European presence even prior to actually meeting any Europeans. If European agents had an impact even when they were not physically present, how can we be sure that what was observed at the time of contact was indeed how things were in pre-Columbian times? Many scholars would simply assume that we can never be sure, and by so doing, they dismiss all possibilities of knowing how these societies operated in pre-Columbian times. Such as stance is misleading insofar as it supposes a radical break between an essentialized, authentic past and a present reality contaminated by European influence. The impact of Europeans was neither instantaneous nor absolute. It often involved long periods of struggle, accommodation, and negotiation. Changes were not always as rapid and radical as to render impossible the task of identifying native forms of organization and cultural patterns. In fact, from the perspective of the longue durée, certain "structures"—environmental conditions, productive strategies, sets of social relations, cultural practices—persisted for long periods of time, changing only very slowly (Braudel 1958). To assume that the European conquest of America changed the history of Amerindian peoples forever, dividing it into an authentic past and a somewhat spurious present, is to project our own Eurocentric biases. Without in any way underrating the dire consequences of the European conquest, one could legitimately argue that for many Amerindian peoples the expansion of the Huari, Aztec, and Inka empires was equally cataclysmic. Such dramatic events are in the nature of historical processes, but they are in no way more "historical" than other, less radical changes.

If we accept that history is an ongoing process rather than an intermittent succession of dramatic events interspersed by periods of historical paucity, then the statement that it is not possible to know how Amerindian societies were organized in pre-Columbian times is spurious. As Sahlins (2005: 5) has aptly put it: "Whatever the compulsions of the global forces, whatever may be inherent in foreign things or propagated with them, the people are not simply determined by them, since they also bring their own understandings to bear upon the encounter." At most, it could be said that native relations of extreme dependence were analogous to those brought in by the European invaders. Analogies between the cultural baggage of the conquered and the conquerors were not uncommon at the symbolic and discursive levels, as Whitehead (2003b: x) has suggested. Analogous perceptions such as these led to a mutual recognition of resemblances between their respective forms of servitude and facilitated the passage from one to the other. As Myers (1990: 121) observed: "To the degree that change had taken place, it was an aboriginal accommodation to changing conditions rather than change based upon a European model." In other words, it was an indigenization of European practices rather than the Europeanization of Amerindian practices (Sahlins 2005: 6).

Furthermore, as Lovejoy (2000[1983]: 21) has noted for Africa, Europeans were able to integrate Amerindian peoples into the colonial slave trade only "because indigenous forms of dependency allowed the transfer of people from one social group to the other." Given that native forms of extreme dependency coexisted for long periods with slave raiding that was oriented to supply the colonial markets, the contact situation provides an unrivaled opportunity to test the extent to which native models of capture and servitude differed from European imports (see Whitehead, forthcoming: 3-4, for a similar argument). The modifications that native and foreign forms of servitude underwent when adopted by the Other, including the conflicts that arose from such (mis)appropriations, do not represent a "contamination" of pristine pre-Columbian institutions. Rather, they are the best indicators of what specific forms indigenous practices of servitude assumed.

Thus, this book is an exercise in reconstructive history, a first attempt to identify the constitutive elements of relations of extreme dependency in the American tropics, by which I understand not only the hot lowlands between the Tropics of Cancer and Capricorn but also those areas outside the tropics yet close enough to them to have tropical climates. This vast region stretches from southern Florida in the north, to the Grand Chaco in the south, and from the piedmont of the Andes and the Central American mountain ranges in the east, to the Antilles and the coasts of Guiana and Brazil in the west. The bulk of the study is constituted by the analysis of six indigenous societies dispersed throughout this region, to wit: the Calusa of southern Florida, the Kalinago of Lesser Antilles, the Tukano of northwest Amazonia, the Conibo of eastern Peru, the Chiriguaná of eastern Bolivia, and the Guaicurú of the Grand Chaco (Map I.1). These societies were all composed by a variety of regional groups that shared a similar cultural matrix, and were sometimes, though not always, linked by common linguistic affiliation. They had blurred boundaries and were far from being culturally homogenous. None of these polities had the kinds of institutions that underpinned state formations such as those of the Maya, Aztec, and Inka. But neither were they egalitarian societies, adamant on questions of personal autonomy and averse to vertical forms of authority like those characteristic of present-day tropical America. More importantly, they all had developed, or were in the process of developing, supralocal forms of authority, often under the guise of hereditary paramount or regional chiefs. The degree of power and authority wielded by these supralocal leaders varied widely, however.

