Jump to ContentJump to Main Navigation
Gender in Amazonia and MelanesiaAn Exploration of the Comparative Method$

Thomas Gregor and Donald Tuzin

Print publication date: 2001

Print ISBN-13: 9780520228511

Published to California Scholarship Online: May 2012

DOI: 10.1525/california/9780520228511.001.0001

Show Summary Details
Page of

PRINTED FROM CALIFORNIA SCHOLARSHIP ONLINE (www.california.universitypressscholarship.com). (c) Copyright University of California Press, 2018. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in CALSO for personal use (for details see www.california.universitypressscholarship.com/page/535/privacy-policy-and-legal-notice). Subscriber: null; date: 17 January 2019

Comparing Gender in Amazonia and Melanesia: A Theoretical Orientation

Comparing Gender in Amazonia and Melanesia: A Theoretical Orientation

(p.1) One Comparing Gender in Amazonia and Melanesia: A Theoretical Orientation
Gender in Amazonia and Melanesia

Thomas A. Gregor

Donald Tuzin

University of California Press

Abstract and Keywords

This chapter explores the sources and theoretical implications of remarkable similarities between societies in Amazonia and Melanesia, a comparison which combines the universalist and localist traditions and is the base of anthropological studies. It establishes a common discourse among scholars working with different cultures, stimulates new perspectives on findings from particular cultures, and allows searching for general principles. The chapter suggests that gender is of great importance in the context of many of the small-scale cultures of Melanesia and Amazonia, i.e. the societies of Amazonia and Melanesia are gender inflected. Further, in both Amazonia and Melanesia, the self-concept, social identity, and the anatomy and physiology of the human body are intertwined with theories of conception, maturation, depletion, and death. The comparison also indicates the basis of procreative symbolism in both the cultures. This comparison affords an exceptional opportunity to explore fundamental questions about the conceptualization and examination of the human condition.

Keywords:   Amazonia, Melanesia, gender, procreative symbolism, cultures, anthropological studies


Approximately one hundred years ago anthropologists identified what was to become an intriguing, enduring mystery of culture history: the question of the sources and the theoretical implications of remarkable similarities between societies in Amazonia and Melanesia. A world apart and separated by forty thousand or more years of human history, some of the cultures in the two regions nonetheless bore striking resemblances to one another. In both Amazonia and Melanesia, the ethnographers of the period found societies organized around men's houses. There the men conducted secret rituals of initiation and procreation, excluded the women, and punished those who would violate the cult with gang rape or death. In both regions, the men told similar myths that explained the origins of the cults and gender separation. The resemblances were such as to convince anthropologists of the day, including Robert Lowie, Heinrich Schurtz, and Hutton Webster, that they could only have come about through diffusion. Lowie flatly declared that men's cults are “an ethnographical feature originating in a single center, and thence transmitted to other regions” (1920, 313).

The diffusionist school of anthropology waned soon afterward, and for a long period so did interest in the puzzling resemblances of specific societies in the two regions. Nonetheless, during this period anthropologists continued informally to remark upon the similarities in regions that were separated by such a vast gulf of history and geography. The parallels included not only men's cults but also similar systems of ecological adjustment; egalitarian social organization; flexibility in local- and descent-group composition and recruitment; endemic warfare; similar religious, mythological, and cosmological systems; and similar beliefs relating to the body, procreation, and the self. So (p.2) striking were such resemblances that, at times, as we review them today, they overshadow obvious differences between the regions and obscure the fact that not all subregions or societies lend themselves to this comparison.

In September 1996, in Mijas, Spain, the Wenner-Gren Foundation for Anthropological Research convened a week-long international symposium to examine similarities and differences between Amazonian and Melanesian cultures. This volume is based on that inquiry. Our intention is to examine some of the many dimensions of the Amazonia-Melanesia comparison by focusing on gender, which is perhaps the most noticeable area of resemblance between cultures of the two regions. The authors, all accomplished Amazonianists and Melanesianists, contribute to this goal by examining specific questions that derive from their own research and, above all, by comparing their findings with those drawn from societies in the other cultural region. That is to say, comparativism informed the collective enterprise and its component studies from their inception. This feature of the project distinguishes it from anthropology's long tradition of edited volumes in which, for the most part, contributors describe and theorize about a given topic in a particular ethnographic context, with little regard to cross-cultural comparison per se (e.g., Lambek and Strathern 1998; but see Strathern and Stewart 1998 in the same volume). By contrast, through argument or example, each author in the present volume examines larger theoretical issues of method and epistemology that are suggested by Amazonia-Melanesia comparisons.

The Comparative Method and its Predicaments

Our focus on Amazonia and Melanesia is explicitly comparative. But at its most basic level, all anthropology is comparative. There is no way to talk about other cultures and their institutions without, at least implicitly, comparing them to other cultures. When we speak of a society as having “men's cults,” for example, we have in mind similar organizations in other societies. For all its centrality to anthropology's disciplinary identity, however, the comparative method has had a troubled history. A brief review of that history situates the present volume within the debate and provides a broader theoretical context for the specific comparisons that are the substance of our project.

