Quadripartite structures

Categories, relations, and homologies

in Bush Mekeo culture

Quadripartite structures
Categories, relations, and homologies
in Bush Mekeo culture

MARK S. MOSKO
Hartwick College

The right of the

University of Cambridge
to print and sell
all manner of books
was granted by
Henry VIII in 1534.
The University has printed
and published continuously
since 1584.

CAMBRIDGE UNIVERSITY PRESS

Cambridge
London New York New Rochelle
Melbourne Sydney
 CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo, Delhi

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521105385

© Cambridge University Press 1985

This publication is in copyright. Subject to statutory exception

and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.

First published 1985

This digitally printed version 2009

A catalogue record for this publication is available from the British Library

Library of Congress Cataloguing in Publication data

Mosko, Mark S., 1948-
Quadripartite structures.
Bibliography: p.
1. Mekeo (Papua New Guinea people) I. Title.
DU740.42.M67 1985 306'.0899912 84-19906

ISBN 978-0-521-26452-5 hardback

ISBN 978-0-521-10538-5 paperback
Dedicated to the memory of Kaiva Muniapu
Are we satisfied, then, that everything
is generated in this way -
opposites from opposites?

- Plato's Phaedo
 Contents

List of figures, tables, and maps page viii

Preface xi
1 Introduction: the problem and the people 1
2 Between village and bush 21
3 Body and cosmos 38
4 Sex, procreation, and menstruation 60
5 Male and female 73
6 Kin, clan, and connubium 100
7 Feasts of death (i): de-conception and re-conception 150
8 Feasts of death (ii): the sons of Akaisa 182
9 Tikopia and the Trobriands 200
10 Conclusions: indigenous categories, cultural wholes,
and historical process 234

Appendixes
1 Village resources derived from bush resources 250
2 Ingestion and ingestibles 251
3 Categories of food 254
4 Work and nonwork skills 256
5 Categories of human dirt 257
6 The myth of Foikale and Oa Lope 258
7 The afinama myth 265
Notes 271
Bibliography 278
Index 289

vn
 Figures, tables, and maps

Figures
1.1 The trajectories of Nature and Culture page 5
1.2 Fipa mythical and social values 6
1.3 Hohfeld's fundamental legal relationships 8
1.4 The structure of the Klein group in mathematics 9
2.1 The sphere of ordinary transfers 26
2.2 The sphere of extraordinary transfers 36
3.1 Water transformation: boiling 42
3.2 Water transformation: roasting 42
3.3 Food transformation: boiling 43
3.4 Food transformation: roasting 43
3.5 Food transformation: drying 44
3.6 Food transformation: ripening 45
3.7 Work transformation: house building 47
3.8 Nonwork transformation: hunting 47
3.9 Culinary work transformation: boiling food 48
3.10 Culinary nonwork transformation: roasting food 48
3.11 Sweet/unsweet body transformations 50
3.12 Blood and flesh synthesis: capacities for work and nonwork 51
5.1 Female ritual cycle 85
5.2 Male ritual cycle 89
5.3 Alternating gender categories 93
6.1 Atsi atsitsi terminology (patrilateral) 105
6.2 Atsi atsitsi terminology (matrilateral) 106
6.3 Ipa ngaua terminology (matrilateral) 107
6.4 Ipa ngaua terminology (patrilateral) 108
6.5 Ipa ngaua terminology (Ego's and descending generations) 110
6.6 Composite lineage history (Nganga clan) 119
6.7 Pisaua friendship network (Amoamo tribe) 131
6.8 Children-of-"first cross-cousin," or "second-cousin," marriage 135
6.9 Akaila public marriage compensation exchange 136
6.10 Agnatic, cognatic, and affinal bloods 140
6.11 Bush Mekeo marriage system (i) 143
6.12 Bush Mekeo marriage system (ii) 146
6.13 Bush Mekeo marriage system (iii) 147

viii
 Figures, tables, and maps ix

7.1 Mortuary feast-givers and -receivers 172

7.2 Mortuary-feast categories and clan identities 176
7.3 De-conception and re-conception of grandmothers'clan bloods 178

Tables
1. Village and tribal populations 15
2. Offices, lineages, and subclans of Nganga residential clan 118
3. Friend and nonfriend kofuapie betrothals and elopements 133

Maps
1. The Bush Mekeo 14
2. The Bush Mekeo and their neighbors 17
 Preface

Not without humor, Bush Mekeo villagers will occasionally retell the
story of how their ancestors first came to be known as the "Bush" Mekeo.
Whenever a government patrol entered the area during the early years of
contact, they say, their ancestors hid in the bush until the strangers had
left. Once, upon entering a deserted village, a patrol officer remarked,
"Oh, so these must be the 'Bush' Mekeo, because they are always hiding
in the bush." Figuratively speaking, the Bush Mekeo have remained "in
hiding" ever since. In his classic study, The Melanesians of British New
Guinea, Seligmann does refer to "a small but uncertain number of vil-
lages on the middle reaches of the Biaru River [which] must be consid-
ered to constitute an ethnographical annexe to Mekeo" (1910:311); but
now, even after nearly a century of contact with Europeans, the Bush
Mekeo are still essentially unknown to the outside world. Although there
have been numerous in-depth studies of their closest neighbors, virtually
nothing substantively new concerning the Bush Mekeo themselves or
their culture has appeared. This book is partially intended to help fill this
lacuna and bring the Bush Mekeo, as they would say, "out of hiding."
This book, however, also attempts something rather more theoretical
and, for that reason, potentially fruitful in other ways. In the course of
struggling to interpret Bush Mekeo tradition in my own thought as a
"total social phenomenon," a structure of an unanticipated form gradually
took shape. It became clear that the meanings of many (if not most) of the
cultures diverse contexts are ordered by and through it. That structure,
as it turns out, is generally fourfold or quadripartite. But with the specific
inner operations of its working among the categories of the culture, it is
more accurately characterized in the terms of homologously bisected du-
alities. This book is, then, principally devoted to revealing this particular
structure and its logically consistent ramifications throughout Bush
Mekeo culture.
Nonetheless, I was inevitably led to a detailed exploration, in much the
same terms, of several related Oceanic cultures - most notably, the clas-
sic cases of Tikopia and the Trobriands. The results of those inquiries are
contained in this volume as a separate chapter. Finally, emboldened,
XI
Xll Preface
perhaps, with these (to me) ethnographic and comparative successes, I
have ventured still further. I propose that, in a manner previously un-
suspected, the structure of bisected dualities characterizes - or underlies,
informs, and links together - a number of fundamental, but otherwise
heretofore disconnected, formulations within social anthropology itself,
particularly those dealing with the relations of myth to ritual and of struc-
ture and synchrony to history and diachrony. Therefore, in addition to
bringing the Bush Mekeo out of hiding, this book represents an effort to
reveal what perhaps has long been hidden within two of the most well-
documented ethnographic cases on record and within a few of the more
notable sectors of anthropological tradition as well.

I feel by now a particularly keen sense of indebtedness to the many

persons who have helped and .encouraged me along the way. Although I
shall never be able to recompense them adequately, I should now like to
acknowledge these debts and express my gratitude.
The writings of Edmund Leach, Claude Levi-Strauss, Marshall Sah-
lins, and David Schneider among living anthropologists stand out clearly
as my main theoretical inspirations. My greatest intellectual debt of a
more personal and immediate sort is to my doctoral adviser, Professor
Stephen F. Gudeman. The exceedingly high standards for quality, integ-
rity, and thoroughness he demands, not so much of others as himself,
have done more to inspire and guide me to think like a social an-
thropologist than he might ever guess. Professor Eugene Ogan conscien-
tiously served the no doubt tedious role of my principal theoretical foil;
thus, in addition to keeping me laughing and moving, he at least tried to
keep me honest. Others whose scholarly support and encouragement I
cannot fail to mention are Robert C. Kiste, Alan Rew, Paul Wohlt, Mar-
ilyn Strathern, David G. Baker, John M. Ingham, Mischa Penn, Richard
L. Haan, and Laurie Lucking.
My fieldwork and dissertation write-up were generously supported by
a predoctoral fellowship from the National Institute of General Medical
Sciences (Grant No. 01164). Hartwick College has also financially helped
in bringing this book into being by covering some of my costs of revision
and production. Ron Embling skillfully worked to complete the numer-
ous figures contained in the text. Dot Parmerter and Georgette Corrao
meticulously typed the several drafts of the manuscript.
Among those deserving foremost credit for their contributions to this
study are the Bush Mekeo villagers themselves. The patience, generosity,
and tolerance they so often displayed were truly astonishing. For the
benefit of those Papuans who might someday read this book, I would like
truthfully to declare that never were my family and I without food among
the Bush Mekeo. I must especially acknowledge the friendship so freely
 Preface xiii
offered my family and me by Pavivi Menga, Mangemange Muniapu,
Menga Piomaka, Ameaua Tsibo, Thomas Ae, Peniamo Peniamo, Peter
Keanga, Apou Kaengo, Piomaka Fala, and Marcello Apou. Particularly, I
would like to thank the village women collectively in appreciation of their
many kindnesses to my wife.
A great many others outside the Bush Mekeo from 1974 to 1976 also
provided invaluable assistance in one form or another. Among them are
Bill and Antje Clarke; Paul and Ruth Wohlt; Paul and Agi Kipo; Bishop
Vangeke; Fathers Boudaud, Didier, Diaz, and Bouseau; Sister Christine;
Epeli and Barbara Hau'ofa; Michael Monseel-Davis; Eliza Marshall;
Nigel and Joan Oram; Andrew and Marilyn Strathern; Jeff and Laura van
Osterwick; the staff of the National Archives; Waigani Lodge; WCA Bor-
oko; Bereina Government Offices; and the country order department at
Steamy's. I thank them all for their time, energy, resources, and hospi-
tality.
My parents have given me their unflagging support and encourage-
ment (moral and financial) at every stage in spite of their silent misgivings
about so unlikely a profession. My gratitude for their wisdom and con-
stancy goes very deep. My wife, Cassandra, has been a constant source of
undivided support, encouragement, and inspiration. Her toils and sacri-
fices in the field and afterward have allowed me immeasurable freedom,
without which I would have been devastated long ago.
Lastly, I would like to acknowledge my wide-ranging indebtedness my
to Bush Mekeo confidante, Kaiva Muniapu. Kaiva took it as a personal
mission to teach me his people's customs and to see that I did not leave
without understanding them to his satisfaction. But above everything
else, Kaiva taught me the meaning of trust and friendship and, by his own
example, what it means to be a Bush Mekeo man. Because Kaiva's un-
timely death in May 1977 prevented his ever seeing this book - the final
realization of his efforts and dreams as much as my own - it is especially
fitting that it be dedicated to him.

M. S. M.
 1
Introduction: the problem
and the people

This account of a Papuan culture is avowedly structuralist. In this view,

ethnographic description and explanation essentially consist of translating
the meanings of indigenous culture categories into our own language, and
constructing in the process a model of the total culture (Schneider 1972,
1980). For non-Western cultures like the Bush Mekeo, meanings cannot
be assigned or adduced either a priori or ad hoc from Western concepts.
Rather, meaning, as argued by Saussure in terms of linguistic value
(1959), is neither random nor piecemeal, but systematic and logical. It
resides in the interrelations among indigenous categories, in their rela-
tions of difference and similarity, in the underlying structure of ideas.
Moreover, the meanings of particular cultural elements are inseparable
from the wider synchronic "whole" or "totality."
The notions of meaning, indigenous category, structure, and cultural
whole are thus central to my treatment of Bush Mekeo traditions.
In the current "post-structuralist" era (Kurtzweil 1980; see also Fried-
man 1974) there has been a tendency for these conceptions to be super-
seded by reemphases upon social action, history, and diachrony. Un-
doubtedly, the revival of Marxian approaches (e.g., among others,
Friedman 1974; Worseley 1968; Harris 1968; Sahlins 1972; Bourdieu
1977; Godelier 1977) is largely responsible. Although certain elements of
this development are necessary and welcome, others are premature if not
regrettable - premature in that some of the most valuable and useful
insights deriving from the structuralist perspective have been passed over
without yet receiving adequate opportunity for empirical verification, and
regrettable in that the risks have consequently increased of seriously
distorting our conception and understanding of the essential nature of
cultural systems and how they are constituted. Namely, the contempo-
rary historicist approaches tend implicitly or explicitly to deny or ignore
both the analytical validity of indigenous categories and that cultures can
profitably be seen to consist of total integrated systems of ideas.
Quadripartite structures in anthropological perspective
It is also in response to these challenges, then, that I offer the following
structuralist interpretation of Bush Mekeo culture. Through the sequence
1
2 Quadripartite structures
of my chapters, that culture unfolds as a synchronic whole. In the process,
I focus upon the replication of a particular quadripartite structure evident
among the categories impinging upon various cultural and social contexts
or domains of village life. By unearthing this replicative structure or
pattern as I move from one context to another, the culture of the Bush
Mekeo and the meanings embodied in it are represented as a series of
homologies or metaphors.
The notion of structural replication has a long tradition in anthropologi-
cal theory and can be traced back to the very founding of the discipline
late in the nineteenth century. The systematic replication of relationships
within a single culture is fundamental, I think, to Durkheim's concep-
tualization of "collective representations" and Mauss's idea of "total social
phenomena." It is also central to Hertz's classic studies of religious polar-
ity and mortuary ritual. Structural or patterned replication is implicit as
well in the Boasian tradition, as evidenced by Benedict's "configura-
tionalist" theory and its intellectual cognates. More modern and explicitly
structuralist anthropological insights following chiefly from the works of
Evans-Pritchard, Levi-Strauss, and Leach have sustained and refined the
pursuit of culture's systematic nature through the internal replication of
form. Noteworthy examples of this approach include Douglas (1966),
Bulmer (1967), Burridge (1969a), Strathern and Strathern (1968, 1971),
M. Strathern (1981), Fernandez (1974), J. J. Fox (1971a, 1971b, 1973,
1974, 1975, 1980a), Ostor (1980), Jamous (1981), Gell (1975), Schneider
(1969, 1972, 1980), Kelly (1977), S. Hugh-Jones (1979), C. Hugh-Jones
(1980), Needham (1962, 1973, 1979), Dumont (1970), Tambiah (1968a,
1968b, 1983), Gudeman (1976), Vogt (1969), Sahlins (1976), Bourdieu
(1973), and Shapiro (1981). My description and interpretation of Bush
Mekeo culture in the following chapters should be generally viewed,
then, as a continuation of this core anthropological tradition.
It is probably fair to say that the greater share of anthropological studies
in this tradition has focused on binary or dualistic forms. Nature/Culture,
Sacred/Profane, Right/Left, Male/Female, Life/Death, Above/Below,
This World/Other World, and Wife-Giver/Wife-Receiver are among the
more familiar (Durkheim 1915; Hertz 1960; Needham 1973; Leach 1954,
1964, 1966a; Levi-Strauss 1969a; J. J. Fox 1971b, 1973, 1974, 1975; Lancy
and Strathern 1981). Indeed, of considerable significance here, one re-
cent observer has noted that the culture of the neighboring Central
Mekeo is particularly marked by a wide and complex assortment of bi-
nary-category oppositions (Hau'ofa 1981:289-91). Nonetheless, rep-
licative structures with more than two elements or relations have also
been proposed (cf. Needham 1973, 1979). These more complex struc-
tures, although still formally reducible to binary oppositions, have been
in terms of their cross-cultural significance predominantly either tripar-
 Introduction 3
tite (e.g., Levi-Strauss 1963a, 1966b; Leach 1964; Douglas 1966) or quad-
ripartite (e.g., Levi-Strauss 1963a; Foster 1974; Leach 1958a, 1961). With
triadic structures, the third element is typically added to the binary pair
in order to "mediate" or "resolve" the opposition between them. With
quadripartite structures, a similar function is performed by the inversion
or reversal of the initially opposed binary pair. Obviously, binary, tripar-
tite, and quadripartite structures are not mutually exclusive in a for-
malistic sense (Hammel 1972). Clearly, also, the notions of contrast, op-
position, contradiction, reversal, and inversion are implicit in all three
kinds of structure. Unfortunately, as it has been noted, the theory of
reversal in anthropology "is still rather more random than formal" (Foster
1974:346; cf Geertz 1972:26; Leach 1954, 1961:132-6; Levi-Strauss
1963a; Gell 1975:335-8; Fortes 1970; Kelly 1977).
In any case, the kind of structure that I show to be replicated through-
out Bush Mekeo culture is not simply binary nor triadic but quadripartite.
Categories distinguished and mutually defined as belonging to the same
set systematically come in fours. Each fourfold category group is initially
composed of a single binary opposition (X' : Y"), which is itself
bisected by its own inverse or reverse (Y' : X "). The complete category
set can thus be expressed in terms of a double analogy:
X' : Y" :: Y' : X" (1)
This is the structure of bisected dualities that, I shall argue, systematically
underlies the category distinctions of Bush Mekeo culture and that con-
stitutes the homologous or metaphorical relations among the culture's
varied contexts.
It is of some considerable significance to indicate preliminarily that this
particular structure of bisected dualities is isomorphic with several other
important quadripartite structures that have been proposed both within
and without anthropology. First, Levi-Strauss's cryptic formulation of the
underlying structure of myth corresponds almost precisely with my no-
tion of bisected duality:
Fx(a) : Fy(b) = Fx(b) : Fa-ify)
Here, with two terms, a and b, being given as well as two functions, x and y, of
these terms, it is assumed that a relation of equivalence exists between two
situations defined respectively by an inversion of terms and relations, under two
conditions: (1) that one term be replaced by its opposite (in the above formula, a
and a - I); (2) that an inversion be made between the function value and the term
value of two elements (above, y and a) (Levi-Strauss 1963a:228; see also Levi-
Strauss 1967; Leach 1970:62-86).
The only major difference between Levi-Strauss's equation and my own
concerns their respective ranges of application. Levi-Strauss has re-
4 Quadripartite structures
stricted his efforts in this context largely to mythological or narrative texts
(cf. Maranda and Maranda 1971). Furthermore, he has characteristically
dealt with the mythical literatures of many societies - all of Amerindian
myth in his Mythologiques, for example - as a total corpus rather than
with the myths alone (or, more preferably perhaps, the myths alongside
the related institutions) of a single cultural tradition. The most notewor-
thy exception here is probably his celebrated essay on the Tshimshin
myth of Asdiwal (1967). As convincing as Levi-Strauss's general approach
to mythology might appear to some (to others not), it has nonetheless
tended to leave aside the structural integrity of separate cultural tradi-
tions. Also, because he draws the myths for his transformational groups
from across cultural boundaries, Levi-Strauss's efforts have too often suf-
fered for lack of empirical verification by comparison with nonmythical
materials of the same traditions (Maybury-Lewis 1969; Burridge 1967;
Willis 1967). My own exploration of structures of bisected dualities, by
contrast, is initially restricted to a single cultural system, the Bush
Mekeo, and involves comparisons between mythical as well as non-
mythical contexts: conceptualizations of space and time, physiological
process, gender roles, social organization, leadership and authority, ritu-
al, etc. Geertz (1973, 1974; see also Schneider 1972), for one, has argued
for such a hermeneutic for similar reasons, but he stops short of seeking
formalistic interconnections in favor of highly elaborated exegesis, or
"thick description," of native text. Inasmuch as the ideas impinging upon
these various contexts of Bush Mekeo culture all conform to the quad-
ripartite structure of bisected dualities, the total culture - not merely its
parts in isolation nor in comparison with analogous parts of other tradi-
tions - constitutes a "whole" or transformational group in its own right.
In another theoretical treatment of mythical thinking, Godelier (1971)
posits a similar structure, but one of subs tan tively specific elements,
namely, the possible "trajectories of analogical linkage" between Nature
and Culture (Figure 1.1). The form of these relations, however, is still
that of a bisected duality. The trajectories are thus four in number. Com-
bined in a kind of vectorial algebra of the imagination, "[they confer] on
mythical discourse and mythical thought their inexhaustible polysemia
and symbolic richness" (1971:100). Any particular mythical projection can
be thereby plotted along one of the indicated trajectories of the graph. In
this way, all the possible kinds of understanding that can be mythically
created between Nature and Culture may be characterized:

I. Culture > Nature

II. Nature » • Culture
III. Culture < • Culture
IV. . Nature * * Nature
 Introduction
Trajectory I
NATURE ^ CULTURE

IV I III

NATURE ^ CULTURE
Trajectory II

Figure 1.1. The trajectories of

Nature and Culture. (From M.
Godelier, "Myth and History,"
New Left Review, vol. 69, p. 100.
Reprinted by permission.)

such that
I : IV :: II : III
Of course, one has to accept here not only the universality and signifi-
cance of the Nature/Culture dyad (Levi-Strauss 1969b; cf. MacCormack
and Strathern 1980), but also, in Godelier's handling, the reality of such
reified entities as Myth, Society, and History. The alternate and em-
pirically verifiable notion of a structurally integrated cultural whole is
thus still lacking. Nonetheless, Godelier's formal convergence with Levi-
Strauss upon the structure of myth in terms of bisected dualities is
provocative.
Attempting to overcome empiricists' critiques of structuralism gener-
ally and Levi-Strauss's handling of myth in particular, Willis (1967) has
examined the systematic distinctions embodied in Fipa mythology against
the backdrop of the wider system of Fipa social organization. He dis-
covers that the mythical structures are homologous with the more readily
verifiable structures of the sociopolitical system, and the form they share
in this instance is that of bisected dualities. In both contexts, relations of
complementary opposition are systematically inverted through being at-
tributed contrasting values, positive (+) and negative (—) (Figure 1.2).
Mythically and socially, Fipa sometimes attribute the "Head" positive
values (maleness, intellect, authority, seniority) and the contrasting
"Loins" negative ones (femaleness, sexuality, reproduction, juniority),
whereas other times the "Head" receives the negative values (lightness,
fewness, weakness, constraint) and the "Loins" the positive ones (heav-
iness, numbers, strength, fellowship). Thus,
(+) Head : (-) Loins :: (+) Loins : (-) Head
This represents a decisive step in the directions I am suggesting. First,
Willis is struggling to make structural models more empirically satisfacto-
ry. He is notedly successful in this particular case because, second, he
 Quadripartite structures

Figure 1.2. Fipa mythical and social values. (From R. G. Willis, "The Head and
the Loins: Levi-Strauss and Beyond." Man [n.s.] 2:524.)

deals with diverse contexts of the culture together as if that culture were a
total system. And third, the exact structure he posits for the Fipa is
convergent with that characteristic of the Bush Mekeo culture, that is, as
a system of bisected dualities.
In his essay "Structure and Dialectics," Levi-Strauss (1963a:232-41)
gives at least a passing indication of these possibilities. Exploring the
relation between myth and ritual, contra Malinowskian wisdom, he most
vigorously advocates the comparison of myths and rituals from different
societies out of their respective contexts (see also Foster 1974:346-7).
Nevertheless, he does briefly mention the possibility of establishing
structural homologies between myths and rituals of a single society.
Moreover, in the two ethnographic cases of this sort he discusses, he finds
just such a correspondence between the structure of twofold opposition
that generates the myths - his analogic model given above - and the
patterning of the rituals (cf. Gell 1975:341-6). Unfortunately, Levi-
Strauss has never fully developed this specific procedure elsewhere.
The study of ritual separate from myth has of course also preoccupied
generations of anthropologists. To many today, the foremost classical au-
thority on ritual is Arnold Van Gennep (1960). Van Gennep observed that
rituals in different societies frequently followed the general pattern of
"rites of passage," characterized by a tripartite sequence of separation,
transition, and incorporation. Modern symbolist explorations of ritual
largely continue to follow Van Genneps formula, as, for example, in the
splendid works of Victor Turner (1967, 1969). However, it is Leach
(1961:132-6) alone who so far has recognized the essentially fourfold
character of rites of passage in the form of a double opposition - sacred
(transition) time versus profane time, and separation (sacralization) versus
aggregation (desacralization) - or, in my terms in the form of a bisected
duality:
sacred : profane :: separation : aggregation
The structures of myth and rites of passage can thus be seen as homolo-
gous. Foster (1974) reaches virtually the same conclusion with respect to
myth and ritual generally in her comparison of Navaho and American
sacred activity. This heretofore barely recognized convergence could well
have a major bearing on the classical issue of the relation between myth
and ritual. Although such prominent figures as Robertson Smith (1957),
Jane Harrison (1903, 1912), Durkheim (1915), Radcliffe-Brown (1939),
 Introduction 7
Malinowski (1948), Kluckhohn (1942), and Spiro (1964) have argued as to
the relative priority of one over the other, all have been taking predomi-
nantly substantive elements only into consideration. Neither viewpoint
has been satisfactorily supported with ethnographic materials (Levi-
Strauss 1963a:232-41). And this, I strongly suspect, is because the per-
ceived congruities are structural rather than substantive, and empirically
verifiable perhaps only in the context of comparing rituals and myths of
the same tradition, at least in the preliminary stages of research.
Several modern descendant adaptations (acknowledged or otherwise) of
Van Gennep's original formulation, which have received considerable
recognition, exhibit the same logical scheme. By what amounts to project-
ing rites of passage onto the level of whole societies or major segments of
societies undergoing "millenarian movements," for example, Burridge
isolates three phases additional to a fourth period of stable tradition and
old rules: (1) doubt and uncertainty (old rules in doubt), (2) orgiastic and
other activities (no rules), and (3) new rules (1969b: 165-70). Thus, the
phases of millenarian movements are homologous with rites of passage:

sacred : profane :: segregation : aggregation

no rules : old rules :: old rules in doubt : new rules affirmed
To take a second example dealing with the same class of phenomena
under the alternate rubric of "revitalization movements," Wallace (1956)
develops virtually the same quadripartite1 temporal ordering:
period of steady period of period of
cultural distortion ' state " increased individual stress ' revitalization

Still other quadripartite structures have been proposed by anthropolo-

gists for application to non-Western materials. Douglas's (1970) graph of
"grid" (private vs. shared classifications) and "group" (control of other
people vs. being controlled by other people) is one recognizable instance
of a double opposition. The evident intent in this case is to plot and
classify the overall comparative similarities between different cultural
traditions rather than to portray the homologous relations between cate-
gories indigenous to particular cultures as the logic by which the very
nature of those separate traditions is made manifest (i.e., as wholes).
Significantly, also, part of the meaning of Douglas's "group" dimension
had been explored elsewhere by Hohfeld with his four fundamental types
of legal relationship (Figure 1.3). However, Hohfeld's quadripartite for-
mulation is composed of substantive elements that of course are very
likely not to be found all together beyond the boundaries of Western
jurisprudence. And in any case, it constitutes only one part of the totality
of Western culture.
 Quadripartite structures

Person A Person B
I. Demand-right < • Duty
II. Privilege-right No-demand-right
III. Power Liability
IV. Immunity No-power
Figure 1.3 Hohfeld's fundamental legal relationships. (From
E. A. Hoebel, The Law of Primitive Man, p. 48. Cambridge,
Mass.: Harvard University Press, 1954. Reprinted by permis-

The possibility inevitably arises that these several instances of bisected

dualities in the theory of social anthropology are logical devices specific to
our own manner of viewing the world, including the perception of other
cultures, rather than inherent in the nature of those other systems. Are
we as anthropologists, in our efforts to interpret other cultures free of
ethnocentric distortion, ourselves guilty perhaps of inadvertently impos-
ing upon them the very forms of logic and order characteristic of our own
culture (Hallpike 1976)? All the versions of quadripartite structures I have
mentioned are in essence compatible with the theory of the transforma-
tional or "Klein group" in mathematics (Figure 1.4). The suspicion in-
creases still more because other social sciences have also fixed upon the
same class of transformational structures in the contexts of their investiga-
tions: the relations of phonology, syntax, and semantics in linguistics
(Jakobson 1948; Chomsky 1957), and the INRC (identity, negation, reci-
procity, correlation) group of operations of Piagetian developmental psy-
chology (Piaget 1949, 197O:31n), for example. Closer to home, Andriolo
(1981) has recently shown by applying ethnosemantic techniques to myth
and history as a single domain that conventional anthropological contrasts
between these two in dualistic terms are inadequate for capturing the full
breadth of issues involved. Specifically, two intersecting dimensions -
replication versus differentiation and vector versus field - are required.
In this one critical context, then, anthropology is revealed to be itself
built upon a foundation of bisected dualities. In a more comprehensive
examination, Auge (1982) plots the whole of modern and traditional an-
thropological theory along two cross-cutting axes constituted of the op-
positions "symbol versus function" on the one hand and "evolution versus
culture" on the other.
The ultimate significance of these several structural parallels in an-
thropology and related disciplines, and the philosophical implications of
reflexivity to which they give rise, go considerably beyond the eth-
nographic limitations of this work. But still they bear upon it. Although,
indeed, in the process of interpreting the culture of the Bush Mekeo it
will be (as it always is) impossible not to draw upon constructs of our own
 Introduction

Figure 1.4. The structure of the Klein group in

mathematics. (After M. Barbut. From Michael
Lane [ed.]: Introduction to Structuralism, p.
368. Introduction and compilation © 1970 by
Michael Lane. Reprinted by permission of
Basic Books, Inc., Publishers.)

tradition, it nevertheless remains at least partially an empirical question.

This can be briefly illustrated here with an example taken from the recent
literature pertinent to the Bush Mekeo and other Melanesian cultures.
Brunton (1980) attempts to cast doubt upon the various symbolic inter-
connections in ida ritual that Gell establishes in his book, Metamorphosis
of the Cassowaries (1975), dealing with the Umeda culture of Papua New
Guinea. Gell elicits the meanings of the ida on the basis of two kinds of
"lexical motivation": "sematic motivation" or polysemy, where a word has
multiple meanings, metaphorical or metonymical; and "morphological
motivation," where a meaning is obtained from the compounding of dis-
tinct lexical elements (Gell 1975:121-3; cf Ullmann 1963). Conceding
that Gell produces the most sophisticated and detailed attempt of this sort
to date, Brunton is still not convinced. In the process of questioning some
of the ethnographic instances of motivation posed by Gell, Brunton
focuses upon the issue of formulating a convincing picture of the cultural
whole, as I have argued earlier in this chapter. He emphasizes that Um-
eda villagers themselves, on the one hand, offer few reliable clues; they
typically do not engage in exegesis of their rituals and whenever pressed
give highly individualistic, tentative, and varying interpretations. And,
he argues, on the other hand, that additional rituals of the same system,
by which Gell's symbolic analysis through comparison might be em-
pirically verified, are lacking (cf. Geertz 1974). Thus, judging Gell's ac-
count deficient on these counts, Brunton proposes that this and other
extreme instances of presumed logical order in cultural and religious
systems are the product of imposing an Aristotelian discipline foreign to
many Melanesian situations. Presumably, should an account of im-
pressive logical ordering in a Melanesian culture be offered, supported by
both sufficient native exegeses and a broad range of diverse rituals, and
10 Quadripartite structures
supplemented with analyses of even additional contexts of that culture,
then the sort of skepticism voiced by Brunton should be overcome. In any
case, the question of the relative orderliness and structuring becomes one
of empirical investigation as well as theoretical presupposition.
My interpretation of Bush Mekeo culture is intended to answer this
challenge on both counts. I shall show, first, that meaning and order are
intrinsic to the very nature of the culture-as-constituted (Sahlins 1976);
second, that meaning resides in the relational values (including, in Gell's
terms, the lexical motivations) of the indigenous culture categories; third,
that the order consists of a total system of conceptualizations ramifying
throughout the culture structured in the form of bisected dualities;
fourth, that because these relations are comprehensive of diverse contexts
or domains of the culture, they are not merely artifacts of the outsider's or
analyst's thought processes; and fifth, that comparisons with other histor-
ically related cultures as analogous wholes suggest the widespread dis-
tribution of this particular structure in the Oceanic cultural sphere. As
Willis aptly puts it, "What is needed . . . is not less structuralism but
more - at the 'grass-roots' level of ethnography" (1967:531).
Overall, then, my aim is to unravel the structural relations and replica-
tions among the indigenous categories of Bush Mekeo culture as a total,
unified system. Nonetheless, I do ultimately focus upon the categories
and relations underlying Bush Mekeo society, particularly as they are
represented in gender roles, kinship, clanship, and marriage, and in
rituals and feasts of death and mourning. To reach this point, however, it
is necessary to elucidate certain categories from other spheres that are
preliminarily invoked in the cultural ordering of social distinctions and
relations. Also, as implied here, the numerous contexts of the culture as I
describe them become increasingly complex. Therefore, it will prove
useful first to articulate fully the basic structure in one beginning context
of the culture and then to move on to others related to it. Not arbitrarily, I
have selected indigenous conceptualizations of space and time to serve in
the next chapter as this beginning context. Some of the key oppositions in
Bush Mekeo culture that express categories of space and time are vil-
lage/bush, inside/outside, resource/waste, ordinary/extraordinary, above-
ground/belowground, and, very importantly, the bisected subdivisions of
these.
Chapter 3 proceeds to reveal similar classificatory distinctions among
material things as they are ideally situated with respect to categories of
space and time, and especially the human body. In addition, the notion of
transformability expressed in terms of hot and cold will be introduced
here. Some of the relevant contexts that will be discussed in this chapter
include culinary procedures, work and nonwork, eating, digestion, excre-
tion, health, illness, and curing. A few of the pervasive categorical opposi-
 Introduction 11
tions are inside/outside of the body, sweet/unsweet, and bloody/blood-
less, along with their respective subdivisions.
Chapter 4 focuses upon a major ideological contradiction that is impli-
cated when certain relations coalescing around food, eating, digestion,
excretion, bodily health, and growth or life are transposed in the culture
to the context of sex, birth, and reproduction. In this case, conceptualiza-
tion of the latter is similar to that of illness and death and the reverse of
health and life, rather than, as one might expect, the other way around.
By a close comparison of the indigenous theory of conception with those
ideas pertaining to excretion, menstruation, and parturition, the apparent
categorical contradiction is resolved in the culture, and the homology of
eating, health, sex, and reproduction is preserved.
The terms of this contradiction and its resolution are further repre-
sented in the indigenous classification of gender roles and in the activities
appropriate to each. In Chapter 5, I show, through an analysis of revers-
ible ritual state alternations, that Bush Mekeo adult males and females
resolve basically the same kind of contradiction arising from their comple-
mentary sexual identities. In this way, the bisected categories male and
female are shown to be homologous with the categories discussed in the
previous chapters.
In the last three substantive chapters, the various persistent category
distinctions revealed in other contexts of the culture are shown to be
represented in the classification of Bush Mekeo social relationships: cog-
natic kinship, agnatic clanship, kofuapie affinity, exchange friend, and
hereditary clan office. In Chapter 6, I show that embedded in Bush
Mekeo society there exists a contradiction resembling that discovered in
the comparisons of ideas concerning eating and sex and female and male,
and this contradiction is expressed in the classification of interpersonal
relations and in the exchange of marriage compensation. Briefly stated,
the contradiction takes the following form: All members of the ideally
endogamous tribe are "one blood" and acknowledged relatives to one
another, but marriage rules prohibit persons of one blood from inter-
marrying. Resolution of this contradiction at the societal level is achieved
in rituals and feasts having to do with death and mourning. Parallel with
the resolutions noted above, the deceased and survivors are "de-con-
ceived" or "re-conceived" (Mosko 1983), both replicating and reversing
the conceptualized transmissions of blood that individuals and groups
accomplished earlier at the moment of conception in sexual reproduction.
The death ritual and feasting described at length in Chapters 7 and 8 are,
in other words, homologous with, or a metaphor of, marriage and re-
production and the other contexts of the culture expressed in terms of the
same juxtaposed, categorically bisected dualities.
Once I have completed my description and interpretation of Bush
12 Quadripartite structures
Mekeo culture as a "total social phenomenon" within this schema of
persistent bisected dualities, I shall explore its potential comparative use
by reexamining other well-documented Oceanic cultures that are lin-
guistically and historically related to the Bush Mekeo. For this purpose, I
have selected the Trobriand Islands from Melanesia and Tikopia from
Polynesia. As far as possible, I employ both classic and contemporary
materials. In each case, nevertheless, a number of significant and here-
tofore unrecognized ethnographic and interpretive insights are proposed.
Finally, in the light of these empirical investigations, I reconsider in
the last chapter the several theoretical issues I have already raised here.
Most especially, I shall argue that the structure of bisected dualities
characteristic of Bush Mekeo and related cultures (if not elsewhere) pos-
sesses a particular efficacy for analytically encompassing such distinctly
diachronic phenomena as social action and historical event. Indeed, a
structuralist approach such as the one I am suggesting here, following
Sahlins (1976, 1981) and Kelly (1977), is neither antithetical to considera-
tions of history and diachrony nor merely complementary to them.
Rather, a prior consideration of any culture as a structured whole can well
make definitive, positive, and otherwise unique contributions toward re-
solving issues distinctly historical in nature. In other words, the value of
synchrony even for studies of diachrony in anthropology (Saussure not-
withstanding) has not yet been exhausted.
The data-gathering methods on which this study is based consist largely
of standard ethnographic procedures (e.g., Malinowski 1922:1-25; Evans-
Pritchard 1962:64-85). Over the twenty-six months of my stay among the
Bush Mekeo (April 1974 to June 1976), I acquired a certain proficiency in
the indigenous language; conducted intensive interviews; participated in,
observed, and recorded all aspects of village life; conducted a survey and
census of all households in one village; constructed genealogies linking
the members of the community to one another; and kept a daily journal. A
major share of the material presented in the following chapters, however,
consists of idealized statements about village life, and of articulations of
the indigenous culture categories gleaned through intensive interviews
with knowledgeable key informants. It is worth emphasizing that these
last-mentioned kinds of empirical information are particularly amenable
to the task of constructing a synchronic model of the traditional cultural
system as manifested in the contact situation.
One important digression concerning empirical method, and change
and continuity as well, is at this moment in order. Bush Mekeo social life
has in several respects dramatically changed from the aboriginal condition
over the past ninety or so years of contact with Europeans. Nevertheless,
there does exist in the contemporary situation a very strong and pervasive
continuity with the past. This particular cultural tradition, moreover,
 Introduction 13
retains a remarkable degree of coherence despite the multiple exigencies
of the colonial presence. Recent ethnographic and historical studies ex-
plicitly confirm this impression for the neighboring Central Mekeo
(Hau'ofa 1971, 1981; Stephen 1974), and I think it is all the more demon-
strable with respect to the Bush Mekeo, given their generally less intense
and less direct exposure to external forces.
But the issue here is still to a considerable degree theoretical as well. It
concerns the Saussurean precept of the priority of synchrony over di-
achrony, or as Levi-Strauss (1966a) has stressed, the priority of structure
over event (see also Sahlins 1976 regarding the contrast of culture-as-
constituted vs. culture-as-lived). I incorporate into my analysis certain
elements of Bush Mekeo tradition that, as a result of historical changes,
are no longer exercised, but that continue to be viable in the conscious-
ness of villagers and thereby retain their significance in the contemporary
culture. The conceptualization of "warfare" (aoao) as distinct from its
present-day nonoccurrence is an obvious example.2 In the precontact
past, warfare was both an intrinsic ideational component of the total
culture and an occasional event. With pacification, of course, warfare in
the latter sense has disappeared. It no longer occurs as an event, and I did
not witness it in the field. But in the former sense, warfare continues to
play its traditional role. Indeed, its conceptualization still contributes
significantly to the definition and comprehension of other elements in the
contemporary culture (many of which are observable as events), and they
to it. I would argue in the same vein, moreover, that no complete under-
standing of any postpacification developments and changes in Bush
Mekeo social life could be achieved without some appreciation of the
nexus of ideas concerning traditional warfare and its place in the structure
of the total cultural system. Once more, it is upon this specific theoretical
issue that I dwell, especially in my concluding remarks in the final
chapter.

The Bush Mekeo

Before I begin the actual juxtapositioning of the categories and relations
indigenous to the culture, however, it might well prove useful to orient
the reader to the Bush Mekeo in terms already familiar. Thus, in the
remainder of this introductory chapter, I shall locate the Bush Mekeo
very generally in our own conceptions of space and time. I shall describe
who they are, where they live, what conditions they lived under aborig-
inally, and finally how those conditions have changed, broadly speaking,
with contact and colonial rule up until the time of Papua New Guinean
independence.
The Bush Mekeo are an Austronesian-speaking riverine people of the
main island of New Guinea. Today they live much as they did before
14 Quadripartite structures

Map 1. The Bush Mekeo.

Europeans arrived, in consolidated villages scattered along the middle

reaches of the Biaru River and its tributaries that drain the Mount Yule
range and empty into the Papuan Gulf (Map 1). This area, measuring
approximately 250 square miles,3 is predominantly swamp or grassland,
depending on the season, except that toward the northern edge near the
foothills of the mountains the terrain changes to virgin and secondary rain
 Introduction 15

Table 1. Village and tribal populations

Amoamo tribe Kuipa tribe

Engeifa village 228 Ameiaka village 170
Ioi village 130 Apanaipi village 578
Maipa village 225 Papangongo village 181
Inaukina village 192
Piunga village 137
Totals 583 1,258

forest. The total population of the Bush Mekeo - 1,841 persons as of 1970
(Bereina 1970-1) - has been traditionally divided among two political
units or "tribes": the Amoamo, among whom I conducted most of my
fieldwork, and the Kuipa. Each tribe speaks a dialect of a common lan-
guage. The two dialects are distinguished only by three or four dif-
ferences of sound, and they are mutually intelligible. As far as I was able
to ascertain, the respective cultures of the two groups are virtually homo-
geneous. Nonetheless, a perpetual state of war prevailed between the
two tribes aboriginally, and they did not intermarry. Traditionally, then,
each tribe was essentially endogamous, and peace reigned only within it.
Open warfare is now, of course, outlawed, but otherwise these intertribal
relations and the verbalized attitudes that are consistent with them still
characterize the contemporary Bush Mekeo scene.
Each tribe consists of several villages. In general, Bush Mekeo villages
are considerably smaller, simpler in plan, and more widely scattered than
the comparatively metropolitan villages of the Central Mekeo. The cen-
susfiguresfor Amoamo and Kuipa villages as of 1970 are listed in Table 1.
Traditional subsistence activities include swidden agriculture along the
banks of rivers and streams, hunting, and fishing. The stable crops are
sweet potato, banana, taro, and coconut. Fish and game (wild pig, casso-
wary, wallaby, and bush fowl and other birds) abound in the bush and
supply villagers with a relative abundance of protein. Pigs as well as dogs
are domesticated and roam the village freely, but they are eaten only
infrequently and principally on ceremonial occasions.
Before the introduction of village constables and councillors under the
colonial regime, there existed no overlapping village-level authority. In-
stead, power and authority were (and still are) largely vested in several
categories of specialized, hereditary clan office. Ideally, each clan repre-
sented in a village possessed its own peace chief, peace sorcerer, war
chief, and war sorcerer. Eldest sons according to rule succeeded their
fathers. Although the qualities of diffuse and effective leadership were not
16 Quadripartite structures
irrelevant to hereditary clan authority, the Bush Mekeo pattern does not
conform to the stereotype of the Melanesian "Big Man" (Sahlins 1963), as
has been noted also of the Central Mekeo (Hau'ofa 1971, 1981). As with
official succession, clan recruitment and membership are conceptualized
in terms of patrifiliation and descent. The several patrilineal clans of the
tribe are ideally exogamous, but the tribal entity, as mentioned above, is
basically endogamous.
The immediate neighbors of the Bush Mekeo include the culturally and
linguistically related Central Mekeo and Waima-Kevori (Roro) tribes of
the subcoast and coast, respectively, the non-Austronesian-speaking
Toaripi of the Papuan Gulf, and the Kuefa and Kunimaipa of the moun-
tains to the north (Map 2). Traditionally, interethnic relationships be-
tween these groups were dominated by intermittent peace and war. Dur-
ing eras of peace, the Bush Mekeo participated in trade exchanges with
their non-Bush Mekeo neighbors. From the mountain peoples, the Bush
Mekeo received weapons and tools of polished stone, dogs' teeth, and
feather ornaments. The Bush Mekeo in turn supplied their subcoastal and
coastal trade partners with some of these articles, along with smoked bush
meat, cassowary bone implements, barkcloth, and bird of paradise skins
they produced themselves. Shell ornaments of several varieties, lime,
clay pottery, and areca nut were traded in the opposite direction, passing
through the Bush Mekeo on their way from the coast to the mountains.
The Amoamo Bush Mekeo4 were first contacted by Europeans in the
year 1890 following a raid they perpetrated on the Central Mekeo (Veke)
village of Ififu that left twenty of the latter dead. The administrator of
British New Guinea himself, Sir William MacGregor (1890:88-90), hap-
pened to be patrolling in the Central Mekeo at the time. On hearing news
of the attack, he trekked through the swamps to the Bush Mekeo and
"pacified" the Amoamo without further bloodshed.5 From that time until
Papua New Guinea achieved self-government and independence in the
mid-1970s, the Amoamo have been officially administered by the British
(up until 1906) and the Australians.
Both these administrations, particularly the latter, have been noted for
their relatively lenient and paternalistic approach to "native affairs" and
"development" (Mair 1970). This is especially the case as regards the
Bush Mekeo - not out of administration benevolence, however, so much
as neglect. Surrounded by swamps on three sides and the perilous moun-
tain range on the fourth, the Bush Mekeo have been physically insulated
from many of the direct outside influences experienced by their nearest
Papuan neighbors. Unlike the Central Mekeo, for example, the Bush
Mekeo never had to pay the annual "head tax" until they were brought
into Kairuku Local Government Council jurisdiction in 1963. They were
never required to plant either coconuts or rice for government or mission.
 Introduction 17

Map 2. The Bush Mekeo and their neighbors.

Villagers were supposed to dig and maintain latrines and to bury their
dead in cemeteries away from the village, but for many years these pol-
icies and others were neither regularly practiced nor enforced. After
nearly thirty years of contact, one patrol officer noted that the govern-
ment still lacked much control among the Bush Mekeo (Humphries
1923:233). According to my own informants, the only serious intrusion
18 Quadripartite structures
government agents made into village life (beyond maintaining the peace)
until World War II and into the 1950s involved the recruitment of adult
males to carry loads for infrequent patrols going into the mountains to
pacify the Kuefa and Kunimaipa peoples.
Unquestionably, the Catholic mission (Sacre Coeur) has had a greater
impact on village life than government, but here again these effects have
been restricted to a limited sphere. Unlike their Protestant counterparts
established elsewhere along the Papuan coast, the priests, nuns, and lay
brethren of the mission historically adopted a comparatively more en-
lightened, tolerant, and patient approach to traditional Bush Mekeo vil-
lage life (Dupeyrat 1935; Belshaw 1951). One of the priests stationed at
Beipaa in the Central Mekeo would periodically patrol the Bush Mekeo
villages, but several months might pass between visits. Thus, by 1920,
the same patrol officer mentioned above notes that the mission was hardly
more influential than government in the villages. Aside from converting
the people to Christianity, introducing the sacraments, and more recently
providing health care and primary education (gains that are indeed im-
pressive), the church seems to have been only fairly successful in coun-
teracting the few traditional practices that are incompatible with its ten-
ets: adultery, polygyny, and sorcery, in particular. The one traditional
area where the mission has been considerably more successful is in
lessening the severities of physical abuse directed against widows, wid-
owers, and other persons in mourning. Moreover, mission activity in the
Bush Mekeo seems not to have been particularly intense until sometime
in the 1940s or thereabouts, when the Protestants of the London Mission-
ary Society worked up the Biaru River from the gulf and tried to establish
a foothold in the Bush Mekeo from which they could strike up into the
mountain villages. It was at this point, apparently, that the Beipa'a priests
focused their zeal on the Bush Mekeo.
Despite pacification, depopulation, colonial administration, and mis-
sionization, then, exposure to the outside world for the Bush Mekeo up
until the time of World War II was generally mild in comparison with the
experiences of their immediate coastal and subcoastal neighbors (cf. Ste-
phen 1974).
The war and its aftermath, of course, had a great impact on the lives of
villagers, and change since that time has been ever-accelerating. More
and more the Bush Mekeo have been brought into the stream of local,
regional, and national developments. During the war, many of the men
who are now among the elders had served as carriers for the Allies on the
"road" to Bulldog and Wau, and along the Kokoda Trail. The rapid
growth of Port Moresby and other urban centers in the country since then
has offered new opportunities for wage employment. Most of the men
now living in the village spent at least one or two of their bachelor years in
 Introduction 19
the towns doing menial labor of various sorts before returning home to
marry. During the 1960s, the Bush Mekeo began to vote in local and
national elections. It was during this period also that villagers planted
their first cash crops (coffee and cacao) and first attempted to run local
businesses. A number of trade stores were opened in each village, and a
few local entrepreneurs purchased outboard motors for their canoes to
transport passengers and cargo to and from the Gulf. In the middle and
late 1960s several villages even cooperated to clear and maintain an air-
strip. For a year or so a small Port Moresby airline firm sent in weekly
flights to buy garden produce, which was later to be sold to the urban
expatriate population. Early in the 1970s, Department of Agriculture
agents organized villages and clans in the Bush Mekeo to begin work on
cattle-raising projects. In 1972, after two earlier and unsuccessful at-
tempts, a seasonal feeder road linking the Hiritano Highway from Port
Moresby to the border of the Bush Mekeo (Papangongo village) was
finally completed. Villagers now had unprecedented access to the town,
and especially to its lucrative areca and betel trade.
Although these several recent innovations and changes have had a
major impact on Bush Mekeo village life, it is noteworthy that almost
every one of them has been short-lived. By 1974-6, the coffee and cacao
were completely neglected and overgrown, the airstrip had reverted to
grassland, few motorized canoes still operated on the rivers, only a couple
of trade stores intermittently opened for business, and the cattle projects
never progressed beyond the state of marking land boundaries. The Port
Moresby areca market remained the principal source of cash, but only a
few of the Bush Mekeo villages produced much beyond their own subsis-
tence needs.6 Those who did produce a surplus often could not transport
it to market except in the dry season when the market was already glutted
and prices were their lowest. Given these disappointments after a decade
or so of enthusiasm for economic "development," it is perhaps not sur-
prising that village life among the Amoamo Bush Mekeo in the early and
mid-1970s was perceptibly dominated by a low-key revival of traditional
forms, especially as regards subsistence production, and mourning and
feasting obligations and reciprocities.
It can be justifiably argued that, in spite of recent changes, the postwar
era has generally witnessed an extension of the prewar contact pattern.
Exposure to the outside world for the Bush Mekeo has continued to be
markedly more tempered and more gradual than has been the case for
many other Papuan groups. Throughout their entire history, the Bush
Mekeo have had none of their lands appropriated, no expatriate planta-
tions have been established nearby, and there has been a general lack of
European business and enterprise in the area. Until my family and I
arrived in 1974, no European had even come to live among the people.
20 Quadripartite structures
To their own chagrin, in short, the Bush Mekeo have been historically
relegated to the backwaters of Papuan development. The waves of abrupt
change that struck the coast and subcoast populations under colonial rule
were hardly more than ripples by the time they reached the Bush Mekeo.
In most respects, villagers have yet to be brought "out of hiding" to join
the rest of Papua, the nation, and the modern world.
 2
Between village and bush

Bush Mekeo villagers comprehend their own location in the world and in
the universe very differently from the way we Westerners might view
either ourselves or them. Indeed, the Bush Mekeo see themselves and
their world in quite distinct terms. It will be most useful, then, to begin
the description of Bush Mekeo culture and society by examining their
notions about the ordering of space and time and their own place in it
(Durkheim and Mauss 1963; Evans-Pritchard 1940; Fortes 1970). I shall
introduce a number of indigenous ideas that nonetheless approximate
certain familiar paired concepts fundamental to anthropological thought.
"Village" and "bush," for example, roughly parallel Levi-Strauss's opposi-
tion of culture with nature (1966b, 1969a, 1969b). Similarly, what I shall
distinguish as "ordinary" and "extraordinary" corresponds in many re-
spects to the Durkheimian separation of the sacred from the profane. I
hope to show that human social life as the Bush Mekeo view it is predi-
cated upon the transferral of certain objects between the village and the
bush that surrounds it, and that there are two separate spheres of these
transferences: the ordinary and the extraordinary. These two spheres are
distinguished according to divergent temporal frames as well as by partic-
ular subdivisions of space that cross-cut the village/bush axis. At the end
of the chapter, I shall argue that space and time for the Bush Mekeo
represented as ordinary and extraordinary spheres of village/bush object
transfer are fundamentally structured according to the bisection of dual
oppositions. This initial conclusion will set the stage for my handling of
other related contexts of indigenous cultural and social interrelationships,
for in subsequent chapters I shall deduce structures of homologous form.
Permit me one cautionary note at the outset. In certain respects, con-
nections between some of the indigenous categories I discuss may initially
contradict our intuitive English-speaking (and perhaps even our an-
thropological) understandings. For instance, the village is described as an
"outside" place in opposition to the bush, which is "inside." This will
seem all the more paradoxical as, in the language of the Bush Mekeo, the
center space of the outside village is termed its "abdomen." Hopefully,
this warning and my explanatory remarks in the text will prevent undue
confusion.
21
22 Quadripartite structures

Village and bush - outside and inside

Two distinct kinds of places, "village" (paunga) and "bush" (ango aonga),
compose all the world as the Bush Mekeo view it. The village is, of
course, the human place. Were it not for the human labor of clearing it
out of the bush, there would be no village. Humans, say the Bush Mekeo,
are "village beings/creatures/things" (paunga aunga). They live at the
village and eat, sleep, and reproduce there. Indeed, the overwhelming
share of daily human interactions are village affairs. Children's play, ado-
lescent courting, adult-female domestic activities, and the public pro-
ceedings of married men are all conducted at the village. The more
infrequent and spectacular events of Bush Mekeo life - marriage ex-
changes, installations of hereditary officers, mourning rituals and feasts,
and even intertribal warfare - are planned and staged here as well. More-
over, just as the interests of humans are concentrated at the village during
their lives, the soil of the village eventually and permanently encloses
their bodies in burial. In short, the village is the locus of Bush Mekeo
social life.
But the village is not a place limited to humans. Other things are also
village beings/creatures/things: domesticated animals (pig, dog, fowl, and
nowadays the cat), various plants (including coconut, areca, and bread-
fruit), and a whole range of human-fashioned material objects (dwellings
of several sorts, household implements and tools, weapons, etc.). Hu-
mans, then, are not the exclusive claimants to village status.
A well-defined croton hedge bounds the periphery of the village from
the other, larger part of the world - the bush beyond. The bush is divided
into numerous named and geographically differentiated places: primary
and secondary forest of the mountains and coastal plain, garden, grass-
land, swamp, and river or stream. Although humans do not live in these
places, the bush provides the natural setting for other beings and things.
These include the many species of wild animals (atsi), nondomesticated
and domesticated plants, and mineral resources. In addition, some cate-
gories of spirit beings and dreaded human sorcerers are said to inhabit
certain parts of the bush. All these are "bush beings/creatures/things"
(ango aonga aunga).
The village and its inhabitants are far from self-sufficient. Although
human life as the Bush Mekeo know it is predicated upon the conceptual
and physical separation of themselves and their place from the surround-
ing bush, villagers depend upon the bush for an astonishing array of
material resources. The bush serves, moreover, as the final repository for
the "rubbish" or "wastes" (kamakama) of village life. Thus, various cate-
gories of village and bush things are transported between the two places
despite their separation and distinction otherwise.
 Between village and bush 23
These transfers between village and bush are conceptualized and or-
dered in terms of "inside" and "outside," by further spatial partitioning of
the village and of the bush, and by the temporal regulation of the daily
cycle.
The language of the Bush Mekeo opposes the village and the bush as
"outside" (afanga)1 and "inside" (aonga), respectively. Ango aonga, the
indigenous term for bush, literally means "land inside." An alternative
expression for village is ango afangai, or "land outside." As with placing
anything inside a receptacle, a person going from the village to the bush
"goes inside" (isa ekoko). Moving in the opposite direction from bush to
village is described as "going outside" (isa ebualai). Furthermore, the
creation of a new village involves clearing it "out" of the bush, as I
implied earlier in this section.
This construal of the village as outside and the bush as inside reverses
the natural or intuitive judgment of many English speakers. For most of
us, I think, the town (or city) is generally "inside" and the country is
"outside." We live "in town," but upon heading for the countryside we
"get out of town." Now in certain other English-speaking contexts, the
reverse would seem also to hold. Rural dwellers, for example, live "in the
country." But the expression "out in the country" is equally acceptable,
whereas living "out of the country" (i.e., "in town") and "out in town" are
less, if at all, intuitively sensical.
Nonetheless, at least one context in English is comparable to the Bush
Mekeo associations of bush to inside and village to outside. For English
speakers, one unambiguously goes "into the woods" (or forest) and comes
"out of the woods" to town. For example,
Lets hide in the woods,
or
We're not out of the woods yet.
Perhaps these familiar expressions will serve some readers as mnemonic
devices in grasping how, in Bush Mekeo culture, village and bush are
analogously related as outside to inside.

The daily cycle and ordinary village/bush transfers

In the hum and activity of village life, there is a temporal patterning and
regularity that synchronize transfers of things between village and bush.
The two directions of flow are linked to a daily cycle and thus segregated
from one another. In the early morning, villagers remove accumulated
wastes to the bush, that is, from outside to inside. As evening approaches,
villagers carry fresh or raw resources in the opposite direction, from bush
to village, from inside to outside.
24 Quadripartite structures
Two of the more prominent features of the morning routine involve
bodily elimination and the clearing of the village. Only nursing infants,
who do not yet know better, and the domesticated animals that roam the
village may defecate and urinate on village grounds. All others must
retreat into the surrounding bush. Moreover, this collective movement is
linked to the initial hours of each day.2 With the first hint of dawn,
solitary villagers in a staggered procession head for the secluded parts of
the bush nearby (maka), empty their abdomens of feces and urine, and
several minutes later reappear at the village to get on with the other
activities of the new day. However, if urination only is called for, it is not
necessary to seek such privacy - remaining in the sight of others, a man or
woman needs only place him- or herself in breaks of the bounding croton
hedge and aim into the bush.
While the sporadic procession into and out of the bush continues, the
members of each household energetically devote themselves to cutting
new sprouts of kunai grass from village grounds with bush knives. After an
hour or so of this, women and girls scoop the cuttings along with other
discarded rubbish (areca rinds, coconut husks and shells, animal feces,
food scraps, firewood chips, etc.) into piles on the flat central area of the
village. The term for this clear open space free of trees, dwellings, or
other structures is paunga inaenga, or "abdomen of the village."3 The
females of each household then empty the village abdomen of its collected
rubbish. They deftly carry the debris to the edge of the village and toss it
over the croton hedge into the bush. Each village household is responsi-
ble for the grounds surrounding its house, including that portion of the
village abdomen immediately in front of it.
The clearing of people's and the village's abdomens of wastes therefore
characterizes the early-morning routine.
Late afternoon and evening contrast with morning by object movement
in the opposite direction (Appendix 1). With village clearing complete by
midmorning, most villagers set off for other pursuits in the more remote
parts of the bush. Some members of each household head for their gar-
dens. A few men of the village often go hunting or fishing, alone or in
small groups, deep into the forest. Parties of women and girls may spend
the day wading through creeks or along the river bank with their nets
searching for prawns or fish. A few people will stay back at the village -
the old and feeble, the very ill, newly delivered babies and their mothers,
brides, and those in mourning seclusion - but, for the most part, they,
staying in their houses, are not to be seen. Even the pigs and dogs leave
the village to forage in the surrounding bush or, in the case of the latter,
to hunt with their masters in the forest. As the afternoon wears on, those
who left earlier begin straggling out to the village burdened with their
 Between village and bush 25
quarry: vegetables, meat and fish, firewood, and various materials for
house and canoe manufacture or other village industries.
During the remaining hours of evening and into the night, with the
stock of resources replenished, the village and its inhabitants begin the
process of accumulating new rubbish and wastes in their respective ab-
domens for bush disposal the next day. Women rekindle their hearths and
prepare the evening meal. Married men gather on chiefs' clubhouses
discussing events of the day while they wait for their wives to serve them
their food. Bachelors haunt the margins of the village together hoping for
the chance to impress their loves, and young children play. After the men
have been fed, the women serve food to the others of their own house-
holds and to the pigs; finally, they too eat. Women frequently carry an
extra bowl of cooked food to a neighboring relative's house. As darkness
gradually overtakes the village, the adults relax with their talk, tobacco,
and areca. The drone of voices and laughter sinks, and, eventually, every-
one retires to a night of sleep.
Thus, each day resources are taken from the inside bush to the outside
village, and the next morning wastes of the village are sent back into the
bush. Resources move in one direction, wastes in the opposite direction.
But because the two movements occur in distinct periods of the day, and
village wastes are thrown away immediately before fresh resources are
brought from the bush, the chance for the mixing of the two kinds of
objects is minimized. Moreover, resources and wastes, whether inside
the bush or outside at the village, are spatially separated from one an-
other; that is, resources are extracted from regions of the bush where
village wastes are not deposited (and vice versa), and bush resources are
stored in parts of the village where wastes derived from them do not
accumulate (and vice versa). Village and bush are therefore not uniform
places.
This last point requires some elaboration.
Inverted outside and everted inside
By virtue of the daily transfers of objects between village and bush,
outside and inside domains are bisected by a reversal or inversion of each,
such that the outside village has its own inside place (i.e., an inverted
outside) and the inside bush has its own outside (i.e., an everted inside).
The Bush Mekeo village is ideally erected around the central, elon-
gated abdomen. Clubhouses of the resident clan chiefs stand prominently
at the ends. Domestic dwellings, platforms, and other structures are
arranged in parallel rows facing one another across the width of the ab-
domen. The abdomen itself is entirely clear of permanent buildings and
features.4 Even the towering coconut and areca palms of the village are
I / adjacent bush

Figure 2.1. The sphere of ordinary transfers. Bush resources —> village resources; village
resources —» village wastes; village wastes —> bush wastes.
 Between village and bush 27
planted behind the lines formed by the facades of adjacent buildings
(Figure 2.1; cf. Hau'ofa 1981:49-76). Materials brought from the bush at
evening are taken to the peripheral part of the village. Here, on the raised
floors of platforms, houses, and clubhouses, the resources are processed,
transformed, and consumed or otherwise utilized (Chapter 3). Only the
wastes and by-products of resources originating in the bush are tossed to
the ground and, in the morning, swept to the abdomen of the village.
The abdomen as part of the village is outside, but in a number of
contexts including this one the abdomen is regarded as inside. Rubbish
swept to the abdomen goes inside (isa ekoko). The abdomen is then
primarily or initially an outside place, and secondarily inside. It is as if the
village were topographically inverted, turned or pushed within itself,
separating and distinguishing the abdomen from the rest of the village
around it. Villagers may sweep their rubbish into the abdomen, but for all
that it is still outside, that is, occupying village space and not that of the
bush.
Designating the village abdomen as inside may appear at first paradox-
ical. However, the rule that enjoins excretion directly in the bush sug-
gests that the abdomen of a human being is homologously conceived as
inside the body only insofar as it is an inversion of space outside the body.
If other nonbodily rubbish and wastes must be swept into the village
abdomen before final deposition in the bush (I frequently wondered while
in thefield),why do villagers not also deposit and collect their own bodily
excreta on the village abdomen and thereby integrate more closely both
processes of waste removal? The answer, I think, is that feces and urine
collected in (sic) the abdominal cavity are already outside the body. By
emptying their abdomens directly in the bush, bodily wastes move from
outside to inside homologous with the transfer of nonbodily wastes from
village to bush.
In subsequent chapters dealing with indigenous theories of food and
nutrition, sexuality, reproduction, and certain categories of social rela-
tionship, I shall be describing other relations whereby the abdomen (or
stomach, pregnancy, womb, mother) is similarly and systemically concep-
tualized in Bush Mekeo culture as an "inverted outside."
The bush as a spatial entity includes its own analogously "everted"
region - the band of land (maka) used for waste disposal immediately
surrounding the village. There is no direct transfer of things, due at least
to human intervention, between this and the more remote part of the
bush. Resources extracted from the latter are taken only to the peripheral
inhabited village; and abdominal wastes are deposited only in the former.
The absence of intrabush transference contrasts with the regular move-
ment of things directly between the two village zones, that is, from village
periphery to village abdomen. Yet, because of this, villagers indirectly
28 Quadripartite structures
transport resources from the remote bush to the adjacent bush after they
have been transformed at the village into wastes. Therefore, although
wastes moved from the village to the bush (from outside to inside) reverse
the direction by which bush resources are taken to the village (from inside
to outside), the complete process is also unidirectional:

bush village village bush

resources resources wastes wastes
remote peripheral village adjacent
bush village abdomen bush
Moreover, each crossing of the village/bush threshold is a movement from
inside to outside if inverted and everted subdivisions are included:
inside —» outside —> inverted outside —» everted inside
The daily cycle, then, expresses a series of homologies systematically
relating conceptualizations of space (village and bush, inside and outside)
with things (resources and wastes):
bush village village bush (2)
resources resources wastes wastes
remote peripheral village adjacent
(3)
bush village abdomen bush
inverted everted
inside : outside : (4)
outside inside

Extraordinary village/bush transfers

The daily unidirectional transfer of ordinary resources and wastes is not
entirely comprehensive; that is, there exists another opposing sphere of
village/bush intertransfer that is set apart. To the extent that the things
and beings, human or otherwise, involved in daily transfers are ordinary,
the beings and things implicated in this other, temporally more irregular
sphere are extraordinary. Also, where everyday transfers are unidirec-
tional, the displacements of extraordinary beings and things are hi- or
ambidirectional. Nonetheless, the categories of place involved in this
extraordinary sphere are homologous with those of the ordinary everyday
resource and waste cycle. The village/bush duality, in other words, is
similarly bisected, but along an additional spatial dimension: above-
ground versus belowground.
Human beings (papiau) possess either "spirits" (tsiange) or "souls"
(laulau), depending upon whether the body is dead or alive, respectively.
Human spirits and souls inhabit and animate bodies and their leavings,
particularly the bloody ones. Because they incorporate a spirit or soul,
human blood and flesh (ifa) are considered "dirty" (iofu) and "hot" or
 Between village and bush 29
"potent" (tsiabu) for causing sickness, death, and other misfortunes, es-
pecially if they are eaten (Chapters 3-5). Also, a person's soul can be
attacked in certain ritual contexts if a knowledgeable specialist secretly
acquires some of that person's blood or the blood of his or her recently
deceased relative. The blood or flesh of a dead human being can also be
used in certain other ritual contexts to attack nonrelatives.
Villagers never send the blood or flesh of their own bodies in any form
to the bush. Rather, upon leaving the body or upon the body's death,
human blood and flesh are placed and contained within certain specific
regions of the village, termed "holes' (ine). In this way, villagers isolate
human spirits and souls from everyday affairs and minimize the potential
for misfortune. And once placed in a village hole, human blood and flesh
are never casually removed.
Blood and bloody tissues excreted by the living, as for example in
childbirth and chronic illness, are buried in catchments dug underneath
domestic dwellings. Wherever else human blood and flesh may fall on
village grounds, villagers scoop up the soil with wooden digging sticks or
shovels and turn it over and underground. People bury the bodies of the
deceased in "graves" (ongo) dug in the soil of the village abdomen or
beneath their houses.5 By contrast, resources and nonbloody wastes and
excreta of daily life are not buried in these underground village holes.
The village also embodies a catchment hole for human blood and flesh
aboveground - the rear compartment (ialiali) of chiefs' clubhouses. For
months on end, adult males who are preparing or practicing various types
of aggressive ritual (war sorcery, courting, garden, etc.) and adolescent
bachelors who are learning these arts seclude themselves in the compart-
ment away from women, children, and ritually unprepared men. These
male ritual initiates perform their extraordinary acts by manipulating
charms, weapons, or other objects (depending upon the category of ritual
involved) that are contaminated with, or contain, bloody human remains.
These bloody charms and weapons are never removed from the rear
compartment unless they are to be employed. Conversely, everyday re-
sources and wastes are systematically kept out. (This aspect of male ritual
separation will be discussed in greater detail in Chapter 5).
The bush also contains a number of holes segregated from the sphere of
ordinary daily transactions with the village.
One category of practicing ritual specialist, the "peace sorcerer" (un-
gaunga) (see Chapters 5 and 6), is excluded from the compartment of the
chief's clubhouse and, indeed, from the entire village. Peace sorcerers
base their operations in special ritual dwellings called fauapi, which are
erected away from the village in the adjoining bush (maka). Similarly to
village specialists, peace sorcerers prepare, manipulate, and store their
nefarious charms and paraphernalia in their fauapi dwellings. These
30 Quadripartite structures
charms and other objects include stolen bloody human remains among
the critical ingredients. No ordinary villager - man, woman, or child -
would dare enter a sorcerers bush dwelling unless, in the case of an adult
male, he first takes the appropriate and elaborate precautions. Moreover,
the rules that isolate the contents of sorcerers' dwellings from ordinary
resources and wastes are identical to those I mentioned earlier in this
chapter concerning the ialiali compartments of peace chiefs' clubhouses.
The more remote bush contains its own category of holes. These holes
harbor two kinds of "spirits" (tsiange) in animal forms. The spirits of dead
humans take the form of deadly poisonous snakes {afi, death adder; and
auama, Papuan black snake), which live in holes under trees, rocks, or
fallen logs. Another nonhuman kind of spirit, termed ongokapu, takes the
form of certain animals known as faifai. Pythons, eels, catfish of several
varieties, some species of frog, fresh-water prawns, and perhaps several
other species are commonly regarded as faifai.6 Faifai animals, like poi-
sonous human spirit-snakes, live in holes in the ground, or in holes un-
derwater (beneath waterfalls, in depressions of the stream bed, under
stones, or under logs left by the flood).
These underground and underwater holes of the remote bush and their
inhabitants are also associated with the blood and flesh of death. The flesh
of poisonous spirit-snakes and faifai animals is said to be exceptionally
bloody. Supposedly, blood copiously flows when they are cut open. The
spirits animating poisonous spirit-snakes are, of course, those of dead
humans. Ongokapu spirits offaifai animals are also associated with death.
The term ongo means "grave." Also, villagers say that although ongokapu
spirits in faifai animal form do not have poison like human spirit-snakes,
"they have other ways of killing you." In addition, human and ongokapu
spirits may both abandon their animal bodies of their own volition and
become invisible.
Villagers insist that ongokapu spirits are distinct from spirits of the
human dead. Virtually nothing concrete is known or ventured concerning
the origin of ongokapu, except that they did not originate from the death
of humans. Nonetheless, ongokapu, or rather their corporeal faifai animal
forms, are conceived as beings of the bush analogous to human villagers.
Villagers speak euphemistically of ongokapu and faifai as "bush peo-
ple" (angoaonga papiau). Among their various faifai guises, ongokapu
may appear in peoples dreams as humans, but with white skins, long
straight hair, a large head, or some other gross deformity. Ongokapu keep
their own domesticated "pigs,"7 and, like humans, they are social beings.
Villagers consider the more expansive and permanent bush holes (be-
neath waterfalls, deep pools in the river bed, etc.) as ongokapu "villages,"
and the boulders strewn nearby as ongokapu "houses." Many people
have said to me they have heard ongokapu singing, dancing, drumming,
 Between village and bush 31
and feasting at their bush villages like humans do at theirs. Ongokapu
communicate with one another in their own spoken language, and when
they appear to villagers in dreams they take human shape and speak the
Bush Mekeo language. Some faifai species (python, prawn) are explicitly
recognized to shed their skins and grow new ones like humans reportedly
did in mythical times (see Chapters 6-8). Ongokapu sometimes evince
certain humanlike emotions such as anger, much as offended human
spirits of the dead do in response to human trespass and the theft of their
pigs. Consequently, they too can cause villagers to suffer illness, run
amok, and die. One specific illness caused by ongokapu, for example, is
"python sickness."8 An invisible snake wraps itself around and constricts
its victim, so movement is impossible. Similarly, when a sorcerer un-
earths a fresh grave, the corpse will supposedly rise and try to wrap its
arms about him.
Lastly, pregnant women who mistakenly eat the flesh of faifai animals
risk giving birth to a faifai baby. Ongokapu are said to "spoil the baby"
(imi ebalifua) in these cases, so the baby is born dead and, often, with a
large head or other faifai deformity. A middle-aged woman of the village
gave birth to just such a stillborn hydrocephalic faifai infant during my
stay. At another village, a woman reportedly gave birth to a faifai in the
form of a dead snake along with her healthy and otherwise normal baby
boy. According to indigenous categories, faifai animals possess a certain
limited and perverted potency for interbreeding with human females. I
shall return to this issue offaifai sexuality momentarily.
In sum, all the things and beings villagers associate with holes in the
village and in the bush are bloody, are human or humanlike, and possess
extraordinary and spiritual powers. In virtually every case it is at least
implicit that the extraordinary or spiritual powers are dangerous to ordi-
nary humans living at the village. Sorcerers and other ritual specialists,
the spirits of the dead, ongokapu, human corpses, faifai, and the eviscer-
ated blood of the latter and living humans must, therefore, be isolated in
holes away from the sphere of ordinary daily village life.
Aboveground versus belowground
Each of the four places in the world resulting from the bisection of the
village and the bush according to the ordering of ordinary resource and
waste transfer has its corresponding hole or holes. Moreover, no one of
the four spatial categories may have a hole appropriate to one of the
others. Human villages (along with the bands of bush land immediately
adjacent to them) are purposefully erected where ongokapu and faifai are
known not permanently to reside. Peace sorcerers may not erect their
fauapi at the village nor in the remote bush, and so on. These correspon-
dences between categories of holes and the other categories of space are
32 Quadripartite structures
maintained in part according to a distinction of aboveground versus un-
derground. The hole or holes appropriate to each kind of place at the
village and in the bush are either aboveground or underground, but
never both. Remote bush holes occupied by poisonous spirit-snakes and
faifai are underground (or underwater). The peace sorcerers' fauapi in
the waste-disposal region adjacent to villages are aboveground. laliali
clubhouse compartments of the occupied peripheral village are above-
ground. Finally, graves are dug in the ground beneath the village
abdomen.
Graves and catchments for human blood and flesh dug directly under-
neath domestic dwellings might be thought initially to occupy the periph-
ery of the village; thus, in contradistinction to the pattern described
above, the periphery of the village would appear to include holes both
aboveground and belowground. But graves and catchments underneath
houses are homologous rather with the ground underneath the village
abdomen. Human corpses may be buried in either location without ap-
parent discrimination, except that, when burial occurs below the village
abdomen, a small house or platform is erected over it to seclude the
mourners (see Chapter 7; cf. Guis 1936:127). When bodies are buried
beneath peripheral village houses, mourners will occupy the house di-
rectly above. More significantly, the bloody tissues collected in holes
underneath houses fall from the abdomens of postpartum and chronically
ill villagers straight down. Corpses and bloody tissues alike, in other
words, fall into underground holes directly beneath village and bodily
abdomens. Although domestic dwellings and the rear compartments of
chiefs' clubhouses9 stand together in the periphery of the village, holes
dug in the ground for catching blood and flesh under the former are
homologous with holes underneath the central village abdomen. This
homology, moreover, is comparable to the daily emptying of feces and
urine from human abdomens directly into the bush rather than first into
the village abdomen (see above; cf. Young 1971).
Thus, there is a hole, or category of holes, associated with, but set apart
from, each of the four bisected village and bush places. The locations of
these holes above- or belowground (or underwater) serve to bisect the
village and the bush along a distinct vertical dimension that, although it
involves extraordinary beings and things as opposed to ordinary ones, is
nonetheless congruent with the horizontal spatial orientation of ordinary
resource and waste location and transport:
. #1
inside :
.,
outside ::
inverted
.,
everted
. .,
...
(4)
L
outside inside
remote peripheral village adjacent ,~v
bush ' village " abdomen ' bush
 Between village and bush 33
bush village village bush ,^
underground aboveground underground aboveground
Just as village and bush each have their own inside and outside regions,
village and bush each possess corresponding holes both aboveground and
belowground. Furthermore, moving from left to right, it will be noted
that underground and aboveground holes are horizontally next to one
another just as each inside and outside place as defined in the ordinary
daily sphere is next to its opposite.
Although holes of all categories are spatially isolated from one another
and systemically set apart from the sphere of ordinary, everyday village
life, there is a certain amount of interchange and transfer between them.
Extraordinary beings and things occasionally move or are moved from one
kind of hole of the village or the bush, aboveground or belowground, to
another. This has been implied a number of times already in the foregoing
discussion. Peace sorcerers leave their fauapi bush houses near the vil-
lage and steal human flesh and blood from graves and catchments under-
neath human and village abdomens. Peace sorcerers also require faifai
animal tissues from remote bush underground holes for some of their
deadly practices, and it is claimed they also keep poisonous snakes in
their fauapi houses and entertain spirit-snakes there. When a peace sor-
cerer temporarily retires from his practice, he stores his various materials
in his chiefs clubhouse compartment, or he buries them secretly under-
ground in the remote bush. As sometimes happens, especially since the
introduction of shotguns, a peace sorcerer shot wandering away from his
fauapi is buried in a grave dug in the remote bush. Under certain condi-
tions, a ritual specialist isolated in a clubhouse compartment at the village
will go live with a peace sorcerer in his fauapi. Bachelors and widowers in
the clubhouse compartment, who have ritually prepared themselves for
courting ritual but who lack the appropriate spells or charms, are often
reputed to serve as helpers or apprentices of peace sorcerers who will
help them in return. Indeed, any adult male who takes the ritual steps
enabling him to enter the ialiali may also (and with relative safety, appar-
ently) enter a peace sorcerer's fauapi. Like peace sorcerers, ritual spe-
cialists of the peace chiefs clubhouse compartment must have the bloody
remains of poisonous snakes or faifai animals for some of their techniques.
In one kind of destructive garden magic, the specialist plants faifai re-
mains in the ground of remote bush gardens, for example. War sorcerers
and other specialists secluded in clubhouses do use certain kinds of
bloody human excreta (e.g., desiccated human penis, semen, menstrual
blood), but these are neither taken from graves nor from catchments
underneath houses of the village. Finally, while I was living at the village,
a faifai (python) was once killed trying to raid a village chicken coop. Its
34 Quadripartite structures
body was immediately buried in a grave dug underneath the village
abdomen.
These examples illustrate a number of significant aspects about the
transfer of beings and things between different holes of the extraordinary
realm.
First, only extraordinary beings such as sorcerers, ritual specialists, and
human or nonhuman spirits intervene in or instigate these events - not
ordinary villagers, who are actually threatened by their occurrence. The
man who buried the python after it was killed, for instance, was a peace
sorcerer, although he was not practicing at the time. Also, in the particu-
lar case of a sorcerer shooting reported to me, the accused murderer was
widely regarded in the area as a peace sorcerer himself.
Second, this extraordinary sphere is distinctly "male" or "maleish."
Only adult male humans may become sorcerers or any of the other kinds
of ritual specialist I have mentioned who practice in either bush fauapi or
clubhouse compartments. Faifai species and spirit-snakes have undeni-
ably phallic connotations. They possess elongated bodies, live in holes,
and have as regards the former a certain sexual potency with human
females. Although the ordinary sphere of daily resource and waste trans-
fer is not exclusively "female" or "femaleish," it is the only context of the
two whereby women actively participate, intervene, and predicate object
movement from place to place (e.g., carrying garden produce home to the
village, preparing food, sweeping rubbish into the village abdomen, and
depositing it in the bush). Conversely, in the extraordinary-realm, village
women are only passively involved; they are always acted upon by men
and other maleish beings, never the reverse. I shall elaborate upon vari-
ous additional aspects of this gender distinction in relation to the ordinary
and extraordinary spheres at a number of points in subsequent chapters.
Third, transfers of extraordinary beings and things must cross over the
space ordinary humans occupy in the daily routine. As these beings and
things move from hole to hole, they present a danger to the human
population of the village. Whenever this occurs, villagers take immediate
action to return the threatening beings and things to holes where they
will not present further danger. On the other hand, transfers of ordinary
beings and things from place to place in the daily cycle presume move-
ment entirely exclusive of holes. Thus, the ordinary and the extraordinary
spheres are maintained separate, or set apart from, one another.
Fourth, movements between holes are ambidirectional, not unidirec-
tional like transfers of the ordinary sphere. The diverse kinds of transfers
between holes listed above illustrate that extraordinary beings and things
may move in either direction, that is, parallel with resources and wastes
or in reverse. Also, transfers between holes are not necessarily ordered in
linear sequence. Some of them "skip" or "short-circuit" steps of the daily
 Between village and bush 35
resource/waste spatial progression. Faifai that have left their under-
ground bush holes and come to the village, for example, are killed and
then buried in ground beneath the village abdomen rather than in an
aboveground hole (i.e., clubhouse compartment) of the village periphery.
Fifth, like the ordinary sphere of transfer, the extraordinary sphere has
a beginning and an end, but in the latter case this boundary exists at the
village rather than in the bush. In the daily routine, recall, resources are
taken from the remote bush to the village, then finally deposited in the
bush adjacent to the village. There is no direct, ordinary transfer between
remote and adjacent bush zones. In the extraordinary sphere, objects are
moved in both directions between remote and adjacent bush, but not in
either direction between village periphery and village abdomen. This is
because ritual specialists in clubhouse compartments do not rob graves in
order to acquire their bloody human remains. A war sorcerer, for exam-
ple, must take the penis for his charm from the corpse of an enemy slain
in battle, that is, from a dead body that has not yet been buried under-
ground. Nor is there any burial in the village abdomen of things and
beings taken directly out of the clubhouse compartment. This distinction
of extraordinary as opposed to ordinary village/bush relations is repre-
sented in Figure 2.2. The coincidence of this boundary at the point where
the village is bisected will have significant implications in the context of
social relationships conceived through males as opposed to females
(Chapters 6-8).
In these several respects, the sphere of extraordinary transfers between
holes oppose and contradict the ordinary, everyday, unidirectional
sphere of resource and waste transfer. Holes are spatially set apart from
everyday village life, and the two realms are ordered and conceptualized
in contradistinction to one another. Nonetheless, the two are homolo-
gously structured, that is, according to the bisection of the village/bush
duality:
:: :
remote bush : .,? ii adjacent bush (3)
village abdomen
Ordinary Sphere
(daily time, hole-less places)
. ., ., inverted everted /AS
inside : outside :: ., : . .i (4)
outside inside
Extraordinary Sphere
(irregular time, holes)
bush village village bush ,_*
underground ' aboveground " underground ' aboveground.
In more abstract terms, both spheres take the same form:
X' : Y" :: Y' : X" (1)
 remote bush

r adjacent bush

peripheral village

f village abdomen
ialiali

>
j H grave
fauapi
S-4 i
J i

\
\
\
Ti i
/
v\
faifai house/village

Figure 2.2. The sphere of extraordinary transfers. Bush belowground -^a—^village

aboveground; village aboveground <—b—•> bush aboveground; bush belowground ^c—> village
belowground; village belowground ^d—» bush aboveground; bush belowground - ^ e ^ bush
aboveground.
 Between village and bush 37
I must emphasize that the form of homologously bisected dualities I
have revealed here also structures the categories and relationships of the
remainder of Bush Mekeo culture and society. This assertion must, of
course, be tested against empirical observation, and it is to this task I
devote the following chapters. However, even at this preliminary stage of
the investigation, there is one sense in which it is possible to foresee this
conclusion. The village and the bush, as well as the spheres of ordinary
and extraordinary transfers that together crosscut or bisect them, are
utterly comprehensive. Beyond them - in the world traditionally oc-
cupied by the Bush Mekeo - there is nothing else. All seemingly "other"
contexts of Bush Mekeo cultural and social life are in essence themselves
integral aspects of the same categorical relations. Whatever additional
relationships, discriminations, boundaries, or oppositions the Bush
Mekeo have carved out of their experience, all are embedded in and part
of the same unified reality, and they in turn are structured accordingly.
Moreover, if the culture of the Bush Mekeo is indeed so ordered and
systematic, as I am assuming, there should be a certain homogeneity of
cultural or semantic content as well as structure whether the specific
contexts in question are village versus bush, ordinary versus extraordi-
nary, or, as we shall soon see, body versus nonbody, plant versus meat
food, bloody versus nonbloody, food versus sex, female versus male, kin
versus nonkin, and so on.
In the next chapter, keeping these ideas in mind, I shall expand upon
the various conceived subsidiary processes by which villagers transport
objects between village and bush and transform them from diverse sorts
of ordinary resources into equally diverse kinds of wastes.
 Body and cosmos

The ordinary and extraordinary spheres in Bush Mekeo culture are predi-
cated upon the bisection of the inside bush and the outside village. How-
ever, every spatial transfer in either of these spheres involves the
bisection of yet another inside/outside duality - the human "body"
(kuma). As the world is composed of village and bush, there is an outside
to the body and an inside. I suggested twice earlier that the outside and
the inside of the body have inverted and everted parts - abdomen and
excrement, respectively - that are homologous with the village abdomen
and the adjacent bush. Thus, everything transported between village and
bush either remains outside the body, goes into the body and stays there,
passes to the abdomen, or comes out of the body. Moreover, bodily
tissues set apart in holes of the village and bush are extracted from holes
of the body. The overall purpose of the present chapter, then, is to
delineate the various aspects of this homology between conceptions about
space and the body in both the ordinary and extraordinary spheres of
Bush Mekeo culture.
My handling of indigenous conceptualizations of bodily processes will
be considerably more complex than was the case with village/bush rela-
tions. This is unavoidable because every object transfer relating to spatial
distinctions also involves a corresponding transformation or change of
state, and these transformations cannot be described independent of how
they relate to the inside and outside of the body. Also, the body, or rather
distinctive aspects of the body, is variously conceived as either the agents
of transformation, the objects of transformation, or, in some cases, both.
Therefore, it will be necessary to distinguish systematically among nu-
merous transformers and the correlative transforms in a certain variety of
contexts. Most critically, the interrelation of these contexts involving
conceived bodily states and processes will be dependent upon the distinc-
tions among four categories: "sweet" (mitsia), "unsweet" (etsiu), "hot
dirt" (iofu tsiabu), and "cold dirt" (iofu ekekia). Sweet and unsweet as two
types of the clean1 are opposed by the two types of the dirty - hot and
cold.