Subsistence patterns and other forms of livelihood also varied. The sea-oriented Kalinago and the river-oriented Tukano and Conibo were conspicuous horticulturalists who complemented gardening with fishing and, to a lesser extent, hunting. In contrast, the Calusa and Guaicurú did not practice agriculture; their livelihood depended largely on fishing and hunting, and on the agricultural produce they exacted from tributary populations. The Chiriguaná stand as an intermediate case: horticulturalists with a strong emphasis on hunting, gathering, and, to a much lesser extent, fishing.

Two common denominators distinguish this set of societies—namely, being adapted to tropical environments, and having slaveholding institutions. It should be emphasized that this sample is not representative of the indigenous societies of the American tropics in general. Kept out are those indigenous peoples—similar to the present-day Txicão, Matis, and Parakanã (Menget 1988; Erikson 1986; Fausto 2001)—who took captives from their enemies, incorporating them as full "citizens" through marriage and adoption instead of holding them as slaves. It also excludes Amerindian societies like the ancient Tupinambá of the Brazilian littoral, who practiced intratribal raiding and the taking of captives for ritual purposes rather than to obtain wives or slaves (Carneiro da Cunha and Viveiros de Castro 1985). And, obviously, it also excludes those societies that did not practice any form of slave raiding. It is essential for the reader to keep these facts in mind. Nothing is more alien to my intention than to contend that all indigenous tropical American societies were engaged in slaveholding, or in other forms of servitude, at the time of contact.

For my purposes in this analysis, and in the interest of making tighter comparisons, I have excluded Amerindian slaveholding societies from other geographical regions of America, such as those from the northwest Pacific coast, greater Lower Columbia, Great Plains, southwest borderlands, and southeast Atlantic coast of North America. I am persuaded, however, that the forms of servitude observed in these societies were very similar to those found in the American tropics (see, for instance, MacLeod 1928; Mitchell 1984; Perdue 1993; Ruby and Brown 1993; Donald 1997; Brooks 2002; Minges 2003; Hajda 2005). And I am convinced that most of the conclusions I arrived at in this study could just as well apply to these similar nonstate societies.

When selecting the societies to include in this study, I took into consideration four criteria. I favored societies that were contacted in the fifteenth and sixteenth centuries over those contacted in later times, to avoid the risk of including societies possibly altered in fundamental ways by the indirect influence of the colonial experience playing itself over long periods of time. I chose societies for which there was abundant historical documentation so as to be able to gauge the reliability of the historical evidence, using multiple independent sources. I also favored societies that have been the object of modern ethnographic studies, over those that have not, to guarantee access to cultural information that could throw light on little-documented ancient practices. Lastly, I made an effort to include societies from as broad a range of tropical areas as possible, to ensure greater representativity. Not all the societies in the sample comply with the aforementioned four prerequisites, however; they nevertheless all comply with at least three of them. Although subject to improvement, the resulting sample, I believe, represents faithfully the diverse patterns of servility found in native tropical America. However, some aspects of the sample can profit from further clarification.

If one looks at Map I.1, it becomes apparent that societies from Central America and the central regions of tropical South America are excluded from the sample. Indeed, the societies that were included are all located in what could be considered the borderlands of the American tropics. There are solid reasons for this choice. The Amerindian societies located along the Pacific coast of Central America and along the Amazon River and its larger tributaries were among the first to disappear or radically change as a consequence of the spread of European epidemics through long-distance native trading networks and, later on, of the ravages of European slaving expeditions. Coupled with the fact that information from earlier ages was too scant to permit an in-depth analysis, their altered status excluded them as good candidates to incorporate in the sample. In contrast, societies for which there is more abundant historical information were located in what has been dubbed the "tribal zone"—that is, "the area affected by the proximity of a state, but not under state administration" (Ferguson and Whitehead 1992: 3). The Calusa and Kalinago, who were fought over by Spaniards, Frenchmen, Englishmen, and Dutchmen; the Tukano and Guaicurú, who were disputed by Spaniards and Portuguese; and the Conibo and Chiriguaná, who were engaged in a constant struggle against the Spanish colonial empire, were all in this kind of situation.