Let us begin with the influential essay “The Limitations of the Comparative Method of Anthropology,” published by Franz Boas (1940 [1896]) exactly one hundred years before our conference. A watershed in American anthropology, the work defined what was wrong with the “comparative method” of the preceding era and set the course for what anthropology was to be, all the way to the present. American anthropology entered a new age with Boas's essay, but it was to be an age of turbulence. Through it all eddied the dilemma of anthropology's perennial, century-long “crisis” regarding the comparative method (Barnes 1987, 119).

(p.3) Boas's famous complaint with the so-called comparative method was that its Victorian practitioners assumed, without warrant, that in human affairs like effects spring from like causes, and that the occurrence of similar traits in different cultures implies participation in a universal, orthogenic, evolutionary process. Not so, wrote Boas. Rather, similar phenomena can arise from quite dissimilar processes: “The identical result may have been reached on … different lines of development and from an infinite number of starting points” (1940 [1896], 274). Without clear delineation of the underlying processes, without knowledge of cultural and historical particularities, comparison between cultures is impossible—except, perhaps, when the cultures being compared possess known historical affinities and occur within the same geographical vicinity.

In other words, Boas did not object to comparison per se but to comparison that is uncontrolled, uninformed, preconceived, and prejudicial. Certain laws do exist, he averred (p. 276), “which govern the growth of human culture, and it is our endeavor to discover these laws.” Ultimately, a mature anthropology would have to be a blend of history and science, a harmony of the particular and the general; but “for the moment,” priority had to be given to “the study of specific sequences in delimited areas” (Harris 1968, 259). Boas concluded his essay with a clear message for our discipline (1940 [1896], 280). “The comparative method,” he charged, “notwithstanding all that has been said and written in its praise, has been remarkably barren of definite results, and I believe it will not become fruitful until we renounce the vain endeavor to construct a uniform systematic history of the evolution of culture, and until we begin to make our comparisons on the broader and sounder basis [of documented cultural-historical processes]. Up to this time we have too much reveled in more or less ingenious vagaries. The solid work is still all before us” (emphasis added).

In the century since this was written, much of that “solid work” has been done, in the form of fine-grained ethnographic description and analysis. And yet, although good ethnography necessarily entails classification, and hence comparative assumptions, the specific enterprise of comparison remains shadowy and even a bit disreputable, and its results, according to Evans-Pritchard (1965, 27), are meager and controversial. “When anthropologists generalize, they do so on the basis of cross-cultural comparison, but the rationale of their use of comparative data seldom bears close examination” (Leach 1968, 344).

Living With the Predicament

Generally speaking, there have been two types of comparison. The first is broad or even universal in scope, such as one finds in the works of J. G. Frazer, Sigmund Freud, Ruth Benedict, Claude Lévi-Strauss, George Peter Murdock, (p.4) Leslie White, and their respective followers. Whether idealist or materialist, these approaches seek to account for traits and processes common to all humanity, or they appeal to an aspect of human nature (e.g., psychic unity of mankind) to justify a global sample in the testing of very general propositions. Although this type of comparison pursues some of the most interesting questions of all, its practitioners sometimes draw fire for their imperial attitude toward cultural evidence: decontextualizing, overselecting, teleologizing, and taking a generally dogmatic view of things. More technically, as Evans-Pritchard put it in a still vital essay (1965), broad comparisons fail to ensure that “the units of comparison” are of equivalent value. “Are ‘monogamy’ among the Veddahs of Ceylon and ‘monogamy’ in Western Europe units of the same kind?” he asked (1965, 19). If not, how can we compare them?

In a more recent article, Hobart argues that what passes for comparison is woefully simplistic: that the objects compared are prejudged as to their similarity and that there are no standards for declaring them equivalent. He makes the claim that even what appear as cultural universals are, when examined closely, culturally unique. Hence he asks: “Everywhere animals and people eat. Is this not a universal which underwrites all translation?” (1987, 39). It is not, says Hobart—who then proceeds to show that there are at least eight terms for “eating” in Balinese, and that the terms and the concept of eating vary with caste, politeness, personal health, whether the eater is known or a stranger, and so forth (1987, 39).

To an extent, this critical view of comparison rests on a concept of culture in which institutions and symbols are inseparable from their context, or at least difficult to tease apart. For example, in Patterns of Culture, Ruth Benedict famously argued that the diversity of culture results from “a complex interweaving of cultural traits. The final form of any traditional institution … depends upon the way in which the trait has merged with other traits from different fields of experience” (1959 [1934], 37). She went on to examine the guardian spirit complex in North America and concluded that when looked at closely its forms varied so much from one another that there was no “single” guardian spirit complex. Rather, the complex took different forms “according to the other traits of the culture with which it was most closely associated” (p. 39). How could one isolate comparable units if they were so embedded in local traditions? Benedict's resolution of this dilemma was to compare not single cultural traits but deeper essences or patterns (such as the “Dionysian” and the “Apollonian”) that were said to be characteristic of cultures and cultural regions.