38
 Body and cosmos 39

The clean, the dirty, and the human

In at least two significant respects, sweet and unsweet are differentiated
from one another relativistically. First, all living creatures in the world
are visualized as distinguishing between what they regard as sweet and
"like" (mitsia), and what they consider unsweet and "dislike" (etsiu). But
what is sweet and unsweet varies between species and, in certain condi-
tions, even between members of the same species. The village as a place
to live is generally sweet to humans and domesticated animals, for exam-
ple, and the bush is unsweet. However, undomesticated animals find the
bush sweet and the village unsweet. So among all the diversity of living
beings, the sweet and the unsweet vary with regard to spatial and ecologi-
cal habitat, food preferences, intraspecific behaviors, interspecific in-
teractions, night-versus-day activity, weather, season, and so on. The
complete depiction of Bush Mekeo animal taxonomy, then, would partly
involve discriminations according to complementary sets of sweet and
unsweet for each animal category.2
Second, sweet and unsweet are relative as to context. With respect to
the same being or category of being, a thing may be sweet in one context
but unsweet in another. A mans food and his axe, for example, are both
sweet to him, but only in different situations. Food is sweet for eating, the
axe is sweet for working in the garden. But axes are unsweet for eating
and food is unsweet for felling trees. Furthermore, food and axes are both
unsweet as objects for sexual gratification; here, by contrast, a man's wife
is sweet to him.
Thus, categories of animals and humans are differentiated from one
another according to contrasting sets of sweet and unsweet predisposi-
tions relativistically ascribed to them in all contexts of their respective
lives.
This extreme relativity of sweet and unsweet might give the misleading
impression that the Bush Mekeo categorization of humans and animals
(and particularly the boundary separating them) is arbitrary or capricious.
For two reasons, however, this is not so. First, what is sweet or unsweet
to each category of human being - male and female, young and old, kin
and nonkin, married and single, etc. - is standardized by traditions (ka-
ngakanga) inherited through many generations from the ancestors who
lived long ago. Whenever I queried a cultural observance, the first and
seemingly automatic explanation informants offered was simply, "That is
the way our ancestors [au apoutsi, or "old men"] did it." To do as the
ancestors did is itself sweet to their present-day descendants. When vil-
lagers portray in their actions sweet and unsweet dispositions appropriate
to their respective statuses, it is "good" (verlo); when they do not, it is
40 Quadripartite structures
"bad" (abala). The sweet and the unsweet carry a strong component,
then, of the ideal and the moral for human beings. The only animal
species that possess inherited traditions are the faifai and ongokapu, or
"bush people." In this respect, the sweet and unsweet constellations of
humans and bush people are distinct from all other living creatures.
Second, the conceiving of an opposition between clean and dirty abso-
lutely differentiates human beings from all animals, including bush peo-
ple. Other creatures of the village and bush are seen by villagers to show
no sign of distinguishing dirty things from clean things. Human feces and
flesh, for example, are cold and hot dirt, respectively, and therefore are
entirely inedible to humans. Yet, if and when they have the opportunity,
village pigs and dogs will eat either, as well as the flesh and feces of their
own species, as if they were sweet. All other nonhuman living things
including faifai animals supposedly fail to distinguish any category of dirt
beyond the clean. Thus, the conceptualization of the dirty in opposition to
the clean serves to uniquely differentiate human beings from all non-
human beings.
In this connection, it should be noted that for humans there is generally
no relativity between hot and cold dirt comparable with that of sweet and
unsweet. Hot and cold dirty things are uniformly designated as such for all
humans in all contexts but one: mortuary feasting. I shall return to this
important point later in this chapter and in Chapters 7 and 8.
With these preliminary remarks about the contrasts between sweet,
unsweet, and hot and cold dirt kept in mind, I turn to the relations
between conceptualizations of the body and of the world.
Village resources: sweet/unsweet
The village and its resources are sweet to human beings. The bush and its
resources, on the other hand, are generally unsweet. Nonetheless, vil-
lagers are largely dependent upon the unsweet resources of the bush to
provide them a considerable portion of their total assemblage of sweet
village resources. Appendix 1 gives a partial but fairly representative list
of these village resources. Even articles that the Bush Mekeo do not make
themselves from local bush resources - clay pottery, some kinds of shell
and feather ornament, polished stone axes and adzes, etc. - are recog-
nized as originating in the bush elsewhere.
Unsweet bush resources are adapted to village life by the application of
human "skills" (etsifa), which transform them into sweet village re-
sources. But because there are different kinds of village and bush re-
sources and human skills are so diverse, there are many different ways of
transforming the unsweet into the sweet. Nonetheless, among all these
there are fundamentally two general ways that unsweet things are or-
dinarily transformed into sweet things and integrated into village life.
 Body and cosmos 41
Villagers are scrupulous about what they will and will not allow to enter
their bodies. Some unsweet bush resources are made sweet for humans
and passed into their bodies by "eating" (earn), "drinking" (einu), or
"chewing" (euwa), for example. In this context, it is significant that "cook-
ing" (emitsi) is the predicate of sweet (mitsia). Other unsweet things are
also made sweet, yet they remain outside the body and are not ingested.
In this distinction, things sweet for putting inside bodies are unsweet for
keeping outside, and things sweet when kept outside bodies are unsweet
for ingesting.
The variety of things sweet for eating, drinking, and chewing is im-
pressively large, and includes the many kinds of garden food and animal
meat along with water, tobacco, salt, milk, areca nut, betel pepper, lime,
and certain ritual and curative substances (see Appendix 2).3 Of these,
only the first three - "plant food" (fokama), "meat" (tsitsi), and "water"
(ivi) - are considered critical for daily bodily sustenance. The daily meal,
even to the most discriminating palate, ideally consists of chunks of plant
food and meat boiled and served in water and coconut oil. Every house-
hold strives to sustain its members and guests in this way as often as
possible.
Plant food, meat, and water are "big" or "important" (apoutsi) and
unlike the other kinds of ingestibles because the human body is under-
stood to require at least minimal quantities of each to survive. Although
villagers often remark that sorcerers and other ritual specialists "eat or
drink nothing" as part of their rigorous "tightening"4 procedures, they
will acknowledge that even these practitioners must eat some food ("just
one bite") and drink some fluid. This also suggests that no one of the three
staples is a possible substitute for either of the other two. A person could
eat enormous quantities of plant food with water, for example, yet die of
meat starvation. This will be explained more fully in a subsequent section
of this chapter.
Food and water transformations
The everyday method of "boiling" (eanga or epanganga) chunks of plant
food and meat in clay pots transforms these "raw" (maisa) staples from
unsweet to sweet so they can be eaten. However, boiling is only one skill
among four of food preparation. Staple vegetables and meat can also be
"roasted" (euma) directly over smoldering coals. Some kinds of plant food
can be consumed without prior cooking once they "ripen" (eaipa) and
mature. Most meats can be more or less indefinitely preserved by con-
stant "smoking" or "drying" (eongongo). Thus, to comprehend the indige-
nous processes of transforming the unsweet to the sweet in the context of
ingestion, it will be necessary to first compare these four skills with one
another and later examine their contrasting implications for bodily suste-
42 Quadripartite structures
cold/hot
(no fire/fire)

unsweet „ sweet
(wet, cold) " (wet, hot)

Figure 3.1. Water transformation: boiling.

cold/hot
(no fire/fire)

unswe »sweet
(wet, cold) (dry, hot)
Figure 3.2. Water transformation: roasting.

nance. The analysis that follows has been most significantly inspired by
Levi-Strauss's article "The Culinary Triangle" (1966b; see also Tambiah
1968a).
The features that critically differentiate sweet from unsweet as regards
food preparation and transformation are hot versus cold in addition to wet
versus dry.
The simplest place to begin is with water. Water is "wet" (ekome) and
may be drunk whether it is hot or cold. But villagers will rarely drink
plain cold water because it is also unsweet. To be sweet for drinking,
water must be boiled as the broth in which food is cooked (or nowadays as
tea), or it must be taken from coconuts or wild banana stalks that have
been roasted. With boiling, water becomes sweet and remains hot (Fig-
ure 3.1). With roasting, however strange it may seem, water becomes dry
as well as sweet (Figure 3.2). Whether by boiling or by roasting, un-
sweetened cold water is transformed to sweet hot water by "fire" (ito).
Fire is hot, and it is this "hot" that changes the unsweet cold water to
sweet hot water. It is also significant to note now, for I shall return to this
point later in the chapter, that sweet and unsweet states of water are
temporary. Allowed to cool away from fire, the sweet water reverts to
being cold and unsweet.
Most raw staple foods, both plant and animal (Appendix 3), can be
 Body and cosmos 43
cold/hot
(no water/water)

unsweet sweet
(dry, cold) ^ {wet, hot)

Figure 3.3. Food transformation: boiling.

cold/hot
(no fire/fire)

unsweet > sweet

(wet, cold) (dry, hot)
Figure 3.4. Food transformation: roasting.

transformed from unsweet to sweet when they are boiled, together or

separately, in water. After the hot of fire transforms the water from cold to
hot and unsweet to sweet, the hot of the water transforms the chunks of
food from cold to hot and unsweet to sweet. Like the water they are
cooked in, boiled foods are wet. Also, boiled foods allowed to cool before
they are eaten will revert to being cold and unsweet, but they will remain
wet (Figure 3.3).
Nearly all the categories of plant and animal food that are ordinarily
boiled before eating may also be roasted (Appendix 3). Here, the hot of
the fire directly transforms the pieces of raw cold unsweet food to hot
sweet food that is edible. Unlike food transformation by boiling, with
roasting there is no mediation by hot water. Thus, roasted plant food and
meat are dry. When left uneaten, roasted foods will remain dry, but they
revert to being cold and unsweet to humans (Figure 3.4).
The remaining two categories of food processing - ripening, and smok-
ing or drying - are not sweet/unsweet transformations as such (i.e., by
mediation of hot), but they do involve analogous changes of state from wet
to dry, or dry to wet.
In smoking or drying, butchered meat (never plant food) is placed atop
a wooden platform approximately a meter or so over a low, smoldering
fire. The fire is hot, but the meat remains cold and unsweet, or un-
44 Quadripartite structures
cold/heat
(no fire/fire)

unsweet • unsweet
(wet, cold) (dry, cold)
Figure 3.5. Food transformation: drying

transformed. Smoking merely dries the meat. Indeed, smoked meat is

considered still raw. In this instance, fire's temperate "heat" (pan-
gaingai), not its hot (tsiabu), dries the meat. Smoked meat must be subse-
quently boiled in water like other raw meat for it to be transformed into
hot, sweet, wet, and edible food (Figure 3.5).
Villagers cultivate a number of plant foods that may only be eaten raw
(coconut, pineaple, paupau, watermelon, sugarcane, almond fruit, ripe
banana, etc.). The fruit of these plants is unsweet, cold, dry, and inedible
while it is immature, but allowed to ripen in the garden it becomes sweet,
villagers say. This requires some qualification.
According to my fieldnotes, it is the "sun" (tsina) that produces this
change of state, and I once presumed (quite erroneously, I think) it to be
the hot (tsiabu) of the sun rather than its heat (pangaingai) that ripened
the plant foods. Lacking further and more conclusive data, this in-
terpretation must be revised because ripening, like smoking but unlike
cooking, does not make these foods hot (tsiabu); they remain cold.5 The
fully ripened fruit of these plants, I am arguing, is said to be "sweet" (sic)
only insofar as it is wet. Otherwise, it is untransformed and unsweet.
Here is an instance of the relativity of sweet and unsweet, so I must be
extremely cautious. In its unripened, unsweet condition, the fruit is dry.
Ripened, it becomes merely wet. This change corresponds to the simul-
taneous change of seasons. Villagers clear their gardens at the beginning
of the "dry season" (April to October). After the fallen trees have been
fully dried in the sun toward the end of the dry season, they are burned
off. At the onset of the "wet season" (November to May), the gardens are
planted. The fruit of edible raw plants thus ripen during the season when
there is comparatively less direct sunlight but more heat from the sun as
well as more moisture. It would seem that ripened fruit is changed from
dry to wet by the heat of the sun, according to the culture, much as wet
raw meat is dried by the heat of fire, except in reverse (Figure 3.6). Wet
ripened plant food and dry smoked meat are therefore both cold and
unsweet.
 Body and cosmos 45
cold/heat
(no sun/sun)

un sweet < . unsweet

(dry,coId) (wet, cold)
Figure 3.6. Food transformation: ripening.

In a later section of this chapter, I shall discuss another critical feature

of the category sweet as regards food and water, namely, hot for blood
synthesis. The fact that ripened plant food and smoked meat are alike cold
and unsweet in this context and boiled and roasted foods are not, accord-
ing to indigenous Bush Mekeo views, supports my interpretation above
regarding ripened foods as unsweet.
The four categories of food processing and transformation can be dis-
tinguished from one another in the following terms:

Sweet Unsweet
roasting : boiling :: ripening : smoking (6)
hot dry : hot wet :: cold wet : cold dry (7)

The sequence of categories in these homologous statements is intentional.

I shall return to this point as well as to other relations of these culinary
skills once I have described the indigenous theory of digestion and blood
synthesis. For the moment, it is merely necessary to note that sweet foods
(roasted and boiled) are transformed by hot (tsiabu) without and with
water, respectively; and unsweet raw foods (ripened and smoked) are
changed by heat (pangaingai) with and without water, respectively.
Nonfood transformations: work and nonwork
Many categories of unsweet bush resources that are transformed into
sweet village resources are not cooked. Nor are they intended for eating
or for other kinds of bodily ingestion. Nonetheless, hot associated with
skills other than cooking is still the active, transforming agent.
In their pristine bush condition, stone, clay, shell, animal skin, feather,
bone, bamboo, wood, bark, leaf, vine, and so on are unsweet to villagers,
but upon their retrieval these things can be made into village resources
that are sweet to human beings (tools, weapons, charms, pottery, dwell-
ings, canoes, etc.). Moreover, these categories of village resources are,
with respect to manufacture and use, all sweet outside as opposed to inside
46 Quadripartite structures
the body. Each inedible village resource is made and used according to
specific skills. However, these skills are distinguished from one another as
either "work" (pinaunga) or "nonwork" (aisepinaunga, or simply etsifa
"skill").6 This distinction also applies to skills employed in culinary pro-
cesses described in the last section. Boiling food and smoking meat are
work skills, and roasting and ripening are nonwork skills. Other categories
of work and nonwork skills are listed in Appendix 4.
Work and nonwork skills are both hot, but they differ critically in how
the body participates. With work, the body becomes hot through its own
exertion, and this exertion directly transforms unsweet things to sweet
things. With nonwork, on the other hand, bodily exertion is cold or
irrelevant for transformation. The significant transformative hot of non-
work is nonbodily.
This contrast between hot bodily work and hot nonbodily nonwork
transformations can be best illustrated by comparing examples of each:
house building, and hunting with bow and arrow, respectively.
As a man's house shows signs of age and delapidation, it becomes
unsweet to him. The idea of a fine new house becomes correspondingly
sweet. The members of his household, sometimes with help from rela-
tives, gather building materials from the bush: timber for pilings and
frame, palm or bamboo for flooring, split sago branches for walls, kunai
grass or sago leaf for roofing, and split vine for joining all the various parts.
These bush things are initially unsweet. Gathering, transporting, trim-
ming, fitting, and binding them are separate skills. By virtue of their
planned combination, hot bodily exertions transform the unsweet bush
materials into a sweet village resource: the complete dwelling (Figure
3.7). The tools and other implements the house builders employ in their
work are also sweet village resources that have been transformed by
bodily work skills in their manufacture. Thus, the sweet products of work
conserve the hot of the active body and are themselves hot for transform-
ing other unsweet bush resources to sweet village ones in appropriate
work contexts.
To any European who has tried it, stalking game in the tall rain forest
(even nowadays when shotguns are used) is hardly less a physical effort
than any one of the skills employed in building a house. To Bush Mekeo
conceptions, however, hunting of all kinds is nonwork. As he stalks
through the steamy forest, the hunter's body does become perceptively
warm, but this is only temperate heat (pangaingai) and not hot (tsiabu).
The hunter successfully retrieves game another way, with another non-
bodily hot. It was explained to me this way by one of the best hunters in
the village:
Pigs [and other game] fear people. People are unsweet to them. That is why they
run away . . . so he [the hunter] possesses hot (tsiabu) hunting charms (kabukabu
 Body and cosmos 47
cold/hot

unsweet < •sweet

(bush resources) (village resources)

Figure 3.7. Work transformation: house building.

^ebodV

cold/hot

unsweet sweet
(toward humans) (toward humans)
Figure 3.8. Nonwork transformation: hunting.

tolina) and spells (menga). . . . The pig's unsweet becomes sweet toward humans
and it does not run away. It does not matter that you go walkabout every day. You
must carry hot charms or you will never shoot a pig.

This skill of hunting then consists in manipulating nonbodily hot objects

outside the body upon other objects likewise outside the body.
Hunting spells and the ingredients of charms are secret. The charms
are made from certain combinations of dried bush leaves (fuka). The
ingredients are unsweet to the hunter for eating but sweet to him for
hunting, and they are carried outside the body around the throat. The
manufacture of charms and the learning of spells are nonbodily or non-
work skills. And the bush resource they transform - the pig - is unsweet
to the hunter for ingestion; that is, he will not eat his own prey but
distributes it instead among his relations.
Hunting and other kinds of nonwork contrast with work skills such as
house building in that the relevant source of hot is nonbodily. Nonethe-
less, and this is a critical point, nonwork not less than work presumes a
distinction between the inside and the outside of the body. For work, hot
emanating from inside the body, or of the body itself, transforms things
outside the body from unsweet to sweet. With nonwork, hot manipulated
48 Quadripartite structures
cold/hot

unsweet > sweet

(raw food) (cooked food)
Figure 3.9. Culinary work transformation: boiling food.

cold/hot

unsweet 4 > sweet

(raw food) (cooked food)
Figure 3.10. Culinary nonwork transformation: roasting food.

outside and apart from the body transforms things that are likewise out-
side the body from unsweet to sweet (Figure 3.8).
The categories work and nonwork relate back to food transformation.
Fire and the sun are not the only sources of hot and heat involved in
culinary processes. Bodily work (gardening, carrying, skinning, the mak-
ing and fueling of fire, and the other manual exercises of cooking) and
nonbodily nonwork skills (especially gardening ritual) are also hot for
transforming inedible unsweet plant food to edible sweet plant food.
Similarly, game becomes sweet edible meat partially by the work of
carrying, butchering, and cooking after the nonwork of hunting has been
successful (Figures 3.9 and 3.10).
Bodily work and nonbodily nonwork alike are hot for transforming
unsweet bush resources into sweet village resources of both ingestible
and noningestible categories. The question remains: What criteria differ-
entiate those things that are sweet for taking into the body and those that
are sweet in contexts outside the body? The answer will involve examin-
ing indigenous theories of bodily digestion and blood synthesis according
to still other transformations of unsweet to sweet by mediation of hot, and
relating these back to the culinary skills of preparing and transforming
food.
 Body and cosmos 49

Food digestion, blood synthesis, and dirt

Cooking transforms inedible unsweet food and water into sweet edible
food and water outside the body. Once these transformed substances are
ingested, they are themselves hot for transforming the inside of the body,
or rather specific parts of it, from unsweet into sweet. It is in these terms
that villagers express hunger, thirst, and satiation.
In the language of the Bush Mekeo, "hunger" for plant food is ex-
pressed literally as "abdomen unsweet" (ina etsiu). After eating hot trans-
formed plant food, the abdomen becomes sweet (ina mitsia), and hunger
is "satiated." "Throat unsweet" (also etsiu), or "meat hunger," indepen-
dently becomes "throat sweet" (also mitsia) when hot transformed meat is
eaten. Villagers will often communicate their sensations of plant food and
meat hunger and satiation by gesturing to the appropriate parts of the
body - abdomen and throat, respectively.
Thirst for and satiation with hot boiled water are perceived in the
throat, like meat, but not entirely in terms of unsweet and sweet. Instead,
"thirst" is expressed as "throat closed" (also ekupu), and lack of thirst or
"satiation" as "throat unclosed" or "open" (aiso aisekupu). Nonetheless,
drinking hot sweet water does transform the inside of the body generally
(aonga) from unsweet into sweet (Figure 3.11).
Transformations of the body's abdomen, throat, and inside by ingested
hot food and water are homologous with transformations of food and water
outside the body by hot fire, work, and nonwork. These internal bodily
transformations are likewise only temporary. Whenever hot transformed
plant food, meat, and water are not ingested over several hours, the
abdomen, throat, and inside of the body, respectively, revert from the
sweet to the unsweet condition. Untransformed (i.e., raw, ripened, or
smoked) foods and water are cold rather than hot. Even if they are eaten
in substantial quantities, abdomen, throat, and bodily inside will remain
unsweet. Noningestible things that may be transformed and sweet to
humans in other contexts external to the body (e.g., a tool, house, article
of clothes) are likewise cold for transforming the body and its parts to
sweet. Also, because each category of sweet ingestible is hot only to the
specific body part, no one of them can be substituted for one of the
others. Some of all three sweet and hot staples are thereby regularly
required to sustain the life of the body.
Hot transformed plant food, meat, and water, beyond assuaging hun-
gers and thirst, sustain the life of the body by mixing together to form
"blood" (ifa), or, perhaps closer to the indigenous view, elements of plant
food and meat mix with water to form blood. In contrast, unsweet things
such as untransformed food and cold water and noningestibles like tools
and weapons are cold for synthesizing blood. The blood within one's body
is sweet, and the more blood the more sweet the body. Blood is itself one
50 Quadripartite structures
cold/hot
(no fire/fire)
MEAT

cold unsweet hot sweet

(raw meat) (cooked meat)

throat unsweet < • throat sweet

cold/hot
(no fire/fire)
PLANT FOOD

cold unsweet * hot sweet

(raw plant food) (cooked plant food)

outsidebodv^
~uSde body

abdomen unsweet abdomen sweet

cold/hot WATER
(no fire/fire)

cold unsweet . hot sweet

(fresh water) (boiled water) o u t s idebody_

body inside unsweet body inside sweet

Figure 3.11. Sweet/unsweet body transformations.

kind of meat or flesh (tsitsi), but distinct from the rest in that the latter are
made from it; that is, sweet blood is hot for making all other meat tissues.
Blood transforms them from unsweet to sweet, and they too increase, so
the body contains a great deal of wet, hot, bloody flesh. By restricting the
 Body and cosmos 51
cold/hot
(no fire/fire)

cold unsweet ^ hot sweet

(raw food (cooked
& water) food & water)

cold unsweet hot sweet

(throat/ (throat/
abdomen/ abdomen/
body inside) body inside)

cold unsweet ^ hot sweet

(lack of (abundance
blood) of blood)

cold unsweet hot sweet

(lack of (abundance
flesh) of flesh)

cold unsweet - ^ hot sweet

(non-work) (work)

&\&

Figure 3.12. Blood and flesh synthesis: capacities for work and nonwork.
52 Quadripartite structures
intake of transformed food and water, less hot blood will be synthesized,
and there will be less of all other hot sweet meat tissues. The body
generally becomes unsweet, thin, and dry. When the body possesses
plenty of blood, it is hot for work. When it lacks much blood, it is cold for
work but nonetheless capable of performing nonwork skills. Left entirely
without any one of the hot transformed staples over a prolonged period,
there will be no blood, and then the body will die (Figure 3.12).
This last point suggests that illness and death are reversals of the bodily
states of health and life defined by an abundance of blood. This can be
demonstrated through consideration of the category dirt, mentioned only
briefly at the opening of this chapter, and its relation to the processes of
digestion and excretion.
Villagers acknowledge that all bodily excreta are by-products or resi-
dues of the things that previously went into the body. These excreted
things are all termed "dirt" or "dirty" (iofu).
Villagers distinguish between those dirty things that are bloody and
hot, and those that are bloodless and cold. The members of these two
categories are listed in Appendix 5. Cold bloodless feces and urine are
regularly deposited in the bush. Other bloodless dirt of the body (sweat,
areca spittle, nail parings, hair, etc.) that are not excreted from the body's
abdomen are swept into the village abdomen and then discarded into the
bush.7 Blood does not ordinarily emerge from the body, but when it does
it is dirty and buried underneath the village abdomen.
Dirt, whether hot and bloody or cold and bloodless, is not the only
indigenous category of inedibility. Unsweet things, it will be remem-
bered, are also inedible. However, unsweet things are either transforma-
ble into sweet ingestibles or they are transformable and sweet to villagers
in contexts other than ingestion. Inedible bloody and bloodless dirt in all
its forms is a final state - the temporal end product of transforming and
ingesting food and water and excreting their residues.
Nonetheless, the categories bloody and bloodless dirt are conceptually
related to the categories sweet and unsweet. For although dirt must
never be eaten by humans, villagers told me numerous times, "dirt is
very much the same thing as food." This paradox can be explained by
comparing indigenous ideas regarding the differential effects of dirty,
sweet, and unsweet things upon the health, illness, and curing of the
body.
Health, illness, and curing
Transformed outside the body, hot sweet food and water synthesize the
body's blood after they are eaten. The body's blood is also hot and sweet
as well as wet. Blood and the meat or tissues it synthesizes, in turn,
remain inside the body and are not ordinarily excreted. Therefore, inges-
 Body and cosmos 53

tion of all three staples - plant food, meat, and water - transforms them,
or certain elements of them, into the blood and flesh of the inside of the
body:
Outside body Inside body
hot sweet food and water > hot sweet wet blood and flesh
The indigenous Bush Mekeo definition of bodily life or health, of
course, involves precisely this transformation of outside ingestibles into
inside blood and flesh. However, the life and health of the body predi-
cates a separate transformation whereby things inside the body are ex-
creted to the outside. Even foods and water transformed by the hot of fire
contain some unsweet residues that are cold for synthesizing blood. These
residues are ordinarily excreted from the abdomen to the outside of the
body as cold feces and urine as well as the other types of bloodless dirt.
The eating of cold unsweet bloodless elements of transformed food and
water is comparable in its effects upon the body to eating raw foods and
water that have not been transformed by fire. The eating and drinking of
these cold unsweet substances together will not synthesize blood for the
body, so they will be excreted as cold bloodless dirt. But because cold
water is a member of this category, the body will become temporarily
wet. The excretory transformation that is subsidiary to health and life,
then, involves regular excretion of cold and bloodless dirty wastes:
Abdomen Body inside-out
(inverted outside) (everted inside)
cold unsweet food and water > cold bloodless dirt
In the condition of health or life, then, the four categories of things
associated with blood synthesis and excretion correspond to particular
conceptualized spaces or regions of the body:
inside : outside :: inverted outside : everted inside (4)
blood : sweet food :: unsweet food : bloodless dirt (8)
Health or life also constitutes a temporal state. Occasionally, villagers
become "ill" (eisaoa; literally, "see inside"), and sometimes that illness
can be cured, restoring the body to health. But other times illness results
in death. Thus, illness, like health, is a temporal bodily state.
Villagers diagnose illness by observing the excretion of blood from the
body. In the state of health, a persons blood is hot and sweet while it is
inside the body. Once blood comes out of the body during illness, it
immediately becomes dirty. This dirty blood, however, remains hot be-
cause it can transform the health of still other persons into illness, and
they in turn will also excrete hot dirty blood.
According to native etiology, excreted human blood8 can cause illness
54 Quadripartite structures
in one of two general ways: (1) by ingestion as a "poison" (ipani), or (2) by
its ritual manipulation with respect to the intended victim's captured
"soul" (laulau). In either case, the hot bloody dirt reverses the bodily
transformations conducive to health, as described above. It will be in-
structive to examine both techniques of causing illness in this light.
After hot dirt is administered as a poison and the patient or victim
consumes cooked food and water, the hot sweet elements of the latter are
not synthesized into blood of the inside of the body. Instead, they are
excreted from the abdomen as blood. Simultaneously, cold unsweet ele-
ments of ingested food and water are not excreted as feces and urine but
rather are retained inside the body. Here it must be noted for later
reference that with illness the inside of the body becomes dry for the lack
or loss of wet blood, but it becomes internally wet with the fluids that
would otherwise be excreted as urine.
Thus, illness by poisoning is precisely the reverse of health:
Outside body Inside body
cold unsweet food and water > cold wet bloodless dirt
Abdomen Body inside-out
(inverted outside) (everted inside)
hot sweet food and water > hot dirty wet blood
inside : outside :: inverted outside : everted inside (4)
bloodless wastes : unsweet food :: sweet food : bloody dirt (9)
Hot dirty blood can also contribute toward illness in humans when it is
ritually manipulated from afar, that is, when the intended victim does not
actually ingest it. There are a number of conceptually distinct skills of this
sort, but they all have certain features in common. I shall use the skill
termed mefu to illustrate that the bodily transformations these techniques
are thought to cause are identical with those of poisoning.
With mefu, the soul of the victim is captured in a "bottle" (immature
coconut shell, gourd, clay pot) with a specimen of his or her bloodless
bodily dirt. The bottle also contains the hot bloody dirt of a dead human
being. As the bottle is hung over a fire, the hot dirty blood attacks the
captured soul; and as the soul suffers, so too does the body exactly as
described above when it is poisoned. The victim will excrete blood and
retain the substances that would ordinarily be excreted as feces and urine.
Some mention too should be made here concerning indigenous skills
for curing illness, for these demonstrate conclusively the opposed concep-
tualization of health and illness.
Most native "medicines" (mulamula) are plant materials extracted from
wild species of the bush. Raw areca nut, ginger, and chili, however, are
also medicines, and they are cultivated in most household gardens.
 Body and cosmos 55
Knowledge of which plant species may be used as medicines is generally
secret. Medicines, including areca, 9 ginger, and chili, are ordinarily un-
sweet; that is, they are cold for synthesizing blood and sustaining health
as they are not food. Yet, medicines are hot and sweet when they are
administered to people who are ill. From what I have been able to gather,
most medicines are boiled in water before they are consumed. Chili and
ginger, however, are both hot when eaten raw. Medicines work by cool-
ing the foreign blood that caused the illness and, in the case of poison,
purging it from the patient's body in his/her urine. Blood in the urine is a
sign of recovery from illness. Once the foreign bloody dirt has been
cooled and removed, the body will revert once more to the condition of
health. The patient will thereafter cease excreting his/her own blood and
resume excreting accumulated bloodless wastes, which will include the
medicines as well:

Outside body Inside body

hot sweet medicine and water > hot sweet blood
Abdomen Body inside-out
cold sweet medicine and water -» cold wet bloodless dirt
inside : outside :: inverted outside : everted inside (4)
blood : sweet medicine :: unsweet medicine : bloodless dirt (10)

To inflict illness on another person by the practice of mefu, the victim's

soul is initially captured by the theft of a trace of his or her cold bloodless
dirt, such as an article of worn clothing, discarded areca rind, cigarette
stub, etc. These leavings, usually termed "skin dirt" (fa iofu), can be used
to bring about illness only in the person they were obtained from. Also,
bloodless skin dirt cannot be used as a poison on anyone, because it is
cold. In these respects, a person's cold bloodless dirt is distinct from his
or her hot bloody dirt. It will be worthwhile to pursue this issue in this
and related contexts, for by doing so it is possible to establish fully the
significance of the indigenous distinctions between dirty and clean, and
hot and cold.
Villagers do not eat the bloodless excreta of their own bodies, or for that
matter of animals, because, as they explain it, these things are dirty.
Lacking blood elements of sweet food, cold bloodless dirt cannot contrib-
ute to blood synthesis or sustain bodily health. However, neither can it
alone cause illness when it is ingested or heated in isolation over a fire by
a mefu practitioner. Bloodless dirt is thus also cold in the context of
causing illness. Nonetheless, it can aggravate or complicate an illness that
is already in progress when it is eaten, just as it can be mixed with hot
bloody dirt as in mefu to the same effect (i.e., the victim's body will
56 Quadripartite structures
assimilate rather than excrete it). Cold bloodless dirt is effectual for caus-
ing illness, therefore, only when it is mixed with foreign hot dirty blood
inside the victim's body, or outside the victim's body but inside a sealed
bottle.
If cold bloodless dirt is eaten by a person who is not already ill, it
merely will pass out the body from the abdomen as do cold unsweet
elements of food and water. But unlike the latter, cold bloodless dirt
alone cannot even make the body wet with water. For these reasons,
apparently, there is no risk to humans when they eat the flesh of village
pigs and dogs that customarily do eat the feces of toddlers left on village
grounds or of adults deposited in the bush. None of the cold bloodless dirt
these animals eat is incorporated into their flesh. However, villagers take
every precaution that their domesticated animals never have the oppor-
tunity to eat hot bloody dirt in any form, human or faifai in origin. This is,
in fact, one native explanation for burial underground. Village animals10
are themselves immune to eating hot bloody dirt. But humans who eat
the flesh of animals containing it are not.
Although bloodless dirt is cold for causing illness, as are sweet and
unsweet things, villagers say they do not eat the former because they can
never be certain merely by looking that it does not contain traces of hot
bloody dirt. In this respect, then, cold bloodless dirt is distinct from both
categories of clean things - sweet and unsweet. Moreover, villagers never
know for certain that they are not already in the initial stages of an illness.
Here, cold bloodless dirt is not eaten so that fewer complications will
arise, if such is the case. In this additional respect, cold bloodless dirt is
distinct from sweet food and water but like unsweet things; that is, un-
transformed food and water and clean noningestibles can also aggravate
an illness if they are eaten, though perhaps to a lesser extent.
Turning away from this last point, it is now appropriate to compare the
notion of cold bloodless dirt with that of "rubbish" (kamakama).
Inedible village resources, as described above, are the products of
transforming bush resources from unsweet into sweet by mediation of the
hot of work and nonwork. These sweet objects are hot in their own right
for use and transformation in other kinds of work and nonwork. However,
just as the hot of sweet transformed food is only a temporary state, trans-
formed noningestible objects are only temporarily sweet and hot for their
intended uses. Things break, wear out, deteriorate, and so on.
In this process, human artifacts do not merely revert to an unsweet
condition. Through their use, these things accumulate deposits of dirt of
either the hot bloody or bloodless kinds. Weapons of war and ritual
charms of certain categories are dirtied by contact with human or faifai
blood. Other things - tools, clothes, string bags, areca rinds, culinary
 Body and cosmos 57
articles, etc. - accumulate an outer deposit of cold bloodless skin dirt.
And it is these that are distinctly termed rubbish. In my example of house
building earlier in the chapter, the old dilapidated house is not just
unsweet to its occupants; it is dirty to them. Moreover, as village re-
sources become dirty either way, they become dry. So, although non-
ingestible village resources do not pass into and out of human bodies, in
use they are transformed from clean to dirty. However, it is the kind of
dirt they accumulate - hot blood or cold bloodless - that determines
whether they are regarded as hot or cold for causing illness and death,
and correspondingly whether they are to be removed from the village
abdomen to the bush as rubbish or isolated within holes of the village.
The total range of material objects that are not ingestible by humans is
therefore categorized and ordered with relation to space and time as
edible things are with respect to the human body. It is in these terms that
I wish to pull together my conclusions for this chapter.

Body homologies
I must first return to the paradoxical assertion that "dirt is very much the
same sort of thing as food."
On the one hand, hot sweet food and water and hot bloody dirt are
temporal and final states, respectively, of one another. The former are
transformed into blood in the state of bodily health. With the advent of
illness or death, the blood is excreted as hot bloody dirt. Both, moreover,
are hot, yet possess transformative capacities that are the reverse of one
another in terms of health and illness causation when they are consumed.
On the other hand, cold unsweet food and water and cold bloodless dirt
are also temporal and final states of one another, respectively, but both,
being cold, lack transformative capacities in the contexts of health and
illness when they are ingested.
In my comparison of these four categories, I have tried to show that
each is associated with a particular region of the body in the circumstance
of health and with a different part of the body in the condition of illness.
Bodily health and illness, then, are the reverse of one another:
inside body : outside body :: inverted outside : everted inside (4)
Health
blood : sweet food :: unsweet food : bloodless dirt (8)
Illness
bloodless dirt : unsweet food :: sweet food : bloody dirt (9)
I have noted above that when humans ingest substances of these four
categories, the relative wetness or dryness and hot or cold of the body are
variously affected with the absorption or loss of blood or water. It is in
58 Quadripartite structures
these terms that the four culinary processes are homologous with the
categories sweet, unsweet, hot dirt, and cold dirt. The culinary skills can
be characterized:
Boiling food makes it hot with wet blood elements.
Ripening food makes it cold with wet water.
Roasting food makes it hot and dry with loss of blood.
Smoking food makes it cold and dry with loss of water.
By their effects upon the body, the four categories of ingestibles and
noningestibles can be characterized:
Sweet food and water make the body hot with wet blood.
Unsweet food and water make the body cold with wet water
Bloody dirt makes the body hot and dry with loss of blood.
Bloodless dirt makes the body cold and dry without water.
Therefore, culinary categories and categories of substances relative to the
body are homologous with one another in terms of the oppositions wet
versus dry and hot versus cold, and with the indigenous conceptualization
of the body itself as a bisected duality.

inside body : outside body :: inverted outside : everted inside (4)

roasting : boiling :: ripening : smoking (6)
blood : sweet :: unsweet : bloodless dirt (8)
hot dry : hot wet :: cold wet : cold dry (V)
By now it must be abundantly clear that Bush Mekeo views concerning
the human body are also homologous with those concerning the world or
cosmos. The duality of the body's inside and outside is bisected through
eversion and inversion, respectively, as is the bush/village duality. Thus,
the four categories of space and the things associated with them described
at length in the previous chapter can be added to the above series:
remote peripheral village adjacent ,,
bush ' village " abdomen ' bush
bush village village bush ,-,
resources ' resources " wastes ' wastes
The opposed states of human bodily health and illness can now also be
shown to correspond with the ordinary and extraordinary spheres of vil-
lage life. In the sphere of ordinary spatial and temporal relations, re-
sources and wastes lacking human blood are passed between the inside
bush and the outside village. As an aspect of those same daily transfers,
humans in a state of health likewise pass edibles and excreta that lack
human blood into and out of their bodies. On the other hand, the inside of
the body is the reservoir and source of the one substance - human blood
 Body and cosmos 59
- that, once it emerges during illness, is distinctly associated with holes of
the extraordinary spatiotemporal sphere. Human blood outside one body
and passed into another, or the mixing of bloodless human excreta with
foreign human blood in holes, are extraordinary events no less than is the
transferral of hot dirty human blood between holes. In any case, there are
material crossovers between the inside of bodies and the inside of holes.
Moreover, these extraordinary events or transfers are all similar in their
irregularity as regards time.
The conceived opposition of health and illness also corresponds to the
opposition between the ordinary and the extraordinary in terms of one
final distinction: the directionality of transformation or transfer. Sub-
stance transformation in the processes of healthy blood synthesis and
bloodless excretion are unidirectional, as are village/bush transfers of the
ordinary sphere of space and time:
Inverted Everted
Inside Outside outside inside
bush resources —* village resources —» village wastes —» bush wastes
blood <— sweet food unsweet food —» bloodless dirt

Conversely, transformations indicated by the indigenous theory of illness

are ambidirectional like extraordinary transfers between holes:
Inverted Everted
Inside Outside outside inside
bush village village bush
underground < • aboveground underground *—• aboveground

i t t

bloodless
inside body outside body < > abdomen excreta

t I
From what would appear to be every significant dimension of contrast
and opposition, indigenous Bush Mekeo ideas about the human body
constitute a cultural representation of the cosmological ordering of space
and time, and vice versa.
 4
Sex, procreation, and menstruation

Bush Mekeo views of human sexuality and reproduction are closely tied
to conceptualizations of the spatiotemporal order and the indigenous the-
ories concerning food, blood synthesis, excretion, health, and illness.
Conception, for example, involves transfers in and out of bodies and
transformations between sweet and unsweet as well as the distinction
between clean and dirty. Also, babies are conceived in the abdomens of
their mothers. However, the processes of sexuality and reproduction as
the Bush Mekeo know them seem initially to contradict the relations
between the categories I have revealed in the contexts previously de-
scribed. Oral ingestion of any hot bloody dirt is said to lead to illness and
even death. Semen is one such poison and will produce the same un-
desired result if it is eaten. Yet, semen ingested by a woman in sexual
intercourse may well have the exact opposite effect; that is, it can result in
the conceiving of a child or new life, rather than illness or death. This is
one further way dirt, at least in the form of male semen, may paradox-
ically sustain the life of a body, much as may food, even though the two
are categorically opposed in the context of eating. Furthermore, the
native understanding of menstruation is implicated in this particular con-
tradiction. According to villager claims, women will not menstruate spon-
taneously but only as a result of sexual intercourse.
This chapter attempts to clarify these issues and resolve the apparent
contradictions among them by articulating the explicit ideological tenets
of the indigenous theory of human procreation.
Sexuality in the ordinary and extraordinary spheres
"Semen" (ilaila) contains, or is constituted of, "male blood" (ifa man-
guenga). Like other bloody substances of the human body, semen is
regarded as wet, hot, and dirty. And consistent with categorical distinc-
tions of space and time, semen is never ordinarily deposited inside the
bush. This prohibition extends even to the performance of otherwise
legitimate sexual intercourse between spouses in the bush away from the
village.
Villagers claim that men and women are most cold or vulnerable to the
60
 Sex, procreation, and menstruation 61
hot of sorcery at the moment of coitus. Certain kinds of peace sorcery
involve capturing the victim's soul in a bottle; from the bottle, the blood
and spirit of a dead human are ritually induced back into the victim's
body.l During sexual intercourse, the bodies of the male and female are
open, allowing the transfer between sorcerers charm or bottle and their
own bodies. Supposedly, a sorcerer haunting the bush will not attack
when his intended victim is in the company of even one other person, for
then his efforts will be in vain. But the mating couple in the bush is easy
prey. Married men and women need never take this risk in their gardens
or elsewhere in the bush, for they may engage in sexual intercourse in
their houses of the village where they are insulated from the approach of
sorcerers. This does not mean all sexual unions outside marriage must
occur in the bush, however. Single girls and their lovers (whether men
married to other women, or bachelors) may also sexually open their
bodies at the village. A man must sneak into a girl's mosquito net at night
after others in her household have fallen asleep. Only the scandalous and
socially disruptive adulterers (i.e., men and their lovers who are married
to other men)2 lack an opportunity to unite sexually at the village. There-
fore, adulterers have no choice but to "do like pigs and dogs" in the bush.
But they do so at the risk of their lives. Except for adulterers, then,
human sexuality does not violate the ordinary spatial relations that restrict
bloody residues of the human body to the village.
However, it seems that semen is transferred between holes and bodies
of the extraordinary sphere. My data are incomplete on this score because
much of the relevant information is secret, but it was suggested to me that
semen is used in the ritual charms or bottles of some sorcerers and other
practitioners (see Williamson 1913:274). One possibility is that peace
sorcerers ritually acquire semen when they catch an adulterous couple
mating in the bush, as described above. In any case, semen captured by
sorcerers will be taken to their respective aboveground holes (fauapi bush
houses or ialiali clubhouses at the village) for manipulation and storage
before it can be ritually introduced into the bodies of other villagers as hot
bloody poison.3 Thus, adulterous sex and the theft of men's semen by
ritual specialists belong to the extraordinary sphere of object transfer.
Ordinary sexuality, particularly between married persons, however, is
symbolically tied with the ordinary daily activities surrounding food, eat-
ing, and excretion. A woman's vaginal secretion (ivi, or "water") is dirty,
but it is neither hot nor bloody like semen. It is cold and bloodless. Still it
is dirt, so villagers claim they practice no form of oral sex for fear of eating
either kind of dirty thing. Nonetheless, villagers copulate in the same
place they store, cook, and eat food - inside their domestic dwellings.
Moreover, it is immediately adjacent to the household hearth that women
give birth and endure the many months of the postpartum memengoa (see
62 Quadripartite structures
below and Chapter 5). It may seem curious that the Bush Mekeo, who are
so particular as regards what they eat and do not eat, should so casually
prepare and eat their food in the same place dirty sexual fluids of inter-
course and parturition are released from their bodies.
People prepare food and eat with their hands. During the periods of
married life when a husband and wife are sexually active with one an-
other,4 the wife is expected to cook food for her husband. So husband and
wife, especially the latter, must take precautions that their hands are not
contaminated with sexual dirt. Mates simply do not handle one another's
genitals in the act of sex. Thus, sexual and food exchanges between hus-
band and wife coincide spatially and temporally, but they are isolated
from each other with respect to their bodies.

Sexuality and bodily transformations

Although sexual fluids are dirty, sexuality in other respects involves cate-
gories of, and transformations (or the absence of transformations) be-
tween, sweet and unsweet. The act of adult coitus is sweet for both males
and females, whereas abstinence is unsweet. Indeed, it is human adult
heterosexuality only that is seen as sweet. Sex among children, between
adults and children, among adults of the same gender, and between
humans and animals is unsweet. However, among adult humans of op-
posite gender, intercourse is ideally sweet only with nonrelatives, that is,
with persons who are actual or potential spouses. Conversely, sex is
unsweet between men and women who are classified as kin or relatives
and prevented from marrying by the rules of exogamy.5
My extended examination of the categories sweet and unsweet in con-
texts other than sex (e.g., cooking, eating, blood synthesis, excretion,
work, nonwork) drew upon the conceptions of hot or transformability, and
cold or nontransformability. Hot and cold are also involved in the distinc-
tion of sexual sweet and sexual unsweet. Adult humans of opposite gender
are generally hot for transforming one another's unsweet condition (as a
result of prior abstinence) to sweet. Children, persons of the same gen-
der, and animals are sexually cold in this sense. Similarly, adult opposite-
gender nonrelatives are sexually hot to one another, and relatives are
sexually cold. Later chapters dealing with kinship, clanship, marriage,
and mortuary ritual will explore in depth the latter dimension of sexual
transformability versus nontransformability with respect to relatives and
nonrelatives. Here, I shall concentrate instead upon the indigenous con-
ceptualizations of bodily sweet/unsweet transformations predicated by
the act of heterosexual congress between adults who are assumed to be
nonrelatives and married, and upon the implications these notions have
for procreation.
Villagers describe the physical sensations of sexual intercourse and
 Sex, procreation, and menstruation 63
abstinence according to the same category distinction - sweet versus
unsweet - they use for food satiation and hunger, but with respect to a
different part of the body - the "skin" (fanga). In the condition of sexual
abstinence or quietude, the skins of men and women are unsweet. The act
of sexual intercourse transforms their skins to sweet. Then, after a while,
the sweet skins will revert to the unsweet condition of abstinence. With
cooking, it is the hot (tsiabu) of fire (ltd) that transforms unsweet inedible
food and water, or food alone, into sweet edible food. With sex, there is also
a "fire"; that is, the term for "vagina" in the language of the Bush Mekeo is
ito, also. In addition, as mentioned above, women's vaginal secretions are
water. When a woman is not engaged in sex, her vagina is cold. But with
the exertions of sexual penetration, her vagina becomes hot, transforming
her skin and the skin of her mate from unsweet to sweet. Apparently, the
bodily exertions of the sexual act (etsita) are not a "skill" (etsifa) of either the
work or nonwork types, but linguistically at any rate "sex" and "skill" do
appear to be related (see Chapter 3).
The hot of sex possesses additional transformative capacities on the
bodies of men and women that relate directly to procreation, and the
same category distinctions can be seen again. However, sex is here the
inverse transformation of eating. By eating hot cooked food, the body
synthesizes sweet blood and flesh that remain inside the body. Converse-
ly, the hot of sex transforms the sweet for the body's internal blood to
unsweet, so male and female excrete semen and womb-blood, respec-
tively, into the "womb" (imi afunga; literally, "child's place") of the lat-
ter's abdomen. In this sense, sex is homologous with illness. Sweet is
transformed to unsweet so blood is excreted from, rather than retained
by, the inside of the body of male and of female.
Nevertheless, in certain other respects, sex is significantly distinct from
illness, as it was defined in Chapter 3. It is in this context that the
apparent contradiction of women ingesting hot bloody and dirty semen is
partially resolved. First, the male's semen as well as the female's womb-
blood go from the inside of their respective bodies to the women's ab-
domen, which is outside (inverted outside) her body. Technically, then,
semen and womb-blood are not excreted to the everted inside of either
their bodies, as blood is during chronic illness. Second, the male's ejacu-
lated semen is not strictly inside the female's body like poisons that have
been orally ingested and cause illness; his semen, once more, is outside
(inverted outside) her body. Lastly, the loss of male and female bloods
from the inside of bodies in coitus is not in the initial case the result of
either ingesting hot bloody dirt or of ritually manipulating it external to
the body with personal bloodless leavings. The secretion of womb-blood
to the woman's abdomen is not a direct consequence of insemination per
se. Rather, both transfers are the joint consequence of body transforma-
tions activated by the hot of sexual exertions upon the woman's vagina.
64 Quadripartite structures
My comments below concerning menstruation will add further clarity
to this issue of the conceived relations between sex and illness.

Procreation versus menstruation

Although sex with respect to the bodies of the mating couple reverses the
process of synthesizing blood from food yet is distinguished from illness,
sex is nonetheless homologous with blood synthesis or body sustenance
and the health of newly conceived infants. According to the indigenous
theory of procreation, a human infant is "begun" or "conceived" (engatna)
by the mixing of equal quantities of father's and mother's sexual bloods.
As is fairly common elsewhere in Melanesia, the father's semen is under-
stood to dry, solidify, or coagulate the mother's wet amorphous womb-
blood and to give form to the fetus (Chapter 5; Mosko 1983). It will be
remembered that the hot sweet blood that sustains the human body after
birth is synthesized from the proportionate mixing of two distinct catego-
ries of food - meat and plant food - containing elements of blood that
have been transformed by fire. The transformation of plant food and meat
when they are mixed together in water by boiling is a woman's skill. The
transformation of food separately without water by roasting over fire is a
man's skill, but, alone, roasting is cold for synthesizing the body's blood
or transforming it from unsweet to sweet or from cold to hot. In procrea-
tion, equal proportions of drying male semen and amorphous female
womb-blood are mixed together with water (vaginal secretion) and trans-
formed together by the hot of sexual exertions of the woman's fire-vagina
to create the blood and flesh of a human body - the newly conceived
infant. Neither semen alone nor womb-blood alone is hot for procreation.
Therefore, with respect to the body of a human fetus, sex is homolo-
gous with blood synthesis and life or health consequent to eating food.
These ideas concerning procreation are closely connected to the indige-
nous theory of "menstruation" (ngaua eisa; literally "sees the moon").
Informants scoffed at my arguments that a single act of coitus, or even
several acts performed sporadically, is sufficient to impregnate a human
female. Instead, they claimed, a great many acts of sexual intercourse are
necessary before the mixed sexual bloods are hot or "enough " (ekaina) for
conception. Sex performed only intermittently is cold or "not enough"
(aisekaina) for conception, and thus the sexual bloods will come out of the
woman's abdomen as she menstruates. A couple must then reinitiate the
process of accumulating enough hot semen and womb-blood in the wom-
an's abdomen to make a child. When a couple engages in sex with enough
intensity, the woman will not menstruate. The semen and womb-blood
will remain in her abdomen, and, after three "moons" (months) of intense
sexual activity have passed without menstruation, she will be acknowl-
edged as "pregnant" (inaenga or inainga, or "in her abdomen").
 Sex, procreation, and menstruation 65
As bloody hot semen does not emerge from inside a man's body spon-
taneously, neither do babies nor menstrual blood from a woman's ab-
domen. On the one hand, some six months 6 after a woman is declared
pregnant, she can expect to give birth. Sexual bloods accumulated in the
woman's abdomen are hot for making a baby, and thus they are retained
by her abdomen. While the fetus is being conceived and afterward during
the woman's pregnancy, however, cold water (amniotic fluid) also collects
in the amniotic sac. When the sac breaks, the cold water cools the hot of
the baby and associated tissues so they leave the mother's abdomen. In
addition, certain delivery and postpartum skills, termed memengoa, are
applied to the same effect.
Menstruation, on the other hand, which occurs when conception, preg-
nancy, and birth do not, seems to be conceptualized in similar terms.
Although indigenous views preclude the notion of "Virgin Birth" (Leach
1966b), they also rule out the possibility of virgin menstruation. It is
argued that women do not menstruate except as a consequence of sexual
intercourse, which is cold, not enough, or insufficient to achieve concep-
tion and pregnancy.
My informants stated that prepubescent girls do not menstruate be-
cause they have not yet initiated sexual relations. One man explained it to
me this way:
Before a girl's breasts enlarge, she does not menstruate. She is not sleeping with
any lovers. When her breasts do enlarge, bachelors will begin courting and using
their love charms on her. [With courting ritual] one bachelor will change her
mind away from the words of her fathers and brothers, so she will secretly admit
him to her mosquito net at night after the others are sleeping. Afterward the girl
will menstruate. So when the people see a girl's breasts grow, they know the
bachelors will court, and she will sometimes let one of them sleep with her. If she
does not copulate too much, she will not become pregnant. She will only menstru-
ate. Everyone knows single girls are having sex.
It is also claimed that menstruation is "like illness but not 'true' [or real]
illness," although menstruating women excrete blood. One way true ill-
ness originates is from the ingestion of foreign hot dirty blood that trans-
forms both the victim's blood from sweet into unsweet, and cold bloodless
residua of digestion from unsweet into sweet. Hot blood and dirty semen
will cause true illness in precisely these terms if they are eaten. The
transformation that generates menstrual blood, however, seems to be
somewhat more complicated. As a woman participates in sexual inter-
course, semen deposited in her abdomen is technically outside her body.
The hot of intercourse itself - not the hot of semen - causes her to
transfer her own womb-blood to her abdomen. Because semen and
womb-blood are both hot, they remain together in the abdomen until
they are cooled. Unfortunately, I did not ask my informants for detailed
66 Quadripartite structures
clarification on this point. I strongly suspect, however, that collected
semen and womb-blood are cooled like the hot and fully developed fetus
by cold water, but of another sort. Women generally are described by
villagers (male and female alike) as cold and wet, whereas men are com-
paratively dry and hot (see Chapter 5). After participating in sex, assum-
ing there is no conception, a woman will menstruate. As menstruation
approaches, her body becomes increasingly wet with fluid; she bloats.
This retained fluid ("cold water?") is lost approximately at the time she
menstruates. Therefore, it seems as though cold water retained inside the
woman's body may be eventually released to the abdomen, where it cools
the accumulated hot sexual bloods. Thus, they become unsweet to her
body and come out of her abdomen, just as amniotic water cools the hot
fetus and placenta immediately before birth.
If this bit of speculation does indeed correspond to the native view,
then it can be understood why menstruation (and child-birth, for that
matter) is not true illness: first, because semen and womb-blood con-
tained by a woman's abdomen are outside, not inside, her body; second,
because neither condition requires techniques of curing in order to evac-
uate the foreign blood or restore health; and third, because illness-caus-
ing blood that emerges from the body with the cooling water of urine is
considered a sign of curing rather than continued illness.
I should mention that the one question for which I have been unable to
find a satisfactory cultural answer, based upon my data, is how villagers
account for the temporal connection between the menstrual and lunar
cycles.
Otherwise, a number of additional observations are consistent with the
above conceptual relations between menstruation and sexual intercourse.
Aside from the pleasurable or sweet aspect of intercourse, married cou-
ples typically want many children7 and actively strive in this direction at
the appropriate times. The quantity of semen that must be deposited in
the wife's abdomen over three months to conceive and so avoid menstrua-
tion is considerable. Yet, husbands and wives are not sexually active with
one another for most of their married lives. A woman must not have
relations with her husband while she is pregnant, nursing, or menstruat-
ing. At these times, the husband may continue his sexual activity, but it
must be with other women (particularly another wife, or single girls or
widows who are not in any of the same conditions). Most men, however,
have only one wife, and move to the chief's clubhouse at night to sleep
upon their wives' pregnancies.8 Older, more conservative men of the
village still maintain that a man and a woman who sleep together, yet
claim they have no sex, are "lying" (bifonga). Supposedly, no one can
resist this temptation (or transformation). Fathers of young husbands thus
chase their sons away and force them to sleep at the clan clubhouse as
 Sex, procreation, and menstruation 67
soon as their daughters-in-law become pregnant. Once enough semen has
been deposited for conception to be assured, no more must enter a wom-
an's abdomen. And, of course, she will not menstruate.
A woman's sexual abstinence continues long after parturition, and,
again, according to Bush Mekeo culture, menstruation does not occur.
Birth and postpartum techniques (memengoa) mentioned above help
force the baby, placenta, and birth-blood out of the mothers abdomen.
This process of evacuation, however, continues until the child is weaned.
It is imperative that parents in the meantime refrain from sexual congress
for several reasons. First, the semen of the husband can spoil the moth-
er's milk and make the nursing infant ill. Second, birth-blood clinging to
the skins of mother and infant can make the husband seriously ill if he
accidentally ingests any of it following physical contact with them. Third,
the husband would traditionally devote the postpartum period to ritual
pursuits that typically involve the handling and manipulation of such
other hot dirty and bloody substances as ingredients of charms. Traces of
these potential poisons remain on his skin. In the act of sex, they could
inadvertently enter his or his wife's open body and make either or both of
them seriously ill. Or, even more dangerous to the health of wife and
child, birth-blood on their skins could later be mixed with the other
substances of the man's ritual charms. Should he then use these charms,
his wife and child would also be attacked according to the principles of
mefu ritual, as previously discussed.
Still other observations are relevant to the prohibition of sexual inter-
course during pregnancy and following birth. Over my stay in the village,
I gathered the distinct impression that many women try to extend the
nursing of their infants for as long as possible, or rather for as long as their
husbands will permit them. Some women admit that they are reluctant to
wean their babies because the experience is so painful to them (i.e., the
mothers). Bush Mekeo mothers wean their babies "cold turkey," as we
would say, by rubbing chili juice on the nipples. Chili is hot and quickly
transforms the child's sweet regard for mother's milk9 to unsweet. Mar-
ried men, by contrast, seem typically anxious to resume sexual relations
with their wives. At least from what I was able to observe, it is at the
husband's insistence that nursing is terminated and the reproductive cy-
cle reinitiated. The couple now risks, of course, the chance that the
woman will menstruate as a consequence of sexual intercourse. But if
there is enough semen deposited in her abdomen, she will conceive
instead.
These aspects of the sexual prohibition lasting from conception through
weaning will be described more systematically in the next chapter. For
now, it is merely necessary to note that, while husband and wife do not
have sexual relations, according to the culture there will be no menstrua-
68 Quadripartite structures
tion and there will be no procreation. However, once sexual relations are
resumed at other times, there are two possible results: (1) the conception
of a human fetus, but no menstruation, or (2) menstruation, but no
conception.
Women must not menstruate
It is implicit in these indigenous beliefs about sexuality and reproduction
that village women can pass their entire lives without ever excreting
menstrual blood, at least theoretically. This should not be altogether
surprising, because menstruation among women of preindustrial societies
is evidently quite rare (Harrell 1981). Indeed, it is of critical significance
to all Bush Mekeo villagers that women do not. Women traditionally lack
sop (soap) for washing hot dirty menstrual blood from their hands. This
poses a potential threat to the women and to all others, particularly the
men. As it was told to me, "We [the men] never know for sure what they
[the women] have been touching." Men even insist that women serving
food carry the bowls above waist height to ensure that none of the rubbish
from their grass skirts, perhaps containing traces of menstrual blood, falls
into their food. Disgruntled or adulterous wives are notorious for poison-
ing their husbands by putting menstrual blood into their food. For this
reason, men who are suspicious of their wives will not eat the food they
cook.
Women themselves are caught in an additional and somewhat different
dilemma. They must not handle or cook food while they menstruate, but
they must at the same time keep the fact of their menstruations secret. A
single girl conceals her sexual liaisons from her elder brothers and parents
to avoid a beating upon discovery. Her declining to cook food for men,
which she is expected to do each day, is tantamount to admitting men-
struation and, therefore, recent sexual encounters. Married women along
with single girls must also be very circumspect. From the time they were
very young, they were warned by their mothers that men of the village
are constantly on the alert to acquire menstrual blood for their courting
charms. If a man includes a particular woman's menstrual blood in his
charm and attempts to use it in courting, she will become gravely ill and
die. Married women, like single girls, essentially announce to one and all
that they are menstruating if they cease cooking and carrying food to the
men for no other apparent reason (i.e., illness, pregnancy, postpartum
seclusion).
To make matters even more awkward for the women, there is no cul-
turally prescribed place where they may safely dispose of menstrual
blood. It might be predicted that menstrual blood should be buried un-
derground at the village as other hot bloody and dirty human tissues are.
But because even domestic dwellings are raised off the ground on posts,
 Sex, procreation, and menstruation 69
the woman who tries this may be easily discovered. Whenever I inquired
of female informants what they did with their menstrual blood, I received
evasive replies. Many village men say they do not know what women do
with menstrual blood, except that they hide it. A few men ventured that
women collect the blood in dry leaves and secretly bury it nearby in the
bush.10 If so, the disposal of menstrual blood contradicts the category
distinctions by which the ordinary and extraordinary spheres of vil-
lage/bush transfer are ordered. Not only would bloody human excreta be
thereby transported to the bush, they would be buried in an underground
hole of the adjacent bush.
Clearly, from the perspectives of men and women alike, the best cul-
tural solution to the problem of what to do with menstrual blood is simply
to ensure that women do not menstruate in the first place. Indigenous
conceptualizations provide for this ideal. Unmarried sisters and daughters
need only be prevented from sexually entertaining lovers at night. The
situation with wives is more complicated, but to the same effect. They
must abstain from sex at the inception of pregnancy until the child is
born, nursed, and weaned. Then they must go the other way, indulging
in sexual relations with their husbands with enough intensity to ensure
conception and pregnancy once more. And this is precisely what we find,
as discussed at length in Chapter 5.

The story of Amaka

Thus far, by examining villagers' explicit views and responses, I have
tried to show that the apparent contradiction, whereby women's health is
not threatened by sexual transfers of semen, is resolved in the culture of
the Bush Mekeo by two alternate and opposed outcomes - reproduction
or new human life, on the one hand, and menstruation, with its poten-
tiality for generating illness and death, on the other - such that ideally the
former and never the latter occurs. These relations are expressed rather
explicitly in the mythical tale of Upstream Amaka (Amaka Kaenga) and
Downstream Amaka (Amaka Tsibo). My analysis of this myth will con-
clude this chapter.
One day long ago, upstream Amaka [AU] was fishing in his canoe for meat for
his wife and child when, in the afternoon, he became tired. He tied his canoe to
the shore and went to sleep. During the night, the river flooded and swept the
canoe downstream. By morning, AU had drifted to where the river nears the coast
and slows. Here, his canoe gently bumped along the shore. AU awoke and saw a
garden planted there. He walked into the garden and he saw downstream Amaka
[AD]. The two men asked each other their names, as they had never met before,
and discovered that they were namesakes (wau). So the two men started calling
each other, "Ah, my namesake" (a usage of endearment).
AU explained that he lived "up there" and had drifted down in his canoe. AD
70 Quadripartite structures
called his wife and told her to prepare food for his namesake. She skinned and
sliced a taro and laid it on the ground to be warmed by the sun. AD's wife did not
cook the taro; she served it raw. AD told AU to go ahead and eat the food his wife
had "cooked" [sic] for him. AU said to wait a moment, that he had to defecate first
[he was lying]. He went to the bush and made a fire. Quickly the bush caught fire.
When AD and his wife saw the fire, they were very afraid. AU returned and told
them not to be afraid, for it was only fire (ito). They must put the taro into the fire
to cook it. AU gave AD and his wife some of the cooked taro, and they ate it. Not
having eaten cooked food before, the two of them vomited. Thereafter, when they
ate cooked food, however, they did not vomit it up.
Then AU told his namesake he wanted a chew. AD gave him areca nut and
betel pepper, and then he inserted his limestick into his wife's vagina (ito) and
passed it to AU. AU again excused himself to go to the bush, and secretly he used
some lime powder (apu) hidden in his hair instead to chew with the areca and
betel. When he came back to the garden, his mouth was bright red. AD saw the
red spittle drip out of his mouth, and he screamed in fear, "Namesake, blood is
coming out of your mouth." AU told him not to be afraid, it was not blood. Then
AU explained how to use lime rather than his wife's vagina to chew.
While the two men sat chewing, AD's wife, who was pregnant, went into labor.
Her husband took a bamboo knife to cut open her abdomen and take out the baby.
This was how babies were born according to the downstream custom. As he
approached his wife, AD started to cry. AU asked him why he was crying, and AD
explained what he was about to do. AU told him to stop and see what he could do.
He brought a short stick and told the woman to sit on it while he squatted behind
her, squeezing her abdomen to force the baby out of her vagina. Then AU boiled
some water. He told AD he must not do any more than this, and he told AD's wife
she must do all the memengoa things - wash her body and the baby's skin with the
hot water, rub her abdomen with a heated bamboo tube to force out the remain-
ing afterbirth, cut the umbilical cord with bamboo knives, etc.11 "Do all these
things," AU told AD, "and you can keep your wife instead of killing her when she
gives birth.
And so this is how the people learned about fire, chewing, and birth.
The two male characters in this story are in certain respects equated as
one person. In others, however, they are systematically differentiated
and contrasted as two. They are both named Amaka, and, upon meeting,
reciprocally assume the egalitarian namesake relationship. But the two
previously lived unaware of, and apart from, one another. Both AU and
AD possess skills that employ hot or heat in contexts where men interact
with their wives: transforming food, chewing, and delivering babies. Yet
the respective skills of the two men differ. AD eats food warmed by his
wife in the heat of the sun, chews areca and betel with secretions from his
wife's vagina, and kills his wife in order to deliver her baby. AU, on the
other hand, cooks food with the hot of fire, chews with lime, and assists in
parturition without killing the mother. 12
Part of the meaning of the myth comes from the temporal contrast of its
segments. The sequence of cooking and eating, chewing, and giving birth
 Sex, procreation, and menstruation 71
reveals a progression among categories of things from unsweet to sweet,
and from cold bloodless dirt to hot bloody dirt. AD's transformative skills
(or lack of them) not only contrast with AU's, they are inconsistent among
themselves. AD fears fire (ito), but he is not afraid of putting dirty secre-
tions from his wife's vagina (ito) into his mouth. He fears the bloodless
spittle coming out of AU's mouth, but he does not fear the blood put on his
hands from killing his wife when she gives birth. Inevitably, this blood on
his hands would be put into his own mouth. As a result of his peculiar
skills that use cold or inappropriate transforming agents rather than ap-
propriate hot ones, AD risks putting inside his own body the things that
ought to be kept outside (i.e., the three categories of inedible substances
- unsweet, cold bloodless dirt, and hot bloody dirt) and does not put
inside his body what ought to be there (i.e., the one category of edible
substances - sweet). In AU's terms, that is, in terms of the proper or-
dered categories, AD is thoroughly confused.
AU in opposition to AD employs the appropriate and ordered sources
of transformative hot in each of the contexts where men interact with
their wives, and so he attempts in instructing AD. With fire, AU trans-
forms unsweet food to sweet for eating. Next, AU uses lime to mix with
areca and betel to transform the chew from unsweet to sweet. Alone,
areca and betel are unsweet for eating or chewing, but mixed with lime,
which is hot, betel and areca are made sweet for chewing (Chapter 5).
Lastly, by using the skills oimemengoa, AU enables AD's wife to remove
the hot bloody and dirty infant, afterbirth, and birth-blood from her
abdomen and yet live. Moreover, AU's memengoa skills allow AD's wife
to retain her own blood inside her body, and, because she alone touches
the bloody things of birth, AD neither gets them on his hands nor into his
own body. In short, AU's set of skills allows ingestion of sweet ingestibles
and systematically prevents ingestion of the three categories of noninges-
tibles.
AU's and AD's respective skills further contrast in their final results.
AD's confusions, beginning with cooking (sic), eating, and chewing, lead
to death simultaneous with birth. AU's ordered and enlightened con-
sistency in cooking, eating, and chewing culminate in life without death at
the moment of birth. The myth is thus a "charter" in the Malinowskian
(1948) sense for proper orderings of things. Specifically, it demonstrates a
systematic consistency between the contexts of cooking, eating, chewing,
and parturition that otherwise might not be apparent. This consistency or
homology is nothing other than the pervasiveness of the category distinc-
tions sweet/unsweet/cold dirt/hot dirt, and of the transformations be-
tween them through application of the proper sources of hot with respect
to the outside, inside, inverted outside, and everted inside of the body
that I have revealed earlier in this and preceding chapters.
Nonetheless, although the myth establishes that cooking food with fire,
72 Quadripartite structures

chewing with lime, and giving birth with memengoa techniques are ho-
mologous and proper transformations for humans, it also sets up an op-
position between the third of these skills and the first two in terms of the
reversibility of the bodily transfers involved. Incidentally, it brings us
back to the specific problems posed at the beginning of this chapter
concerning the native theories of human sexuality and menstruation, and
it provides further elements to their resolution. When AD and his wife
eat food cooked by fire for the first time, it comes back out as "vomit"
(unga). On first observing AU chew with lime, AD sees blood [sic] coming
out of his mouth. Both eating cooked food and chewing with lime for the
first time here suggest one feature of the indigenous theory of menstrua-
tion; namely, semen from a single act of intercourse is not enough for
conception. Similarly, food that is vomited contributes nothing toward
blood synthesis and bodily health, just as chewing areca, betel, and lime
is understood to contribute nothing to bodily sustenance.13 Vomiting,
expectorating after chewing, and menstruation all involve ambidirectional
transfers. Food goes into the mouth and then comes out of the mouth as
vomit. Areca, betel, and lime go into the mouth and come out as red
spittle. Semen goes into the abdomen and comes out of the abdomen as
menstrual blood (or part of menstrual blood). The point is that childbirth
also constitutes an ambidirectional reversal of this kind: Semen goes to
the woman's abdomen and comes out of it as a baby. But here there has
been a contribution to blood synthesis, bodily sustenance, and health.
Indeed, although semen makes no bodily endowment to the mother, it
does to the body of her and her mates baby.
Thus, menstruation as a result of insemination is like vomiting and
chewing. Although it contains hot dirty human blood and represents
ambidirectional bodily transfer, it is not, as they are not, a sign of illness.
 5
Male and female

The discursive examination of the indigenous theory of human sexuality

and conception contained in the previous chapter revealed a number of
significant distinctions between men and women that are relevant to the
systematic conceptualization of gender differences and roles in Bush
Mekeo culture. In other words, the relative contributions men and wom-
en are considered to make to reproduction correspond with certain other
activities and transformative capacities that are likewise differentiated
according to gender. The purpose of this chapter will be to describe more
fully this indigenous division of labor in terms of the categorical opposi-
tions and analogies that characterize male and female adulthood. My
argument proceeds as follows. First, I outline the cyclic ritual alternation
of adult females between procreative and contraceptive states according
to the opening and closing, respectively, of their bodies. I then describe
how adult males, like females, alternate between procreatively opening
and contraceptively closing their bodies, except that the female pattern
reverses the male pattern in certain critical respects. Next, I show that
the bisection of the male/female duality is homologous with the distinc-
tions previously elicited from other contexts of the indigenous culture:
village versus bush, body-inside versus body-outside, hot versus cold,
sweet versus unsweet, and so on. Finally, I demonstrate that the oscilla-
tions of men and women between ritual conditions in village life are given
further meaningful expression by the counterposition of two myths taken
from the Bush Mekeo corpus.1
The ritual cycle of the female
Among the Bush Mekeo, human reproduction is ideally initiated and
thereafter sustained within the bounds of "marriage" (amange). Sexual
encounters between bachelors and unwed maidens are strongly discour-
aged by parents and elder siblings, but only in part for fear of conception
and pregnancy out of wedlock. However, once males and females are
married by either formal arrangement or elopement, sexual intercourse is
vigorously encouraged. Indeed, the newlywed couple is expected to con-
ceive their firstborn as soon as possible in a fashion consistent with the
73
74 Quadripartite structures
prevailing notions regarding procreation. This initial phase of marriage is
termed "new marriage" (amange manga).
Once a bride leaves her family to live with her groom and others of his
household,2 her activities focus upon her sexual and reproductive capaci-
ty. In short, she is made hot for conceiving a child. On the day of her
arrival and each day thereafter, the bride's new female in-laws, under the
authority of the groom's own mother, attend to her toilet in every detail to
make her sexually sweet to the groom. Her body hair (buibui)3 including
her pubes is removed, and the hair on her head (fufu) is combed out and
rubbed with coconut oil. She is bathed in hot water. Reddened coconut
oil is applied to her skin. She is dressed in a leaf skirt colorfully decorated
with the insignia of her husband's clan (cf. Seligmann 1910:320-30). She
wears tortoiseshell earrings and a shell nose plug and fresh aromatic
flowers are slipped into her woven arm- and legbands.4
For as long as she remains in a newlywed condition, the bride does no
work. She may not leave the confines of her husband's house except for
the purpose of bodily elimination. Even then, she must be accompanied
by a female member of the household. During most of the daytime hours,
the bride simply sits alone and motionless inside a mosquito net hung
indoors awaiting the frequent visits of the groom. But late every after-
noon, the bride, attended by her mother-in-law, emerges to sit on the
veranda facing the central village abdomen in full view of villagers head-
ing home from the bush. For several hours, she sits draped with the shell
and dogs' teeth ornaments her own parents have reciprocated to the
groom's relatives as part of the protracted marriage compensation (kaua)
exchanges (see Chapter 6). As villagers express it, everyone, particularly
the bride's own relatives, can see that she is being well cared for by her
new in-laws, getting enough food to eat, and becoming fat.
These three objectives relate directly to the brides newlywed diet,
termed "engorging" (engoupa), and to the general expectation that she
conceive as soon as possible. The primary responsibility for the bride's
engorging falls, once more, on her mother-in-law. However, other mar-
ried women living at the same village are also expected to assist. These
other women include the kinswomen and clanswomen of the groom and
the wives of his kinsmen and clansmen.5 After returning from their gar-
dens in the afternoon, each of these women is expected to bring a bowl of
boiled plant food for the bride to eat. The mother-in-law may cook food,
too, but her main role is to see that her daughter-in-law eats the food all of
them have prepared and brought. The bride is not allowed to eat raw or
ripened plant food.
Ordinary Bush Mekeo etiquette dictates that one need take only a
single bite when food is offered so as not to offend the host and hostess.
Not so for the bride! She must eat all the food, bowl after bowl, her
 Male and female 75
mother-in-law places before her. When she feels satiated (i.e., when her
abdomen is sweet, ina mitsia), her mother-in-law will massage her ab-
domen. This manual manipulation supposedly stretches the brides stom-
ach and transforms her abdomen from sweet to unsweet so she will hun-
ger for, and be able to eat, more food.
The first time I actually observed bridal engorging, I was honestly
amazed at the quantity of plant food the bride was required to consume.
Despite the woman's courageous attempts to appear pleasant in the pres-
ence of her assembled in-laws, she was in obvious discomfort. I could only
laugh when, after five or six bowls, she drew from her string bag a bottle
of seltzer tablets purchased from a trade store. Small wonder that once-
comparatively-slim maidens in a matter of only two or three weeks follow-
ing marriage dramatically show the anticipated results of engorging! They
become visibly "fat" (efaka), with "plenty of skin" (fanga ealongai) and
"plenty of blood" (ifa ealongai).
The body of the bride is affected by engorging in other ways as well.
Most of the food she eats is plant food, and all of it is boiled in water. In
Chapter 3, I described how cooked plant food is considered hot for trans-
forming the abdomen from unsweet to sweet, and similarly how hot water
transforms the generalized inside of the body from unsweet to sweet. As a
result of engorging, then, the bride's abdomen and the inside of her body
become exceedingly sweet. And because she drinks the broth her food is
cooked in, her body is rendered considerably wet.
Engorging is understood in these terms to promote the conception of a
fetus in the bride's abdomen in conjunction, of course, with the mixing of
released womb-blood and semen. For her part, the engorged bride is hot
for procreation. Because many acts of intercourse and the consequent
deposition of much womb-blood and semen are required for conception,
the bride is expected to perform sex with the groom several times each
day - indeed, as often as possible for the minimum period of three
months necessary to ensure pregnancy and avoid menstruation. Blood
from the groom's semen and blood from the bride's engorging gather in
the same part of the bride's body (her ina or abdomen), which is lin-
guistically connected with the category "mother" (ina). The more fre-
quently the newlyweds perform sex, the more hot the engorged bride
becomes for procreation. Thus, each night and for most of each day the
bride remains at the sexual disposal of the groom. As we shall see below,
the groom does not wander too far nor too long from his bride.
The frequent bathing of the bride in hot water is understood to increase
the frequency of newlywed mating. The exertions of sexual intercourse
transform the bride's skin from unsweet to sweet. Without interference,
it would take some considerable time for the bride's skin to revert to an
unsweet condition, that is, when sex would be again sweet and desirous to
76 Quadripartite structures
her. However, the hot water of her frequent baths reverses the sexual
transformation of her skin back to unsweet, so she will be inclined to
repeat the act of coitus with markedly increased frequency.
The bride's procreative assimilation of large quantites of cooked plant
food, water, and semen contrasts with the radical restriction in the quan-
tity and variety of meat she may eat. This meat restriction also has definite
cultural implications for the bride's fertility. While they are single, girls
may consume moderate amounts of meat, but once they marry and be-
comes brides, it is said, "they eat no [sic] meat." Actually, this statement
should be qualified to mean that brides should eat merely enough meat of
particular categories to sustain their bodies minimally and avoid meat
starvation. But a bride may eat no meat whatsoever of animals that are
regarded either as faifai bush people or associated with them for fear of
giving birth to a faifai baby. Staple "big" meats - pig, cassowary, barra-
mundi, bush fowl, etc. - are strictly prohibited as well. Thus, only mini-
mal quantities of "small" meats - smaller fish and birds, mostly - with
one exception6 are deemed appropriate to the brides diet. The brides
minimal consumption of small meat here hardly balances with her exagge-
rated engorging of cooked plant food. While she was still single and not
yet a bride, her diet was balanced with meat proportionate to plant food.
Therefore, upon marriage a girl's balanced diet is replaced with the im-
balanced diet of little meat and much plant food. Moreover, the food she
eats as a bride is provided to her by people of entirely different rela-
tionships as regards her sexuality and fertility.
This is particularly evident as regards meat. Relatives of a single girl are
especially concerned that she not conceive or give birth out of wedlock.
Also, it is her male relatives who provide her with most of the meat she
eats. They will give her moderate amounts of both big and small varieties,
most often the former. Occasionally, her male relatives may even give her
some meat from an animal of a species associated with faifai, but which
was not taken from a bush hole and does not have an excess of blood.7
Upon her marriage, her own relatives cease supplying her with food of all
kinds. The minimal amount of small meat she as a bride may eat is
procured by her groom, and he alone. Every few days, the groom sets off
to hunt or fish specifically for her. When he returns with a catch, his
mother will cook and serve bits of it to his bride.
By comparing these indigenous rules concerning meat and fertility, it
seems a single girls male relatives hunt and fish so that blood-potent
elements of meat and complementary blood-potent elements of plant food
mix in balanced proportion to transform and sustain her throat, abdomen,
and, therefore, her blood and flesh while guarding against the mixing of
semen and womb-blood in her abdomen. By contrast, the mixing of the
grooms semen with the bride's womb-blood initially derived from her
 Male and female 77
engorging of cooked plant food is encouraged in the brides abdomen,
such that the mixture of blood-potent elements from meat and plant food
associated with both her throat and her abdomen is sharply discouraged.
The intended results in each case are correspondingly opposed as well.
On the one hand, a single girl's capacities for her own blood formation,
bodily sustenance, and, thereby, her work are maintained while her fer-
tility is minimized. On the other, the brides own blood formation, bodily
sustenance, and work capacity are minimized while her capacity for pro-
creation is maximized.
Of particular interest here is the relation of meat and semen. By pursu-
ing it in some detail, I hope to establish a homology between food catego-
ries and categories of sexually differentiated human bloods.
Semen and bush meat generally are male-derived. Male relatives of a
single girl give her moderate amounts of meat but no semen, and she is
supposedly infertile. Thus, there are no excess blood-elements from plant
food to form womb-blood and mix with semen contributed by nonrelated
male lovers. This dietary balance of moderate meat and plant food that
promotes and sustains work would seem to be procreatively cold, or
contraceptive. A bride, however, ingests maximal amounts of semen and
plant food and minimal meat to mix for procreation. The bride's meat food
balances with only a part of her generous plant-food intake, enough to
minimally sustain her throat and body but not to allow her to work. The
excess or surplus blood-elements derived from plant food associated with
her abdomen can mix instead in the form of womb-blood with the abun-
dant semen of the groom. The two procreative bloods form and consoli-
date the body of an entirely new person in her abdomen. In this event,
sexually hot semen replaces or substitutes for edible but procreatively
cold or contraceptive meat.
These observations finally allow the establishment of a number of con-
nections between categorizations of food and distinctions relevant to sex.
First, the opposition of meat and plant food corresponds to the distinction
between male and female, respectively. A simple syllogism can illustrate
this sufficiently. The data concerning newlywed ritual presented so far
indicate:
(a) meat : plant food :: semen : womb-blood
And, of course:
(b) semen : womb-blood :: male : female
Therefore:
(c) meat : plant food :: male : female
Although these three statements might initially resemble the series of
homologies displayed in Chapters 2 and 3, they must not be construed as
78 Quadripartite structures
belonging to the same set, at least as they stand. They are written out
here merely to enable the deduction of statement (c).
Nonetheless, by a further elaboration of the categories in statement (c)
above, certain distinctions regarding food and sex are homologous with
one another and with the series of bisected dualities analyzed in previous
chapters. Secondly, then, the binary oppositions meat/plant food and
male/female are each further bisected into four categories. On the one
hand, there are contraceptive and procreative meats, and contraceptive
and procreative plant foods; on the other, there are contraceptive and
procreative male bloods, and contraceptive and procreative female
bloods:
male male female female . .,
contraceptive ' procreative " procreative ' contraceptive
Food categories
!. „ cookable noncookable /ir . x
big meat : small meat :: , , : , L c , (12)
plantL rfood plant food
Sexual blood
male : u ,, , female /1ON
i j ii j semen :: womb-blood : , , , , , (13)
body blood body blood
I shall be returning a number of times to these categories and relations as
I trace out the remaining phases of the female ritual cycle and contrast it
with the complementary male ritual cycle. Before proceeding, however,
it should be noted that contraceptive foods and bloods correspond with
inside and procreative foods and bloods with outside, in terms of the
body-and-space homology:

inside : outside :: inverted outside : everted inside (4)