Location in the interstices or peripheries of the colonial space allowed these societies to resist European domination, retaining their political autonomy for longer periods of time. Taking advantage of the rivalry between European powers, they negotiated better conditions for themselves by pitting one colonial state against the other, changing sides whenever it suited their interests. More importantly, because of their long-term resistance, these societies became of great interest to colonial authorities, who wrote many more reports and other works about them than they did about societies that rapidly succumbed to colonial rule. It has been argued that these societies are a creation of the states against which they were pitted and, thus, that such features as supralocal forms of authority, intensive warfare, and nascent forms of servitude were the result of colonial relations (Ferguson and Whitehead 1992). As we shall see, however, these features were already present in these societies at the time of contact and cannot be attributed to European presence, however indirect. If captive slavery appears as a salient social feature in these societies, it is not because they were tribal creations—which they were—but rather because, having resisted European domination for longer periods, we know more about them than we do about societies that were exterminated in the years immediately after contact. In short, if there is a bias in the sample, it is not due to methodological oversight. Rather, it is the result of a historical situation stemming from unique geopolitical conditions.

Honoring my definition of the notion of time of contact, the sources consulted not only were the earliest available in each case but also were written throughout the period in which the societies of the sample still retained their autonomy. This meant that in some cases the sources to be consulted covered a period of two or even three centuries. To make the amount of literature I surveyed more manageable, I confined the documentary research to published documents. The major task of collecting historical information was carried out at the Library of Congress—especially in the Rare Book and Special Collections Reading Room—during a period of two years (August 2001-July 2003). I reviewed a wide range of documentary sources by all sorts of Spanish, French, Portuguese, Italian, and English writers. Among this mass of records, I conferred priority to documents produced by persons who actually witnessed the events they recounted—conquistadores, missionaries, and explorers reporting about their own experiences—or by lay and religious historians who relied on oral or written accounts obtained directly from eyewitnesses. I excluded most secondary sources—for example, the works of nineteenth- and twentieth-century historians—except when their accounts were based on materials no longer available to present-day scholars. To help readers keep track of the period to which these old sources correspond, I have included in the references that appear in the text not only the year of publication of the edition consulted but also (in square brackets) the date in which the author witnessed or reported the events described. All translations of quotes in languages other than English are mine.

That European sources are prejudiced and biased, and thus totally unreliable concerning knowledge about the history of Amerindian societies, is a frequently voiced remark. Such statements ignore the fact that, as Whitehead (2003b: ix) has pointed out, there is no such thing as an "unprejudiced" or "unbiased" historical source. Historical sources are always "situated"; that is, they are firmly anchored in the place and time they were written and, above all, reflect the personal trajectories and agendas of their authors. Denying the possibility of learning about the history of Amerindian societies using European sources would be tantamount to denying the possibility of knowing the history of any people through any kind of source. This is not to say that we should accept at face value all the information provided by these accounts or that we should espouse the notion that there is only one history. Being aware of the historical juncture in which a document was written, and of the political inclinations or personal agenda of its author, is crucial to the detection of potential biases or downright fabrications. To avoid using unconfirmed or dubious data, I have discarded data that are not verified by at least two independent sources. Whenever I make use of unconfirmed but otherwise suggestive data, I indicate it in the text. Additionally, in the Appendix, I discuss the reliability of thirty-eight of the most frequently cited authors. These methodological procedures have contributed to a better evaluation of the sources and, I hope, to a better depiction and interpretation of Amerindian servile institutions.