For Boas and Evans-Pritchard, the solution to the dilemma of comparable units was to limit the scope of comparison to a small number of well-studied, mutually related cultures in a local region. Its success, while perhaps modest and seemingly parochial to the region (e.g., Nadel 1952; Eggan 1950; Watson 1963; but see Tuzin n.d.), is due to methodological controls over the manipulation (p.5) of variables and constants—controls made possible by historical and geographical proximities. And yet, by its very controls this method begs the question that has vexed anthropology from the start: how does one account for similarities between cultures that are historically and geographically unrelated?

In recent years, the plight of comparison has worsened. Poststructuralist critics now challenge the ontological validity of ethnographic findings and the very possibility of positive knowledge in the human sciences (Clifford and Marcus 1986; Hobart 1987; Holy 1987). If cultures are islands unto themselves or “texts” composed in the imaginations of pseudo-observers, if all classification and generalization are nothing but the exercise of Western hegemony and arrogance, if, in short, all is vanity, then comparison would be at best impossible and at worst immoral. This nihilistic rhetoric, though extreme, is not entirely new. As sober and reasonable an anthropologist as Evans-Pritchard questioned whether cultures could be compared at all, or even if they could, whether there was anything to be learned that was not fatuous or tautological. Should we not question, he wondered, “whether the basic assumption which has so long been taken for granted, that there are any sociological laws of the kind sought; whether social facts, besides being remarkably complex, are not so totally different from those studied by the inorganic and organic sciences that neither the comparative method nor any other is likely to lead to the formulation of generalizations comparable to the laws of those sciences” (1965, 33).

Despite this quote, which catches Evans-Pritchard at his most querulous, few anthropologists would be willing to carry the argument, especially the position of the postmodernists, to its logical conclusion of radical cultural relativism. To do so would mean the end of anthropology. It would even mean the end of ethnography; for, pace Holy (1987, 3), who exaggerates the distinction between “description” and “generalization,” all description employs classes of ideas external to the phenomena, beginning with the words we use. It is an inductivist fallacy to see generalization as always and necessarily proceeding from description, rather than as the dialectical process it is. The fact is, clarifying our comparative methods and assumptions could not fail to improve our descriptions.

By its nature, the present volume challenges the more extreme assertions of poststructuralism, as well as those who would argue that comparison is futile. Balinese “eating” and other curiosa notwithstanding, the human experience is sufficiently similar to make comparison possible. The diversity that Benedict finds in the guardian spirit complex is real enough, but the various expressions subsumed by this diversity are nonetheless comparable. Indeed, by pursuing such a comparison, she learns more about the cultures in question and the way their institutions are interwoven. With respect to local context, then, and to the definition of phenomena, there is much to be learned in the process of (p.6) comparison. The editors and contributors concede, however, that the epistemological criticisms of the comparative method properly test the mettle of anthropology, as regards its claim to scientific seriousness. The issues these criticisms raise intrude upon our project by way of problematizing the ontological status of ethnographic description, regional classification, and cross-cultural generalization (Holy 1987; Fardon 1990; Strathern 1990). The fact remains, however, that the poststructuralists, although modish and with theoretical agendas of their own, are but the latest players in the century of turmoil surrounding the comparative method in anthropology.

Amazonia, Melanesia, and the Comparative Method

In addressing the substantive and theoretical issues associated with comparison, this volume stands on an unusually solid foundation of scholarship. To begin with, the quality of ethnographic coverage in both Amazonia and Melanesia is of a very high order, sufficient even to Boas's stringent standards. In addition, each region has produced a sizable literature of areally and topically focused collections. In Amazonia, one notes David Maybury-Lewis (1979) on “dialectical” societies, Peter G. Roe (1982) and Jonathan D. Hill (1988) on myth and history, Hames and Vickers (1983) on ecology and adaptation, Peter Rivière (1984) on basic structural models among lowland societies, Kenneth M. Kensinger (1984) on marriage systems, Greg Urban and Joel Sherzer (1986) on native discourse, and, most recently, a series of comparative works under the general editorship of Kenneth M. Kensinger, entitled South American Indian Studies and Working Papers on South American Indians. Melanesia has produced similar studies, including distinguished collections on sex and gender (Brown and Buchbinder 1976; Herdt 1982, 1984), ritual and cosmology (Lawrence and Meggitt 1965; Stephen 1987; Gewertz 1988; Herdt and Stephen 1989; Stewart and Strathern 1997), social organization (Watson 1964; Glasse and Meggitt 1969; Cook and OʼBrien 1980), leadership (Godelier and Strathern 1991), and regional culture (Lutkehaus et al. 1990; Strathern and Stürzenhofecker 1995). The historical dimension of Amazonian and Melanesian societies, so central to Boas's critique, is also increasingly accessible (e.g., Hill 1988; Ferguson and Whitehead 1992; Gewertz 1983; Gewertz and Errington 1991; Knauft 1993). Thanks to this vigorous work, Amazonian and Melanesian data are extraordinarily rich, and their value to wider ethnological questions has been enormous. In short, we know much about Amazonian and Melanesian cultures in context.