Newlywed activity continues until the couple conceives a child and the
bride becomes "pregnant" (inainga). As previously mentioned in this
section, the onset of pregnancy is indicated by the passage of three con-
secutive months with sex but without menstruation. The pregnant wife is
no longer confined to her husbands house. She may bathe in cold water,
her body is neither anointed nor ornamented as before, and she is not
publicly displayed each afternoon. Very importantly, engorging also
ceases abruptly. During her pregnancy, a wife eats less plant food (all of it
is still cooked) and more meat, including big or contraceptive varieties
that are hot for the synthesis and sustenance of her own blood and flesh.
This more moderate and balanced pregnancy diet enables her to work.
She can now cook and carry food and participate with other women in the
daily clearing of the village abdomen. It is at this time also that she and
her husband begin clearing their first new garden together or perhaps
gather bush materials for a new house of their own. As pregnancy ap-
 Male and female 79
proaches its end, it is critical that the wife devotes considerable work
toward gathering and stockpiling a large supply of firewood underneath
her house. This fuel will be burned at night during the postpartum vigil
kept by her husband.
The relations of a newlywed woman to her husband and to others also
change significantly during the pregnancy phase. Sexual intercourse be-
tween husband and wife terminates as he begins sleeping on the chiefs
clubhouse in the company of other men, and she in the family dwelling
with other women and children of the husband's household. By not eating
faifai meat and taking no additional semen into her abdomen, the preg-
nant woman prevents conception of a faifai baby as well as menstruation.
The husband no longer exclusively supplies his wife with meat. She par-
takes of meat other men bring to the village and share among resident
households, and her husband likewise distributes his quarry. And be-
cause she is now capable of, or hot for, work, the pregnant wife herself
participates with other women in the daily reciprocation of cooked food
between related households.
In summary, the bride's procreative configuration of indulgent sex and
maximal plant food and minimal meat intake, focusing upon exchanges
between the bodies of bride and groom, is overthrown during pregnancy
to resemble in some respects the contraceptive working regimen of the
single girl. Only now, with a growing baby in her abdomen, it is ties to
her husbands relatives, male and female, instead of to her husband ex-
clusively or to her own relatives, that dominate her life.
These numerous aspects of a woman's pregnancy phase are compatible
with her bodily condition. By eating a variety of meats and plant foods in
moderate amounts, her throat as well as her abdomen is regularly trans-
formed and made sweet. Though sexually continent, her skin remains
sweet by being bathed in cold water; the cold water cannot transform her
sweet skin to unsweet. She still drinks the hot broth of her boiled food
even after the end of the newlywed phase, but she may also drink cold
water. Her body thus remains wet and sweet as it still retains much skin
and blood.
Simultaneously, however, the pregnant woman is regarded as particu-
larly cold or vulnerable to illness caused by exposure to the hot of certain
categories of male ritual: peace and war sorcery, mefu, bird of paradise
sorcery, courting ritual, aggressive garden ritual. In this one critical re-
spect, the bodies of pregnant women remain the same as the bodies of
brides; namely, they are open. In both cases, some things are regularly
ingested and other things are regularly excreted in moderate or excessive
quantities. With their bodies open, the hot bloody and dirty substances
used in the categories of male ritual mentioned above may easily enter.8
Moreover, because hot bloody and dirty substances (i.e., semen and
80 Quadripartite structures
womb-blood, or the developing fetus itself) are already contained by their
abdomens, it is not necessary for male ritual practitioners to introduce the
nefarious ingredients of their charms into the open bodies of their bridal
or pregnant victims. For this reason, it was explained to me, husbands
take special precautions that their unconfined pregnant wives are never
even remotely exposed to the hot of ritual specialists. I suspect (but was
never explicitly told) that the same reasoning figures in the seclusion of
newlywed brides.
The category open as opposed to closed, as we shall see, is critical to
differentiating the cyclical ritual alternations of males as well as females.
Husband and wife make certain special preparations as the time of
delivery draws near. By then, the wife should have a great stock of
firewood on hand. The husband must erect a fence of bamboo poles
beneath their house to keep village animals away. Within this enclosure,
he digs a hole. Inside the house and directly above this hole, his wife will
give birth and remain throughout the entire postpartum period. Also, a
pair of bamboo knives is fashioned in advance and kept at hand. The wife
continues working all the way up to the day she gives birth.
"Birth" (emauni) and the associated memengoa procedures mentioned
in the Amaka myth (Chapter 4) follow the rupture of the amniotic sac and
the release of its water. As upstream Amaka taught downstream Amaka,
the husband participates up to the point the baby finally emerges. While
his wife squats next to the fire of the household hearth that transforms her
skin to unsweet, he positions himself behind her with his shins along her
back. Reaching around in front, he forcefully works her abdomen with
downward strokes of his hands. These exertions help transform her ab-
domen from sweet to unsweet with respect to the baby so it slides out.
Once it does so, the new father must immediately depart, for he must
have no contact with the infant or the birth-blood. Also, this is the last
contact he will have with his wife for many months.
Close female relatives of either the wife or the husband who possess the
knowledge and experience assist from this point on. Careful not to touch
the baby or any of the birth-blood, one of these women severs the um-
bilical cord with the bamboo knives and deposits the placenta inside the
hole dug underneath the house. The umbilical cord is not tied. The
mother rubs some of the birth-blood on the baby's skin. If it is a boy, she
rubs blood on its shoulders. If it is a girl, she rubs it on the baby's head.
This is said to strengthen and make hard the respective parts of the body
people of different gender use for carrying loads later in life. Then the
mother washes herself and the baby in hot water. After warming her
hands over the fire, the mother gently manipulates the infant's head into
the desired round shape.9
Only the mother herself may touch the baby or any of the other bloody
 Male and female 81
things that come out of her abdomen in giving birth. Other women pre-
sent take every precaution against soiling their hands with the dirty birth-
blood because they must continue to handle and cook food for their
households and for the men gathered on the chief's clubhouse. The fa-
ther's abrupt departure ensures that there is even less chance of him
coming into contact with the blood of birth. Mother and infant, however,
have this hot dirty blood on their skins. The mother must take certain
steps so as not accidentally to eat any of this dangerous birth-blood. She
will do no work. Especially, she cannot cook food for herself or for others,
so other women must cook for her. Also, she must not touch the food she
eats with her hands. Instead, she holds long narrow wooden sticks to
pierce the pieces and place them in her mouth.
It will be remembered from the last chapter that memengoa skills
include the downward stroking of the mothers abdomen with a heated
bamboo tube. This stroking facilitates the ejection of all traces of after-
birth and birth-blood. The same technique (also termed memengoa) is
applied to the site of pain in the curing of illness to force the offending hot
bloody dirt that is causing the pain out of the body. If, after childbirth,
any of the procreative bloods remain in the mother's abdomen, she may
subsequently die. Memengoa procedures, then, employ the hot of fire
and physical exertion to make the mother's abdomen unsweet with re-
spect to the afterbirth and birth-blood so they will come out. Memengoa
sources of hot also contribute toward drying and closing the mother's
body (see below).
The father of the newborn infant maintains a nightly vigil outside the
house to protect his wife and child from sorcerers or other ritual spe-
cialists stealing any of the afterbirth or birth-blood collected in the
ground. He burns a large share of the accumulated firewood to illuminate
the region immediately surrounding the house. The vigil, along with the
mother's abdominal stroking, will be abandoned five or six weeks after
parturition when no more blood flows from the mother's vagina and the
afterbirth underground is believed to be dry and disintegrated.
Memengoa procedures and the father's separation from his family con-
tinue afterward, however. Hot dirty substances still cling to the skins of
mother and baby. As a consequence of nursing, the new baby excretes
wet feces. These wet infant feces are said to be like those of sick people;
that is, they are hot for causing illness in anyone who inadvertently eats
any traces of them. Healthy weaned children and adults excrete dry feces
that, alone, are cold for causing illness. Because nursing is not inter-
rupted until the child is toilet-trained around the age of one-and-one-half
or two, the mother constantly recontaminates her hands as she cleans up
after the baby. Without sop (soap), this hot dirt remains on the mother's
and infant's skin as long as the latter suckles.
82 Quadripartite structures
Although the infant's feces are hot for causing illness, apparently, they
are not regarded as bloody. Nonetheless, these wet bloodless feces are
similar to the wet bloody feces of the chronically ill, for infants and the
infirm have both previously ingested substances that are in a certain
respect analogous with one another, that is, substances inside-out of other
human bodies. However, unlike those bloody things that can cause illness
when eaten, "milk" (kuku) is not dirty. Rather, milk is regarded as sweet
to babies and unsweet to other people. Indeed, instead of fostering illness
and death, milk nourishes human infants and promotes their growth,
health, and life.
Here there are a number of correspondences between nutritive milk
and procreative semen. In their respective contexts, milk and semen
emerge from one body and are deposited into another. Also, both milk
and semen are hot for promoting or sustaining new life. In this respect,
milk and semen are distinct from other dirty human effluvia. Although
milk is not a form of human blood as is semen, it does possess foodlike
elements for blood synthesis. And, of course, milk and semen are the
same color - "white" (kelo) (cf. Turner 1967).
During a woman's postpartum phase, meat relates to her milk in a way
reminiscent of the connection between meat and semen for the newlywed
bride. A nursing mother eats no contraceptive big meat. Instead, she eats
moderate amounts of small meat - above the minimal sustenance level
supplied by her husband or his relatives, balanced with moderate
amounts of boiled plant food. Her body remains generally wet, cold, and
open, yet she does no work. Excess blood-elements from food eaten by
the nursing mother, it seems, mix in her body and are diverted into milk
production. It is interesting to reflect that during their pregnancies wom-
en do not entirely suspend their diet of eating small meat. 10 If a nursing
mother either eats phallic faifai meat or receives semen in sexual inter-
course, her milk will be "spoiled" (ebalifua) and cause symptoms of illness
in the baby. This suggests a connection between a woman's abdomen and
her breasts. It will be remembered here that girls become fertile when
their breasts enlarge. Seligmann also made similar observations among
the culturally related Motu and Koita peoples of the Port Moresby region.
He writes:
The advent of catamenia and the development of the breasts occur together, and I
gathered that this led to the belief that theflowconsisted of blood, which in some
way had come from the breasts, and certainly the fetus was supposed to be formed
of blood derived from the breasts. (1910:84)

According to rules of postpartum memengoa, as long as a nursing mother

remains sexually continent, her vagina leading to her abdomen will grad-
ually close. Consequently, she will neither conceive nor menstruate.
 Male and female 83
Moreover, by not eating contraceptive big meat that would dry her
breasts, she postnatally nurtures her child with milk analogous to the way
semen out of the father's body prenatally contributed to the child's con-
ception. And just as the father's semen dries or coagulates the mother's
womb-blood to form the child, over the course of nursing the infant's
body is progressively dried on a diet of milk. Nonetheless, because the
mother is exclusively dependent upon social ties through the father for
what appears to be the critical ingredient of lactation (i.e., small meat),
the father indirectly continues his previous procreative role (cf. A. Strath-
ern 1972:13-14).
Villagers claim that nursing infants are "not like other human beings"
in several respects. For example, it is often remarked that they can nei-
ther talk nor walk, they are not toilet trained, and they know no rules. In
these ways, babies are "like pigs." n After their first several months of
nursing, Bush Mekeo babies' feces become progressively drier as an in-
creasing variety and quantity of raw plant food and cooked vegetables and
meat are eaten. Finally, once a child walks, is toilet trained, begins to
produce and understand speech, and plays with the other children, it is
time for it to be weaned. This is accomplished when the mother rubs hot
chili juice on her nipples. From that moment onward, the child is ex-
pected to grasp, comprehend, and obey social rules appropriate to age,
sex, and other considerations, just as other human beings do. Of critical
importance in this regard, and emphasized by informants, weaned chil-
dren must be prevented from ingesting dirty things. Parents often ex-
pressed to me their concern that their young children learn and obey the
rules of dirt avoidance and safe handling.
In the terms villagers use to conceptualize them, the gradual processes
of postpartum memengoa, weaning, and dirt enlightenment are quite
similar to upstream Amaka's systematic reordering of downstream
Amaka's confused distinctions between outside, inside, inverted outside,
and everted inside of the human body. Babies and downstream Amaka
and his wife, we might say, are similarly "weaned" or "withdrawn" from
ingesting improper substances (i.e., unsweet and dirty things) and intro-
duced to proper cooked ingestibles by transformations according to vari-
ous hot mediations. It is noteworthy in this context that just as down-
stream Amaka and his wife vomit cooked food the first time they eat it,
human infants are typically understood to vomit on their first introduction
to each new food.
While protractedly nursing their young, village mothers are isolated
from all adult males. The bodies of mothers and suckling infants are wet
and open. Hence, they, like brides and pregnant women, are cold and
vulnerable to the hot male influences of semen, faifai animals, and male
ritual. But because hot and wet dirty feces are being continually re-
84 Quadripartite structures
deposited on their skins, mother and infant are reciprocally hot for caus-
ing illness in men who have not adequately protected themselves from
this influence (see next section).
As the suckling child has drunk more and more of its mother's milk and
eats more and more solid food before it is weaned, the bodies of mother
and infant gradually become drier, less sweet, less cold to male ritual, and
more closed. With weaning and the end of memengoa ritual, these pro-
cesses are complete. In addition, because dirt is being deposited on their
skins no longer, mother and child become cold for causing illness in
others. Mother and child may leave the house of their confinement, and
the nuclear family may be reunited. Husband and wife may now sleep
together at home and resume sexual relations with the possibility of later
conceiving their next child. They can garden, otherwise work together,
and so on. The wife, however, does not resume newlywed ritual with all
its exaggerations. Having borne a child already, there is no pronounced
anxiety that she conceive immediately to reaffirm and validate the marital
relationship. Thus, she eats a moderate balanced diet of plant food and
meat (both big and small) rather than engorging. Nonetheless, although
she and her husband are not expected to perform sex nearly as often as
they did as newlyweds, they indulge themselves at least enough so she
eventually will conceive again and not menstruate. The woman's body
becomes relatively wet, sweet, and open as opposed to dry, unsweet, and
closed. Correspondingly, the wife is cold or vulnerable to male ritual
attack but hot for procreation and work. When she does conceive a child,
she will once more obey the rules of pregnancy just as she did for her
firstborn, and so on.
In conclusion, these ideal Bush Mekeo conceptions of the adult female
reveal cyclic bodily alternations and transformations between open and
closed, wet and dry, sweet and unsweet, and cold and hot as regards
working or not working, vulnerability or immunity to aggressive male
ritual, and procreation or contraception (Figure 5.1). The rules that ac-
company these fluctuations regulate also the flow or lack of flow of diverse
categories of things into and out of her body. And doing so, women
actively open and close the set of boundaries that differentiate the inside,
the outside, the inverted outside, and the everted inside of their own
bodies from the bodies of their husbands, developing fetuses, suckling
and weaned offspring, and, indeed, all other villagers.
The ritual cycle of the male
We have seen that semen and meat particularly, but other male-related
or male-associated things also, are critical elements in the progression of
the female ritual cycle. Indeed, it has been implicit that married men are
themselves intimately tied to the transformations of their wives in much
 Male and female 85
opening ^ ^ closing

wetting ^ ^ drying

sweet ^ ^ unsweet

procreatively h o t ^ ^ procreatively cold

contraceptively cold ^ ^contraceptively hot

Figure 5.1. Female ritual cycle.

the same terms, but with a temporal emphasis upon aggressive male
ritual rather than on reproduction. I will now focus on this complemen-
tary adult-male ritual cycle.
The newlywed groom eats a varied, moderate, and balanced diet of
meat, big and small, and cooked plant food. He drinks the broth his food
is cooked and served in. As a result of ingesting these things, his body
regularly excretes alimentary by-products. The groom is therefore ren-
dered generally sweet, wet, and open, like the bride, with a moderate
excess of blood inside his body. But because the groom does no work,12
his excess blood in the form of semen is diverted into the abdomen of his
bride during their frequent sexual encounters.
Villagers summarily describe the groom as hot for sex, work, procrea-
tion, and nonwork hunting and fishing, but cold for categories of nonwork
such as sorcery, mefu, courting, and aggressive garden ritual. These latter
skills prohibited to the groom involve the manipulation of hot bloody and
dirty things. Lacking these ingredients, hunting and fishing charms are
not hot to the groom who uses them. The other dirty ritual charms,
however, are hot to the groom with his ritually cold and open body, and
they can make him ill or perhaps die. The skills of hunting and fishing,
then, are compatible with the grooms bodily condition and sweet to him.
With a wet, open, sweet body, the groom, or any other adult male
engaged in regular sexual relations, is cold and easy victim to hot war
sorcery worked upon him. Traditionally, war is the greatest threat to
every adult-male villager, at least in the opinion of present-day villagers.
According to the aboriginal system, attack can come at any moment, and
enemies are ruthless in attempting to decimate entire Bush Mekeo vil-
lages. In the event of surprise attack, ritually cold grooms are rendered
heavy and slow by the sorcery of the attackers. They are unable to dodge
projectiles or escape, and they are totally unprepared to manipulate and
wield war sorcery back upon the enemy forces for fear of it transforming
their own bodies. Because the most effective weapons are already cov-
ered with the blood of previous victims, sexually active men cannot safely
use them even in defense. Moreover, if a groom or sexually active man is
86 Quadripartite structures
successful in killing an enemy during a raid, homicidal blood will cover
his skin and unavoidably enter his open body and kill him. As one war
sorcerer remarked, "In the days long before, a man always eating and
sleeping with his wife was a dead man." Because attacking forces usually
have the strategic advantage of surprise and superiority in numbers,
drawing their forces from several villages to attack just one village, it is
imperative that as many adult men of a village as possible be on continual
defensive alert. It was explained to me many times that this constant
threat of war lies behind the anxiety that all married men impregnate
their wives with all possible haste once sexual relations are initiated or
resumed. After conception is affirmed, men ideally suspend sexual rela-
tions entirely and follow a contrasting and countervailing path enabling
them to participate in warfare. This alternate adult-male course is termed
ngope, or "tightening."
Most ethnographers of the Mekeo have rendered ngope to mean "fast-
ing" (Seligmann 1910:333; Guis 1936:35, 176; Williamson 1913:170; Ste-
phen 1974:107). "Fasting" is the gloss English-speaking villagers give to
the term ngope, and systematic restriction of food consumption is indeed
an important element of ngope ritual among the Bush Mekeo. While men
practice ngope, vegetable foods and meat are unsweet to them, so they
"eat nothing" [sic] or, rather, "just one bite" at each sitting - the amount
of food deemed necessary for absolutely minimal body sustenance. But
fasting is neither the only nor the most significant element of ngope ritual.
Hau'ofa has alternatively rendered the Mekeo term ngope as "sacred" or
"set apart," while still emphasizing the element of self-denial (1981:58,
236-9). But among the Bush Mekeo, as I suspect among the Mekeo,
ngope is more complicated than just fasting, and its "sacredness" is rather
more precisely definable in the terms of the culture. The variety of foods
consumed under ngope, for example, is markedly reduced to one kind of
plant food and one kind of meat. And what little food is eaten is always
roasted and dry, never boiled. Alternatively, ngope practitioners regular-
ly consume considerable quanitites of raw chili and ginger. But chili and
ginger are not plant food (fokama), they are purgative "medicines"
(mulamula) - substances that ordinarily are cold and unsweet as food, but
otherwise are hot and sweet for the elimination of lingering traces of dirty
alimentary by-products.13 Married men practicing ngope ritual almost
constantly chew areca and betel with lime for similar reasons (see the
Amaka myth discussed in Chapter 4 and my comments below). According
to other rules of ngope ritual, a man must not drink water from free-
flowing sources. Only minimal quantities of hot fluid from roasted co-
conuts and roasted wild banana tree stalks may be sipped. The ngope
practitioner may bathe only on hot water to keep his skin unsweet. To the
 Male and female 87
same bodily effect, sexual intercourse is strictly prohibited to him, and he
spends most of his time sitting next to a fire.
As a result of these various factors, abdomen, throat, skin, and the body
generally become unsweet. The body appears thin with little skin, little
blood, and little flesh. Bodily excreta, with the exception of areca spittle,
gradually diminish. Also, the body becomes dry. "With no sweat to give,"
a man is cold for work and does none (Guis 1936:176). His body is closed:
"Nothing goes in, nothing comes out, nothing can go through." There-
fore, he is hot for ritual nonwork skills.
Elsewhere in the culture, however, the term ngope refers to "tighten-
ing," as when two poles or sticks are bound together with vine so as to
close the space or opening between them. As part of his ngope regimen, a
man wears a bark belt severely tightened around his midsection, giving
him to European eyes a wasplike appearance. This notion of tightening, I
think, comes closer to summarizing the several indigenous aspects of
ngope ritual than do the more narrow gloss of "fasting" or the Western
concept of "sacred."
By rigorously tightening his body, a man prepares himself for practic-
ing a number of nonwork skills: war, peace, or bird of paradise sorcery,
and courting and aggressive garden ritual. The charms used in these male
arts contain hot bloody and dirty substances. Tightening essentially pre-
vents the contents of the charms from entering the practitioners own
body while he prepares and manipulates them. Otherwise, they would
enter and transform him as his own victim. Tightening, in other words, is
a ritual prophylaxis against the effects of hot bloody and dirty things. To
him who has tightened himself, the charms are cold or impotent. They
cannot enter. But to others who have not tightened themselves - whose
bodies are wet, sweet, open, and ritually cold - the charms are hot and
potent. Moreover, by the very act of manipulating his charms, the practi-
tioner deposits hot bloody dirt on the outside of his body, on his own skin
and hands. In this way, his own person is hot to nontightened villagers.
Therefore, he must not wander freely through the village or even the
bush for fear of unintentionally making victims of others. He must remain
secluded in the extraordinary aboveground holes or places of village or
bush, or run and hide whenever the chance of accidental encounter
arises.
Let me now return to the predicament facing the groom once his wife is
impregnated. Ever fearful of surprise enemy attack, he cannot immedi-
ately handle the charms, weapons, and other paraphernalia of war with
his open body. Tightening is a gradual process. It requires approximately
six months to make a man's body sufficiently dry, unsweet, and ritually
hot so that its inside is closed from its outside. During the course of his
88 Quadripartite structures
wife's pregnancy, the tightening groom becomes increasingly capable of
performing the dangerous ritual male arts, but initially his body is still hot
for work. Therefore, in a flurry of garden clearing immediately following
the recognition of his wife's pregnancy, he can cut down the largest of
forest trees before he becomes cold for work. Afterward, his wife, as her
pregnancy progresses, can perform the tasks that remain to be done - the
lighter clearing, planting, and tending. Because hunting and fishing are
nonwork skills, the groom can continue to supply his wife and others with
meat, at least as long as he does not contaminate his hands with hot
bloody dirt.
It is certainly no coincidence that, according to the culture, the mo-
ment hot bloody dirt emerges from his wife's body in parturition ideally
coincides with the completed closing of the inside of the man's body from
its outside. His abrupt departure from the scene and his ensuing avoid-
ance of wife and child are motivated in part by considerations for their
welfare. The husband can now safely handle hot ritual charms because
they are cold to him. But if he does so with any birth-blood remaining on
his hands, his wife and infant child will sicken and perhaps die when he
later attempts to use the charms on other persons. Just as the tightened
husband, without danger to himself, keeps vigil so that peace sorcerers do
not steal any of the birth-blood or afterbirth to kill his wife and child, he
ensures that he himself does not unintentionally do the same with their
blood on his own hands. Incidentally, this consideration further illumi-
nates the mythical imperative that downstream Amaka not handle bloody
substances of birth and memengoa (see Chapter 4).
Obviously, the entire adult-male population of a village cannot simul-
taneously and indefinitely tighten. Whereas bachelors and widowers are
expected to tighten themselves continuously, married men possess an-
other alternative geared to the ritual oscillations of their wives. During a
woman's pregnancy, postpartum memengoa, and nursing phases, her hus-
band tightens his body preparing for or engaging in exclusively male-
oriented nonreproductive pursuits. Marital procreativity regularly inter-
rupts the husband's male ritual, but only infrequently and for relatively
brief periods. At any one time, most men of the village, married and
unmarried, are in a more or less tightened condition of war-readiness.
Unless an actual battle is planned, the advent of weaning each child
corresponds with the return of its father to indulgent sexual relations with
its mother. However, it is necessary that a man continue tightening him-
self one season - roughly six months - after he has last handled his ritual
charms. Until then, traces of hot bloody and dirty substances remain on
his skin. Sexual intercourse before that time would prematurely open the
bodies of a man and his wife, and this could result in their deaths. Thus,
 Male and female 89

closing^ ^ opening

drying ^ ^ wetting

unsweet ^ ^ sweet

procreatively cold ^ ^ procreatively hot

contraceptively hot ^ ^ contraceptively cold

Figure 5.2. Male ritual cycle.

the husband abandons his ritual activities but continues to tighten until
his child is weaned.
This achieved, he and his wife may now safely sleep together and
resume regular sexual intercourse. He may drink moderately, eat food in
greater quantity and variety, and so on. The husband's body thereby
reverts to a comparatively wet, sweet, open, and ritually cold condition.
However, for procreation and for working (particularly garden work), he
is once more hot. After a theoretical minimum of three months, his wife
may conceive again. Then he will begin to tighten himself once more and
so repeat the male ritual cycle (Figure 5.2).
For the Bush Mekeo, tightening ritual suggests more than a physical
setting apart or isolation of husband from wife. In several respects, tight-
ening represents a general withdrawal of men from the company of wom-
en and the corresponding symbolic accentuation of the opposition of male
to female. This is illustrated most dramatically during the informal initia-
tion of "boys" (imi) into "bachelorhood" (koaekongo) (cf. Seligmann
1910:256; Williamson 1913:273; Guis 1936:17-42; Hau'ofa 1981:130).
Sometime between the ages of eight and twelve, boys leave their moth-
ers, sisters, and other female relatives and begin tightening in the com-
pany of other men. A bachelor initially may not cast eyes upon a woman
nor even children of either sex who are in daily contact with their mothers
and female relatives. What little bachelors do eat and drink is prepared by
men only and according to the strictly male culinary skill, roasting. At
first, bachelors are confined to the rear compartment of their chiefs
clubhouse where women may never enter. Moreover, the only males,
married or unmarried, with whom bachelor novices may interact there
are those who have also completely separated themselves from women by
tightening. It is these men who teach the bachelors the hereditary secrets
about sorcery and other ritual that only male members of the clan may
share. Once their initial confinement is over, bachelors may emerge from
the rear compartment of the clubhouse, at least at night, but they may
never cross the village abdomen - a distinctly feminine part of the village.
90 Quadripartite structures
Male tightening thus represents categorical opposition of male and female
as much as it involves extinction of interaction between men and women.
It must not be forgotten, however, that women, too, analogously dis-
tinguish and separate themselves from men during the closing phase of
postpartum memengoa ritual. I shall return to this point below.
Gender differentiation and areca nut chewing
The contrasts between indigenous conceptualizations of male and female
can be substantially sharpened by digressing briefly into the rules that
involve the chewing of areca nut.
Areca (mat) is unsweet for ordinary eating, as it is categorically not even
a plant food. However, when it is mashed and boiled in water and drunk,
areca is a medicine that is sweet and hot for the curing of some kinds of
illness, particularly of the abdominal region. In illness, recall, what is
ordinarily sweet to the body becomes unsweet, and what is unsweet
becomes sweet. As a medicine, boiled areca, like chili and ginger, is hot
for reversing the transformations of illness and returning the body to a
state of health. Specifically, areca allows the body to retain the blood
elements of food inside and to excrete the bloodless elements of food
along with the invading foreign bloody dirt to the outside.
When areca is chewed, it has similar connotations of purging sub-
stances from the body. Villagers assert that when they chew, they swallow
(i.e., "eat") none of the pulp or juice; all is spit out of the mouth. Alone,
areca is unsweet even for chewing. But when it is mixed with betel
pepper and lime, it is transformed to sweet for chewing. Nonetheless,
chewing is distinctively sweet to males and females only at particular
points of the respective ritual cycles. Only men and women who are
married may chew. Maidens and tightened bachelors may occasionally
experiment with chewing, but they must keep this fact, along with their
sexual exploits, secret from their elders. The rules that specify when
married people may and may not chew indicate additional conceived
gender differences in the culture and, further, that chewing is closely
associated with adult sexuality whereas not chewing corresponds with
sexual contraception.
These relations are partly reflected by chewing's conceived effects
upon the adult body. Chewing is said to be hot for making the abdomen
feel sweet as though it were satiated or full with plant food, even when
none has been eaten. Chewing will not similarly transform the throat to
sweet in simulation of meat satiation. Rather, chewing dries the mouth
and thereby closes the throat as if meat had been eaten. Moreover,
chewing areca is said to transform the skin from sexually sweet to unsweet
because it dries from the lack of water or sweat.
A man and woman publicly chew for the first time when, as newlyweds,
 Male and female 91
they take up residence together as groom and bride. Thus, in the period
they engage in intensive sexual relations, they chew. Indeed, for the rest
of their married lives, chewing will be sweet for men and women when-
ever sex is also sweet for them. Married men who have been sleeping
with their wives gather each evening on the exposed portion of their
chiefs clubhouse and chew areca and betel with lime once they have
finished eating the boiled food brought to them by their wives. At home,
their wives eat and then chew as well. Because chewing transforms the
skins of husbands and wives from sweet to unsweet immediately following
the eating of food and just prior to retiring, it encourages sexual inter-
course by making their skins unsweet and desirous of sex. Also, it will be
remembered from the Amaka myth that downstream Amaka encouraged
his namesake to chew improperly with the dirty secretions of his wife's
vagina after they finished eating food. And when upstream Amaka cor-
rectly chewed with lime instead, downstream Amaka mistook the red
spittle for blood. In my analysis of the myth, I argued that proper chewing
with lime represents body category distinctions that are opposed to the
improper and potentially dangerous ingestion of dirt that may accompany
sexual encounters between males and females. Thus, in addition to mak-
ing married partners desirous of sex after eating and before sleeping
together, chewing enables men to avoid taking female sexual fluids into
their bodies, and it enables women to receive semen in their abdomens
but separate from the inside of their bodies. In more general terms,
chewing promotes sex but preserves the conceived bodily boundaries
separating procreative male and procreative female.
The categories male and female differ further with respect to chewing
during the periods of their married lives when they are sexually absti-
nent. When married men begin to tighten their bodies ritually, they
radically intensify their chewing, as paradoxical as that might appear.
From the time women become pregnant until they wean their infants, on
the other hand, they abstain entirely from chewing.
Intensified chewing for the tightened male accentuates or exaggerates
the differences between himself and the sexual female. His body is drier
and more closed, and his skin considerably more unsweet. He becomes,
therefore, supermasculine - not for sexually inseminating a female, but
for ritually depositing bloody dirt from his charms into his cold, wet,
sweet, and open victims. In the sense that male ritual is an analogue or
metaphor of male sexuality, the tightened male, although he chews and
sexually abstains, actually becomes more intensely sexual.
In the same way intensified chewing further accentuates males as dis-
tinct from females, abstaining from chewing differentiates females from
males. Earlier I argued that sexually active women, by chewing, keep
semen - a male substance - out of the inside of their bodies (i.e., re-
92 Quadripartite structures
stricted to their abdomens). By not chewing, it would seem logical that
some similar male substance is not kept from the inside of women's
bodies. In this case, because I unfortunately lack precise ethnographic
details, I will suggest that two such male substances are permitted inside
a woman's body when she is pregnant and nursing. When she is pregnant,
this would be contraceptive big meat. And when she is nursing, this
would be lactative (or procreative) small meat. During pregnancy when
she does not chew, a woman eats big contraceptive meat to prevent
further impregnation and sustain her body and the body of her develop-
ing fetus, respectively. Following the birth, however, she eats small meat
so there will be milk for lactation. To chew areca now would dry up her
milk. Analogously, if she performed sex or ate faifai at this time, then
semen or faifai blood could get into her milk, spoil it, and make the baby
sick.
Female inversion/eversion of male
I have shown that in their married lives men, like women, cyclically
alternate between two extremes characterized in terms of the same binary
oppositions: dry versus wet, unsweet versus sweet, closed versus open.
Generally speaking, however, it is clear that husbands and wives repre-
senting male and female "move" simultaneously in opposite directions.
At the birth of his child, the tightened husband is at his most masculine,
his body closed, dry, and unsweet. The wife at this time is at her most
feminine, her body open, wet, and sweet. Later, when the wife weans
her child in anticipation of resuming sexual relations, her body is rela-
tively closed, dry, and unsweet. And at the moment her husband
achieves sexual union with her, his body is relatively open, wet, and
sweet. Thus, all four categories of gender differentiation are temporally as
well as conceptually contrasted with one another with the cyclical alterna-
tions of the marital relationship.
Viewed from another perspective, however, these four gender catego-
ries are all analogous to one another. In each case, there are interpersonal
transmissions of bloody or blood-potent substances. Open and pro-
creatively hot males transmit semen to open and procreatively hot
females. Closed and procreatively cold males transmit dirty blood to their
victims in aggressive male ritual. Open and procreatively hot females
transmit womb-blood to their developing fetuses, and closed and pro-
creatively cold women transmit milk to their nursing infants. There is,
furthermore, a distinct sense in which the transmissions implicated for
the two female categories are dependent upon male transmissions, that is,
to provide either womb-blood or milk to their progeny, married women
are dependent upon men for semen and small meat. By contrast, trans-
missions indicated for the two male categories are not predicated upon
 r
male
(separated from female)
hot male ritual cycle >

male
(in contact with female)
r
female
(in contact with male)
cold fem<lie ritual cycle

V
•

female
(separated from male)

closing opening opening closing

drying wetting wetting drying

unsweet sweet sweet unsweet

procreatively cold procreatively hot procreatively hot procreatively cold

contraceptively hot contraceptively cold contraceptively cold contraceptively hot

f 11
1
^^
^ ^
heterosexual
reproduction

Figure 5.3. Alternating gender categories.

f
94 Quadripartite structures
prior and analogous transmissions to them from women. For Bush
Mekeo, there is, in other words, a priority of the category male with
respect to the category female.
This priority of male over female relates, once again, to the opposition
of hot to cold. To clarify this point and to relate the categorization of
gender to other cultural contexts, it will be necessary to address the
apparent context-sensitivity or context-relativity associated with each of
the four gender categories.
Closed male and closed female, as I have shown, are hot in the context
of contraception and aggressive male magic, but cold in the context of
procreation. Open male and open female, conversely, are cold for con-
traception and aggressive male ritual, but hot for procreation. Thus, each
gender category is simultaneously both hot and cold with respect to the
two opposed contexts.
However, the terms hot and cold are often used in a somewhat differ-
ent and context-insensitive manner to distinguish the category male and
men generally from the category female and women generally. Irrespec-
tive of context, villagers sometimes flatly declare that men are hot and
women cold. For example, although an open, sexually active man may be
described as cold in the context of contraception or aggressive male ritual,
and a closing sexually inactive woman as hot in this same context, the
former as a male is unambiguously hot in opposition to the latter, who as a
female is cold irrespective of context (Figure 5.3). This usage of hot and
cold may indeed seem partially contradictory, but it nonetheless reveals
the underlying homology between the indigenous classification of gender
and gender roles, on the one hand, and categorizations of sexually dis-
tinguished bloods, food categories, spatial and body relations, and so on,
on the other:
closing male opening male : : opening female closing female (14)
hot dry hot wet :: cold wet cold dry (7)
male blood semen :: womb-blood female blood (13)
male male female female
contraception procreation procreation contraception (11)

cookable noncookable
big meat small meat : (12)
plant food plant food
roasting boiling :: ripening smoking (6)
blood sweet food : : unsweet food bloodless dirt (8)
peripheral village
remote bush adjacent bush (3)
village abdomen
inverted everted
inside outside : (4)
outside inside
The terms of the final analogy listed here - inside and outside - reveal
perhaps more distinctly than elsewhere the systematic conceptual rever-
 Male and female 95
sal that specifies the opposition of female to male and, in particular, the
cultural designation of the latter as prior to the former. The categories
"procreative womb-blood" and "procreative opening female" are inver-
sions, respectively, of "procreative semen" and "procreative opening
male." Correspondingly, "contraceptive female blood" and "contracep-
tive closing female" are eversions of "contraceptive male blood" and
"contraceptive closing male." In a strictly formal sense, of course, the
reverse is also true, but in the culture of the Bush Mekeo one set of
categories (those homologous with the female) are always defined subse-
quently to the other (those homologous with the male), and not vice
versa. This is not to be simply interpreted, however, as a systematic
"male bias" in the culture, such that males and male-associated things and
categories are valued more than females and female-associated things and
categories, although in certain contexts this may indeed be the case (cf.
Weiner 1976; Ardener 1975; Rosaldo and Atkinson 1975). Rather, I am
drawing attention here to the observation that things and categories ho-
mologous with maleness are viewed in the culture as logically prior to
things and categories associated with femaleness.
In the next two chapters dealing with categories of social interre-
lationship (e.g., kinship, clanship, affinity), I shall develop this point
more fully. Before I conclude this chapter, however, it will be helpful to
anticipate this extension or expression of the gender divisions into the
realm of social relations by counterposing and analyzing the two myths
that serve as "charters" at both levels simultaneously. It should be noted
that the cultural priority of maleness just mentioned is explicit in these
two myths. The first myth, dealing with the life of a man named Foikale,
tells how men, who long ago lived in the bush, came to live in the village
with women. The other myth purportedly explains how village women
once left their menfolk and became flying foxes of the bush. Both myths
are regarded as particularly "big" or "important" (apounga) by knowl-
edgeable villagers. I have documented full versions of both as appendix-
es, including Guis's (1936:189-93, 220-6) accounts, which were current
within a generation of contact with Europeans.

The myth of Foikale

In the earliest recorded rendition of the myth of Foikale (see Appendix 6),
Fi5 a man named Foikale lives with other men in caves underneath the
mountains (an underground hole of the remote bush). Foikale and his
comrades utterly lack gardens, plant food of any kind, hunting and meat,
fire, clay pottery, water, and women. Instead, they eat a kind of soil
(pkimo) baked in the sun, and give birth themselves without women or
even knowledge of women.14 Wandering through the bush one day,
Foikale stumbles upon the garden of Oa Lope,15 of whom he was not
previously aware. In each of their sequential encounters thereafter, Oa
96 Quadripartite structures
Lope imparts to Foikale knowledge of one of the things he has lacked in
his bush existence - gardens and plant food, fire, clay pots, water for
cooking, meat and hunting - and, each time, Foikale returns to his com-
rades of the cave to share the new discoveries with them. Whenever
Foikale and the cave men eat a new food for the first time, they vomit it,
as in the Amaka myth. When Foikale first sees a woman, Oa Lopes wife,
he asks, "What is that?" At the end of the myth, Oa Lope gives Foikale
his own daughter in marriage, his "last" and "great" gift, so that Foikale's
children will be numerous and his village (sic) prosper. Thereafter, youths
of Oa Lope's group marry daughters of Foikale's, and the men of Foikale's
group take wives from Oa Lope's.
Versions Fu and FUi of the Foikale myth closely follow the basic outline
of F( but with some variation of detail worth mentioning. Fu begins with
Foikale in a similarly deprived bush existence, except that he eats raw
plant materials rather than soil or cooked plant food. Upon their meeting,
Oa Lope bestows on Foikale knowledge of gardening, woman's vagina
and heterosexuality, houses, villages, fire, cooked food, areca chewing,
and clothes. Oa Lope does not explicitly give Foikale his daughter, but
the story concludes with Foikale and his male companions moving to Oa
Lope's village and enjoying all the new amenities, including women,
given by Oa Lope.
In version, FiU, the bushman Foikale acquires gardens, fire, roasted
plant food, and, lastly, water out of the pierced ground from Oa Lope.
Everything Oa Lope gives Foikale in all three versions is sweet, associ-
ated with village life, and, in F( and Fu, with women, sexuality, and
procreativity explicitly. FiU makes no reference to women per se but
otherwise duplicates elements of either Fiy Fu, or both. Foikale's initial
remote bush existence, lacking proper cooked foods, fire, pottery, water,
chewing, housing, clothes, villages, and women, is strictly unsweet and
strictly male. The final village life he adopts, by contrast, has all these
things, and it is overwhelmingly sweet and heterosexual.
Several features in the three versions of this myth repeat elements of
the Amaka myth. Foikale initially warms his food (sic) in the sun (F{).
Foikale receives fire and cooked food from Oa Lope (Ft, Fu, FiU), and he
vomits the first time he eats each new food category (Fit FiU). After
Foikale becomes knowledgeable of woman's vagina and its possibilities for
sexual intercourse, he misuses it, as downstream Amaka did, for chewing
with areca and betel, and he is frightened, after properly chewing with
lime, of blood (sic) coming out of his mouth (Fu). Foikale and his male
bush companions acquire wives who will give birth to their children as a
result of sexual intercourse, but here without the complication of killing
the mother (Fit Fh). Finally, the Foikale myth like the Amaka myth
systematically differentiates bloody and bloodless dirty things from un-
sweet and sweet clean things as they relate to the village/bush distinction.
 Male and female 97
Although Foikale is still a man of the Bush, he twice distinguishes
himself as nonetheless human from what he mistakes for "snakes" (ba-
nanas and rising smoke), that is, from nonhuman people that like faifai
also live in holes of the bush (F{). Similarly, Foikale is at first frightened
by fire (FJ, or asks if fire will bite him (Ff). Oa Lope also ensures that
Foikale does not chew dirty vaginal secretions with betel and areca (F£i).
In other words, Foikale abandons an unsweet bush life, free of bloody
dirt, for sweet village amenities wherein bloodless dirt arising from asso-
ciation with women is prevalent. However, Oa Lope, like upstream
Amaka, teaches Foikale how to distinguish significant categories of things
in terms of ordering the relationships between men and women, and
between groups of people related through males (Fi? Fu).
Foikale's two life styles, first in the bush and then in the village, ex-
press certain features of the contrasting phases of the male ritual cycle
described above. Foikale's strictly male, nonsexual or sexually cold, un-
sweet, dry, and closed bush existence corresponds symbolically with the
tightened contraceptive condition. Foikale's life after receiving Oa Lope's
gifts corresponds in turn to the sexually hot, wet, sweet, and open pro-
creative phase of a man's married life. As Bush Mekeo males alternate
between these two extremes in their adulthood, they express or portray
the alternative life styles of the mythical figure Foikale.
The story of afinama
The other myth I will discuss here describes the origin of flying foxes
(afinama; also known as fruit-bats) (see Appendix 7). It, too, deals with
alternate conceptual modes of interrelationship parallel to the distinction
of village and bush, but this time the focus is upon the female. It will be
worth mentioning at the outset that flying foxes eat only raw plant food
(breadfruit, banana, pawpaw) and migrate in enormous numbers each
evening during the dry season (May through November) from swamps to
the forest, gardens, and villages where their foods grow.
The three recorded versions of this myth begin very much where the
Foikale tale leaves off- men and women are living together at the village.
Each day, men leave their women at the village and hunt pigs, casso-
waries, and wallabies in the bush (Ai? Au, AiU). Rather than sharing their
catch with the women, the men either roast and eat all the meat them-
selves before returning home and lie to the women that they caught
nothing (A{, Au), or the men give the meat to the women to cook but still
eat all of it themselves, leaving the women only the broth to drink (Aiit).
As days go by like this, the women become meat-hungry, angry, and
shamed by the men (Ai? Au, AUi). The women discuss their resentments at
a meeting.16 Leaving behind their male children while the men are gone
hunting (Af) and throwing everything (blankets, skirts, cooking pots, etc.)
of theirs away (Aif), the women fashion wings of coconutflockand fly "like
98 Quadripartite structures
birds" to a tall tree (A{, Aip AiU) at the end of the village (AiU). Sitting right
side up on the branches, their vaginas are visible from below, so they all
hang upside down (Ai9 Au, AiU). The men return and try to get the women
and girls to come back down, crying and offering them meat, but the
women refuse (Af, Aii9 Aiit). The men cut the tree down, and the women,
now flying foxes, fly away (Ai9 Au, AiU).
At this point, the three versions end with conflicting results. On the
one hand, because the meat of flying fox is "woman's flesh," it is declared
dirty and inedible (A^). On the other, a pregnant women (Au) or a "weak"
(apoke)17 woman (AiU) crashes to the ground and dies. Men discover that
the meat of flying fox is not dirty by feeding it roasted to a boy who does
not afterward sicken. It is then declared that men may eat this new kind of
meat (Aw AiU).
Meat figures prominently in this myth in more than one respect. By
denying women big contraceptive meat, men of the village lose to the
bush their females and any children they might otherwise bear in the
future. It is significant that in one (Au) and perhaps a second (AUi) version,
a pregnant woman, unbound with meat but still bound by semen to men
at the moment of flight, is unable to leave, falls, and dies, and the men eat
her flesh. These versions express once again the symbolic relations of sex
to meat and semen, but also the novel proposition that women, unbound
with meat and with semen to men, may escape. In turn, they become
potentially edible small meat themselves. It is implicit that if men had
supplied their women with meat, they would not have lost them or their
progeny to the bush, and the question of eating woman's flesh would not
be problematic, at least in this context.18 By contrast, in the Foikale
myth, there was no suggestion that Foikale's male flesh was edible when
Oa Lope first encountered him in the bush (F{, Fu, FiU). Yet the edibility
of female flesh in the form of flying fox remains a possibility once women
leave their men for the bush.
The afinama myth, like the Foikale myth, places the two alternative
ritual female conditions into the context of village in opposition to bush.
Women hold their meetings on ground of the village abdomen, then flee
to the uncultivated tree standing in the nearby or adjacent bush at the
end of the village. The women who become exclusively female flying
foxes of the adjacent bush, therefore, significantly compare with the ex-
clusively male Foikale of the remote bush before he met Oa Lope. Also,
the women, who at the beginning of the myth gather in the village ab-
domen and then at the end fly to the bush, invert and evert, respectively,
the end and the beginning positions of the males in the Foikale myth.
Women thus end up occupying the bush, and men are left residing alone
at the village. In terms of village and bush, men and women have effec-
tively exchanged positions. The categories village and bush are conse-
quently bisected by the male/female distinction.
 Male and female 99
The mythical focus upon meat here once again recalls the pivotal role of
meat in the ritual of the female ritual cycle and suggests a relation be-
tween this myth and female ritual comparable to the relation between the
Foikale myth and male ritual. When the women asflyingfoxes leave their
possessions, their men, and their village for the adjacent bush, they
replicate the contraceptive withdrawal of women from men during the
postpartum and nursing phases. Women, like men, have two opposed
conditions or states - one procreatively enjoined with men, sexually hot,
sweet, open, and wet; and the other contraceptively separate from men,
sexually cold, unsweet, closed, and dry.
This digression into myth, I think, confirms the main points made
throughout this chapter regarding the complex, complementary, and
cyclic oscillations implicit in male and female ritual and the conceptualiza-
tions of gender difference. Married adult Bush Mekeo men and women
ritually and symbolically "do much the same thing" as one another and
roughly according to the same schedule. Nonetheless, women "do it" as a
reversal of the male. With respect to the culturally ordained logical pri-
ority of male to female, Foikale and his comrades at the beginning of the
first myth exist entirely separate from women and female-associated
things and qualities. The flying foxes at the end of the second myth,
however, do not absolutely escape the world of village males they leave
behind. Nor do they take up the same underground place of residence
occupied by Foikale initially.
Moreover, the two myths suggest opposed conceptual affinities of
opening and closing to ordinary and extraordinary spheres of village life
for males and for females. In the Foikale myth, males open; in the
afinama myth, females close. In terms of spatial transfers, the two myths
together express unidirectional transfers of the ordinary sphere of village
life. There is movement first from remote bush to peripheral village, and
then from village abdomen to adjacent bush. In their respective ritual
cycles, however, males and females also "move" the other way - men
close and women open - and in these phases both males and females are
compatible with bi- or ambidirectional spatial transfers between holes of
the extraordinary sphere of village life. Apparently, then, the myths es-
tablish male opening and female closing as unidirectional processes com-
patible with the ordinary sphere, and their ambidirectional reversal (i.e.,
male closing and female opening) as compatible with the extraordinary
sphere. And once more, the categories male and female are distinguished
in precisely reversed terms.
 6
Kin, clan, and connubium

Based upon his reading of Seligmann's Mekeo ethnography, Levi-Strauss

observes:
The plan of their social organization is a subtle and complex symmetry, and the
historical vicissitudes to which its componenet elements have been exposed have
never succeeded in abating its strictness. (1969b:77)

Although there are differences between Bush and Central Mekeo social
organization chiefly as a consequence of differences in scale, the "plan" in
each case is, I think, virtually the same. The avowed purpose of this
chapter is to elucidate as explicitly and simply as possible this subtle and
complex symmetry that structures Bush Mekeo social relationships. To
achieve this end, I relate the various categories of Bush Mekeo social
classification to one another and to other contexts of the culture (Leach
1976; Schneider 1972, 1980; cf. Lounsbury 1965). I begin by describing
and interrelating the set of kin categories in the indigenous language, the
various idealized levels of unilineal clan fragmentation and residence, and
the formalized interclan affinal relationships. Then I trace the logical
interplay between these conceived cultural formations and the rules of
marriage and marriage compensation. In anthropological jargon, the ide-
alized system I describe is one that prohibits marriage between all first
cousins while permitting (if not prescribing) marriage in alternate genera-
tions between children of first bilateral cross-cousins who belong to differ-
ent patrimoieties (i.e., between classificatory second cross-cousins). In
view of comparison, this formulation of the basic Bush Mekeo system in
its overall structural properties resembles the classic "Aranda'-type sys-
tems of Australia except that it lacks named sections or subsections but
possesses as alternatives both "Hawaiian"- and "Dravidian"-type kin-
term classifications.
Much more importantly for my purposes, however, I reveal a number
of structural homologies between the categories of social relationship and
the other contexts of category distinction in the culture I have already
elucidated. The entire nexus of Bush Mekeo symbols regarding reproduc-
tion is of particular significance here. The indigenous theory of human
100
 Kin, clan, and connubium 101
reproduction, recall, involves the resolution of a critical contradiction at
the level of the mating couple, whereby hot dirty male blood (semen)
assimilated by the woman's abdomen procreates new life rather than
illness and death. A related contradiction is specifically implicated for
human reproduction at the societal level. In spite of exogamous exchange
of women and the incorporation through them of female blood, groups of
purported agnates maintain an ideology of exclusively male or patrilineal
blood relatedness. Alternately phrased, at the level of the mating couple,
male blood (semen) challenges the blood-purity of the female body; at the
societal level, correspondingly, female-derived cognatic blood challenges
the integrity of social groups defined in terms of male blood and agnation.
The societal-level contradiction, then, is between blood relationship con-
ceptualized unilineally through males and blood relationship concep-
tualized bilaterally or cognatically through females as well (see Moore
1964; Leach 1966a; Douglas 1968, 1970; A. Strathern 1972; Denich 1974;
Weiner 1980; Hage and Harary 1981). I shall outline the resolution of this
contradiction in terms of the indigenous culture categories and their un-
derlying quadripartite structure.
Before moving on, it will prove helpful to indicate that the actual terms
and form of this resolution have already been foreshadowed in the coun-
terposing of the Foikale and afinama myths at the end of the previous
chapter. Beyond chartering the conceptual bisection of the category adult
male, the Foikale myth enjoins the creation of a society structured by the
"restricted exchange" (Levi-Strauss 1969b; Needham 1962) of women
between the two groups, the members of each being initially related
exclusively through males. The afinama myth, aside from inversely
bisecting the characteristics of the adult female, describes the severing or
abrogation of ties between affinally related groups of agnates through the
loss of the women previously exchanged. The reciprocal exchange of
women between nonagnates, in other words, is negated by the loss of the
women themselves from male society. How these complementary pro-
cesses expressed even in these mythological terms resolve the contradic-
tion of societal reproduction more precisely in Bush Mekeo culture will
become evident when I turn specifically to the analysis of marriage ex-
change and compensation.
Relationships of blood and nonblood
The evidence of preceding chapters testifies to the fundamental signifi-
cance of the blood/bloodless opposition in virtually every context of Bush
Mekeo culture. Indeed, it may well be said that several of these contexts
are explicitly related to one another in the culture through successive and
homologous bisections of the initial duality of blood and nonblood. All
substances in the world, for example, are distinguished in terms of:
102 Quadripartite structures

ii J sweet :: unsweet : U1 J1 , /ON

blood : /i i j ,. ^ /ui J • * IA bloodless dirt (8)
(blood-potent) (blood-impotent)
And different kinds of human blood are subsequently and homologously
categorized in terms of:
male female female , ^
body-blood ' womb-blood ' body-blood
The realm of social relationships for the Bush Mekeo is likewise initially
conceptualized in terms of blood and nonblood, and subsequently at less
inclusive levels social relationships are systematically distinguished ac-
cording to homologous bisections of the blood/nonblood opposition.
Although the peoples now known as the Bush Mekeo were never politi-
cally unified in any fashion, there seems to have existed among them
traditionally a self-awareness as a single cultural and linguistic community
as distinct from their neighbors - the Central Mekeo, the coastal Roro
and Toaripi, and the mountaineer Lapeka and Kuefa groups. Within the
Bush Mekeo population, the highest level of political organization is the
"tribe," of which there are two - the Amoamo and the Kuipa. In the days
before colonial rule, the two tribes were hereditary and irreconcilable
enemies. "Warfare" (aoao) is expressed in the language in terms of an
exchange of blood between tribes. This ritualized intertribal reciprocity of
blood, moreover, contrasts with the pattern of shifting alliances and hos-
tilities between each of the Bush Mekeo tribes and the other n o n - B u s h
Mekeo groups in the area. In other words, the two Bush Mekeo tribes
were related to one another in war reciprocation of blood in a manner
absent from the less formalized relations of the Bush Mekeo to their
neighbors. Thus, in the context of war reciprocation, the two Bush Mekeo
tribes are traditionally conceptualized as " o n e " or the "same blood" (ifa
alakangamo) in contrast to the "different bloods" (ifa ikoina) of other
groups:

(a) same, different, Bush non-Bush

one blood ' nonblood " Mekeo peoples ' Mekeo peoples
The blood/nonblood opposition also corresponds to the distinction
"own Bush Mekeo t r i b e " / " e n e m y Bush Mekeo tribe" in the context of
reciprocating blood through marriage, sex, and procreation. With superi-
ority in war nearly always on the side of the attacking forces, it was of
utmost strategic importance that the location of villages b e kept secret
from enemy tribesmen. For this reason, my informants argued, the two
Bush Mekeo tribes never intermarried. W o m e n taken from the enemy
group might reveal to their brothers the location of villages of their hus-
bands' tribe. Thus, each tribal unit is traditionally an endogamous group
and remains largely so even after a century of pacification. Moreover,
 Kin, clan, and connubium 103

warfare as it was waged between the two tribes was never practiced
within a tribe. In contradiction to Tylor's dictum, the Bush Mekeo do not
marry whom they fight and do not fight whom they marry. l
In recognition of many generations of endogamously transmitting blood
cognatically among themselves, members of a tribe regard each other as
one or the same blood as distinct from the different blood of the enemy
Bush Mekeo tribe:
(b) same, different, own Bush enemy Bush
one blood ' nonblood " Mekeo tribe ' Mekeo tribe

The boundary of own blood/nonblood for war, in the context of intermar-

riage, is bisected.
I shall have more to say below regarding the conceptual relations of
blood conceived in war exchange and blood conceived in peace and mar-
riage exchange.

Blood relatives or kin

Blood shared among members of a tribe as a consequence of common
reproductive ancestry is expressed in terms of bilateral kinship or, sim-
ply, "relatives" (atsi atsitsi).2 Relatives are said to be "born together" or
"born in lines" (mauni panini) and to share blood by virtue of the mixing
and transmission of procreative bloods from the parents to children con-
sistent with the prevailing theory of reproduction described in Chapters 4
and 5. By extension, people who are connected with one another through
consecutive bilateral parent-child links are one and the same blood.
Moreover, just as ones own body blood is sweet to oneself, one's own
relatives by blood are sweet. Relatives, in other words, "like" (kernitisa)
one another, and ideally they should not fight.
At the level of the tribal unit, blood and relatives or kin are concep-
tualized uniformly. However, within the endogamous tribe villagers rec-
ognize a number of additional discriminations or distinctive categories of
blood and relatives that correspond with what the anthropologist custom-
arily regards as a system of kinship nomenclature and classification. The
major system of the Bush Mekeo qualifies as a variant of the generational
or "Hawaiian" type (Lowie 1968; Kroeber 1909; Spier 1925). Beyond
describing the main outlines of this system, I shall indicate a number of
terminological alternatives for cousins suggestive of a "Dravidian" classifi-
cation that also have profound sociological significance, particularly as
regards the dimensions of blood relationship (cf. Scheffler 1971).
In its most literal or narrowest extent, the indigenous term for rela-
tives, atsi atsitsi, refers to full siblings born of the same mother and
father: atsi or "their senior sibling(s)," and atsitsi or "their junior sib-
ling^)." 3 Because all full members of a sibling set acquire their respective
104 Quadripartite structures
blood(s) from the same mother and father at conception, senior and junior
siblings are one and the same blood.
Half-siblings ("one father, different mothers," or vice versa) of both
sexes are also terminologically senior or junior to one another. Here,
"seniority" (fakania) and "juniority" (eke) are assigned, not on chronologi-
cal birth order, but on the chronological sequence by which the parents
not in common married the parent in common of the two sets of offspring.
For example, all children of a mans second wife are junior in relation to
their siblings born of the same man's first wife, even if (as often happens
in polygynous unions) the former are born before the latter. Bilateral first
cousins are also terminologically senior and junior siblings, but here se-
niority and juniority are assigned according to the relationship of their
respective parents, who of course are siblings to one another. An elder
sister's children, for example, are all senior siblings of her junior brother's
children. It is in these ways that senior and junior sibling terms are
extended and assigned to persons of the same generation who are still
more distantly related. Second cousins are distinguished as senior and
junior according to the seniority and juniority of their grandparents who
are siblings, and so on.
Although all siblings as relatives irrespective of genealogical distance
are, like full siblings, considered one and the same blood in many con-
texts, villagers do recognize differences among them that correspond to
the distinction of blood/nonblood. Full siblings are unambiguously one
and the same blood. But when I questioned informants regarding other
types of siblings, they readily conceded that, whereas half-siblings, for
example, are one blood by virtue of sharing a parent in common, they are
different or "half" (apie) blood by virtue of not sharing the other parent in
common. Bilateral first cousins similarly share "some" (isa) blood, but
they also possess more different blood than full or half-siblings. Corre-
sponding discriminations in the quantity of blood shared by more remote
siblings are made proportionate to genealogical distance. As I shall show
below, certain of these discriminations bear directly on the regulation of
marriage.
As mentioned above, intergenerational relationships are classified in
terms suggestive of the classic "Hawaiian" system (Figures 6.1 and 6.2).
The terminology employed between Ego's own and adjacent generations
is not reciprocal. Father, male collaterals, and the spouses of female
collaterals in the first ascending generation are all "father" (ama). In the
same generation, mother, female collaterals, and the spouses of male
collaterals are "mother" (ina). In the first descending generation, own
children along with the children of all siblings are "children" (ngau) un-
differentiated for sex. For all remaining intergenerational relationships,
lineal as well as collateral, terms are reciprocal and undifferentiated for
A = ama father
I = ina mother
a = a senior
sibling
t = atsijunior
sibling ngotsingotsi

Figure 6.1. Atsi atsitsi terminology (patrilateral).

A = ama father
1 = ina mother
a - a senior
sibling
t = atsi junior
sibling ngots ingot si

Figure 6.2. Atsi atsitsi terminology (matrilateral).

 ngaua

ipa

ngotsingotsi

Figure 6.3. Ipa ngaua terminology (matrilateral).

 ngaua

ngotsingotsi

Figure 6.4. Ipa ngaua terminology (patrilateral).

 Kin, clan, and connubium 109
sex: "grandparent"/"grandchild" (ufu), "great grandparent"/"great grand-
child" (apapa), and "great great grandparent"/"great great grandchild"
(ngotsingotsi).
These several categories of intergenerational kin relationship also pre-
sume proportionate sharing of blood through bilateral extension along
consecutive parent-child links.
Although personal pedigrees articulated according to this termi-
nological system are usually relatively shallow, most often extending no
further than two or three generations back, it is presumed by everyone
that, after many generations of tribal endogamy, all members of the tribe
are relatives (atsi atsitsi) of one or another category to each other. The fact
that many people do not know their precise genealogical relationships is
simply explained by the presumption that the more remote ancestral
connections have been "forgotten."4 Tribal endogamy then is viewed to
place an outer limit or boundary upon the extension of bilateral-kin termi-
nology, and it gives to the tribal unit the integrity of one or the same
blood.
Within the tribe, however, there exist other boundaries that separate
blood from nonblood relationships. Although these boundaries corre-
spond in part with the recognition of differences in genealogical prox-
imity, they are expressed in the culture in the quite distinct terms of
affinal relationship, intraclan and interclan relationship, marriage pro-
scription and prescription, and an alternate classificatory sibling termi-
nology.
Affines (ipa ngaua)
Terminological affinal relationships (ipa ngaua) exist between persons of
only the same or adjacent generations. According to the rules that regu-
late marriage (see next section below), affines should not be relatives or
share blood of cognation. The reciprocal term for "husband" and for
"wife" is akaua. All the spouse's siblings are termed ipa or "siblings-in-
law," and reciprocally all siblings' spouses are ipa siblings-in-law. The
term ngaua is used with reference and address to "spouse's parent,"
reciprocally to "child's spouse," and between the parents of a married
couple (Figures 6.3, 6.4, 6.5).
In generations ±2, ±3 and ±4, however, Ego and Ego's spouse's
relatives employ the reciprocal kin terms appropriate for generation that
Ego uses for his/her own relatives of these generations. For the second
generation descending and below, Ego's descendants and their spouses
are both given the same reciprocal kin terms in relation to Ego; thus, here
the categories "own relatives" and "spouse's relatives" converge. It is
particularly important to note further that with respect to the second
ascending generation and above, Ego and Ego's spouse stand in identical
 $9
u I u u I u

_9$9£3

N EGO

K6b_K
N I n

u u u u

N = ngaua
n = rnjet/
u = ufu

Figure 6.5. Ipa ngaua terminology (Ego's and descending generations).

 Kin, clan, and connubium 111

terminological kin relations with the same persons. Classificatory rela-

tives of Ego again coincide with those of Egos spouse. Because these
remote genealogical affines are terminologically classified as relatives
rather than as affines, it is categorically implicit that they share some
degree of cognatically transmitted blood. By the same token, moreover,
because Ego's spouse and affines in the same and adjacent generations are
descended themselves from these cognates of Ego's ancestors, some of
Ego's erstwhile relatives are terminologically classified as his/her affines
or nonrelatives who share none of Ego's blood. The intersection of affinal
with kin terms and categories, in other words, suggests that certain cog-
nates who share blood become affines who share no blood. Thus, the
cultural relations between blood and nonblood, on the one hand, and
cognation and affinity, on the other, are reversed:
(c) blood : nonblood :: cognation : affinity
and
(d) blood : nonblood :: affinity : cognation
Comparison of these relations suggests that the blood/nonblood dis-
tinction is bisected by, rather than strictly parallel to, the distinction
between cognatic relationship and affinity. Such a structure is perfectly
compatible with a system of restricted exchange whereby marriage is
symmetrically prescribed or preferred between relatives of particular cat-
egories. I shall deal at length with this point below, for it is critical to the
full characterization of the Bush Mekeo system of exogamic reciprocity of
blood between agnatic groups.

Moiety, clan, and lineage

The categories blood and nonblood, and cognate and affine, are related to
the indigenous notion of unilineal clanship as it is conceptualized at a
number of levels of social organization. The word for "clan" is ikupu.
Ikupu is also the nominative inflection of the verb "to close" (ekupu)
discussed in the preceding chapter with reference to the ritual processes
of male tightening and female postpartum ritual. Villagers say of clans
what they also say of bodies during these male and female closing rituals:
"Nothing goes in, nothing comes out, nothing goes through. ' Part of what
is meant in this context is that a clan of whatever organizational level has
an inside (aonga) and outside (afanga). Every person in Bush Mekeo
society is considered to be inside the clan of his/her father, but outside
his/her mother's and spouse's clans (with significant qualifications; see
below). In terms of recruitment, then, Bush Mekeo clans are patrifilial
(Barnes 1962).
Because all persons inside the same clan are ideally considered to be
112 Quadripartite structures
relatives (atsi atsitsi), they are terminologically classified according to the
appropriate categories described above. Unlike certain cognates of differ-
ent clans, however, persons of the same clan cannot use with one another
the alternate sibling term (ekefaka) that is described later in the chapter.
Clanspeople correspondingly are deemed to be one and the same blood,
but through exclusively male rather than cognatic connections. In the
sense of expressing group unity and solidarity, Bush Mekeo clans are
therefore patrilineal (A. Strathern 1972).
Villagers apply the notion of a clan to several distinct levels of segmen-
tary (Evans-Pritchard 1940; Fortes 1953) or fragmentary (Groves 1963)
organization and closure. The Foikale myth (Chapter 5) expresses clan-
ship at its widest or most comprehensive range - the exogamous "moiety"
(ngopu). Knowledgeable old men can remember hearing from their fa-
thers that the tribe in the precontact past was bisected into moieties, but
none of them can now recall the names (if any) by which they were
known. This form of moiety organization seems also to have been charac-
teristic of neighboring societies that are culturally related to the Bush
Mekeo. Each of the two tribes of the Central Mekeo - the Biofa and Veke
- is composed of two named moieties (ngopu) (Seligmann 1910:320-7;
Guis 1936:39-40). According to Hau'ofa, moiety organization among the
Biofa tribe has since been abandoned, but among the Veke it is still
observed (1981:27-30). Levi-Strauss has pointed out a similar dual orga-
nization among the Motu and Koita villages of the Port Moresby area
traditionally (1969b:76-8; cf. Seligmann 1910:46-58). From Seligmann's
material, it could be argued that at least some of the Roro tribes along the
coast of Hall Sound on the Papuan Coast traditionally had moieties (Selig-
mann 1910:196-204). The data, however, are not altogether complete
(see the discussion of kofuapie and pisaua relationships later in the
chapter).
Because members of the same Bush Mekeo moiety are agnates with
one another, regarded as "one clan" (ikupu alakangamo), and one and the
same agnatic blood, members of different moieties of the same tribe are of
"different clans" (ikupu ikoina) and different agnatic blood. In terms of
the blood/bloodless distinction:
(e) blood : nonblood :: own clan : different
clan :: agnation : nonagnation
This suggests, in turn, that the own clan /different clan and agna-
tion/nonagnation distinctions are related to the own tribe/enemy tribe
and cognation/affinity distinctions discussed above:
(f) own clan : different clan :: own tribe : enemy tribe
(g) agnation : nonagnation :: cognation : affinity
(h) agnation : nonagnation :: affinity : cognation
 Kin, clan, and connubium 113
These categories, moreover, are related in similar ways to the distinc-
tions between intratribal peace and intertribal war, on the one hand, and
exogamy and endogamy on the other. According to the rule of moiety
exogamy, two moieties of the same tribe at peace reciprocate blood
through the exchange of women in marriage. The reciprocation of blood
through warfare and killing between moieties of the same tribe, however,
is prohibited. These latter relations, therefore, are precisely the opposite
of relations between different tribes who cannot intermarry but who are
obliged to fight. Exogamy and endogamy, in other words, are analogous
with peace and war:
(i) exogamy : peace :: endogamy : war
I shall have additional comments to make regarding these categories once
I have finished tracing out the boundaries of the indigenous notion of
clanship.
Every moiety in Bush Mekeo society is comprised of numerous patri-
lineal subunits with varying ranges of social inclusiveness, yet all are
uniformly termed ikupu. First, each moiety contains more than one "dis-
persed clan" distinguished by name. By "dispersed" I mean that all adult-
male members do not necessarily or by rule live in the same village. And
because postmarital residence is viripatrilocal, women of a dispersed clan
who have married out do not reside with their clanspeople unless they
have married into another clan of the same village. All members of a
dispersed clan claim to be one clan and one and the same blood ex-
clusively traced along agnatic connections, even if (as is very often the
case) the precise genealogical links have been forgotten. Also, the mem-
bers of a dispersed clan claim to be more closely related as agnates than
they are to the members of another dispersed clan in the same moiety.
In the tribe where I conducted most of my inquiries, there are four
named dispersed clans. Below I shall argue that this number of dispersed
clans in the tribe is not coincidental but contains important social-struc-
tural significance.
Characteristically, each named dispersed clan contains more than one
"subclan," likewise termed ikupu, distinguished one from another in
terms of senior (fakania) or junior (eke) status (cf. Sahlins 1963). In most
cases, senior and junior subclans of the same dispersed clan are said to
have once been a single subclan without differentiation as to seniority,
but through the process of fission and the public installation of a junior
peace chief - ideally, a "younger brother" of the senior peace chief - the
semiindependent status of the junior subclan has been established (Selig-
mann 1910:342-8; Hau'ofa 1981:184-214). In some cases, the events sur-
rounding the separation are still recalled, evidently occurring within the
lifetimes of surviving villagers. By similar means, incidentally, dispersed
114 Quadripartite structures
clans of the same moiety are said to have first become independent. The
details of these occurrences, however, have usually been lost from the
recollections of old people. Also, the relative seniority of agnatically relat-
ed dispersed clans, as contrasted by that of subclans, seems to be neither
asserted nor deemed socially significant.
Each subclan, senior or junior, of any dispersed clan may be further
subdivided into patrilineal units that are again termed ikupu, but which I
shall call "lineages." Ideally, each subclan possesses at least one lineage of
all four possible types: "peace chief" (lopia), "peace sorcery" (ungaunga),
"war chief" (iso), and "war sorcery" (faika). Typically, the members of a
single lineage can recite specific genealogical connections that unite them
all together as agnates. Each lineage possesses its own hereditary office
ideally passed from father to eldest son in formal public installation cere-
monies of the appropriate type. The officeholder of a lineage at any one
time, then, should be the most senior male of the lineage. Moreover, the
specialized lineages of a subclan are themselves related to one another in
terms of relative seniority. The peace chief and members of his lineage
are senior to the other officeholders and members of their respective
lineages. Apparently, relative status between the junior lineages corre-
sponds only with actual or presumed genealogical seniority.
Details regarding the ritual specializations of the four subclan and lin-
eage officeholders among the Bush Mekeo are nearly identical to those of
the Central Mekeo as described at length by Seligmann (1910:279-301,
342-8; cf. Hau'ofa 1971, 1981:215-88; Belshaw 1951:3; Stephen 1974:6).
Very broadly, peace chiefs and peace sorcerers dominate relations within
the tribe, and war chiefs and war sorcerers direct relations between
tribes. More specifically, the main responsibility, indeed the prerogative,
of the peace chief is to give and receive death-feast prestations on behalf
of his subclan. The peace sorcerer assists the peace chief by "closing the
eyes" (i.e., killing) of persons within the tribe who have shamed or an-
gered the chief and thereby prevented him from properly fulfilling his
feasting duties. Peace sorcerers will supposedly even attack a peace chief
of another clan if the latter publicly shames or angers the sorcerers own
chief. Also, as Hau'ofa has emphasized, the peace sorcerer is the custo-
dian of traditions (1971:162).
War sorcerers perform the ritual, on behalf of the warriors of the clans
and subclans of their tribe, that weakens their enemies before they are
attacked. Also, as the peace sorcerer is the custodian of traditions with
respect to peace affairs, the war sorcerer is the guardian of war traditions.
War chiefs lead the attacking forces of their subclan into battle and
direct the actual killing.
The four hereditary ritual specializations that distinguish the lineages of
a subclan constitute a cultural and social totality. The clan group, which
 Kin, clan, and connubium 115
includes specialists of all four kinds, is totally balanced, that is, it can
participate in all contexts of intratribal and intertribal, or peace and war,
communication.
This official hereditary ritual specialization involves the bisecting of the
peace/war duality along the dimension of chief/sorcerer. The resulting
four types of officeholders and lineages are homologous with the catego-
ries resulting from the bisection of the inside/outside duality discussed in
previous chapters. Peace chiefs and peace sorcerers are both "inside"
insofar as their sphere of influence is restricted to their own bounded
tribe. The relations "outside" the tribe are the special province of war
chiefs and war sorcerers. Peace chiefs never kill within the tribe. Rather,
they secretly appoint a peace sorcerer of their tribe to kill for them. If the
intended victim is in the chief's own clan or moiety, he will most likely
have to find a peace sorcerer of the opposite moiety to do the killing, as
his own moiety sorcerers will be less inclined to kill their own agnate. If
the intended victim is in the moiety opposite to the peace chief he has
shamed or angered, the peace chief will supposedly appoint a peace
sorcerer of his own moiety to do the killing. In either case, the peace
sorcerer who does do the killing will typically be from the moiety opposite
to that of the victim.
In the context of war outside the tribe, these relations between chief
and sorcerer and killing and not killing are reversed (Seligmann
1910:295-8). War sorcerers, unlike peace sercerers, do not kill their
enemies. In fact, as soon as they finish pronouncing their spells in the
bush at the edge of the enemy village, war sorcerers fall to the ground as if
slain themselves. On the other hand, war chiefs in the lead of the attack-
ing forces will direct all the actual killing and participate in it themselves,
as distinct from peace chiefs. Although war sorcerers do not directly
engage in battle, it is they who specify the plans for killing to the war
chiefs, just as peace chiefs direct peace sorcerers to kill. Thus, the two
war specialists reverse the relations of the peace specialists:
inside : outside :: inverted outside : everted inside (4)
peace chief : war sorcerer :: war chief : peace sorcerer (15)

Clans, residence, and working together

The description I have just presented of lineage and office differentiations
of a subclan is an idealized conceptual ordering that does not represent
the actual constituency of even a single actual subclan of the tribe I lived
among and studied most intensively. Every identifiable Amoamo subclan
unit lacks at least one of the supposedly necessary ritual specializations.
Nonetheless, these conceptual distinctions and relations that define the
fourfold division of the subclan are persistently represented in the local-
116 Quadripartite structures
level organization of "residential clans" or simply clans (ikupu) (cf. Groves
1963). The conceptualization of an independent, fully staffed, and ritually
complete patrilineal clan overlaps with co-residence. Together, these
ideas correspond with the relationship termed "working together" (kepi-
naunga kaitsialo) (cf. Hau'ofa 1971:155; A. Strathern 1972:217-18).
Typically as well as according to rule, each Bush Mekeo village is
occupied by two residential clans, each supposedly of opposite moieties.
The kinds of social units that make up a local clan vary in some respects
from village to village. A local clan may include, for example, two sub-
clans sharing the same clan name but distinguished as senior and junior
with their own respective peace chiefs. This is the ideal situation. Some-
times, however, senior and junior subclans reside in different villages. In
a few instances, ritually specialized lineages of the same subclan live apart
from one another. It is not uncommon to find that some members of a
single lineage also reside in different villages. These deviations are evi-
dently the result of clan, subclan, or lineage fission, migration, relocation,
and, as we shall see, fusion. Usually villagers can provide stories that
describe the supposed historical events that have led up to the current
seemingly jumbled-up situation.
Consequently, the composition of local clan groups in terms of adult
married male personnel is characteristically heterogeneous.
The local clans that result following numerous immigrations and emi-
grations typically consist of several heterogeneous units bound together
as one clan through working together. Two considerations are commonly
cited as relevant to the formation of this relationship. One has specifically
to do with the requirement that each local clan have its own peace chief.
The other concerns the necessity that each local clan be represented by a
lineage of each of the other three ritual specializations.
First, when a lineage of peace chiefs dies out, surviving members of the
subclan are left unable to participate in the system of death-feast re-
ciprocation. This is a most untenable situation for them. After one of their
number dies, the survivors have no means of ending the restrictions of
their mourning, because only an installed peace chief of their clan can
give or receive mortuary-feast prestations (Chapters 7 and 8). Thus, a
subclan without its peace chief must work together as one clan with
another subclan whose peace chieftainship is still intact and functional.
Usually, the two units will live together in the same village. Moreover, it
is typically the case that the two subclans will not claim an agnatic ances-
try in genealogical terms. Numerous local clans among the Bush Mekeo
are, in fact, composed of such heterogeneous units who cooperate or work
together explicitly for feasting purposes.5 The groups that have lost their
own peace chieftainships, however, will frequently assume the clan
names of the groups they join who still possess them.
 Kin, clan, and connubium 117
Secondly, a number of local clans incorporate subclans that work to-
gether as one clan even while the constituent subclan units still possess
their own installed peace chiefs. In these cases, one or both of the fused
subclans lacks one or more of the other three ritually specialized authority
positions. For example, one subclan may not have its own war chief and
war-chief lineage, whereas another subclan, living in the same village or
elsewhere, which has no known agnates, may have a war chief but lack a
peace sorcerer. The two subclans might then settle in the same village
and work together as one clan with the ideal and total complement of
ritual and lineage specializations. Sometimes the two subclans will retain
their previous clan names, other times they will both go by the clan name
of only one of them.
One means of establishing a ritually complete clan at the local level is
simply to recruit a knowledgeable specialist of the missing type from
outside the clan to found a new lineage within the clan. Sometimes spe-
cialists recruited in this way come from another subclan, local clan, or
dispersed clan of the same moiety, particularly as regards peace chiefs.
However, specialists of the other categories are in most instances initially
recruited after first having married in. Not only do relationships of work-
ing together bind heterogeneous subclans and lineages at the local level,
very often they are predicated by affinity and cognation rather than by
agnation.
These processes of subclan and lineage fusion and the recruitment of
nonagnates and even affines describe in general terms how subclans and
residential clans, by working together at the local level, preserve their
ideal lineage structure with a complete fourfold complement of ritual
specialization among their personnel. However, local clans consolidated
in these ways might seem to contradict the fundamental idea of the purely
agnatic closed social unit. But regardless of how they are initially related,
once lineages and subclans begin working together, they are supposed to
cease intermarrying among themselves. Different nonagnatic bloods
thereby become one agnatic blood, and persons of different clans become
one and the same clan. Because the local clan continues to marry ex-
ogamously, it remains closed vis-a-vis other clans.
To illustrate the persistence of ideal clan structure as it is manifested at
the level of local-clan organization, the ritual specializations of the lin-
eages composing Nganga clan of Akabe village6 as it existed in 1976 are
portrayed in Table 2. The abbreviations NM3, NF2, etc., refer to sepa-
rate lineages of Nganga Manga, Nganga Angai, and Nganga Faka subclans
(see Figure 6.6 and following text). No one of these three component
subclans possesses all four types of lineage speciality. Wherein publicly
espoused ties of agnation uniting these subclans in this case are not de-
monstrable and in many instances are known to be otherwise, the case of
118 Quadripartite structures

Table 2. Offices, lineages, and subclans of Nganga residential clan

Iopia Ungaunga Iso Faika

(peace chiefs) (peace sorcerers) (war chiefs) (war sorcerers)

NM3, NF2, NM5, NF8 NM3, NF6 NA3, NA4

NF7, NA1,
NA3

Subclans: NM = Nganga Manga; NF = Nganga Faka; NA = Nganga Angai.

Nganga clan is typical of other local clans in the Bush Mekeo. Notwith-
standing, each local clan operates as though it were one agnatic blood.
Despite all the vicissitudes of traditional warfare; pacification; coloni-
alism; rampant epidemics and depopulation; migrations in and out; and
intraclan disputes over endogamous marriages, sorcery suspicions, and
official successions, Nganga clan has successfully achieved and maintained
a fully functioning approximate ordering of the ideal fourfold balance of
offices and lineage specialization through the processes of subclan and
lineage fusion in terms of working together.

The oral history of Nganga clan

This fourfold balance is one dimension of the strictness Levi-Strauss men-
tions regarding Mekeo social organization, and an example of what he
more generally discusses elsewhere as the primacy of structure over
event (1969b:77, 1966a:66-74; see also Sahlins 1976). As such, and be-
cause I will discuss this specific theoretical issue at length in Chapter 10
in the light of my conclusions, it will be instructive to digress briefly and
outline here the composite oral history of the internal relations of Nganga
clan. This account is based upon interviews I conducted with numerous
villagers, both in and out of Nganga clan, and upon certain critical genea-
logical details recorded by Father Egidi (n.d.) near the turn of the cen-
tury. The temporal relations of the lineages and subclans discussed herein
are diagramed in Figure 6.6.
At the time of contact (1890, or thereabouts), Nganga clan was com-
posed of Nganga Manga and Nganga Faka subclans working together.
Villagers today claim, however, that several generations before Nganga
Manga had joined with Nganga Faka, Nganga Manga was itself composed
of two agnatically unrelated subclans - Nganga Manga (NM) and Nganga
Angai (NA) - who had begun working together as one clan.
NA is said to have first migrated into Bush Mekeo territory following
the legendary breakup of Afai village (Seligmann 1910:315-16). When
they arrived in the Bush Mekeo, NA included the lineages of senior and
junior peace chiefs, senior war sorcerer, and senior war chief. They
 NGANGA MANGA N G A N G A FAKA NGANGA ANGAI
SUB-CLAN SUB-CLAN SUB-CLAN

1 f 2 i\ 5 1 2 -i I > 6 7 ) 1) 11 12 13 14 15 16 17 1 > • \ «j 3 1 II

> <
c > < >
X X X X X X
(1890)

+
(Amo)
X
X
N f e (Veke)

(1976)
Lf Le If le Uf Lf Lf Ube le le le Ubf Ff/Uf ? F F Le Lf Le Ubf Ff Fe le ? ?