Key to the central agenda of historical anthropology is the incorporation of indigenous understandings, ontologies, cosmologies, and historicities to capture native perspectives on historical events and social practices. Insofar as possible, I have strived to incorporate indigenous viewpoints in order to overcome the apparent resemblances between analogous Amerindian and European forms of servitude and identify the unique features of indigenous institutions; a task easier said than done. The ethnocentricity of European colonial agents left little mental room to appreciate, let alone take into consideration, indigenous perceptions. Native points of view have nevertheless been preserved in four types of information that caught the attention of Europeans, because of either their exotic appeal or their pragmatic value: myths, shamanic beliefs, political discourses, and language. I have made use of all these types of data whenever possible, especially of early colonial dictionaries, which proved to be an invaluable source of information on how Amerindians conceptualized what chroniclers labeled as "slaves" and "servants." More importantly, I have avoided "etic" explanations of Amerindian warfare and slavery, in favor of an "emic" approach that relies heavily on Amerindian notions and conceptions (see below).

Following Rivière's call (1993) for the "amerindianization" of key anthropological concepts as a means of achieving a better understanding of Amerindian socialities, I have opted not to begin this book with a definition of slavery as a means to assess whether Amerindian servile practices and institutions conform to it—a top-down kind of strategy. Instead, I have adopted a bottom-up approach by first presenting a brief account of the histories of domination of the capturing societies selected for discussion, emphasizing the regional systems of interethnic power relations of which they were part (Chapter 1). This is followed by a critical examination of the three different regimes of capture and servitude that I have identified as having existed in the region. Captive slavery was practiced by all the societies considered in this study. But whereas among the Kalinago and Conibo it was the main and only form of extreme dependency (Chapter 2), among the Tukano and Chiriguaná it was combined with the subjugation and attachment of servant groups (Chapter 3). And among the Calusa and Guaicurú it coexisted with the subjection of neighboring peoples as tributary populations (Chapter 4). I then explore the sociologies of submission that were characteristic of these societies, focusing on the five aspects that have been singled out in objections to the existence of slavery in native tropical America to see whether they are supported by the historical data (Chapters 5-7). Once the degree to which these objections may be valid has been determined, I explore the ideologies of capture on which Amerindian regimes of servitude are based. These ideologies, equating war captives to prey and captive slaves to pets, facilitated the integration and, later on, assimilation of captive slaves (Chapter 8). Such ideologies are sustained by Amerindian eco-cosmologies that posit that the vital cosmic force that energizes all living beings in the world is finite and scarce, a notion that has given rise to what I call the Amerindian political economy of life. I argue that Amerindian warfare in general, and captive slavery in particular, can be understood only within this ideological framework (Chapter 9). Only then do I discuss—in the light of extant theories of slavery and other relations of extreme dependence—whether native tropical American forms of servitude conform to those practices that in other times and places have been labeled "slavery" (Conclusions).

Here, however, I would like to begin from the end. The outright conclusion of this study is that native slavery in tropical America was not a colonial product. Rather, it predated the European presence even though its character differed in several respects from slavery as it was practiced in other historical periods and geographical areas. The uniqueness of native tropical American forms of slavery and servitude resides in their being an extreme expression of the Amerindian political economy of life; a political economy based on the widespread notion that vital energy is finite, generally fixed, scarce, unequally distributed, and in constant circulation. In such an economy, humans, animals, and other beings are conceived of as competing among themselves to accumulate as many potentialities of life as possible through mutual predation and capture. Potentialities of life wrested from enemy Others can assume various forms. In the context of interhuman competition, they adopt the shape of actual persons—generally young women and children of both sexes—or of life forces contained in bodily trophies, magical objects, ritual paraphernalia, and sacred effigies. The logic of Amerindian warfare, I would suggest, lies in this complex of ideas. However, it should be noted that even though all native tropical American societies can be characterized as "capturing societies" engaged in a ruthless competition for potentialities of life, not all capturing societies can be characterized as slaveholding societies. This book seeks to understand Amerindian forms of extreme dependence in societies that practiced captive slavery.

Fernando Santos-Granero is a staff researcher at the Smithsonian Tropical Research Institute in Panama and has worked with the Yanesha of the Peruvian Amazon for three decades.