The “solid work” of ethnography and ethnology in Amazonia and Melanesia has established the informational base that is necessary for testing methodological principles and criticisms. This is a crucial precondition of the enterprise of this volume, for it enables the contributors vastly to enrich their own ethnographic descriptions with insights drawn from a rich, reliable, and (p.7) relevant literature. The stage is thus set for the kind of informed comparative study that Boas called for and would approve.

The strategies of comparison utilized by our contributors, however, would have seemed novel to Boas. In aggregate, the authors combine both the “universalist” and “localist” traditions mentioned above. Although the project shares the former's interest in similarities between historically and geographically unrelated cultures, it “controls” that comparison by favoring for attention instances that are, indeed, similar—or that, at any rate, appear to be so. This does not imply a lack of interest in differences between cultures of the two regions. As previously noted, some differences examined here are quite germane to the comparison; moreover, systematic differences between the regions may be construable as similarities at a higher level of abstraction, in that the relations between things may be similar, even if the things themselves are not. What our approach does imply, however, is that a search for differences is pointless, unless such differences occur and are made meaningful in the context of similarities. As Evans-Pritchard remarked (1965, 25), “institutions have to be similar in some respects before they can be different in others.” This of course begs the question of how “similarities” are to be demonstrated and understood between unrelated cultures; hence the need for epistemology.

The contributors to the volume follow “localist” traditions of comparison as well as universalist models. For the most part, they compare richly contextualized constructs—systems, processes, and relationalities—rather than isolated traits. And instead of exploring the entire panoply of comparative possibilities, they sharpen their focus on gender, thereby to advance the methodological objectives of the project. As explained below, gender is a powerfully integrative topic that reflects the interests of many scholars who work in Amazonia and Melanesia.

In summary, comparison is the bread and butter of anthropology. It is inherent in the act of classification, by which we identify unfamiliar behaviors, describe institutions, and communicate the results of our work to others. We cannot describe one society without having others in mind, for comparison is the recurring element of our basic analytical tools. Comparison establishes and refines a common discourse among scholars working with different cultures (cf. Strathern and Stewart 1998, 251); it stimulates and provokes new perspectives on findings from particular cultures; and it allows us to search for general principles through controlled comparison. Comparison elevates the level of our work to the quest for principles of human life that transcend any one culture, even as it accepts the importance of culture in forming people's interests and the views they have of others. Without comparison, we risk miring our work in exotica and in the description of the particular: “without systematic comparative studies, anthropology will become only historiography and ethnography” (Radcliffe-Brown 1951, 16). The prospect is far worse than Radcliffe-Brown supposed. For, without comparison—without systematic (p.8) observation, classification, and generalization—anthropology will become nothing at all. Hence the need to articulate and exemplify new approaches to the old problems of anthropological comparison.

Sex, Gender, and Related Dimensions of Comparison

In Search of “melazonia”

One of the contributors to this volume, Stephen Hugh-Jones, has suggested that a comparative perspective on the cultures of Amazonia and Melanesia presupposes an imaginary place, a place which he felicitously dubbed “Melazonia.” Within the borders of Melazonia, we would find cultures whose characteristics make their comparison useful and interesting. Clearly, there are many dimensions of comparison that would define citizenship in Melazonia, including similar environments and subsistence regimes, patterns of leadership, and shamanistic religion. For us and many of our colleagues, however, the resemblance among the societies in Melazonia that stands out most dramatically is gender. This is a useful basis for a comparative study, since gender is a topic at the forefront of anthropology. It is an inherently integrative subject, bringing together intellectual perspectives derived from such diverse areas as human biology, environmental studies, psychology, social anthropology, and the humanities. Above all, the territory of Melazonia, when its frontiers are defined by patterns of gender, seems to provoke fascinating insights about the cultures studied by our contributors and, more generally, about the human condition.