Figure 6.6. Composite lineage history (Nganga clan). L = lopia peace chief; / = iso war chief;
U = unguanga peace sorcerer; F = faika war sorcerer; Ub = uibina assistant peace chief; x =
lineage extinct; * = emigrant lineage; + = immigrant lineage; f = fakania senior; e = eke
junior.
120 Quadripartite structures
lacked a peace sorcerer and peace sorcerer's lineage. Soon afterward,
apparently, the war chief lineage died out.
One of the lineages of the newcomers (NA2) exchanged sisters with a
junior lineage of Nganga Manga (NM2) who was already living in the area.
At some point, all of NA settled with NM. The two began working to-
gether as one clan and ceased intermarrying thereafter. Although this
arrangement permitted NA and NM to approximate more closely the
ideal fourfold complement of offices and lineages, the new Nganga clan
still lacked a peace sorcerer.
Subsequently, NM subclan lost both its senior and junior peace chief
lineages. The last male senior peace chief (NM1) had only daughters and
no sons. Whenever a peace chief lacks sons, he may pass the office to his
daughter, but she may not in turn pass it to her own son, for it must stay
inside the clan. In this instance, the last male senior peace chief of NM
installed his eldest daughter, and she her younger sister. Upon the death
of the latter, the senior peace chieftainship went unoccupied.
The junior peace chief of Nganga Manga (NM2) eloped with a woman of
Nganga Angai, violating the rules of local-clan exogamy and incest, reset-
tled with an unrelated clan of a neighboring village, and became that
clan's junior chief.7 For several generations thereafter, Nganga Manga
subclan had no peace chiefs of its own, but it participated in feasting
exchanges through Nganga Angai's peace chieftainships.
Some generations previous to European contact, a large clan of
Lapeka8 origin, now known as Nganga Faka (NF), had migrated to the
Bush Mekeo, intermarried heavily within the tribe, and established itself
as an independent local clan. Among its original settlers, Nganga Faka
included lineages of senior (NF1) and junior (NF12) peace chiefs, five
lineages of war sorcerers (NF2, NF7, NF8, NF10, NF11), two lineages of
junior war chiefs (NF4, NF5), and seven additional lineages now uniden-
tified as to hereditary ritual specialization. A group of Kunimaipa (non-
Austronesian-speaking) mountain people happened upon the Nganga
Faka village and attacked, mistaking them for a different enemy group. In
this battle and the first epidemics of European disease to sweep through
the area soon thereafter, twelve of Nganga Faka's sixteen recorded lin-
eages were exterminated. The survivors included members of three
Nganga Faka war sorcerer lineages (NF2, NF7, NF8) and one war chief
lineage (NF5). Lacking a peace chief to make feasts for their own dead,
lacking also a peace sorcerer, and in need of allies to help them avenge
the many deaths, the NF survivors relocated near Nganga Manga's newly
established village, Akabe, and soon became one clan, working together
with NM and NA.
As a result of a dispute over clan endogamy in another village, two
brothers of the war chief lineage of a clan there left and settled with the
 Kin, clan, and connubium 121
Nganga Fakas. One of the brothers was already married to a NF woman.
In the years sometime after European contact when the surviving Nganga
Faka war chief lineage (NF5) produced daughters only and no sons, the
descendants of these two brothers became the war chief lineage of
Nganga Faka clan (NF6). Since then, some members of this war chief
lineage have intermarried with people of their clan of origin.
Up until pacification, Nganga local clan seems to have concentrated
especially on recruiting lineages and individuals with hereditary war spe-
cializations. But after the colonial government put an end to war, the
strategic emphasis shifted toward peace sorcery as a means to avenge
outstanding war debts and the many postcontact deaths thought caused
by peace sorcery, as wave after wave of foreign disease swept through the
area (Mosko n.d.). Until this time, Nganga clan had no peace sorcerers of
its own. In a series of complicated maneuvers over the next sixty years or
so, the Ngangas have attempted to fill this lacuna. I mentioned earlier the
case of Nganga Manga losing its junior peace chief (NM2) to another clan
of the tribe. This other clan had two peace sorcery lineages, one senior
and one junior. A bachelor of the junior peace sorcery lineage tried
unsuccessfully to elope with the wife of a man of the senior peace sorcery
lineage of his own clan. Although the adulterer and his elder brother were
themselves competent sorcerers, they could not remain in the same vil-
lage with the rest of their clan. Both men moved to live at Akabe village
with the elder married brothers wife's people, Nganga Faka. Later, when
the elder brother was formally installed as an official peace sorcerer, the
three Nganga subclans (NM, NF, NA) sponsored the event staged at
Akabe. Finally, Nganga clan had its own peace sorcerer and peace sorcery
lineage. Although uninstalled, the younger brother was also competent to
practice peace sorcery. And so for a time Nganga clan had at least one
representative for each of the four offices, and all four categories of spe-
cialized lineage. Although the two brothers and their families lived among
the Ngangas, they were known as the subclan Nganga Katsina,9 and they
worked together with the rest of Nganga local clan.
Unfortunately for the Ngangas, however, their new peace sorcerers
died before they could turn their hereditary secrets over to their sons.
Several years later, the sons' mothers also died, and they returned to the
natal clan and village of their fathers. I happened to be working in the
Bush Mekeo when the last of the sons departed.
Nganga clan was not left without peace sorcerers for long, however.
Over the last several years, the Ngangas had acquired two other peace-
sorcery lineages. The last possessors of war sorcery in NF8 were killed in
battle shortly before pacification, and their hereditary secrets and knowl-
edge died with them. But some of their descendants survived. Among
them, one man, who is now quite old, is, in the opinion of people of
122 Quadripartite structures
neighboring villages, a "new peace sorcerer" (ungaunga mamanga). Al-
though he publicly denies that he is a peace sorcerer, he certainly plays
the role. Periodically, he goes "on patrol," and it is rumored he visits
practicing peace sorcerers all over the Mekeo region. It has been widely
circulated that a few peace sorcerers from the Biofa and Veke tribes of the
Central Mekeo have unsuccessfully tried to kill him and that he has
allegedly shot more than one of them with his shotgun. 10 These are
indications to villagers that he too is a peace sorcerer. The people of
Akabe village, however, know that he is not a "true" or "real" peace
sorcerer. They suspect that he only knows the use of poisons. In any case,
this old man has created a measure of renown in the area as a peace
sorcerer. Nganga local clan has seemingly generated a new peace-sorcery
lineage from within.
Yet another peace-sorcery lineage (NM5) is in the process of establish-
ing itself in Nganga local clan, this time from without, through intermar-
riage. An old peace sorcerer and uninstalled heir to a peace chieftainship
from a Central Mekeo village had taken a wife from one of the other Bush
Mekeo clans in the early 1920s. When other peace sorcerers of his own
clan coveted his claim to the peace chieftainship and supposedly killed
several of his young children, he moved to live with his wife's people
among the Bush Mekeo. Later, his eldest daughter married the war chief
of Nganga Manga (NM3), and her father eventually moved to Akabe
village to live with her. The old mans only surviving son left the village
for several years to work in the Papuan Gulf on oil-exploration crews.
When the son returned to the Bush Mekeo, he stayed at Akabe village
with his father and sister, then married a woman of NA and settled with
Nganga clan. It is said by some villagers that the old man passed on much
of his sorcery knowledge and paraphernalia to his son before he died.
Through his Bush Mekeo mother, the son is a categorical junior sibling to
three of the adult men of Nganga Angai subclan, and sister's daughter's
husband of the senior male of NA3 through his second wife. Moreover,
Nganga clan paid the marriage compensation for this second marriage.
Thus, the son of the old peace sorcerer has numerous and distinct ties to
both Nganga Manga and Nganga Angai subclans of Nganga local clan. He
has come to identify himself and his children with Nganga Manga (NM5),
and they in turn have been so identified by the rest of the village. Most
significantly, this man consistently and energetically participates in mor-
tuary-feasting exchanges as a Nganga clansman. Now that he is reaching
the peak of his active maturity, the rest of Nganga clan depends more and
more on him to perform the services of resident peace sorcerer - thwart-
ing the attacks of foreign peace sorcerers and "rascals," curing and divin-
ing, providing counsel to peace chiefs in the minutiae of feast perfor-
 Kin, clan, and connubium 123
mance, and so on. Indeed, it was he who buried the python faifai that had
trespassed the village (see Chapter 2).
There have been a few other recent personnel modifications to fully
round out the representation of lineage specializations in Nganga clan.
The last official war sorcerers of NA4 died after leaving their knowledge in
trust with a man of NA3 to pass on to their sons once they became old
enough to understand. That the trustee has failed to do so for nearly thirty
years now and is suspected of passing this knowledge to his own sons
instead - thus founding a new war-sorcery lineage - lays at the root of
many troubles in Nganga Angai subclan and in Nganga clan generally.
Also, the extinct senior peace chieftainship of Nganga Manga (NM1) (see
above) has been revived in the war chief lineage of that subclan (NM3).
The son and heir to the war chieftainship received the subclan peace
chieftainship, and his father's younger brother assumed the title of war
chief, which has since passed to that man's eldest son.

Clanship and women's children

The Bush Mekeo employ the idea of clanship in yet another and some-
what paradoxical context where the connecting relatives are female rather
than male. Thus, in addition to full membership and entitlement in the
clan estate of his/her father, each Bush Mekeo villager retains certain
reciprocal auxiliary rights and obligations toward members of his/her
mothers clan and the children born of female members of his/her own
clan. In the terms of the indigenous culture, these relationships con-
stitute the category papie ngaunga or "woman's children. "11
When you ask villagers the name of their clan, almost invariably they
respond with the name of their own or father's clan. But if you also ask
them if they are members of a particular clan that happens to be that of
their mothers, they will consistently respond in the affirmative. In appar-
ent deviation from the rules of patrifilial and patrilineal clanship, a wom-
an's children are papie ngaunga but nonetheless regarded as one clan
with her supposed agnates and clanspeople.
The term papie ngaunga is used reciprocally between agnatic members
of a clan and children born of the clan's women. Villagers' articulations of
the rights and obligations inhering in relationships of papie ngaunga are
very similar to those between presumed agnates and members of the
same clan, except for the transmission of secret male ritual knowledge,
succession to subclan and lineage office, and rights to certain kinds of
property (land, dancing ornaments and insignia, etc.). In this and the next
two chapters, I shall discuss the more important and significant contexts
of papie ngaunga interrelationship vis-a-vis agnation: marriage regulation
and compensation, and the reciprocation of mortuary feasts.
124 Quadripartite structures
The history of Nganga clan recorded above illustrates how subclan and
lineage fusion and the recruitment of nonagnates are typically initiated
along lines of affinal connection. Marriage and affinity, of course, generate
papie ngaunga relationships. Villagers claim that usually in cases of non-
agnate recruitment, men do not join or "go inside" the clans of their
wives; rather, their children become new full members by papie ngaunga
through their mothers. In any case, a man signals his intention of joining
his wife's, mother's, or any other clan in his own lifetime by participating
in that clan's marriage and death-feast exchanges and not those of his natal
clan (see below in this chapter and Chapter 7). I shall argue that cases of
recruitment and subclan fusion among nonagnates along papie ngaunga
lines are common because papie ngaunga are already conceptualized in
the culture as one clan and one blood but traced through females.
This, of course, seriously qualifies the notion of a clan at every level of
fragmentary and segmentary organization as closed. Ideally, a clan is
closed by virtue of shared descent through males exclusively. But a clan is
also simultaneously open by tracing papie ngaunga auxiliary clanship and
shared blood through women; that is, the ideally closed clan requires the
intrusion and extrusion of blood from and to other closed units, making it
thereby open as well. By means of heterosexual reproduction, the bloods
of clans of the tribe flow into and out of each other. Women of one's own
clan give birth to agnatic members of other clans, and women of other
clans give birth to all members of one's own clan. Thus, clanspeople
similarly trace agnatic blood only through their fathers, and, excepting
full siblings, trace nonagnatic or cognatic blood differently through their
mothers.
Clanship in terms of papie ngaunga, however, is not traced as far
genealogically as clanship in terms of agnation. Ideally, the notion of one
clan is traced through father to incorporate all members of father's
moiety. Papie ngaunga traced through mother, on the other hand, is
restricted to the membership of her dispersed clan and not to the rest of
her moiety. Reciprocally, children of women of a dispersed clan are papie
ngaunga, but not women's children's children.

An alternate sibling term - ekefaka

All persons who regard themselves as one clan whether by known or
presumed agnation, by co-residency and working together, or by papie
ngaunga relationship also regard themselves as kin or relatives and ac-
cordingly may use the appropriate kin terms in addressing and referring
to one another as described above. It is curious, however, that classi-
ficatory siblings who consider themselves one clan may not use a certain
alternative sibling term, ekefaka, which other genealogically known rela-
tives of the same generation, who are not one clan, may use. Inasmuch as
 Kin, clan, and connubium 125

I shall in the following section define the term ekefaka when used be-
tween persons of opposite sex as "potential spouse," it will be useful to
characterize it first in its cognatic applications (A. Strathern 1972:16-17).
The term ekafaka is said by some villagers to be a compound of the two
terms eke or junior and fakania or senior, perhaps implying equality
between siblings of this category as opposed to the inequality of siblings
who may not also be classified as ekefaka (see above; cf. Hau'ofa
1981:158). Only siblings who are neither members of the same patrilineal
clan or moiety nor papie ngaunga to one another may reciprocally use the
term ekefaka. Although all first cousins and the children of parallel cous-
ins are never ekefaka, the children of bilateral first cross-cousins who are
members of clans in opposite moieties may consider each other ekefaka.
The ekefaka sibling category, in other words, is consistent with a "Dravi-
dian'-type terminology.
The rules that bound and distinguish the ekefaka category from siblings
who must always use the senior and junior sibling terms (a-and atsi-,
respectively) suggest that Bush Mekeo considerations of clan, moiety, and
papie ngaunga override or contradict the notions of known genealogical
relationship and of one blood cognatically defined (cf. Lounsbury 1965;
Scheffler and Lounsbury 1971). A- and atsi- include many people of one
blood and the same generation among whom genealogical relationship is
at best presumed and often known to be otherwise. Members of the
category ekefaka, on the other hand, include persons who are cognatically
one blood according to known genealogical connection, but who are none-
theless tentatively excluded from that class. Viewed from the perspective
of the progression of generations through time, genealogically defined
nonrelatives along with descendants of certain genealogically known rela-
tives obligatorily remain one blood and siblings distinguished according
to seniority. Others, oftentimes tied closer genealogically by cognatic
blood, become ekefaka and potentially nonrelatives sharing no blood.

The rules of marriage

This paradox or contradiction involving the category ekefaka is directly
related to, and resolved in part by, the Bush Mekeo system of marriage
exchange.
Marriage exchanges are ordered according to positive and negative
rules that circumscribe a system structurally similar to the "Aranda"-type
(see Egidi 1912; Radcliffe-Brown 1930; Levi-Strauss 1969b; R. Fox
1967:195-9; Sahlins 1976:29). Two persons of opposite sex and opposite
moiety may marry if neither of them is of the same clan as the other's
mother, that is, if they are not papie ngaunga. The category of permissi-
ble spouses thus includes all persons of opposite gender who are or may
be considered ekefaka siblings to one another.
126 Quadripartite structures
Villagers themselves do not articulate their marriage preferences in
these terms, for marriage is also allowed between persons who are not
relatives to one another and not therefore ekefaka except as members of
the same endogamous tribe. Also, in contrast to "Aranda -type systems in
Australia, there are no named sections or subsections among the Bush
Mekeo. Rather, idealized rules of marriage are phrased in terms of a
series of rules - some proscriptive, some prescriptive - with considerable
redundancy among them. To demonstrate the overall character of the
system as well as its homology with other dimensions of Bush Mekeo
culture, it will be necessary to summarize each of these rules and com-
pare them to one another in terms of the indigenous categories. Before
proceeding, I should mention that the rules as expressed here and by the
Bush Mekeo apply without distinction as to the gender of Ego.

Marriage rule I: cognatic proscription

Perhaps the most frequently repeated regulation prohibits intermarriage
among bilateral kin and blood relatives: "Our relatives are one blood; we
human beings do not marry our relatives" (Aika atsika ifa alakangamo,
ika bapiau aika atsika aekamange). On a number of occasions, I remarked
to my acquaintances that many villagers are married to known relatives
and that it is generally understood that ultimately everyone of the tradi-
tionally endogamous tribe is a relative to everyone else. To this I was told
that marriage is prohibited between relatives with "one grandparent"
(ufu alakangamo) in common, i.e., between all siblings as far as first
cousins, and between parents, on the one hand, and between their own
children, children of their full and half-siblings, and children of spouses
full and half-siblings, on the other. Second cousins and other siblings of
the same generation beyond this genealogical distance of one grandparent,
however, unless otherwise excluded (see next section), are marriageable.

Marriage rule II: agnatic proscription

Overlapping in part this bilateral exclusion, villagers also expressly pro-
hibit marriage between persons who are agnatically one clan. Members of
the same named lineage, subclan, dispersed clan, and moiety, or who are
working together as members of the same local clan, may not marry.
According to this restriction, although cognatic blood is exchanged ex-
ogamously between opposite moieties and clans, presumably shared ag-
natic blood stays inside the clan. As an agnatic entity, the clan at all levels
is thus closed vis-a-vis parallel-clan units of the opposite moiety.

Marriage rule III: papie ngaunga proscription

Marriage is also prohibited among papie ngaunga who belong to different
moieties. Villagers say that when their clanswomen marry into other
 Kin, clan, and connubium 127
clans, the skin (fanga) of the clan "goes outside" (ebualai) (Chapter 2).
The children of these women become papie ngaunga to members of their
mothers' clan, and vice versa. Although papie ngaunga belong to clans of
opposite moieties agnatically defined, they are one clan and one blood
either through their mothers or through their clan sisters and clan
daughters. Thus, papie ngaunga are also prohibited from intermarrying
according to the rules of clan exogamy. Again, it should be emphasized
that papie ngaunga relationship in one generation is not transmitted to
the next, that is, neither to the children of the daughter of a clanswoman
nor to members of mother's mother's clan. These people, it is acknowl-
edged, have their own mothers and clanswomen's children, respectively,
relating them as papie ngaunga to people in still other clans. The rule
regarding papie ngaunga, then, excludes for members of each sibling set
a particular clan of the opposite moiety (i.e., their mothers) and the
collection of children born of all their clanswomen. Although each clan is
agnatically closed with respect to clans of the opposite moiety, it is cog-
natically open in relation to them by virtue of blood relationship through
women for the length of one generation.
Viewed simultaneously, marriage rules I, II, and III proscribe a range
of blood relatives that corresponds precisely in the category of siblings
who cannot use the alternate ekefaka term. All persons of the same gener-
ation who must use the asymmetrical a- or atsi- forms of address and
reference12 cannot marry. Siblings beyond this range who may employ
the ekefaka alternative reciprocally are thereby not prohibited and may
marry. To this extent, Bush Mekeo kinship nomenclature and the bound-
aries of clan and papie ngaunga categories are isomorphic with marriage
proscription.

Marriage rule IV: kofuapie prescription

In contrast to these forms, villagers also express their rules of marriage in
prescriptive terms: "We marry people of different blood, we marry our
kofuapie"13 (Ika ifa ikoina kamange, ika kofuapiemai kamange). It is on
this point that a minor controversy in the literature of the Mekeo area
generally has turned.
Virtually all observers of the Central Mekeo have reported that vil-
lagers claim they ideally or preferably marry their ufuapie, but there
seems to be confusion among the ethnographers as to just who the
ufuapie are. Moreover, it has been empirically discovered that villagers
marry ufuapie (however they are defined) only to a numerically limited
extent. This has caused trouble for some observers, and each has in-
terpreted the discrepancy somewhat differently.
Father Guis, the earliest ethnographer in the area, reports that ufuapie
are regarded as "part of our blood. " He infers also that ufuapie is an
128 Quadripartite structures
enduring relationship between neighboring villages, not among social
units within a single village. However, Guis also asserts that clans of a
single village form an "ufuapie group.' Excepting marriages by elope-
ment everyone must marry ufuapie, Guis says. Additionally, he observes
that ufuapie raise pigs and dogs for one another and exchange them at
death feasts (1936:39).
Based on his visit to the area at the turn of the century, Seligmann
(1910:349-58) challenges Guis on several points. He suggests that ufuapie
is a relationship between ikupu or "clans," or between what I have
termed dispersed clans. But he unnecessarily complicates the issue by
incorrectly using the indigenous term pangua for dispersed clan. Accord-
ing to all subsequent investigators of the Central Mekeo, the word pa-
ngua, like the cognate Bush Mekeo term paunga, means "village," and
each village is ideally composed of two agnatically unrelated clans (see
Chapter 2; Belshaw 1951:2; Stephen 1974:3; Hau'ofa 1981:30). As regards
marriage regulation, Seligmann reports that it is merely considered bet-
ter for ufuapie to marry one another; it is neither necessary nor obliga-
tory. Seligmann also notes that preferred ufuapie marriages do not nu-
merically predominate. However, his confusion over the meaning of the
term pangua, and thereby of ufuapie, would partly account for this dis-
crepancy as well as his argument with Guis. Moreover, although he con-
cedes that ufuapie reciprocate death-feast prestations, Seligmann denies
that ufuapie raise pigs and dogs for one another as Guis reports (Selig-
mann 1910:364).
Visiting the area nearly fifty years later, Belshaw agrees that tradi-
tionally only ufuapie were permissible spouses, but he argues against
Seligmann, claiming that the term ufuapie refers rather to "people who
are in close patrilineal relationship" than to special relationships between
patrilineal clans or villages (1951:2-4).
More recently, Stephen clarifies pangua to mean village rather than
any unit of an ikupu or clan, and she describes ufuapie as a ceremonial or
feasting relationship between one ikupu and one or more ikupu (1974:3,
22).
Lastly, Hauofa concurs with Stephen as regards the designation of a
pangua as village, but he goes on to differentiate ufuapie from another
interclan relationship known as auai or "soulmates" (Hauofa 1981:162-
3). Ufuapie, according to Hau'ofa, refers to special relationships between
co-residential clan groups for reciprocal political, ceremonial, and feasting
purposes, and only between groups who can marry. Auai, on the other
hand, is a personal relationship between individual adult men of localized
clans tied together by ufuapie group relationship. Over four generations
in Beipa'a village, observes Hauofa, between 30 and 37 percent of all
marriages (considering only patrilineal clan affiliations of grooms' and
 Kin, clan, and connubium 129
brides' fathers) are between ufuapie. But if the ufuapie oipapie ngaunga
are also included, then 67 percent of villagers marry their ufuapie.
Hau'ofa concludes that Biofa villagers in the present and in the past have
married ufuapie with greater frequency than other investigators have
suspected (Hau ofa 1971:157, 1981:160-83).
By a comparison with Bush Mekeo conceptualizations, I think it likely
that these several problems and inconsistencies in the literature of the
Central Mekeo have arisen, first, from a failure to characterize accurately
the notions ufuapie and auai, and second, from not appreciating the
relation of this distinction to the two institutionalized alternate means of
initiating marriage, that is, by "betrothal" (onge apua) as opposed to
"elopement" (kepiau). Briefly, among the Bush Mekeo traditionally, peo-
ple who are pisaua (cognate with Central Mekeo auai) to one another
must marry only by betrothal. Members of kofuapie related groups who
are not also pisaua are permitted to marry by elopement. In any case, in
each generation there should be at least one betrothed marriage intact
between subclans whose peace chiefs are bound as friends. My treatment
of these issues with respect to the Bush Mekeo should help illuminate
also the systems of marriage exchange among their neighbors, including
the Central Mekeo.
In the Foikale myth (Chapter 5), it will be remembered, the existence
of clanship, indeed the creation of Bush Mekeo society in its present
form, is coincident upon the reciprocal exchange of women between two
otherwise unrelated agnatic groups. Just as the notion ikupu or clan may
imply any of several levels of fragmental organization and segmental con-
ceptualization, the term kofuapie is used in reference to "other" or "op-
posite clan" categories and entities of correspondingly varying scale. At its
widest extent, kofuapie may refer to any and all clan units that are not
claimed to be agnatically related to ones own clan. In this sense, kofuapie
means "clan (at any level) of the opposite moiety," or the opposite moiety
as a unit. From the viewpoint of the members of each residential clan
seeking spouses, all others of the tribal unit are either one clan, one
agnatic blood, and unsuitable for marriage; or they are kofuapie other
clans, different agnatic blood, and permissible for marriage irrespective of
living in the same or different villages. Clearly, papie ngaunga are
kofuapie according to this definition, but they are an exceptional case, for
they are simultaneously one clan and one cognatic blood traced through
females.
When villagers refer to kofuapie in certain contexts, however, it is also
clear that they mean a particular clan of the opposite moiety, usually the
other clan that resides in the same village. Bush Mekeo villages, as I have
already mentioned, ideally contain two local clans, each comprising either
one subclan or more than one subclan working together with the com-
130 Quadripartite structures

plete assortment of specialized lineages. Kofuapie describes the rela-

tionship between these two local clans and in this context excludes the
other clans of the opposite moiety.
The notion of pisaua or "friends," as I prefer to gloss it, is related to,
and often coincides with, the notion of kofuapie or other clan, but it is not
synonymous. Nor, I think, is Hau'ofa's construal above quite right. Mem-
bers of each subclan distinguish certain other subclans among their
kofuapie (in any of the above senses) as pisaua or aua friends. Only peace
chiefs may initiate these friend relationships with one another. "Ordinary
men" (ulalu) and officeholders of the other types neither initiate nor
possess friends independent of those of their subclan peace chiefs.
Rather, members of a subclan are friends through their peace chief to the
subclanspeople of his friend (cf. Hau'ofa 1981:162-3).
Friendships between peace chiefs and between subclans are hereditary
along agnatic lines. Any one peace chief may be the friend of one or
several other subclan peace chiefs. Two friends and their respective sub-
clans may reside in different villages, but ideally "first friends" (pisaua
manga) live in the same village. Thus, within a residential clan composed
of heterogeneous subclans and lineages working together, the friend(s) of
one peace chief and his subclan may well be different from those of
another allied peace chief and his subclan. Indeed, a senior and a junior
peace chief of the same clan should not both be friends to the same other
peace chief. Nonetheless, it is expected that their respective friends will
be peace chiefs and subclans of the same clan, and that the senior chief of
the one clan will be friend with the senior peace chief of the other, and
likewise for the junior peace chiefs of their respective subclans. Pisaua
friendship as distinct from kofuapie relationship ties senior peace chiefs
and senior subclans with one another, and junior peace chiefs and junior
subclans with one another. Thus, no one peace chief or subclan can be
friends with all kofuapie peace chiefs and subclans in the tribe.
The network of friend relationships of the Amoamo tribe as of 1974-6 is
represented in Figure 6.7. Close examination of this network reveals a
number of important organizational features inherent in interclan rela-
tions. First, all friends are kofuapie, but not all kofuapie are friends.
Second, some friend relationships involve a subclan that has become
extinct. Nonetheless, these ties are maintained in perpetuity by the peace
chief of another subclan of the same local or dispersed clan at feasts and
other ceremonies in the name of the last officeholder. Third, there are
several instances whereby one subclan is friend to two subclans that are
working together in the same local clan. The peace chiefs and subclans
involved in these particular relationships, however, are of the same se-
niority status. Fourth, in only one case is a senior peace chief (Nganga
 Kin, clan, and connubium 131

KUIPA @<t
(B. Mekeo)
KUIPA @
| (B. Mekeo)

BIOFA @<t
(C. Mekeo)/

INEUCLAN

NGANGA CLAN ^Ineu (f)

N lneu(e)

\ / \lnemay
Initsi village
N. Angai (f) (

N. Angai (e) # (
Akabe village
N. Manga( \ ) Olapu

) Inafufu (f)
N. Faka(f)(
) Inafufu (e)
\ N. Faka (e) (

OLAPU CLAN
)Olapu (f>
\Olapu (e) \
/

AMOAMO TRIBE Etsetse (f)

^Ngongofobina #

Etsetse (e)

\\ ETSETSE CLAN
s VEKE @
X (C. Mekeo)
\ Iko village

Figure 6.7. Pisaua friendship network (Amoamo tribe), (f) = senior fakania sub-
clan; (e) = junior eke subclan; @ = intertribal friend; 0 = extinct friend; # =
extinct subclan.

Angai) friend to a junior chief (Etsetse). Here, however, members of the

two Etsetse subclans and their peace chiefs periodically dispute their
relative seniority. Fifth, first friends do not always live in the same vil-
lage. Informants clarified these discrepancies by claiming that long ago
first friends did live in the same village, but during the ensuing years of
war and migration they separated and resettled apart from one another.
132 Quadripartite structures

At mortuary feasts, hereditary first friends still take ceremonial prece-

dence regardless of co-residence. But otherwise friends who live in the
same village "really do make like first friends" to one another. Sixth, it is
apparent that the traditional dual organization of the society into discrete
moieties has broken down, if indeed it ever existed (Levi-Strauss
1963a: 132-63). That subclans of two local clans are friends to each other
and to the same other subclans makes the existence of actual exogamous
moieties now impossible. I shall presently show, however, that the fact
that there remain four discrete, named dispersed clans in the tribe is
significant and compatible with dual structuring at another level.
As Guis (1936:40) observes for the Central Mekeo, marriage among the
Bush Mekeo is initiated in either of two ways: (1) by "betrothal" (onge
apua) in advance of cohabitation, or (2) by "elopement" (kepiau) (see
Seligmann 1910:363n, where he acknowledges this same distinction). Ide-
ally, everyone should marry by betrothal with friends, but if people do
marry kofuapie who are not friends, then elopement with a nonfriend is
appropriate.
In the old days, my acquaintances told me, a young man could acquire
a wife through betrothal only if his father and other close relatives had
plenty of valuables (kefu) and surrendered them in advance, that is, by
betrothal. Failing this, bachelors had to employ courting ritual on a girl of
their own choosing who was not a friend and elope with her to the bush.
Usually, the quantity and quality of shell, dogs teeth, and other valuables
exchanged in marriage compensation differ considerably in the two cases.
Elopement among friends is considered very bad form and, until rela-
tively recently, often results in the girls return to her parents until be-
trothal is arranged with the transfer of marriage compensation. The par-
ents and relatives of a girl "stolen" by a kofuapie youth who is not a friend
may still try to take her back, but it is not the same as if she had been
taken by a friend; they may allow her to stay with her paramour. Betrothal
is inappropriate between people who are not friends but only kofuapie.
They will exchange marriage compensation some time after the elope-
ment and establishment of a married household.
The percentages of intratribal marriages contracted before 1965 in
these terms among living adults are represented in Table 3. Based on my
village census of 45 total marriages, 9 (20%) are by betrothal and 36 (80%)
by elopement. Of the betrothals, 7 (78%) are between friends, and 2
(22%) are between nonfriend kofuapie. Correspondingly, 6 (16.6%)
elopements involve friends, and 30 (83.4%) involve nonfriend kofuapie.14
It should be noted that 2 of the 6 elopements between friends involve the
remarriage of widows, and 2 more involve the taking of polygynous sec-
ond wives. Neither type of these marriages usually results in the public
exchange of marriage compensation. Also, of the 16 marriages in the same
 Kin, clan, and connubium 133
Table 3. Friend and nonfriend kofuapie
betrothals and elopements

Betrothals Elopements
Friends 7 (78%) 6 (16.6%)
Nonfriend kofuapie 2 (22%) 30 (83.4%)

period with people from other tribes and ethnic groups, none were be-
tween friends, and all were by elopement.15
Between 1965 and 1976, nearly all marriages (28 of 31) were initiated
by elopement.16 Among these, there is one elopement between friends,
two friendship betrothals (one of these involving a new intertribal friend),
and one betrothal between persons who were nonfriend kofuapie. Catho-
lic missionaries and teachers nowadays successfully encourage young peo-
ple to choose their new spouses. Marriages, it is hoped, will then have a
greater chance of proving amicable and stable. Additional factors, es-
pecially moneys increasing availability and its entry into marriage trans-
actions along with the postcontact escalation of sorcery and courting ritu-
al, have contributed to this recent change. Also, since approximately
1965, the percentage of marriages with persons outside the tribe has
increased (16 of 31 total, or 51.6%). These figures suggest that until
recently the Bush Mekeo have been marrying in much closer conformity
to the rules of kofuapie and friend prescription than some earlier investi-
gators had suspected. The problem until now has involved, again, a
failure to portray accurately the indigenous categories kofuapie and
pisaua and how they articulate with expressed marriage rules and the
alternate means of contracting marriages within the single rubric of mar-
rying kofuapie.
Marriage rule V: sister exchange
With this issue now clarified, I will make note of the one remaining rule
by which villagers claim they regulate marriage in their society. Accord-
ing to this rule, "sister exchange" (pitonge kaua, kengami kaua) is prohib-
ited; that is, two men of different clans should not marry one another's full
sisters at the same time, and even if they are contemplating marriage at
different times, the second must not be initiated until compensation has
been paid in full for the first. Villagers frown upon sister exchange, they
say, because there is no public marriage compensation or exchange ac-
companying it. With sister exchange, each of the groups would be both
"hanger" or wife-receiver and "eater" or wife-giver to the other. They
would be the "same" (ikopoangaimo), and "compensation" (kaua) would
thereby be already finished. These assertions will acquire more meaning
134 Quadripartite structures
below when I discuss the exchanges involved in marriage compensation,
and it will be shown that the prohibition of sister exchange in these terms
has unexpected structural significance. I should add here, however, that
ideally sister exchange whenever it does occur is never the result of the
betrothing of friends in marriage; that is, it may only result from elope-
ment.

Marriage in alternating generations

Consistent with other systems that have been likened to the "Aranda'-
type, marriages between friends and kofuapie among the Bush Mekeo
may be repeated in alternating but not in adjacent "generations' (with
one qualification; see the concluding section of this chapter). Here I shall
discuss the implications this feature of the Bush Mekeo system has for the
public exchange (akaila) of valuables in marriage compensation.
For Ego, the rules of kin classification and the rules of marriage coin-
cide such that persons of the same generation who are not reciprocally
ekefaka are excluded from the marriageable category. Beyond this radius,
opposite-sex relatives of the same generation within the tribe are or may
use the term ekefaka, and they are potential spouses. Furthermore, this
boundary is isomorphic with the line separating own moiety, clan, and
papie ngaunga, on the one hand, from marriageable kofuapie and friends
on the other. From either perspective, Ego (male or female) may not
marry a person of mother's clan (i.e., where father married) but may
marry into a different clan of mother's moiety, particularly where father's
father married.17 Figuring strictly along a single patriline, then, the rules
and categories of the Bush Mekeo system enjoin (or at least allow) mar-
riage between persons of two clans of opposite moieties in alternating
generations.
Although papie ngaunga may not themselves marry one another, the
children of same-sex papie ngaunga may marry. Or, stated in somewhat
different terms that structurally mean the same thing for the sake of
comparison, first bilateral cross-cousins may not marry, but the children
of first cross-cousins in different moieties may marry (Figure 6.8) (see also
Williamson 1913:275).18
As the members of each generation mature, marry, and reproduce,
prohibited papie ngaunga relatives are succeeded by offspring who are
not papie ngaunga but ekefaka and marriageable. Conversely, ekefaka
who do marry in one generation are succeeded in the next by papie
ngaunga who cannot marry. Clans of opposite moieties correspondingly
and alternately open and close their boundaries with respect to one an-
other. Stated in more familiar anthropological terms, certain cognates in
one generation are transformed into potential affines in the next. And
conversely, affinity in one generation is transformed into cognation the
 A1

i I
Figure 6.8. Children-of-"first cross-cousin," or "second-cousin," marriage.
 spouse's mother's clan
eaters

*from perspective of groom

Figure 6.9. Akaila public marriage compensation exchange.

 Kin, clan, and connubium 137

next. Consequently, affines in that generation are transformed back into

cognates in the succeeding generation.19

Marriage compensation
These complementary processes are represented most graphically in the
exchanges by which members of different clans publicly give and receive
"marriage compensation" (kaua).20
Every occasion of public marriage compensation must involve the par-
ticipation of four parties ideally exemplified21 by father's elder brother
and mothers elder brother of the groom, and the fathers' elder brother
and mother's elder brother of the bride. The two fathers' brothers and
their respective clans are termed the "owners of the mother" (ina ngome).
The mothers' brothers of the bride and groom and their clanspeople are
called "women's children" (papie ngaunga). Together, the two clans al-
lied with the groom are known as "compensation hangers" (kaua kopa).
Hangers give, in two separate prestations, valuables22 to the "compensa-
tion eaters " (kaua ani) allied with the bride. The two payments from the
hangers to the eaters cannot be made on different occasions, and every
effort is usually taken to ensure that both prestations include equal pro-
portions of the same categories of valuables. Groom's father's brother
gives to the bride's father's brother, and the groom's mother's brother
gives to the bride's mother's brother. The eaters usually reciprocate pres-
tations to the hangers, but only in token amounts and qualities compared
with those they receive. Wife-giving eaters, in other words, are "above'
or superior to wife-receiving hangers (Figure 6.9).
From the perspective of the groom, the clan identities of the four
exchanging exemplars and clans are as follows:

own clan —> spouse's clan mother's clan —> Y > i

* mother s clan
And from t h e perspective of the bride:

own clan <— spouse's clan mother's clan <—

, , ,
mother s clan
Furthermore, both perspectives can b e summarized as follows: (a) At t h e
level of interclan relations:

own clan : spouse's clan :: mother's clan : V > 1 (16)

mother s clan
(b) in terms of the compensation exemplars:
father's spouse's mother's spouse's . _,
brother ' father's brother " brother ' mother's brother

and (c) in terms of relational blood categories:

138 Quadripartite structures

,, , . cognatic nonblood
agnatic : nonblood :: cognatic :. . /1OX
ui J / a> \ ui J (same
v moiety, (18)
blood (afhnes) blood , . '
other clan)
Consistent with the rules of clan, moiety, and papie ngaunga exogamy, no
two of the four exemplars' roles may be held by one person or even by
agnatically defined members of the same clan. Also, two distinct clans
from each moiety should be represented. Thus, proscriptive rules of
marriage and the pattern of marriage compensation exchanges presume
that the two moieties of a tribe are bisected minimally into two dispersed
clans each. One clan of a moiety will hang or give while the other will eat
or receive at the same gathering:
clan A1 - * clan B1, clan B2 -+ clan A2
(Here, as elsewhere in this work, upper-case letters refer to moieties, and
numbers refer to named clans of those moieties.)
Homologous with the relations depicted above,
clan A1 : clan B1 :: clan B2 : clan A2 (19)

own clan : spouse's clan :: mother's clan : V > i (16)

mother s clan
father's spouse's mother's spouse's . ^
brother ' father's brother " brother " mother's brother
ii i . cognatic nonblood
agnatic nonblood cognatic , . ,,_.,.
:
ii j : / rr- \ '•'- ui J (same moiety, (18)
v v ;
blood (affines) blood . . /
other clan)
For several reasons, I have hesitated glossing these public transactions
as "brideprice " or even "bridewealth." Villagers insist they do not trans-
fer valuables on the occasion of marriage for the purpose of buying or
paying for the bride. Rather, the hangers compensate the eaters for the
loss of certain rights in the children to be born of the bride that will be
transferred to the hangers. More to the point, the occasion on which
compensation is made is referred to as "doing/making/manipulating
blood" (ifa kekapaisa). I mention these ethnographic details because they
shift the focus and scope of the Bush Mekeo marriage system to incorpo-
rate the blood identities of the prospective offspring, the exchanging
exemplars, and their clans, along with the blood identities of the bride
and groom.
The expression "manipulating blood" refers to any social encounter
where relatives or affines violate or contradict the rules by which they
should interact with one another, particularly when quarreling or disput-
ing is involved. The term for "quarrel" or "dispute" is kelele or kengenge
- literally, reciprocal "loss" or "crossing over" (enge),23 as on a bridge.
Marriage-compensation exchange as one kind of manipulating blood typ-
 Kin, clan, and connubium 139

ically involves a fair degree of quarreling. Villagers explicitly emphasize

that cognates, agnates, and affines should never openly quarrel or dispute
with one another, nor should they do things that they know might lead to
quarrels. When they do deviate from the ideal in these respects, it
amounts to denying the fact of their relationship. Brothers, for example,
should never quarrel; they should help one another with house building,
cooperate in gardening, share food cooked by their wives, never commit
adultery with one another's wives, and so on. When one brother fails in
any of these things, he virtually says to the other, "You, my brother, are
not my brother. We are different, not the same, blood." When villagers
manipulate blood in quarreling, as in this example of brothers, the im-
plicit or explicit denial of blood between them is acknowledged as a
"fiction" or "lie" (pifonge); the two men are still one blood even though
they act as if they are different bloods.
Marriage compensation, as one context of manipulating blood, also has
this element or quality of fiction making about it. The cognatic blood
relationships of bride and groom to the clans of their fathers' mothers and
their mothers' mothers are not flatly negated in marriage compensation;
they are manipulated so as to publicly create the impression that they
have been negated. Bride and groom are thus each considered to proceed
the rest of their lives with all four of their overlapping ancestral cognatic
bloods traced to their grandparents, but, in the context of marrying one
another and reproducing together, a fiction is created such that each has
only two bloods - those of their respective father's and mother's clans
(Figure 6.10). This point will assume considerable significance when in
the next two chapters I describe the indigenous meanings of funerary
ritual.
In these terms, marriage compensation resolves the contradiction of
human reproduction whereby purportedly closed agnatic clans, circum-
scribed by one blood traced exclusively through males, are open to other
clans by virtue of the exogamous exchange and transmission of different
blood through women. For a couple to cohabit as husband and wife, their
respective relatives should ideally have first betrothed them and com-
pleted the compensation exchanges. Also, cohabiting husband and wife
should be of non- or different bloods in relationship to one another. Thus,
the parents and compensation exemplars of a groom and his bride ideally
belong to four distinct named dispersed clans representing distinct ag-
natic bloods, that is, two bloods for each of the two moieties. Children
born of this marriage will possess cognatic blood of all four of these clans
represented in the persons of the exemplars of the marriage compensa-
tion. However, when the children grow up and themselves marry, only
two of the four blood identities will continue to be deemed relevant; the
other two will be manipulated so as to be irrelevant. With reference to
Figure 6.10. Agnatic, cognatic, and affinal bloods, x = children of bride and groom; Y ~
potential spouses.
 Kin, clan, and connubium 141
Figure 6.10, if the exemplars for the marriage of the bride and groom
were members of clan A1 (groom's fathers brother), B2 (grooms mothers
brother), B1 (bride's father's brother), and A2 (bride's mother's brother),
then exemplars for the marriages of the children (X) of this marriage will
be of clans and bloods A1 (father's brother) and B1 (mother's brother). The
other two clan and blood identities that were represented in the marriage
of their parents are not represented among the exemplars of the chil-
dren's own marriages. According to the rules of marriage prescribing
someone of different blood(s), the two lost bloods of the children will be
those of their ekefaka and potential spouses (Y), or persons with B2 and A2
bloods from father and mother, respectively. Relations of shared blood
and cognation, in other words, are being manipulated for the sake of
marriage into relationships of different blood and affinity (Mosko 1983).
Significantly, any kin nomenclature employed between a man and woman
before they marry is abandoned on appropriate occasions after paying
compensation and is replaced by the terminology of affines who are im-
plicitly not related by blood.24
Still with reference to Figure 6.10, the two public compensation ex-
changes should be understood in the following terms. By the transfer of
valuables from the groom's father's brother to the bride's father's brother,
the groom is affirmed to possess agnatic blood of his father's clan (A1) but
not the cognatic blood of his father's mother's clan (B1). Conversely, it is
simultaneously affirmed that the bride possesses agnatic blood of her
father's clan (B1) but not her father's mother's clan (A1). The exchange of
valuables between mothers' brothers works similarly. The groom is af-
firmed to possess the cognatic blood of his mother's clan (B2) but not the
cognatic blood of his mother's mother's clan (A2), and the bride is affirmed
to possess the cognatic blood of her mother's clan (A2) but not the cognatic
blood of her mother's mother's clan (B2). The bloods acknowledged for
the groom are those denied for the bride, and vice versa. After the
exchanges are completed, bride and groom can be regarded as persons of
different blood and, therefore, legitimately potential spouses.
By giving and receiving marriage compensation, patrilines of clans of
opposite moieties alternate between affirming and denying cognation
consonant with intermarriage in alternating generations. Thus, patrilines
of clans in opposite moieties that open to one another in one generation
by mixing and sharing cognatic blood in relations of papie ngaunga close
in the next to permit reciprocal intermarriage. Moreover, although one
patriline of a clan is closed to the patriline of another clan in the opposite
moiety, other patrilines of the same two clans may be open to one an-
other. I shall return to this point in the concluding section of this chapter.
On the occasion of exchanging marriage compensation, two clans of
each moiety are present, but they do not exchange valuables directly.
142 Quadripartite structures
Rather, they participate in separate exchanges. Nevertheless, while each
party denies or negates cognatic ties with different clans of the opposite
moiety, cognatic ties between clans of the same moiety (i.e., between
own and mother's mother's clans for both bride and groom) are ex-
tinguished. Clans of the same moiety, in other words, can maintain the
ideological claim of purely agnatic relationship and closure without
contradiction.
Finally, although each clan as a unit is open to all other clans in the
society through procreation and the relations of cognation it generates, by
participating in all four roles of marriage compensation, each clan
eventually reasserts its distinctiveness and closure vis-a-vis the other
clans both in and out of the same moiety.
Before I move on to other implications specific to the Bush Mekeo
system, it is important to note that structurally, according to Figure 6.10,
paired giving and receiving compensation exemplars are alternatively
bound to one another in reciprocal cognatic relationships of papie
ngaunga. Groom's father's brother is papie ngaunga to bride's father's
brother, and groom's mother's brother is papie ngaunga to bride's moth-
er's brother. The persons who manipulate blood and suspend cognatic
blood relationships between bride and groom are, in other words, them-
selves exemplary of those categories of cognatic blood relationship. Be-
cause they are cognatically open in their mutual relationships, they are
the appropriate persons to close them fictively for members of the next
generation.

Homologies of the system

I would now like to complete my description of the Bush Mekeo rela-
tionship system by articulating some of the more obvious homologies
between its categories and those of other contexts in the culture that have
been discussed at length in previous chapters.
First, let me mention again that there are four actual named, dispersed
clans (each including one or more local clans) in Amoamo tribe - precisely
the number of exogamous social units required for the system to work
according to its rules. It is curious that none of my informants ever
explicitly declared that there should be just four clans in the tribe. None-
theless, I think it is clear that the fact that there are four clans demon-
strates once more the Levi-Straussian tenet concerning the priority of
structure over event. The structure, here in its minimal or simplest form,
involves some of the implications that have been drawn for the Australian
"Aranda'-type systems; namely, that all first cousins are prohibited from
marrying while classificatory second cross-cousins are permitted to mar-
ry, and that the relevant named categories are four patrilineal clans with
papie ngaunga ties between them instead of either bilateral kin terms
s
PT3
CD
144 Quadripartite structures
only or eight matrilaterally transmitted marriage classes or subsections
(Figure 6.11). As noted in the previous section, these relations between
clans of a tribe can be expressed as:
own clan : spouse's clan :: mother's clan : spouse's mother's clan (16)
All members of own clan are one strictly agnatic blood and unequivocally
inside the same clan together. Their respective mothers and spouses,
however, are both in these terms outside own clan and not one agnatic
blood. Members of mother's clan, however, are one cognatic blood with
their papie ngaunga in own clan. Their blood has come from the outside;
that is, the blood of mother's clan is outside blood inverted with respect to
own clan. In contrast, members of spouse's clan are unambiguously out-
side own clan, for as affines their blood is neither agnatic nor cognatic; it is
different blood. Members of spouse's mother's clan, finally, are one ag-
natic blood, and therefore inside the same clan qua moiety as own clan.
But their blood is also everted from inside the moiety outside (i.e., into
spouse's clan). The agnatic blood of spouse's mother is, therefore, everted
inside blood in relation to own clan.
In this context, it is significant also that each clan opens its boundaries
vis-a-vis other clans through their skins (women), whose procreative
womb-blood goes from inside their own clans and bodies outside and then
inside other clans. Moreover, in terms of the hot/cold opposition, own
and spouse's clans are hot for sex and reproduction where mother's and
spouse's mother's clans are cold (in relation to own and spouse's clans).
In precisely these terms, interclan relations are homologous with the
category distinctions that pervade the other contexts of the culture I have
described. This is most obvious perhaps in the comparison with spatial
relations, the categorizations of substances, culinary techniques, sexually
distinguished bloods, and the division of adult labor according to gender:
inverted everted
inside : outside (4)
outside inside
remote peripheral village adjacent
(3)
bush village abdomen bush
blood : sweet :: unsweet bloodless dirt
roasting : boiling :: ripening smoking
male female
semen :: womb-blood (13)
body-blood body-blood
closing opening opening closing
(14)
male male female female
clan A1 : clan B 1 :: clan B 2 clan A2 (19)
spouse's mother's spouse's
UWI1 LI till (16)
clan clan mother's clan
 Kin, clan, and connubium 145

father's spouse's mother's spouse's . -,

brother ' father's brother " brother ' mother's brother
i! , .. cognatic nonblood
agnatic nonblood cognatic , . , x
: :: :
ui J / £c> \ ui J (same moiety, (18)
v v
blood (affines) blood . . ;' '
other clan)
Several dimensions of the total relationship system I have already de-
scribed have yet to be fully articulated with this general pattern: se-
nior/junior differentiations among subclans; the double bisection of each
subclan into four ritually specialized lineages; and marriage rule V, which
forbids sister exchange of a narrowly defined sort.
The Bush Mekeo system as depicted in its most elementary or essential
form does not as yet actually represent a complete model of the total
society. As elaborated thus far in Figure 6.11, the model presumes: (1) no
division of clans into senior and junior subclans; (2) that each clan inter-
marries with the clans in the opposite moiety only in alternating, and not
in every succeeding, generation; and (3) that every marriage is an appar-
ent case of sister exchange. The model must therefore be elaborated
further to accommodate these formal requirements, with the result that
each clan is divided into senior and junior subclans, and each subclan is
doubly bisected into four lineages.
Beginning with the marriage rule for friend prescription in every gen-
eration, members of a patriline in one clan may not marry into a certain
clan of the opposite moiety because they are papie ngaunga, but same-
generation members of a parallel patriline of the same clan who are not
papie ngaunga to that opposite moiety clan may marry there. Thus, the
rules of friend prescription and papie ngaunga proscription combine to
bisect each clan into two groups (Figure 6.12).
The rule forbidding simultaneous sister exchange similarly requires the
bisection of the clan, but along another dimension. If a man marries a
certain woman of a clan in the opposite moiety, his wife's brother and his
sister may not also marry until after there has been compensation for the
first marriage. Because the mans sister cannot marry into her mother's
clan of the opposite moiety, she must marry a man of her brother's wife's
clan but not her brother's wife's full brother, and likewise for him. Thus,
each of the two patrilines of every clan represented in Figure 6.12 must
be further bisected so that full sibling brother and sister do not marry a
full sister and brother. This results in a total of four patrilines or lineages
per clan (Figure 6.13). In each generation, members of two lineages of a
clan will marry into one clan of the opposite moiety when the two other
lineages of the former clan marry into the remaining clan of the opposite
moiety. Of any two same-clan lineages intermarrying in the same genera-
tion with the same clan of the opposite moiety, each of the former will
K S u
b—
I
tj TO
k>-| l<^ -r
kM

CD
0)
bC
 Kin, clan, and connubium 147

Figure 6.13. Bush Mekeo marriage system (iii).

marry into distinct patrilineages of the latter. Each of the four clans then
would ideally possess four lineages.
It might be initially proposed that either one or the other of these
bisections of the clan unit corresponds with the division of each clan into
senior and junior subclans. This could not be the case, however, for the
model just elaborated is premised upon marriages contracted according to
the expressed ideal, that is, according to betrothal between friends who
must be of the same status level. Friends in opposite moieties should be
both senior, or both junior. Moreover, the entire network of friend rela-
tionships at each status level should be duplicated in the nonoverlapping
friend network of the other status level. Each clan is thus ideally bisected
into senior and junior subclans, and then both of the latter are doubly
bisected as described above. Altogether, the model specifies that each
tribe has two moieties, each moiety has two clans, each clan has two
148 Quadripartite structures
subclans, and each subclan has four patrilines or lineages. I am arguing, of
course, that the four patrilines or lineages here correspond with the
fourfold ritual specialization of lineages and offices within the subclan:
peace chiefs, peace sorcerers, war chiefs, war sorcerers. The distinctions
between different categories of official male ritual and specialized lin-
eages, in other words, are predicated in part upon the category distinc-
tions also underlying the system of marriage regulation just elaborated.
Moreover, in these same terms, specialized lineages and offices are ho-
mologous with the differentiations among clans comprising the entire
tribe:
own clan : spouses clan :: mother's clan : spouse's mother's clan (16)
inside : outside :: inverted outside : everted inside (4)
peace chief : war sorcerer :: war chief : peace sorcerer (15)
The latter of these can now be added to the other homologies of the
culture listed previously.
To me, this is a surprising discovery, for this specific relation has not
been suggested as far as I am aware in the ethnographic literature for the
immediate area of Austronesian-speaking Papua where very similar cate-
gories are prevalent. The internal composition of the Bush Mekeo subclan
as a unit and the corresponding divisions of ritual specialization within it
are representations of the total society and of the world.
The derivation of a model for Bush Mekeo society in these terms indi-
cates a certain unexpected additional correspondence with the prevailing
generational kin and affinal terminologies. According to the Bush Mekeo
marriage system as I have just elaborated it above and in Figure 6.13,
lineages considered as units may reciprocally intermarry in alternate gen-
erations. However, the full completion of every marriage cycle such that a
marriage can be repeated exactly between the same two lineages ideally
takes four generations. Exogamous-lineage intermarriage in one genera-
tion might seem to be repeated in the next alternate generation, except
that the sex identities of each participant are reversed rather than repeat-
ed. Thus, a total of four, not two, generations of reciprocal intermarriage
defines each marriage cycle. A classificatory scheme that, with respect to
Ego, locates bilateral kin of exactly four generations in both ascending and
descending generations is perfectly consistent with this stipulation. And
this is precisely what I described earlier in this chapter for the indigenous
Bush Mekeo kin and affinal nomenclatures. The system of cyclical mar-
riage and fourfold lineage ritual specialization is, in other words, struc-
turally isomorphic with the classificatory nomenclature of the language.
Moreover, that this latter terminology specifies precisely four reciprocal
generational categories suggests an added ramification of the more gener-
al and pervasive quadripartite cultural pattern. Apparently, the Bush
 Kin, clan, and connubium 149
Mekeo identify four major categories of intergenerational relationships
according to the same "plan" by which they recognize four of virtually
everything else in their world.
These last several findings, oddly enough perhaps, bring me back to
consider the relation of the person to the total society as conceptualized in
the culture. The integrity or completeness of the tribe as a unit is con-
stituted by four systematically differentiated agnatic bloods representa-
tive of the tribe's four discrete named, dispersed clans. Although every
person in the tribe is identified with only one clan of the tribe by strict
agnation, it is implicit in the rules of kin, papie ngaunga, and cognatic
blood classification and in the rules of marriage prescription and proscrip-
tion that every person is a complete personification of the tribe and,
indeed, the whole social universe. From his or her four grandparents,
every person inherits at conception and birth all four bloods of all four
tribal clans. The person then is not just a discrete element of larger social
wholes. By blood, the person is a total representation of those social
wholes. In other words, the person is a social persona tied in systemat-
ically differentiated relationships of shared blood with all other similarly
constituted social personae in the society.
We have seen in this chapter that villagers reach a critical point in their
lives when compensatory exchanges are made on their behalf to create
the fiction of disavowing two of four ancestral bloods provisional to mar-
riage and reproduction. With death, each person reaches another critical
point in his or her life. In the next chapter, I shall show that similarly
significant disavowals of ancestral blood are finally expressed in the rituals
of mourning with the periodic reciprocation of elaborately staged mortu-
ary feasts between the different clans of the tribe.
 7
Feasts of death (i): de-conception and
re-conception

To this day, the most authoritative anthropological treatments of the

symbolism and ritual of death remain Hertz's essay "The Collective Rep-
resentation of Death" (1960:27-85) and Van Gennep's study The Rites of
Passage (1960). Although these classic interpretations differ widely, in
one sense the two are very similar. Hertz and Van Gennep each convinc-
ingly claims to unearth a particular structure or pattern of cross-cultural if
not universal validity from among the diversity of mortuary ritual in the
societies they study. Where they differ is in the exact constitution of their
respective structures.
Although I cannot but agree that it is possible to identify some ele-
ments of both Hertz's and Van Gennep's formulations in the ethnographic
elements of Bush Mekeo mortuary ritual and feasting, my principal in-
terest and intentions in this chapter lie elsewhere. Throughout this work,
I have tried to reveal a number of structural homologies pervading all
contexts of the indigenous culture, and I shall continue to do so as con-
sistently as possible here regarding Bush Mekeo conceptualizations of
death and mourning. Therefore, I shall not be explictly concerned with
establishing relations between corpse, soul, and survivors, or with por-
traying relations between funerary and other rites of passage, except
insofar as they contribute toward eliciting the structure of category dis-
tinctions found generally throughout the culture and specifically in mor-
tuary ceremony and feasting.1 In particular, I shall show that in the
process of performing their funerary rituals, Bush Mekeo villagers finally
affect what is only fictively and tentatively accomplished with marriage
compensation, as discussed in the previous chapter; that is, villagers
resolve the contradiction of blood versus nonblood relationship at the
collective level, thereby conceptually providing for the structural con-
tinuity and coherence2 of their society and the distinctiveness of its com-
ponent social units (Mosko 1983).
The agents and methods of bodily death
As villagers view it, the world offers many (if not too many) ways for them
to die. The Bush Mekeo term for "death" and "dying" is mae. People can

150
 Feasts of death (i) 151
die from the bite of poisonous snakes, chronic illness and starvation,
infection from wounds, violent homicide, maiming by wild boar and cas-
sowary, falling trees, and faifai and crocodile attack. There is, however,
no cultural recognition of "natural death" or even "death from old age," as
we view it. All deaths are understood to be rather the consequence of
manipulations of hot bloody and dirty substances by certain agents upon
their victims.
The agents of death may be human or nonhuman, or spirit or nonspirit.
Peace sorcerers in league with ancestral spirits of the dead are accorded
the blame for death by snakebite and for many fatal illnesses. Certain
other symptoms of disease that may lead to death are caused by faifai
animals or their humanlike spirits, the ongokapu bush people. Death in
battle or from self-inflicted wounds, falling trees, and wild boar and cas-
sowary attack, however, are deemed the handiwork of war sorcerers and
war chiefs in control of ancestral spirits. Crocodile specialists (oala)3 are
blamed for all cases of crocodile attack. The different kinds of poisoning
require one or another category of human or spirit poisoner: peace sor-
cerers, disgruntled wives, adulterers, war chiefs, war sorcerers, and spir-
its of homicide victims.
The substances that agents of death manipulate with respect to their
victims include the flesh or blood of dead human beings, menstrual
blood, afterbirth and birth-blood, the flesh of faifai animals, and "poi-
sons" (ipani). Indeed, all these substances are categorically defined as
poisons if they are physically ingested; all of them are hot, bloody, and
dirty. This is even true of the class of the nonhuman and non-faifai
traditional poisons, which include inanimate mineral and plant sub-
stances. Although the ingredients of these particular poisons are secret,
two I know of - a certain red ochre called "Akaisa's lost blood" (Akaisa
enge ifa) and an orange mushroom called "Akaisa's penis" (Akaisa lako) -
are deemed the blood and flesh of the mythical culture hero Akaisa (see
Chapter 8). The substances manipulated by agents of death upon their
victims, then, include hot bloody dirty things from various human or
humanlike bodily sources.
The ritual or manual insertion of hot bloody dirty substances into the
victim's body reverses the state of health and life and so produces illness
and death. In these terms, death is the opposite of life. With the pro-
cesses of cooking, eating, digestion, work, reproduction, nursing, male
tightening ritual, and female postpartum memengoa, I have shown that
forms of hot such as fire, cooked food and water, work, sexual exertion,
and so on are understood to transform people's bodily conditions from
unsweet into sweet (or vice versa, depending upon the specific context)
toward sustaining life:
152 Quadripartite structures

inside : outside :: inverted outside : everted inside (4)

blood : sweet :: unsweet : bloodless dirt (8)
Agents of death similarly transform the sweet and unsweet states of their
victims with their hot dirty substances, but to the precisely opposite
effect, that is, to illness and death:
inside : outside :: inverted outside : everted inside (4)
bloodless dirt : unsweet :: sweet : blood (9)
Victims of these latter transformations are regarded by other villagers as
"crazy" (kafoko). What is sweet and unsweet to the chronically ill, for
example, is opposite to the sweet and unsweet of the healthy. Cooked
food becomes unsweet and the lack of food sweet, so sick people do not
eat. Or if they do eat, either they vomit, or the bloodless wastes of food
they have consumed accumulate in their bodies while the blood-potent
elements of food are excreted. Also, their own blood becomes unsweet
rather than sweet, and it too is excreted rather than remaining inside the
body. Moreover, the victim's "soul" (laulau) is attacked by the "spirit"
(tsiange) of the offending hot bloody dirty substances. It too becomes
unsweet for keeping inside the body, and it leaves. At this point, the
victim will die. Death then occurs according to the same relations that
sustain life, but in reverse.
Death to villagers, however, is much more than a physiological phe-
nomenon pertaining to the individual. If death were merely an aspect of
the individual, villagers might simply discard the corpse or, at best, bury
and keep vigil over it until the flesh had dried and decomposed, and then
be done with it. That villagers do not do this, and that each death is
attributed to the actions of an agent, indicates that death in Bush Mekeo
culture is a preeminently social fact. Indeed, when news of a death is first
spread from village to village within the tribe, people characteristically
express it in terms of an entire clan dying - "Nganga [clan] is dead"
(Nganga kemae), or "Ineu [clan] is dead" (Ineu kemae) - rather than
stating the name of the deceased individual.
Throughout most of this chapter, I shall be examining Bush Mekeo
mourning and feasting ritual as predicated upon this fundamental indige-
nous conceptualization of death as a social, and not just an individual
event. As such, the culture categories relevant to death will be elucidated
in their relations to other contexts of village social life.

The aftermath of death: village hot and village cold

While its resident clans are free of death and mourning, the village and
village life are described as hot (tsiabu). When the village is hot, re-
sources and wastes are regularly passed between the village and bush.
People work, so there is plenty of food to eat. Children play unin-
 Feasts of death (i) 153
hibitedly. Individuals, households, and clans reap the maximum enjoy-
ment their world generally has to offer. Each evening while the village is
hot, the air is filled with singing, chanting, and drumming as bachelors
dressed in all their finery court and woo their loves. Occasionally, a young
couple will be betrothed or elope and marry, and afterward their respec-
tive clans will gather and exchange marriage compensation in animated
proceedings. When the village is hot, life in general is sweet; it is lived to
the fullest, and the people are "happy" (engama).
The event of a death in any one of the resident village clans, however,
changes all this, for death4 too is hot, but of another variety. As such, it
systematically transforms the sweet lives of all villagers into ones unsweet
and full of "sadness" or "pain" (etsetse) instead of happiness. Activities
representative of life in general are suspended. Ordinary household and
village occupations cease immediately. In particular, the singing, drum-
ming, and colorful dress of courting are replaced by mourning and other
activities connected with death. In short, the village, once hot and full of
life, has become cold by the transformative hot of death, and it will
remain so until death is completely removed at the final death feast.
The initial reaction to death at the village provides several important
clues for interpreting the complicated events that are to follow. Upon the
discovery of death in their midst, men and women of the village, both
young and old, rush crying and shrieking in near pandemonium to the
scene of death and gather around the body of the deceased. As the de-
ceased's many relatives arrive, they push their way through the throng
and embrace the body. Amid the tears, conventionalized "dirges" (an-
gopa) are sung. Before too long, the grief-stricken begin to differentiate
themselves around the corpse roughly according to their relationships
with the deceased. Female relatives - agnates and cognates - and the
surviving spouse visibly assume the most prominent and active parts,
which they will play throughout the proceedings. They crowd closest to
the dirty corpse and remain in contact with it, touching and stroking it,
while the men (except perhaps the sons of the deceased and the widower,
if the deceased was female) fall back. This is one context where villagers
overcome their fear of hot and dirty substances and willingly allow them
onto their skins.
That it is women more than men who come into contact with the dirty
corpse is significant. When I observed these events on a number of
occasions, bereaved women forcefully struck their arms and breasts with
their fists in time with the beats of the dirges they sang. Father Guis with
undisguised horror reports that traditionally the female relatives, holding
sharpened pieces of shell, would lacerate their heads and bodies with
these motions and roll frenzied on the ground (1936:121). Under the
mission's influence, these more extreme acts of bereavement have been
154 Quadripartite structures

curtailed, except that women still shave their heads and go through the
motions of drawing blood. Women, it will be remembered, are likened to
the skin of the clan, both their own and mothers' clans. At death, hot dirt
from the skin of the corpse, which is assumed to be bloody, and hot
bloody dirt from inside the clan are symbolically placed on the outside of
the clan - onto its skin or women. Such is the hot of death. It makes
women and clans do as sweet what otherwise would be unsweet to them,
namely, excrete blood from the inside to the everted outside.
These women and those male relatives of the deceased who have either
touched the corpse or touched the bloody skins of the grief-stricken wom-
en must take special precautions thereafter not to allow unwittingly any of
the hot bloody and dirty substances into their own bodies, or they might
sicken and die. Also, they must take care not to endanger others with
their dirty persons. Among the Central Mekeo at Beipa'a, at one point,
contaminated mourners had to live in the bush, as the feasting cere-
monies that remove the dirt had been banned by the mission (Hau'ofa
1981:22).
As preparations are made for burial an hour or so after the initial
outpouring of grief, blood relatives of the deceased differentiate them-
selves from other villagers who are not related by blood. Clanspeople,
papie ngaunga, other resident close cognates, and the surviving spouse
remain gathered together close to the body. Because all these relatives,5
male and female, are reckoned one or the same blood as the deceased
and/or the clan of the deceased, they too "die." In other words, when
they mourn for the loss of their blood relative, they grieve for the loss of
their blood, for the loss of part of themselves. Collectively, these blood
relatives are the "owners" (ngome) of the death.
The living spouses of the owners of death are not themselves one blood
with the deceased,6 and they in contrast do not collectively die. These
people are therefore not owners of the death; they are collectively termed
"in-laws" (ipa ngaua) or, nowadays, in English, "laborers." Very gener-
ally, male and female laborers perform the work that their mourning
spouses and spouses' blood relatives cannot, but that is nonetheless nec-
essary to alleviate their sorrowful and dirty condition. For example, if the
deceased happens to die away from his/her natal village, the male la-
borers carry or canoe the body back to the owners' village. Laborer males
prepare the corpse for burial, dig the grave, and finally bury the body
there. Laborer women perform other work on behalf of the mourning
women. They wipe tears away from their faces, prevent them from
wounding themselves too severely, and assist with shaving their heads.
Also, female laborers will henceforth bring food from the gardens, cook
and serve it to all the people assembled in the village, and each morning
clear the village abdomen of rubbish.
 Feasts of death (i) 155

In the performance of these tasks, laborers also contract dirty sub-

stances from the corpse and mourners on their own skins. As soon as their
tasks are completed, laborers, unlike mourners, wash their skins with hot
water and solvents (mulamula) of dirt. This dirt is not their blood; it is not
part of themselves.
Soon after the initial outpouring of grief, bachelors are sent to neigh-
boring villages of the tribe to inform other blood relatives living or visiting
there of the death and to summon them to the village of the owners.
These blood relatives typically include female members of the deceased's
clan, papie ngaunga, and other cognates of the deceased. The spouses of
these persons accompany them to the village of death and assume the role
of laborers. Thus, by the time all have arrived, there are basically two
categories of people: owners tied by blood relationship to the deceased,
and laborers who are not tied by blood to the deceased or deceased's clan.
This division corresponds precisely with the distinction of blood versus
nonblood discussed at length in the previous chapter. Moreover, here, as
in the context of agnation, cognation, and affinity, the blood/nonblood
distinction crosscuts the line differentiating the two moieties from one
another. Some of the deceaseds blood relatives are members of his or her
own clan and moiety; others through papie ngaunga relationships are
members of clans in the opposite moiety. Correspondingly, whereas the
laboring spouses of the former are ideally members of the opposite
moiety, spouses of the latter who serve as laborers will belong to the other
clan of the same moiety as the deceased (or the clan of owners' spouses'
mothers). There are, then, four categories of participants gathered imme-
diately after death: mourners or owners in the same and opposite moiety,
and laborers in the same and opposite moiety with respect to the de-
ceased. Indeed, these four categories at the time of death are isomorphic
with the four categories of persons who gather to give and receive mar-
riage compensation and to fictively rearrange cognatic and agnatic blood
identities for brides and grooms:

same- opposite- opposite- same-

moiety : moiety :: moiety : moiety (20)
owners laborers owners laborers
,, , L. cognatic non-blood
agnatic :
non-blood ::
cognatic :
, . /1OX
ui j / a> \ ui J (same moiety, (18)
v
blood (afhnes) blood . , '
other clan)
, spouse's mother's spouse's ,,„,
own clan : , :: . : A. > i (1")
clan clan mother s clan
As we shall see, the exchanges undertaken among the participants at
death ceremonies, particularly at feasts, have very nearly the same pur-
156 Quadripartite structures

pose as marriage-compensation exchanges - the relinquishment of certain

blood ties and the affirmation of certain others.
The surviving spouse of the deceased plays a unique role in mourning
proceedings. Villagers claim that a husband and wife become one blood
once they have produced children. Assuming they did produce offspring
before the spouse died, the widow or widower, like any blood relative, is
expected to show signs of being sad or pained at the loss of her or his
blood and to observe mourning. But the surviving spouse is also an in-law
and not related by agnatic or cognatic blood traced through the deceased
to the owners. For this reason, apparently, many suspicions are focused
upon the widow or widower, and they typically undergo the harshest of
treatment at the hands of the owners. In the first few days following
death, the widow or widower may be beaten or otherwise physically
abused by the owners. The widow or widower must not diplay the
slightest resistance. Also, they are given only minimal quantites of in-
ferior food. Thereafter, the roles of widow and widower diverge some-
what. A widow must sit nearly motionless hidden inside the house she
shared with her husband (underneath or in front of which his body is
buried). She cannot speak above a whisper nor receive visits from her
own relatives except at the permission of the owners. And she may leave
only for the purpose of bodily elimination under cover of blackened bark-
cloth blanket that hides her dirty skin. Widowers are similarly confined
during the day in a bachelor's house (kofu) walled up for the occasion, but
at night they may emerge to assist in the grave-site vigil.
Villagers explain that widowers and widows must suffer these abuses
because of their profound sadness and pain now that death has made life
unsweet for them. But also, it seems, the surviving spouse is in some
respect held accountable for the death. This does not mean that the
surviving spouse is blamed for actually killing the deceased, however.
Instead, widows and widowers might be accused of not giving food to, or
generally not caring for, the deceased in his or her lifetime, of lying when
they cry, of intending to marry someone else, or of having committed
adultery. Only very rarely is the surviving spouse indicted of having
conspired to kill the deceased. Widows, in particular, might be suspected
of having put menstrual blood in their husbands food; widowers, of care-
lessly exposing their wives to their hot ritual charms or their own persons
when it was unsafe to do so. Even if the other mourners are firm in their
belief that sorcery, for example, was the cause of death, they may accuse
the widow or widower of having allowed the sorcerer access to the de-
ceased. The widow or widower then suffers at the hands of the mourning
owners in most cases, not because they are thought to have poisoned or
killed their spouses, but because they did not keep the vigilance neces-
sary to prevent others from doing so.
 Feasts of death (i) 157

A further differentiation in the roles of widow and widower relates

directly to their sexual identifies. Inside the walled bachelors quarters, a
widower ritually tightens himself. Upon emerging at night, he is fully
armed. He assists in the grave-site vigil by patrolling the outskirts of the
village ready to kill or chase away any peace sorcerers attempting to rob
the grave or claim additional victims. As his body closes, the widower
becomes more and more immune to the hot of the sorcerer's charms and
spells, and he is increasingly capable of committing homicide without
allowing his victims blood into his own body. When the vigil is sus-
pended a few months later, the widower continues to tighten in anticipa-
tion of avenging his wife's death by secretly killing with peace sorcery or
poison the wife or close relative of the suspected murderer or the person
who hired the murderer. Often, it is said, a widower will kill a woman of
an enemy tribe so her relatives will feel the sadness and pain he does.
Because only relatively few men know the secrets of peace sorcery, poi-
son, and other forms of ritual killing, most men once they become wid-
owers must enlist the aid of a practicing peace sorcerer or other specialist.
In return for this service, widowers must apprentice themselves to the
specialist for a considerable time; or at least this is what others in the
village suspect. Some people go so far as to say that all widowers become
peace sorcerers for a time before they are freed from mourning and
remarry. Being a woman, a widow cannot tighten, but she does close her
body, as do postpartum mothers, peace sorcerers, and widowers. It is
significant in this comparison that the roles of widow and widower corre-
spond with the homologous categories of postpartum mother and father,
respectively (see Chapters 4 and 5).
Assisted by the male laborers and the widower, the male blood rela-
tives of the deceased keep a nightly vigil over the grave until the blood
and flesh of the corpse have rotted away, dried, and dissipated. This may
take as long as two or three months. The vigil is maintained to prevent
peace sorcerers from robbing the grave, for it is in this way that peace
sorcerers are said to acquire the nefarious hot dirty bloody things essen-
tial to their arts. However, the deceased's male relatives and in-laws are
not simply concerned with thwarting the practice of peace sorcery gener-
ally. Rather, they are protecting all blood relations of the deceased, in-
cluding themselves, from collectively becoming the victims of a particular
category of peace sorcery that, in its techniques, is similar to mefu de-
scribed in Chapter 3 (cf. Hau'ofa 1981:221-5). To work mefu upon his
intended victim, it will be recalled, the peace sorcerer need only obtain a
specimen of the victim's bloodless dirty leavings (an article of clothing,
areca spittle, feces, cigarette stub, hair, etc.) to control his or her soul.
The peace sorcerer can kill the soul of his victim whose leavings he has by
mixing them with hot bloody residues of corpses. But because his victim's
158 Quadripartite structures

own leavings are bloodless, they are not hot for killing anyone else. It is
only when the sorcerer acquires the bloody remains of a dead human that
others are threatened, particularly the persons related by blood to that
dead human. Thus, with comparative speed and ease, a peace sorcerer
can decimate an entire clan or network of blood relations if he can obtain
some blood from the dead body of one of them. With the blood of my
dead brother, for example, a peace sorcerer has the capacity to kill me.
My blood and my brother's blood are one. Under the peace sorcerer's
direction, the spirit of the deceased forces his/her survivors to mistakenly
ingest some of the hot blood on their skins that has not been washed off
since the death, and one by one they will die. Whenever many people of a
clan or village die within a short period of time, as apparently happened
during the early contact era when villagers were exposed to many newly
introduced European diseases, this is among the prevalent explanations
(Mosko n.d.). The rationale of the grave-site vigil, then, illustrates that
blood out of the body of the deceased, unlike cold bloodless excreta of the
living, is conceptualized as blood out of the collectivity formed by the
bodies of the surviving relatives of the deceased. Therefore, it is essential
that this shared blood be protected from theft or manipulation that could
generate further death.
The grave-site vigil, as already mentioned, is abandoned when the flesh
and blood of the corpse are thought to have dried and disintegrated.
Afterward, nothing is left of the deceased's body that can be used by
peace sorcerers to exterminate the deceased's blood relatives. Only the
bloodless bones remain. Ideally, the abandonment of the vigil should also
coincide with the staging of the first of two death feasts - the katsiamore,
or 'burial feast. "7
Performance of the burial feast follows very closely the general outlines
of the subsequent final mortuary feast (kumau or umupua). In both cases,
the peace chief of the deceased's subclan gives a prescribed variety of raw
plant foods and meat to his friend or friends. The peace chief who receives
these foods distributes them for consumption among the people affiliated
with his subclan. To the feast givers or owners of the death, the food
represents the blood or flesh of the deceased. At both the burial feast and
the final death feast, the giving away of the deceased's blood in the form of
food releases some of the survivors from mourning. Moreover, death
feasts of both kinds are reciprocal between friends. When the feast receiv-
ers subsequently have deaths themselves, they are obligated to return
identical quantities of the same categories of food to their friends who
gave them a feast earlier. In short, the final mortuary feast, which puts an
end to mourning for those who have died, replicates the overall design of
the initial burial feast. Therefore, it is unnecessary for me to describe
them both in equal detail. Inasmuch as the ritual and symbolism of the
 Feasts of death (i) 159
final death feast are more elaborate than the burial feast, I shall concen-
trate on interpreting it at length, keeping in mind that my comments,
unless otherwise indicated, are also applicable to both. In this way, I hope
to avoid undue repetition.
Nevertheless, it will be worthwhile to briefly mention those features
that are unique to the burial feast and differentiate it from the final death
feast discussed below. Planning and preparation for the burial feast begin
as soon as the body has been set in its grave beneath the village abdomen
or under the house of the deceased. None of the deceased's blood rela-
tives may exert themselves to provision the feast. Under the authority of
the deceased's subclan peace chief, the female laborers are sent to the
gardens to work and thereby provide enough plant food for the owners'
burial feast. The male laborers are sent hunting and fishing each day. The
meat they bring back from the bush is butchered and smoke-dried on a
stick platform over a low fire to preserve it for the burial-feast exchanges.
The male laborers are also charged with bringing the necessary firewood
to fuel this fire, and a few of them watch over it at night with some of the
male mourners as part of the grave-site vigil.
Laborers explain that they work, hunt, and fish to provision the burial
feast in order to make the mourners or owners of the death happy (en-
gama), to release them from the sadness and pain of death. Recall that the
term engama is used also for the notions "to begin" or "start' as well as
"to conceive" (see Chapter 5). However, only some of the owners are
relieved at the end of the burial feast. After release, these ex-mourners
may resume work and other activities preparatory to the final death feast,
at which time the remaining mourners will be similarly absolved of their
funerary observances and made happy. The determination of who leaves
and who remains in mourning after the burial feast will be discussed later
in the chapter.
Prior to the scheduled day of the burial feast, the officiating peace chief
does not send invitations to his friends as he will do at the final death
feast. Thus, unless a friend of the owner lives in the same village or is
present in the role of laborer or owner himself (through papie ngaunga
relationship to the clan of the deceased), the principal feast-receivers will
not be present. In any case, because the friends of the feast-giving peace
chief are not formally invited, the burial feast is marked by the absence of
the extravagant and colorful singing, dancing, drumming, and courting
that characterize the final mortuary feast. Thus, the prohibitions upon
singing, dancing, colorful dress, and courting that were immediately im-
posed on the village with the discovery of death continue after the burial
feast is over up to the performance of the final death feast. The village, in
other words, remains cold.
Just in terms of sheer numbers, the large majority of the deceaseds
160 Quadripartite structures
surviving blood relatives are released from the more rigorous mourning
observances once they carry the burial-feast prestations to their friends at
their friends' villages. These owners and all the laborers may now pick up
their lives pretty much where they left them before being transformed by
death. If they came from other villages, they are allowed to return home.
They may work, hunt, fish, and engage in other activities for the sake of
their own households, but the released owners particularly must begin
also to prepare for the final death feast by planting extra gardens, per-
haps, or breeding pigs and dogs.
If the same clan experiences another death in the meantime, its mem-
bers and papie ngaunga affiliates and all their spouses must gather once
more as owners and laborers and stage another burial feast. This points to
one of the main features that distinguishes the burial feast from the final
death feast. A burial feast should be given for each death, but a single final
death feast may be given with respect to several deaths in the same clan
or subclan. A series of deaths and burial feasts typically precedes the
exchange of one final death feast.

The final death feast: village cold and village hot

It should be clear already that peace chiefs play a prominent role in all the
activities associated with death, mourning, and release of mourning
through reciprocal feasting. Turning now to the final death feast, it will be
useful to begin with a brief account of the peace chief's focal position in
this context, for indeed peace chiefs are described as the "owners of
death" (mae ngome aunga).
At feasts, peace chiefs are assisted in their work by a number of func-
tionaries: "string-giver" or "assistant chief" (uibina); "knife" or "food
server" (atsiua); "bossman, ' who relays orders from the chief to the la-
borers; and the peace sorcerer, who advises the peace chief regarding
feasting traditions and punishes those individuals who disobey the chief's
legitimate wishes. All these functionaries derive their authority from that
of the peace chief (Hau'ofa 1971:162-4). But peace chiefs are not merely
the owners of death as regards mourning and feasting. According to the
culture of the Bush Mekeo, all deaths are ultimately traceable to the
authority of peace chiefs, and through them to the culture hero Akaisa.
Peace sorcerers, war chiefs, and war sorcerers must ideally obtain the
sanction of peace chiefs to do their killing. As the nominal owners of clan
lands, peace chiefs serve as the intermediaries between ordinary human
villagers and extraordinary beings of the bush - the ongokapu bush peo-
ple. And when the latter threaten any of the former with illness or death,
it is left to the peace chief to intervene. Thus, although peace chiefs
should never themselves kill other human beings, the agents of death
perform their tasks with the explicit or tacit approval of peace chiefs.
 Feasts of death (i) 161
Therefore, the right and responsibility to perform death feasts on behalf
of his subclan is only a part of the peace chiefs role as an owner of death.
As the exchange of the burial feast draws near, the peace chief must
decide who among the deceaseds blood relatives will be released from
mourning and who will continue to mourn as either oaifa or akemoku
official mourners until the final mortuary feast is completed.
For every death, one or more persons must be appointed oaifa. Oaifa
must be married, or at least have been married. The peace chief may
select any of the deceased's blood relatives, and he need not restrict his
choice to within the clan. For example, if the deceased is a woman, her
son or daughter belonging to a different clan may be chosen to serve as
oaifa. Ideally, the chief should choose from among the closest surviving
relatives (e.g., parents, sons, daughters, full siblings, half-siblings, first
cousins).
Oaifa can be of either sex, but it is my distinct impression that women
are chosen more often than men. Villagers themselves often speak of oaifa
in the generic as women, and, although there is no rule preventing men
from becoming oaifa, the other category of official mourner, akemoku, is
reserved exclusively for male blood relatives of the deceased. The dis-
tinguishing mark of an akemoku mourner is that he is released from
mourning along with the others at the burial feast except that he must
allow his beard (buibui) to grow. Men who are akemoku must keep their
beards long until the final death feast, at which they contribute one pig
each to the officiating peace chief for him to give away. The feast-receiv-
ing peace chief will then remove the injunction against depilation. There-
fore, one way that feast-giving peace chiefs guarantee there will be
enough village pigs for their feasts is by appointing enough men to serve
as akemoku. Because women cannot grow beards, and their pigs (or por-
tions of the pigs they jointly own with their husbands) cannot be used in
their own clans' feasts, women cannot become akemoku but oaifa only.
Oaifa mourners, by contrast, are not required to donate a pig for their
own release at the final death feast. In addition, akemoku men, like ex-
mourners after the burial feast, may regularly wash their bodies and hunt,
fish, and work in preparation for the final death feast, which releases the
predominantly female oaifa. Thus, men as akemoku contribute signifi-
cantly toward provisioning the final death feast, which they could not do if
they instead were appointed oaifa. This gender bias implicated by the
selection of akemoku and oaifa will have some relevance to the interpreta-
tion below of the final death-feast exchanges.
The mourning ordeal of oaifa deserves special comment. Although
oaifa are neither characteristically blamed for the death of their relative
nor beaten, their experience is much the same as that of the widow or
widower. They must speak only in whispers; eat so as not to touch their
162 Quadripartite structures

food with their dirty hands; smear charcoal over their bodies; refrain from
washing; shave their heads; wear black, woven arm-, leg-, and waist-
bands; wrap themselves in black barkcloth blankets; remain hidden in the
dwelling of the deceased near the grave; do no work; abstain from sex; and
wear skin dirt (fa iofu) relics about the neck. These relics (not worn by
akemoku after the burial feast) contain cold bloodless bodily residues
taken from the corpse before burial (e.g., loincloth, fibers of grass skirt,
string bag, limepot, hair, nail parings, teeth).
Finally, the one feature that distinguishes oaifa mourners from wid-
ows, widowers, and akemoku concerns the practice of bafu, "food absten-
tion." Bafu is distinct from ritual male tightening ngope and from post-
partum memengoa procedures. Oaifa mourners practicing bafu eat mod-
erately in most respects. Bafu involves the strict prohibition of just one
type of food, e.g., banana, coconut, sweet potato, wild pig, cassowary,
wallaby, and so on. 8 At the time of appointment, each oaifa mourner
chooses as his or her bafu one food customarily shared with the deceased
during the latter's lifetime. If, for instance, I as a good spearsman always
gave my brother mullet to eat, when I become oaifa upon my brothers
death I will not eat mullet. Mullet will by my bafu. Ideally, no two oaifa
mourners will choose the same bafu with respect to the same deceased.
What the deceased lacks in death the oaifa collectively lack as well. As
they abandon all other signs of mourning upon their release at the final
mortuary feast, oaifa may resume eating the food that had been bafu.
Therefore, along with the wearing of the skin dirt bundles and the prohi-
bition against washing, bafu serves to identify each of the oaifa mourners
personally with the death of their blood relative.
In some respects, final death-feast preparations are begun with the
performance of the burial feast. The majority of blood-related survivors
are released from mourning at the burial feast with the explicit objective
of allowing them to work toward completion of the final death feast,
thereby ridding the clan of death entirely. Before he gives them leave to
return home, the officiating peace chief will approach the relatives of the
deceased who came from other villages and who are known to own pigs
and dogs, and he will ask them to raise an animal specifically for the final
mortuary feast he is planning. Those persons who remain at the subclan's
natal village must also take steps to increase their herd. Because the
burial feast has significantly depleted their garden resources, old gardens
must be enlarged and new gardens planted so there will be enough plant
food to feed themselves, their pigs, and their many anticipated guests.
Thus, a great deal more plant food is cultivated following a burial feast
than would be required merely for subsistence.
The subclan peace chief possesses the authority to impose "markers"
(onge apua)9 on extra gardens planted at his direction and on the coconut
 Feasts of death (i) 163
and areca palms owned by members of his subclan, preventing their
exploitation until the feast is ready. If the subclan suffers another death
during this interim period, yet another burial feast must be performed
and additional oaifa and akemoku mourners appointed. The extra provi-
sions of garden food, pigs, and dogs set aside for the final death feast may
be thereby exhausted, so final-death-feast preparations must be reiniti-
ated. A clan or subclan often experiences such a setback between the time
one burial feast is completed and resources accumulate sufficiently to
hold the final death feast.
When the feast gardens have reached maturity and there are enough
fully grown village pigs and dogs, the peace chief will send for women of
his subclan living in other villages and the subclan's papie ngaunga affili-
ates. These people will return with their spouses - the laborers or in-laws
- to the village of the owners. As the peace chief and his functionaries
dictate, the laborers will work for the feast owners. Now, however, the
owners, excepting the oaifa mourners still in confinement, can join with
the laborers in some of the preparatory activities. This is particularly
evident as regards men and the provision of meat from the bush. As the
most senior member of his subclan and nominal owner of the land, the
peace chief organizes his clanspeople, the papie ngaunga, and the la-
borers into large cooperative hunting and fishing expeditions into the
bush (Seligmann 1910:292-3; Williamson 1913). At the appointed spot,
the chief begins the drive by ritually appealing to the ancestral clan spirits
including the recently deceased, the culture hero Akaisa, and Akaisa's
son, Tsabini, to chase many animals into their nets. Several such hunting
and fishing forays are usually necessary to procure enough bush meat for a
final death feast.
Otherwise, male and female laborers are called upon to perform the
same tasks they did for the burial feast. The women clear the village
grounds, procure garden food, cook and serve food to the men, and, along
with the men, pound sago. The men for their part construct new sitting
platforms to house the guests, make needed repairs on the chief's club-
house and other structures, gather coconuts and areca nut, supply fire-
wood, and attend to the smoking of the accumulated bush meat.
While so engaged, the laborers are fed several times each day some of
the food they have procured from the bush. By Bush Mekeo standards,
they eat exceedingly well.10 Also, they are regularly given areca and betel
to chew. These observations bear upon the conceptualizations of the
relations between the owners and the laborers.
The laborers must never complain or hesitate in their work, nor may
they violate the rules of eating on the chief's clubhouse (no smoking or
chewing while others are eating, no throwing rubbish over the railings,
etc.). If a laborer does do one of these things, he or she and all the others
164 Quadripartite structures

of the same sex will have to compensate (kaua) the officiating peace chief
or his functionaries by paying a fine. The work laborers do on behalf of the
owners is also termed compensation (kaua), and both work and fines are
likened specifically to marriage compensation (kaua). This suggests that
the laborers, who are the owners' affines, are of different ancestral bloods
from them.
Although the work and fines contributed by laborers are privately
viewed by them as burdensome, an atmosphere of "play" or "fun" (bakau)
tends to prevail publicly. Linguistically, the root for compensation (kaua)
and play (bakau) is the same: kau. Villages have a proverb that reads,
"When they compensate (or pay), they play" (isa kekaua aisama isa ke-
bakau). Laborers' paying and playing, once more, make the owners
happy.
When all the laborers are gathered prior to feasts, a special form of
paying and playing is staged. It begins when, absentmindedly or other-
wise, one of the laborers casually expresses a desire for some particular
kind of food. The owners take this to imply the laborers think they have
not been entirely well fed. The next day is set aside for all the laborers to
"eat mullet" (or whatever food had been mentioned). In the morning, the
laborers set off to collect and bring back enormous quantities of the
indicated food. Either seated on the village abdomen or on their plat-
forms, the laborers are ordered to eat all the food piled before them after
it has been cooked. Collectively failing to do so or vomiting in the at-
tempt, the laborers will have to pay substantial fines of shell, feather, or
dogs' teeth valuables to the owners. By means of this and other play
diversions, it is argued, the laborers offer the owners occasion to forget
their sorrows temporarily and be happy.
In this particular kind of playing and paying, the laborers of both sexes
assume a role reminiscent of the newlywed, engorging bride. They are
obliged to eat all the food they are given, and it seems that they are only
given procreative varieties and never contraceptive ones. Because the
laborers are all married, ideally their own relatives have exchanged mar-
riage compensation with the owners to express the lack of shared an-
cestral blood, just as brides are different blood from their grooms. More-
over, the laborers make the owners happy (engama), just as brides partly
through engorging food conceive (engama) members of husband's clan
(i.e., their offspring). Quite clearly, I think, this playing between owners
and laborers symbolically expresses the relation between own clan and
groom, on the one hand, and spouse's clan and bride, on the other.
As the date for the feast approaches, the laborers are directed to make
the last touches on constructing and repairing platforms for the visitors
and generally getting the village in order for their arrival. Collective
hunting especially intensifies so there will be enough bush meat. The
bossman will dispatch small parties of male laborers to neighboring vil-
 Feasts of death (i) 165
lages to retrieve the pigs and dogs that will be contributed by owners
resident there. Ideally, in the case of a male deceased, the village pigs to
be slaughtered are those that were raised by him or by his full brothers
and sons. Because these are usually exhausted at the burial feast, the pigs
of akemoku and other male relatives (parents, first cousins, etc.) of the
deceased are substituted. The peace chief should provide the dogs him-
self, but he can also request a pig or dog from any other married adult
male related to the deceased or to the subclan or local clan of the de-
ceased. In any case, it is up to the owners of the animals to decide
whether the latter will be killed and given away on any particular occa-
sion. Usually, the peace chief has negotiated for rights to pig and dog
meat well in advance of the death feast, when the animals are still
immature.
When all is ready for the final mortuary feast, the peace chief invites his
friends (pisaua) by sending his assistant peace chief to each of them with
the carcass of one of his dogs11 and a branch of areca nut. After his wife
has cooked the meat, the invited peace chief serves it to the men of his
residential clan on the clubhouse and distributes the areca nut among
them afterward. He does not himself consume either, however. The areca
nut and dog are both symbolic of the office of peace chief; the areca nut
has the added significance of sexuality and procreation (see Chapter 5).
Kofuapie peace chiefs in the moiety opposite that of the feast-giving peace
chief who are not friends of the latter will not be formally invited to come
to the feast, but they will come nonetheless as part of the feast-receiving
retinue led by the peace chief(s) who have been formally invited.
When the appointed day arrives, separate parties of friends and
kofuapie guests from the rest of the tribe led by their peace chiefs gather
in the bush along the paths leading to the cold village of mourning.12 The
men and women who are to dance at the feast richly adorn their bodies
with reddened oil, face paint, flowers, feathers, boldly colored barkcloth
and grass-skirt garments, and valuables. In addition, the men carry drums
and hide their courting charms underneath their tightening belts. Once
fully adorned, though, the guests do not immediately enter the village
where the feast is to be held. The first friend of the senior feast-giving
peace chief must be the first to enter with his people, and he must be
given compensation (onge kau) for his effort. If there is more than one
senior chief of the same residential clan making a feast together, their
respective first friends may both await payment of onge kau. Onge kau
can only be paid in the form of smoked or dried bush meat. When the
host peace chief hears that his friend is in the bush, he sends him one or
more joints of meat, admonishing him to come and make his village hot
again. Singing, drumming, and dancing, the first friend advances several
yards until his clanspeople go no farther. The peace chief must listen to
his people, for it is they who will eat the compensation meat and have to
166 Quadripartite structures

provide it again to their chief when it is their turn to make their own final
death feast. With each payment, the party of the first friend advances
closer and closer to the village until the people do not hinder the advance
of their peace chief.
Eventually, the first friend arrives at the end of the village opposite that
of the feast-giving chief's clubhouse. If the first friend happens to live in
the same village, he enters from his own end, and his party explosively
passes into the central village abdomen. They sing, dance, and drum at
highest pitch, proceeding slowly to the opposite end of the village and
back again. Those who do not dance - children, the elderly, and others
who do not choose to - make for the homes of their relatives resident in
the village or for the platforms constructed to house them. There they
watch the spectacle. Other friends and kofuapie groups now begin to
arrive and join the resplendent mass. Each time the moving crowd in the
abdomen swells, a new crescendo is reached. As Seligmann observes, the
widows, widowers, and oaifa still in mourning are saddened by the noise
(1910:360). Nonetheless, the boisterous singing, dancing, and drumming
continue throughout the remainder of the day, all night, and into the next
day until such time as the feast-giving peace chief begins to make his feast
prestations.
The arrival of the guests renders the village of the mourning clan hot,
where before, immediately following death, it was made cold. This de-
serves some additional comment, for dancing, singing, and drumming at
final death feasts entail important meanings that have not been reported
by other observers.
Married and single people of both genders participate. The designs for
colorful dress and ornamentation of the men and women are hereditary.
Men of the same clan (or subclan) are said to appear indistinguishable
from one another, but men of different clans are visibly distinct. Single
girls and married women also adorn themselves with the hereditary insig-
nia of their clansmen or husbands, respectively, but their personal identi-
ties are not visibly concealed from the men. The hereditary and visible
accoutrements, not to mention the skills of singing and dancing, are
regarded as sexually sweet to people of opposite genders in clans of
different moieties. Hereditary clan emblems or insignia (kangakanga and
auafangai; see Seligmann 1910:320-34) and styles of ornamentation pro-
vide clues to the participants as to who is and who is not a potential sex
partner or mate. The themes of the songs (pike) deal with love and
courtship. Indeed, some of the songs are love spells that have entered the
public domain. Other love or courting spells are secret and are uttered
silently by the men as they dance and stare into the eyes of their para-
mours. All that distinguishes fellow clansmen from one another to the
women, aside from their individual skills in singing and dancing, are the
 Feasts of death (i) 167
hot love charms and spells that are not visible but are nonetheless at play.
Furthermore, each man's face is marked by a black line drawn across the
bridge of his nose or down his cheeks. This is the only black detail of dress
on either men or women. This black paint is mixed from the fluid of the
man's love charm. Thus, a man's glance transforms or "changes the mind "
of his prospective lover and makes him sexually sweet to her rather than
unsweet. Another technique to the same effect involves putting a trace of
charm fluid on a woman's grass skirt as she brushes by.
All the while this goes on, lines of women and men from different clans
weave in and out of one another as the whole assemblage moves back and
forth along the village abdomen in an apparent simulation, I shall argue,
of the sexual act, and particularly the mixing of procreative bloods
(Williamson 1913:281-5). The Bush Mekeo themselves regard this danc-
ing and singing at final death feasts as the most sublime expression of
collective happiness there is in village life. Once again, the term for
happiness, engama, is used also for the notion of procreating new life in
sexual reproduction. By their presence and explicit ritualized reenact-
ment of the sexual act, the friends and other guests represent societal
procreative hot, happiness, new life, and reproduction in contrast to the
procreative cold (i.e., contraception), sadness, and death of the feast-
giving owners.
As the dancing continues, more subdued and highly formalized ac-
tivities are underway on the clubhouses of the resident peace chiefs. The
kofuapie peace chief resident in the same village as the owners entertains
the other peace chiefs and men of their moiety who are not dancing or
serving as laborers. At the order of the feast-giving peace chief, the
female laborers take them bowls of cooked food to eat.
On the clubhouse of the feast-giving peace chief, peace chiefs and
peace sorcerers, together with knowledgeable old men of the same resi-
dential and dispersed clan, gather to advise him on how he should pro-
ceed. These men are collectively known by the honorific title au akaisa
("Akaisa Men"; Chapter 8). "Ordinary" or "poor men " (ulalu) of the feast-
giving clan are not entitled to participate in these important delibera-
tions. Sometimes at feasts, the Akaisa Men are called upon to adjudicate
grievances and trouble cases to which the occasion of the feast has given
rise, especially, it seems, when a new peace chief is to be installed and
conflicting claims of legitimacy are expressed. The presence of the peace
sorcerers in the village during the feast, moreover, contributes toward
making the village of the mourning clan hot.
The final death feast actually consists of two ceremonies. The first to be
performed is termed "body" (kumau). Body exchanges are followed by
"charcoal" or "black carrying/wearing" (umupua). The body ceremony
replicates most of the details of the burial feast, but charcoal carrying is
168 Quadripartite structures
unique to the final mortuary feast (cf. Hau'ofa 1981:169-83; Seligmann
1910:359-62).
The officiating peace chief signals his intent to begin the body proceed-
ings by sending to his first friend at the other end of the village additional
joints of smoked bush meat (onge kau) with the request that he club to
death the village pigs and dogs that are to be given away. When his own
clanspeople are satisfied that he has received enough bush meat, the first
friend passes through the abdomen of the village to the front of the feast-
giving chief's clubhouse, where the animals are staked to the ground.13
The dogs are passed onto the clubhouse, and the pigs are carried behind
or beside it. Laborers carry the largest of the slain pigs, called the "dirty
pig" (iofu kuma), to the steps. The owning peace chief now performs the
body ritual proper. He rubs blood from out of the dirty pigs mouth onto
the skin-dirt relics of each of his subclan's deceased, which have been
worn by the oaifa, widows, and widowers. This identifies the blood and
flesh of the dirty pig with the bodily dirt of the deceased and similarly
with the persons of the mourners who have kept these things and re-
frained from washing their skins. Afterward, the bloodied skin-dirt arti-
cles can be either burned or merely cast away in the bush;14 they are
rubbish now.
The dirty pig, however, is taken behind the clubhouse and set next to
the other pigs. Here the male laborers singe off the hair of each animal,
cut out the anus and colon for the female laborers to wash in the river,
wash the skin, and begin butchering. First, and very importantly, a large
square or patch of skin and fat is cut away from the back of each village
pig, and only then is the rest of the body dismembered. The head,
backbone, and patch of tailskin of the dirty pig are left attached as one
piece. This will be given specifically to the peace chief's first friend. Each
of the other friends of the feast-giving chief will also receive a similar
compound joint from another pig (head, backbone, tailskin). Remaining
village pigs are fully dismembered.
Only some joints from each pig will be used by the peace chief to make
his feast. A husband and wife share ownership of their pigs. The ventral or
"lower" half of each pig (comprised of lower jaw, chest, one foreleg, and
one hindleg) belongs to the wife. These parts cannot be given away for her
husband's feasts. When one of their pigs is killed for a feast, she gives her
half to her own brothers, and they share it among themselves and mem-
bers of her clan. The dorsal or "upper" half (head, backskin, hindskin, one
foreleg, and one hindleg) belongs to the husband. He negotiates with his
peace chief over which parts he wishes to keep for his own personal
exchanges15 and which parts the peace chief will give away on behalf of
the entire subclan. When a man gives his peace chief a particular piece or
pieces of village pig meat for the subclan's death feast, his peace chief is
 Feasts of death (i) 169
obliged to give back to him an identical piece or pieces of meat once their
friends reciprocate with a feast of their own. However, if the man is
akemoku at a particular final death feast, his peace chief is entitled to take
all the upper half of his pig and is not obliged to return it later.
The backskin, liver, stomach, and small intestine of each village pig
unquestionably belong to the peace chief. When the laborers have
finished butchering the pigs, these portions with all the others intended
for the peace chiefs exchanges are taken onto his clubhouse. Laborers
also carry to the peace chief the accumulated garden produce and smoke-
dried bush meat. With all the food now piled on the floor, the peace chief
begins to do his work under the watchful eyes of the peace sorcerers and
other Akaisa Men who must observe that all proceeds without mistake or
disturbance. This is the most solemn part of the body ceremony. Fes-
tivities in the village abdomen are curtailed, and no one may scream out
or distract the peace chief from what he is doing. First, he slices the
patches of backskin from the village pigs into wide strips. Assisted by his
string-giver assistant, the peace chief binds with vine a piece of small
intestine, liver, and stomach to each strip of backskin. These tied bundles
are termed iunge fanga. Only publicly installed peace chiefs and, in
somewhat different circumstances, peace sorcerers, war chiefs, and war
sorcerers may cut iunge fanga backskin and organs. Indeed, it is by
publicly grasping the hand of his chosen heir and directing him in the
cutting oiiunga fanga at feasts that a peace chief relinquishes his office in
favor of his successor. The other three types of Akaisa Men similarly
transfer their hereditary positions to their heirs, but not at mortuary
feasts. Next, to each iunge fanga bundle the peace chief ties the carcass of
one dog,16 and this unit is termed ikufuka. Each of the peace chiefs and
peace sorcerers in the opposite moiety (i.e., both friends and nonfriend
kofuapie) will be given one bundle of ikufuka meat.
Once the ikufuka bundles are tied and sorted, the officiating peace
chief begins to assemble the complete prestations he will give to his
friends. The male laborers by now have planted bamboo poles in pairs on
the village abdomen. Each pair of poles, termed kou17 according to Selig-
mann (1910:361), is joined at the top. 18 The peace chief steps down from
his clubhouse into the abdomen and, as the string-giver passes him the
food, he hangs it from the branch stubs left on the shafts of the bamboo.
Laborers encircle the area with sticks to keep village pigs and dogs away.
First, the string-giver hands the chief a bunch of each kind of plant food
(coconuts, sweet potatoes, sago, bananas, sugarcane, taro, etc.). Next, he
passes along dried joints of bush meat (preferably pig, but cassowary,
wallaby, and mullet may be included). Then the head-backbone-tailskin
piece of the dirty pig is hung from the apex of the kou, and to it, finally,
the ikufuka is tied. All this food is termed pange. It includes essentially
170 Quadripartite structures
the same categories of plant food and meat as are given away at burial
feasts. I shall delve deeper into the meaning of pange foods below.
The entire tribal population waits, looking on in silence until all is
ready with the kou. Then the feast-giving peace chief walks up and strikes
the kou with the knotted end of a coconut leaf as he calls out, "Kaiva,
yours." Next, he steps to another kou and repeats the act, calling the
name of his second friend, and so on for each of his remaining friends. If
two or more peace chiefs of the same residential clan are working together
and giving their feasts at the same time, each separately prepares and
gives food to his own friends in this way.
Then nonfriend kofuapie peace chiefs and all peace sorcerers are given
their food. As the string-giver hands his peace chief the remaining bun-
dles of ikufuka meat from the clubouse, the peace chief calls out the
names of the intended receivers. These peace chiefs and peace sorcerers
come forward and claim their ikufuka meat. They do not receive plant
food or extra outer meat, but their food, ikufuka, is nonetheless cate-
gorically pange. If both senior and junior peace chiefs of the same clan are
making their feasts together, the senior peace chief gives ikufuka to se-
nior kofuapie peace chiefs and sorcerers, and the junior peace chief to
junior kofuapie peace chiefs and sorcerers. Thus, the peace chiefs of each
feast-giving clan send one or the other category of body prestation, or
pange, to all their friends, kofuapie peace chiefs, and peace sorcerers in
the tribe.
The first friend of the feast-giving peace chief now performs the second
segment of the final mortuary feast, the umupua charcoal or black carry-
ing/wearing ceremony. Owners who were released from mourning at
previous burial feasts (including the bearded akemoku) escort the official
oaifa mourners and the widows and widowers out of hiding and into the
circle of laborers standing on the village abdomen. The mourners appear
with charcoal smeared on their skin; hence, the name of the ceremony.
Each oaifa mourner carries a bit of the hafu food that heretofore had been
prohibited.
The first friend of the feast-giving peace chief stands in front of the kou
structure facing the assembled mourners, his own clanspeople standing
behind him with the pange foods emerging from their midst. Of the oaifa
mourners, he remarks that each man or woman has avoided one food
since the death of his or her relative. Admonishing the onlookers not to
shame19 this person when they see him/her subsequently eating this
food, the first friend takes the bafu food from the oaifa s hand, waves it
over his/her head, and places it inside the oaifa s mouth. The assembled
friends of the owners burst forth, shouting "Ahhhhhhhhhh." According to
this formula, the first friend lifts the barkcloth blankets away from the
oaifa, severs the woven mourning bands, dabs on red or yellow facial
 Feasts of death (i) 171
paint, and adorns them with flowers and aromatic leaves worn in court-
ing. With the replacement of each sign of sadness with one of happiness,
the gathering repeats the cry, "Ahhhhhhhhhh.''20
In identical fashion (excepting the removal oibafu), the feast-receiving
first friend relieves the widows and widowers of their mourning burdens
and announces that they may let their hair grow and remarry. Also, the
first friend cuts away the tuft of beard from the face of every akemoku
mourner so that henceforth they may depilate their bodies. Lastly, wid-
ows, widowers, and akemoku are given colorful signs (cf. Seligmann
1910:362-3).
And so the long mourning ordeal ends for the owners. Dirty relics of
the deceased of the feast-giving clan have been discarded, along with the
last traces of blood and other dirt from the skins of the survivors. Those
who were sad are now made happy, those who had died are now returned
to life. As the friends and other kofuapie on the village abdomen resume
their singing, dancing, drumming, and courting, the ex-mourners join in.
They celebrate with their friends and kofuapie through the night and into
the next day. The village, once made cold by death, is hot and full of life
again.
As soon as the charcoal-carrying ceremony is concluded, clanspeople of
the feast-receiving peace chief(s) take their food from the kou to their own
clubhouse. Clansmen's wives are apportioned meat, which they immedi-
ately cook. The wife of the receiving peace chief will herself cook the
ikufuka meat. Wives of other peace chiefs and peace sorcerers cook the
ikufuka meat their husbands received. The next day, after the festivities
are over and everyone has gone home to their own village, the various
peace chiefs and peace sorcerers distribute all the food among the mem-
bers and affiliates of their respective residential clans.
The principal feast-receivers - that is, the friends of the feast-givers -
give some meat and plant food to every adult married member of the
residential clan and to their married papie ngaunga, excepting those who
were also among the owners of the feast. These people will share this
pange feast food out among the members of their respective households.
Ikufuka meat, comprised of iunge fanga village-pig backskin and organs
and dog, will not be consumed by villagers in their houses. Every peace
chief and peace sorcerer who received ikufuka will serve it to the married
male members of his own residential clan and their married male papie
ngaunga on the clubhouse. But peace chiefs and peace sorcerers will not
themselves eat any ikufuka (see Chapter 8). In general, then, just as feast-
givers include clanspeople of the deceased and their papie ngaunga,
feast-receivers include the clanspeople of the friends and nonfriend
kofuapie of the feast-giving chief and their respective papie ngaunga
(Figure 7.1).
172 Quadripartite structures

papie papie
ngaunga ngaunga
of clan of of feast-
deceased receiving
friends

feast-givers feast-recievers
(ngome owners) [pange eaters)

i
Figure 7.1. Mortuary feast-givers and -receivers.

Feast foods, bloods, relations, and de-conception

I have already mentioned that pange feast foods exchanged among friends
at burial and final death feasts are said to be the blood and flesh of the
deceased and, thereby, of all the owners of the death. In other contexts,
the word pange means "losing" or "lost." As the blood of the deceased,
then, pange foods represent the "lost blood" of the feast-giving clan and
their papie ngaunga. Being such, pange foods are dirty (iofu) to the
owners and must not be eaten by them. Adult villagers informed me that
they had to keep a close watch on their young children who are among the
owners at feasts to make sure that they did not eat pange, for it could kill
them just as dirty things can kill anyone who eats them. And inasmuch as
the owners collectively are one blood with the deceased, pange repre-
sents their own flesh and blood.
Moreover, in keeping with the comments above, pange would seem to
represent specifically female as distinct from male blood or flesh. At the
announcement of death, it is predominantly women who contaminate
their skins with blood and dirt and who recapitulate in mourning obser-
vance the features of the postpartum memengoa. It is women also who are
characteristically selected to serve as oaifa. During the final mortuary
feast, it is inside the female part of the village - the abdomen - that the
feast animals are killed, that blood of the dirty pig is merged with the skin
dirt of the deceased and the survivors, that pange foods are transferred,
and that dirt and other signs of death are removed from the mourners'
bodies. Also, as I have shown in numerous contexts, the abdomen is
systematically associated with the specifically female notion of mother.
One of the curious features of Bush Mekeo death feasting is that, to the
feast-receivers who actually eat the food, pange is not considered dirty.
Rather, to them it is not pange; it is clean and edible food once it has been
cooked. This evaluation of pange foods recalls in a general sense the
contradictions whereby semen can generate both death and new life and
 Feasts of death (i) 173
certain cognatic blood relatives can marry and reproduce as nonrelatives
(Chapters 4-6). Specifically, persons in friend and kofuapie relationships
who may exchange blood through intermarriage and heterosexual re-
production also reciprocate one another's blood over death. The sim-
ilarities here between feasting, reproduction, and marriage are, I think,
not superficial or coincidental but representative of homologous category
distinctions I have described throughout Bush Mekeo culture.
It is worth noting first that there is a total of four categories represented
among the givers and receivers of each death-feast exchange. Also, the
givers and receivers at any one death feast include members of all four
clans within the tribe. Although some members of each patrilineal moiety
are givers, other members of the same moiety will be receivers. More-
over, it is not always the case when some members of a particular clan (or
subclan) can eat pange food that all members of the same clan (or subclan)
may do so at the same time. The rules concerning who may and who may
not eat a particular pange prestation, then, do not always coincide with
the divisions separating moieties and clans. Nonetheless, the relations
between the four feast-giving and feast-receiving categories are homolo-
gous with the relations between clans as represented in marriage-com-
pensation exchanges.
The clan of the deceased and their papie nguanga belonging to the
clans of the opposite moiety are owners or feast-givers together. Persons
belonging to neither of these categories may eat the pange feast food. The
feast-giving papie ngaunga may not eat the food their own peace chiefs
receive for their clans. Moreover, it is precisely among the two categories
of feast-giver that marriage is prohibited. The two kinds of feast-giver,
then, are comparable to the relation of own clan : mother's clan.
Relations between the two categories of feast-receiver are analogous
with those between the two categories of feast-giver. There are, first,
members of clans in the moiety opposite to that of the deceased (i.e.,
friends) who are not papie ngaunga of deceased's clan, and, second, their
papie ngaunga who are not among the clan of the deceased. The two
categories of pange feast food eater are also prohibited from marrying
among themselves. Indeed, according to the rules of marriage, feast-
givers and feast-receivers who belong to opposite moieties are the pre-
scribed or preferred spouses of one another. Feast-receivers, then, or the
eaters of pange feast food, are the spouses and laborers of the two catego-
ries of feast-giver. They eat pange food that they receive through the
peace chiefs of their own or mother's clans comparable with the two clans
of marriage-compensation eaters.
Therefore, from the perspective of the clan of the deceased, the rela-
tions between all categories of feast-giver and feast-receiver are homolo-
gous with distinctions between intermarrying clans:
174 Quadripartite structures

papie-ngaunga
feast-receiving . of feast-receiving
i r r i i i papie ngaunga r . T , ,
clan oi mends; laborers :: c ^ r mends; laborers /r,nX
j j : r i r oi clan oi : r . (21)
deceased tor clan oi , , lor papie ngaunga
, j deceased r * C
deceased of clan oi
deceased

own clan : spouse's clan :: mother's clan : , , , (16)

mother s clan
clan A1 : clan B 1 :: clan B 2 : clan A2. (19)

Each category of feast-giver contributes food that is presented both to

members of the moiety opposite to it and to members of the same moiety.
In terms of the above homology, feast-givers jointly give food to their
respective spouse's clan and spouse's mothers clan. Feast reciprocation,
in other words, is isomorphic with the structure of marriage-compensa-
tion exchange.
To specify more precisely the meaning of feasting among the Bush
Mekeo and its relation to marriage compensation, it will be necessary to
delve deeper into the symbolism of pange feast food itself. How exactly
does pange represent the blood or flesh of the deceased and/or the sur-
vivors? What is the relation of these notions to the indigenous theory of
conception and reproduction at both the individual and societal levels?
Namely, how does the blood of death in pange prestations relate to the
bloods of procreation, kinship, clanship, and marriage exchange?
During burial feasts and the body ceremony of final mortuary feasts,
the plant foods included in pange are ideally taken from gardens planted
by the deceased before his or her death. These vegetables are the product
of the deceased's work and gardening nonwork in life. Had death not
intervened, these foods would have sustained the deceased's blood, flesh,
and body through further life. The plant foods in death feast pange, then,
represent the blood-elements of the deceased's "unlived life," that por-
tion of living the deceased had provided for him/herself before death
occurred.
Except for ikufuka backskin, organs, and dog, the remainder of pange
includes outer raw meat from village pigs and bush animals. Village pig
meat in pange is wet; the smoked bush meat is dry. 21 Village pig meat is
taken from the upper half of the animal, and bush meat is taken from the
lower half. Together, these joints of outer pange comprise the outer body
of a dead creature, half village and half bush. To each half, however,
attaches a slightly different meaning. At a burial feast, this outer body of
pange is that of the recently deceased. At a final death feast, one such
outer body represents the composite blood and flesh of all the people in a
subclan who have died since the clan performed its last final death feast.
 Feasts of death (i) 175

In either case, the giving away of this outer body releases from mourning
persons belonging to both categories of blood-related survivors.
The upper half of the pange meat must be taken from a village pig.
Ideally, meat of the lower half is composed of bush pig, but meat of other
large bush animals is sometimes substituted when there are not enough
bush pigs caught. Pig (kuma) is ideally suited for both purposes. The term
for pig (kuma) is also the term for body (kuma). Also, the set of categories
descriptive of pig bodies is the same for human bodies; that is, every part
of pig anatomy is assigned the same term for the homologous part of a
human being. This is not entirely so in all details for marsupials, birds,
fish, reptiles, and amphibians. Nevertheless, although some pigs inhabit
the village alongside human beings, other pigs forage wild in the bush and
lead dramatically different lives. Pigs of the village are fed cooked food.
Pigs of the bush eat their food raw. Villagers regard their own pigs "like
children." They require considerable care, attention, and affection. Vil-
lagers correspondingly liken themselves to "fathers" and "mothers" of
their pigs. Village pigs have personal names. Traditionally, Bush Mekeo
women suckled their baby pigs (Guis 1936:62). Even nowadays villagers
are extremely sad when the time comes to kill one of their own pigs, and
they cry. Village pigs even have their own clans (ikupu) auxiliary to those
of their human fathers. Bush pigs killed for feasts, on the other hand,
although they are taken from the land of the feast-giving clan and belong
to that clan, are not regarded as adoptive children. They are not suckled,
individually named, or mourned for; nor do they possess clans. Thus, the
two kinds of pigs approximate human beings to different degrees.
These and other contrasts between village and bush pigs correspond to
sexually assigned human attributes. The feeding and daily care of village
pigs is largely the responsibility of women. Only men hunt bush pigs.
Male village pigs are castrated at weaning, and as a result they become
like village sows in their maturity: Courting (bakai) and mating are un-
sweet to them, but eating is sweet. Thus, all village pigs, male and
female, become fat and "femaleish." Although eating is sweet to bush
boars, so also are courting and sex with female pigs. Thus, the bodies of
bush pigs, even bush sows, are comparatively much leaner than village
pigs and more "maleish."
At death feasts, the two kinds of sexually differentiated meat - femaleish
village pig and maleish bush pig - are exchanged between categories of
persons whose relationships traced through the ancestral bloods of the
deceased are correspondingly differentiated according to sexually distinct
bloods. Feast-givers give female meat to people related to them through
female blood (cognates, of spouse's clan), and they give male meat to
people related to them through male blood (agnates, of spouse's mother's
clan). With reference to Figure 7.2, group A1 (clan of the deceased) gives
 papie ngaunga papie ngaunga
of clan of deceased of feast-receiving
friends

spouse s
own clan spouse's clan mother's clan mother's clan

I
Figure 7.2. Mortuary-feast categories and clan identities.
 Feasts of death (i) 177
female village meat to B1 of the opposite moiety (spouse's clan), and bush
meat to A2 of the same moiety (spouse's mother's clan). Group B2 gives
bush meat to B1 (their spouse's mother's clan), and village meat to A2
(their spouse's clan). According to the indigenous theory of conception,
each individual is born with the four bloods of his/her grandparents, or
their respective clans. Marriage compensation creates the public fiction
whereby two of these cognatic bloods for the individual (i.e., father's
mother's or spouse's clan blood, and mother's mother's or spouse's moth-
er's clan blood) are acknowledged, but for the purpose of exogamous
marriage and reproduction are abrogated. At death feasts, symbolic
equivalents of these cognatic bloods of the deceased are returned to the
clans of their origin - back to spouse's clan and spouse's mother's clan
(Figure 7.3). What was provisionally done with marriage compensation is
finally done at death feasts. The ancestral cognatic bloods of the deceased
are separated, and the deceased as a human being is "de-conceived." For
the survivors and especially the descendants, the deceased's de-concep-
tion at feasts effectively permits the severing of their cognatic blood rela-
tionships with the feast-receivers and pange eaters as traced through the
deceased. In short, as death feasts de-conceive the deceased, they "re-
conceive" the deceased's blood relatives. The giving and receiving of feast
food redistributes blood among clans of the same and opposite moieties
that had earlier exchanged it in the conception of the deceased. Blood
transmitted through feasting reverses blood transmitted through sex and
reproduction (Mosko 1983).
Returning to my observations stated at the end of the last chapter
regarding the conflict of blood relationship viewed agnatically as well as
cognatically, the reciprocation of the two outer portions of pange feast
food has the most profound and significant of sociological implications.
Following from the rules of exogamy and the indigenous theory of pro-
creation, agnatic blood will be dissipated through the skin (women) of a
clan into other clans, including the other clan of the same moiety; and the
nonagnatic bloods of other clans will be correspondingly assimilated into
one's own clan unless there is some countervailing process of returning
and recapturing lost bloods. Through intermarriage, ideally closed clans
are cognatically open to one another's blood. This is true even of clans
belonging to the same patrilineal moiety. The purely agnatic constitution
of the clan and moiety is thereby compromised. Reciprocal feasting pro-
vides the appropriate, final, and lasting resolution. Once a person dies,
the survivors can send back his or her cognatic bloods whence they came,
thereby contributing to the closing of the boundaries between the clans of
both feast-givers and feast-receivers consistent with the ideology of ag-
natic purity within the clan and moiety.
178 Quadripartite structures

A1

the
deceased

Figure 7.3. De-conception and re-conception of grandmothers' clan bloods.

Happiness, conception, de-conception, and re-conception

This interpretation of the two outer portions of pange meat as bloods of
de-conception and re-conception sheds considerable new light on numer-
ous associated aspects of burial and final death feasts (cf. Hertz 1960;
Goody 1962; Hicks 1976; Weiner 1976; Huntington and Metcalf 1979;
Bloch and Parry 1982; Bloch 1982).
 Feasts of death (i) 179
The single term engama is used to convey the essentially similar ideas
of conceiving human fetuses with blood, de-conceiving the deceased, and
re-conceiving the blood relationships and blood-identities of the survivors
and the feast-receivers. It is significant in this context that all categories of
ex-mourners join the courting festivities immediately upon their re-con-
ception at the end of the final death feast. The nonsexual, nonreproduc-
tive, symbolically dead mourners have been transformed back into the
sexual, reproductive, living, and happy. Needless to say, this transforma-
tion occurs inside the village abdomen. There, it might be said, the entire
tribe is symbolically re-conceived in its essence, as an endogamous unit or
totality.
In a similar respect, constituent units of each exogamous clan within
the tribe simultaneously establish or reaffirm their integrity as one agnatic
blood vis-a-vis other clans through feasting re-conception. Individuals
and patrilines, it will be remembered, can be recruited to clans as a result
of either sexual reproduction and conception or reciprocal feasting re-
conception. Through food exchange and feasting, an adult-male nonag-
nate may join a clan by participating at feasts as though he were an agnate.
Villagers, in fact, justify or explain the presence of genealogical nonag-
nates among their clan members in just this way. Similarly, ancestrally
unrelated clan units (subclans, lineages) can fuse and become agnatically
one blood and one clan by working together in feast performance. By the
same token, residential clan unity and cooperation for feasting is de-
stroyed when clanspeople violate the rules of clan exogamy and intermar-
ry. In fact, the standard complaint against such marriages is that they
"spoil the feast." Clans, in other words, can recruit their members by re-
conceiving them at feasts with food just as they can by conceiving them in
heterosexual reproduction.
There is still the question of the appropriateness and meaning of using
raw meat in the two outer pange prestations. The feast-receivers them-
selves must cook their feast food before they may eat. In Bush Mekeo
culture, cooked food is symbolic of blood relationship. Agnates and cog-
nates should reciprocate and together consume cooked food on a daily
basis. They should never exchange raw food. Even the notion of "adop-
tive kinship" (ibafaka; literally, "made fat") implies blood relationship
through cooked food. At the final death feast itself, feast-givers feed their
non-yet-re-conceived feast friends and kofuapie guests with cooked food
before they give them raw pange. By contrast, when feast-givers give to
feast-receivers raw pange at the end of the feast, they are expressing the
re-conceived absence of blood relationship between them. And when the
feast-receivers eat the cooked pange foods among themselves, they of
course express their blood relationships to one another.
The raw outer meats that friends exchange at feasts are among the
contraceptive varieties prohibited to brides strenuously seeking concep-
180 Quadripartite structures

tion with their husbands. At death feasts, clans that have strenuously
reciprocated procreative blood to create cognatic relationships through
their female members exchange contraceptive blood in the form of meat
to negate these same cognatic relationships through women.
Viewed in this light, there is a certain appropriateness to the female
qualities associated with mourners, whether they are men or women
(Bloch 1982; Weiner 1976). Women, for example, tend to predominate as
oaifa. Women are the outer skin of the clan. Oaifa, widows, and wid-
owers alike allow their own blood and the bloody dirt of the corpse to
remain on their skins until they are re-conceived. In the meantime, these
bloods are on the outside or skin of the clan. It will be remembered that,
according to legend, the ancestors of present clans did not die, they just
shed their old skins and grew new ones. Just as female blood (i.e., womb-
blood) of a clan goes outside and into other clans through its skin or
women in sexual reproduction, female blood of death and mourning goes
back outside a clan and into other clans through its skin or women.

Feasting homologies
The terms I used at the end of the last chapter in the context of marriage
compensation are relevant here. At feasts, the members of the clan of the
deceased become one strictly agnatic blood to one another, and unequiv-
ocally inside the same clan. In relation to them, their feast-receiving
friends and spouses and feast-giving papie ngaunga are both outside the
clan of the deceased. But the latter are nonetheless one cognatic blood
with the clan of the deceased and, therefore, outside inverted in relation
to them; whereas the former are unambiguously outside and entirely
different blood in relation to the clan of the deceased. Lastly, the feast-
receiving papie ngaunga of the feast-giver's friends are one agnatic blood
inside the same moiety of the clan of the deceased, but this relationship is
everted through clans of the opposite moiety.
In these terms and the others I have noted above, the four categories of
feast-giver and feast-receiver are not only homologous with category dis-
tinctions among clans; they are also homologous with the other fourfold
conceptual discriminations that pervade the indigenous culture. I shall
list several of the more directly pertinent of these here:
. ., _, inverted everted ,.,
inside : outside :: ., : . •i (4)
outside inside
papie n aun a
clan of feast-receiving Z Z PaPie
j j : r . T :: of clan of : of feast-receiving (21)
deceased friends , , r. i
deceased mends

own clan : spouse's clan :: mother's clan : , , , (16)

mother s clan
agnatic non-blood cognatic cognatic . „,
blood ' (affines) " blood ' non-blood
 Feasts of death (i) 181

closing opening opening closing

(14)
male male female female
male female
semen : womb-blood : (13)
body blood body blood
roasting boiling : ripening : smoking (6)
blood sweet : unsweet : bloodless dirt (8)
bush village village bush
resources resources wastes wastes (2)
remote peripheral village adjacent
bush village abdomen bush. (3)

It is also significant that the ordered articulation of marriage and mortuary

exchanges in these terms parallels the processes of ritually opening and
closing the bodies of adult males and females. When men and women
procreatively open their bodies together, just as when clans simul-
taneously open their boundaries, there is life. When men and women
ritually close their bodies and clans feast, however, there is an absence of
life, or death; but death with the potential for creating new life.
One additional point follows along this vein. If the exchanges of blood
in feasting reverse earlier transmissions of blood between clans through
sex and reproduction in marriage, then these two spheres of interclan
relationship can be distinguished from one another in terms of extraordi-
nary and ordinary, respectively. With the conception and birth of its
members, a clan unidirectionally assimilates the bloods of other clans.
And with the death and de-conception or re-conception of its members, a
clan ambidirectionally returns the bloods of other clans. Although sepa-
rated in time, death feasting reverses conception and reproduction. It is,
furthermore, consistent with this reasoning that the exemplars of the
unidirectional marriage compensation (as the preliminary to reproductive
exchanges of blood) are ordinary villagers or poor men where peace chiefs
and other hereditary officials have no formal role, and that the exemplars
of ambidirectional feasting exchanges are extraordinary villagers, that is,
peace chiefs.
The following chapter will be expressly devoted to this and related
extraordinary aspects of death feasting among the Bush Mekeo.
 8
Feasts of death (ii): the sons of Akaisa

It has taken me this long in my interpreting of Bush Mekeo death and

mourning ritual to address the significance of the roles played by peace
chiefs specifically at death feasts and by Akaisa Men generally in the total
culture. Also, I have not as yet closely examined the definitive element of
the complete death-feast pange prestation - the ikufuka meat. This chap-
ter will be devoted to analyzing these issues, for, as I will presently show,
they are not only intimately connected with one another, but they relate
directly to the culture hero Akaisa and thus involve the most central
mythical symbols of all Bush Mekeo culture. Furthermore, the charac-
teristic structure of bisected dualities will become evident in this context
as in the others I have discussed.
It will be remembered that ikufuka consists of the intact carcass of a
dog and iungefanga (a strip of backskin and pieces of liver, stomach, and
small intestine from a village pig). At final mortuary feasts, owning peace
chiefs give ikufuka to each of the kofuapie peace chiefs (including their
friends) and peace sorcerers of the tribe. Because senior and junior peace
chiefs of the same clan ideally make their feasts together, the senior peace
chief gives ikufuka meat to senior kofuapie peace chiefs and peace sor-
cerers, and the junior peace chief to junior kofuapie peace chiefs and
peace sorcerers. Kofuapie peace chiefs of the same status level, whether
they are friend or nonfriend, should return identical quantities of ikufuka
meat when they make their own feasts. Ikufuka is thus precisely re-
ciprocal between kofuapie peace chiefs and, through them, their peace
sorcerers. Without belaboring this point, these reciprocities replicate the
idealized structure of marriage exchange.
Peace sorcerers and the other two kinds of hereditary subclan official -
war chiefs and war sorcerers - also exchange ikufuka among themselves,
but in different contexts - at their respective public installation cere-
monies. The installations of peace sorcerers, war chiefs, and war sorcerers
are not mortuary feasts. Only peace chiefs use the occasion of their sub-
clans' or clans' death feasts to install their heirs to office. In the same
respect that the three other officials do not make death feasts, they do not
have friends in their own right like peace chiefs do, and so only ikufuka is
reciprocated between them when they install their heirs.
182
 Feasts of death (ii) 183
It is noteworthy also that whereas peace chiefs may give ikufuka many
times during their tenure in office (i.e., as often as their clans give death
feasts), peace sorcerers, war chiefs, and war sorcerers each give ikufuka
independently only twice - once when they assume their offices from
their fathers, and later when they install their sons.

Ikufuka meat and the blood of Akaisa Men

Knowledgeable villagers claim that iunge fanga, the backskin and internal
organs of ikufuka, is the blood or flesh of forebearers of the clan official
who gives it away. He may never eat the iunge fanga or ikufuka1 he gives,
nor indeed may he eat any he receives or that comes his way otherwise
from one of the other officials of his clan or his mother's clan. Incumbents
to office informally designate their intended successors (ideally eldest
sons) by preventing them from eating ikufuka while they are small boys
and before they have been formally installed. To eat ikufuka even once,
irrespective of its source, forever disqualifies a man from being of au
akaisa, or Akaisa Man, status. Men who have eaten ikufuka are doomed
to be "poor" or "ordinary men" (ulalu) for life. The prohibition for Akaisa
Men against eating ikufuka extends to their wives. If either one eats
ikufuka even mistakenly, the officeholder automatically relinquishes his
position.
The ritual specializations of Akaisa Men have been described in Chap-
ter 6. Akaisa is the name of the principal culture hero of the Bush Mekeo
who originally ordained the fourfold division of clan authority and gave
the people the institution of the death feast. Present-day Akaisa Men are
regarded as the descendants of the first officials personally appointed by
Akaisa in the mythical past. In the field where I inquired of many details
concerning traditional death feasting, peace sorcery, warfare, and so on,
my informants often referred me to one or another event described in the
Akaisa saga. Therefore, to appreciate fully the significance of ikufuka at
mortuary feasts, I, too, shall have to refer to certain features of the myth
of Akaisa. With villagers themselves pointing the way, I shall regard the
Akaisa myth in part as a "charter" (Malinowski 1948) for many of the
elements of the death feast of concern here. Seligmann published a few
versions of this myth current among the Roro and Central Mekeo at the
turn of the century (Seligmann 1910:304-9). Here I shall translate the
version I collected myself among the Bush Mekeo in 1975, interspersed at
points with commentary provided by my informants. The complete myth
is composed of two segments or sequences.

Akaisa myth 1: inauguration of hereditary offices

Long ago, an old woman lived in the village. She had neither husband nor sons to
give her meat. Whenever the men came back with game, she was only given
184 Quadripartite structures
bones, or nothing at all. Akaisa was a spirit (tsiange). He heard her cries of sadness
and assumed the form of a young boy. The woman adopted him as her son.
The spirit, Akaisa, was not born of human beings, but he was adopted
by a woman. Even in adoption, Akaisa only had a mother and no living
father or siblings.
Akaisa went hunting with the men the next day, pitching his "father's" [deceased
husband of Akaisa's mother] torn net away from the conjoined nets of the other
men. When all the animals came into Akaisas net but none went into the nets of
the other men, they beat Akaisa and took the mature animals, leaving him only
with the babies. On the way back to the village, Akaisa breathed (eunge) spells
into the anuses of the baby animals, and they grew to full size. Then Akaisa carried
his meat to the village and gave it to his mother.
Akaisa's "breath" (iunge) accompanied by "spells" (menga) whispered
into the animals' lower abdominal tracts (ina) made the small animal
bodies grow large. [Note that iunge or breath is linguistically identical
with the part of ikufuka that includes internal organs of the abdomen -
iunge fanga.]
The next day, Akaisa went hunting again, and the same thing happened. The men
beat Akaisa and stole his big animals. Akaisa blew into the baby ones to make
them grow, and he gave them to his mother. On the third day, Akaisa stayed
home and rested, and the men returned from hunting with nothing. When his
mother expressed her curiosity, Akaisa explained what the men had done to him:
"It would have been alright to take the meat, but not to beat me," he said. To
convince her, the next day he went hunting again; the men beat him, stole his
meat, and so on. Then, on the following day when Akaisa rested, they came back
with nothing again.
Men get bush meat only by taking it from Akaisa. All bush meat comes
from Akaisa; it is Akaisas. Without him there is no meat. So why beat
Akaisa? [Note also that the meat in question is contraceptive meat: pig,
cassowary, wallaby.]
The next morning, Akaisa told his mother he was tired of forest meat and wanted
to eat fish. He went fishing alone and returned that night, not only with enough
fish (angai mullet) for his mother but enough for all the people of the village.
When the people told his mother they wanted more fish, Akaisa said that the next
day, when the men rested from hunting, the women would go with him fishing.
Some would come back in the evening with fresh fish, and others would stay over
at the river, smoke the remaining fish, and return the following day with Akaisa
and his mother.
In the morning, Akaisa poled the women across to the opposite bank of the
river, and there the women fished with their nets. Akaisa went upstream out of
sight, changed himself into a large fish (boku barramundi), swam between the legs
of the wading women, and examined the vaginas of each. When he came back to
 Feasts of death (ii) 185
them in the form of a human boy, the women told him of the large fish that had
molested them. Akaisa said they should have caught it so their husbands could eat
it that night. When the women were ready to go burdened with their fish in the
afternoon, Akaisa made them stand in two groups. First, he carried the women
with large vaginas to the village bank; then he returned for the women with small
vaginas. On crossing the river with them on board the canoe, the river flooded
and they were swept downstream. Akaisa called back to the women with large
vaginas that the women with him were his wives now.
Barramundi is contraceptive meat. Women with small vaginas have not
yet given birth. The firstborn of a woman is typically smaller than subse-
quent children. Husbands of the women with small vaginas thus as yet
have no children. [Note that if the women had caught Akaisa in the form
of a fish, their husbands could have eaten his flesh.]
The women with large vaginas returned home and report that Akaisa has stolen
some of the wives of the men. After traveling some distance and running aground,
Akaisa and his wives make a camp to spend the night. Akaisa makes a fire for them
to cook the fish, telling them that, because they will never see their husbands
again, they had better eat the fish. But the women do not eat. During the night,
the camp rises to form a mountain high above the coastal plain. In the morning,
the women can see their husbands coming to take them back from Akaisa. When
the men try to cross the river, Akaisa makes them fall unconscious.
"Unconscious" (mae kafu) is one form of death (mae), as when a peace
sorcerer "hits," "kills," or renders "unconscious" his victim. [Note that
Akaisa's wives do not eat the fish, but they do cook it, just as the wives of
Akaisa Men cook but do not eat ikufuka.]
After a while, Akaisa brings the men back to life and dares them to try to kill him.
They try and again fall unconscious. After they are revived again, Akaisa tells
them that, if they want to kill him, they must stand in two groups. So he separates
them and they fight each other, friend (pisaua) against friend. Akaisa made them
do this with his hot powers. The men speared and clubbed one another, and many
were slain with arrows. Blood was everywhere. Akaisa called down to them,
saying they were relatives (atsi atsitsi) and one blood, so he bid them get up, and
he asked them if it was good for them to fight each other. They replied, "no, it is
bad. We came to take our wives away from you, but you make us angry and fight
each other." Akaisa then informs them that whenever they want tofighthim, they
instead will fight each other.
Long ago, our ancestors did not die. Nowadays whenever people be-
come angry and shamed with the killers of their relatives or wives, or with
the people who hired them to kill, really it is Akaisa [who killed them].
Affirming that they now understand, Akaisa has the men stand in one group
(ngopu). Akaisa gives them some customs (kangakanga). He throws down the
things [charms and other paraphernalia] for men who want to be war chiefs, peace
sorcerers, war sorcerers, or peace chiefs.
186 Quadripartite structures
Akaisa proclaims that peace chiefs must decide if the war chief and war sorcerer
will go to fight, and they must obey their peace chief. The peace chief, Akaisa
says, is very kind to his people and to anybody else. Sometimes he must keep the
people from fighting, other times he allows them to fight. When people make
trouble for the peace chief, then he calls them to fight and be killed. Then Akaisa
tells the peace chiefs what they must do when their people are killed [i.e., how to
make death feasts]. To the "last men" Akaisa gives peace sorcery things. He tells
the peace sorcerers they must listen to their peace chiefs, do their work only when
the peace chiefs tell them, when the peace chiefs are shamed or angered. Peace
sorcerers must look after the peace chiefs.
These men were the first Akaisa Men. The peace chief is the most
senior (fakania), and the other three are junior (eke) and must listen to
their "elder brother/' The hereditary customs they are given include the
ritual paraphernalia each is to use.
Once more, Akaisa tells the Akaisa Men to stand in two groups and to fight. This
time none of them gets hit because Akaisa has told them how to dodge the arrows
and spears, that they must fight this way to avoid being killed. They keep on
fighting, and Akaisa tells them to stop, but they do not listen. Akaisa tells the
peace chiefs to spread their lime on the ground to make the people stop fighting,
and they do.
It is by tightening that men are light and nimble enough to dodge
projectiles in war. By spreading lime powder on the ground, peace chiefs
stop (intratribal) fights. If fighting still continues, warriors of any clan
involved must pay a fine to the peace chief, or his sorcerer will be sent to
kill them (cf. Hau'ofa 1981:215-88).
Now Akaisa says that if the people are strong enough to catch the things of death
he throws down to them, they can live again after they die. First, he throws the
light [dry] skins of snakes and prawns, and these the people catch. They say
"Ahhhhhhhh." Then Akaisa throws the heavy [wet] bones of dead animals and
humans. If they catch these, their spirits (tsiange) after death will go to Akaisa,
and he will send them back to their bodies. But if the men do not catch them, they
will die forever. As the bones fall, the people are frightened by the noise and let
them drop to the ground, but they pick them up as though they had caught them
and call out, "Ahhhhhhh."
When the people say "Ahhhhhhh" the second time, they are lying
(pifonge) to Akaisa, but he knows. This is why snakes and prawns are born
with new skins after they die, but other animals and humans are not born
again after they die. Spirits of the dead stay with Akaisa. [Note that the
Akaisa Men, by catching the skins, are given some form of rebirth, but by
dropping the bones they are also denied rebirth in some other form.]
Akaisa then enjoins all the people to return to their village and live with these new
customs, each man doing his own way as peace chief, peace sorcerer, war chief, or
war sorcerer. Akaisa says he does not like that the people will die, and he talks
 Feasts of death (ii) 187

about the old woman, his mother, staying alone, poor, and sad. That was why he
came out of the bush, stayed with her, and went hunting with the men. But the
men were greedy. It was alright that they took his meat, but they beat him and
shamed him. That is why Akaisa will stay away from the people, and the people
will stay away from him. Akaisa says that when people die, they will die forever. If
they had not beaten him, he would stay with them so that when any of them die,
he would make them live again.
So ends the first segment of the myth. Akaisa appoints as au akaisa the
men who come to him for their wives. All four types of Akaisa Men are
designed to help resolve the issues of life, death, and new life. In this and
other significant respects, Akaisa calls to mind the mythical figures Amaka
and Oa Lope. But because I am primarily concerned with peace chiefs
and the interpretation of mortuary feasts, I shall restrict my comments
here to points most critical to these purposes.
At the beginning of the sequence, the men of the village take Akaisa's
big contraceptive meat and leave him with small animals. Akaisa's breath
and spells are blown into the anuses of the small animals, making them
grow into large contraceptive meat. After Akaisa demonstrates that meat
the men steal comes only through him, he steals the women with small
vaginas to become his wives, leaving the men to keep only the women
with large vaginas. The women Akaisa steals, as well as their husbands,
have no children. The women Akaisa does not steal have already given
birth. Significantly, Akaisa's own adoptive mother is among the childless
village women he takes away.
The women with large vaginas return home with their fish for their
families. Akaisa's wives cook, but he and they do not eat their fish, and
neither do the husbands of Akaisa's stolen women. When the angry hus-
bands come to kill Akaisa and take back their wives, Akaisa kills them
three times and revives them three times. Akaisa then bestows upon
these men his own powers concerning life, death, and, in part, new life.
This sequence of killing and reviving parallels the earlier sequence
involving the hunting of game animals. Akaisa kills game animals three
times, and, after the big ones are forcefully taken from him, he grows the
smaller ones with his breath. Then, after resting a day, he kills and grows
the small animals a fourth time. In other words, Akaisa's final bestowing
of au akaisa status upon the men, signaling their new social respon-
sibilities and prerogatives, is a comparable fourth life or rebirth. The
husbands of childless women become Akaisa Men, or Akaisa's human
representatives and successors.
In this version of the myth, Akaisa sleeps on the mountain with the
childless wives. But it is uncertain if these women return to their hus-
bands and implictly bear children afterward as a result of conception with
Akaisa's semen, or not. If so, then all Akaisa Men and the firstborn male
188 Quadripartite structures
descendants of Akaisa Men share some of Akaisa's blood and would be, in
a sense, "fathers/brothers/sons-in-Akaisa" to one another. Otherwise, as
in the three versions Seligmann cites, the human skin and human body
(or bones) Akaisa throws down are those of the women he has stolen
(Seligmann 1910:306-8). This would leave the au akaisa husbands en-
tirely without wives, suggesting that they are to reproduce their heirs
without women. Such au akaisa heirs would thus be sons of their mothers
only by adoption, like Akaisa himself. In this sense, the relations of all
Akaisa Men, the heirs of Akaisa Men, their wives and mothers, and Akaisa
and his adoptive mother are comparable to the notion of "Virgin Birth"
(Chapter 4; Leach 1966b).
This divergence in the myth as regards the fate of the wives with small
vaginas and their husbands' progeny recalls the conflicting conclusions of
the afinama myth (Chapter 5) whereby men either eat or do not eat the
flesh of women who, as flying foxes, have left their husbands, brothers,
fathers, and sons for a high place in the bush. 2 In the first Akaisa myth
segment, on the one hand, Akaisa Men sexually reproduce their heirs
through their wives whom Akaisa has inseminated. On the other, Akaisa
Men who Akaisa has likewise appointed heirs, but without wives and
sexual reproduction. As we shall see later in the chapter, these two in-
terpretations are not entirely irreconcilable. In either case, all ancestral
Akaisa Men are linked or likened to one another and to Akaisa, but in
contrast with ordinary men; that is, men who are the husbands of women
with large vaginas who have born children by sexual means only, who
have eaten their wives' cooked fish, and who did not receive any of
Akaisa's powers. Thus, Akaisa Men share or trace a quasiagnatic ritual
relationship originating in Akaisa along lines of father to eldest son that
transcend or crosscut the distinctions between clans and moieties. At one
point in the myth, after making the men stand in two groups, Akaisa
reminds the Akaisa Men that they are all relatives and one blood to one
another. In other words, Akaisa Men, irrespective of kin, clan, and
moiety affiliation, are all quasiagnates of Akaisa defined by not eating a
certain meat (fish cooked by wives); by dying and being given new life by
Akaisa; by reproducing in this way without wives or with women already
inseminated by Akaisa; by having mothers through adoption only; and by
mediating in some fashion the death, de-conception, and limited re-
generation or re-conception of ordinary human beings.
A significant element of this last-mentioned Akaisa Man attribute,
namely, mortuary de-conception and re-conception, in light of my earlier
interpretation and analysis of the outer pange feast prestation, surrounds
the contrast of skin and bones that Akaisa throws to the men below. They
catch the skin and say "Ahhhhhhh," as do the friends at final death feasts,
but they receive only qualified or incomplete new life. The bones fall to
 Feasts of death (it) 189
the ground, and the people get permanent bodily death. Again paralleling
the afinama myth, the men both catch and do not catch female flesh
falling from above. For the moment, I simply draw attention to these
perhaps confusing mythical features of de-conception and regeneration or
re-conception and proceed to the second and final sequence of the Akaisa
saga, which focuses upon the interactions between Akaisa and his younger
brother, Tsabini. The relations of the first segment just outlined, along
with several others dealing with death, de-conception, and re-concep-
tion, are again highly prominent.
Akaisa myth 2: Akaisa and Tsabini
Tsabini was Akaisa's younger brother. Tsabini was not very skillful or clever
(etsifa) like Akaisa. One day, Tsabini told his wives they must go visit Akaisa.
When Tsabini's wives entered Akaisa's village, they just saw a young boy chewing
areca and betel on Akaisa's clubhouse, and they told him he must go get Akaisa.
They did not recognize him because he had taken this youthful shape. Young
Akaisa went to the garden and informed his wives that they must bring back
plenty of food for his younger brother's wives. In adult form, Akaisa returned to
the village awaiting his wives' return. When Akaisa's wives came back, they told
Tsabinis wives that Akaisa was always playing (bakau) tricks like that, sometimes
looking like a boy, other times like a man. Tsabinis wives apologized for shaming
him. When he asked why Tsabini had not come himself, the women said that
Tsabini did not want to.
The next day, Akaisa went hunting and returned with many dead pigs, casso-
waries, and wallabies. After the meat was cooked and the wives of Tsabini ate
some, they vomited it up. They agreed the meat was good, however. They had
never eaten meat before, 3 as Tsabini and his people ate instead the skins of raw
almond fruit (kamai) for meat. When Akaisa brought back meat the subsequent
day, Tsabinis wives ate some and did not vomit this time. Because Tsabini did not
come with them, Akaisa told them they would go back the next day. Before they
left, he asked if Tsabini had a son. One woman said yes, so Akaisa told her Tsabini
must name him after him, Akaisa. Akaisa killed some of his village pigs to carry
back to Tsabini, and because there was more meat than they could carry, Akaisa's
wives went along to help.
When the women arrived at Tsabinis village, they cooked and ate some of
Akaisa's meat. When Tsabini saw the village meat, he asked what it was, and his
wives explained to him that they grew it at Akaisa's village. They said that they
had been eating bad meat [skins of kamai] while Akaisa's people had been eating
good meat. When Tsabini ate some of the good meat, he too vomited.
In the morning, Tsabini went "hunting" and brought back many almond fruit.
His wives cooked some and served it to Akaisa's wives, who ate some and
vomited. The remainder they placed on a smoking rack because they did not like
it [it was unsweet to them]. When Tsabini saw that they were not eating his meat,
he explained that he had no animals for meat. When Akaisa's wives left for home
burdened with skins of almond fruit, Tsabini told them Akaisa must name his son
Tsabini, after him.
190 Quadripartite structures

Akaisa saw the almond fruit and asked what they were. His wives said they
were Tsabini's "meat," but Akaisa said they were rubbish and threw them away.
When Tsabini heard of this, he was shamed by his brother's rejection of his gift.
He sent some of his people to spy on Akaisa to learn how Akaisa acquired his good
meat. They reported that Akaisa had a fence or place in the bush. He merely
opened the way inside, and as the animals come out, he shot them, then closed
the way when he left. Now Tsabini understood why he had no good meat, so he
released all Akaisa's wild pigs, cassowaries, and wallabies in the forest.
When Akaisa learned of this, he was angered and claimed he would have shown
Tsabini how to hunt meat if he would have come, visited his brother instead of
sending his wives, and asked him. Akaisa decided to play a trick (bakau) on
Tsabini and pay (kaua) him back. Akaisa told his people to clear a garden for him.
For their work, he gave them a feast of village pig meat. He told everyone they
must pass it on to Tsabini that Akaisa gave the people his own mother to eat.
Later, when Tsabini asked how Akaisa helped his people, he was told about the
garden and Akaisa's giving the people his mother's flesh to eat. So Tsabini had a
garden made and really did kill and cook his own mother to feed the people who
worked for him.
When Tsabini afterward learned that Akaisa had played this trick on him, he
was shamed and angered, and he wanted to kill Akaisa's son, Tsabini, his own
namesake. Because Tsabini had no sorcery of his own, he stole young Tsabinis
loincloth and gave it to Afupa [a large burrowing lizard of the coastal plain; a/ai/ai
species] to make mefu sorcery. Akaisa's son, Tsabini, sickened and died. Then
Akaisa killed Tsabinis son, young Akaisa, with peace sorcery.
At one level, and I think a superficial one, this portion of the myth
relates to the origin and escalatory nature of "jealousy" or "quar-
relsomeness" (pikupa) arising from the conflict between "equality" and
"inequality" intrinsic to the sibling relationship (cf. Hau'ofa 1981). I
would like to focus here on other elements.
Tsabini shames and angers Akaisa when he sends his wives without
him. When Tsabini steals Akaisa's meat, Akaisa is shamed and angered.
Tsabini and Akaisa, apparently, have different mothers, and for stealing
his meat Akaisa tricks Tsabini into killing his own mother and feeding her
flesh to his people, so that Tsabini is also shamed and angered. Finally,
the two brothers kill one another's sons, their own namesakes, and, im-
plicitly, themselves.
In this sequence, it seems, food, especially meat, is equated with moth-
er, with son, with brother, and, through the namesake relationship, with
oneself. These relations repeat those expressed in the first segment of the
Akaisa myth, except in reverse. There, meat is associated with Akaisa's
mother, Akaisa's installed sons or successors, and, through the same-
name relationship between him (Akaisa) and them (the Akaisa Men), with
himself. Here, rather than giving meat to his mother to eat, Akaisa kills
his "mother" in the form of a village pig for other people of his group to
 Feasts of death (ii) 191

eat. Rather than creating sons, Akaisa's son is killed. And rather than
creating namesakes, namesakes are killed.
In the first segment of the myth, there is a recurrent theme of death
followed by regeneration. So far in the second sequence, we have seen
only death. In its conclusion, to which I now turn, however, the issues of
de-conception and re-conception are addressed once more.
After Tsabini killed his son, Akaisa did not immediately bury the body. Carrying
his son in his arms, Akaisa wandered to a mountaintop from which he could not
see the places where he lived with his son and where his son died. Sight of these
places made him sad. En route to the mountain, Akaisa cried in mourning as his
sons flesh rotted from the bones and the juices dripped onto Akaisa's skin. Atop
the mountain, Akaisa placed the bones on a rack and dried them over a fire. At
night, the bones turned into wallabies,4 jumped down from the rack, and ate the
grass growing nearby. At dawn, the wallabies jumped back onto the rack and
turned into bones once more. After several nights, Akaisa noticed that the bones
of his son were eating the grass, and he cried.
Two boys playing with grass spears came upon Akaisa at the mountaintop.
Akaisa asked them if there was plenty of these two kinds of grass, and they said
yes. They brought some to him, and thereafter Akaisa ate only this as his food, just
as the bones ate grass. The two boys stayed with Akaisa, and he asked them if their
father was still alive. They said he was. Akaisa told them to bring the father to
him. When he came, Akaisa told him the story of how his son died, and he asked
him to make a garden for him there on the mountain so he, Akaisa, could make a
feast. When the plants started to grow, the bones turned into wallabies and ate
the fruit. The man informed Akaisa of this, and Akaisa told him to put up his
hunting net at night to catch the bones. The man did this, and some of the
wallabies got into the net while others jumped over it and escaped. In the morn-
ing, Akaisa saw the wallabies, which he had never seen before. Also, he noticed
that some of the bones on the rack were missing; therefore, he knew that the
wallabies were his son's bones. The hair was burnt off the dead wallabies, the
meat cooked, and some was fed to the two boys. Because they did not sicken and
die, Akaisa knew it was good meat and told the man he would use it for his feast.
Akaisa sent word to all the peoples that he would make a feast, and he told the
Mekeo and Roro they must come and sing and dance. When they arrived, Akaisa
made the first mortuary feast, and to the peace chiefs he gave pange from the flesh
and blood of wallaby. Thereafter, Akaisa was released from mourning, and he left
the people forever; but he did not die himself.
When young Tsabini is killed, Akaisa carries the body until it rots and
drips his son's bloody dirt onto his skin. Akaisa cries and mourns over his
son as oaifa. The sight of the places of his sons life and death makes him
sad. Ordinary food his son ate before his death becomes Akaisa's bafu
prohibited food, so Akaisa only eats the grasses that his son (as wallabies)
eats in death. Akaisa does no work and does not hunt, but he gets others
to work and hunt for him so he can make the final death feast. This is why,
192 Quadripartite structures

villagers argue, some of the relatives of the deceased are released from
mourning at the preliminary burial feast - to work and hunt so that the
oaifa mourners can eventually be released at the final mortuary feast.
At Akaisa's feast, part of his sons body (the bones) represented by
wallabies is given to the peace chiefs for distribution among the ordinary
people of their clans. Present-day villagers keep and bury the bones of
dead clanspeople. Thus, when Akaisa gives away the bones of his son as
wallabies, he dissociates from himself that which, upon the death of ordi-
nary villagers today, is not given away or dissociated from clanspeople
through de-conception at feasts. So at the end of the myth, Akaisa is left
with no sons born of his own wives, and he does not even have his son's
bones. This result is comparable to that of the first segment of the myth
when, according to one variant of the story, Akaisa Men lose their wives
in death and thereby lose the prospect of producing sons and heirs of their
own through sexual reproduction. In other words, Akaisa gives the blood
and flesh of his sexually conceived son to his appointed or asexually
conceived sons or heirs who, themselves, lack sexually conceived sons of
their own. They, in turn, give Akaisa's son's blood and flesh (and that of
their own sons as well) for eating to ordinary people who do have sexually
conceived offspring.
I argued in the previous chapter that the outer portion of pange given
at mortuary feasts represents the cognatic bloods of the deceased derived
from both of his or her grandmothers (father's mother and mother's moth-
er), or, alternately viewed, the bloods of spouse's clan and spouse's moth-
er's clan. Akaisa's giving of wallaby meat - his own son's blood and flesh -
is also connected in the myth with the giving of a grandmother's blood
and flesh. Just before Tsabini, Akaisa's bother, kills Akaisa's son, young
Tsabini, Akaisa makes a feast with the flesh of his mother's body, or young
Tsabini's grandmother. Moreover, it was in retaliation or compensation
(kaua) for Akaisa's tricking him into killing his own mother that the broth-
er, Tsabini, kills his own namesake. The myth in effect equates the flesh
or blood of Akaisa's sexually conceived son with that of Akaisa's mother
and, from the first segment of the myth, with the blood and flesh of
Akaisa's wives.
In sum, the myth portrays Akaisa giving away in initiatory feasts to his
successors the blood and flesh of his mother, son, and wives. Might
Akaisa's extraordinary asexually reproduced heirs, the Akaisa Men (peace
chiefs, peace sorcerers, war chiefs, and war sorcerers), likewise recipro-
cate among themselves the blood and flesh of their respective mothers,
sons, and wives when they feast or install their chosen successors? If so,
how does this relate to the issues of death and de-conception, on the one
hand, and of regeneration or re-conception, on the other?
 Feasts of death (ii) 193

Ikufuka meat
Answers to these questions can be provided by returning to the issue of
the meanings attributed to the separate parts of ikufuka prestations that
peace chiefs and other Akaisa Men reciprocate at death feasts and official
installations. Ikufuka meat, it will be remembered, consists of iunge fan-
ga organs and backskin strips taken from femaleish village pigs, and of the
meat of dogs.
The internal organs of iunge fange - liver, small intestine, and stomach
- are all part of the pig's ina or abdomen. As in the contexts of spatial
relations and the human body, the ina or abdomen of a pig is not just
simply an inside part of the body, but rather an inverted outside of the
body. As I have shown in Chapters 2 through 5 as well, the ina or
abdomen is conceptually associated with sex, conception, birth, assimila-
tion of blood-elements of plant food, and the female physiognomy gener-
ally. Most significantly in the present context, perhaps, the word ina is
also the relationship term for the category "mother."
Because ordinary villagers never officiate at death feasts or installa-
tions, they never themselves have the opportunity to give away the blood
or flesh of mothers clan, and upon their own deaths the blood of their
mothers and mother's clans will not be represented in the outer portion of
pange or in ikufuka meat. Peace chiefs and other Akaisa Men, however,
do give away the flesh of ina (i.e., abdomen, mother, or mother's clan) in
the form of iunge fanga abdominal organs during their own lifetimes. In
the act of being installed by cutting and giving away iunge fanga, suc-
cessors to au akaisa office ritually de-conceive themselves from the cog-
natic blood of their mothers and/or mother's clans. At the same time, of
course, they are re-conceived as quasiagnatic descendants of Akaisa. Here
it is worth recalling from the first segment of the Akaisa myth that Akaisa
blows iunge or breath into the abdomen of each animal he later gives to
his mother. At first, the animals are immature, but Akaisa's breath causes
the bodies of the animals to grow to full size, or, closer to the indigenous
understanding, Akaisa's iunge is hot for bodily growth and sustenance.
That Akaisa appears repeatedly in the myth as a boy is certainly significant
here also. Specifically, Akaisa's youth symbolizes the tremendous capaci-
ty of the young to grow and mature, but not (as contrasted with the fertile
female) to reproduce and bear offspring. Akaisa's mother eats the flesh of
the animals grown by Akaisa's hot iunge, and this growth capacity is
incorporated into her own blood and flesh. Still, it is growth capacity
neither for giving and producing new life, that is, for sexual fertility and
procreation, nor for eternal life without death. Although Akaisa's mother
is female, she is stated to be old and thereby presumed to be reproduc-
194 Quadripartite structures

tively sterile. The meat Akaisa gives her - the meat she eats to sustain her
life, the meat she incorporates into her own flesh - is contraceptive meat.
Consequently, the iunge fanga meat ordinary men receive from Akaisa
Men at feasts and installations - the meat they eat and incorporate into
their bodies - is likewise contraceptive meat. It gives them life, or the
temporary life-sustaining capacity of bodily growth, but it does not give
them immortality, that is, neither eternal life for the individual nor the
capacity to reproduce.
Moreover, the eating of iunge fanga abdominal organs denies to ordi-
nary men one kind of reproductive immortality that is the specific pre-
rogative of Akaisa Men, namely, the capacity to reproduce themselves in
the persons of their publicly installed successors, generation after genera-
tion, without the sexual contributions of women.
Taken together, these several points add new interpretive insight to
Akaisa's feast occurring in the second segment of the myth. Akaisa gives
to the ordinary people the flesh of his mother to eat in the form of village
pig. It is through this mythical feast of his mother's body, I suggest, that
Akaisa gives to the people his own iunge breath possessing the contracep-
tive capacity for bodily growth. In this context, it is logical that Akaisas
mother is understood as adoptive. When peace chiefs or other Akaisa
Men give iunge fanga to the ordinary men of their respective clans, they
"give iunge' (iunge ebia or iunge ehinia). Thus, as the blood and flesh of
Akaisa Men's forebearers, iunge fanga abdominal meat is at least in part
Akaisa Men's mothers' blood and flesh, and by giving it away Akaisa Men
de-conceive themselves as distinct from mothers' clans.
The strips of pig backskin that are tied to these pieces of iunge fanga
abdominal organs had bothered me for a very long time. Particularly in
the early stages of analyzing the culture, I found it difficult to reconcile
the binding together of outside flesh (the skin) with internal flesh (the
organs). This puzzle tended to throw off my whole analysis as I had
developed it so far. The pig's backskin, I was long convinced, simply did
not belong with the internal organs of iunge fanga.
The problem here stems initially, of course, from viewing the abdomi-
nal region as a principally inside place in complete opposition to the skin,
an outside place. The ina or abdomen, it will be remembered, is concep-
tualized in the culture rather as an inverted outside, or, as I tried to
describe it in Chapter 3, as an inversion of outside space into the body.
Thus, although something contained by the abdomen might appear to be
inside or internal, in the categories of Bush Mekeo culture it is something
technically outside the body. In this respect, then, the abdominal cavity
consists of outside skin pushed in upon itself. The inconsistency of bind-
ing outside skin with inside (sic) organs is thus revealed to be merely
apparent and misconceived. Peculiar as it might seem to a Westerner's
 Feasts of death (ii) 195
intuitions, for the Bush Mekeo skin and abdomen are both fundamentally
or categorically outside parts of the body and thereby compatible ingre-
dients of iunge fanga feast prestations.
Moreover, skin no less than abdomen is consistently associated with
femaleish qualities elsewhere in the culture. As a result of newlywed
procreative engorging, women acquire an abundance of skin. Female
members are described also as the skin of a clan. Death-feast village pigs
from which iunge fanga is taken are viewed as femaleish in contrast to
maleish wild pigs, whether they are biologically male or female.
Homologies of these categories in the context of social relationships are
also revealing. Ina or mother's clan and cognatic blood correspond with
ina or abdomen and inverted outside; and spouse's clan and affinal blood
correspond with outside. The definitive linking relation in each case is
female, and the clan and blood identities correspond respectively with
opposite moiety and with blood relations consequent upon exogmaous
exchange:

inside : outside :: inverted outside : everted inside (4)

remote peripheral village adjacent
bush village abdomen bush (3)
spouse's
own clan : spouse's clan :: mother's clan : mother's clan (16)
clan A1 : clan B 1 :: clan B 2 : clan A2 (19)
agnatic nonblood cognatic cognatic
:
blood : (affines) :: blood nonblood (18)

Therefore, employing the same logic that represents mothers clan and
cognatic blood with the giving of iunge fanga abdominal organs, iunge
fanga backskin of village pigs given by Akaisa Men represents the giving
back or de-conception of spouses clan and affinal blood. Further support
for this contention is provided later in this section where I discuss the dog
portion of ikufuka meat that is joined with iunge fanga.
Before I turn to consider this remaining portion of feast food, it will be
informative to draw attention to a separate additional symbolic value
attached specifically to the backskin strips of iunge fanga. Based solely on
my own impressions of villagers' attitudes, statements, and actual feast
performances, the ritualized cutting, care, and disposition of backskin
strips are the most intense and focused events of village life. It is backskin
more than abdominal organs, for example, that most villagers are likely to
mention or discuss. It is backskin ritual that is the most solemn of all the
proceedings. Finally, it is backskin with its thick layer of fat that the
commoner old men and ancestors had always found the most sweet.5
These observations concerning the special regard for iunge fanga back-
196 Quadripartite structures
skin strips, along with the principal question they give rise to - namely,
why is backskin so emphasized in the context of death feasts and installa-
tions? - are issues of an order different from that which has concerned me
elsewhere in this and the previous chapter. Hence, they will lead to a
slightly different sort of answer, as follows.
I remember an instance where one of the clans of my village received
some iungefanga as part of ikufuka meat from the peace chief of another
village. After their friends had departed for home, the men had a lively
discussion over whether the iunge fanga backskin and organs their chief
had just received were in fact originally derived from a village pig and
legitimate, or taken from the back of a wild pig of the bush. No one
present had seen the pig killed. The whole question arose simply because
the backskin part of the iungefanga bundle had no thick layer of fat, as is
purposefully cultivated in village pigs. Typically, wild pigs are very lean
by comparison. On the basis of the backskin, then, most men of the clan
claimed these iungefanga organs as well as backskin came from a bush pig
and not a village pig. And, furthermore, they planned on ignoring the
obligation to repay this counterfeit iunge fanga and the rest of the pig it
came from to their friends at their own upcoming feast.
When you think about it, and probably when the Bush Mekeo think
about it too, the internal organs of village pigs and bush pigs are visibly
indistinguishable. Some external indicator or sign is needed to differenti-
ate them. The pig's backskin is the only visible and uniform part of its
body that can definitely lay this question aside. And for this reason, I am
arguing, iunge fanga backskin has assumed its provocative role in the
culture.
With the dog portion of the ikufuka prestation, the analogous case is
never an issue. All dogs (auke), unlike some pigs, reside only at the
village. Bush Mekeo dogs are unique in this respect among all the local
animal species, and, consequently, they live in closest proximity to
human beings. Perhaps it is not surprising, then, that the incorporation of
dog meat into ikufuka at feasts and installations has other special
meanings.
I have already mentioned that peace chiefs and other Akaisa Men
ideally provide their own dogs at feasts and installations. All village men
own their own dogs outright, rather than sharing ownership with their
wives as they do with regard to pigs. Also, men tend to their dogs by
themselves. Dogs are not given cooked food of their own; they eat the
rubbish and leftovers that human villagers do not eat. Unlike femaleish
village boars, male dogs are not castrated. Like men, but unlike women
and pigs, dogs hunt for meat in the bush. Generally, then, even the fertile
bitch of the village is as humanly maleish as the castrated village boar is
humanly femaleish.
Indigenous conceptualizations of the bush wallabies, matsi and mani,
 Feasts of death (ii) 197
closely parallel those of the village dog. The two local species of wallaby
are said to be brothers - matsi, the smaller, is senior, and mani, the
larger, is junior. Wallabies, like dogs, of course, are not castrated. They
also eat rubbish (grasses) and do not live in both village and bush like pigs.
Whenever a hunter kills a wallaby in the bush, he breaks its hind legs and
leaves the body on the trail to pick up and carry home on his return trip.
One of the first Europeans to penetrate the Mekeo area in the late nine-
teenth century recorded that when a peace chief killed two village dogs to
honor him, its legs were first broken (Chalmers 1887:12). As far as I am
aware, of all animal species only wallabies' and dogs' legs are broken in
this manner.6 Villagers add that wallabies are "Akaisa's dogs" (Akaisa nga
auke). Therefore, just as Akaisa mythically gave his dogs as his son's
bones, Akaisa Men give away their dogs at death feasts and installations to
represent their wives' bloods, which were sexually incorporated at con-
ception into the bodies of their sons. With the dog portion of ikufuka,
then, Akaisa Men de-conceive and re-conceive their sons and themselves
with respect to their sons' mothers' bloods (i.e., their own wives' bloods).
Indeed, villagers say that at their own installations, "au akaisa kengama"
or "Akaisa men are begun/de-conceived/re-conceived/happy."
This condition where Akaisa Men have lost cognatically derived and
transmitted female blood is reminiscent of the original mythically ap-
pointed Akaisa Men who lost their wives, and with them the prospect of
sexually produced offspring. According to the two alternate interpreta-
tions of the myth's first segment, it will be remembered, appointed
Akaisa Men implicitly must either produce their au akaisa heirs asexually
(because they lack wives) or obtain their heirs as a result of Akaisa's, and
not their own, inseminating of their wives. In either case, it works out to
the same result. The giving away of de-conceptive or re-conceptive
bloods represented in ikufuka meat leaves peace chiefs and other Akaisa
Men without the blood of their mothers and without the blood of their
wives through the persons of their offspring. So, in effect, all categories of
blood relationship through females and all female blood in the persons of
Akaisa Men and their hereditary patrilines are negated; strictly quasiag-
natic blood- and relationship-in-Akaisa remains. For this latter reason,
Akaisa Men cannot ever eat ikufuka regardless of who gives and who
receives it. This, moreover, allows Akaisa Men as such to reproduce
themselves asexually without women, by de-conception and re-concep-
tion rather than by sexual conception, and along strictly male lines of
succession.
To the extent that the giving of ikufuka meat by Akaisa Men is viewed
in the culture in these terms, it relates back to marriage compensation;
specifically, to a third personal exchange of valuables from groom's father
to bride's father mentioned briefly in Chapter 6. In opposition to the two
public marriage-compensation payments, this third marriage transaction
198 Quadripartite structures
is done privately (onge oake) by the two fathers in hiding. Similarly,
although the whole tribe looks on during feasts and installations, the
giving of ikufuka by one Akaisa Man to another is a private and personal
act by comparison with the collective exchanges of the remainder of
pange foods at mortuary feasts. Moreover, to the extent that the giving of
ikufuka meat legitimates the titles of heirs to Akaisa Man status through
de-conception and re-conception, it is very much like a marriage
compensation.
Thus, with pange feast prestations there are three transactions that are
isomorphic with the three transactions of marriage-compensation ex-
change. The various parts of pange food lost or given at mortuary feasts
symbolically represent different categories of blood, within the mem-
bership of a clan, that have been acquired cognatically from other clans
through women by marriages contracted in preceding generations, or, in
the case of Akaisa Men, in the same generation. Akaisa Men unam-
biguously represent in their re-conceived persons the pure agnatic blood
of their respective clans. They are closed to female bloods of other clans
by de-conceptive and re-conceptive feasting in their own lifetimes. Ordi-
nary clanspeople must wait until they die for final de-conception, or their
relatives die for re-conception, but then it is only partial because their
mothers' bloods are not then lost. Paradoxically, however, Akaisa Men
share a pure quasiagnatic blood connection among themselves, even
though they may belong to otherwise agnatically unrelated moieties.