Gender-inflected Societies

Gender is of great importance in the context of many of the small-scale cultures of Melanesia and Amazonia. Though we are wary of essentializing assumptions that lead to the stereotyping of cultural regions, the societies of Amazonia and Melanesia are arguably gender inflected (Lindenbaum 1987, 222), perhaps more than any other areas in the world. That is, gender roles and their attendant ideas about sexuality appear as templates for many other domains of culture. Human sexuality is projected upon nature so that flora, fauna, and natural objects have anthropomorphically sexual qualities. Ritual systems, cosmology, leadership, warfare, self-concepts and images of the human body, kinship, and perceptions of the environment are genderized: thought of and conceived in terms that are linked to masculinity, femininity, and human sexuality. Spiraling upward and away from the specific images of human sexual physiology and anatomy, sex and gender in Melazonia infuse broader worldviews about the nature of the human condition. Hence Stephen Hugh-Jones (Chapter 11), using material from the Barasana, sees the embodiment of gender identity as a meditation on the interdependence of (p.9) men and women. For both, “the human body and its various parts—vocal apparatus, gut, bones and genitals—are all tubes. Through the couplings of these tubes and the passage of various substances—food, water, air, sound, semen, blood, feces, children—along their interiors, the flow of life is ensured.” In both Amazonia and Melanesia “sexuality and reproduction flow into each other and are part of a broader confluence of indigenous philosophies about the cosmic sources of life, health, and growth and of decay and degeneration” (Jolly, Chapter 8).

The sexualized cosmos inhabited by many Amazonian and Melanesian cultures is by no means limited to these regions. What makes it remarkable is the degree to which it obtains there. Lawrence E. Sullivan (1988), for example, provides us with an extraordinarily detailed summary of Amazonian religious systems. In bringing together the mythology of the Amazonian region, Sullivan adduces seven basic summarizing principles, one of which is sexuality: “the mode of being human is distinguished by sex from the very beginning; i.e., the very consciousness of sex is sacred.” Peter Roe, who has also attempted an integration of Amazonian symbol systems, is unequivocal. In an extended discussion of gender (1982, 265–273), he notes that “the most meaningful distinctions in these societies are those between male and female” (p. 265), and that gender is the source of most of the metaphors of social interaction as well (p. 266). It is easy to adduce specific Amazonian examples. Among the Desana and other Tukanoan peoples of the northwest Amazon, the entire cosmos is both sexualized and genderized. The Sun god's piercing rays are equated with phallic male sexuality while the earth and a uterine paradise below it are identified with femininity (Reichel-Dolmatoff 1971).

Similar ideas shape the cosmos in Melanesia. Lindenbaum could have been referring to Amazonia when she observed (1987, 222), “Papua New Guineans live in a gender-inflected universe in which the polarities of male and female articulate cosmic forces thought to be located in the human body; indigenous theories of human reproduction contain within them an implicit recipe for social reproduction.” In Melanesia, animals, plants, common objects, and sacred implements are often described as having a gender identity and even sexual motivation (Tuzin 1972, 1992a; Meigs 1984). Individuals, as well as cultures, are conceived largely in terms of masculine and feminine categories. Rituals marking puberty and the life cycle are redolent with gender imagery. Sexuality is an intense cultural focus, one often associated not only with procreation and pleasure but with illness and symbolic contamination. Female sexuality is considered by men to be dangerous and yet is imitated by them in rituals of symbolic menstruation, procreation, or (less commonly) parturition. Sex and gender in both Amazonia and Melanesia may have a dreamlike quality, in which, it would seem, the extremes of fantasy are realized in conduct.

(p.10) We shall see that the contributions to this volume make the case for a Melazonian cultural universe that is redolent with all kinds of sexual imagery Stephen Hugh-Jones (Chapter 11) describes the sexualization of foods, fauna, and flora within the symbolism of initiation. Paul Roscoe (Chapter 12), in his comparative chapter examining a broad sample of cultures in the two regions, looks at how ideological issues of sex and aggression are conflated, demonstrating that sexual prohibitions adhere to dangerous and aggressive ventures. Philippe Descola (Chapter 5) examines the gender categories that classify plants, animals, and spirits. Margaret Jolly (Chapter 8) looks at the equations of food and sexuality. Thomas Gregor and Donald Tuzin cite numerous similar metaphors from a range of cultures (Chapter 14). And Michael Brown (Chapter 9) describes how gender and sexuality may structure the process of religious transformation.

In considering the gender-inflected universes of many Amazonian and Melanesian peoples, the task of comparison must confront two fundamental questions. The first is the question of essentializing the two regions. Do all South American and Melanesian cultures construct their cosmos with the same single-minded attention to sex and gender? And what does it mean to the comparison of the two regions if some or many of them do not? One of the contributors to this volume, Philippe Descola (Chapter 5), argues that gender and sexuality are far less important in Amazonia than in Melanesia. In an original and challenging article, he maintains that gender is not of particular relevance to Amazonian cultures. Rather, Descola maintains, sex and gender are subsumed by more significant social and symbolic categories such as the relationship of in-laws and consanguine kin, and even the predator-prey relationship that distinguishes the perception of nature of some Amazonian peoples. These themes, he believes, are the essence of Amazonian cultures, which are thereby quite different from those in Melanesia, preoccupied as they are with sex and gender. For Descola, Melazonia is a small place indeed, and only half populated at that.