The mythical Akaisa

To the extent that peace chiefs and the other three categories of Akaisa
Men are males lacking female blood and cognatic relationships, they call
to mind the strictly male characters of the Foikale and afinama myths
(Chapter 5). Foikale and his comrades initially had no females, yet re-
produced themselves asexually by shedding their skins. Later, they ac-
quired women and all other aspects of sweet village life, including sexual
reproduction, from Oa Lope. In the afinama myth, when women became
flying foxes and left their husbands and sons alone in the village, males
were symbolically de-conceived and re-conceived to their original
Foikale-like condition. Unquestionably, these two myths are transforma-
tions of the Akaisa tale. All three stories deal with the relations of village
and bush, male and female, male relationship and female relationship,
and agnation as opposed to cognation and affinity in the context of clan
exogamy and sexual reproduction. Among the neighboring Roro peoples
described by Seligmann (1910:304-9), Akaisa is named Oa Lope. This
confirms that the character, Oa Lope, in the Foikale myth, who gave to
purely agnatic bush males village life and women, is the same character,
Akaisa, who gave villagers both the Akaisa Men, including peace chiefs,
 Feasts of death (ii) 199

and death. In other words, Akaisa-Oa Lope mythically bestowed upon

groups of agnatically related men both women, by which the men could
sexually conceive or reproduce themselves and their clans, and Akaisa
Men, by which they could inversely de-conceive themselves of female
derived bloods and re-conceive or reclaim agnatic purity in distinction
from other clans.
It was mentioned previously that men and patrilines of various male
ritual specializations may change their clan affiliations by means of par-
ticipation in feasting obligations with a clan they are not agnatically relat-
ed to and thus become one clan with them. There would not then neces-
sarily be any temporal stability to intraclan relations without identifiable
and hereditary relations between different clans at each status level (Levi-
Strauss 1969b:77-8), and vice versa. Certain relations within and be-
tween clans, therefore, must be held constant across generations. And
this is what I have tried to show in this and the preceding chapter as
regards the formalized hereditary friend relationships between peace
chiefs and their subclans in the contexts of both exogamic intermarriage
and reciprocal mortuary feasting, and as regards the homologous distinc-
tions between subclan peace chief, peace sorcerer, war chief, and war
sorcerer:
papie n aun a
clan of feast receiving :: Z Z VaVie
, j : r• i of clan of : of feast-receiving (21)
deceased mends , , r. ,
deceased friends
peace chief : war sorcerer :: war chief : peace sorcerer (15)

own clan : spouse's clan :: mother's clan : V » i (16)

* mothers clan
closing opening opening closing . ,
male male female ' female
. ., ., inverted everted ,,
inside : outside :: ., : . j (4)x
outside inside.

From the perspective of ordinary villagers, their moieties, clans, sub-

clans, and lineages are both open and closed at any one moment as
regards the intermingling of cognatic and female-related bloods through
conception and the heterosexual recruitment of agnates. But from the
perspectives of the four categories of hereditary Akaisa Men, the
moieties, clans, subclans, and lineages are perpetually closed, and their
boundaries do not overlap. In this dynamic balance, the continuity of the
society and its distinctive quadripartite plan or structure through time are
assured.
 9
Tikopia and the Trobriands

The foregoing chapters have been devoted to interpreting diverse con-

texts of Bush Mekeo social life and to building in the process a model of
the total culture. With the ethnographic treatment of the indigenous
conceptualizations of death and mortuary feasting, that goal is now ren-
dered virtually complete. The structure that characterizes Bush Mekeo
culture overall, as I have shown, is consistently quadripartite in form.
However, the implications for comparison and culture theory that arise
from this derivation have yet to be considered. It is to these complemen-
tary issues that this and the concluding chapter are respectively ad-
dressed.
For the present purpose of comparison, I have chosen the cultures of
Tikopia and the Trobriand Islands. These two unquestionably represent
the most extensively documented and well-known ethnographic cases in
the Oceanic sphere, if not in the entire anthropological record. Firth and
Malinowski, the original ethnographers, are both regarded even today as
among the most meticulous observers and recorders to have ever lived.
Their classic functionalist descriptions have been supplemented by more
modern and diverse interpretations, some but not all of them sympathetic
with a general structuralist perspective like my own. And particularly in
the case of the Trobriands, the original ethnographic corpus has been
substantially augmented and clarified with more recent firsthand field-
work.1 For these several reasons, then, Tikopia and the Trobriands each
represents a preeminent challenge for structuralist interpretation along
the lines of my analysis of Bush Mekeo culture.
Most importantly, however, I have selected these two cultures for
comparison because they are linguistically and historically related to that
of the Bush Mekeo. Bush Mekeo, Tikopians, and Trobrianders are alike
Austronesian speakers, and in many respects their cultures embody cog-
nate conceptualizations of many domains. Therefore, by these com-
parisons I do not need necessarily to make assumptions about the char-
acter of the Human Mind. Although these could well be made, many
more exhaustive studies of total cultural and social systems such as I am
suggesting here would first have to be completed. And in the meantime,
200
 Tikopia and the Trobriands 201
there remain several compelling epistemological reasons concerning the
analytical validity of indigenous categories and cultural wholes that grant
a relative autonomy to cultural structures separate from mental configura-
tions due to the unique involvement of the former with history and social
process. These will be discussed at some length in Chapter 10. Therefore,
I disavow any charges of "mentalism" that might arise from misunder-
standing the fundamentally cultural basis of these explorations.
By the same token, I do not restrict my comparisons to principally one
context or domain of these cultures - say, male initiation ritual (Allen
1967), sex roles (Meggitt 1964), or leadership types (Sahlins 1963). Nor do
I assume one context of any culture to be necessarily dominant, as, for
example, structures of exchange (Rubel and Rosman 1978). Both these
other strategies tend to involve essentially the same problems charac-
teristic of many historicist efforts discussed in the concluding chapter;
namely, it is oftentimes falsely assumed a priori that certain specific sec-
tors of different cultures are either uniformly real isolates, or, alter-
natively, more significant and determining than the others. Rather, as I
have consistently tried to show for the Bush Mekeo, it is the structure of
each total cultural system that is itself dominant over the content of any
one of its own constituent domains. Also in the cases of Tikopia and the
Trobriands, these specific structures again appear to be systematically
quadripartite. Lastly, I must emphasize that these comparative exercises
must be seen as provisional or suggestive "sketches" only and as strictly
exploratory. They are not at all intended to serve as definitive surrogates
for fuller, more detailed, and exhaustive analyses on the scale of my own
handling of Bush Mekeo culture. A more thorough treatment of these
classic systems understandably lies well beyond the limitations of this
work.

Tikopia
A fair amount of interest has recently developed in the structuralist in-
terpretation of Tikopian culture (Levi-Strauss 1963b; Leach 1962b; Eyde
1969; Park 1973; Hooper 1981). From his own functionalist perspective,
Firth (1969, 1981) has of course resisted these efforts, so that the situation
here, as with the Trobriands corpus, has become quite complicated. Nev-
ertheless, the careful juxtapositioning of indigenous categories and rela-
tions pertinent to many diverse but interconnected contexts of the culture
- for example, conception theory, rites of passage, clan and political
organization, and ritual authority and activity - reveals a pervasive quad-
ripartite ordering consistent with a structure of bisected dualities.
Tikopian views on conception seem partially to invert those of the Bush
Mekeo. In Tikopia, the father supplies the substance or material incorpo-
rated into the fetus. The Female Deity, the apotheosis of womanhood
202 Quadripartite structures
inside the mother's womb, "shapes" or gives "form" to the father's semen
to make the fetus. Thus, the male component of conception can be seen as
"matter," the female component as "form." Additionally, Firth in his own
rendering states that the fetus is so formed into four categories of body
parts: limbs, head, body, and genitals (Firth 1981:51-2, 1936:479-81).
The same bodily distinctions seem also to be preserved upon the recrea-
tion of Tikopian souls, following their mythical annihilation, as they are
transformed from amorphous blood again into atua spirits by the Female
Deity after death (Firth 1967a:338).
The pairings of male with female and matter with form are represented
in other distinctly social contexts of the culture. The most salient social
units - clans, lineages, and households - are essentially constituted of
shared patrilineal connections to common male ancestors. But clans, lin-
eages, and households are also articulated to one another according to
female-based or matrilateral ties. These I shall describe in more detail
momentarily. Thus, it would appear that although the dominant groups of
Tikopian society consist of identities of shared male substance (semen),
the integration of those units into a total, ordered, and integral body or
system (namely, all of Tikopian society) with four broad members or parts
(the clans) is accomplished by the conceptualization of structured rela-
tions along female ties. The indigenous theory of conception that counter-
poses matter and form, in other words, is metaphorically represented in
the intra- and interclan arrangements, respectively, that mark the society
as a whole (cf. Firth 1936:575).
In traditional times, the four clans (kainanga) were nonlocalized, non-
exogamous, and ranked in the order of Kafika, Tafua, Taumako, and
Fangarere. Fangarere, although the lowest in rank, is nonetheless closely
associated with Kafika, the highest. Tafua and Taumako, both of inter-
mediate rank, are correspondingly associated with one another (Firth
1936; Leach 1962b:275). Hooper has assembled the various mythico-ge-
nealogical "charters" of these interclan relations with the discovery that
Kafika stands as "mothers brother" (tuatina) to the significant ancestors of
the other three clans. The ancestor of Tafua clan is sister's son of the chief
of Kafika and the son of Kafika "female chief." The Taumako ancestor is
sister's son of a subsequent Kafika chief, but not also the son of a chiefly
Kafika woman. And the ancestor of Fangarere is son of a Kafika woman,
but his own father and others of his clan had been decimated through
misfortune; therefore, he was adopted by Kafika (Hooper 1981:17-22).
Again, relations between patrilineal units of Tikopian society are formed
or shaped by mythical matrilateral ties to the ranking clan. Although
these myths do not deal in all respects with the founding of the modern
clans nor with the clans themselves and all their separate subdivisions,
they do serve as a "charter" for interclan relations, and in particular for
 Tikopia and the Trobriands 203
the spiritual and ritual predominance of Kafika over the other clans (cf.
Firth 1981:61-2).
Each of the four clans possesses its own hereditary "chief" (ariki) re-
sponsible for performing ritual and political activities on behalf of mem-
bers of their respective clans and, to varying degrees, the wider commu-
nity. In this latter respect, the clans and chiefs are associated with a
somewhat pronounced specialization of function. The chief of Kafika dom-
inates the religious ritual performed on behalf of the entire society. More
secular or civic activities are regulated by the chief of Tafua. Taumako
clan and its chief are associated with fierceness and aggression and have
specialized in warfare. Finally, the chief of Fangarere clan, consistent
with mythical heritage, is connected with public misfortunes - epi-
demics, hurricanes, canoes lost at sea, etc. (sorcery?) (Firth 1936,
1961:28-31, 1967b, 1970; cf. Hooper 1981:28-38). There are, again, four
essential administrative tasks, each predominantly managed by the chief
of one clan. It could be argued that there is here a distinct complemen-
tarity of function in terms of homologously bisected dualities:
religious religious secular secular
administration : administration :: administration : administration
of peace of misfortune of peace of misfortune
Kafika : Fangarere :: Tafua : Taumako

In the traditional Tikopian social order, there are groupings and politi-
co-religious statuses other than clans and chiefs that also appear to be
significant. Overall, the total society exhibits four distinct levels or scales
of segmental integration. First, and most comprehensively, all of Tikopia
is bound into a single community of kin, based on the ideology of shared
cognatic ancestors. Second, Tikopia is segmented into the four non-
localized patrilineal clans, each with its own chief to perform the various
ritual, political, and economic functions discussed above. Third, each clan
is segmented into patrilineages (paito), each headed by its own elder.
Members of a lineage, however, do not typically reside under a single
roof, but rather live in scattered households consisting of a fraternally
based joint family. The household represents the smallest-scale segmen-
tal unit of Tikopian society, for a total of four such levels of integration
(Goldman 1970:372-5).
One further, somewhat different, realm where a similar ordering seems
also to be clearly evident in the culture is in the division of the island into
districts. Leach first called attention to the issue of the four districts -
Uta, Namo, Faea, and Ravenga - which, he claimed, are related, but not
directly so, to the four clans (Leach 1962b:275). Firth has responded that
the distinctions among districts vary with context and are a historical
coincidence, unrelated to the number of clans. However, his explanation
204 Quadripartite structures

encompasses a complex and unexpected, but nonetheless quadripartite,

ordering among district units:
For social purposes generally, only two districts, Faea and Ravenga, are spoken
of, in constant relations of cooperation and competition. In the most inclusive
sense, Ravenga is a broad name for a unit of this name plus Uta, Namo and Tai,
all of which can be treated separately for land holding and economic purposes,
while for ritual purposes Uta and Namo may have separate status and be ranked
with Faea and Ravenga in complex contrasts. (Firth 1981:48; emphasis added)
Although Faea and Ravenga are, therefore, seen in social contexts as
binarily opposed, there are in others - economics, land tenure, and ritual
- clear signs of quadripartite district totalities: Ravenga plus Uta, Namo,
and Tai, on the one hand, and Uta plus Namo, Faea, and Ravenga on the
other. Moreover, Firth notes that although the simpler dual-district
structure "fits ritual clan pairings in contexts other than local group refer-
ence" (Firth 1981:48), in contexts of dart matches, dance festivals, ritual
foods, head and body symbolism, relations of restraint and freedom in
kinship, and alternation in human and spirit worlds, the pairing of clans
referred to here is double, meaning therefore a total of four clans. There
does seem to be, then, a structural convergence of four districts, variously
defined according to different contexts, with the existence of four clans
and the quadripartite or double-paired patterning of other domains of the
culture. However, this does not necessitate a substantive convergence
such that any one district corresponds exactly in personnel with a particu-
lar clan. It is this that Firth appears to insist upon, but it is nonetheless an
altogether different issue. Rather, it is the structure ramifying across
contexts that seems to be significant.
As mentioned in an earlier paragraph, chiefs are not the only culturally
important functionaries. Again, we find a complete fourfold division of
politico-religious authority. Patrilineages are first distinguished in terms
of ariki (chiefly) and non-ariki (nonchiefly). Each of these two categories is
subsequently bisected again to produce a total of four types of official. In
chiefly lineages, there are the ariki chiefs themselves, and the maru
"executive officers" of the chiefs who are their agnates and held responsi-
ble for social control (Firth 1976:289, 1970:36). Among the non-ariki
lineages, there is a corresponding distinction between purefai kava "ritu-
al elders" who are entitled to perform the kava ceremony, and those pure
elders who perform the same secular functions as the others but who are
not entitled to serve kava (Firth 1970:57-8).
Other religious rituals performed by chiefs and elders on various occa-
sions on behalf of their respective groups embody strikingly similar con-
figurations. The kava ceremony, being fundamental to them all, is there-
by most instructive on several counts. Tikopians divide the kava rite
wheresoever it is performed into four elements:
 Tikopia and the Trobriands 205

1. Laying out pieces of bark-cloth (te mano e fora).

2. Recital of a formula, using root and stem of a kava plant as ritual apparatus (te
kava e tar6).
3. Libations of cups of kava liquid (te kava e ringi).
4. Throwing away portions of food and betel materials from specially prepared
packages or bunches (te kai e pe). (Firth 1970:200)

There are also four categories of participants - officiant, kava maker,

cupbearer, and audience or assembly, present or not, on whose behalf the
ritual is being performed (Firth 1970:202, 223). Moreover, the officiant
recites the formulas and offers the libations in a definite order to four
categories of god and ancestral spirit - to the officiant's principal god, to
other gods prominently involved in the rite, to the officiant's predeces-
sors, and to the matrilateral tutelaries (Firth 1970:222). Food offerings are
frequently made in conjunction with kava ritual, and again the pattern of
four is apparent. For the banana-ripening kava ceremony and the ritual of
the sacred canoes, for example, offerings of fourfold portions of food are
systematically displayed before the gods and ancestors and then con-
sumed (Firth 1970:227-8; 1967b:105, 115-16, 132-3). Numerous other
instances of quadripartite orderings of things and sequences are man-
ifested throughout Tikopian ritual (see Firth 1967a, 1967b, 1970).
Elsewhere, Firth provides evidence that intergroup relations at the
level of lineages are also fourfold. Each Tikopian man is principally con-
nected with four such patrilineages:

A man 'belongs' to his father's group, is closely linked with his mother's group,
cooperates with his wife's brother's group, and owes other obligations to his
sister's husband's group, that is, his sister's son's group. Here the man is en-
meshed in a web of which the strands are four independent patrilineal kinship
groups. (Firth 1936:372; cf. Eyde 1969:48)

I suggest that the fact that there are four such groups so linked in every
man's personal network is not coincidental.
Even before a Tikopian reaches maturity and marries, and then also
after his or her death, the same quadripartite social network is evident,
although somewhat different groups are involved on separate occasions.
Postnatal ceremonies for a firstborn child mark "the creation [formation?]
of new kinship groups' with respect to the child. The four significant
parties in this instance include the fare nana (female cognates of the
child's father), the fare masikitanga (female cognates of the child's father),
the child's tuatina (mother's brothers), and men of the child's own or
father's patrilineage (Firth 1967a:46-59; see also Eyde 1969:50).
At male initiation, four groups or categories are distinguished for each
initiate. There are two groups of "those who perform " the ceremony -fax
matua people of the initiate's father's group, and fax tuatina people of the
206 Quadripartite structures
initiate's mother's brother's group - and two groups of "cooks" (fai soko)
- the husbands of female members of the initiate's group, on the one
hand, and their sons, the sister's son's of the initiate's group, on the other
(Firth 1936:433, 551; cf. Hooper 1981:3). It is noteworthy also that during
the initiation ceremony, mother's brother performs the superincision op-
eration upon the initiate's penis, thereby qualifying him for sexual rela-
tions and eventually marriage. Homologous with the Female Deity's pro-
creatively forming the body of a fetus into four members, the mother's
brother forms the social identity of his nephews vis-a-vis his own and
other groups prerequisite to manhood and marriage (cf. Hooper 1981:2-
13).
At the marriage ceremony, there are again four clearly distinguished
categories of persons between whom there are a complex series of trans-
actions. Here, however, the points of reference for the various parties are
the husband and wife as a unit together and, presumably, their children-
to-be. There are (1) the kinship group of the husband, (2) the "cooks"
(men married to women of the husband's group, and men married to
daughters of women of the husband's group), (3) the kinship group of the
wife, and (4) the kinship group of the wife's mother's brother (Firth
1936:544-63, 1939:322-4; cf. Eyde 1969:56-61).
In mortuary ritual as well, the same sort of general pattern of exchange
seems to obtain, but not as distinctly as in the other contexts, due in part
to the sheer complexity of the proceedings and to the admitted in-
completeness of Firth's data in this case. The formal presentation essen-
tially involves reciprocities between the various participants as mother's
brother and sister's son. A classification of the entire mortuary body into
four categories oipaito lineage or family does nonetheless emerge. There
are, first, tau pariki, "mourners" (the family of the father of the deceased,
and more distant kin); second, tama tapu, "sacred children" (children of
the women of the mourners' group and their fathers, the "cooks"); third,
tukunga tanata, "burial party " (family of the mother of the deceased); and
fourth, tangi soa, "spouse-lament wailing group" (family of the wife of the
deceased) (Firth 1939:324-31, 1936:213).
In these several pronounced life crises of Tikopians, then, the pattern-
ing of four social groupings to constitute in each case a complete set
underscores the pervasiveness of quadripartite categorical structuring. In
following with classic formulations of rites of passage in terms of death and
rebirth, as suggested by Hooper (1981:11), these Tikopian life crises
would be perhaps better understood as a series of ritual reformulations of
the fourfold social identities of the respective initiates (see also Chapter 1;
Van Gennep 1960; Turner 1967; Bloch and Parry 1982).
Eyde has revealed other instances of quadripartite ordering in Tikopian
culture principally dealing with social space. The plan for a typical Tiko-
 Tikopia and the Trobriands 207
pian house, he notes, consists of a rectangle divided into a "left,"
"female," "profane" half and a "right," "male," "sacred" half. Each of
these halves furthermore possesses two corners and two ranked posts.
The world to Tikopians, like their houses, is also divided into quarters
marked by four major wind directions, each the home of an important
clan god and source of a seasonal wind. These winds are further associated
with male and female deities as they receive their devotions in the season-
al ritual cycle of the Work of the Gods. Marae, the ritual dance place at
the center of Tikopia, is laid out in quadrilateral form like houses and like
the world. At the corners, backrests or seats are positioned for the prin-
cipal gods of the four respective clans. On most occasions, they serve as
the stationing points for the chiefs and their clansmen. On others, when
women participate, people of the same sex gather on opposed sides -
women on the side of Kafika and Fangarere, and men on the side of Tafua
and Taumako (Eyde 1969; cf. Firth 1969).
In Eyde's handling, these instances of patterned spatial ordering are
shown to be isomorphic with social arrangements - household, lineage,
and, most suggestively, for the society as a whole. For example, at the
rebuilding of Kafika temple, the two clans paired together on Marae as
"female" (Kafika and Fangarere) serve as "cooks" to the "worker" clans of
the "male" side on Marae (Tafua and Taumako). From these contrasts and
other evidence, Eyde draws the outlines of a model of restricted marriage
alliance between the mythical and legendary figures of the chiefly lin-
eages of the four clans consistent with the prevailing ranking of Kafika
first, Tafua second, Taumako third, and Fangarere last2 (Eyde 1969; cf.
Firth 1969:68-9).
Levi-Strauss (1963b:24-9) has suggested that other dimensions of Tiko-
pian culture are similarly classified in a quadripartite, "totemic" manner.
The four principal edible plant species - yam, coconut, taro, and
breadfruit - are each associated with one of the four clans in order of rank.
So, too, are particular species offish or eel (Firth 1970:175-8; Hooper
1980:30). These foods are representative of the respective clan deities
(atua), especially those of the chiefs. Their classification, moreover, ex-
hibits both an additional extension of the body metaphor and the specific
structure of bisected dualities. Yam and taro, for example, are the
"bodies" of the deities of Kafika and Tafua, and breadfruit and coconut are
the "heads" of the Taumako and Fangarere deities, respectively. Also,
whenever a dolphin is stranded, it is divided into four parts and appropri-
ately allotted; Kafika and Tafua are entitled to the two middle sections of
meat, and Taumako and Fangarere to the two extremities (see also Firth
1930).
Firth, of course, has tried to refute the various suggestions such as these
that Leach, Eyde, Levi-Strauss, and others have forwarded, arguing, for
208 Quadripartite structures
example, that the existence of four clans and other "fours" in the culture is
not symbolically necessary or significant. He cites the case whereby the
chief of Tafua clan converted to Christianity in the mid-1920s (Firth
1981:48-9). Previous to 1925, or thereabouts, there existed a "double
pairing" in the ritual exchanges of Marae between Kafika and Taumako, on
the one hand, and Tafua and Fangarere, on the other. But after the chief of
Tafua clan converted, Firth argues, the others resisted the possibility of
hiving-offsome major lineage segment to form a fourth clan. Moreover, he
points out, the ritual double pairing was preserved without such corre-
spondence to clan numbers or boundaries. Essentially, Fangarere con-
tinued as one clan for social purposes, but split to perform ritual re-
ciprocities within itself so that the critical Kafika-Taumako pair could
remain intact.

So while the dual pairing concept remained of prime importance for ritual ex-
change in this religious context, it was four major transactional units that were
needed, not necessarily four clans. For a quarter of a century, while the four clan
structure operated in general social, economic and political contexts over the
whole community, in this particular ritual context the pagan Tikopia got along
very effectively with only three clans. (Firth 1981:48-9)

However, the pagan Tikopians still got along only with a total of four
ritual categories of participants. Furthermore, I think it significant, con-
trary to Firths perspective on this fact, that Tikopians did not create a
new clan, for a total then of five clans. Whether or not Tafua participates
in the pagan ritual with the others, it still constitutes the fourth clan of
Tikopian society for social, economic, and political purposes. This quad-
ripartite form or structure of social classification is in each context the
same, and essentially so, evidently, even if the exact content of the result-
ing substantive or categorical divisions are not always perfectly coinci-
dent. Here, as elsewhere, Firth would seem to be insisting upon too strict
a series of convergences of content before he will grant any of these
quadripartite formations analytical validity. Nonetheless, Firth himself at
one point concludes, "Leach's view of the significance of this number four
can be supported to some extent if it he looked at as a double pairing'
(Firth 1981:48; emphasis added), or, in my terms, a fourfold structure of
bisected dualities. In any case, I suggest that it is focusing upon the
homology of form in the interrelations among the categories of different
contexts rather than the literal coincidence of the categories themselves
that consistently provides the most effective way of conceptualizing the
total integration of the culture. This issue, of course, goes well beyond the
interpretation of any one culture, and I will return to consider it again at
length in Chapter 10.
 Tikopia and the Trobriands 209

The Trobriands
The ethnographic situation for the Trobriand Islands has become consid-
erably more complex than with Tikopia. On the one hand, several eth-
nographers have gone back to the Trobriands and collected additional
firsthand information expanding on Malinowski's corpus (Powell 1960,
1969a, 1969b; Uberoi 1962; Montague 1971, 1973, 1983; Weiner 1976,
1978, 1979, 1980; Hutchins 1980). On the other, a considerably larger
body of commentary and reinterpretation has been generated by a sub-
stantial community of anthropologists, including many who have not
themselves done Trobriand fieldwork (Mauss 1967; Leach 1958a, 1961,
1966b; Robinson 1962; Levi-Stauss 1963a; Sider 1967; Tambiah 1968b,
1983; Eyde 1976; Brunton 1975; Tooker 1979; MacDougall 1975; Keesing
1981). We are left with an embarrassment of riches, but also in many
instances a multiplicity of conflicting ethnographic facts and interpreta-
tions. In my own efforts to reveal the quadripartite conceptual classifica-
tions of Trobriand culture that seem to be significant, I hope to make a
step toward clarifying at least some of the complicating issues that seem to
mark the literature so dramatically in this case. Again, this analysis is to
be understood as neither exhaustive nor definitive, but instead as explora-
tory and suggestive. I shall examine the following contexts of Trobriand
culture: the indigenous notions of the person, procreation, alimentation,
the category father, clan organization and interclan relations, rank and
chieftainship, marriage and mortuary exchange, kula and wasi exchange,
and gender symbolism.
The Trobriand conception of the person is elaborated in several inter-
connected ways. These all exhibit, however, a consistent scheme of classi-
fication according to logical structures of bisected dualities. In Mal-
inowski's original rendering, the age distinction of child and adult
crosscuts that of male and female to produce a fourfold division of any
village population into "male child" (monagwadi), "female child' (inag-
wadi), "man" (tau), and "woman" (vivila). Each of these statuses, howev-
er, is further subdivided into four developmental stages, underscoring
the Trobriand cultural concern for growth (Weiner 1976:20-2). A male
child begins life as a "fetus" (waywaya) and then progressively develops
into an "infant" (pwapwawa), "child" (gwadi), and "male child" (monag-
wadi). For a man, his "youth" (toulatile) is followed by his being a "ma-
ture man" (tobubowa'u), a "married man" (tovavaygile), and an "old man"
(tomwaya). Female children similarly progress from being fetus, infant,
and child to female child; and women from "girl" (nakapugula or
nakubukwabuya) to "ripe woman" (nabubowa'u), "married woman"
(navavaygile), and "old woman" (numwaya). Thus, there are four major
age-gender categories, and within each of these four, assuming the per-
210 Quadripartite structures

son lives long enough, four subdivisions of age and growth correlated with
gender (Malinowski 1929:60; cf. Leach 1958a; Powell 1969b).
According to Montague's recent inquiries, Trobrianders in the process
of growing into complete adult men and women are understood to devel-
op powers or capacities that relate them to their respective spheres of
activity in life, both productive and reproductive. The adult male is con-
stituted of a solid or hard bodily strength with an internal power or "noise
force' (miegava). The qualities of a complete adult woman, by contrast,
consist of a "quasiliquid" or "squishy' body with an external store of
power (doba fiber skirts and banana-leaf bundles). The strength of a man's
body allows him to perform the substantial tasks of gardening, and his
noise force enables him to control nonsubstantial gardening activities
such as weather and garden "magic" (see Eyde 1983). Alternatively, with
her soft body a woman can substantively give birth, and with her store of
doba wealth she can give nonsubstantial birth or rebirth to living people
in mourning (Montague 1983).
Thus, among adults of both genders, Trobriand villagers seem to exhib-
it four powers or capacities - male substantial and nonsubstantial, and
female substantial and nonsubstantial. Viewed in this way, adults of op-
posite sex are inverted or reversed analogues of one another. Structurally
similar formulations of Trobriand gender ideas have been proposed by
Tooker (1979) and Tambiah (1983; see below; cf. Weiner 1976, 1977,
1978). Therefore, in these terms the Trobriand system of adult gender
roles is structured much like that of Bush Mekeo villagers, according to a
bisected duality.
The Trobriand view of procreation has, of course, generated one of the
most colorful controversies in all of anthropology. It will be necessary for
me to touch on only the more salient points here to reveal a heretofore
unrecognized quadripartite dimension to the indigenous theory of
conception.
Malinowski's original report was that a subclan ancestor of a woman
transformed into a spirit-child (waywaya) enters the woman's body and
animates her menstrual blood (agu buyavi), which in turn constitutes the
bodily substance of the child (Malinowski 1929:173-95). As Leach (1966b)
points out, Malinowski subsequently qualified this declaration of the ig-
norance of physiological paternity with the recognition of a rather more
"complicated attitude towards the facts of maternity and paternity. Into
this attitude there enter certain elements of positive knowledge, certain
gaps in embryological information. These cognitive ingredients again are
overlaid by beliefs of an animistic nature, and influenced by the moral and
legal principles of the community . . . " (Malinowski 1932:31). Subse-
quent inquiries by Austen (1934), Powell (1968), and Weiner (1976, 1978)
have also revealed complementary paternal as well as maternal contribu-
 Tikopia and the Trobriands 211
tions: that through sexual intercourse with a man, a woman's vagina is
"opened"; that a mans semen (momona) "checks" or "stops-up" a wom-
an's menstrualflow;that a man's sexual activity or semen "forms," "coag-
ulates," "shapes," or "moulds" the mother's menstrual blood into the
fetus (see Leach 1966b; cf. Spiro 1968).
These data suggest that a father's procreative influence as "co-contrib-
utor" involves two components that parallel the two maternal compo-
nents. Each parent, in other words, contributes a material and a non-
material element. A child receives substantive menstrual blood and a
nonsubstantive spirit by virtue of its mother. And from its father a child
receives substantive semen in sexual intercourse that, although not itself
incorporated into the body of the fetus, does seemingly help give it its
nonsubstantial form and appearance (see Weiner 1976:122-3, 1978; cf.
Eyde 1983; Tooker 1979). It should come as no surprise, then, that com-
pletely formed persons, whether male or female, in the Trobriand view
possess both substantive and nonsubstantive bodily components, as de-
scribed above. Also, Powell's report - that the complementary emphases
upon mother's blood and spirit-child, on the one hand, and father's semen
and appearance or form, on the other, are contrasted respectively as
"men's talk" by one's matrilineal relatives and "women's talk" by pa-
trilateral relatives - suggests that the disputed issues of maternity and
paternity are seen by Trobrianders themselves according to the interests
of the parties involved (Powell 1968; see also Austen 1934:107, 153). It is
not unreasonable to assume maternal kin would emphasize their own
interests in a child, perhaps to the exclusion of the interests of the child's
paternal kin (and vice versa). Why the maternal perspective is paradox-
ically described by Powell as "men's talk" and the paternal as "woman's
talk," I cannot now explain (Powell 1968). However, as I move to consider
other areas of Trobriand culture, additional male /female dimensions of
contrast will be shown to correspond similarly with distinctions of sub-
stance and nonsubstance, form and spirit, as well as temporary and
permanent.
It should be noted here with this quadripartite Trobriand theory of
conception that the male and female components of structure or form and
amorphous matter, respectively, parallel the cultural associations of gen-
der in conception for the Bush Mekeo and invert those of the Tikopia,
although these latter two cases are both generally identifiable as patrilin-
eal in terms of social organization. Moreover, it seems in this view, fol-
lowing Leach (1966b), that Trobriand conception theory is not the sort of
ethnographic anomaly it has been for so long perceived to be. Rather, it is
a clear and systematic transformation of other seemingly less exotic and
less incredible views of gender complementarity, reproduction, and so-
cial organization in other traditions. Specifically, in this context, the
212 Quadripartite structures

quadripartite Trobriand view of procreation has been shown to be struc-

turally isomorphic with the analogous views of related Pacific cultures.
What would seem to be equally anomalous to Westerners, but, curi-
ously, has not created the same sensation, is the reported Trobriand
assertion that "eating is not regarded as indispensable to life" (Malinowski
1929:441). I suggest this item of belief is related to, and can illuminate
further, the indigenous notions of procreation as just discussed. As Mal-
inowski notes,
in part they have no idea that there is such a thing as physiological need for
alimentation, or that the body is built up on food. According to them, one eats
because one has appetite, because one is hungry or greedy. The act of eating is
very pleasant, and it is a suitable expression of a joyful mood. Large accumulations
of food, their formal distribution (sagali) and, at times, their immediate, though
not public, consumption form the core of all native festivities and ceremonies
. . . Yet meals are never taken in public, and eating is altogether regarded as a
rather dangerous and delicate act. (1929:441)
For Trobrianders, food has definite associations with the role of father.
A father nurses his infant children by giving them mashed food during the
time they are suckled by their mothers (Malinowski 1929:20, 207-8;
Weiner 1976:123-30, 1978:178). The food a father gives then and later is
taken from the subsistence gardens of his own labor (Weiner 1976:125).
The child's outward physical appearance or shape continues to be
moulded by the fathers food as well as by the names and decorations the
father gives it and the beauty magic that its father's sisters bestow
(Weiner 1976:127-9, 133, 1978:182). Ideally, also, the land upon which
adult men grow food is acquired for use through the father (see below).
Clearly, the reported ambivalence Trobrianders have for food and eating
recalls contrasting perspectives of the fathers role in conception. Food,
associated with the father, is like semen. It, too, moulds, forms, or struc-
tures the physical appearance of his children, but it is not necessarily
taken up into the blood or flesh of their bodies. And neither would food
seem to communicate eternal baloma or spirit. Indeed, as I discuss be-
low, when the body dies and loses its shape, the structuring paternal
influences of semen, sexual intercourse, food, magic, and so on are
negated.
The construction of a satisfactory model of Trobriand social organization
has proven to be nearly as elusive as a convincing rendering of the indige-
nous theories of conception and alimentation. The question of the four
"clans" (kumila), first seriously raised by Leach (1958a), persists as partic-
ularly puzzling. According to the most recent authoritative source, "in
fact, why four clans, instead of two or three or even eight, remains
problematic" (Weiner 1976:60). In trying to solve this riddle, I shall
 Tikopia and the Trobriands 213
generally follow Leach in arguing that "Trobrianders need four categories
to display the workings of their society, and that the four matriclans fulfill
this purpose" (Leach 1958a: 141). I would qualify this, however, to the
effect that it is not only four clans that fulfill this purpose, but rather the
structure of bisected dualities that homologously divides many other con-
texts of the culture as well.
Trobriand Islanders classify themselves and all of humanity into four
named, ranked, exogamous, and nonresidential categories or matrilineal
clans. All members of a clan apparently never assemble for corporate
activity. Indeed, all that seems to bind them together is the shared identi-
ty of a name and certain totemic prohibitions (Malinowski 1929:496-503,
1948:112-3). The mythological "charter" for the relations between the
clans concerns the emergence of their respective totemic animals from
the hole, Obukula. First, iguana, representing Lukulabuta clan,
scratched the ground from underneath and surfaced. The next to emerge
were dog (Lukuba clan), pig (Malasi clan), and finally the animal of Luk-
wasisiga clan, variously identified in different versions of the myth as
crocodile, snake, or opossum. Initially, dog was the "chief," or animal of
high rank (guyau). But although pig and dog played together, dog
smelled, licked, and ate a bush fruit likened to excrement. Malasi thereby
became chiefly and higher ranking than Lukuba and the other clans (Mal-
inowski 1929:498-9; Weiner 1976:52).
In spite of its wide currency in anthropological circles, until now it has
remained an unremarked ethnographic fact that this myth embodies a
number of categorical bisected dualities. There are, for example, two
animals - pig and dog - that are both mammals of the aboveground
associated with high, chiefly status, and two others - iguana and snake or
crocodile - that are subterranean reptiles associated with low, "com-
moner" status (tokay). Mythically speaking, there are thus two chiefly and
two commoner clans. Also, the two commoner animal species and reptiles
(excepting opossum) are associated in Trobriand culture with immortal
spirits (baloma) of the nonsubstantial underworld and with women,
whereas the chiefly mammalian species are associated with mortal human
ancestors that emerged to the substantial world aboveground and with
men (Tambiah 1983). Moreover, despite Malinowski's own perplexity
(1948:113), the logic of the exact sequence of emergence seems to be
significant. Of the first pair, it is the second animal (dog) rather than the
first (iguana) that initially assumes the rank of chief. Subsequently, how-
ever, the first animal (pig) of the second pair is associated with high rank
and takes the status of chief from dog. In other words, the opposition of
the /irs£-to-emerge - commoner versus second-to-emerge-chief for the
first animal pair is resolved through its inversion in the second pair -
214 Quadripartite structures
first-to emerge-chief versus second-to-emerge-commoner. The formula of
bisected dualities thus sheds new light on an old problem.
Montague (1971) has examined further the articulation or intersection
of the clan principle in Trobriand society with the principle of rank. She
notes Malinowski's observation of three 3 ranked classes - guya'u,
"chiefs"; gumguyau, "nobles"; and tokay, "commoners" - as well as a
fourth category, the "pariahs" of Bwaytalu, Ba'u, and Suviyagila villages
(Montague 1971:355; Malinowski 1929:30, 499-500, 1922:67; see also
Malinowski 1966:385; Powell 1960:129). There are here obviously two
high-ranked and two low-ranked classes; and they crosscut distinctions
between clans. Although the data are still fragmentary, it appears that
different subclans of the same clan are ranked in different classes, and that
subclans of different clans are, or can be, ranked in the same status. These
four rank divisions evidently play a role in marriage regulation. Although
marriage is ideally exogamous for clans, excepting chiefly officeholders, it
is preferentially endogamous as regards rank (Malinowski 1929:80, 83,
457-8). Chiefs and village cluster leaders (see discussion below), howev-
er, by virtue of polygynous unions may marry both exogamously and
endogamously as regards rank class. Thus, in the traditional society, there
appear to be four ranked classes with at least some significance for mar-
riage regulation, just as in the same respect there are four clans.
The quadripartite division of Trobriand society is also reflected in
Leach's classification of localized subclan hamlets (katuposula). He argues
that the kinship nomenclature consists of categories reflecting non-
genealogical criteria of residence rules, age and generation status, and
yam-exchange obligations.4 Earlier in the chapter, I noted the signifi-
cance the child/adult distinction has in Trobriand culture, and it is im-
plicit also here. From the viewpoint of the male child, the social world is
divided up into two categories: "outsiders" and "people like us." The
former includes the males of two subclan hamlets related to Ego as are
father's father and father's sister's husband; the latter includes members
of Ego's own (father's) hamlet of domicile and mother's brother's hamlet.
There are as a result four types of hamlet managers or landowners where
Ego has and has not rights to land, and four types of landowners'
daughters, both marriageable and nonmarriageable. Upon his becoming
an adult, marrying, and assuming his own yam-exchange obligations, a
man's social universe contains again four distinct types of subclan hamlets:
(1) tama hamlets, and (2) kada, tuwa, and bwada hamlets ("people like
us"); and (3) affines' hamlets and (4) tabu hamlets ("others") (Leach 1958a;
cf. Lounsbury 1964, 1965; Sider 1967).
Focusing particularly upon the potentials for hierarchical relations in
obligatory yam exchanges, Powell also reveals a somewhat different but
nonetheless structurally similar quadripartite portrayal of the relations
 Tikopia and the Trobriands 215
between groups, which he claims is symbolized instead in the four clans.
He takes for his social units corporate but not necessarily co-residential
subclans (dala). The spectrum of relations between subclans, whether
they be of the same or different clans, is "equality, super-ordination, sub-
ordination, and non-interrelation" (Powell 1969b:596-7). These distinc-
tions, moreover, have their counterparts in the indigenous scheme of
social classification. Thus, similar again to Leach, Powell elsewhere re-
ports a quadripartite distinction between (1) tomakava, the hostile "unre-
lated"; (2) veivai, affinal competitors "on the other side"; (3) veyo, com-
petitive matrilineal kin "on the same side"; and (4) tabu, matrilineal and
affinal nonrivals through whom ritual rivalries are established (Powell
1969a: 199).
According to somewhat different criteria still, Weiner has isolated a
subtle and multiple taxonomy of relational categories that embodies addi-
tional significant quadripartite social divisions. In one perspective, the
Trobriand universe is first bisected into tomakava, nonclanspeople or
"people different from us," and veyola, clanspeople or "people like us."
Among the former, some are lubela or lugebu, "origin nonclanspeople"
(i.e., people of other clans whose founders mythically emerged from the
same hole); the remainder are simply tomakava, nonclanspeople whose
ancestors did not emerge together. Among clanspeople, on the other
hand, there is a parallel distinction between veyola, origin kin whose
founders emerged together, and kakaveyola, nonkin whose ancestors
emerged separately (Weiner 1976:53-4). From the perspective of Ego,
however, relationships are classified into four categories according to
mutually inclusive and exclusive criteria. First, Ego is identified by blood
and spirit with a particular subclan (veyola tatola or veyola mokita). Sec-
ond, Ego possesses a larger circle of clanspeople corresponding with
origin kin. Third, all other clanspeople are designated as kakaveyola,
nonkin. And fourth, all members of the other three clans are identified as
nonclanspeople (tomakava) (Weiner 1976:53-5).
Alternately informed by the intersection of the same /different clan and
class distinctions discussed above, Montague has presented the Trobriand
social universe as divided in terms of the kinds of interrelations that can
exist between subclans, and, again, "there are four clear possibilities: 1,
two dala [subclans] share clan identity but not class rank; 2, they share
rank class but not clan identity; 3, they share neither; and 4, they share
both" (Montague 1971:362-3). In Powells and Weiner's terms discussed
above, these distinctions have their terminological counterparts in the
language: kakaveyola, veivai, tomakava, and veyola, respectively (Powell
1969a: 199; Weiner 1976:53-5).
Weiners recent data contain still another indigenous quadripartite so-
cial division of exchange relationships within the Trobriand clan as a unit.
216 Quadripartite structures
Each clan includes true blood kin, origin kin, and nonkin (with whom
there is no exchange relationship) as above, and a fourth category of
clanspeople, keyawa, who are linked by the same affinal relationships to
members of other clans (Weiner 1976:55-63). It might be added that
Weiner explicitly recognizes the potential structural significance of the
congruence here between the four Trobriand clans and these four kinds of
clanspeople, but she hesitates in carrying it further (Weiner 1976:59-60).
Nevertheless, although Leach, Powell, Weiner, and Montague have
each produced a somewhat different view of the Trobriand social uni-
verse, they have in trying to follow indigenous distinctions all found it cut
up in roughly the same way, that is, consistently into four parts. The
pattern of bisected dualities thus appears to be systematically employed
in sorting out and differentiating important social relationships. That any
one of these particular constructions deviates from or distorts native tax-
onomies or fails to represent them accurately is, of course, an empirical
possibility. My own suspicion is that if the ethnography has been correct,
then each of these realms as described does capture a significant but
particular slice of the total meaning of Trobriand culture. Again, further
direct empirical investigation should help answer this question.
Powell's and Weiner's studies have also helped clarify the pattern of
residence and local organization. In most contexts of exchange, as indi-
cated above, the prominent social units are matrilineal dala subclans or
lineages (Malinowski 1929; Leach 1958a; Powell 1969a, 1969b; Weiner
1976). Members of dala possess a common identity based upon shared
descent from the same named ancestral beings (tabu) associated with the
same place of origin, blood, ritual paraphernalia, and lands. Women as
girls grow up in their fathers' dala. Upon marrying, they move to their
husbands'. As boys, men live with their fathers. But in adolescence or
adulthood they may change their residence, depending principally on
whether they are in line to succeed an elder brother or maternal uncle as
subclan hamlet manager. If this is not the case, the proper place for a man
to live is at the home of his father (Powell 1960, 1969a, 1969b; Weiner
1976:154-5).
Trobriand villages (valu) consist of one or more than one section or
subclan hamlet. Members of a subclan are recognized as "owners" at the
village of their subclan hamlet, whether they individually reside there or
not. The corporate affairs of a subclan are directed by the subclan "man-
ager. ' When there are two or more owning subclans in a village, one is
always recognized as senior in status, and its manager is headman for the
whole village (see my qualifications discussed later in the chapter). De-
mographically as well as politically, two or more villages are congregated
into "compound village communities " or "clusters." Evidently, all four
clans are represented in the populations of separate clusters. This is, I
 Tikopia and the Trobriands 217
think, significant because marriages tend to be endogamous to the cluster
both in frequency and in explicit preference. Of all the subclans within a
village cluster, some are of high or chiefly rank (guyau), and others are of
low or commoner status (tokay). Among the owning subclans of the clus-
ter villages, the highest ranking (always of guyau status) is recognized as
chiefly and ascendant over all the others. The successful manager of the
highest ranking of these chiefly subclans serves as acknowledged leader
for the entire village cluster. Finally, traditional Trobriand society is
divided into eight named "districts." At least in the northern districts of
Kiriwina and Tilataula, it seems, certain chiefly subclans that rank highest
within their own clusters enjoy a certain ascendancy over other clusters of
their respective districts. Their leaders are also recognized as district
chiefs (Malinowski 1922:49-62, 1929:7-20, 1965:3-12, 345-69; Powell
1960; Weiner 1976:141-67.
Overall, then, it appears there are a total of four recognized hier-
archical levels of territorial and political integration: subclan hamlet, vil-
lage, village cluster, and district. And each level has a corresponding
category of political authority: hamlet manager, village headman, cluster
leader, and district chief.5
This particular hierarchical classification embodies a number of features
that would appear to be homologous with critical distinctions of the myth
of clan emergence discussed earlier in the chapter. According to Powell,
district chiefs and village-cluster leaders and their subclans are always of
chiefly or guyau rank, but otherwise for villages and subclan hamlets
their headmen and managers might just as likely be of tokay commoner
rank. Thus, as among the clans in the myth of origin, there are two
categories of high or chiefly rank and two categories of low or commoner
rank. The myth then serves as a "charter" not only for the existence of
clan and rank distinctions; it "charters" also the overall political integra-
tion of Trobriand society. Moreover, according to legend, there were
originally eight chiefly subclans, each having subclan segments residing
elsewhere in the islands (Weiner 1976:44). And it seems also there were
originally eight chiefs (Weiner 1976:45). Powell's report that the Trobri-
ands are divided into eight native districts thus takes on a new signifi-
cance (Powell 1960:221).
As mentioned previously, the important social units for marriage are
exogamous subclans and the largely endogamous village clusters.
Weiner's refinements of the interconnected rules of marriage and yam
harvest and mortuary exchanges reveal additional evidence of heretofore
unrecognized, but nonetheless highly significant, quadripartite pattern-
ing in these contexts. First, Weiner clarifies the much-discussed stipula-
tion of "patrilateral cross-cousin marriage" as rather a strategic preference
between persons and their tabu, that is, with a person of a different
218 Quadripartite structures
subclan but the same clan as Ego's father. Marriages that conform to this
rule (for either a male or female Ego), and a preponderance do, stablize or
renew important keyawa exchange relationships between subclans of the
same clan. As Weiner observes, neither she, Malinowski, nor Powell
recorded a single instance of marriage with "own fathers sisters
daughter," and in the figures she gives there is apparently no noteworthy
incidence of, or advantage to, marrying into father's own subclan. There-
fore, if Ego's spouse is of a different subclan but same clan as Ego's father,
fundamentally four subclans, two in either of two clans, complete the
necessary affinal set (Weiner 1976:187). The relevant social universe of
each Ego in Trobriand society as deduced by the rules of marrage as
stipulated so far necessitates the existence of clans, as Weiner points out,
but also involves the pairing of clans, each of which is bisected again to
produce four subclans in all (Weiner 1976:52-60, 174-87, 1979:343-5).
A related quadripartite classification of subclans is identifiable in the
distribution and juxtapositioning of keyawa and non-keyawa relationships
between members of both the same and different clans, as a result of
marriage to a person of father's clan (tabu), consistent with the rule or
preference cited above. On the basis of Weiner's material, these distinc-
tions play a critical role in the organization of mortuary distributions and
the consequent regulation of overall social continuity. Among one's own
clanspeople, there are some who are related as keyawa and others who
are not. Similarly, among nonclanspeople, there are keyawa and non-key-
awa. The same-clan/different-clan contrast, in other words, is bisected by
that of keyawa/non-keyawa (Weiner 1976:55). As a result of marrying
tabu, another quadripartite social division is generated.
A degree of uncertainty still remains in the marriage rules, however,
that might well have profound implications for overall societal integra-
tion. It remains unclear, for example, whether a brother and sister can
both simultaneously marry tabu and into the same other subclan or clan.
Indeed, if every adult, male and female, did so marry, then the system
(excepting the polygynous marriages of chiefs and cluster leaders) of pa-
trilateral second-cousin marriage, as it has been depicted, would in actu-
ality be one of symmetrical instead of asymmetrical exchange (cf. Weiner
1976, 1979). As I turn to the examination of marriage exchanges and
mortuary rituals, I shall focus upon additional ethnographic data that
partially aid in resolving the implicit contradictions here with the rules of
marriage in a way still consistent with the overall structure of the culture
in quadripartite terms.
By her efforts to demonstrate that the relation of "father" (tama) to
"child" (latu) is one of patrifiliation as well as affinity, Robinson
(1962:127-37) shows that in the complicated series of exchanges following
marriage there are four principal parties - two representing the groom,
two representing the bride. For the groom, there are the father and
 Tikopia and the Trobriands 219
mother's brother supported by their respective subclan members, and
similarly for the bride (see also Weiner 1976:177-8). Malinowski dis-
cusses eight named marriage gifts between the bride's "family" and the
groom's "family." The first three of these he lists (katuvila, pepei, and
kaykaboma) are given by the bride's family to the groom's family. These
latter people reciprocate with two more gifts (mapula kaykaboma and
takwalela pepei). Then the bride's family contributes the first harvest of
yam food (vilakuria) to the groom's family. Finally, the groom's family
gives fish (saykwala) and valuables (takwalela vilakuria) to the bride's
family (Malinowski 1929:89-94). There are altogether, then, four group-
ings of marriage gifts. The bride's family gives two, and the groom's family
returns two (Malinowski 1929:89-94). Moreover, of the eight named ex-
change categories, there appear to be four that are essential - (1) pepei or
kaykaboma, (2) vilakuria, (3) saykwala, and (4) takwalela pepei or tak-
walela vilakuria (Robinson 1962:136; cf. Eyde 1976:244) - two of which
are either given or received by either of the exchanging parties.
From Malinowski's and Robinson's discussions of these transactions, it
is clear that some of them are understood as directly reciprocal for one
another. However, Robinson observes that there are long-term obliga-
tions implicated outside this immediate exchange nexus that serve to link
up the various kinship groupings across generations. And it appears to be
the case that there are four of these to the set. First, the bride's family gift
of vilakuria yams constitutes the initial annual yam exchanges. Second,
the groom's father gives valuables in takwalela to his son's wife's family in
indirect reciprocation, third, to his own wife's family for having sent him
annual yam prestations throughout the course of his own marriage. And in
the process, too, the groom's father ensures on behalf of his son that,
fourth, he will also receive yam gifts from his wife's relatives, thereby
obligating his son to contribute valuables at the marriage of the son's son
(1962:153-5).
These particular long-term reciprocities are entirely consistent with the
way marriages between tabu are ordered among subclans of the same and
different clans, as portrayed above in light of Weiner's material. Through
his own and his sister's marriages, a man plays complementary exchange
roles with respect to yams and valuables in relation to members of two
different affinal groups. To his "own' sister's husband's subclan he will be
(1) a giver of yams and (2) a receiver of valuables. Correspondingly, to his
wife's subclan he will be (3) a receiver of yams and (4) a giver of valuables.
The same roles would be replicated in form for other annual marriage
exchanges of yams and valuables (Malinowski 1929:121-9, 1966; Weiner
1976:146-53). As a result of the asymmetry of these transactions, there
are a total of four categories of marital exchange role. I shall show below
how this particular quadripartite structure of reciprocity is also evident in
other contexts of Trobriand exchange, notably kula and wasi.
220 Quadripartite structures
The annual yam payments themselves constitute a major and dramatic
part of the native horticultural economy. Upon closer examination, yet
another fourfold distinction is evident. For any particular exchange of
yams, there are two categories of giver and two of ultimate receiver.
Generally, the annual yam exchanges are corporate subclan affairs (Powell
1969b: 184-7). However, particular men are recognized as the principal
givers to particular women on behalf of their whole group. A woman and
her husband receive two major gifts of yam, from (1) her "own" brother
and his wife, on the one hand, and (2) her father and mother, on the
other. The couple given this food (3) shares several baskets of it (kovisi)
among the husband's classificatory subclan sisters prior to having the
remainder stored in his own yam house (Weiner 1976:149, 199; Mal-
inowski 1966:189). Finally, the husband will (4) contribute the yams he
has stored to help support hamlet and village collective activities (e.g.,
leaders' "feasts of merit, " canoe building, mortuary distributions) (Powell
1969b:584).
The yams for these particular transactions are taken from only one type
of food garden, kaymwila. Trobrianders plant three other types as well.
First, a man will plant kaymata, the larger main garden, for the hamlet
manager who grants him rights to land at the village of his residence.
Second, a man cultivates a number of taro gardens, gubwauli, the plants
of which he will give to his kinswomen. And third, a man produces the
principal foods for his family's subsistence in gubakaueki gardens (Weiner
1976:137, 140, 146, 199-202, 1978:178; cf. Malinowski 1966:58). Thus,
there are four categories of garden: two kinds for exchange yams, and two
kinds for other produce and subsistence.
The same pattern appears in still other dimensions of gardening. Re-
gardless of whether the separate types of plant food are for exchange or for
subsistence, yams at least are additionally distinguished in terms of four
stages of growth: "garden plots, " "ownership, " "at yam house," and "bas-
kets of yams" (Weiner 1976:140-5). Yam storehouses (bwayma) in the
village, of course, have four sides (Malinowski 1966). And Eyde has com-
mented upon the significance of the quadrilateral structure of poles (kam-
kokola) erected in gardens as well as the rectangular layout of the gardens
themselves (Eyde 1983; Malinowski 1966). Altogether, there are several
separate sets of indigenous quadripartite categorical distinctions in the
context of food production, exchange, and consumption.
I now turn to the examination of Trobriand mortuary ritual. For here I
believe the inner dynamic of the culture's pervasive structure is most
evident. In the course of this examination, I shall propose a somewhat
novel interpretation of the rites that is nonetheless not inconsistent with
the several instances of quadripartite ordering I have already discussed
with respect to indigenous notions of the person, procreation, marriage
rules, and yam exchange.
 Tikopia and the Trobriands 221
For Malinowski, "the whole mortuary ritual [was], in fact, perhaps the
most difficult and bewildering aspect of Trobriand culture" (1929:148).
He thus left only the barest outline of the customary proceedings. Quite
fortunately, Weiner (1976, 1979) has supplied a recent and full account
based on her own fieldwork in Kiriwina. Her fundamental argument6 is
that a villager's death suddenly disrupts interrelations among the relatives
of the deceased. Trobriand mortuary ritual and exchange therefore serve
to restore and regenerate those ties that death had so abruptly severed
(Weiner 1976:21-2, 219). My own interpretation is rather the inverse of
this: Although death does represent a rupture of social ties among the
living, it is only the beginning of a process of social dislocation. The
involved series of mortuary rites serves not to restore those relationships
but instead to complete their total dissolution. Eventually, however,
those ties should be renewed, but with the creation of new marriages
among the survivors rather than with mortuary exchanges. I shall show,
then, that Trobriand mortuary ritual predicates an undoing of social rela-
tionships initially created by exogamous intermarriage that is analogous to
Bush Mekeo mortuary de-conception. A close reading of the eth-
nographic material will bear this interpretation out.
According to Weiner, the critical mortuary exchange is the woman's
ceremony, termed lisalabadu. This is one of several types of sagali dis-
tribution. The word sagali means "to divide," "to settle accounts," or "to
reclaim ownership" (Weiner 1976:62, 1979, 1980). At the women's cere-
mony, there are two main social categories: givers and receivers. Ideally,
each of these two categories is broken down into two subcategories.
Givers thus include "owners" (toliuli, women of the deceased's subclan)
and their clan keyawa, and receivers include "workers' (toliyouwa,
spouse of the deceased and spouse's subclan kin, and deceased's fathers
subclan kin) and their respective clan keyawa (Weiner 1976:74). Con-
sistent with the preference for marrying someone of father's own clan, the
women's mortuary ceremony is characterized as an exchange between
members of two distinct clans, each minimally represented by two sub-
clans (Weiner 1976:63, 1978:180, 1979:343). However, as it turns out, the
personnel who give and receive overlap considerably. Owners' brothers'
wives, brothers' children, and kin of the owners' fathers may help in the
giving, and spouses and children of the male owners and workers may
well receive (Weiner 1976:63, 74, 104-15). Again, consistent with the
avowed preference for patrilateral second-cousin marriage, at least two
clans will thereby be represented.
Nevertheless, other data suggest that the number of distinct clans is
necessarily or ideally more than two. As part of the tadabali distribution
that precedes the women's mortuary ceremony, women's wealth in the
form of bundles of banana leaves is given to the male subclan kin of the
deceased's spouse and their married and unmarried children. Persons of
222 Quadripartite structures

these two categories will continue to observe the signs of mourning until
the women's mortuary ceremony. They are, of course, members of differ-
ent clans. They receive their bundles from two other categories of persons
who are also of different clans from one another - clanswomen of the
deceased, and daughters of the deceased's subclansmen. According to
Weiner, both kinds of person who receive bundles on this occasion are
also nonclanspeople of the deceased's clanswomen (1976:74). Thus, it is
certain that there are at least three clans represented. And although the
details are not complete, there is some implication that the daughters of
the deceaseds clansmen who give bundles represent members of an-
other, fourth, clan, assuming they did not themselves marry a man of
their father's clan (1976:75; cf. Weiner 1979:343-5).
What is needed to resolve this issue, as well as several others, as I
noted earlier, is further clarification of the idealized rules of marriage with
patrilateral second cousins and of the manner by which the resultant
social categories enter into mortuary ceremonial. In the meantime, how-
ever, it appears that there is some evidence for a full complement of
members from all four clans in Trobriand mortuary rituals.
Indeed, this seems to be the case at Omarakana traditionally. During
major public ceremonials - including mortuary rituals whereby the obli-
gations incurred by several deaths are fulfilled - the members of the
representative clans pair off into two dyads:

The clan most numerous in the local population would be teamed with the least,
and the remaining two would be teamed together also. In the Omarakana cluster
the Malasi were grouped in this way with the Lukulabuta as the most and least
numerous, and the Lukuba with the Lukwasisiga, giving a fairly even division of
the population into "moieties." I was assured however that the clans would line
up differently elsewhere, and I have no other evidence of any moiety organiza-
tion. (Powell 1969b:603n)

His final comment notwithstanding, this double pairing of clans at

mortuary ceremonials cannot but be of significance, given the added
evidence of bisected dualities in this and other contexts of the culture. It
is interesting to note, incidentally, that at Omarakana, at any rate, Malasi
(pig) and Lukulabuta (iguana) represent the first-to-emerge of both the
first and second pairs of animals in the myth of clan origin, and Lukuba
(dog) and Lukwasisiga (snake/opposum/crocodile) represent the second-
to-emerge of the two origin pairs. Moreover, the Malasi-Lukulabuta and
Lukuba-Lukwasisiga pairs each mythically includes one chiefly and one
commoner clan. And further, in terms of the corporate status of the
residential groups, each of these "moieties" at Omarakana contains one
high-ranking or chiefly subclan - the Tabalu of Malasi clan, and the
Bwaydaga of Lukwasisiga clan. Members of the other thirty-seven sub-
 Tikopia and the Trobriands 223
clans are of commoner status (Powell 1960:124; Malinowski 1929:30-4,
1965:86).
At their lisalabadu mortuary exchanges, the owners with their helpers
"reclaim" elements or artifacts (blood, identity, personal names, coconut
and betel palms, decorations, rights to land) that had gone from their
subclan to members of other subclans through the deceaseds relations
with those other persons - the workers and those who receive with them.
On the day of the women's mortuary ceremony, some sixteen7 different
kinds of transactions are accomplished (Weiner 1976:105-15). The first
twelve are marked off as a set in that bundles of dried banana leaves are
given, but no colorful fiber skirts. The receivers of the bundles in all these
exchanges are being compensated for various contributions and services
they have bestowed upon the deceased and owners, and for the relin-
quishment of subclan rights the deceased had invested in them before
death. The exchanges of bundles appear to culminate in kaymelu, which
is the most "important" of the lisalabadu transactions. Here, the female
owners give bundles to the wives of all their kinsmen who have ever given
them raw yams and taro.8 Men's sisters thereby reexpress the prior rights
of the subclan to the property or artifacts of their brothers that they had
loaned (mapula) to their children by virtue of working and living with
them and their wives. This is entirely consistent with Weiner's specifica-
tion of the meaning of sagali as "to reclaim," based upon a second trip to
the Trobriands (1976:246, 1979).
The three remaining "important" ceremonial exchanges (for a total of
four such) in the concluding set of the women's mortuary ceremony in-
volve men's wealth (axe blades, shells, and clay pots) and colorful wom-
en's skirts that have been made from old and dirty leaf bundles. In ka-
lilakuvili, the women, who had played the role of giver at some point in
the proceedings but who belong to different clans and thereby also played
receiver, have the signs of mourning removed from their persons. These
are replaced by the colorful skirts that signify release from mourning
(Weiner 1976:112). Next, in kalakeyala kakau, clanspeople of the de-
ceased carry wealth to the deceased's spouse. Women carry colorful skirts
again, and men carry male wealth. The skirts are distributed to the
spouse's kinswomen, and the male valuables are taken by the hamlet
manager of the spouse's subclan. And last, in kalakeyala kapu, women's
skirts and male wealth are given to the kinswomen and hamlet manager of
the deceased's father (Weiner 1976:113-14).
The quadripartite ordering of these most important mortuary rituals is
revealed even more clearly when the cultural and social contexts are
examined more closely. Of these four transactions, the separateness and
integrity of the clan of the deceased vis-a-vis other collectivities are
graphically enacted four distinct times, but in different ways. First, in
224 Quadripartite structures
kaymelu, the wives of men who have previously contributed plant food
are distinguished from those men's sisters and kinswomen. In ka-
lilakuvili, second, women who gave bundles in kaymelu who are not of
the clan of the deceased are now given colorful skirts by women who are
of that clan and thereby so distinguished also. Third, in kalakeyala kakau,
the deceased's clan is differentiated from the clan of the deceased's
spouse. And finally, the clan of the deceased is contrasted with that of the
deceased's father in kalakeyala kapu. Overall, as well, the clan identities
of villagers who still have viable ties, either patrilaterally or affinally to the
clan of the deceased in spite of the deceased's death, are emphasized in
the first two major exchanges, and the clans of villagers who, because of
the death of the deceased, no longer have viable relations with the de-
ceased's clan, are demonstrated in the latter two major exchanges. And
finally, in the first pair of important transactions (kaymelu and ka-
lilakuvili), the giving and receiving parties begin with mixed clan mem-
bership and then culminate in the latter pair (kalakeyala kakau and ka-
lakeyala kapu) with parties of unmixed clan membership. Incidentally, it
is probably significant here that the women's wealth given for unmixed
villagers is colorful skirts made anew from old and dirty leaf bundles.
Moreover, the women's mortuary ceremony is held some six or eight
months after the death - roughly the same amount of (reverse?) time that
Trobriand pregnancies last (Weiner 1976:62; Austen 1934:108-10).
Although I have differed with Weiner as regards some of the details in
certain contexts, she is generally correct insofar as the men and women of
the subclan of the deceased "reclaim" their dala identity and property in
mortuary ceremony, as far as it goes. Those elements or artifacts that had
been temporarily lost or loaned to the subclans of other clans over the
lifetime of the deceased by virtue of marriage and patrifiliation are re-
turned or claimed upon the deceased's death in mortuary exchange
(Weiner 1976:21-2, 55-7; cf. Weiner 1976:81-90, 116-20, 163, 246,
1978:175, 180-3, 1979:334-5, 345, 1980:81). But those members and
hamlet managers of other subclans who had been so blessed by the de-
ceased simultaneously surrender or "de-claim" those same rights. The
unambiguous and systematic boundaries and divisions of all the various
subclans and clans represented are thus publicly expressed.
Indeed, the very integrity of the society's major component units is in
this way reaffirmed. Far from bridging across or mending the ruptures
created by death, mourning ritual completes the process. Trobriand soci-
ety thereby appears to be structured in a manner analogous to the patri-
lineal Bush Mekeo. And in that ordering, mortuary ritual evidently plays
a similar role, that is, of de-conception. In the Trobriands, exogamous
matrilineal clans and subclans create ties with others and, indirectly
through these, with subclans of their own clan. These ties are initially
 Tikopia and the Trobriands 225
contracted as marriage relationships, but ultimately they produce viable
intergenerational patrilateral, male, or cognatic connections across clan
and subclan boundaries. In a persons procreation and growth and in the
contribution to these that a person makes in other persons, personalized
elements of subclan essence and property are distributed along both
matrilineal and patrilateral lines. When a person dies, the members of
his/her subclan and clan retake possession of that which, although it was
corporately theirs, they had lost in the life of the deceased. The mourners
and other relatives of the deceased in other subclans and clans correspon-
dingly relinquish that which they had acquired rights to in the life of the
deceased but was not corporately theirs. In death and the performance of
mortuary ritual, the components of a person that had been mixed in
relationships to persons of other subclans and clans are unmixed and
returned to the subclan and clan of their origin. The overall result, of
course, is to give the total society and its constituent social units a pro-
found and distinct symbolic sense of continuity, or, as Weiner has chosen
to describe it with a somewhat different meaning, a sense of social re-
production (cf. Weiner 1976, 1977, 1978, 1979, 1980).
Trobriand mortuary ritual, then, undoes what was done in earlier mar-
riage exchanges. As with the Bush Mekeo, marriage and affinity translate
nonrelatives into affines and cognates, and death and mortuary rites
translate kin back into nonrelatives. But for those intraclan keyawa and
interclan relations to be perpetuated beyond this, the cycle must be
repeated from the start. New marriages must be negotiated. In this re-
spect, my interpretation differs substantially from Weiner's.
However, she herself observes that, even if the full series of mortuary
transactions is properly concluded, a collective relationship between the
ceremonial givers and receivers (i.e., beyond the bounds of merely per-
sonal ties) can be maintained only if those ties are entirely recreated with
altogether new marriages generative of new relationships of cognation
(Weiner 1976:57, 81, 119, 163, 1978:182-3, 1979:345, 1980:81; see also
Tambiah 1983:173). Hence the rule or preference for marrying tabu, or
persons of fathers clan. And in this same vein, Powell has observed that,
with a death of a chief, the affinal and political relations between groups
are abolished unless they can be successfully reestablished through the
contending heir's individual marriages (Powell 1980:131).
In this view of Trobriand social continuity, the interrelationship of
kinship, marriage, and mortuary institutions involves a balance of sub-
tractive as well as additive processes. This follows from the underlying
fourfold structure of bisected dualities. But for the "long run" (Sahlins
1981), a more consistent balance of factors is envisioned and maintained,
that is, consistent logically and with the data at hand.
Indeed, numerous other ethnographic imponderabilia can now be
226 Quadripartite structures
made intelligible. First, in mortuary ritual, there are two kinds of article
for persons of each gender given by owners to workers. Women give raw
unmanufactured leaf bundles, and from them processed or manufactured
fiber skirts are made and given again. Men by comparison give raw yams,
and, from the obligations that arise from these, manufactured veygua
valuables are given. The raw bundles of banana leaves are themselves
further classified in terms of a separate fourfold ordering. There are old
and new, and clean and dirty, bundles. For men's manufactured wealth,
there seem also to be four main varieties: shell armbands, shell necklaces,
clay pots, and axe blades (Weiner 1976:91-5, 104-15). But it is more
significant, I think, that the logic of four clans appears to be directly tied
in with the doings and undoings of marriage and mortuary exchanges.
However, until the ethnographic details of the marriage rules are clar-
ified, this cannot be concretely demonstrated. Still, the recognition that
mortuary ceremonies undo what is done in marriage and procreation
helps to solve further the riddle of supposed "Virgin Birth." Although
males may make contributions of semen and form to their offspring, these
are of mere short-run significance. Given that the bodies of their children
eventually die, the contributions of fathers over the long run are negated.
The matrilineal spiritual and substantive essences of mothers, subclans,
and clans can be thereby claimed without contradiction as the only lasting
and persistent or real contributions. Again, similar to the Bush Mekeo
case of a seeming conflict of same versus different blood relationship,
Trobrianders can assert that in life there is a continuity of form and
appearance of fathers and children ("women's talk") along with that of
blood and spirit between mother and child ("men's talk"). With death, it
is only the latter that persists (Weiner 1978:182; cf. Weiner 1976:122,
193).
The treatment accorded a deceased male's yam house and his bones
now also become intelligible. Upon his death, a man's yam house -
symbolic of his position in life - is disassembled, signifying the dissolution
of his personal network of relationships to members of other subclans and
clans (i.e., father's, wife's, and sister's husband's relatives; hamlet manag-
er; village headman; etc.) (Weiner 1976:179, 215). Even more curiously,
the sons of a male deceased suck the fluids from the bones of their father's
exhumed corpse in the process of its decay. Before distributing them to
their own relations, including the widow, and also to the deceased's
father's relatives, the sons clean the bones of flesh in the sea - a place
associated with maternal baloma spirits (Tambiah 1983). From here, the
spirit of the deceased is understood to travel across the water to Tuma,
the land of the dead. Then the deceased's clanswomen wash the sons'
mouths and purify their hands (Malinowski 1929:155-7; Weiner 1976:81-
4), perhaps symbolically "reclaimirig" the deceased's blood and flesh on
 Tikopia and the Trobriands 227
behalf of his subclan (Malinowski 1929:155-7; Weiner 1976:81-4). By
their acts, the sons seemingly help "dry" the bones and separate the wet,
formless, bloody parts from the dried, formed parts. Quite explicitly,
Trobrianders reason that the deceased's sons perform these disagreeable
operations because their father fed them mashed foods and cleaned away
their feces and urine when they were infants (Malinowski 1929:20-1,
156-7). These latter paternal tasks contribute to the child's formation that
had begun in procreation. In their father's death, the sons would appear
to be inversely reciprocating the service. Fathers control the formation of
their children, and the children control the deforming of the father. In
either case, the paternal or male relationship is concerned with the di-
mension of form as distinct from spirit or blood.
In the months or years of mourning, the relatives of the deceased who
are bound by relations other than shared subclan identity (i.e., through
males as opposed to through females) wear the dissembled bones. That
the body of the deceased is indeed understood to be undergoing a process
of deformation (along with the collectivity of survivors) is partially ex-
pressed in the bones being frequently passed from one relative to another
(Malinowski 1929:149; Weiner 1976:81-4). After the concluding sagali
exchange, the fully dried bones are entombed in a cave or grotto associ-
ated with the deceased's subclan, and particularly its ancestral baloma
spirits (Weiner 1976:84; Tambiah 1983). Just as the deceased's bones are
given a final rest, relations between the deceased's subclan kin and non-
subclan relatives are thereafter abrogated, unless, of course, they are
renewed by the creation of new social ties through males. Incidentally,
this precise interpretation of the rites of the bones as well as my general
perspective on Trobriand mourning ritual receives additional and power-
ful theoretical support in light of Hertz's classic formulation (1960). The
soul, the body, and the society of survivors experience parallel journeys
after death - in this case, a fate of separation, disintegration, de-concep-
tion, and reclamation or return to the sources of origin.
In mortuary ceremonial and in the several other relevant contexts of
the culture I have examined so far, ties through males have systematically
complemented ties through females. Although the constituent descent-
based units of the society - subclans and clans - are conceptualized in
terms of matrilineally transmitted blood and spirit essences, the rela-
tionships that link those diverse units are structured according to shared
cognatic and affinal interconnections, that is, through males (Montague
1971:357). A symbolic dimension of Trobriand political organization,
heretofore unrecognized, linking fathers, hamlet managers, cluster lead-
ers, and chiefs, thereby suggests itself. I outlined earlier in the chapter
the organization of hamlets, villages, village clusters, and districts, largely
on the basis of Powell's material. He viewed the village (valu) as the
228 Quadripartite structures
principal residential unit (Powell 1960, 1969b). On the basis of her more
recent data, however, Weiner has convincingly argued that the significant
co-residential collectivity in most contexts is rather the subclan hamlet
(katuposula), of which there may be one or more than one in any particu-
lar village (1976:38; Leach 1958a). I wish to suggest that, in relation to
their respective followings, hamlet managers along with village-cluster
leaders and district chiefs (but not village headmen) constitute meta-
phorical representations of the Trobriand notion of "father." There would
be then a total of four such types of " father" in Trobriand social
organization.
This is particularly evident in the various food-exchange contexts of the
society. I have already described how food is central to the conceptualiza-
tion of the relationship of the father to his children. The relation of a man
to the manager of the hamlet where he resides is similar. Unless he is the
likely heir to the manager of his own subclan hamlet, a mans proper place
of domicile is at his fathers hamlet. There he would be entitled to plant
his first yam-exchange garden (kaymata). The produce of this first garden
is consequently given to a mans own father.9 In this way, a man initially
establishes rights to use that land. Annually thereafter, however, he vali-
dates his rights to garden and house land by making a yam-exchange
garden (kaymata) for the hamlet manager who controls that land, or for
his own father as intermediary. Of course, it is the hamlet manager who
performs the gardening ritual throughout the growing cycle on behalf of
all the resident males. The hamlet manager thus stands in the relationship
of "father" to the resident males who make gardens for him (Malinowski
1966; Powell 1960, 1969b; Weiner 1976:42-3, 140-56).
Cluster leaders and district chiefs are also given yams by the members
of subclans of other clans. Because of their chiefly guyau status, leaders
and chiefs are entitled to marry polygynously. Leaders tend to select their
wives from the subclan hamlets of other clans represented in their own
clusters. Chiefs acquire the majority of their wives from subclans outside
their own cluster. Just as in the case of ordinary marriages, the leaders'
and chiefs' wives' relatives are expected to plant special yam-exchange
gardens annually. At least four of each wife's male relatives - her father,
her mother's brother, her brother, and some other relative (unspecified)
- are expected to plant one garden each and to give the produce to the
leader or chief (Malinowski 1929:24-31; Powell 1960:132-6, 1969b:587-
92; Weiner 1976:201-2, 1978:178). These particular prestations to leaders
and chiefs, however, are not classified in the language the same as other
annual yam exchanges (kaymwila) are between nonchiefly affines. Rather,
these are termed kaymata, as are the gardens and produce given to father
and to the manager of one's hamlet (Weiner 1976:42, 140-6, 201-2,
1978:178).
 Tikopia and the Trobriands 229
There is, therefore, a fourfold category of men who receive produce
from kaymata yam-exchange gardens in Trobriand society: tama fathers,
hamlet managers, cluster leaders, and district chiefs. In this view, a chief
or a cluster leader, like a hamlet manager, is not exactly a "glorified
brother-in-law of the whole community" as Malinowski (1966:192) sup-
posed, but a metaphorical father.10 In indigenous terms, the significance
of the societal relations between subclans and hamlets of a cluster and
district is restricted to neither affinity, politics, nor the individual life
cycle, but expressly indicates patrilateral filiation at the collective level
(Fortes 1959; Robinson 1962; cf. Levi-Strauss 1969b; Weiner 1976:20-1).
As with fathers in the contexts of procreation, gender distinction, and
mortuary ritual, the relations of hamlet managers, cluster leaders, and
district chiefs appear to mould, form, or structure the otherwise amor-
phous or unconnected clan bloods of the local populations.
Moreover, these political ties must be regularly maintained and re-
newed, as must paternal ties at a number of levels. In sexual intercourse,
fathers give shape or form to the bodies of their children; hamlet manag-
ers similarly organize and coordinate the relations of the co-resident
males (these latter being of other matrilineal subclans or bloods) as fathers
to sons, just as leaders and chiefs shape or form the relations between the
various subclans and bloods of their children. A father continues to form
or shape his children's physical appearances after they are born by giving
them food; managers, leaders, and chiefs are responsible for sharing food
they have stored with their supporters, particularly on ceremonial occa-
sions. A man's father is his principal source of magic, especially for beauty
(form and appearance); managers, leaders, and chiefs are the sources of
magic that affect all their respective communities (McDougall 1975:61-
2). Following a father's death, ties to his subclan or hamlet manager can
be maintained only if they are renewed by another marriage; with the
death of a leader or chief, relations among subclans of a cluster or district
are maintained only if an heir successfully replicates the marriages of his
predecessor, selecting his wives from the same appropriate subclans.
The structural ties of Trobriand society predicated by hamlet manag-
ers, cluster leaders, and district chiefs as metaphorical fathers of their
followings, in other words, are transitory. This is distinctly similar to the
impermanence of bodily form or appearance that each child acquires in its
conception and lifetime from its father, as contrasted by the permanent
qualities inherited from the mother (blood and spirit). Overall, then, for
Trobrianders as for the Bush Mekeo, the configuration of society is ho-
mologous with indigenous quadripartite ideas about the human body.
Finally, I think it must be significant that two of these kinds of father -
cluster leader and district chief- are always identified with chiefly guyau
rank as distinct from commoner tokay status, whereas the other two -
230 Quadripartite structures
hamlet manager and tama father - are not (see above; Powell 1960;
Weiner 1976:44-6). There exists in this classificatory scheme of meta-
phorical fathers another instance of bisected duality.
Brunton (1975) has proposed that the character of the Trobriand rank
and political system, rare if not unique for Melanesia, is tied to the kula
trade in valuables (vaygua). Although I disagree as to the basis of this
relation, surely it is significant that for every kula participant there are
four temporally and contexually separate roles he must play vis-a-vis his
partners in order to participate effectively at all. With regard to the
exchanges of the shell valuables that have been most extensively de-
scribed (Malinowski 1922), a man will play all the roles of (1) giver of
armshells (mwali), (2) receiver of necklaces (soulava), (3) receiver of
armshells, and (4) giver of necklaces. There are, in other words, four
different kinds of kula exchange occasion for each participant. A man will
play the appropriate type of giver or receiver depending upon which of
his partners he is exchanging with, and where (i.e., at his or his partners
place).
Moreover, it seems that aboriginally there were four kinds of vaygua
valuable: the two types of shell valuable, along with axe blades (beku) and
clay pots (kuliya). The role these latter played in kula unfortunately re-
mains obscure (Malinowski 1922; Weiner 1976:179-83). Of course,
veygua are used also in contexts of marriage payment, reciprocation for
annual yam harvests, and mortuary rituals. It is thus not surprising that
the organization of kula transactions reflects these others.
Furthermore, Powell reports that, in addition to classifying the
armshells as female and the necklaces as males, there may be a subdivi-
sion of armshells (if not necklaces too) into subcategories "male" and
"female" (Powell, quoted in Eyde 1976:245n). The ethnographic details of
kula have not yet been satisfactorily worked out (Leach and Leach 1983),
but quadripartite distinctions in it are clearly prevalent.
Nonetheless, one additional implication of Trobriand kula exchange in
the sense of metaphorical fatherhood seems almost irresistible. Mac-
Dougall (1975) and Tooker (1979) have revealed that kula exchange repre-
sents a metaphorical or quasisexual relationship between men (see Mal-
inowski 1922:81-104; Tambiah 1983). However, high-ranking district
chiefs and village-cluster leaders have tended to dominate Trobriand kula
exchange (Malinowski 1922; Powell 1960, 1969b; Uberoi 1962; Brunton
1975). Indeed, their personal successes partially legitimate their titles to
office. Thus, in their quasisexual kula intercourse specifically, chiefs and
leaders give form to their respective district and cluster followings; that
is, in kula they mould together the blood of different subclans and clans
much as tama fathers in their literal sexual exertions give form to the
blood of their children of different clans.
 Tikopia and the Trobriands 231
Culturally, there is more to the Trobriand chiefly participation in kula
than just politics as such in response to the physical and social environ-
ment (cf. Brunton 1975). In the same respect, too, I would suggest that
hamlet managers' quasisexual actions as garden magicians are critical in
their role of forming or shaping cohesive local followings of men of bloods
of other clans and subclans (Malinowski 1966; Eyde 1983). These meta-
phorical connections of sexual intercourse, kula, and garden activity,
then, broaden the distinctly paternal symbolism of the four types of
Trobriand "father" proposed above - tama father, hamlet manager, clus-
ter leader, and district chief.
Another Trobriand exchange institution that also exhibits a quadripar-
tite ordering is the ceremonial reciprocation between inland and lagoon
villagers of garden produce for fish,11 termed wasi. At harvest time,
inland villagers bring garden produce to their lagoon partners. Sometime
thereafter, the lagoon dwellers go fishing on behalf of the former, who
carry the catch home. Trobrianders wasi in this way only for major sagali
distributions. Thus, the direct recipients of the vegetables or fish do not
consume them themselves, but use the foods they receive to feed others.
There would then seem to be four parties involved in wasi exchanges -
the initial givers and receivers of the fish and vegetables, and those to
whom they are finally presented for consumption (Malinowski 1922:42-3;
see Eyde 1976:243-4).
Virtually all the domains of Trobriand culture I have considered so far
have involved in one way or another the distinction of male versus female.
Even the contrast of wasi fish and vegetable exchanges just discussed
parallels this gender distinction (Eyde 1976). However, these iso-
morphisms rarely appear to be simply binary. Indeed, most of them
suggest more complicated quadripartite orderings. For literal males and
females, as we have seen, the body is bisected differently. Men possess
externally manifested, substantial hardness or strength and an internal,
nonsubstantial noise force, whereas women possess comparatively non-
substantial or squishy bodies and external stores of substantial wealth and
power (Montague 1983). It is the fourfold pattern implicit here that ap-
pears to link together several otherwise disconnected domains of the
culture that have been carefully examined quite recently by others. Be-
cause these contemporary analyses have tended to be extremely in-
volved, I can only refer here to their most salient features.
First, with respect to the plan of Trobriand village organization, Mal-
inowski contrasted the central area with the peripheral street as male to
female (1929:10). Levi-Strauss, however, has pointed out a more "com-
plex [i.e., double] system of oppositions." The outer ring of domestic
houses for commoners is conjoined with the peripheral street, and these
two spaces together are associated with the female, marriage, and cooked
232 Quadripartite structures
food. By contrast, the inner ring of yam houses and the enclosed central
area are jointly associated with the male, bachelorhood, and uncooked
food (1963a:136-7). Keesing has suggested further that at the milamala
harvest festival when the baloma spirits return to the village, these and
parallel relations (regulated sex/unregulated sex, moon/sun, dark/light,
above/below, and so on) are systematically reversed (1981:359-61; Boon
1972:127). A similar ceremonial inversion of categories in fact seems to be
involved in the women's mortuary ceremony (Weiner 1976:104-5).
Investigating the attributions of male and female mythical beings with
malevolent powers, Tambiah (1983) isolates a set of four significant cate-
gories: male sorcerers (bwaga'u), female witches (mulukwausi and
yoyova), nonhuman malignant spirits (tauva'u), and human ancestral spir-
its (baloma). The nonhuman malignant spirits have exchange rela-
tionships with male sorcerers, and baloma spirits are reincarnated
through human females. These parallels are crosscut by space, however,
as male sorcerers and female witches are associated with "aboveground,"
and nonhuman and human spirits with "belowground."
Tooker's (1979) efforts to reveal the links between Trobriand views of
kula exchange with sexual symbolism have yielded a very similar fourfold
classification, such that male sorcery is to kula as female witchcraft is to
procreation. And focusing upon the complementation of the sexes in the
context of principal subsistence occupations, Eyde derives yet another
but convergent set of distinctions: "The ritual of gardening (male) is to the
earth (female) as the ritual of fishing (female) is to the sea (male)"
(1976:246).
I cannot here explore further these very abstract semantic and struc-
tural interconnections. However, as the data and analytic techniques
have improved, the more clearly has the pervasiveness of homologous
quadripartite classification in Trobriand culture become evident.

This comparative digression beyond Bush Mekeo culture into the related
traditions of Tikopia and the Trobriands is merely a first step, but none-
theless a productive one. In spite of the ethnographic differences among
the three, analogous institutions - many of them clearly cognate - have
been shown to be systematically arranged according to essentially the
same quadripartite plan. Moreover, this structuralist perspective on each
culture as a total system of indigenous meanings has shed new light on
many old ethnographic problems. For Tikopia, conception theory; clan
and district organization; ritual specializations and roles of chiefs and
elders; and life-crisis rites are the most noteworthy. And for the Trobri-
ands, conception theory again; gender distinctions; clan and subclan orga-
nization; marriage regulation; mortuary ritual; rank and political integra-
tion and the roles of fathers, managers, leaders, and chiefs; and yam,
kula, and wasi exchange deserve mention.
 Tikopia and the Trobriands 233
Using this approach, it is conceivable that similar results could be
obtained by examining yet other cultures, Oceanic and beyond. Sahlins's
recent reanalysis of his original Fijian material, for example, clearly rep-
resents a fourth case from the Austronesian sphere where significant and
diverse contexts of the culture are ordered according to bisected dualities
(1976:24-46). The cultures of Palau and the Motuans appear also as likely
candidates (Mosko 1980:325-30). And in a study of a different sort based
on linguistic materials, Blust has even generated a model for early Aus-
tronesian social organization that strikingly conforms with the exact quad-
ripartite formula at issue (1980a, 1980b; Goodenough 1955; Fox 1980b).
Finally, outside this range, Hage and Harary (1981) have revealed quad-
ripartite conceptual structures in the culture of the Melpa of the New
Guinea Highlands that are compatible with group theory in mathematics,
as described in Chapter 1 (see also Hage 1976; cf. Rubel and Rosman
1978).
Clearly, that such structural convergences may exist between different
cultures, either related or unrelated, could well have implications beyond
just Oceanic ethnography. And it is to this possibility that I turn in the
next and final chapter.
 10
Conclusions: indigenous categories,
cultural wholes, and historical process

Two kinds of interconnections have been shown in the preceding pages to

ramify throughout Bush Mekeo culture, giving it both meaning and form.
First, categories articulated together in one context (the body, for exam-
ple) are invoked also in others (space and time, adult-male and -female
ritual, social organization, death and mortuary ritual, etc.). Second, the
sets of relations by which the categories of the culture are juxtaposed
uniformly exhibit a structure of bisected dualities. These diverse contexts
thus constitute semantic metaphors and structural homologies of one an-
other (Levi-Strauss 1963a:83). The same may well also be ventured, on
the basis of my comparative sketches, of Tikopia and the Trobriands. The
purpose of this chapter is largely to explicate these findings. However, in
the process of doing so, a number of collaterally important implications
for cultural and social theory arise. These involve the analytical efficacy of
indigenous culture categories, the nature of cultural systems as inte-
grated, meaningfully structured totalities, and the relation of these struc-
tures to historical process. As I summarize my conclusions regarding
Bush Mekeo culture, I shall directly address these issues as well.
Many of the significant semantic relations upon which this analysis is
based have consisted of both kinds of "motivation" described by Gell
(1975) and discussed in Chapter 1; that is, certain critical indigenous
categories are either polysemous, or they are constituted of words com-
pounded of other words. Examples of polysemy would include the sys-
tematic employment of spatial metaphors in body, waste/resource, vil-
lage/bush, and clan-relationship classifications. In an identical way, the
concept engama variously means procreative conception and mortuary
de-conception and re-conception, as well as beginning in general. So also
does ina refer to the ideas of mother, bodily abdomen, and village ab-
domen; ito to vagina and fire; ikupu to clan and closed; etc. The analysis of
compound words has been especially revealing in the analysis of highly
specialized contexts of the culture. To give a few examples, oa, "law,"
plus if a, "blood," make "law of blood," or official mourner; pa or fa,
"skin," plus nge, "lost," make "lost skin," or feast food; kofu, "club-
house," plus apie, "half" or "other half," make "other clan"; ina, "moth-
234
 Conclusions 235
er," plus ngome, "owner," make "owners of the mother," or father's
relatives; eke, "junior," plus/afca, "senior," make sibling of symmetrical
status; if a, "blood," plus kekapaisa, "making" or "manipulating," makes
"manipulating blood," or quarreling. In addition to these, numerous non-
motivated metaphors have been described: semen/male/closing/meat/dry/
unsweet/bush, womb-blood/female/opening/plant food/wet/sweet/village,
and so on. Altogether, these relations among indigenous categories ex-
emplify the semantic interconnections of the cultural system.
The structural relations of the culture, alternatively, consist of quad-
ripartite conceptual contexts systematically arranged in dual and crosscut-
ting oppositions. In each context, one binary pair is bisected by its own
inversion. I have expressed these formal homologies of the culture in
terms of a structure of bisected dualties:
X' : Y" :: Y' : X"
These semantic and structural sorts of interrelationships together give
rise to a pair of considerations at once empirical and theoretical; namely,
the analytical validity of the indigenous categories, on the one hand, and
the nature of the cultural system as a whole, on the other. Although these
two are very much interdependent issues, I shall attempt to consider the
critical aspects of each separately.
By virtue of the many instances of semantic interrelation I have iso-
lated, the enormous number of disconnected meanings in the culture are
substantially reduced. The model of the total culture remains at this
stage, of course, an extremely complex functionalist system, with an im-
pressive number of categories and meanings. However, the number of
these is smaller than if there were no polysemy, word compounding, or
metaphor, or, rather, if these various kinds of sign relationships were not
regarded as significant. Where all this leads is to the observation that for
Bush Mekeo culture, as for others, there appear to be a relatively few key
or central symbols that crosscut diverse contexts or domains and establish
the dominant themes of the culture. In a famous quotation, Evans-Pritch-
ard has said, "As every fieldworker knows, the most difficult task in social
anthropology field work is to determine the meanings of a few key words,
upon an understanding of which the success of the whole investigation
depends" (Evans-Pritchard 1962:80). It is through the ethnographic de-
cipherment of the semantic interconnections of the cultures key catego-
ries that the culture as a whole and its various elements then become
intelligible.
The possibility of viewing a culture in this general way has been vari-
ously proposed by others in terms of "key" or "core symbols" (Ortner
1973), "epitomizing symbols" (Schneider 1980:135-6), "dominant sym-
bols" (Turner 1967), and "root metaphors" (Pepper 1942; see also Leach
236 Quadripartite structures

1958b; Hallpike 1969; Douglas 1966; 1970; Firth 1973:262-98; Sahlins

1976:210-21; Bourdieu 1977:114-24; Piaget 1970:44-5). All these view-
points suggest that certain domains or sectors from all of a culture are in
one or another way dominant or privileged; that is, they go a considerable
distance toward semantically bringing the diverse and otherwise dispar-
ate cultural contexts together.
For the Bush Mekeo, these dominant symbols might initially seem to
be those that deal with the body: inside, outside, and their inversions;
blood, nonblood, and analogous substances; sweet, unsweet, and dirty;
male and female; open and closed; etc. These metaphors of the body
could also be seen as extended into contexts of classifying social relations,
(e.g., kinship, gender, ritual) and space and time. Moreover, these meta-
phors do not involve merely piecemeal transfer, but instead, as I have
emphasized throughout, they are systematic and logical.
Although the body metaphor notion frankly does reflect my own earlier
intuitions about the culture, methodological and theoretical considera-
tions have led me to be more cautious of this and other alternatives. At
the juncture of isolating a set of dominant symbols for a particular culture,
Ortner (1973) proposes two possible procedures; I shall suggest a third.
First, as implied above, one context - the body, in this case - is identified
by the analyst as containing the key or dominant symbols, and the other
contexts of the culture as possessing extensions of them. The standard
indicators by which key symbols can be isolated are, following this
procedure:
1. The natives tell us that X is culturally important.
2. The natives seem positively or negatively aroused about X, rather than
indifferent.
3. X comes up in many different contexts. These contexts may be behavioral or
systemic: X comes up in many different kinds of action situation or conversa-
tion; or X comes up in many different symbolic domains (myth, ritual, art,
formal rhetoric, etc.).
4. There is greater cultural elaboration surrounding X, e.g., elaboration of
vocabulary, or elaboration of details of X's nature, compared with similar
phenomena in the culture.
5. There are greater cultural restrictions surrounding X, either in sheer
number of rules, or severity of sanctions regarding its misuse. (Ortner
1973:1339)
All these indicators apply to some degree to the Bush Mekeo case. How-
ever, the third one listed is perhaps closest and most consistent with my
own perspective of the nature of cultural systems and indigenous categor-
ies.
The second procedure, best exemplified by Schneiders (1980) treat-
ment of American kinship, begins by analyzing the whole system of cate-
 Conclusions 237
gories and searching it for the context or domain wherein the various
symbolic relations and distinctions are most clearly represented. The
analyst then selects one as epitomizing all the rest.
In somewhat different ways for each of these two procedures, one
context is selected as dominant or epitomizing, and the others identified
as metaphoric extensions of it. My hesitation and suspicion here, despite
my intuitions, is that both involve some degree or element of analytic
caprice. In the first case, nonmetaphoric and metaphoric usages are ini-
tially isolated on the basis of their frequency of occurrence in the culture,
and then these quantitative observations are given the qualitative distinc-
tion as "key," "core," "dominant," etc.; there is, in other words, a meth-
odological and epistemological leap of sorts from quantitative measures to
qualitative significance. According to the second procedure, the selection
of the epitomizing symbol(s) is neither an issue explicitly pertinent to
constructing a model of the culture solely from the data as presented nor
necessarily a salient attribute of the culture itself. Rather, epitomizing
symbols are convenient but nonetheless potentially arbitrary literary or
expository devices, which may well empirically lack the broad signifi-
cances attributed to them (see Ortner 1973).
A third procedural possibility, which I am proposing here, is at once
more modest and more ambitious. The signs or symbols of Bush Mekeo
culture are semantically ordered in part through their replication across
cultural contexts, but there is no single symbol or context that can neces-
sarily be identified as dominant over the others. For Bush Mekeo culture,
one could propose, as I did quite tentatively above in a sort of modified
Marxian perspective on the labor theory of value, that it is the experience
of ones own body that is the first and foremost reality. All others would
be secondary refractions of it according to how the body is exerted upon
them (or vice versa) (cf. Bourdieu 1977:114-24). Quite the opposite is
involved, of course, from a conventional Durkheimian viewpoint, where-
by social forms are the primary roots of experience, and others, such as
those of the body, are derivative of them (cf. Douglas 1970). Based upon
other Bush Mekeo evidence, there is perhaps yet another alternative -
that the experience of living in bounded, consolidated villages sur-
rounded by the bush shapes conceptualizations of all space and time, in
the first instance, and subsequent to that, conceptions of the body and
society (M. Strathern, personal communication).
Long-standing, conflicting theoretical propositions could no doubt be
advanced in support of all three of these views. My own considered
thinking, however, is that no one of these either empirically or the-
oretically eliminates the value absolutely of the others. Notwithstanding,
a telling feature they all share is that they implicitly assume relations that
give deterministic qualities to the culture-as-lived over the culture-as-
238 Quadripartite structures

constituted (Sahlins 1976). The one overridingly inescapable ethnogra-

phic conclusion in the Bush Mekeo case, however, is that there are
consistent, pervasive, structural homologies across the contexts of the
culture, and, furthermore, that these are prior to any specific manifesta-
tion or content. "Matter and form, neither with any independent exis-
tence, are realized as structures, that is as entities which are both em-
pirical and intelligible" (Levi-Strauss 1966a: 130). Therefore, of the two
kinds of interrelations I have focused upon in building a model of Bush
Mekeo culture, it is structural as distinct from semantic ordering that is
preeminent. Otherwise stated, although the culture consists of semantic
interrelations among its various categories, these latter are ultimately
shaped by their structural counterpositioning dictated by the dimensions
of the total system as an integrated whole. The structural whole is greater
than the semantic parts.
This seems the appropriate point to return to Brunton's (1980) dis-
claimers, introduced in Chapter 1, to the effect that order perceived in
Melanesian religions and cultures by Western anthropologists is often-
times an extraneous misconception. First, it is essential to differentiate
methodologically the phenomenological "reality" of cultural and social life
(the culture-as-lived, in Sahlins's terms) from models the anthropologist
makes up after it in order to comprehend it (the culture-as-constituted)
(Sahlins 1976; see also Levi-Strauss 1963a:277-345). Regarding models,
there are, of course, better and worse ones. The standard test of the
adequacy of any particular model in social anthropology depends upon
how well it satisfies four requirements:

First, the structure exhibits the characteristics of a system. It is made up of several

elements, none of which can undergo a change without effecting changes in all the
other elements.
Second, for any given model there should be a possibility of ordering a series of
transformations resulting in a group of models of the same type.
Third, the above properties make it possible to predict how the model will react
if one or more of its elements are submitted to certain modifications.
Finally, the model should be constituted so as to make immediately intelligible
all the observed facts. (Levi-Strauss 1963a:279-80; see also Piaget 1970:3-17

These four requirements are variously implicit in Brunton's reserva-

tions pertaining to the plausibility of Gell's (1975) analysis of ida ritual
among the Umeda; namely, that there is an absence of both native ex-
egesis and analyses of other rituals from the same cultural tradition for the
sake of comparison and support. I am confident, however, that my treat-
ment of Bush Mekeo culture adequately answers Brunton's charges by
showing that the order conveyed is not exclusively the product of my own
manipulations upon data from selected spheres of the cultural whole,
given my own presumed inclination to "oversystemize. ' On the one
 Conclusions 239
hand, I have at numerous points established and reinforced my in-
terpretations by juxtapositioning empirical observations with statements
of native exegesis. Of course, in the field on many more occasions, it was
just such kinds of statements villagers made that initially led me along the
directions I have followed here. Only limitations of length have prohib-
ited including more than a minor proportion of them in the preceding
chapters.x On the other, I have not restricted the scope of my inquiries to
any one ritual nor to rituals of any one sort. Indeed, I have revealed a
structure of bisected dualities in many diverse contexts of the culture,
ritual and otherwise. On both scores, then, the adequacy of my overall
model in accurately reflecting significant systematic features of Bush
Mekeo culture-as-constituted is thereby reaffirmed and strengthened.
Brunton's contention, however, goes considerably farther. He argues,
"Anthropologists have tended to ignore differences in the degree of co-
herence and elaboration in Melanesian religions" (1980:112) partly as a
result of underestimating the degree to which indigenous political com-
petitiveness encourages individualized and inconsistent statements of
symbolic construction. Quite regularly, for example, Bush Mekeo vil-
lagers offered inconsistent and seemingly individualistic responses to my
queries, and their own interactions just as often exhibited considerable
variation. However, as I have tried to show, it is just these sorts of
discrepancies - but seen as conventionalized rather than individualized -
that involve systematic and significant contradictions of the culture as a
whole. The various bundles of conflicting data regarding blood and non-
blood in contexts of consumption and reproduction; of gender roles; of
classifying kin, clan, and affinal relations; and of marriage and mortuary
feasting exchange are essential to portraying the very nature of the order
that characterizes the culture. It is upon these hardly "individualistic"
divergences that the whole system appears to turn and constitute itself as
an ordered system, albeit fraught with contradiction. Although it is un-
deniable that villagers do fall upon opposing sides of conceptual issues in
the unfolding of the political process, this observation does not alter the
validity of the conceptual relations I have established. To suggest other-
wise, as Brunton would seem to, is, I think, to confuse what Levi-Strauss
has distinguished as a mechanical model (and a very simple one at that) for
a statistical model; or, at the very least, to apply inappropriately the
standards of evaluation of the latter to the former (1963a:283-9). In other
words, whatever actions they might take to "maximize" their individual
or political advantage, Bush Mekeo villagers (and I suspect other Melane-
sians too) have very little recourse but to the system of signs and mean-
ings that comprise the public cultural domain. The seemingly "indi-
vidualistic" interpretations proffered before anthropologists may well be
anything but that - they may have meaning only in relation to other
elements of the wider cultural whole.
240 Quadripartite structures
An even more serious problem underlies Brunton's attack, that is, the
attempt to link varying and discrepant indigenous statements of a re-
ligious nature through reified categorical distinctions of extraneous origin
(e.g., degree of orderliness, supernatural vs. political, sexual "conflict"
vs. interclan "warfare'). In such areas, of course, native exegesis is by
definition impossible because the latter, foreign notions are brought to
the data by the observer. Nevertheless, this is the inherent risk of an-
thropological approaches that employ conceptual elements other than
indigenous culture categories in their analyses. In the process of utilizing
such analytical tools, the very sorts of relations that give the system
meaning are either extracted out of context and distorted or ignored.
And, of course, any eventual characterization of properties pertaining to
the whole system is likely to be deceptive.
A recent ethnographic study possessing the greatest ethnological bear-
ing upon the understanding of Bush Mekeo culture illustrates most clear-
ly the dangers of this kind of approach compared with the sort I have
attempted here. I am referring to Hau'ofa's (1981) study of the neighbor-
ing Central Mekeo, which I have had numerous opportunities to cite. In
the course of his analysis of Mekeo society, Hau'ofa investigates and
elaborates upon several semantic domains of the culture, many of which I
have also discussed at considerable length (e.g., inside vs. outside, senior
vs. junior agnate, senior vs. junior clan branches, single vs. married, in-
law relationships, ufuapie and aua relationships [analogous to Bush
Mekeo kofuapie and pisaua], chiefs and sorcerers vs. commoners) as well
as others (e.g., visible vs. invisible, good vs. evil). However, in trying to
make these distinctions intelligible, Hau'ofa chooses to relate them not
solely or systematically to one another, as dictated by the categories
themselves, but to extraneous Western notions. Following Bateson
(1958), he employs the contrast of complementary and symmetrical du-
alism. Because he sees both evident to some degree in Mekeo society,
Hau'ofa claims Mekeo villagers are fundamentally ambivalent about social
life. Nevertheless, in contrast to Iatmul society, it is the complementary
form of dualism that is declared predominant. Thus, Hau'ofa dissolves the
diverse semantic contexts of Mekeo culture and society under a compre-
hensive principle of "inequality" (1981:289-302).
Indeed, to an English-speaking outsider, relations of senior and junior
siblings and clans, of wife-giver and wife-receiver, and of chief, sorcerer,
and commoner, etc., all seem to involve variations of political "in-
equality. " This supposed uniformity, however, does not have any clear
counterpart in indigenous understandings. As far as I am aware, in both
Mekeo and Bush Mekeo cultures there is no single category opposition
such as "equal" versus "unequal" applicable to all of these contexts, much
less to contexts of inside versus outside, visible versus invisible, and so
 Conclusions 241
on. Indeed, in his enthusiasm to articulate the elements of Mekeo society
to the single idea of "inequality," Hau'ofa has inadvertently distorted the
very character of the categories and relations he is investigating. Two
ethnographic cases from Bush Mekeo are most obviously noteworthy, and
from what I understand they would apply equally well to Mekeo. First, in
his analysis of ipangava affinal relations, Hau'ofa (1981:130-53) recog-
nizes a uniform inequality between wife-givers and wife-receivers, such
that the former are categorically dominant over the latter. Among Bush
Mekeo, however, there is a status difference between some ipa ngaua
affines, but not all. One's spouse's "real" opposite sex siblings, as distinct
from more distant "brothers-" and "sisters-in-law," are not expected to
exhibit any of the asymmetric behaviors; theirs are ideally purely sym-
metrical ties. And for more distantly related wive-giving affines (spouse's
parents, cousins, clan members, etc.), these people are understood as
being themselves one's own wife-receiving affines. As the total context
indicates and as villagers themselves testify, the result is that these affinal
asymmetries altogether balance each other out for the system overall, and
personally for individuals on different occasions.
When Hau'ofa extends the notion of "inequality' into the context of
feasting reciprocities in order to characterize relationships between
ufuapie groups and ana "soulmates," the identical confusion arises
(1981:154-83). As with the Bush Mekeo, feast reciprocities between any
two partners are strictly identical in terms of the types and quantities of
foods exchanged. Nevertheless, in his zeal to find "inequality" here,
Hau'ofa emphasizes that, in the early and late phases of any one feast, the
givers are first "supplicants" and then "superiors" in relation to their
guests, who are correspondingly on this occasion "superiors" and then
"supplicants. " Later, when the guests have their own feast, all these roles
will be reversed: The earlier feast-receivers will now become giving "sup-
plicants"/"superiors," and their earlier hosts will play the roles of receiv-
ing "superiors"/"supplicants. " Hau'ofa has chosen to gloss these rela-
tionships as "symmetrical inequality" (1981:177-80). For me, it is
tempting to posit here another dimension of quadripartite structuring in
Mekeo culture. However, the data do not allow either, as there is no
indication of any indigenous recognition that I am aware of that the roles
of feast-giver and feast-receiver are anything but merely symmetrical and
directly reciprocal. Even in Western terms, moreover, reciprocal or sym-
metrical relationships of "inequality" would seem to balance out to
"equality," if anything, and not "inequality." Hau'ofa has evidently been
misled by nothing other than the fact that Mekeo feast reciprocation
between ufuapie and aua is not immediate but delayed. Yet, this cannot
be enough to characterize the institution as one based on some principle
of "inequality," or even "equality," for that matter (see Young 1982).
242 Quadripartite structures
More generally, the difficulty is not that the ethnographer has sought
an order to Mekeo society, but that he has misconstrued that order in
terms of Western meanings, ironically in many instances with the same
sorts of Western meanings suggested by Brunton. In trying to fit Mekeo
constructions into the Western notion of "inequality," Hau'ofa ends up
ethnocentrically distorting the semantic content of Mekeo culture catego-
ries, and in the process he is led away from developing the structure
intrinsic to the culture-as-constituted. Thus, after the dualistic tendency
of the culture's categories is noted, the theoretical significance of this is
laid aside and the whole hung instead on the singular notion of "in-
equality." Indeed, had the potential coexistence of both Batesonian no-
tions of complementary and symmetrical dualism been more vigorously
pursued, a more dynamic kind of structural ordering for the society and
the culture might have been realized - one encompassing the systematic
bisections of the dualisms that were observed. As a result, also, Hau'ofa
might well have discovered a more sound ethnographic and theoretical
basis for framing his psychological argument of Mekeo ambivalence about
social life generally.
Therefore, although going some considerable distance in isolating
many of the important categories of the culture, Hau'ofa's effort creates a
skewed conception of the whole of Mekeo culture and society, one that
casts back upon the constitutive meanings of the indigenous categories
and distorts them. By contrast, for Bush Mekeo culture I have aimed at
making the indigenous concepts and categories intelligible through their
given ethnographic relations to one another and as they are evident in the
structural ordering of the culture as a whole, not as they relate to Western
distinctions.
Quite different theoretical considerations lead to precisely the same
conclusion. The last decade or so has witnessed the resurgence of histor-
icist perspectives in anthropology that tend to disavow the view of a
culture as a meaningful and structured systems of signs. These raise once
more, but from another quarter, the question of the analytical efficacy of
native constructions and of cultures constituted as wholes. Judging on the
basis of the Bush Mekeo case alone, these oversights are critical. Histor-
icist perspectives typically substitute foreign for indigenous categories
and thereby risk abandoning the hope of capturing the essential features,
semantic and structural, of the phenomenon that is to be described,
interpreted, and explained, that is, of culture. Moreover, they fail to
appreciate the manner in which the noting of a structured cultural whole
is capable of encompassing and comprehending historical events. The full
theoretical dimensions of these issues go well beyond the bounds of this
relatively modest ethnographic effort. But given their contemporary sig-
nificance in anthropology, it is certainly appropriate that they be explored
 Conclusions 243
in light of my findings for Bush Mekeo culture. For indeed, the full
theoretical value of the structure of bisected dualities has yet to be stated.
The divergence represented by history and structure immediately re-
calls the Saussurian distinction of langue versus parole, variously re-
phrased in modern anthropological terms as structure versus event, syn-
chrony versus diachrony, or culture-as-constituted versus culture-as-lived
(Saussure 1959; Levi-Strauss 1966a:66-74, 231-44; Sahlins 1976, 1981;
Schneider 1980). On the side of langue, the concern is with abstracted
models of systematic relations among signs, among meanings, among
ideas. Alternatively, the emphasis for parole is upon action, event, and
behavior, upon historical or chronological determinations rather than
upon system, ultimately upon concepts external to the object of study
(i.e., those of the outside investigator's own cultural or scientific tradition)
(Harris 1980; cf. Geertz 1973:3-30).
The appropriate response to the historicist challenge, which is sus-
tained by the empirical treatment of Bush Mekeo culture in the preced-
ing chapters, is that the anthropological endeavor is of necessity first to
make intelligible the systems of meaning and ideas that constitute the
cultures of diverse societies. Only after the system of meaning of any
particular culture has been thus synchronically elicited is it possible,
secondarily, to account for the acts and events diachronically generated
by that culture without obliterating in the process those very conceptual
relations and structures that lend them their own systematic character. It
was perhaps the case at one time that functionalist anthropologists ne-
glected historical questions merely because they did not know how to
incorporate them into their analyses (Leach 1954:282-7). However, this
is no longer so (if it ever really was). Whereas the critical problem for
many remains selecting one approach to the virtual neglect of the other
(e.g., Harris 1980), for others including myself it involves instead recogni-
tion of the validity of both synchronic and diachronic views in their proper
perspective, that is, according to the Saussurian precept of the priority of
the former over the latter. Even Godelier, who has been perhaps the
most insistent upon welding together structural and historical perspec-
tives, argues that, in a correct Marxian framework, the study of the ori-
gins and development of a structure must methodologically come subse-
quent to a consideration of that structure's internal functioning (Godelier
1970:340-58, 1971). To undertake what is logically the second stage of
investigation (i.e., history or diachrony) without first completing the syn-
chronic analysis of the culture-as-constituted in its own terms would leave
little option but to invoke the analyst's own externally imposed catego-
ries. The object of study - the system of signs and meanings - would
thereby be almost certainly distorted. The issue, reduced to its essence,
is this: "An exact knowledge of the system under investigation and of its
244 Quadripartite structures
structure is a necessary condition of any studies of the dynamic aspect of
facts, i.e., of historical or genetical studies" (Schaff 1978:14). To under-
stand the changes of a system, one must first specify what the system is
that undergoes those changes.
At the level of social action, there is no inherent logic or meaning
except in the context of a coherent and consensual system of signs (Sahlins
1976; Geertz 1973:17). Were it to be assumed otherwise, the conception
of cultures as integrated totalities, each in their own right, would evapo-
rate, leaving piecemeal residues unconnected except for their temporal
coincidence. Clearly, considerations of social action and history need not
be inconsistent with, nor mere alternatives to, structuralist constructions.
In their accurate, concrete, and meaningful forms, they logically follow
from prior synchronic considerations. Bush Mekeo lineage and clan histo-
ries, as described at length in Chapter 6, for example, become intelligible
only within the total conceptual framework of the intersecting dual con-
trasts of peace versus war and chief versus sorcerer, which is, further-
more, logically consistent with the structure of bisected dualities that
ramifies throughout all of the culture. Similarly, as discussed in Chapters
1 and 6, the Bush Mekeo conception of warfare (aoao), although it has
been for long now obsolete as an event, is an inescapable fact or reality in
contemporary village life. The meanings of other ideas in the culture such
as peace sorcery, courtship, marriage relations, etc., are partially shaped
by it, and vice versa. And so too are many actual occurrences and devel-
opments throughout Bush Mekeo postcontact history up to the present.
This does not require that from an analytic viewpoint practical consid-
erations are not involved, or that they are to be excluded at the expense of
the "merely symbolic." Rather, the notion of practicality itself can have
meaning only against the fine texture of a specified system of signs, that
is, only when it is contextualized by a wider set of culturally conditioned
understandings. This is, I think, the essence of Sahlins's attack upon
theories of practice in his book, Culture and Practical Reason (1976),
which I have found many opportunities to cite; namely, "there is no
material logic apart from the practical interest, and the practical interest
of men in production is symbolically constituted" (p. 207). Jamous's cri-
tique of Bourdieu takes this general position as well (1981:140-2).
An anthropology of history is not thereby deemed impossible. Quite
the reverse. In a more recent work interpreting and explaining the se-
quence of events surrounding Captain Cook's experiences in early Hawai-
ian history, Sahlins traces a convincing dialectic between, on the one
hand, the preconceived and developing cultures-as-constituted of both
the Hawaiians and British and, on the other, the chronological unfolding
of their mutual interactions (Sahlins 1981; see also Sahlins 1983). Repre-
sented as both exemplars and contradictions to prevailing symbolic val-
 Conclusions 245

ues, events did contribute to revaluing or restructring the culture. Nev-

ertheless, at no point did any of the players act free of the values embo-
died in their contemporary world views. The culture-as-constituted at
every step lent significance to the culture-as-experienced. Although the
interactions may have contributed to the reshaping of categories and
relations of the culture, their unfolding and interpretation were inevitably
screened through preexisting conventional understandings. Structural re-
lations remain, then, prior to diachronic ones. Again, as Sahlins remarks,
the world is experienced as already segmented by relative principles of signifi-
cance; and even if experience proves contradictory to people's categorical presup-
positions, still the process of redefinition is motivated by the logic of their cultural
categories. The innovative value is still a relationship between signs and cannot be
determined directly from the "objective" properties of the referents. (1981:70)

My analysis of the structure and meanings of traditional Bush Mekeo

culture contained in this volume should be seen as thus representing a
first and essential step toward examining the course of development of
that system in postcontact history. Indeed, the most suggestive the-
oretical possibility deriving from this work goes beyond the relatively
simple observation that the categories of the culture are structurally or-
dered to the recognition that the precise logical form of that structure has
a specific appropriateness for the kind of historical analysis Sahlins envi-
sions. For him, synchrony through the agency of its inherent contradic-
tions opens itself to diachrony. In my model of Bush Mekeo culture,
contradiction is explictly embedded in the pairing of conceptual contrasts
systematically bisected by their own inversions.
In the wider view, there appear to be altogether two sorts of "contra-
dictions" possible, however. Their relation, moreover, seems to be in-
strumental in the articulation of synchrony and diachrony. On the one
hand, conceptual elements contradict each other; on the other, they as
idealized forms are contradicted by observed deviations from them in
action. Leach's classic exploration of the logical and empirical implications
of mayu-dama relationship in Kachin society isolated both of these kinds
of integrative discrepancies (Leach 1954). And it is around these two also
that many contemporary studies in Melanesia continue to revolve.
A particularly illustrative case in point is Kelly's study in structural
contradiction in the social organization of the Etoro of the southern high-
lands of Papua New Guinea (Kelly 1977). Rather than deny the com-
patibility of structure and event, Kelly demonstrates the theoretical vir-
tue of their interconnectedness. Speaking of social structure specifically,
he "envisions an entirely different relationship between elements than
that entertained by either descent or alliance theorists" (1977:5). Social-
structural principles involving conflicting rules of action, he finds, are
246 Quadripartite structures
segregated from one another. Obligations that appear, then, at a super-
ficial level as not inconsistent are at a deeper level contradictory. In the
Etoro case, rules of patrilineal descent conflict with those of siblingship,
and vice versa.2 There is as a consequence additional discrepancy and
inevitable contradiction between these rules and their observance. None-
theless, in the unfolding of social action, deviation in one context or
domain may represent compliance relative to another. And, paradox-
ically, statistical deviations from idealized rules are viewed as public en-
dorsements of shared structural understandings.
This view of structure is rooted in the very notion of contradiction
itself. It is thus differentiated from more conventional perspectives in-
formed by assumptions of simple consistency, structural or otherwise. In
this respect, Kelly's analysis is clearly assimilable to the kind of in-
terpretation of Bush Mekeo culture I have offered in terms of homolo-
gously replicated bisected dualities. His fundamental proposition, that
behavior incompatible with rules in one domain can be consistent with
rules in another, has essentially the same implications as when I have
shown how opposed indigenous categories are bisected by their own
inversions. The so-called real world of action and experience will only
rarely if ever perfectly conform to the cultural or idealized models of that
reality. Any attempts to classify the real world will thereby generate
ideational anomalies.
It is at this juncture, however, that my argument deviates considerably
from Kelly's (1977:298). To account conceptually for the empirical experi-
ence of these anomalies, new categories are, or will be, generated. The
expression and resolution of the contradictions are not necessarily to be
discovered only through action, but also in the conceptual realm of the
culture-as-constituted. This view quite closely resembles that of Leach
(1964) and Douglas (1966) - that it is the anomalies of structure arising at
the boundaries of distinct categories where the structures themselves are
at once threatened and redefined. Nonetheless, my view differs slightly
from theirs too in that with Bush Mekeo culture it is not merely a single
intermediate third category that performs this function, but again an
additional binary inverted version of the original pair resulting in a com-
plete quadripartite structure. Bush Mekeo contradictions in the contexts
of conception and alimentation, gender roles, social structure, and mortu-
ary feasting are resolved through their own inversions.3
Like Kelly's model of structure, then, my own rests on the notion of
contradiction or inversion; unlike it, however, contradiction is resolved in
further conceptualization, and not necessarily through action alone. Still,
it is experience of the world that partially generates the impetus for that
further conceptualization. As with Leach and Douglas, furthermore, my
view of structure is based upon the inevitability of conceptual anomalies
 Conclusions 247
as a result of simply binary discriminations and experiences; but unlike
theirs, it involves quadripartite rather than tripartite classifications.
There is an additional aspect of structures such as I have been discuss-
ing that potentially extends even further the analytical linking of structure
to event. Structures of bisected dualities, it seems, may well possess a
special capability to encompass real-world deviations from cultural pre-
cepts in an entirely comprehensive manner. Conceivably beginning with
any simple binary pair, there will be considerable correspondence with,
as well as deviation from, both terms in the process of employing them to
interpret experience of the world, as argued by Kelly. The correspon-
dences will reaffirm the epistemological "truths' of the initial pair, but
not so easily the deviations. The experiential manifestations of either
category of the pair that do not exactly correspond to the alternate term as
well may generate, in both instances combined, a new additional pair of
terms that inverts the first, for a total of four terms altogether.
As a simple hypothetical (and I hope not too provocative) example of
the logic being suggested here, there are, in our culture, two gender
categories ordinarily recognized: female and male. Further significant or
idealized paired qualities are also conventionally associated with one or
the other gender respectively: short versus tall, gentle versus aggressive,
weak versus strong, and so on. Living experience of women and men,
however, occasionally belies these correspondences. Some men are seen
as shorter, more gentle, or weaker than women generally, and so too are
some women oftentimes noticed to be taller, more aggressive, and strong-
er than many men. A simply binary distinction is therefore inadequate to
accommodate all commonplace experiences of gender variation. Instead,
four categories would be required: conventional female, conventional
male, feminine male, and masculine female.
Examples of this logical mode from Bush Mekeo culture that are per-
haps most easy also to visualize would include spatial distinctions of the
world and of the body. In the former context, there exist the bush and the
village, but there is also a portion of the bush adjacent to the village
where village wastes accumulate, and a region of the village where bush
resources, once processed into wastes, are collected. In the context of
bodily discriminations, there are, additional to an inside and an outside,
regions where outside space is inverted into the body (i.e., the abdomen)
and where substances of the body's inside are everted to the outside (i.e.,
excreta).
The quadripartite distinctions in each of these examples are utterly
comprehensive. Seeming exceptions to, or deviations from, our standard
cultural classification of gender are encompassed elsewhere by inverted
culture categories. There is thus no experience of gender in American
culture that does not easily fall within one or another of the four pos-
248 Quadripartite structures
sibilities. So too are instances of spatial and bodily anomaly in Bush
Mekeo experience systematically accommodated by quadripartite distinc-
tions. In this manner, I propose, structures of bisected dualities may
possess a particular efficacy.
The theoretical implications of this potential for the relation of struc-
ture to event or synchrony to diachrony are distinctive. Structure and
synchrony are not necessarily irreconcilable to historical events and di-
achrony (cf. Saussure 1959; Godelier 1977; Auge 1982). Not only are the
two perhaps enjoined, as I have argued, but structure remains prior to
event. Deviations and exceptions, or even change for that matter, are not
fortuitous but follow from the structure of symbolic meanings. It seems
intrinsic to the phenomenon of human experience that the world is not
experienced except through cultural or conceptual precepts and symbolic
classifications.
These properties together provide some grounding for the empirical
possibility that the structure of bisected dualities may not be unique to
Oceanic cultures. In Chapter 1-, I discussed analogous formulas prevalent
in traditions of Western scholarship. In anthropology, these included
established structuralist theories of myth from Levi-Strauss, Leach, God-
elier, and Willis, Van Gennep's classic formulation of ritual along with its
many modern elaborations, Andriolo's characterization of the relation of
myth to history, Hohfelds analysis of legal relations, Douglas's scheme
for classification for grid and group, and Auges Marxian reconstituting of
the anthropological field itself. Beyond anthropology, the most notewor-
thy instances of structures homologous with these are from Jakobson's
patterns of phonological features, Piagetian developmental psychology,
and the applications of the "Klein group' in mathematics. These last two,
more than most, perhaps by linking together the logical operations of
identity, negation, reciprocity, and correlation, have been viewed as the
most widely generalizable. There are certainly still others (Piaget 1970).
One possible implication of this convergence, of course, is that my having
found essentially the same structure represented systematically in the
non-Western traditions of the Bush Mekeo, Tikopians, and the Trobri-
ands is due, not to the intrinsic nature of the latter, but rather to my own
conditioning and adherence to a pattern of thought (conscious or uncon-
scious) widespread in Western culture and scholarship. If such is indeed
the case, then it is essentially to the same point, namely, that there is at
least one culture so ordered - our own, or at least a significant part of it!
More likely, though, the detailed examinations of Bush Mekeo, Tikopian,
and Trobriand cultures contained in the previous chapters should lay
aside most serious doubts that bisected dualities are not somehow impor-
tant in ordering relations among their respective constituent categories.
I have tried to suggest how structures of bisected dualities might be
 Conclusions 249
capable of encompassing behavioral deviations and exceptions in experi-
ence and in history. Deviations of action, from this perspective, are pro-
grammatic, just as conceptual ambiguities and contradictions are respec-
tively encompassed and resolved. Does it finally become a credible
possibility to describe, interpret, and explain cultures as wholes, or total
social phenomena, rather than leaving it as an assumption that such is the
case based upon the close examination of only parts of a culture? It seems
not unexpectable that structures of bisected dualities with their special
potential for both conceptual and behavioral comprehensiveness and
wholeness might well typify more cultural systems than has heretofore
been recognized. It is, of course, virtually certain that many cultures
manifest fully effective structures of other sorts - simple binary systems,
triadic ones, etc. - and conceivably even more diversity in the manner by
which those structures articulate with historical events. Undoubtedly, the
question will remain as much empirical as theoretical. Perhaps, the most
prudent course would be to investigate first the extent to which bisected
dualities structure the cultures of other Austronesian-speaking Pacific
societies. If this structure is as pervasive in Bush Mekeo, Tikopian, and
Trobriand cultures as I have suggested, one would expect it to be identi-
fiable to a greater or lesser extent in cognate systems. What would be
needed thereafter would be similar models of cultures from many culture
areas, aimed too at uncovering the structures that give them each their
distinctive sense of wholeness or totality while preserving as best as possi-
ble the systematic meanings of their respective indigenous categoriza-
tions. It should be noted that the contexts of the cultures I have selected
for exemplification in the preceding chapters are in most cases not trivial
ones. They concern not only politics, economics, kinship, and religion,
but whole ways of integrating total cultures in their own terms.
It is in this general direction that I have here taken a few very tentative
steps. Therefore, besides finally bringing the Bush Mekeo out of hiding
and into the anthropological world, I hope to have revealed a useful
structural plan for describing, interpreting, and comparing cultures
generally.
 Appendix 1
Village resources derived from bush
resources

garden foods
meat of game animals
houses, platforms, and clubhouses
weapons
tools of bone, wood, tooth, steel, and stone
clothes (loincloths for men, grass or sago-leaf skirts for women)
arm- and legbands
bark belts
combs
string bags
water for cooking and washing
mats
barkcloth blankets and sleeping nets
nets for fishing and hunting
dancing drums
flutes
canoes, paddles, and rafts
firewood
areca nut, betel pepper, and lime
tobacco
vegetable dyes
ornaments and valuables of bone, shell, feather, and tooth
ingredients (fuka) for ritual charms
medicines (mulamula), etc.

250
 Appendix 2
Ingestion and ingestibles

Ingestive process Category of ingestible

eani eating fuka medicines
tsiale native tobacco
nao twist tobacco
sugar sugar
tsitsimalu salt
fokama plant food
[traditional]
akole sweet potato
angeange taro
kokou taro
foa banana
konga coconut
ongoi breadfruit
ongoi breadfruit seeds
aisa sago
kabatsi greens
kamai almond fruit
anifa pandanus nut
ili sugarcane
paupau papaya
[introduced]
banana ripe banana
orantsi orange
lemoni lemon
tomato tomato
plaur bread and scones
laitsi rice
nanatsi pineapple
kauni maize
meloni watermelon
pumpkin squash
pinutsi peanut
onioni onion

251
252 Appendix 2

eani eating tsitsi meat

(continued) [traditional]
kuma domesticated and bush pig
matsi, mani wallaby
pio cassowary
auke dog
kungaka cuscus
apai bush fowl
aina egg
angai mullet
kokolo chicken
anika prawn
kola tree iguana
afinama flying fox
oala crocodile
maipa duck
mokoa, bull catfish
kangaua eel
engo cockatoo
ainapa hornbill
moa frog
inema rat
tsinotsino glider
opo bird of paradise
boku barramundi
(and several other unidentified species of grub,
amphibian, and marsupial)
[introduced]
bull corned beef and mutton
kampai camp pie
maka tinned mackerel
einu drinking gugu breast milk
fuka medicines
ivi water
[traditional]
ivi river water
kemoa wild banana stalk fluid
konga coconut milk
imenga broth of boiled food
konga coconut oil
[introduced]
tsi tea
cofi coffee
loli water soft drinks
milak cow's milk
bia beer
 Ingestion and ingestibles 253

uiski whiskey
rum rum
euwa chewing fuka medicines
max areca nut
awaka betel pepper
apu lime
 Appendix 3
Categories of food

Plant foods

Parts Sex Body

Food category Preparation eaten of eater sustenance

sweet potato B/R S/N M/F H

cooking banana B/R N M/F H
ripe banana raw N F C
cabbage B F C
small taro B N M/F H
big taro B N M/F H
coconut raw N F C
coconut oil B N M/F H
pineapple raw N F C
papaya raw N F C
tomato B S M/F H
pitpit B/R N M/F H
maize R N M/F H
squash B N M/F H
watermelon raw N F C
sugarcane raw N F C
litsi raw S F C
almond fruit raw N F C
breadfruit B/R N M/F H
breadfruit nuts B/R N M/F H
sago B/R N M/F H
pandanus B N M/F H
rice B N M/F H
sugar raw M/F C
salt raw M/F C

Note: B = boiled; R = roasted; S = skins; N = not skins; M = male; F = female;

H = hot; C = cold.

254
 Categories of food 255

Bush meat foods

Bush Dirty Unsweet Big/

Animal species location Preparation parts parts small

wallaby O B/R/D HAF U s

bush pig O B/R/D HAF U b
mullet O B/R/D HAF U s
prawns O/I B/R A? S? s
eel O/I R SAFTOU b
cuscus o B/R HAF u b
barramundi o B/R/D HAF u b
crocodile O/I B/R/D AF us b
tree iguana o B/R/D HAF u b
cassowary o B/R/D HAF u b
large flying birds o B/R/D HAF u b
small flying birds 0 R HAF u s
glider o B/R HSAF u s
bird of paradise o B/R HSAF u

C/3
grubs o B s
beetle o R u s
bush fowl o B/R HAF u b
catfish O/I R SAFTOU u s
buli catfish O/I R SAFTOU u s
rat o R HAF u s
cassowary egg O/I B ST s C/3
bush fowl egg o B ST s s
snakes O/I SAFTOU b/s
small lizards O/I SAFTOU s
frogs O/I R SAFTOU s
ground iguana O/I SAFTOU b

Note: O = outside of bush holes; I = inside bush holes; B = boiled; R = roasted;

D = smoke dried; H = hair, feathers, scales (buibui); S = skin; T = meat and
blood; O = organs; U = bones; A = anus and colon; F = feces; b = big meat; s =
small meat.
 Appendix 4
Work and nonwork skills

Work skills
garden clearing, burning, planting, weeding, harvesting
carrying
chopping firewood
construction of houses, platforms, clubhouses
canoe manufacture
fencing
grass cutting
rubbish removal
digging
butchering
wood carving
boiling food
bark pounding
tool manufacture (axes, adzes, knives, weapons, drills, pounders, etc.)
valuable manufacture (shell armbands, feather ornaments, strings of dogs'
teeth, strings of cowrie shells, etc.)
fire making

Nonwork skills
hunting with dogs, spear, shotgun, bow and arrow, nets
fishing with spear, bow and arrow, hook and line, nets
ritual for hunting, fishing, gardening, curing, peace sorcery, war sorcery,
courting, mefu, pig husbandry
drum making
dancing
drumming
singing
roasting food
chewing areca and betel
childbirth
marriage-compensation exchange
mortuary-feast exchange

256
 Appendix 5
Categories of human dirt

Bloody dirt Bloodless dirt

blood feces
flesh urine
semen skin oil
menstrual blood saliva
placenta fufu head hair
umbilical cord buibui body hair
pus nail parings
tears
vomit
areca spittle

257
 Appendix 6
The myth of Foikale and Oa Lope

Version Ft*
A long time ago in the region of Inauakina (a Bush Mekeo Village), there
was a village (sic) where there were only men. No women, no girls,
nothing but men. They gave birth by themselves (men to men), and upon
being born they were already strong and able to search for their nourish-
ment.
These men were savages. They did not wear clothes, lived in caves in
the side of a hill, and did not even make gardens. Their only nourishment
was a kind of earth called onkimo; they dried it in the sun and swallowed it
that way.
The principal personages of this curious village were Foikale, Koikoi-
pike, Kapankoupike, Maimaipike, and Ikangopo. They were very famous.
Foikale was the chief.
In the same region, there was a family of which the husbands name was
Oa Lope, and the wife was Oini. They had several children. Oa Lopes
family completely ignored Foikale's village, and the latter had no acquain-
tance with Oa Lope.
One day, Foikale was going walk-about and came to a fence constructed
of split and interlaced bamboos. Never had he seen such a thing. He
made a tour of the border, found an opening, entered, and was stupefied
at the sight of a garden of magnificent bananas. This was Oa Lope's
garden. Keeping a distance, he cried, "Psh! Psh!" to see if the stalks of
bananas would run away, but they did not. They stayed in place, immo-
bile. Then he advanced and walked in the garden, wondering what this
might well be. Thus, he arrived at the foot of a banana tree, of which a
ripened bunch was detached and on the ground. Foikale stopped. He was
afraid. He thought it was a snake, and he said, "Psh! Psh!" But it did not
move. He came up to it, and with a stick he drew one of the bananas
away. It allowed itself to be caught without difficulty. Foikale felt it,
smelled it, and hazarded a bite with his teeth. "My, but this is good." And
he swallowed it. He took a second one and ate it too.
* Translated from Guis 1936:220-6.

258
 The myth of Foikale and Oa Lope 259
But as this nourishment was new to him, he vomited it up.
Foikale was not to be beaten, and he recommenced to eating, finding
the bananas more and more delicious. When he had eaten his fill, he
shook the tree, and a bunch fell down. He caught it, covered it up, and
returned triumphantly to his village (sic). He rolled away the heavy stone
placed at the entrance of his cave. His companions were together there
and sleeping. He woke them up and said, "You must eat whatever thing is
good, better than onkimo." But at seeing the bananas, they thought they
were seeing snakes, and they ran away out of the cave.
Foikale called them back, shouting, "Nonsense to be afraid, you go
away although it does not move!" And so to encourage them, he ate one in
front of them, and they returned. But Foikale, anticipating that they
would vomit, placed them back to back so that they would not take a
dislike to them mutually. Then he gave everyone two bananas. They ate
them, and they tasted good. But after several minutes, there was a uni-
versal collapse, as Foikale had thought. His greatly angered companions
accused him of having deceived them, but he responded that he also had
done it thus, and that they must come to try a second time, the bananas
will stay in place. They allowed themselves to be persuaded, and indeed
they were not disturbed by them at all. They looked at one another with
an air of complacence, clicking their tongues and rubbing their stomachs
in sign of satisfaction. They asked him questions without end, and
Foikale, with the help of many details, described his last walk-about.
When Foikale had left Oa Lope's garden, Oini, the latter's wife, went
there to work, and she was very frightened on seeing the footprints,
which were not those of her husband and her children. She went to tell
Oa Lope, and the two of them returned and installed themselves in the
garden to take up guard.
Several days later, Foikale, dressed in the manner of his first adven-
ture, went directly along his new path toward the garden. This time he
perceived plumes of smoke, and great was his fear: "This time for sure
this is a snake, it walks in the air, it flies!" And trembling, he squatted
behind a trunk of a fallen tree. From his hiding place he saw two people
seated near the stalk of this serpent. His fear increased, especially when
these two people rose and came toward him. He thought himself lost, and
he uttered a cry. Oa Lope and Oini, in their turn, fell to the ground with
fear. They overcame their fear and, approaching, asked, "Who are you?"
"I am Foikale. And you, who are you?"
"I am Oa Lope, and this is Oini, my wife."
"Your wife? What is that?"
Oa Lope thought that Foikale was being funny, and he did not respond
to this question, which Foikale had posed very seriously.
"Where do you come from?" asked Oa Lope.
260 Appendix 6
"I come from over there, from the foot of the mountain. My village is
very big," responded Foikale.
"Come with me," said Oa Lope. And all three went together.
But when they approached the fire, Foikale was afraid, and he asked,
"What is this? Is it something that will bite me?"
"No," said Oa Lope, "it is fire. It is a very good thing. It is with this that
we cook food. Watch."
And Oa Lope told his wife to put on the embers some green bananas.
When they were cooked, he took one, removed the peel, and gave it to
Foikale, who did not care to eat it. Oa Lope reassured him and ate one in
front of him. Then Foikale tasted his but vomited it a little afterward.
However, instructed by his first experience, he began again and found
that the bananas cooked were better than green. Oa Lope told him that
this depended on the kind and indicated to him that some were eaten raw
and some were cooked. At the moment of departing, Oa Lope said to
him, "You return in three days. Now take away this bunch of bananas and
this ember. On arriving in your cave, lay the ember on the ground with
some dry leaves and small sticks on top. Blow, and the flame will come,
and henceforth you will have fire, which you must not allow to die. But
tell your friends not to sit down on top of it."
Foikale, completely happy, returned to his village where the cooked
bananas and the fire, especially, after having caused a moment of distrust
and panic, obtained the greatest success. Foikale, filled with pride, ex-
plained everything to them.
True to the appointment, three days later Foikale was again at the place
of Oa Lope. He saw Oini squatting close to the fire, looking from time to
time into a great bowl that was in the middle of it. "What is this large
bowl in the fire?"
"It is a kettle. Oini makes it with clay, soil. It allows us to hold water
and boil vegetables."
And Oa Lope had him drink a little of the sauce from the bowl. Foikale
said that it was very good and that this was much better than the soil of his
village. Oa Lope made him a gift of a pot and gave him some explanations
necessary for using it. Foikale took it away and made known his new
marvel at this place.
After three days, on another visit, Oa Lope taught Foikale about the
spear and its use. He took it with him to hunt a wallaby. On seeing this
beast, Foikale trembled with all his members.
"Why are you shaking?" asked Oa Lope.
"I am afraid of that man with hair, over there."
"It is not a man, it is a wallaby (matsi). Aren't there any at your place?"
"I've never seen one," said Foikale.
"Ah, well, you come watch now," said Oa Lope, and, brandishing his
 The myth of Foikale and Oa Lope 261
spear, threw it into the stomach of the wallaby, which fell down dead. Oa
Lope grabbed it, scorched off its hair, cut it open, cleaned it, cooked a
piece, and gave it to Foikale, who declared that it was better than ba-
nanas, even cooked. Oa Lope killed three other wallabies and gave them
to Foikale, who took them to his village, cooked them as he had seen Oa
Lope do, and invited everyone to eat. After several grimaces, all the
stomachs (sic) digested this new nourishment with bliss.
Again, three days later, Foikale returned to visit Oa Lope. They were
becoming great friends. Oa Lope said to him, "I have given you many
things, I have taught you many things. Now I am going to make a gift
again. This is the last, because afterward I will have nothing. You see this
person. Ah, well, this is a girl, this is my daughter; I give her to you. You
marry together that your children will be numerous and that your village
will prosper."
Foikale left with the daughter of Oa Lope. He had several children. Oa
Lope on his side had many other children, and the boys of the village of
Oa Lope married daughters of the village of Foikale, and some of the
village of Foikale married the daughters of the village of Oa Lope.
It is thus that we still do, for it is not good for a young man to take a girl
of the same village (sic) as himself.
It is thus to Oa Lope we render intelligence (makarima) because pre-
viously we were aki (savages), had no houses, no bananas, no vegetables,
no spears, no game, no women. We ate nothing but earth. Not us, but our
old, old, old grandfathers. It had been this way a very long time. Now it is
no longer so, and it is Oa Lope who has changed everything.
Version Fu*
Oa Lope knew how to do everyting. He was very clever to make houses,
gardens, fire, cook food. But Foikale lived in the bush, mucked about
there. Foikale did not know how to cook nor make fires, and so he and his
clan lived together and ate only raw food.
So Foikale one day came to Oa Lope's garden, cut down some ripe
bananas, and took the bunch away. However, he did not steal this food
properly so Oa Lope would not detect the theft; he left leaves hanging
down and on the ground. Oa Lope came to his garden later and saw the
footprints and the places where the grass was bent over from someone
walking. He followed the footprints and the grass, came to the banana
tree where the fruit had been stolen, and called out, "Who took my
bananas? "
Foikale had hid himself at the edge of the garden in small trees. Oa
Lope followed the footprints to him.

* Recorded by author.
262 Appendix 6
After Foikale sat in the small trees, he sat down and thought or talked
to himself. First, he considered, "Now that Oa Lope has come, I will have
to give them (the bananas) to him." But then he thought, "When I take
them to Oa Lope, he will see me coming with his bananas and try and kill
me." Foikale had been sitting facing the direction from which he had
come, so he turned his back to where Oa Lope was coming from. But then
he thought if he turned his back to Oa Lope, then Oa Lope would think
he was afraid. So, at this Foikale turned his side to face Oa Lope.
Then Oa Lope came up and called out, "Wau [namesake], where do
you come from?"
Foikale said, "I grew up here (as a tree)."
Oa Lope asked him who took his bananas, and Foikale said he grew up
there so he did not know. Then Oa Lope took him back to his village.
When they got to Oa Lopes house, Foikale saw it, and they went
inside. There Foikale saw Oa Lopes wife. He was surprised, and asked,
"What is this?" Oa Lope said it was his wife. So then Foikale went up to
her and felt her skin all over with his hands. Then he pulled open her
grass skirt, grabbed her vagina, and asked, "What is this?"
Oa Lope said, "Vagina, you do this with it, let me show you." So then
Oa Lope copulated with his wife while Foikale looked on. When Oa Lope
was finished, Foikale copulated with Oa Lope's wife too. Then when he
was done, and after a little while, Foikale was sitting in the house, and he
began looking at all the parts of it from the inside. He asked his namesake,
"What is this?" and Oa Lope said it was his house. And so that is where
Foikale stayed.
Then next day, Foikale went outside the house and started to disassem-
ble it piece by piece, untying all the string on the roof, all the grass, and
all the sticks of the walls. He laid all the pieces on the ground. And when
that was finished, he began to build it again using the pieces of the old
house. For the roof, he went and cut new grass, and put it on. So now
Foikale knew about houses.
The first time Foikale came to Oa Lope's, he never knew about fire.
One day, Oa Lope made a big fire in the bush out of dried timber, sticks,
and so on, and Foikale saw it and got afraid. But then Oa Lope told him,
"Wau, it is fire, do not be afraid, women use it to cook food." So Oa
Lopes wife cooked some food and they ate it together. But when Foikale
finished eating, he vomited out the food because before that he only ate
uncooked food.
Later in the evening, Oa Lope asked Foikale to have a chew with him.
Foikale had never chewed areca with lime before, so he took a limestick,
stuck it up inside Oa Lopes wife's vagina, got the juice on it, and put it in
his mouth with areca nut. Oa Lope told him, "No, we do not do it that
way." As he himself chewed, Oa Lope's mouth got red, so when Foikale
 The myth of Foikale and Oa Lope 263
saw this, he was surprised, thought it was blood, and asked, "What is
that? Blood?"
Oa Lope said it was not real blood, but mouth-blood (akeifa) or chew
(auwa). "So we do not use vagina. Instead we use lime and betel pepper
with areca."
After Foikale lived with Oa Lope a while, Foikale brought his people
out of the bush to Oa Lopes village. They built houses, but they did not
wear loincloths. So Oa Lope taught Foikale to wear a loincloth. Because
Oa Lope taught Foikale to wear a loincloth, how to make fire, cook food,
build houses, and chew areca, Foikale showed his people how to do these
things. They were bushmen (ango aonga aunga), so they built their own
houses and learned how to live in them, and Foikale showed them fire
and how to cook with it. And when they ate, they too vomited from the
cooked food, and Foikale showed them how to chew areca. So the people
of Foikale came to live at the village.
Version FHi*
Foikale went to Oa Lope's garden, and he stole some bananas. Then he
went into the bush adjacent to the garden. When Oa Lope came to the
garden, he saw that some bananas were stolen. Then he noticed footsteps
from the plundered tree to the forest, and he followed them.
Meanwhile, Foikale was sitting in the bush facing the garden. When he
heard Oa Lope call out, that he was coming to look for him, Foikale
thought that it would be better if he did not see Oa Lope, because then he
would try to dodge the spear and reveal to Oa Lope his own fear. So he
turned sideways. Then Foikale thought about it again, and he figured if
Oa Lope speared him in the side, it would kill him, so he turned with his
back to the garden and said, "Let him go ahead and strike/kill me,"
reasoning that this way he would not see what Oa Lope did to him.
When Oa Lope came up close, Foikale made a noise and surprised Oa
Lope. So then Oa Lope saw him and told him that three of his bananas
were taken from his garden, and the thief came this way. Oa Lope asked
Foikale, "Who are you?" Foikale told him that he had always been at this
spot, that he did not have a village or any other place. Then Oa Lope told
Foikale to carry the bananas to the garden house, so they both went to the
garden house. Oa Lope told him to get a stick with a crack in it and a
string to make a fire. Foikale asked Oa Lope, "What is that?" when Oa
Lope started working the vine on the stick and smoke came up. Oa Lope
said it was the smoke of fire. When the fire started burning with flames,
Foikale got afraid and ran away. Oa Lope called him back, and then
Foikale asked what it was, and Oa Lope answered it was fire. They
roasted the bananas, peeled the charred skins, and ate them.
* Recorded by author.
264 Appendix 6
After eating, both men were thirsty (also ekupu). Oa Lope told Foikale
that there was no water around there, only grasslands. But he told his
namesake to stay there while he went to get a stick from a tree called
aloalo, sharpened the end of it, punched the ground with it, and held the
top. He pushed it back and forth, screaming in a loud voice "Foikale
panganga" ("Foikale asks more questions"), and then he pulled the stick
and water boiled out of the ground and continued to rise. Oa Lope put the
water in his mouth and said, "It is good water, you can have a drink."
And Foikale said, "Water is good."
This water brought out of the ground became a big swamp or lake. The
aloalo tree is still growing on Kuipa land.
That is the story of how people acquired water.
 Appendix 7
The afinama myth

Version At*
One day, all the men of the village went to the bush to hunt pig, casso-
wary, and wallaby with their nets and spears. In the evening, they came
back to the village. The women cooked the meat and took it to the
clubhouse where the men were gathered. The men ate all the meat, not
giving any to the women to eat. When the men ate the meat, they called
their sons to eat. All their wives and daughters ate nothing. The
daughters went to stay with their mothers. Inside their houses, mothers
and daughters waited for their husbands and fathers to call them, but they
never did.
The next day, the men went hunting again, doing it the same way, and
again they ate all the meat with their sons, sharing none with their wives
and daughters. Day after day they did it in the same way.
Then one day, when all the men were away in the bush hunting, the
women who remained in the village held a meeting. The husbands of the
married women were not kind to their wives, and the women tried to do
something. That day, one small boy had remained at home in the village.
The women told him to climb a coconut tree and bring down some flock
or netting in order for them to make wings (panina). When all the women
finished making their wings, one of them tried to fly up to the coconut
tree, and she sat up there. All the women looked up and said it was not
good; she should use her feet to hang upside down. After that, all the
women with their daughters flew up to the big tree (apanitsi) with round
leaves. All the women went to that tree. They left one very old woman
behind.
That evening, all the men returned from the hunt, but they did not see
any women in the village. They put their meat inside the houses but saw
no women anywhere. They found only one old woman in a house. She
stayed behind because her eyes were not good. The women had left all
their sons behind, taking only their daughters. When the men asked the
old woman about what had happened, she told them the story: Every day
* Recorded by author.

265
266 Appendix 7
all the men went hunting in the bush, got meat, and brought it back to the
village. After the women cooked it and took it to the clubhouse, the men
ate it all. The wives and daughters got nothing. That is why the women
were shamed. That is why all of them flew up to the big tree. So all the
men looked up to the tree, went over to it, and cried. They told the
women to come down, saying they would give them plenty of meat. But
the women did not come down. They did not want to. "You go get meat,
we only cook it. You eat it, we get nothing. So now we are separated, we
make our own place in the tree, and we are going to leave you now. When
our time comes, we will come up here, and you men will come here and
kill us and eat our flesh."
When the men heard this, they all grabbed axes and tried to cut down
the tree so the women would fall down. But when the tree fell, the
women flew away, saying, "We are going."
That time, the flying foxes [who had been the women] flew to the
swamps and settled on the smaller trees there. When their time comes [at
the beginning of the dry season], they will come here to the forest and
around the village to stay on the apanitsi branch. At the end of their time,
they will go back to the swamps. So the men had no wives, no daughters.
In our grandfathers' days, the old people believed this story. So no one
ate flying fox, because its flesh was flesh of woman. Today, men do not
believe it, so they will eat it. When present-day men were boys, though,
they did believe it and would not eat flying fox.
Version AH*
There were once women called afinama. All the husbands went hunting
with their bows and arrows and nets, and they killed many pigs, casso-
waries, and wallabies. At a place in the bush they ate all the meat, and
when they had finished there was none left for their women, their wives,
back at the village. So they returned empty-handed, with nothing.
The next day, after sleeping, the women asked the men if they brought
back some meat with them for them to eat, but the men said they did not
see any. So then women asked the small boys, and the latter told their
mothers that they caught pig, cassowary, and wallaby but ate them in the
bush.
The men went hunting again, caught many animals, and again the men
and boys ate all the meat in the bush. Back in the village, the women
asked the men for some meat. The men said they had no meat, but the
small boys said they ate plenty of meat in the bush. And then they went to
sleep.
The next day, the men went hunting and killed many animals. But this

* Recorded by author.
 The afinama myth 267
time while the men were gone, the women were talking - "The men
went hunting, caught meat, but ate it in the bush and did not give any to
us. Yet they did give some to the small boys." One old woman told the
others that the next day they all must take coconut flock under each arm
and tie them tight like wings, so when the men come back they will all fly
up to the top of a tree and stay there.
The men did the same that day and came home without meat, as they
had eaten it all in the bush.
The next day, one small boy stayed back in the village with the women
when the men went hunting. The women told the boy that every time the
men eat meat and bring none to the women, and, because they never
gave meat, the women were going to fly up to that tree. So all the women
tied wings onto their arms, and they flapped their wings and flew up to
the tree. Then they flew back to the ground. The women told the boy
when his fathers come back to tell them their wives are up in that tree,
that they had flown up there. From now on they would be like birds, and
so the men can eat them. They were meat.
When the women had flown up to the tree, one old woman, who had
not gone up, looked up and could see their vaginas. She told them it was
bad, to hang upside down instead. When the women did as she said, she
remarked that it was good. They then all came to the ground again.
The women cooked and ate some food. In the late afternoon, the wom-
en tossed away all their things (blankets, skirts, pots, etc.). Everything
they threw away. They told the boy to tell his fathers for them, "Today
you ate meat, and we are shamed. So we are living on top, like the birds,
so we are like meat. You must hunt us like meat." So the women flew up
to the tree.
The boy was left alone, and he cried until the men returned. They told
him to be quiet and asked him where the women were. "Did they go to
the gardens to get firewood?" they asked.
The boy said, "Always we went hunting, shot some meat, and ate it in
the bush, not giving any to the women, so they had none to eat. The
women got ashamed." The boy then told them to look up into the tree and
they would see their women.
One man named Apanitsi went up to the tree, and he told the women
to come down to get some meat. They would eat it together. On hearing
this, the women flapped their wings and started flying all over the place in
the air around the tree, but landed in the tree again. The old woman told
them, "You did not give us any meat. We are shamed. Now we are up in
this tree, and it is finished. We will not come back to be your wives
anymore."
When the old woman said this, all the men cried. When they stopped
crying, they grabbed their axes and cut the tree down, and it fell over.
268 Appendix 7
But the women flew away. One woman, however, was pregnant. Her
body was heavy and not strong, so she fell when she tried to fly away and
died. Then one of the women told the men, "One woman was pregnant,
she fell and died, so all of you look at her. Take her and bury her. But the
rest of us, when you cut down the tree, we did not die, we will not die
now. The pregnant woman was not strong, so she died."
The men cried, and again they asked their wives to come down and eat
meat with them together. Again the women answered they would not
come down, and again the men cried. The women said, "Now we have
become meat, so you hunt us and eat us. We are shamed now, so it is
finished. When it gets dark, we will go looking for our food, and you men
come look for yours - us!" When it got dark, the women flew away,
leaving the men crying; but eventually they went to sleep.
The next day, the men took the body of the pregnant woman who fell
and died, they cut up her body and gave a piece to a bad boy for him to
eat to see if it was good meat. It did not kill him, though. So they waited a
day, and he was healthy. He told them it was good meat. So the men then
asked the women when next they saw them out looking for their food
what their name was now, and the women answered, "Afinama."
Version AiU*
Originally,flyingfoxes were women. Their village was called Aouaou near
the Angabunga River at the foot of the mountains.
The men of Aouaou were going hunting. They caught some pigs, wal-
labies, and cassowaries. When they had the pigs, wallabies, and casso-
waries, they started to return. They celebrated because they were happy,
and they were happy because they had so much game.
One day, arriving at the village, they sat down and quarreled as they
counted the pigs, wallabies, and cassowaries. They counted them, then
divided them up between the chiefs and the hunters. When the division
was finished, they said to their wives, "Take these pigs, wallabies, and
cassowaries. One part you will cook in pots, one part you will roast, one
part you will smoke-dry, and tonight we men all alone will eat them, we
will feast." The women did the cooking as ordered.
At night, the men met on the clubhouse, and they ate. They ate like
cassowaries [i.e., greedily], and they ate even to the last morsel. When
they had eaten to expiring, they laughed at the women and told them,
"You others drink the soup." And always the men did like this.
Their stomachs lacking, the women beneath the clubhouse defied the
men, saying, "Oh, oh, how the smell is good," and they smacked their

"Translated from Guis 1936:189-93.

 The afinama myth 269
tongues. They thought to themselves, "If they gave us just a little bit, we
would very much appreciate it."
One day, the men went hunting again. The women, meeting in the
village, talked and talked, made some reproaches to their husbands, and
said, "Our men are cassowaries, they eat everything, and we others get
only the soup, and this is bad. They scorn us. What if we scorn them?
Let's do something! What will we do?" They thought about it. Then they
called their children and said, "Children, climb up the coconut trees, cut
off the flock, and throw it down." The women gathered the branches and
sewed them together as they do mosquito nets.
One woman said to another, "Attach this covering at the top to your
shoulders, at the bottom to your ankles, and try toflya little, to see if you
fly a little bit." One old woman who had not spoken yet then said, "This is
true indeed! It is bad, our men do not treat us well. Now make like you
pledge. If you like them again, do not do what you come to say; if you do
not like them again, do it. This is your affair alone. Me, I am too old, I rest
here. When your husbands come, they will see. This is their affair."
The other women responded, saying, "Our men we do not like them
much, we scorn them."
Then the woman who had attached the pad of coconut to her shoulders
and to her ankles like wings went to the middle of the village. She shook
her arms and her legs, and she flew. She flew like a bird and rose very
high, then descended and came to rest on the branch of a large tree.
There, like a bird, she perched with her feet on the branch and her head
above. The other women below followed her with their eyes, their stom-
achs shifting with emotion, and they cried with joy. But when they saw
that she was sitting like a bird, they told her, "It is not good thus."
The woman responded, "You also come up high and make me see how
I must hold myself." And one other woman holding wings flew near her,
but she did not hold herself upright on the branch. She hung with her
feet, and held her head down. Then she replaced her arms, and her wings
retreated entirely.
The flying foxes are still like this now.
The other women saw this and said, "This is thus very good." They all
attached wings to themselves and flew to a very high apanghi tree. They
suspended themselves from the branches next to one another in bunches.
This tree was at the end of the village.
That night the men returned. When they saw the first coconuts, they
sang a mourning dirge (agopa). The children heard them, saw them, ran,
met them, and said to them, "You carry pigs and wallabies, it is good. But
who will cook them?"
"Who will cook them . . . and the women?" asked the men.
270 Appendix 7
The children answered, "Look and you will see."
They arrived at the village. The village was cold, no one speaking, no
one shouting, no singing. Then they cried out. Each men said, "Hou!
Hou!" and named his wife.
No women! No wives! They saw the old woman sleeping near a pile of
embers half-expired. They woke her and asked, "Eh! Old woman, where
are the women?" And the old woman leveled her hand toward the summit
of the apanghi tree and said, "Look, they are on that tree. They are angry
with you. You did not give them meat to eat. Then they dressed them-
selves with wings and flew high away, like birds."
The men, their stomachs shaken to the ground, cried. Then they called
their wives, "Beloved women, do not be angry! Come down, eat every-
thing, we others nothing. You see, we are carrying much pig, much
wallaby - we are stuffed, truly. We have angered you, but now we are
thinking and our stomachs are bad because we have done bad."
But the women did not come down. They balanced suspended from the
branches and responded in song. "Koki, Kokikokil" The men cried all
night long.
The next day, the women were still hanging from the tree. Then the
men went into the forest, taking strong sticks. They fitted them to their
axes of stone. Then they returned and began to cut the trunk of the tree to
make it fall. The women made fun of them. The men were furious and
struck the tree with all their might. The tree was going to fall. Then the
women, becoming afinama, cried "Koki, koki, kokf and saved them-
selves on another tree.
One alone did not fly away very quickly. She was held in the branches
and fell to the ground. She was killed. The men picked her up, gave her
to the old woman, and said, "Cook it, perhaps she will be good to eat."
The woman cooked it, and they ate it. It was very good. All the other
afinama from the height of their tree, on seeing this, became very afraid,
and quickly they went to hide in the forest. And because they were afraid,
they never see the day. At night, they come out all together. They eat the
nuts of breadfruit trees and also that which they find in the gardens,
especially bananas. It is thus that afinama were born, and people eat
them, and the afinama eat the fruit of the gardens out of anger at the
men.
 Notes

1 Introduction: the problem and the people

1. Wallace (1956) in actuality posits five phases, but the first (I. Steady State) and the last
(V. New Steady State) clearly are redundant.
2. I shall discuss this specific substantive topic further in Chapters 6 and 10 by
specifically comparing culturally ideal group constitution with documented group
composition, on the one hand, and generally discussing the relation of structure to
event on the other.
3. This region corresponds roughly with the North Mekeo Census Division of the
Kairuku Sub-District of the Central Province. However, not all villages contained
within the North Mekeo administrative area are Bush Mekeo in terms of their cultural
identity and social participation.
4. Kuipa Bush Mekeo villages were contacted one by one over the next twenty or so
years.
5. It seems very likely that the early years of contact with Europeans had at least one
unintended consequence to the same effect, nonetheless - severe depopulation - as
wave after wave of newly introduced disease swept up and down the coast and inland.
See Mosko (n.d.).
6. This is due in part to an infestation of the rhinoceros beetle, which kills the palms. To
date, the Amoamo villages have been the hardest hit.

2 Between village and bush

1. Literally, "on the skin" (fa).
2. Eating the indigenous diet from our very first day at the village, my wife and I very
soon fell into step in this regard.
3. Reporting from among the Central Mekeo, Hau'ofa (1981:42-4) polysemously renders
ina as "bowels," "womb," "beginning," "pregnant," "conceive," and "mother." See
below and Chapter 6.
4. Nowadays, under mission influence, however, a shrine to the local patron saint is
singularly erected in the middle of the abdomen.
5. Under government order, cemeteries are now erected in the bush adjacent to the
village.
6. The first Europeans to visit the Bush Mekeo territory in the late nineteenth century
were initially perceived as invading faifai.
7. The term for "pig (kuma) also has the meaning "body" (kuma). The pigs of ongokapu
spirits are real, nonspirit animals, usually of the species associated with faifai. Thus,
faifai are simultaneously the pigs and the bodies of ongokapu.
8. Tetanus?
9. As far as I am aware, there is no kind of burial beneath clubhouses.

271
272 Notes to pp. 38-69

3 Body and cosmos

1. I was unable to uncover a single indigenous term for "clean" that embraced both
mitsia and etsiu, but English-speaking villagers themselves used this gloss.
2. Although certain details of animal and plant taxonomy are relevant in numerous
contexts discussed in this and subsequent chapters, the complete system lies well
beyond the limitations of this work.
3. Indigenous categories of ingestibles are listed, albeit far from exhaustively, in
Appendix 2.
4. See Chapter 5.
5. See the Amaka myth, Chapter 4.
6. Certain elements of the indigenous Bush Mekeo understandings of work and
nonwork bear a striking resemblance to the orthodox Marxian labor theory of value;
see Marx (1906, 1971); Lefebvre (1968:41-50).
7. People are much more circumspect about disposing of their bloodless body leavings
when they travel to neighboring villages for fear they might be stolen by enemies
and given to sorcerers to cause them illness; see later discussion in Chapter 3.
8. As well as animal faifai blood; see Chapter 2.
9. See Chapter 4 regarding further meanings and transformations with areca and betel
chewing.
10. And presumably bush animals; I cannot say for faifai species.
4 Sex, procreation, and menstruation
1. Cf. Chapter 3 concerning mefu ritual. Peace sorcery will also be explained at greater
length in Chapter 5.
2. Because Bush Mekeo men are potentially polygynous, sex outside marriage for them
is adulterous only when the woman involved is married to another man. A married
man's union with a single girl is not adulterous, according to traditional views.
3. This and related sexual overtones of male ritual specialization, including sorcery, will
be discussed at length in Chapters 5, 7, and 8.
4. Excepting the newlywed phase of marriage, when the bride does not cook; see
Chapter 5.
5. These categories and rules are described in Chapters 6 and 7.
6. It is perhaps significant that human fetuses gestate roughly the same length of time
as in each of the two seasons of the year; see Chapter 3.
7. In private, women who already had many children occasionally expressed consider-
ably less enthusiasm.
8. In prepacification times, especially, when the threat of attack was ever present, there
was additional reason for requiring the husbands of all pregnant and nursing women
to keep distant from their wives; see Chapter 5.
9. Lactation is discussed more fully in Chapter 5.
10. Guis (1936:172) implies that perhaps women will burn menstrual blood - but in the
hearths used for cooking food? Thinking of this possibility, I have devised an
alternative explanation of this ethnographic curiosity since returning from the field,
as follows: Burning of meat or vegetables directly in fire destroys the blood and/or
blood-elements of the foods, making them unsweet and cold for blood synthesis.
Similarly, with menstrual blood, its burning might conceivably destroy its blood-
elements and thereby transform it from a hot bloody to a cold bloodless dirt.
Consistent with this interpretation, a married woman's kitchen hearth accumulates
ashes of her cooking fire throughout her life, and only upon her death are the ashes
overturned onto the village abdomen to be disposed with other rubbish in the
adjacent bush.
 Notes to pp. 70-98 273

11. Memengoa birth techniques are described in the following chapter.

12. Note at the beginning of the myth that upstream Amaka has both a wife and child.
13. Villagers do not "eat" (am) areca, betel, and lime; they merely "chew" (euwa) it and
then spit it all out of their mouths.

5 Male and female

1. Many of the ethnographic details I employ in this chapter can be referenced to
similar practices and conceptualizations in Egidi (1912) and Hau'ofa (1981).
2. Unless the groom is widowed or divorced, it is very unlikely that he will already
have established a household separate from that of his parents or, since becoming a
bachelor (koaekongo), his unmarried clansmen; cf. Williamson (1913:276); Hau'ofa
(1981:57-9).
3. Buibui, a cold bloodless dirt, is regarded as ugly and sexually unsweet to persons of
the opposite sex. Bachelors and married men also depilate their bodies as a matter of
course. All hair, scales, and feathers of animal species are also categorically buibui.
Only humans possess fufu on their heads, and, although fufu is also cold bloodless
dirt, when it is fully combed into a nicely rounded coiffure, it is regarded as sexually
sweet.
4. See Chapter 7 as regards head shaving and beard growth during mourning obser-
vance; cf. Leach (1958b).
5. The indigenous categories of these relationships are discussed at length in Chapters 6
and 7.
6. Wallabies of two varieties, matsi and mani, are also small meat, but brides may not
eat them because they are the flesh of the culture hero Akaisa and, for that reason,
hot; see Chapter 8 and Appendix 3.
7. Single girls need not fear conceiving afaifai baby, villagers insist, unless they are
sexually active enough with human males to conceive simultaneously a normal human
baby. Otherwise, however, single girls who eatfaifai are still vulnerable, just like
other villagers would be, to faifai illnesses.
8. Because hunting, fishing, and nonaggressive garden ritual do not employ hot, bloody,
and dirty things, brides and pregnant women are not cold or vulnerable to them; see
Chapter 3.
9. Presumably, this is another sweet/unsweet transformation.
10. I suspect that pregnant women eat big meat for their own bodily sustenance, and
continue eating small meat to nourish their developing fetuses.
11. According to Guis (1936:62), Mekeo women traditionally suckled baby pigs and dogs.
12. Categorically, the hunting and fishing the groom does to supply his bride with meat
is not "work" (pinaunga).
13. Chili and ginger medicines are also eaten for routine curing of illness, especially
abdominal ailments; see Chapter 3.
14. Villagers steadfastly assert that long ago men lived without being born of women and
without dying; they just shed their old skins "like prawns and snakes" and grew new
skins.
15. Alternately pronounced Oa Rove, Walope, or Oa Ngope; see also the culture hero
Akaisa, Chapter 8.
16. Women, while I was in the village, held their meetings on the ground of the village
abdomen, and the men held theirs of course on clubhouses of clan chiefs.
17. "Weak" is occasionally used as a euphemism for pregnancy in women.
18. The question of the edibility or inedibility of female meat posed here will be of
critical concern to interpreting the ritual of the death feast, so I will wait until
Chapters 7 and 8 to discuss it at length in that context.
274 Notes to pp. 103-137

6 Kin, clan, and connubium

1. By far the majority of contemporary marriages to persons outside the tribe are with
peoples who are not Bush Mekeo, and most of these involve persons who have left
the village and work in the towns and cities of the country.
2. Literally, "senior-junior siblings. '
3. These terms are portrayed in the third-person plural personal possessive; au — "my
senior sibling(s)," atsiu = "my junior sibling(s)," etc. It is important to note that
here, as with all other terminological distinctions excepting "father" (ama) and
"mother" (ina), the sexes of the persons speaking or spoken to are irrelevant. Also,
villagers employ the same nomenclature for both reference and address.
4. It is my distinct impression that precise genealogical knowledge well beyond this
range is commonly possessed, at least by older men. In interviews, villagers typically
have much more information about the pedigrees of others than they have, or admit
to having, of their own. Also, the names of patrilineal ancestors are incorporated into
secret ritual spells (menga) possessed by their male descendants. Among the spells I
was able to record, there were between six and fourteen names, all purportedly
patrilineal ancestors who also owned the spells.
5. Nonetheless, working together in this particular context does connote a particularly
significant form of kin relationship through "de-conception"; see Chapter 7.
6. Clan and village names used in this chapter are pseudonyms.
7. It is interesting to note that, according to Egidi's genealogies, of the five children of
this man, three married back into other lineages of NA.
8. An Austronesian-speaking mountain people.
9. Katsi is the name of their natal clan.
10. This is the man mentioned in Chapter 2.
11. Similar to Fortes's notion of "complementary filiation" (1959). However, as indicated,
Bush Mekeo recognize in papie ngaunga ties more than filiation strictly.
12. And, intergenerationally, they and their respective parents.
13. Kofu = "chief's clubhouse"; apie = "half," in Central Mekeo, ufuapie, and in Roro,
aruabira. See Seligmann (1910:348-65); Guis (1936:39-40); Haddon (1901:358);
Belshaw (1951:4); Hau'ofa (1981:160).
14. These figures do not include 4 cases of marriage within the same local clan or within
the dispersed clan. Apparently, there has always been a minor but persistent
frequency of these marriages; see Seligmann (1910:365).
15. The number of extratribal marriages has gradually increased since pacification along
with the establishment of extratribal friend relationships; see Figure 6.7.
16. The year 1965 was chosen here and above because it marks the first time marriage
compensation was paid with money, and for a betrothal no less.
17. Of course, this does not preclude persons in other patrilines or lineages of Ego's clan
and generation who are not papie ngaunga to Ego's mother's clan from marrying
there; see discussion below.
18. Although it might appear in my figures here that Bush Mekeo are represented as
practicing actual sister exchange in violation of their own marriage rule V discussed
above, it must be noted that these relationships are classificatory, that is, exemplary
of the culture categories; also see below.
19. Compare these complementary Bush Mekeo processes with Schneider's depiction of
two distinct elements of American kinship - the union of opposites in sexual
intercourse, and the separation of identities between parents and children (1980:39-
40).
20. A third private or "hidden" (onge wake) transaction between parents of the bride and
groom will be discussed in Chapter 8.
 Notes to pp. 137-168 275
21. Regarding the notion of "exemplar" in this chapter, see Schneider (1980:96).
22. "Valuables" (kefu) appropriated for marriage compensation include numerous kinds of
shell articles, strings of dogs' teeth, string bags, clay pots, bird of paradise skins,
pigs, and other articles.
23. See the cognate notion, pange death-feast prestation, discussed below and in
Chapters 7 and 8.
24. According to Bush Mekeo conceptions, however, the bride and groom who possess
different bloods upon marriage-compensation payment subsequently become one
blood upon the birth of their first child. The ideology of shared blood in this instance
encompasses persons with cognatic descendants, instead of ancestors, in common.