Descola's complex and rewarding article must be read in full to be appreciated as it deserves to be. We take issue with both his reading of the Amazonian ethnography and, perhaps more relevantly, the level of abstraction by which he adduces what is relevant for the people themselves. Moreover, Descola's own field research, conducted among a relatively nongenderized group, the Achuar, may have influenced his generalizations, or the Achuar may constitute a “limiting case” of the interregional comparison. In any case, as we see it, his search for essences or the “master codes” of Amazonia and Melanesia serves only one of the purposes of comparison. We also look for useful comparisons between particular societies that illuminate our own data and enlarge our thinking. It is no embarrassment to the field of Melazonian studies that some of the cultures of its constituent regions have no comfortable place within greater Melazonia.

(p.11) Our interest in the sexualized worlds of Amazonia and Melanesia engages a second question that is also fundamental to comparison. How can we be sure that the apparent resemblances are meaningful rather than superficial? Are sex and gender as organizing ideas put to similar uses among the cultures of the two regions, or are the resemblances more a matter of surface appearance? The question is best answered by our contributors, who place the similarities and differences in richly detailed contexts. But we also note that in both Amazonia and Melanesia sex and gender are linked to self-concepts, suggesting that we are dealing with similar ways of thinking.

Body, Person, and Self

The Melazonian self is a cultural domain closely associated with indigenous assumptions concerning sex and gender. In both Amazonia and Melanesia, the self-concept, social identity, and the anatomy and physiology of the human body are intertwined with theories of conception, maturation, depletion, and death. Conklin and Pollock (n.d.), writing of Amazonia, note that social personhood, including gender, is acquired as if it were the product of growth, through an incremental process rather than being conferred all at once (cf. Biersack, Chapter 4). Even when social transitions are ritually marked, as is often the case with developmental statuses, the rituals are drawn out, so that the status is acquired over an extended period. Crucial to this process is the notion of shared or exchanged bodily substance: movements of physical material from the body of other individuals “via sexual intercourse, breast feeding, food sharing, physical intimacy, enemy killings [and] interactions with animals.” The process rises in the mythic beginnings of men and women, who are commonly assigned radically different origins (Sullivan 1988: Brumbaugh 1990; Tuzin 1997). Moving through the life cycle, masculine and feminine identity is defined by the physical attributes of the human body, and the bodies with whom one comes into contact. Hence, the cultures of the two regions focus enormous attention on menstrual blood, semen, milk, and hair as markers of identity Self-definition, depending as it does on an incremental process of absorbing the qualities of others, is highly malleable, thus providing an ideological basis for the flexibility of social organization that is so characteristic of both areas (Strathern, Chapter 10). The definition of kinship categories, group identity, and community membership are thereby “biologized” and embodied, and inevitably implicate notions of sex and the physical differences between men and women (Seeger et al. 1979; Turner 1980). These issues correspond to current concerns of Amazonianists and Melanesianists and their efforts to conceptualize similar mutable, sexual, and biologically represented relationships within the cultures of the two regions (cf. Herdt 1981; Poole 1981a; Strathern 1988).

(p.12) When examined closely, how similar are Amazonian and Melanesian images of self and embodiment? For Conklin (Chapter 7), who looks at the parallels of warrior seclusion and pregnancy among the Amazonian Wariʼ, the specific parallels are “so striking that when Amazonian ethnographers read the Melanesian ‘body’ literature, we often get that frisson of uncanny recognition described well by the American baseball player, Yogi Berra, who exclaimed: ‘It's like déjà vu all over again.’”

Our authors examine these parallels of self-definition and gender in a series of comparative studies that emphasize ideas about the embodiment of identity in the two regions. Bonnemère (Chapter 12) and Hugh-Jones (Chapter 11), although using different theoretical perspectives, examine the sexual metaphors that frame masculinity among the Anga peoples in Melanesia and those of the Vaupes region in Amazonia. Biersack (Chapter 4) looks at the magical efforts of young men among the Melanesian Paiela to grow beautiful hair as a marker of growth, sexual attractiveness, and masculine status. Fisher (Chapter 6) finds similar patterns of body imagery among the images associated with both age and gender among the Brazilian Kaiyapo and the Yangorou Boiken of New Guinea. Finally, Jolly (Chapter 8), focusing on sexual desire as well as self-identity, finds resemblances between the northwest Amazon and New Guinea.

Procreative Imagery and Ritual

The embodiment of symbols and self that we note in the cultures of island Melanesia and New Guinea is often focused on rituals and imagery that are imitative of female reproductive physiology and anatomy1 The most vivid examples of this pattern occur in male initiation rituals, in which boys are symbolically gestated and birthed by men. In Australia and New Guinea, the implications of this idea are followed to their logical extreme, as in the passing of initiates through the legs of the men in a form of “anal” birth among the Wikmunkan (McKnight 1975, 94), the equations of breast milk and orally ingested semen among the Sambia (Herdt 1981), and many variant forms of “male menstruation” that occur throughout New Guinea (for example, the Wogeo [Hogbin 1970]).

In Amazonia we find the same pattern in male initiation rituals, albeit in somewhat more symbolic, abstract, but still unmistakable form. Barasana and northwest Amazon initiation, for example, make use of such imagery (S. Hugh-Jones 1979, 132), and in the Upper Xingu boys having their ears pierced (one of the major markers of adulthood) shed “menstrual” blood (Gregor 1985). Amazonia is also famous for that most literalized male procreative ritual, the couvade, in which fathers directly imitate the labor pains and assume the taboos of their pregnant wives. These beliefs and practices, unfettered as they are by the realities of human biology, are, in their way, extraordinary triumphs of the imagination (Shapiro and Linke 1996).

(p.13) Our authors bring a number of very different approaches to the understanding of procreative symbolism. Bonnemère, who notes the parallels of Melanesian and Amazonian initiation, sees female reproductive physiology and anatomy as a natural symbol and an analogy for the “birthing” of boys: “The female body being ‘a body capable of reproducing itself’ … constitutes an adequate model for thinking and operating the maturation of men's bodies, which lack this capacity” In short, for Bonnemère, the imagery of initiation is based on observation, analogy, and logic.

Conklin (Chapter 7) and Hugh-Jones (Chapter 11), utilizing “relational approaches” developed in Melanesia, primarily by Marilyn Strathern (1988), see procreative imagery as suggestive of the mutual dependence of men and women. Conklin looks at the pregnancy-like seclusion of Wariʼ warriors who were said to be filled with the blood of their slain enemies. As a gestating women creates new life, so does the warrior transform enemy substance into vitality and power. Hugh-Jones, examining men's initiation in the northwest Amazon, also sees the procreative symbolism of the men's cult as a statement of dependence, closeness, and even anatomical resemblances. Taken together, the symbolism of the men's cult is a philosophy, “a meditation” about the mutually integrated roles of men and women.

Gregor and Tuzin use a radically different, psychoanalytic perspective, in that procreative symbolism is seen as reflective of masculine insecurity. Only rarely, they note, does women's ritual imitate male reproductive physiology (for the case of Shipibo clitoridectomy, see Roe 1982, 93, 106). Male procreative beliefs both express anxiety over ambiguous sexual identity and attempt to gain mastery over the feminine part of the male self.

We regard the alternate interpretations of procreative imagery as complementing one another. Human life is “overdetermined,” and it is expectable that it can be approached from a variety of theoretical perspectives. What continues to astonish is the extent to which the cultures of Amazonia and Melanesia share a common vocabulary and language of gender.

Men's Cults

Among peoples of Amazonia and Melanesia that have formal men's societies, the attention to sex and gender may reach nearly obsessional intensities. Resemblances between the different regions are compelling and have been noted at least since the time of Schurtz (1902) and Webster (1907). Typically, men's organizations are associated with meeting grounds or men's houses, where men conduct secret initiations and feasts. The cults address similar spirit entities, conceal similar secret paraphernalia and sound-producing instruments, and punish female intruders with gang rape or death. Taken together, the pattern of spatial separation, initiations, and punishment of female intruders constitutes a “complex,” or adherence of traits, that is found widely throughout Melanesia, and in (p.14) at least four major and distantly separated culture regions in lowland South America. That “the men's house complex” (see Gregor 1979, 1985) obtains in both regions is noteworthy. What is even more striking is that the details of the cults also bear close comparison. The cults’ origin myths tell of a time when the women possessed secret and powerful objects, often bullroarers, flutes, or trumpets, and used them to dominate the men. Rallying together, the men forced or tricked the women into giving up the sacred instruments, resulting in the reordering of the society and establishing of patriarchy (see Bamberger 1974, and Lévi-Strauss 1973 for summary discussions of Amazonian variants; see Gewertz 1988 for Melanesia). In both regions, the men share a strategic, potentially troubling secret: that the sounds of the trumpets, flutes, bullroarers, and other instruments associated with the cult are not the voice of spirits, as they allege, but are produced by the men themselves (cf. Metraux 1927).2 This summary hardly does justice to the energy that has generated the array of ideas and symbolic associations linked to the cults. In their density, in their manneristic elaboration, in their fantastical departure from ordinary life, and in their raw, uncensored use of primary-process symbolism, they are among the more remarkable human creations documented by ethnography.

How are we to understand the parallels between the men's cults of Amazonia and those of Melanesia? Until now the comparative method, as applied within the Amazonian and Melanesian culture area, has been the most productive explanatory strategy, at least with respect to the general pattern of men's organizations. In each region the cults vary in the intensity of their initiation rituals, in the degree to which their boundaries are defended, and in the extent to which they shape community life. In the context of Melanesia, utilizing theoretical frameworks that could also be applied to Amazonia, men's organizations have been linked to such variables as local exogamy, warfare, descent ideology, and methods of child rearing (Allen 1967; Meggitt 1964; Whiting 1941). Robert Murphy (1959), working independently with Amazonian data prior to Allen's study, came to remarkably similar conclusions, as if he and the Melanesianists had been laboring in the same vineyard all along. Thus, Allen remarks that he had not known of Murphy's paper at the time of writing, but that the latter's “argument, derived from an analysis of a South American … community, is substantially the same as my own” (1967, 3n).