7 Feasts of death (i): de-conception and re-conception

1. I anticipate taking up the theories of Van Gennep and Hertz as well as contemporary
views of this subject in a subsequent publication.
2. The perspective of social continuity I employ here resembles Weiner's (1980) notion
of "reproduction." However, I do not so hastily discard the notion of reciprocity, and
I view the continuity of Trobriand Island society quite differently from Weiner; see
Chapters 9 and 10; cf. Weiner (1976).
3. This ritual specialization reportedly has been imported from the non-Austronesian
Toaripi peoples of the Papuan Gulf. Among the Bush Mekeo, crocodiles are never
regarded as spirits in animal form, like faifai, but as real animals directed by human
specialists.
4. See sweet/unsweet relativism, discussed in Chapter 3.
5. The surviving spouse is said to have been one blood with the deceased after the two
have produced their first child.
6. Each spouse of the owners remains different blood with respect to the owners and
the deceased, except for their own particular mates among the party of mourners,
assuming they have produced at least one child.
7. Literally, katsiamore translates as "ground-putting." Nowadays, and perhaps in the
past as well, the burial feast is termed "extinguishing the fire" (ito epalo) in reference
to the fire kept burning during the vigil to smoke bush meat. However, the term ito
also means "vagina," and it will be seen below that the related notion "extinguishing
the vagina" is not entirely lacking in significance in this context.
8. Smoking tobacco and chewing areca nut may also become bafu.
9. The term onge apua is used for marriage betrothal; see Chapter 6. But unlike marital
onge apua, garden onge apua involves ritual appealing to spiritual ongokapu bush
people to attack thieves and violators of the taboos. Moreover, the Bush Mekeo
possess no tradition of masked kaivakuku enforcers of the taboo; cf. Seligmann
(1910:299-301, 314).
10. See Chapter 3 regarding the relation of work to eating.
11. Nowadays, fowl are sent with feasting invitations, villagers explain, because the
people are adopting European customs such as the avoidance of dog meat. No
invitations are sent for burial feasts.
12. The arrival of the kofuapie is described at some length by Seligmann (1910:360-2).
13. At burial feasts, male laborers kill the pigs and dogs behind, not in front of, the
clubhouse.
14. Into the adjacent waste-disposal region of the bush; see Chapter 2.
15. For lack of space, I do not describe these personal exchanges at length. But briefly,
men maintain their own hereditary reciprocal interclan exchange or trade rela-
tionships known as ekefaka - the same term used for nonagnatic siblings beyond first
cousins. In some contexts, ekefaka trade partners are likened to ekefaka relatives.
276 Notes to pp. 169-221
Also, ekefaka trade partners are frequently compared to chiefly friends (pisaua).
Peace chiefs have no ekefaka trade partners of their own; they only have friends. All
other men of the subclan, including peace sorcerers, war chiefs, and war sorcerers,
may have trade partners but no friends of their own except by virtue of those they
have through the peace chief. Like friends, trade partners within the tribe recipro-
cate village pig meat when their clans make final death feasts. On other occasions,
trade partners reciprocate raw bush meat and plant food or valuables; cf. Hau'ofa
(1981:154-60).
16. Nowadays, a chicken or a joint of village pig — lower jaw, hindleg, or foreleg - is
used instead of dog.
17. Kou structures are not used in burial feasts; rather, food is stacked in piles on the
floor of the peace chief's clubhouse before it is carried to the friends of the feast-
giving chief.
18. Compare the notion of ngope tightening in the contexts of binding two sticks
together and male tightening ritual (Chapter 5) with the closing of interclan relations,
discussed later in Chapter 7.
19. That is, by remarking that he or she eats bafu when his/her relative has died.
20. Regarding the significance of "noise," see Needham (1967) and Leach (1976:63).
21. Compare the terms of this distinction with that of engorging bride versus tightening
male (Chapter 5).

8 Feasts of death (ii): the sons of Akaisa

1. Au akaisa officeholders therefore may not eat the meat of dogs either, but for a
slightly different reason; see later section of this chapter.
2. More will be said of this comparison below.
3. As regards eating a food for the first time and vomiting, see Chapters 4 and 5.
4. Matsi and mani, two species that are "brothers." The former is elder, and the latter
younger.
5. There would seem to be an indirect reference here to the growth capacity of iunge
fanga, as argued earlier in the discussion of the Akaisa myth. Old men, closer to death
themselves, characteristically prefer the fatty backskin to all other foods.
6. One might easily speculate that wallaby and dog legs are broken to keep them from
running away should they revive, as in the Akaisa myth.

9 Tikopia and the Trobriands

1. I am, of course, enormously indebted to the several ethnographers of the cultures of
Tikopia and the Trobriands, without whose superlative descriptions of indigenous
categories my analyses would not be possible.
2. Firth's (1969:67-8) empirical reservations here, to the effect that paito are not
bilateral, are, I think, telling.
3. Malinowski (1929:30, 1966:385) mentions an indeterminate five or six ranked classes,
but specifies only these four.
4. These yam exchanges were improperly termed urigubu by Malinowski; see Weiner
(1976).
5. Powell (1960) has noted that, after the death of a village leader or district chief, there
may very likely follow a phase where not all four levels are evident as chiefly rivals
vie with one another for high office. This does not necessarily imply, however, that
during this interim there is no significance to these distinctions; see my remarks later
in the chapter.
6. Although the issue here is essentially ethnographic, Weiner's and my own theoretical
positions do differ as well. In contrast to the more or less conventional, structuralist
 Notes to pp. 223-246 277

perspective I have taken, she proposes that Trobriand social life is premised on a
division of "I" and "other," or "personal autonomy" and "social control," such that in
exchange the separation of the two is overcome at the same time that it is reinforced;
see Weiner (1976:211-26).
7. My guess is that the number of sagali exchanges at the women's mortuary ceremony,
as an exact multiple of four, is not a coincidence; see Weiner (1976:105-16).
8. Weiner records that these kaymelu bundles are given to the female owners' kinsmen
(1976:110-12). But, in actuality, it is only the name of these women's kinsmen that is
called. The wives of these men are the ones who take up and keep the bundles. To
view kaymelu as Weiner suggests simply appears wrong, for this would be the only
one of the sixteen exchanges in this series where the givers directly present wealth to
receivers of the same clan. Moreover, to give and receive lisalabadu this way plays
havoc with the notion of sagali as an act of "reclaiming."
9. If he lives with or was raised by another male relative - father's brother, mother's
brother, elder brother, etc. - a man would make his exchange gardens for him; see
Weiner (1976:146).
10. To members of subclans of his own clan, of course, a chief is not "father' but
"elder/senior of the same subclan or clan," and he correspondingly receives other
categories of annual gift: urigubu, pokala, bopokala, or guyapokala; see Malinowski
(1966); Powell (1960); Weiner (1976:204-7). Is there here yet another quadripartite
context of exchange?
11. Trobrianders, like Bush Mekeo villagers, evidently possess distinct hungers for both
plant and animal food; see Chapter 3; Malinowski (1966:43).

10 Conclusions: indigenous categories, cultural wholes, and

historical process
1. See Schneider (1980:123-4) for further methodological remarks along these lines.
2. This recalls quite vividly the analogous contradictions implicated in Bush Mekeo
contrasts of blood and nonblood relationship, both agnatically and bilaterally con-
ceived; see Chapters 6-8; A. Strathern (1972).
3. It is all the more significant that those I have explored involve transformations or
homologous inversions of the same categorical distinctions in the culture: blood versus
nonblood. See remarks earlier in the chapter concerning key or dominant symbols and
root metaphors.
 Bibliography

Allen, M. R. 1967. Male Cults and Secret Initiations in Melanesia. Melbourne:

Melbourne University Press.
Andriolo, K. R. 1981. "Myth and History: A General Model and Its Applica-
tion to the Bible." American Anthropologist 83:261-84.
Ardener, E. 1975. "Belief and the Problem of Women." In J. S. LaFontaine
(ed.), The Interpretation of Ritual, pp. 135-58. London: Tavistock.
Auge, M. 1982. The Anthropological Circle: Symbol, Function, History.
Cambridge Studies in Social Anthropology no. 37 (trans. M. Thorn).
Cambridge: Cambridge University Press.
Austen, L. 1934. "Procreation among the Trobriand Islanders." Oceania 5:102-
18.
Barbut, M. 1970. "On the Meaning of the Word 'Structure' in Mathematics."
In M. Lane (ed.), Introduction to Structuralism, pp. 367-88 (trans. S.
Gray). New York: Basic Books.
Barnes, J. A. 1962. "African Models in the New Guinea Highlands." Man
62:5-9.
Bateson, G. 1958. Naven (2d ed.). Stanford, Calif.: Stanford University Press.
Belshaw, C. S. 1951. "Recent History of Mekeo Society." Oceania 22:1-23.
Benedict, R. 1934. Patterns of Culture. Boston: Houghton Mifflin.
Bloch, M. 1982. "Death, Women and Power." In M. Bloch and J. Parry (eds.),
Death and the Regeneration of Life, pp. 211-30. Cambridge: Cambridge
University Press.
Bloch, M., and J. Parry (eds.). 1982. Death and the Regeneration of Life.
Cambridge: Cambridge University Press.
Blust, R. A. 1980a. "Early Austronesian Social Organization: The Evidence of
Language." Current Anthropology 21:205-47.
1980b. Notes on Proto Malayo-Polynesian Phratry Dualism." Bijdragen Tot
De Taal-Land-en Volkenkunde 136:215-47.
Boon, J. 1972. From Symbolism to Structuralism: Levi-Strauss in a Literary
Tradition. New York: Harper & Row.
Bourdieu, P. 1973. "The Berber House." In M. Douglas (ed.), Rules and
Meanings, pp. 98-110. Harmondsworth, England: Penguin.
1977. Outline of a Theory of Practice. Cambridge Studies in Social An-
thropology no. 16 (trans. Richard Nice). Cambridge: Cambridge Univer-
sity Press.
Brunton, R. 1975. "Why Do the Trobriands Have Chiefs." Man (n.s.) 10:544-
58.
1980. "Misconstrued Order in Melanesian Religion." Man (n.s.) 15:112-28.
278
 Bibliography 279

Bulmer, R. 1967. "Why Is the Cassowary Not a Bird?" Man (n.s.) 2:5-25.
Burridge, K. O. L. 1967. "Levi-Strauss and Myth." In E. R. Leach (ed.), The
Structural Study of Myth and Totemism. ASA Monographs no. 5, pp. 9 1 -
115. London: Tavistock.
1969a. Tangu Traditions. Oxford: Oxford University Press.
1969b. New Heaven, New Earth: A Study of Millenarian Activities. Oxford:
Blackwell.
Chalmers, J. 1887. Pioneering in New Guinea. London: Religious Tract
Society.
Chomsky, N. 1957. Syntactic Structures. The Hague: Mouton.
Denich, B. 1974. "Sex and Power in the Balkans." In M. A. Rosaldo and L.
Lamphere (eds.), Women, Culture and Society, pp. 243-62. Stanford,
Calif.: Stanford University Press.
Douglas, M. 1966. Purity and Danger: Analysis of Concepts of Pollution and
Taboo. London: Routledge & Kegan Paul.
1968. "Couvade and Menstruation: The Relevance of Tribal Societies."
Journal of Psychosomatic Research 12. Reprinted in M. Douglas (1975),
Implicit Meanings: Essays in Anthropology, pp. 60-72. London: Rout-
ledge & Kegan Paul.
1970. Natural Symbols: Explorations in Cosmology. New York: Random
House.
Dumont, L. 1970. Homo Hierarchicus: An Essay on the Caste System (trans.
M. Salisbury). Chicago: University of Chicago Press.
Dupeyrat, A. 1935. Papouasie. Paris: Dillen.
Durkheim, E. 1915 [orig. pub. 1912]. The Elementary Forms of the Religious
Life (trans. J. W. Swain). London: Allen & Unwin.
Durkheim, E., and M. Mauss. 1963 [orig. pub. 1903]. Primitive Classification
(trans. R. Needham). Chicago: University of Chicago Press.
Egidi, V. M. 1912. "Le leggi e la ceremonie del matrimonio Nella tribu di
Mekeo [Matrimonial law and ceremony of the Mekeo tribe]. Revista di
Anthropologia 17:217-29.
n.d. Genealogies of Bush Mekeo villages. Beipa'a, Central Province, Papua
New Guinea: Sacred Heart Mission.
Evans-Pritchard, E. E. 1940. The Nuer: A Description of the Modes of
Livelihood and Political Institutions of a Nilotic People. Oxford: Oxford
University Press.
1956. Nuer Religion. New York: Oxford University Press.
1962. Social Anthropology and Other Essays. New York: Free Press.
Eyde, D. 1969. "On Tikopia Social Space." Bijdragen Tot De Taal-Land-en
Volkenkunde 125:64-70.
1976. "Dualism in Trobriand Culture." In A. Kaeppler and H. H. Nimmo
(eds.), Directions in Pacific Traditional Literature. Bishop Museum Spe-
cial Publication no. 62. Honolulu: Bishop Museum Press.
1983. "Sexuality and Garden Ritual in the Trobriands and Tikopia: Tudava
Meet the Atua I Kafika." Mankind 14.
Fernandez, J. W. 1974. "The Mission of Metaphor in Expressive Culture."
Current Anthropology 15:119-45.
280 Bibliography

Firth, R. 1930. "Totemism in Polynesia." Oceania 1:291-321, 377-98.

1936. We, the Tikopia: A Sociological Study of Kinship in Primitive Poly-
nesia. London: Allen & Unwin.
1939. Primitive Polynesian Economy. London: Routledge & Kegan Paul.
1961. History and Traditions in Tikopia. Memoir no. 33 of the Polynesian
Society. Wellington: Polynesian Society.
1967a. Tikopia Ritual and Belief. Boston: Beacon Press.
1967b. The Work of the Gods in Tikopia (2d ed.). London School of
Economics Monographs on Social Anthropology nos. 1 and 2. London:
Althone.
1969. "Tikopian Social Space." Bijdragen Tot De Taal-Land-en Volkenkunde
125:64-70.
1970. Rank and Religion in Tikopia: A Study in Polynesian Paganism and
Conversion to Christianity. Boston: Beacon Press.
1973. Symbols: Public and Private. Ithaca, N.Y.: Cornell University Press.
1975. "The Skeptical Anthropologist? Social Anthropology and Marxist Views
on Society." In M. Bloch (ed.), Marxist Analyses and Social Anthropology,
ASA Monograph no. 2, pp. 29-60. London: Malaby.
1981. "A Comment," to A. Hooper, Why Tikopia Has Four Clans, Royal
Anthropological Institute Occasional Paper no. 38, pp. 45-70. London.
Fortes, M. 1953. "The Structure of Unilineal Descent Groups." American
Anthropologist 55:17-41.
1959. "Descent, Filiation and Affinity." Man 59:193-7, 206-12.
1970. "Time and Social Structure: An Ashanti Case Study." In M. Fortes
(ed.), Time and Social Structure and Other Essays. London School of
Economics Monographs on Social Anthropology no. 40, pp. 1-32. London:
Althone.
Foster, M. L. 1974. "Deep Structures in Symbolic Anthropology." Ethos
2:334-55.
Fox, J. J. 1971a. "Sister's Child as Plant." In R. Needham (ed.), Rethinking
Kinship and Marriage, ASA Monograph no. 11, pp. 219-52. London:
Tavistock.
1971b. "Semantic Parallelism in Rotinese Ritual Language." Bijdragen Tot
De Taal-Land-en Volkenkunde 127:215-55.
1973. "On Bad Death and the Left Hand: A Study of Rotinese Symbolic
Inversions." In R. Needham (ed.), Right and Left: Essays on Dual
Symbolic Classification, pp. 342-68. Chicago: University of Chicago Press.
1974. "Our Ancestors Spoke in Pairs." In R. Bauman and J. Sherzer (eds.),
Explorations in the Ethnography of Speaking, pp. 65-85. Cambridge:
Cambridge University Press.
1975. "On Binary Categories and Primary Symbols." In R. Willis (ed.), The
Interpretation of Ritual, ASA Monograph no. 3, pp. 99-132. London:
Malaby.
1980a. "Obligation and Alliance: State Structure and Moiety Organization in
Thie, Roti." In J. J. Fox (ed.), The Flow of Life: Essays on Eastern
Indonesia, pp. 98-133. Cambridge, Mass.: Harvard University Press.
1980b. "Models and Metaphors: Comparative Research in Eastern Indo-
 Bibliography 281

nesia." In J. J. Fox (ed.), The Flow of Life: Essays on Eastern Indonesia,

pp. 327-33. Cambridge, Mass.: Harvard University Press.
Fox, R. 1967. Kinship and Marriage. Harmondsworth, England: Penguin.
Friedman, J. 1974. "Marxism, Structuralism and Vulgar Materialism." Man
(n.s.) 9:444-69.
Gell, A. 1975. Metamorphosis of the Cassowaries: Umeda Society, Language
and Ritual. London School of Economics Monographs on Social An-
thropology no. 51. London: Althone.
Geertz, C. 1972. "Deep Play: Notes on the Balinese Cockfight." Daedalus
101:1-37.
1973. "Thick Description: Toward an Interpretive Theory of Culture." In C.
Geertz, The Interpretation of Cultures: Selected Essays by Clifford
Geertz, pp. 3-30. New York: Basic Books.
1974. "From the Native's Point of View: On the Nature of Anthropological
Understanding." Bulletin of the American Academy of Arts and Sciences,
October 1974. Reprinted in J. L. Dolgin, D. S. Kemnitzer, and D. M.
Schneider (eds.) (1977), Symbolic Anthropology, pp. 480-92. New York:
Columbia University Press.
Godelier, M. 1970. "System, Structure and Contradiction in Das Kapital." In
M. Lane (ed.), Introduction to Structuralism, pp. 340-58. New York:
Basic Books.
1971. "Myth and History." New Left Review 69:93-112.
1977. Perspectives in Marxist Anthropology (trans. R. Brain). Cambridge:
Cambridge University Press.
Goldman, I. 1970. Ancient Polynesian Society. Chicago: University of Chicago
Press.
Goodenough, W. H. 1955. "A Problem in Malayo-Polynesian Social Organiza-
tion." American Anthropologist 57:71-83.
Goody, J. 1962. Death, Property and the Ancestors. Stanford, Calif.: Stanford
University Press.
Groves, M. 1963. "Western Motu Descent Groups." Ethnology 2:15-30.
Gudeman, S. F. 1976. Relationships, Residence and the Individual. London:
Routledge & Kegan Paul.
Guis, J. 1936. La vie des Papous [The life of Papuans]. Paris: Dillen.
Haddon, A. C. 1901. Headhunters: Black, White and Brown. London:
Methuen.
Hage, P. 1976. "The Atom of Kinship as a Directed Graph." Man (n.s.)
11:558-68.
Hage, P., and F. Harary. 1981. "Pollution Beliefs in Highland New Guinea."
Man (n.s.) 16:367-75.
Hallpike, C. R. 1969. "Social Hair." Man (n.s.) 4:256-64.
1976. "Is There a Primitive Mentality?" Man (n.s.) 11:253-70.
Hammell, E. A. 1972. "The Myth of Structural Analysis: Levi-Strauss and the
Three Bears." Addison-Wesley Module in Anthropology no. 75. Reading,
Mass.: Addison-Wesley.
Harrell, B. B. 1981. "Lactation and Menstruation in Cultural Perspective."
American Anthropologist 83:796-823.
282 Bibliography

Harris, M. 1968. The Rise of Anthropological Theory. New York: Crowell.

1980. "History and Ideological Significance of the Separation of Social and
Cultural Anthropology." In E. B. Ross (ed.), Beyond the Myths of
Culture: Essays in Cultural Materialism, pp. 391-407. New York: Aca-
demic Press.
Harrison, J. E. 1903. Prolegomena to the Study of Greek Religion. Cambridge:
Cambridge University Press.
1912. Themis: A Study of the Social Origins of Greek Religion. Cambridge:
Cambridge University Press.
Hau'ofa, E. 1971. "Mekeo Chieftainship." Journal of the Polynesian Society
80:152-69.
1981. Mekeo: Inequality and Ambivalence in a Village Society. Canberra:
Australian National University Press.
Hertz, R. 1960 [orig. pub. 1909]. Death and the Right Hand (trans. R. and C.
Needham). New York: Free Press.
Hicks, D. 1976. Tetum Ghosts and Kin: Fieldwork in an Indonesian Communi-
ty. Palo Alto, Calif.: Mayfield.
Hoebel, E. A. 1968. The Law of Primitive Man: A Study in Comparative Legal
Dynamics (2d ed.). Cambridge, Mass.: Harvard University Press.
Hooper, A. 1981. Why Tikopia Has Four Clans. Royal Anthropological
Institute Occasional Paper no. 38.
Hugh-Jones, C. 1980. From the Milk River: Spatial and Temporal Processes in
Northwest Amazonia. Cambridge Studies in Social Anthropology no. 26.
Cambridge: Cambridge University Press.
Hugh-Jones, S. 1979. The Palm and the Pleiades: Initiation and Cosmology in
Northwest Amazonia. Cambridge Studies in Social Anthropology no. 24.
Cambridge: Cambridge University Press.
Humphries, W. R. 1923. Patrolling in New Guinea. London. Fisher & Unwin.
Huntington, R., and P. Metcalf. 1979. Celebrations of Death: The An-
thropology of Mortuary Ritual. Cambridge: Cambridge University Press.
Hutchins, E. L. 1980. Reasoning in Discourse: An Analysis of Trobriand Land
Litigation. Cambridge, Mass.: Harvard University Press.
Jakobson, R. 1948. "The Phonetic and Grammatical Aspects of Language in
Their Interrelations." Actes du VIe Congres International de Linguistics.
Paris.
Jamous, R. 1981. Honneur et Baraka: Les Structures Sociales Traditionales
dans le Rif [Honor and baraka: traditional social structures in the Rifj
(Atelier d'Anthropologie Sociale). Cambridge: Cambridge University Press
and Editions de la Maison des Sciences de FHomme.
Keesing, R. 1981. Cultural Anthropology: A Contemporary Perspective (2d
ed.). New York: Holt, Rinehart and Winston.
Kelly, R. C. 1977. Etoro Social Structure: A Study in Structural Contradic-
tion. Ann Arbor: University of Michigan Press.
Kluckhohn, C. 1942. "Myths and Rituals: A General Theory." Harvard
Theological Review 35:45-79.
Kroeber, A. L. 1909. "Classificatory Systems of Relationship. "Journal of the
Royal Anthropological Institute 39:77-84.
 Bibliography 283

Kurtzweil, E. 1980. The Age of Structuralism: Levi-Strauss to Foucault. New

York: Columbia University Press.
Lancy, D. F., and A. Strathern. 1981. "Making Two's: Pairing as an Alter-
native to the Taxonomic Mode of Representations." American An-
thropologist 83:773-95.
Leach, E. R. 1951. "The Structural Implications of Matrilateral Cross-Cousin
Marriage." Journal of the Royal Anthropological Institute 81:23-55.
1954. Political Systems of Highland Burma. Cambridge, Mass.: Harvard
University Press.
1958a. "Concerning Trobriand Clans and the Kinship Category Tabu." In J.
Goody (ed.), The Developmental Cycle in Domestic Groups. Cambridge
Papers on Social Anthropology no. 1, pp. 120-45. Cambridge: Cambridge
University Press.
1958b. "Magical Hair." Journal of the Royal Anthropological Institute
88:147-64.
1961. Rethinking Anthropology. London School of Economics Monographs
on Social Anthropology no. 22. London: Althone.
1962a. "Genesis as Myth." Discovery 23:30-5. Reprinted in E. R. Leach
(1969), Genesis as Myth and Other Essays, pp. 7-23. London: Cape.
1962b. "Review" of R. Firth (1961), History and Traditions of Tikopia.
Journal of the Polynesian Society 71:273-6.
1964. "Anthropological Aspects of Language: Animal Categories and Verbal
Abuse." In E. Lenneberg (ed.), New Directions in the Study of Language,
pp. 23-63. Cambridge, Mass.: MIT Press.
1966a. "The Legitimacy of Solomon." European Journal of Sociology 7:58-
101. Reprinted in E. R. Leach (1969), Genesis as Myth and Other Essays,
pp. 25-83. London: Cape.
1966b. "Virgin Birth." Proceedings of the Royal Anthropological Institute,
pp. 39-50. Reprinted in E. R. Leach (1969), Genesis as Myth and Other
Essays, pp. 85-110. London: Cape.
1970. Claude Levi-Strauss. New York: Viking Press.
1976. Culture and Communication: The Logic by which Symbols Are
Connected. Cambridge: Cambridge University Press.
Leach, E. R., and J. W. Leach (eds.). 1983. New Perspectives on the Kula.
Cambridge: Cambridge University Press.
Lefebvre, H. 1968. The Sociology of Marx (trans. N. Guterman). New York:
Pantheon.
Levi-Strauss, C. 1963a. Structural Anthropology (trans. C. Jacobson and B. G.
Schoepf). New York: Basic Books.
1963b. Totemism (trans. R. Needham). Boston: Beacon Press.
1966a. The Savage Mind. Chicago: University of Chicago Press.
1966b. "The Culinary Triangle." Parisian Review 33:587-95.
1967. "The Story of Asdiwal." In E. R. Leach (ed.), The Structural Study of
Myth and Totemism, ASA Monographs no. 5, pp. 1-47. London:
Tavistock.
1969a. The Raw and the Cooked: Introduction to a Science of Mythology: 1
(trans. J. and D. Weightman). New York: Harper & Row.
284 Bibliography

1969b. The Elementary Structures of Kinship (trans. J. H. Bell and J. R. von

Sturmer). Boston: Beacon Press.
1973. From Honey to Ashes: Introduction to a Science of Mythology: 2
(trans. J. and D. Weightman). New York: Harper & Row.
1978. The Origin of Table Manners: Introduction to a Science of Mythology:
3 (trans. J. and D. Weightman). New York: Harper & Row.
1981. The Naked Man: Introduction to a Science of Mythology: 4 (trans. J.
and D. Weightman). New York: Harper & Row.
Lounsbury, F. G. 1964. "The Structural Analysis of Kinship Semantics." In
Proceedings of the Ninth International Congress of Linguists, pp. 1073-
90. The Hague: Mouton.
1965. "Another View of Trobriand Kinship Categories." In E. A. Hammel
(ed.), Formal Semantic Analysis, pp. 142-85. American Anthropologist
Special Publication 4 (67), part 2.
Lowie, R. 1968. "Relationship Terms." In P. Bohannan and J. Middleton
(eds.), Kinship and Social Organization, pp. 39-59. New York: Natural
History Press.
MacCormack, C. P., and M. Strathern (eds.). 1980. Nature, Culture, and
Gender. Cambridge: Cambridge University Press.
MacDougall, L. 1975. "The Quest of the Argonauts." In T. R. Williams (ed.),
Psychological Anthropology. Paris: Mouton.
MacGregor, W. 1890. "Report on Inspection of the Mekeo District." British
New Guinea Annual Report 1889-90: 88-90.
Mair, L. 1970. Australia in New Guinea (2d ed.). London: Christophers.
Malinowski, B. 1922. Argonauts of the Western Pacific. London: Routledge &
Kegan Paul.
1929. The Sexual Life of Savages in North-western Melanesia. London:
Routledge & Kegan Paul.
1932. The Sexual Life of Savages in North-western Melanesia. (3d ed.).
London: Routledge & Kegan Paul.
1948. Magic, Science and Religion and Other Essays. New York: Free Press.
1965 [orig. pub. 1935]. Coral Gardens and Their Magic, (2 vols., 2d ed.).
Bloomington: Indiana University Press.
Maranda, P., and E. K. Maranda. 1971. Structural Models in Folklore and
Transformational Essays. The Hague: Mouton.
Marx, K. 1906. Capital, vol. 1 (trans. S. Moore and E. Aveling). Chicago:
Charles H. Kerr.
1971. The Grundrisse (ed. and trans. D. McLennan). New York: Harper &
Row.
Mauss, M. 1967 [orig. pub. 1925]. The Gift (trans. I. Cunnison). New York:
Norton.
Maybury-Lewis, D. 1969. "Science or Bricolage." American Anthropologist
71:114-21.
Meggitt, M. M. 1964. "Male-Female Relationships in the Highlands of
Australian New Guinea." In J. B. Watson (ed.), New Guinea: The Central
Highlands, pp. 204-24. American Anthropologist Special Publication no. 4.
 Bibliography 285

Montague, S. P. 1971. "Trobriand Kinship and the Virgin Birth Controversy."

Man (n.s.) 6:353-68.
1973. Correspondence: "Copulation in Kaduwaga." Man (n.s.) 8:304-5.
1983. "Trobriand Gender Identity." Mankind 14.
Moore, S. F. 1964. "Descent and Symbolic Filiation." American An-
thropologist 66:1308-20.
Mosko, M. S. 1980. "The Legacy of Akaisa: Categories, Relations and Homolo-
gies in Bush Mekeo Culture." Ph.D. dissertation, University of Min-
nesota. Ann Arbor, Mich.: University Microfilms.
1983. "Conception, De-Conception, and Social Structure in Bush Mekeo
Culture." Mankind 14:24-32.
n.d. "Peace, War, Sex and Sorcery: The Response to Pacification among the
Bush Mekeo." Unpublished manuscript.
Needham, R. 1962. Structure and Sentiment. Chicago: University of Chicago
Press.
1967. "Percussion and Transition." Man (n.s.) 2:606-14.
1973. Right and Left: Essays on Dual Symbolic Classification. Chicago:
University of Chicago Press.
1979. Symbolic Classifications. New York: Random House.
Ortner, S. 1973. "On Key Symbols." American Anthropologist 75:1338-46.
Ostor, A. 1980. The Play of the Gods: Locality, Ideology, Structure and Time
in the Festivals of a Bengali Town. Chicago: University of Chicago Press.
Papua New Guinea. 1970-1. "Report on Patrol No. 5 of 1970-71." Bereina,
Central Province, Papua New Guinea: Department of District Administra-
tion, Kairuku Sub-District.
Park, J. 1973. "A Consideration of the Tikopia 'Sacred Tale. '"Journal of the
Polynesian Society 82:154-75.
Pepper, S. 1942. World Hypotheses: A Study in Evidence. Berkeley: Univer-
sity of California Press.
Piaget, J. 1949. Traite de Logique. Paris: Colin.
1970. Structuralism (trans. C. Maschler). New York: Basic Books.
Powell, H. A. 1960. "Competitive Leadership in Trobriand Political Organiza-
tion." Journal of the Royal Anthropological Institute 90:118-45.
1968. Correspondence. "Virgin Birth." Man (n.s.) 3:651-3.
1969a. "Genealogy, Residence and Kinship in Kiriwina." Man (n.s.) 4:177-
202.
1969b. "Territory, Hierarchy and Kinship in Kiriwina." Man (n.s.) 4:580-
604.
Radcliffe-Brown, A. R. 1930. The Social Organization of Australian Tribes.
Oceania Monographs no. 1. Sydney.
1939. Taboo. Cambridge: Cambridge University Press.
Robinson, M. S. 1962. "Complementary Filiation and Marriage in the
Trobriand Islands: A Re-examination of Malinowskis Material." In M.
Fortes (ed.), Marriage in Tribal Societies, pp. 121-57. Cambridge Papers
in Social Anthropology no. 3. Cambridge: Cambridge University Press.
Rosaldo, M. Z., and J. M. Atkinson. 1975. "Man the Hunter and Woman:
286 Bibliography
Metaphors for the Sexes in Ilongot Magical Spells." In R. Willis (ed.), The
Interpretation of Symbolism, pp. 43-75. New York: Wiley.
Rubel, P. G., and A. Rosman. 1978. Your Own Pigs You May Not Eat: A
Comparative Study of New Guinea Societies. Chicago: University of
Chicago Press.
Sahlins, M. D. 1963. "Poor Man, Rich Man, Big Man, Chief: Political Types
in Melanesia and Polynesia." Comparative Studies in Society and History
5:285-303.
1972. Stone Age Economics. New York: Aldine.
1976. Culture and Practical Reason. Chicago: University of Chicago Press.
1981. Historical Metaphors and Mythical Realities: Structure in the Early
History of the Sandwich Islands Kingdom. Association of Social An-
thropologists in Oceania Special Publications no. 1. Ann Arbor: University
of Michigan Press.
1983. "Other Times, Other Customs: The Anthropology of History." Ameri-
can Anthropologist 85:517-44.
Saussure, F. de. 1959 [1916]. Course in General Linguistics (ed. C. Bally, A.
Schehaye, and A. Riedlinger; trans. W. Baskin). New York: McGraw-Hill.
Schaff, A. 1978. Structuralism and Marxism. Oxford: Pergamon.
Scheffler, H. W. 1971. "Dravidian-Iroquois: The Melanesian Evidence." In L.
R. Hiatt and C. Jayawardena (eds.), Anthropology in Oceania: Essays
Presented to Ian Hogbin. Sydney: Angus and Robertson.
Scheffler, H. W., and F. G. Lounsbury. 1971. The Siriono Kinship System: A
Study in Structural Semantics. Englewood Cliffs, N.J.: Prentice Hall.
Schneider, D. M. 1969. "Kinship, Nationality, and Religion in American
Culture: Toward a Definition of Kinship." In V. Turner (ed.), Forms of
Symbolic Action, pp. 116-25. New Orleans: American Ethnological Soci-
ety, Tulane University.
1972. "What is Kinship All About?" In R. Reining (ed.), Kinship Studies in
the Morgan Centennial Year, pp. 32-63. Washington, D.C.: An-
thropological Society of Washington.
1980. American Kinship: A Cultural Account (2d ed.). Chicago: University of
Chicago Press.
Seligmann, C. G. 1910. The Melanesians of British New Guinea. Cambridge:
Cambridge University Press.
Shapiro, W. 1981. Miwuyt Marriage: The Cultural Anthropology of Affinity in
Northeast Arnhem Land. Philadelphia: Institute for the Study of Human
Issues.
Sider, K. B. 1967. "Kinship and Culture: Affinity and the Role of the Father
in the Trobriands." Southwestern Journal of Anthropology 23:90-109.
Smith, W. R. 1957 [orig. pub. 1889]. The Religion of the Semites. New York:
Meridian.
Spier, L. 1925. The Distribution of Kinship Systems in North America. Seattle:
University of Washington Press.
Spiro, M. 1964. "Religion and the Irrational." In J. Helm (ed.), Symposium on
New Approaches to the Study of Religion, pp. 102-15. Seattle: University
of Washington Press.
 Bibliography 287
1968. "Virgin Birth, Parthogenesis, and Physiological Paternity: An Essay in
Cultural Interpretation." Man (n.s.) 3:242-61.
Stephen, M. 1974. "Continuity and Change in Mekeo Society 1890-71."
Ph.D. dissertation, Australian National University.
Strathern, A. 1972. One Father, One Blood. London: Tavistock.
Strathern, A., and M. Strathern. 1968. "Marsupials and Magic: A Study of
Spell Symbolism among the Mbowamb." In E. R. Leach (ed.), Dialectic in
Practical Religion, pp. 179-202. Cambridge: Cambridge University Press.
1971. Self-Decoration in Mt. Hagen. Toronto: University of Toronto Press.
Strathern, M. 1981. Kinship at the Core: An Anthropology of Elmdon, a
Village in North-West Essex in the Nineteen-Sixties. New York: Cambridge
University Press.
Tambiah, S. J. 1968a. "Animals are Good to Think and Good to Prohibit."
Ethnology 8:423-59.
1968b. "The Magical Power of Words." Man (n.s.) 3:175-208.
1983. "On Flying Witches and Flying Canoes: The Coding of Male and
Female Values." In E. R. Leach and J. W. Leach (eds.), New Perspectives
on the Kula, pp. 171-200. Cambridge: Cambridge University Press.
Tooker, D. E. 1979. "Some Basic Trobriand Attitudes About Sex as Expressed
in the Kula Trade." Cambridge Anthropology 5:44-65.
Turner, V. 1967. The Forest of Symbols: Aspects of Ndembu Ritual. Ithaca,
N.Y.: Cornell University Press.
1969. The Ritual Process: Structure and Anti-Structure. Chicago: Aldine.
Uberoi, J. P. S. 1962. Politics of the Kula Ring. Manchester: Manchester
University Press.
Ullmann, S. 1963. "Semantic Universals." In J. Greenberg (ed.), Universals of
Language. Cambridge, Mass.: M.I.T. Press.
Van Gennep, A. 1960 [orig. pub. 1909]. The Rites of Passage (trans. M. B.
Vizedom and G. L. Coffee). Chicago: University of Chicago Press.
Vogt, E. Z. 1969. Zinacantan: A Maya Community in the Highlands of
Chiapas. Cambridge, Mass.: Harvard University Press.
Wallace, A. F. C. 1956. "Revitalization Movements." American Anthropologist
58:264-81.
Weiner, A. B. 1976. Women of Value, Men of Renown: New Perspectives in
Trobriand Exchange. Austin: University of Texas Press.
1977. "Trobriand Descent: Female/Male Domains." Ethos 5:150-74.
1978. "The Reproductive Model in Trobriand Society." Mankind 11:175-86.
1979. "Trobriand Kinship from Another View: The Reproductive Power of
Women and Men." Man (n.s.) 14:328-48.
1980. "Reproduction: A Replacement for Reciprocity." American Ethnologist
7:71-85.
Williamson, R. W. 1913. "Some Unrecorded Customs of the Mekeo of British
New Guinea." Journal of the Royal Anthropological Institute 43:268-90.
Willis, R. G. 1967. "The Head and the Loins: Levi-Strauss and Beyond." Man
(n.s.) 2:519-34.
Worsley, P. 1968. The Trumpet Shall Sound: A Study of 'Cargo Cults' in
Melanesia (rev. ed.). New York: Schocken.
288 Bibliography

Young, M. W. 1971. Fighting With Food: Leadership Values and Social

Control in a Massim Society. Cambridge: Cambridge University Press.
1982. "Review" of E. Hau'ofa (1981), Mekeo: Inequality and Ambivalence in
a Village Society. Journal of Pacific History 17:89-90.
 Index

abdomen (ina), 27, 75, 193, 234, 271; in Aranda system, 100, 125, 134, 143-4
birth, 65-6, 70, 81; in conception, 63- Auge, M., 8, 248
7, 75, 77; in digestion, 38, 49, 52-7, 79, Austen, L., 210
193; and feast prestations, 193-5; in
illness, 90; mythical, 184; see also moth- bachelors (koaekongo): courting, 25; ritual,
er, pregnancy, village, womb 29, 33, 87, 88, 90
aboveground vs. belowground, 28-35 pas- Bateson, G., 240, 242
sim; birth blood, 80; menstrual blood, bathing, 75-6, 78, 79, 86
68-9; mythical, 95-6, 99 Belshaw, C , 128
adultery, 61, 69, 272 Benedict, R., 2
Afai village, dispersal, 118 betel, see areca nut
affines (ipa ngaua), 109-11, 241; clans Biaru River, xi, 14
and, 117, 123; marriage regulation and, big-man, Melanesian, 16
134-6; mother-in-law, 74-5; new mar- birth, 61, 88; catchment, 32; faifai, 31;
riage and, 74-6, 79; sweet vs. unsweet, father's role in, 67, 80-81, 88; and ill-
62; see spouse, kofuapie, marriage, kin ness, 66; and menstruation, 66; mythi-
afinama, myth of, 97-9, 101, 188, 198, cal, 70-2; out of wedlock, 73, 76-7;
265-70 techniques, 80-1; theory of, 65-7; vigil,
Akaisa, 197, 198-9, 273; mortuary ritual 79, 81, 88; see also memengoa
and, 160, 163; myth of, 183-99 passim; blood (ifa), 28, 60, 77, 78; birth, 67, 70-2,
ritual poisons of, 151 80-1, 88; conception, 75-8; death, 29-
Akaisa Men, 167, 183, 197-9, 276; de- 30, 85-6, 153-4; engorging and, 75;
conception and re-conception of, 191-9; faifai, 30; holes and, 31; illness, 52;
mythical origin of, 185-8, 191-2, 198-9; manipulating, 138-42, 234; menstrual,
as relatives, 185, 192, 197-8 33, 65, 68-9, 272; mythical, 70-2, 262-
Amaka, myth of, 69-72, 80, 83, 88, 91, 3; semen, 33, 60, 63-7, 72, 75, 76-9,
96, 187 82, 85, 101; as sorcery ingredient, 27,
ambidirectional, see unidirectional vs. 30, 33, 157-8; vs. nonblood, 52-9, 60,
ambidirectional 272; womb, 60-1, 63-7, 75-8, 82, 92;
Amoamo tribe, 15, 16, 19, 102, 271 and work vs. nonwork, 52
ancestors (au apoutsi), 39, 180 blood reciprocation, 113; in peace, 101,
Andriolo, K. R., 8, 248 103, 124, 157; in war, 102-3
animals, categories of, 39-40; bush, 39, blood relationship, female (or cognatic),
98, 175, 191-2, 196-7, 260-1; faifai, 101, 123-7, 129, 139, 175, 180, 199; in
30-1, 34, 40, 76, 79, see also bush feasting, 175, 180; in marriage compen-
people, spirits; village, 22, 39, 175, sation exchange, 138-42 passim; see also
196-7; Trobriand, 213; see also food, papie ngaunga
human vs. nonhuman blood relationship, male (agnatic), 101,
areca nut, 54, 90-2; as medicine, 54-5, 112, 113, 124, 129, 139, 149, 175-7;
86, 90; plant food and, 90 among Akaisa Men, 185, 192, 197-8; in
areca nut, chewing of, 41, 96-7, 253, 273; feasting, 175-7; in marriage compensa-
spitting, 72; at feasts, 165; mythical ori- tion exchange, 138-42 passim; see also
gin of, 70-2, 96, 97, 262; sexuality and, clan
70-2, 90-2, 96; tightening and, 86-7 blood synthesis, 48-54, 57-9, 60, 72, 85;
289
290 Index
blood synthesis (cont.) bones, 158, 186, 191-2; in the Trobriands,
body sustenance, 41; conception, 63-4, 226-7
72, 76-7, 78-9, 101; minimum for life, Bourdieu, P., 2, 244
86, 87; in nursing, 82, 92 bride, see marriage
blood, Trobriand notions of, 210-11, 226, Brunton, R., 9-10, 230, 238-40, 242
231, 232 Bulmer, R., 2
blood vs. nonblood, 11, 52, 70-2, 277; and burial, 22, 52, 56, 69, 271; faifai 33-4; as
animals, 30-1; dirt and, 52, 71-2, 257; hole, 29; mythical, 191-2; regulation of,
health vs. illness and, 52-9, 65; mortu- 17, 271; see also feasts
ary feast, 155; and sweet vs. unsweet, Burridge, K. O. L., 37
49-52, 71-2; and village vs. bush, 96 bush, 22; adjacent vs. remote, 24, 27-8,
blood vs. nonblood relationships, 101—11, 29-35, 95, 98; beings, 22, 30-1, 39,
112; feasting categories, 172-9, 158, 261, 263, see also spirits, faifai, animals,
198; manipulating, 138-42; marriage human vs. nonhuman; extraordinary
categories and, 117, 124, 127, 138, 140- sphere and, 28-38; ordinary sphere and,
2; in sorcery, 157—8; and working to- 23-8; and sweet vs. unsweet, 39, 47
gether, 117 Busk Mekeo, 1, 10—20; population, 15;
Blust, R. A., 233 villages, 15
body excretions: alimentary, 24, 27, 32, bush people, 30-5 passim, 40, 76, 79, 82,
52-7; birth, 80-1; blood vs. nonblood, 97, 160, 190, 271, 275
33, 52-7, 74, 81-3, 151, 191, see also
blood; of infants, 81-2; mythical, 70, 72, catchment, 29, 32, 33
96-7; pregnancy, 79; sexual, 61, 67-9, categories, indigenous, 1, 9—10, 12—13,
70-1, see also blood, semen, water; skin 201, 216, 234-5, 240, 242, 248
dirt, 55, 163, 168; spittle, 72, 96; tight- change, sociocultural, 12-13
ening and, 86; vaginal water, 61, 63, 64, chewing, see areca nut
70-1, 97; vomit, 70-2, 83, 96, 189 chiefs, see peace chief, war chief
body, 10, 234, 247, 271; and chewing of chiefs, in Tikopia, 203, 204
areca nut, 41, 70-7, 86-7, 90-2, 96-7; chiefs, in the Trobriands, 213-14, 217,
female cycle of, 73-84, 90—5 passim; 227-30, 231
hair, 30, 74, 161, 171; health, illness, children, 68, 104; firstborn, 73, 185; pigs
and curing of, 52-9, 151; inside vs. as, 175; see also relatives, sons
outside of, 45-59 passim, 63, 65-6, 71- clan (ikupu), 15-16, 111, 124, 173-9, 199,
2, 78, 247; as key symbol, 236, 237; 234; dispersed, 113, 114, 117, 124, 147,
male cycle of, 84—95 passim; menstrua- 149; exogamy of, 100, 117, 125, 134,
tion, 60, 64-9; in Tikopia, 202, 207; in 141-2, 199; insignia of, 74, 116, 117,
the Trobriands, 210, 211, 212, 229; vil- 166; and kofuapie, 127-33; lineage and,
lage/bush resources and wastes, and, 38; 114-15, 116-23 passim, 147-9; moiety
and work vs. nonwork, 45-50 passim; and, 112-13, 117, 198; mother's, see
womb, 63; see also sweet vs. unsweet, papie ngaunga; open vs. closed, 129,
abdomen, blood, conception, gender, 134, 139, 141-2, 177-81, 199, 234, 276,
semen, vagina, skin see also de-conception; papie ngaunga
body, inside vs. outside, 27, 38; in body and, 124; of pigs, 175; residential, 115-
sustenance, 48-59; male vs. female and, 23, 170, 171; ritual secrets of, 89, 121;
73-95 passim; and menstruation, 60, segmentary/fragmentary organization of,
64-9, 71, 75, 79; and procreation, 60-8, 111-23 passim, 124; skin of, 126, 144,
79-89; and village vs. bush, 27-9; and 154, 177, 180; subclan, 113, 123, 130,
work vs. nonwork, 45-8; see also body see also senior vs. junior
excretions, food, gender, open vs. clan officials (see also Akaisa Men, peace
closed, ritual chief, war chief, peace sorcery, war sor-
body, open vs. closed, 49, 61, 67, 73-94 cery), 114-17, 118-23, 148, 160, 167;
passim, 99, 181 succession and installation of, 113, 114,
boiling, 41; and digestion, 49; food, 42-5, 120, 167, 182-3, 197-8
57-8, 64, 75; medicines, 55; mythical clans: in Tikopia, 202—3; in the
origin of, 260, 264; water, 42; as work, Trobriands, 212-27 passim
46 clean, 38, 272; see also sweet vs. unsweet
 Index 291
closing, 73, 111, 234; clans at mortuary curing (see also medicines), 41, 52-7, 66,
feasts, 177, 181; and mourning, 157; 81, 90, 273
marriage compensation and, 139-42; customs (kangakanga), 39—40
nursing and, 82, 84; thirst and, 49; cycle, daily, 23-8, 74
tightening and, 87—90; see also opening
clubhouses, chiefs' (kofu), 25-8, 29, 30, dancing, 159, 165, 171
66, 79, 271, 273, 274, 276; during day vs. night, 23-8 passim
feasts, 167, 168, 169, 275; rear compart- death, 28, 69, 70, 150; mythical origin of,
ment of (ialiali), 29, 30, 32, 33, 35, 89 185-7, 198-9, 266, 268, 270; owners of,
cold (ekekia) (see also hot vs. cold, sweet 154-5, 156, 159-60, 172, 173; uncon-
vs. unsweet), 38, 159; body transforma- sciousness as, 185
tions, 49-57, 62-3, 65, 79; culinary death, causes of, 29, 30-1, 33, 53-7, 60-
transformations, 42-5, 49; and gender, 1, 150-2; dirt and, 53-7, 81; faifai, 30-
94; of village, 152-3, 165-6, 167, 171; 1; mefu sorcery, 54—5, 67, 157; men-
work vs. nonwork transformations, 45-8 strual blood, 68, 156; peace chiefs and,
colonialism, influences of, 13; business and 160, 185; peace sorcery, 30, 61, 68-9,
cash crops, 16, 19; Christianity, 16, 18, 114; poisons, 30, 54, 60, 61, 67, 68;
133, 153-4, 271; depopulation, 18, 271; snakes, 30-1; starvation, 41, 52, 76
latrines, 17; pacification, 13, 15, 16, 18; de-conception (engama) (see also concep-
taxation, 16; wage labor, 18-19 tion, happiness), 177-81, 181-9 passim,
colors, 165-6, 170-1; black, 167, 170; red, 234; in Trobriand mortuary ritual, 221,
74, 165, 170; white, 82; yellow, 170 224-7
comparison, methodology of, 4, 9-10, 12, digestion, 49-52
200-1, 212 dirt (iofu), 28, 38, 52, 151, 247; birth and,
conception (engama), theory of (see also 67, 81, 88; burial and, 157; chewing
happiness, de-conception), 11, 60, 64-8, and, 91; cooking and, 68; and edibility,
40, 78, 81; feasts and, 172-3; food and,
92, 164, 167, 211, 234, 271; areca nut 52-7, 60-3, 172; health vs. illness and,
chewing and, 90-2; faifai animals and, 52-7, 63-9; infants and, 81-4;
31, 34, 76, 79, 273; feasting and, 177, memengoa techniques and, 61-2, 65,
194; female's role in, 73-80, 84; illness 67, 70-2, 80-4, 90; menstruation and,
and, 63; male's role in, 85-8, 89; mar- 64-9 passim; mythical origin of, 69-72,
riage compensation and, 139; menstrua- 97; nursing and, 81-2; pregnancy and,
tion and, 64-9 passim; mythical origin 79-80; sexuality and, 60-95 passim; of
of, 96-7, 198-9 skin, 55, 81, 163, 168; tightening and,
conception, theory of (Tikopia), 201-2 87-8; see also blood vs. bloodless,
conception, theory of (Trobriand), 210-12, rubbish
226, 227, 229, 230 dirty vs. clean, 28, 38, 40, 57, 60, 62, 94-
contradiction, see structure and 5, 96-8
contradiction districts, in Tikopia, 203-4
cooking (see also food, boiling, ripening, districts, in the Trobriands, 217
roasting, smoking), 25, 48, 98; and dogs, 15, 24; adultery and, 61; feasts and,
blood relationship, 179; gender and, 64; 127, 162, 164-5, 169-71, 196-7, 275,
menstruation and, 68; mythical origin 276
of, 70-2, 96-7, 260-1, 262, 268; and Douglas, M., 2, 7, 246
nonfood resources, 45; sexual inter- dreams, 31
course and, 61-3 drinking, 41, 49, 75, 79, 85, 87, 96-7, 252
courting, 65, 159; ritual of, 27, 65, 68-9, dry vs. wet, 42, 66, 87, 174, 258; bodily,
165, 166-7 50-2, 53-7, 58, 66, 73-95 passim, 186;
crocodiles, 275 chewing and, 90-2; conception and, 64-
culture-as-constituted vs. culture-as-lived, 5, 75, 83, 90-2; cooking and, 42-5, 57-
10, 13, 237-8, 239, 242-5 8, 174
Culture/Nature, see Nature/Culture dualities, bisected, see structures
culture, and society, 10 dualism, see structures
cultures as wholes, 1-13 passim, 201, 208, Dumont, L., 2
234-44 passim, 249 Durkheim, E., 2, 21, 237
292 Index
eating, 41, 251-2; chewing and, 90-2, mony, 167-70; laborers at, 154-68 pas-
273; engorging and, 74-7, 78, 84; male sim; mourners at, 161-71 passim, 179,
sexuality and, 85; mythical, 70-2, 96-7, 180, 191-2, 234; mythical origin of,
258-9, 262, 263; playing at feasts and, 191-2; peace chiefs and, 116, 120, 130,
163-4; sexual intercourse and, 60-3; 159, 160-71 passim; provisioning of,
tightening and, 86-7 159, 160, 161, 163-5; release of mourn-
ekefaka sibling terminology alternative, ing at, 159-60, 162, 170-1, 179, 191;
112, 124-5, 134, 275-6 umupua ceremony, 167, 170—1
empiricism, 1, 5-6, 9-10, 12-13, 37, 239 feast prestations, 127-8, 198; burial, 158-
engama, see happiness, conception, de- 60, 170, 275; dirty pig, 167-8, 169;
conception ekefaka trade partners, 276; ikufuka,
endogamy, see exogamy and endogamy, 169-71, 172-9, 188, 191, 234; iunge
tribe fanga, 169-71, 183, 193, 195-6, 276;
engorging, 74-8 passim, 84 onge kau, 165; pange, 172-9, 188, 191,
ethnocentrism, 8-10 234
Etoro society, 245-6 female, see gender
Evans-Pritchard, E., 2, 235 Fernandez, J. W., 2
exchange, restricted, 101, 129 fire (ito) (see also vagina, hot), 234; birth
exchanges, coastal/mountain, 16 and, 80; burial feasts and, 275; food
exchanges, food (see also feasts): daily, transformation and, 43-5, 48, 49, 53,
between households, 25; daily, on 64; menstrual blood and, 272; mythical
chief's clubhouse, 25; engorging bride, origin of, 70-1, 96, 259-60, 261-2, 263;
74-7; installation ceremonies for clan sexual intercourse and, 63—4; sorcery
officials, 182-3; mythical, 96-9, 189-91; and, 54; tightening and, 87; water trans-
of postpartum mother, 88; of pregnant formation and, 42-3
woman, 78-9; of single girl, 76-7, 79; in Firth, R., 200-8 passim
Tikopia, 204, 205, 207; in the fish, 184
Trobriands, 212, 219-29 passim; wild fishing, 15, 24, 69, 76, 85, 88; for feasts,
game and, 47 163; mythical, 184-5; see also hunting
exchanges, mortuary, see feasts Foikale, myth of, 95-7, 98-9, 101, 187,
exogamy vs. endogamy, 15, 16 198, 258-64
extraordinary sphere, 21; bodily, 58-9, 61, food, categories of (see also meat, plant
69, 70-1, 99; feasts and, 181; gender food, feasts, sweet vs. unsweet), 41,
and, 34, 99; village-bush transfers, 28- 162, 179, 254-5; and dirt, 52-7, 60-3,
35, 35-6, 58-9, 61, 69; see also ordi- 172
nary sphere food, categories of (Tikopia), 207
Eyde, D., 206, 220, 232 food, categories of (Trobriand), 212, 227,
277
faifai, 22; as agents of death, 151; as bush food, transformations of, see cooking, boil-
beings, 30-5; compared to humans and ing, roasting, ripening, smoking
animals, 40, 272, 275; Europeans as, Foster, M., 6
271; mythical, 97, 190; nursing and, 82;
procreativity with humans, 31, 34, 76, fowl, 275
Fox, J. J., 2
79, 273; sorcery and, 151, 190
fat, 74, 75, 175, 179, 196, 276 friend, chiefly (pisaua), 129-33, 145, 147-
father (ama), 67, 104, 175; see also concep- 8, 199; mythical, 185, 261; role at feasts,
tion, birth, relatives 158, 165, 168, 170; see also kofuapie
father, Trobriand, 211, 212, 216, 218-19, functionalism, 201, 207-8, 235, 243
223-30 passim gardens, and gardening, 15, 24, 44, 48,
fauapi sorcerer's bush dwelling, 29-30, 71, 69, 78, 88, 96; mythical origins of,
31, 33, 61 95-6, 258, 261-2, 263, 270; in prepara-
feasts, mortuary, 11, 32, 40, 150, 241; tion for feasts, 159, 162-3; ritual, 29,
burial, 40, 158-60, 162, 163, 174, 178, 33, 48, 85, 275; sorcery and, 29, 33, 85
192, 276; clan recruitment and, 122-4, gardens, and gardening (Trobriand), 212,
179, 199; distribution of foods after, 171; 220, 228, 231, 232
final, 158, 160-83 passim; givers vs. Geertz, C , 4
receivers, 171, 172-4, 177; kumau cere- Gell, A., 2, 9, 234, 238
 Index 293

gender, conceptualization of, 30-1, 73, Hohfeld, W., 7-8

89-90; areca nut chewing and, 90-2; holes (ine), 29, 38, 58; bodily, 38, 58, 61;
clan identity and, 123-4, 126; cooking and extraordinary village-bush transfers,
and, 64; faifai and, 31; food categories 29-35, 38, 57, 61, 80, 87; mythical,
and, 77-9, 98-9, 175; feast prestations 95-7
and, 172-7, 193, 195; holes and, 29-30; homicide, 33, 34, 85-6, 157
inversion/eversion of, 92-5, 98; hot (tsiabu) (see also heat, ritual, sorcery,
male/female separation and, 29, 68, 83- sweet vs. unsweet transformations), 28-
4, 89-92, 98-9, 272-3; menstruation 9, 42; body transformations, 49-57, 62-
and, 64-9 passim; mourning and, 153- 7, 70-2, 75-6, 79-80, 83; and concep-
9, 161, 170-1, 180; mythical, 69-71, tion, 62-4, 65; dirt and, 38; fire and,
95-9, 259-61, 262-3, 265-70; open vs. 42-5; food transformations, 42-5, 49;
closed and, 94; ordinary vs. extraordi- gender and, 94; life and death, and,
nary spheres and, 34, 99; priority of 152-3, 165-6, 167, 171; work vs. non-
male over female, 94—5, 99; ritual cycle work and, 95-8
of female, 73—84; ritual cycle of male, houses, 24, 25; birth in, 80; burial and, 32;
85-90; sexual intercourse and procrea- faifai, 30; menstruation and, 68; mythi-
tion, and, 60-9, 73, 84, 85-6, 88-90; cal origin of, 96, 262; new marriage and,
sweet vs. unsweet and, 39 74; sexual intercourse and, 61; in
gender, division of labor for, 11; in mortu- Tikopia, 207
ary ritual, 153-4, 157, 159, 161, 163; Hugh-Jones, C., 2
mythical, 69-72, 95-9, 265-70; ordinary Hugh-Jones, S., 2
village-bush transfers and, 24-5, 26-34; Human Mind, 200
procreative, 64-7, 69-70, 73-92 passim; human vs. nonhuman: agents of death,
see also feasts, ritual 151; animals and, 39-40, 273; of bush,
gender, Tikopian conceptualization of, 202 22, 30-1, 34-5; dirt and, 55-6; faifai,
gender, Trobriand conceptualization of, 30-5, 40, 76, 79; feast pigs and, 175;
209-10, 211, 213, 227, 230, 231-2 hunting ritual and, 46-7, 260-1; in
Gennep, A. van, 6-7, 150, 248 myth, 97, 98, 260-1, 273; nursing in-
girls, single, 61, 66, 73, 76, 82 fants and, 83, 175; of village, 22; and
Godelier, M., 4-5, 243, 248 sweet vs. unsweet, 39; see also animals,
good vs. bad, 39-40 spirits
grave, 29; faifai and, 30, 35; as hole, 29, humans, 22; and ordinary vs. extraordinary
32; peace sorcery and, 31, 33; see also transfers, 28—31; and sexuality, 62; and
burial sweet vs. unsweet, 39—40
Gudeman, S. F., 2 hunting, 15, 24, 46, 76, 88, 184; for feasts,
Guis, Fr., 127-8, 132 159, 163, 164, 191-2; mythical origin of,
Hage, P. and F. Harary, 233 96, 189-90, 260-1, 265-9; ritual of, 46-
hair, body (buibui), 74, 161, 171, 261-2, 7, 48, 85, 163, 184, 197
273 hunger vs. satiation, 49, 75, 90
hair, head (fufu), 30, 74, 171, 273 ikufuka, 169-71, 182-3, 193-8 passim,
happiness (engama) (see also conception, 276
de-conception), 153, 268; release from illness, 52-8, 151-2; areca nut chewing
mourning and, 159, 164, 178-9, 197; and, 90; childbirth and, 81-2, 84; dirt
courting and, 167 and, 28-9, 32, 154, 272; faifai, 31; food
Hau'ofa, E., 86, 112, 128-9, 240-2 preparation and, 68-81; menstruation
head, 30, 31, 80, 273 and, 63—9; nursing and, 67; pregnancy
health, 52-9 passim, 72, 82; see also body and, 79—80; sexual intercourse and, 60,
heat (pangaingai), 43, 44-5, 46, 48, 70; 64; spirits and, 28-9, 31, 54
see also hot independence, Papua New Guinean, 16
Hertz, R., 2, 150, 227 inedibility, see dirt, sweet vs. unsweet
history: anthropology and, 1, 5, 8, 242-9; inequality, 240-2
synchrony and diachrony, and, 1, 12, infants, 81-4
243, 248; historicism and, 5, 24, 242; in-laws, see affines
myth and, 5, 8; structure and, 5-13 inside vs. outside, 21, 23; body and, 45-
passim, 118, 142, 242-9 59 passim, 63, 65-6, 71-2, 78; clan
294 Index
inside vs. outside (cont.) and, 173; kofuapie prescription, 127-33,
identity and, 144, 180; feast foods and, 145; papie ngaunga proscription, 126-7,
174-5, 194-6; gender and, 95; and 145; sister exchange proscription, 133-
peace vs. war, 115; social relationship 4, 145-7, 261
and, 111, 126; and space and time, 23— marriage, Trobriand: exchanges, 218-19,
35 passim 230; regulation of, 214, 216-18, 221-2,
inside vs. outside, inverted/everted: body, 224, 228
38, 53-5, 57-9, 194, 247; clan bloods massaging, 75, 80, 81
and, 144; death and, 154; mythical, 71- Marxian approaches, 1, 237, 243, 248, 271
2; village and bush, and, 25-8, 32, 38, Mauss, M., 2
247 meaning, 1, 9-10
inversion, see reversal meat (tsitsi), 40, 42-5, 49-50, 56, 64, 97-
iungefanga, 169-71, 182-3, 193-8, 276 9, 255; big vs. small, 86-8, 82-4, 85,
Jakobson, R., 248 92, 98, 184, 187; for feasts, 162, 163,
Jamous, R., 2, 244 169-70, 174-7; mythical origin of, 96,
189-92, 266-70; village vs. bush, 174-7
Keesing, R., 232 medicines: fuka, 47; mulamula, 54-5, 86,
Kelly, R. C , 2, 245-7 90, 155
kin (atsi atsitsi), categories of, 102-12, mefu sorcery, 54-5, 67, 85, 157-8, 190;
274; and marriage regulation, 126-7, see also peace sorcery
137-42; and mourning, 154-5; see also Mekeo, Central, xi, 2, 13, 100, 114, 271;
relatives, affines, blood vs. nonblood hereditary officers, 16, 114; inequality
relations vs. equality, 240-2; marriage regulation,
kin, categories of, in Tikopia, 205-6 127-31; moieties, 112; relations with
kin, categories of, in the Trobriands, 211- Bush Mekeo, 15, 121, 122
31 passim, 277 memengoa ritual, 61-2, 65, 67, 70-2, 80-
Klein group in mathematics, 8-9, 248 4, 90; see also birth
kofuapie, 127-33, 234, 274; at feasts, 165, men, origin of, 95-7
170, 171, 182; see also affines, friend menstruation, 33, 60, 64-9, 71, 75, 79
Kuipa tribe, 15, 102, 271 methods, ethnographic, 12-13, 15
hula exchange (Trobriand), 230, 231, 232 milk, 67, 82, 84, 92
language differences, 13, 14, 16, 31 mission (Sacre Coeur), 16, 18, 133, 153-4,
Leach, E., 2, 6, 203, 211, 212-13, 214, 271
216, 245-6 model, 238, 239
Levi-Strauss, C , 2, 3-7, 21, 42, 100, 118, moiety (ngopu) (see also clan), 112, 124,
207, 231, 239, 248 132; exogamy and, 112-3, 126; mythical
lime, 70, 72, 91 origin of, 96-8; peace sorcery and, 115
Montague, S., 210, 214, 215, 216
MacDougall, L., 230 moon, 64
male, see gender mother (ina), 75, 104, 175, 193, 234-5,
Malinowski, B., 200, 210-11, 212, 219, 271; Akaisa Men's, 184, 192, 193, 197;
221 killing and eating of, 70-2, 96, 98-9,
marriage, 11, 73; areca nut chewing and, 190—4; see also abdomen
90—2; arrangement vs. elopement, 73, motivation, linguistic, 9-10, 234
96, 127, 129, 132, 274; mythical origin Motu peoples (and Koita), 82, 112
of, 261; new, bride's, 74-8, 82, 84, 90- movements, millenarian or revitalization, 7
1, 164, 272, 276; new, groom's, 85, 87, myth, 3-6, 8, 71; and ritual, 6, 71; struc-
90-1, 164; remarriage, 157, 171; see ture of, 3-6, 213-14, 248
also affines, spouses myths: Afinama, 97-9, 265-70; Akaisa,
marriage compensation exchange (akaila, 183, 189-91; Amaka, 69-72; Fipa (Af-
kaua), 11, 74, 129, 132, 137-42, 274, rica), 5-6; Foikale, 95-7, 98-9, 258-70;
275; feasting and, 155-6, 164, 173-7, Trobriand, emergence, 213-4, 217, 222
197-8; private (onge oake), 198
marriage rules, 62, 100, 113, 125-34; cog- Nature/Culture, 2, 4-5, 21
natic proscription, 126; endogamy, trib- Needham, R., 2
al, 102, 125; exogamy, clan and moiety, Nganga clan, history of, 117-23
100, 117, 125, 126, 179; feast categories nursing, 67, 81, 82-3, 84, 92, 175, 273
 Index 295
opening, 49; and feasting, 177, 181; gen- plant food, 42-5, 49-50, 64, 74-8, 85, 86,
der conceptualization and, 73, 84, 92-5; 169, 254; areca nut and, 54, 90; for
male ritual and, 85-6, 87-8, 91; mar- feasts, 169, 174; mythical origin of, 96-
riage compensation and, 139-42; mythi- 7, 258-64 passim
cal, 97-9; new marriage and, 79-80; poison (ipani), 30, 54, 60, 61, 67, 68, 151
nursing and, 83; sexual intercourse and, Powell, H. A., 210, 211, 214-15, 216,
61, 67, 79, 84, 85, 89; thirst and, 49; see 227, 230
also closing pregnancy, 64, 66-7, 75, 78-80, 268, 271,
ordinary sphere, 21; bodily, 58-9, 61, 69, 273
70-1, 99; feasting, 181; gender and, 34, procreation, see conception, pregnancy,
99; marriage reciprocity, 181; village- birth
bush transfers, 23-8, 58, 61, 69; see procreative vs. contraceptive, 77, 82-4,
also extraordinary sphere 92-4, 97, 99; at feasts, 164, 179-80;
ordinary men, 130, 181, 183, 194, 276 mythical, 184-5, 193-4
Ortner, S., 236 rank, see senior vs. junior
Ostor, A., 2 rank, in Tikopia, 202-3, 207
outside (afangai), see inside vs. outside rank, in the Trobriands, 213-16 passim,
papie ngaunga ("women's children"), 123- 276
4, 274; at death, 155; ekefaka and, 124- raw (maisa), 41, 44, 55, 70, 74, 96, 179,
5; at feasts, 163, 171, 173-4; marriage 261
compensation and, 142-5; marriage reg- relations, structural vs. semantic, 1, 9,
ulation and, 126-7, 128, 134, 142-4; see 235, 238
also kin relationship terminologies, 100, 112, 274;
parents, 68, 73, 103 affinal, 109-10; Dravidian, 124-5; Ha-
pacification, 16, 111, 123 waiian/generational, 103-9, 112, 125
patrilineality, 16, 112; Etoro, 246 relatives (see also kin, ekefaka, blood vs.
peace chief (lopia), 15, 113, 116-17; clan nonblood relations, affines): adoptive,
war officials and, 114-15; at feasts, 159, 179, 184, 188, 194; daily food exchange
160-71 passim, 181, 182-3, 199; female and, 25, 47; mourning and, 154; new
office holders, 120; functionaries of, 160; marriage and, 76; papie ngaunga and,
installation of, 167, 169, 193; lineage 124, 134; sexual sweet-unsweet trans-
organization and, 114-15; mythical ori- formation and, 62, 103; sorcery and,
gin of, 183-7; as owners of death, 160- 158; work and, 46
1; peace sorcerers and, 114-15, 160; residence, rules of, 74, 113, 116; in the
friends of, 129-32, 199; residential clans Trobriands, 214-15, 216
and, 116-17; tribal organization and, resources vs. wastes, 22, 23-8, 29, 45-52,
115 56, 234, 250
peace sorcery (ungaunga), 15; extraordi- reversal (or inversion), structural, 3, 5, 11,
nary sphere and, 29-34 passim, 61; 246-9; de-conception and, 177-81, 191-
fauapi bush dwelling, 29-30, 31, 33, 61; 9, 234; gender and, 99; health and ill-
at feasts, 160, 167, 170, 171; installation ness, and, 57; spatial, 25-8, 35
of heir, 182-3; mefu, 54-5, 67, 85, 157- ripening (eaipa), 41, 43-5, 46, 57-8, 74
8, 190; mythical origin of, 183-7; sexu- rites of passage, 6-7
ality and, 61; spirits of dead and, 151, ritual, 6-7, 9, 238-9; extraordinary sphere
158; subclan organization and, 114, 118— and, 29-36, 61, 99, 181
23; techniques of, 29-30, 33, 54-5, 61, ritual, female: areca nut chewing and, 90-
79-80, 87, 88, 114, 157-8, 271, see also 2; birth, 80-1; cycle of, 73-84; men-
tightening, ritual, poison; tribal organi- struation and, 68-9; mythical, 70-2, 98-
zation and, 115; widowers and, 157 9; new marriage sexuality and, 73-9;
person, 141, 149 postpartum and nursing, 81-4
Piaget, J., 8, 248 ritual, male (see also sorcery, chiefs), 28,
pigs, 15, 30, 83, 271; bush, 30, 46, 175, 33, 41, 79, 274; areca nut chewing and,
196; domesticated, 24, 25, 30, 175, 195; 91-2; bachelorhood, 89-90; courting,
likened to adulterers, 61; ownership of, 29, 33, 65, 68-9, 165, 166-7, 171; cycle
168-9; provisioning for feast, 162, 164, of, 84-92, 97; gardening, 29, 33, 48, 85,
196 275; hunting, 46-7; marriage and, 79-
296 Index
ritual, male (cont.) siblings, 73, 103-4, 190, 276
80, 83; mythical origins of, 96-7, 99; skills (etsifa), 40, 46, 63, 70-1, 87, 189
sexual intercourse and, 67, 165, 166-7; skin, 63; bathing and, 76; chewing and,
tightening, 41, 86-90, 157; work vs. 90, 91; of clan, 126, 144, 146, 154, 177,
nonwork and, 46-7 180; dirty, 55, 81, 163, 168; engorging
ritual, mortuary (see also feasts), 11, 32, and, 75; faifai and, 30, 31; homicide
40, 150, 195; bafu food abstention, 162, and, 86; iunge fanga "lost skin," 167-
170-1, 191; burial, 153-8; de-concep- 71, 183, 193, 194-6; mythical, 31, 186,
tion and re-conception, and, 177-81, 188-9, 198, 262, 273; sexual intercourse
191-9 passim; hearth, 272; kofuapie and, 63, 75-6, 79, 91; shedding of, 31,
and, 165, 167, 170-1; playing, 163-4; 180, 273
taboos, 153-4, 159, 162-3; treatment of smoking, or drying (eongongo), 41, 43-5,
corpse, 153-4, 157-8, 168; widow and 46, 57-8, 165
widower, 156-7, 170-1 snakes (faipo), 30-1, 32, 33, 97, 258-9,
ritual, in Tikopia: firstborn, 205; food divi- 273
sion, 207; kava, 204, 205; male initia- sons, 66, 153
tion, 205-6; marriage, 206; Marae, 207- sorcery (see also ritual, peace sorcery, war
8; mortuary, 206 sorcery), 22, 41, 61, 81, 185; acquisition
ritual, in the Trobriands, 209-33 passim; of, 15, 89, 157, 190; gardening, 29, 33,
mortuary, 218, 220-4, 230, 232 85; in the Trobriands, 232
river (ivi), 30 soul (laulau), 28, 29, 54, 152
roasting (euma) {see also cooking), 41, 43- space and time, conceptions of, 10, 21-3,
5, 46, 57-8, 64, 86, 89; in myth, 98, 58-9, 236, 237, 241; extraordinary
263 sphere and, 28-37; ordinary sphere and,
Robinson, M. S., 218-19 23-8; in Tikopia, 206-7
Roro peoples, 16, 112, 198 spirits (tsiange), 22, 28, 30-1, 34, 151,
rubbish (kamakama), 24-7, 68, 168, 190, 184; bush people, 30-5, 40, 76, 79, 82,
196-7, 272 97, 160, 190, 271, 275, see also faifai;
human, ancestral, 28, 30-1, 33, 151,
sacred vs. profane, 2, 6—7, 21, 86—7 186; in the Trobriands, 212, 213, 226,
sadness (etsetse), 153, 159 227, 229, 232
Sahlins, M., 2, 232, 238, 244-5 spouses, 109; adultery of, 61, 68; chewing
Saussure, F. de, 1, 13, 243 by, 90-1; desire for children and, 66,
Schneider, D. M., 2, 236 67, 73; ekefaka sibling category and,
seasons, wet vs. dry, 44, 88, 97, 272 124-5, 134; food and, 25, 62; mourning
Seligmann, C. G., xi, 82, 100, 128 and, 156-7, 170-1; mythical, 70-2, 96-
semen (ilaila), 60; and conception, 64-5, 9, 261, 265-70; as one blood, 275; poi-
72, 75, 76-9 passim, 85, 101; excretion soning and, 68; postpartum ritual and,
of, 63, 65; ingestion of, 60, 62; male 67, 80-4, 88-9; pregnancy and, 78-80,
ritual and, 33, 61; meat and, 76-8; men- 87-8; sexual intercourse and, 61—2, 66—
strual blood and, 64-6; nursing and, 67, 7, 69, 75-8, 84-5, 88-9; see also af-
82 fines, marriage, papie ngaunga, relatives
senior (fakania) vs. junior (eke), 104, 234, Stephen, M., 128
240; friend network of chiefs and, 130; stomach, 90
classification of siblings and, 103, 124-5; Strathern, A., and M. Strathern, 2
hereditary officials, 114, 186; lineages, Strathern, M., 2
114; marriage regulation and, 147; resi- structuralism, 1-10 passim, 12, 13, 248-9
dential clans and, 116; subclans, 113, structure, and contradiction, 3, 11, 139,
270; wallabies, 197 172-3, 239, 245-9; food, sexuality, and
Sexual intercourse, 60—72 passim, 73—9, menstruation, 60, 101, 172; in Hawaiian
167; extramarital, 61-2, 66, 69, 73, 90, history, 244-5; societal bloods and mar-
262, see also adultery; marital, 61-2, 66, riage compensation, 101, 139-42, 172-
69, 73, 75-9, 84, 85-92; oral, 61, 63; 3; societal bloods and feasting de-con-
sorcery and, 61, 67, 84, 86, 88; as ception, 150, 172-3
sweet-unsweet transformation, 62—6, structure, and history, 5, 7, 8, 12, 13, 118,
75-6, 87, 90-2, 96-7, 99 142, 242-9
Shapiro, W., 2 structures, binary, 2, 3, 132, 240, 242, 248
 Index 297
structures, linguistic, 8, 248 sweet-unsweet, transformations of, 38-59,
structures, mythical, 3-6, 213-14, 248 60-7, 71-2, 75-99 passim, 153-4, 167;
structures, psychological, 8, 248 areca nut chewing, 41, 90-1; bathing,
structures, quadripartite, in anthropology, 75-6, 79, 86; blood synthesis, 48, 49-
1-13 passim, 233, 246 56, 57-8, 152; cooking, 41-5, 48-52,
structures, quadripartite (Bush Mekeo), 70-2, 96-7; curing, 41, 55-6, 90; death,
239, 244, 245-8; bloods and interclan 151-2; drinking, 41, 49, 79, 85, 86-7,
marriage, 137-42, 144; bodily, 38, 45, 96-7; health and illness, 52-9; men-
53-9, 71, 247-8; clan offices and lin- struation, 65-6; nursing, 82, 84; sexual
eages, 115, 116-23, 147-8, 244; culi- intercourse, 62-6, 75-6, 79, 87, 90-2,
nary, 38, 57-8, 77; intergenerational, 96-7, 99; tightening, 86; weaning, 67,
148; interclan feasting, 155, 173-81, 83-4; work and nonwork, 46-8
199; gender and sexuality, 77-8, 94, 98, sweet vs. unsweet, 38, 40, 52, 103, 152,
149; spatio-temporal, 28-31, 32-3, 35- 166
7, 149, 247-8 symbols, key, 235-8; in Bush Mekeo
structures, quadripartite (Tikopia), xi, 201, culture, 236-8
204, 205, 207-8, 248; bodily, 202; chief- synchrony, priority of, over diachrony, 1,
tainship, 203, 204, 207; clans, 202-3, 13, 243-5, 248
204, 207; conception theory, 202; dis-
tricts, 203-4; firstborn ceremony, 205; Tambiah, S. J., 2, 232
food categories, 207; houses, 207; kava thirst, 49
ceremony, 204, 205; kin categories, 205, throat, 47, 49, 77, 79; areca nut and, 90
206; male initiation, 205-6; male vs. tightening, 41, 86-90, 157, 276
female relations, 207; Marae, 207, 208; Tikopia, ix, 12, 201-8, 211, 232; body,
marriage ceremony, 206; mortuary ritu- conceptions of, 202; chieftainship, 203,
al, 206; patrilineages, 205; politico-re- 204, 207; clans, 202-3, 204, 207, 208;
ligious functions, 203, 204 conception, 201-2; districts, 203-4;
structures, quadripartite (Trobriand), xi, firstborn ceremony, 205; gender, 202;
201, 209, 216, 225-6, 229-30, 231-2, houses, 207; kava ceremony, 204, 205;
248; age/gender distinctions, 209-10; kin categories, 205—6; male initiation,
animal categories, 213; bodily, 210; 205-6; male vs. female relations, 202,
bones of deceased, 226-7; chieftainship, 207; marriage ceremony, 206; pa-
213-4, 217, 227-30, 231; clans, 214, trilineages, 205; politico-religious func-
215-16, 219, 221, 222-6; cluster lead- tions, 203, 204; rank, 202-3, 207;
ers, 227-30, 231; conception theory, society, integration of, 203, 208
210-12, 226; de-conception, 224-7; fa- Tooker, D. E., 230, 232
ther, 227-30, 277; food exchanges, 214, tribe, 11, blood reciprocation and, 102-3;
219, 228, 277: gender, 209-10, 213, clanship and, 112-13, 142, 149; at
230, 231-2; hamlet managers, 214, 227- feasts, 173, 179; kin and, 109; marriage
30, 231: kula exchanges, 230, 231, 232; regulation and, 125; the person and,
marriage exchanges, 218-19; marriage 149; populations, 15; see also Amoamo,
regulation, 214, 217-18, 221-2, 224-5; Kuipa
milamala festival, 232; mortuary ritual, Trobriand islands, 12, 202, 209-33 passim;
223-4, 277; myth of emergence, 213, age/gender distinctions, 209-10; animal
222; political integration, 216-17, 227- categories, 213; blood, 226, 227, 229;
30; rank, 213, 214, 215; residence, 216; body, conceptions of, 210, 211, 212,
spirit, 232; substantial vs. nonsubstan- 229; bones of deceased, 226-7; chief-
tial, 210, 211, 231; subclans, 214, 215, tainship, 213-14, 217, 227-30, 231, 276;
216, 221, 222-3, 224; valuables, 219, clans, 212-26 passim; cluster leaders,
230; villages, 231-2; wasi exchanges, 214, 217, 227-30, 231; conception, the-
231; yam exchanges, 214, 219-20 ory of, 210-12, 226, 227, 229, 230; dis-
structures, ritual, 6-7, 9-10, 248 tricts, 217, 276; ethnography of, 209;
structures, social, 5-6, 10, 11, 100-49 father, 211, 212, 216, 218-19, 223, 224,
passim, 150; Etoro, 245-7 227-30, 277; father's clan, 217-18; food
structures, triadic, 2-3, 6, 246-7, 248 and eating, 212, 227, 277; food ex-
succession, rules of, 15, 16 changes, 212, 219-29 passim, 277; gar-
sun (tsina), 44, 70, 95-6, 258 dens, 212, 227-9, 231, 232, 277;
298 Index

Trobriand islands (cont.) vomiting, 83, 189; mythical, 70, 72, 96,
gender, 209-10, 211, 213, 227, 230, 259, 260, 262
231-2; hamlet, 216; hamlet manager,
214, 216, 223, 224, 227-30, 231; kula
exchange, 230, 231, 232; marriage ex- wallaby (matsi, mani), 191, 196-7, 260-1,
changes, 218—19, 230; marriage regula- 272, 276
tion, 214, 216, 217-18, 221-2, 224-5, Wallace, A. F. C , 7
228; milamala festival, 232; mortuary rit- war chiefs (iso), 15, 114-15, 151, 160;
ual, 218, 220-4, 230, 232, 277; mother, installation of, 182—3; and lineage orga-
211, 226, 229; myth of emergence, 213- nization, 114-15; mythical origin of,
14, 217, 232; political integration, 216- 183-7; and peace chiefs, 160-1; residen-
17, 227-30, 276; rank, 213-16; sor- tial clans and, 117; tribal organization
cerers, and witches, 232; spirits, 212, and, 115
213, 226, 227, 229, 232; substantial vs. war sorcerers (faika), 15, 160; and authori-
nonsubstantial, 210, 211, 231; subclans, ty of peace chiefs, 160; and extraordi-
214-24 passim; village clusters, 216-17; nary sphere, 29, 33, 35; installation of,
villages, 216-17, 227-8, 231-2; wasi ex- 182-3; mythical origin of, 183-7; and
changes, 231; yam exchanges, 214, 219- sexuality, 85—6; spirits of dead and, 151;
20, 228, 230, 277 subclan organization and, 114; tech-
niques of, 33, 35, 85-6, 87, 151; tribal
Umeda culture, 9, 238 organization and, 115
unidirectional vs. ambidirectional, 28, 34- warfare (aoao), 13, 15, 16, 244; as blood
5, 59, 72, 99, 181 reciprocity, 102; techniques of, 85—6,
urine, 66 88; World War II, 17-8
vagina (ito) (see alsofire),63, 64, 66, 70- warfare vs. peace, 13, 15, 16, 85-6, 102-
2, 91, 98, 234, 275; big vs. small, 184- 3, 113-15
5, 187; mythical origin of, 96, 262, 267 water (ivi), 30, 41, 42-3, 49-50; amniotic,
valuables (kefu), 16, 74, 132, 275 65-6; mythical origin of, 96-7, 264;
village (paunga), 22-37 passim, 39, 40, nonfreeflowing, 86; vaginal, 61, 63, 64,
116; clearing of, 24, 25, 34, 78; faifai, 70-1, 91, 97, 262
30; hot-cold transformation of, 152-3, weak, 273
165-6, 167, 171, 270; kofuapie at, 127- weaning, 67, 83-4, 88-9
33; mythical origin of, 96-7, 262-3; pe- weapons, 85
riphery of, 25-35; resources, 250; Weiner, A. B., 210, 215, 216, 217-18,
Trobriand, 216-17, 227-8, 232-3 219, 221-4, 227
village abdomen (paunga inaenga), 21, 74, wet, see dry vs. wet, water, blood
234, 273; bachelors prohibited from, 89; widow/widower, 33, 88, 156-7
burial in, 35; clearings of, 24-5, 34, 272; Willis, R., 5-6, 248
at feasts, 166, 167, 169-71, 172, 179; womb, 63, 271; see also abdomen
and myth, 98 work vs. nonwork, 45-52, 256, 272; house
village beings (paunga aunga), 21, 22, 39 building and, 46-7; hunting and, 45, 46;
village vs. bush, 21-34 passim, 40, 60-1, mourning and, 154, 162; new marriage
96-9, 174, 234 and, 74, 77-8, 85; tightening and, 87-
Virgin Birth, 65, 188, 226 90
Vogt, E. Z., 2 working together, 115-23