In the decades that have elapsed since these explanatory efforts, the data have become immeasurably richer. The detailed ethnographies of men's cults and initiation practices now available from Amazonia, especially from the Upper Xingu and Tukanoan groups (e.g., Hugh-Jones 1979, Gregor 1985; Murphy and Murphy 1985) and Melanesia (e.g., Gell 1975; Barth 1975, 1987; Tuzin 1980; Herdt 1981, 1982a, 1982b; Juillerat 1992), allow us to revisit the question in greater depth and in sharper focus. Dealing with one region alone, one might dismiss the similarities in gender roles as the result of a common (p.15) origin among related cultures, or diffusion from a single source. But their appearances nearly as far apart as can be, among historically unrelated cultures, are evidence of similar underlying processes. Following this lead, the contributors to this volume examine men's institutions as a set of relationships and symbol systems that go beyond geographic areas. Bonnemère (Chapter 2) looks at the images of parturition and rebirth that mark initiation in both regions. Biersack (Chapter 4) looks at the relationship of men's cults and patriarchy. Hugh-Jones (Chapter 11), following Strathern's “relational” approach to gender, sees in the symbolism of the cults an expression of the interdependence of men and women. Hill (Chapter 3), in his discussion of the northwest Amazon, supports this perspective, and also provides a useful way of categorizing the intensity of the cults in both Amazonia and Melanesia. Finally, Gregor and Tuzin (Chapter 13) apply a psychoanalytic approach to men's institutions and examine the element of moral ambivalence that lies just behind the evident misogyny of exclusion, sexual opposition, and violence. They argue that the cults arise in Amazonia and Melanesia due to a mix of environmental and social factors that encourage the separation of men and women in gender-exclusive groups and leave the socialization of boys largely to their peers.


It has been just over one hundred years since the publication of Boas's critical paper on the comparative method. The editors and contributors believe that Boas's lessons have been learned by the discipline. This volume embodies a healthy and appropriate skepticism of comparative research that does not examine the comparability of the data it employs; it is animated by an awareness of the need for concepts and methods that are more powerful than before. We believe that Boas would approve our venture. Ours is a novel approach to comparison that goes beyond the small-scale, local-level comparisons Boas and Evans-Pritchard endorsed and yet controls for technology, level of culture, and many other factors. In pursuing this approach, the editors and contributors find, collectively, that there is no single, monolithic Comparative Method, but an unbounded set of comparative strategies—a battery of analytic devices, which, at multiple levels and in different degrees, illuminate both the phenomena we observe and the epistemological profiles of our inquiries. Through exemplifying these various strategies, the editors and contributors hope to enrich our understanding of comparative methodology, directly confront the meaning of the descriptive categories, and inspire each group of specialists with the ideas and findings of the other. To be sure, gender is the focus and content of the Amazonia-Melanesia comparison, as it is pursued here; but, in this context, gender also affords an exceptional opportunity to explore fundamental questions about the human condition and how one goes about conceptualizing and examining it.

(p.16) Notes

(1.) Since Hiatt's (1971) important article identifying this pattern in Australia, the literature has referred to male imitations of female gestation and reproduction in the context of initiation and other rituals as “pseudoprocreative” (e.g., Shapiro and Linke 1996). While we applaud the concept, the term itself strikes us as prejudging of the attitude of the participants and giving rise to a variety of logical problems concerning truth and falsehood (Tuzin 1995). Accordingly, we use the noncommittal “procreation,” and allow the context to determine whether and to what extent it is “pseudo-.”

(2.) In a forthcoming doctoral dissertation, ethnomusicologist Robert Reigle (n.d.) writes of the remarkably similar instruments, melodies, legends, and related practices that exist in Brazil's Matto Grosso and in the East Sepik and Madang provinces of Papua New Guinea. “Most striking,” he comments (personal communication), “are the musical forms and instruments that don't seem to exist anywhere except Brazil and New Guinea.”


(1.) Since Hiatt's (1971) important article identifying this pattern in Australia, the literature has referred to male imitations of female gestation and reproduction in the context of initiation and other rituals as “pseudoprocreative” (e.g., Shapiro and Linke 1996). While we applaud the concept, the term itself strikes us as prejudging of the attitude of the participants and giving rise to a variety of logical problems concerning truth and falsehood (Tuzin 1995). Accordingly, we use the noncommittal “procreation,” and allow the context to determine whether and to what extent it is “pseudo-.”

(2.) In a forthcoming doctoral dissertation, ethnomusicologist Robert Reigle (n.d.) writes of the remarkably similar instruments, melodies, legends, and related practices that exist in Brazil's Matto Grosso and in the East Sepik and Madang provinces of Papua New Guinea. “Most striking,” he comments (personal communication), “are the musical forms and instruments that don't seem to exist anywhere except Brazil and New Guinea.”